<%BANNER%>

Effects of High-Fat vs. High-Carbohydrate Diets on Proximal Gastric Relaxation, Gastric Emptying, pH of Gastric Contents...


PAGE 1

EFFECTS OF HIGH-FAT vs. HIGH -CARBOHYDRATE DIETS ON PROXIMAL GASTRIC RELAXATION, GASTRIC EMPT YING, pH OF GASTRIC CONTENTS AND PLASMA CHOLECYSTOKININ IN THE HORSE By MIREIA LORENZO-FIGUERAS A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2004

PAGE 2

Copyright 2004 by Mireia Lorenzo-Figueras

PAGE 3

Dedicated to my father.

PAGE 4

iv ACKNOWLEDGMENTS I would like to thank Dr. A. M. Merritt, my mentor and guide over these years, for his patient support and scientific guidance. Not only has he given me uncountable opportunities to grow professiona lly, but also has helped me feel comfortable since I arrived in Gainesville, more than four years ago. Special gratitude is also extended to my supervisory committee members, Drs. Colin Burrows, Richard Johnson, Robert MacKay, and Edgar Ott, for willingly sharing their knowledge and for review ing the disse rtation. I am particularly grateful to Dr. Burrows for his advice on the very first manuscript I submitted for publication. For their valuable assistance and guidance in the breath test studies, I would like to thank Drs. Tom Preston and David Sutton. It is also a pleasure to acknowledge Dr. Jean Morisset, who shared his expertise on chol ecystokinin and kindly performed the bioassay of this study. Acknowledgments are also exte nded to Drs. Chris Sanchez and Yong Bai, Hilken Kuck and Jim Burrow for their valuab le time and help. Additional thanks go to Dr. Murray Brown, Kelly Merritt and Dr. Ch arles Woods for permitting me to use the SpeedVac concentrator, Dr. Daniel C. Sharp for his assistance with the gamma counter, and Mrs. Marty Johnson for guidance on the RIA technique. I cannot forget to mention all the horses th at have participated in this study, for being patient with me, and for teaching me the importance of patience, care, and gratitude.

PAGE 5

v I am deeply thankful to the Office of Res earch and Graduate Studies at the College of Veterinary Medicine, Associate Dean Char les Courtney III, and Mrs. Sally OConnell for their support throughout my program. I would also like to thank all the staff of the Department of Large Animal Clinical Sc iences and the Deedie Wrigley-Hancock Fellowship for Equine Colic Research for thei r financial support, and the Florida Parimutuel Wagering Trust Fund and USA Equest rian Federation for providing funding for this study. Also, I express my deepest gratitude to the Fulbright Commission of Spain, which allowed me to take the first step of my American adventure. Finally, I am extremely gratef ul to my family for their love and support, and for understanding patiently all these years I have spent on the other side of the ocean. I am also thankful to all the friends I have met in Gainesville. My life here would not have been so enjoyable without them.

PAGE 6

vi TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES.............................................................................................................ix LIST OF FIGURES.........................................................................................................xiii ABSTRACT....................................................................................................................... xv CHAPTER 1 LITERATURE REVIEW.............................................................................................1 Fat Supplementation: A New Conc ept in Diet Formulation........................................1 The Effect of Diet on Gastric Physiology.....................................................................3 Dietary Regulation of Gastric Emptying...............................................................3 Meal consistency: emptying of liquids versus solids.....................................4 Meal composition: nutrient modulation of gastric emptying.........................5 Nutrient Sensing Initiates Nutrient s Regulation of Gastric Emptying..................6 Existence of nutrient-sp ecific chemoreceptors...............................................7 Sensing of dietary fat......................................................................................8 Sensing of dietary carbohydrates.................................................................10 Fat versus carbohydrate on cont rol of gastric emptying..............................12 Effect of Diet on the Meal-Induced Relaxation of the Proximal Stomach..........12 Nutrients Regulate Gastric Emptying by Neural and Hormonal Pathways........14 Neural pathways...........................................................................................14 Hormonal pathways......................................................................................15 Cholecystokinin: A Key Hormone in th e Regulation of Gastric Emptying........16 Gastric Emptying in the Horse: Current Knowledge and Methodology....................19 The Effect of Diet on Intragastric pH.........................................................................23 Squamous Ulceration of the Proximal Stomach..................................................23 Studies on the Effect of Diet on Intragastric pH in the Horse.............................24 Study Objectives.........................................................................................................24 2 MATERIALS AND METHODS...............................................................................25 Animals.......................................................................................................................2 5 Barostat....................................................................................................................... 25 13C-Octanoic Acid Breath Test...................................................................................27 Intragastric pH Monitoring.........................................................................................28

PAGE 7

vii Plasma Cholecystokinin Radioimmunoassay.............................................................28 Test Meals Composition.............................................................................................30 Phase I Studies.....................................................................................................30 Phase II Studies...................................................................................................31 Study Design...............................................................................................................32 Phase I Studies.....................................................................................................32 Phase II Studies...................................................................................................33 Experimental Procedure..............................................................................................33 Data Analysis..............................................................................................................38 Ingestion Time.....................................................................................................38 Proximal Gastric Compliance..............................................................................38 Intragastric pH.....................................................................................................39 Breath Samples....................................................................................................39 Calculation of gastri c emptying parameters.................................................39 Effect of diet on basal 13C output.................................................................40 Effect of the octanoic acid-loaded diets on 13C output.................................40 3 RESULTS AND DISCUSSION TONE OF THE PROXIMAL STOMACH.........41 Effect of 13C-Octanoic Acid Labeling (Breath Test)..................................................42 Phase I Studies: Pelleted Diets............................................................................42 Phase II Studies: Sweet feed Diets......................................................................44 Influence of Dietary Composition on the Effect of 13C-Octanoic Acid Labeling...........................................................................................................45 Effect of Dietary Composition....................................................................................46 Phase I. Pelleted Meals: Fat Versus Carbohydrate..............................................46 Phase II. Sweet feed Meals: Corn Oil Versus Glucose.......................................52 Unlabeled meals...........................................................................................53 Octanoic-acid labeled meals.........................................................................58 Conclusions.................................................................................................................61 Methodology........................................................................................................61 Relaxation Response...........................................................................................62 4 RESULTS AND DISCUSSION GASTRIC EMPTYING......................................66 Effect of Dietary Composition on Gastric Emptying.................................................67 Phase I. Pelleted Meals: Fat Versus Carbohydrate..............................................67 Effect of diet on basal 13C expiratory output................................................67 Effect of diet on gastric emptying................................................................68 Phase II. Sweet feed Meals: Corn Oil Versus Glucose.......................................75 Breath tests without presence of an intragastric barostat bag......................75 Breath tests with presence of an intragastric barostat bag...........................77 Effect of the Barostat Bag on Gastric Emptying........................................................80 Relation Between Proximal Gastric Relaxation and Gastric Emptying.....................80 Conclusions.................................................................................................................81

PAGE 8

viii 5 RESULTS AND DISCUSSION pH OF GASTRIC CONTENTS...........................83 Effect of Dietary Compos ition on Intragastric pH.....................................................84 Phase I. Pelleted Meals: Fat Versus Carbohydrate..............................................84 Phase II. Sweet Feed Meals: Corn Oil Versus Glucose......................................88 Conclusions.................................................................................................................91 6 RESULTS AND DISCUSSION CHOLECYSTOKININ LIKE-ACTIVITY.........93 Radioimmunoassay.....................................................................................................93 Conclusions...............................................................................................................100 7 SUMMARY AND CONCLUSIONS.......................................................................102 APPENDIX A INDIVIDUAL ANIMAL DATA.............................................................................107 B STATISTICAL TESTS FOR BAROSTAT DATA.................................................125 C STATISTICAL TESTS FOR GASTRIC EMPTYING DATA................................150 D STATISTICAL TESTS FOR PH DATA.................................................................153 LIST OF REFERENCES.................................................................................................160 BIOGRAPHICAL SKETCH...........................................................................................179

PAGE 9

ix LIST OF TABLES Table page 2-1 Approximate composition of the pellet ed test meals used in the Phase I studies....................................................................................................................31 2-2 Composition of the sweet feed meal (S eminole Feed, Blue Ribbon 10) used in the Phase II studies.................................................................................................32 3-1 Comparison of duration of meal inge stion between the present study (A) and that of Lorenzo-Figueras and Merritt113 (B)..........................................................49 3-2 Postprandial variations in volume* of an intragastric bag controlled by an electronic barostat .................................................................................................60 4-1 Comparison of gastric emptyi ng parameters determined by the 13C-octanoate breath test...............................................................................................................69 5-1 Changes in intragastric pH after inge stion of a 0.5 g/kg sweet feed meal alone (control meal) or enriched with either corn oil or glucose.....................................89 A-1 Individual body weight, test meal wei ght (Phase I and II), breath test label composition, and sweet feed supplementation (Phase II)....................................107 A-2 Barostat raw data (Phase I): bag vo lumes for baseline and 2-min postprandial blocks after ingestion of the unlab eled high-fat pelleted meal acid.....................107 A-3 Barostat raw data (Phase I): bag vo lumes for baseline and 2-min postprandial blocks after ingestion of the labeled high-fat pelleted meal................................108 A-4 Barostat raw data (Phase I): bag vo lumes for baseline and 2-min postprandial blocks after ingestion of the unlabeled high-carbohydrate pelleted meal............109 A-5 Barostat raw data (Phase I): bag vo lumes for baseline and 2-min postprandial blocks after ingestion of the labele d high-carbohydrate pelleted meal................111 A-6 Barostat raw data (Pha se I): ingestion time (sec) for the pelleted diets...............112 A-7 Barostat raw data (Phase II): bag vol umes for baseline and 2-min postprandial blocks after ingestion of the unlabeled corn oil-enri ched sweet feed meal.........112

PAGE 10

x A-8 Barostat raw data (Phase II): bag vol umes for baseline and 2-min postprandial blocks after ingestion of the labeled corn oil-enriched sweet feed meal.............113 A-9 Barostat raw data (Phase II): bag vol umes for baseline and 2-min postprandial blocks after ingestion of the unlabeled glucose-enri ched sweet feed meal.........114 A-10 Barostat raw data (Phase II): bag vol umes for baseline and 2-min postprandial blocks after ingestion of the labeled glucose-enriched sweet feed meal.............116 A-11 Barostat raw data (Phase II): bag vol umes for baseline and 2-min postprandial blocks after ingestion of the la beled control sweet feed meal.............................117 A-12 Barostat raw data (Phase II): inges tion time (sec) for the sweet feed diets.........118 A-13 Gastric emptying raw data (Phase I) : parameters of the pelleted diets................118 A-14 Gastric emptying raw data (Phase II): parameters of the sweet feed diets..........119 A-15 Intragastric pH raw data (Phase I): mean pH for baseline and 5-min postprandial blocks after ingestion of the unlabeled high-fat pelleted meal.......119 A-16 Intragastric pH raw data (Phase I): mean pH for baseline and 5-min postprandial blocks after ingestion of the labeled high-fat pelleted meal...........120 A-17 Intragastric pH raw data (Phase I): mean pH for baseline and 5-min postprandial blocks after ingestion of the unlabeled high-CHO pelleted meal...121 A-18 Intragastric pH raw data (Phase I): mean pH for baseline and 5-min postprandial blocks after ingestion of the labeled high-CHO pelleted meal.......121 A-19 Intragastric pH raw data (Phase I I): mean pH for baseline and 5-min postprandial blocks after ingestion of the labeled corn oil-enriched meal..........122 A-20 Intragastric pH raw data (Phase I I): mean pH for baseline and 5-min postprandial blocks after ingestion of the labeled corn oil-enriched meal..........123 A-21 Intragastric pH raw data (Phase I I): mean pH for baseline and 5-min postprandial blocks after ingestion of the labeled glucose-enriched meal..........123 B-1 ANOVA and CL for ingestion times of unlabeled pelleted meals (Phase I).......125 B-2 Shapiro-Wilk test for normality of ingestion times of unlabeled pelleted meals (Phase I).....................................................................................................125 B-3 Bartlett's test for homogeneity of variance of unlabeled pelleted meals (Phase I)...............................................................................................................125 B-4 ANOVA mixed procedure for mean bag volumes of pelleted meals (Phase I)..125

PAGE 11

xi B-5 ANOVA mixed procedure for mean bag volume minus baseline of pelleted meals (Phase I).....................................................................................................130 B-6 ANOVA and CL for ingestion times of unlabeled sweet feed meals (Phase II)..............................................................................................................134 B-7 Shapiro-Wilk test for normality of ingestion times of unlabeled sweet feed meals (Pha se II)....................................................................................................134 B-8 Bartlett's test for homogeneity of variance of unlabeled sweet feed meals (Phase II)..............................................................................................................134 B-9 ANOVA mixed procedure for mean bag volumes of unlabeled sweet feed meals (Pha se II)....................................................................................................135 B-10 ANOVA mixed procedure for mean bag volume minus baseline of unlabelled sweet feed meals (Phase II)..................................................................................137 B-11 Bartlett's test for homogeneity of variance for ingestion times of labeled sweet feed meals (Phase II)..................................................................................139 B-12 Friedmans 2-way ANOVA for ingestion times of labeled sweet feed meals.....139 B-13 ANOVA mixed procedure for mean bag volumes of labeled sweet feed meals (Phase II)..............................................................................................................139 B-14 ANOVA mixed procedure for mean bag volume minus baseline of labelled sweet feed meals (Phase II)..................................................................................143 B-15 ANOVA mixed procedure for mean bag volume minus baseline of unlabeled (BFAT) and labeled (FFAT) corn oil-en riched sweet feed meals (Phase II)......146 B-16 ANOVA mixed procedure for mean bag volume minus baseline of unlabeled (BCHO) and labeled (FCHO) glucose-en riched sweet feed meals (Phase II).....147 C-1 Shapiro-Wilk test for normality of Phase I parameters......................................150 C-2 Bartlett's test for homogeneity of variance of Phase I parameters....................150 C-3 Two-sample t-test for Phase I parameters............................................................150 C-4 Shapiro-Wilk test for normality of Phase II parameters without an intragastric bag........................................................................................................................150 C-5 Bartlett's test for homogeneity of variance of Phase II parameters without an intragastric bag.....................................................................................................150 C-6 Paired-sample t-test for Phase II para meters without an intragastric bag............151

PAGE 12

xii C-7 Shapiro-Wilk test for normality of Phase II para meters with an intragastric bag........................................................................................................................151 C-8 Bartlett's test for homogeneity of variance of Phase II parameters with an intragastric bag.....................................................................................................151 C-9 Repeated measures ANOVA for Phase II t1/2 with an intragastric bag...............151 C-10 Friedmans 2-way ANOVA for Phase II tmax with intragastric bag.....................151 C-11 Repeated measures ANOVA for Phase II GEC with an intragastric bag............152 C-12 Shapiro-Wilk test for normality of Phase II parameters: effect of an intragastric bag on gastric emptying....................................................................152 C-13 Bartlett's test for homogeneity of variance of Phase II parameters: effect of an intragastric bag on gastric emptying...............................................................152 C-14 Paired-sample t-test for Phase II parame ters: effect of an intragastric bag on gastric emptying...................................................................................................152 D-1 ANOVA mixed procedure for mean intragastric pH of pelleted meals (Phase I)...............................................................................................................153 D-2 ANOVA mixed procedure for mean intragastric pH of labeled sweetfeed meals (Pha se II)....................................................................................................155

PAGE 13

xiii LIST OF FIGURES Figure page 2-1 Motility of the stomach measured by an electronic barostat....................................26 2-2 Polyester bag connected to a pr essure line and an inflation line..............................27 2-3 Endoscopic view of the squamous mucosa of the proximal stomach......................34 2-4 Positioning of the polyester bag used to measure intragastric pressure in the proximal stomach.....................................................................................................34 2-5 Endoscopic views showing correct pos ition of the barostat bag within the proximal stomach.....................................................................................................35 2-6 Motility of the proximal stomach was measured with an intragastric bag, inserted through the gastric cannula and c onnected to the electronic barostat by two separate catheters...............................................................................................36 2-7 The pH of gastric contents was measur ed using a pH electrode inserted through the gastric cannula and positioned in the most ventral part of the stomach.............37 3-1 Changes in intragastric bag volume af ter ingestion of th e high-carbohydrate (CHO) pelleted meal with and without addition of 13C-octanoic acid.....................43 3-2 Mean bag volume revealing the effect of intake of a high-fat (A) and a highcarbohydrate (B) pelleted diets (0.5 g/kg) on gastric tone of the proximal portion of the stomach in 6 horses...........................................................................47 3-3 Changes in intragastric bag volume afte r ingestion of either the high-fat meal or the high-carbohydrate (CHO) pelleted meal........................................................48 3-4 Mean volume trace (n=6) of the effect of ingestion of different sweet feed meals (0.5 g/kg) on baseline tone in the proximal stomach...............................................54 3-5 Changes in intragastric bag volume afte r ingestion of a 10% protein sweet feed meal (0.5 g/kg) enriched by either corn oil or glucose (n=6)...................................55 3-6 Changes in intragastric bag volume after ingestion of a contro l sweet feed meal (0.5 g/kg) with and without addition of either corn oil or glucose...........................61

PAGE 14

xiv 4-1 Mean percentage dose recovery (PDR/h) SEM and modeled curve of 13C in breath following ingestion of a highfat pelleted meal (n=5) or a high-carbohydrate (CHO) pelleted meal (n=6)........................................................69 4-2 Mean percentage dose recovery (PDR/h) SEM and modeled curve of 13C in breath following ingestion of a 10% crude protein sweet feed meal (Seminole Feed, Blue Ribbon 10) enriched wi th corn oil or glucose (n=6)...........................77 4-3 Modeled mean % dose recovery curves of the 13C label in the breath of 6 horses after ingestion of a sweet feed meal (cont rol) or the same meal enriched with corn oil or glucose....................................................................................................79 5-1 Changes in intragastric pH afte r ingestion of the high-fat or the high-carbohydrate (CHO) pelleted meal..................................................................85 5-2 Changes in intragastric pH after inges tion of a control sweet feed meal, or the same meal enriched with either corn oil or glucose.................................................90 6-1 The standard curve shows the fraction of 125I-radiolabeled CCK-8 bound to antibodies ( B / B0) at increasing levels of standard CCK...........................................95

PAGE 15

xv Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EFFECTS OF HIGH-FAT vs. HIGH -CARBOHYDRATE DIETS ON PROXIMAL GASTRIC RELAXATION, GASTRIC EMPT YING, pH OF GASTRIC CONTENTS AND PLASMA CHOLECYSTOKININ IN THE HORSE By Mireia Lorenzo-Figueras May 2004 Chair: A.M. Merritt Major Department: Veterinary Medicine Addition of fat, rather than extra carbohydr ate, to the diet has become a common strategy to increase the energy density in ratio ns for high performance horses. Little is known, however, about the effect of fat on equi ne gastrointestinal function. Therefore, the objective of this study was to evaluate and compare the effect of high-fat and highcarbohydrate (CHO) diets on different parameters of gastrointestin al function in the horse. Six adult horses, each with a gastric cannula, were used in two different series of studies. In Phase I, horses were offered 0.5 g/kg of a high-fat (8% fat) or a high-CHO (3% fat) pelleted diets of identical volume, caloric density and prot ein content. In Phase II, test meals consisted of 0.5 g/kg of a sweet feed meal, or this meal supplemented with corn oil (12.3% fat) or an is ocaloric amount of glucose (2.9% fat). Four parameters were measured simultaneously: 1) proximal gastric tone by variations in the volume of an

PAGE 16

xvi intragastric bag, introduced through the gastri c cannula and maintained with a constant internal pressure by an electronic barost at; 2) rate of ga stric emptying by the 13C-octanoic acid breath test; 3) pH of gastric contents by a self-referencing pH probe introduced through the gastric cannula; and 4) plas ma CCK-like activity by a commercial, nonspecific radioimmunoassay kit (Phase I only). Meals with higher CHO content induced a significantly (p<0.05) more prolonged receptive relaxation of the proximal stomach th an those with higher fat content, but the accommodation response was similar. Labeling of meals with the breath test marker modified the receptive relaxation response. Ga stric emptying rates were not significantly different between meals, although those hi gh in CHO tended to empty more slowly initially. A significantly greater increase in intr agastric pH was seen after ingestion of the high-CHO pelleted meal. Radioimmunoassay and bioassay methods failed to detect plasma CCK activity. This study suggests that in the horse, in contrast to most species, dietary fat may not be more suppressive of gastric tone and emptying than CHO.

PAGE 17

1 CHAPTER 1 LITERATURE REVIEW Fat Supplementation: A New Con cept in Diet Formulation The horse has naturally evolved to digest and use high-fiber sources efficiently. These sources are relatively low in energy cont ent, but adequate to satisfy maintenance requirement. Coevolution with humans, t hough, has resulted in in creasing demands on the horse to perform under ci rcumstances that require energy intakes greater than that provided by its more natural diet of fresh roughage.69 Consequently, performance horses are routinely supplemented with high-starch diets to reach their high energetic needs. This common practice can represent a digest ive and metabolic challenge for the horse, and excessive high concentrate intakes are us ually avoided. Concentr ates are rich in soluble carbohydrates, which may overwhelm the digestive capacity of the small intestine, leading to rapid fermentatio n of the grain carbohydrates in the hindgut.29 Addition of fat to the diet has been recei ving considerable atte ntion as a way to increase dietary energy content, and fatsupplemented diets are becoming a common alternative to traditional high-st arch diets. Multiple feed companies are introducing their own commercial high-fat diet s, which may contain as much as 12% of the total dry matter. Although we refer to these diets as hi gh-fat diets (6-12% of total diet or 20-40% of the total calories), they are actually mu ch closer to the norma l diets used in other species.94 Providing energy as fat rather than carbohydrate may avoid some disadvantages of concentrate diets, in cluding the production of potentially harmful volatile fatty acids in the stomach157 and in the large intestine,130 which may cause ulcers,

PAGE 18

2 and colic and laminitis, respectively. Fat may also ameliorate some of the alterations in fluid and electrolyte homeostasis associated with feeding a large, high carbohydrate, concentrate meal,29;188 or help in controlling development of rhabdomyolysis in predisposed horses.214 Since fat is an efficient energy source for exercising muscle, additional benefits of dietary fat on performance have been proposed.94 The glucose-sparing effect of fatsupplemented diets seems to delay symptoms of fatigue94 associated with glycogen depletion,188;231 acidosis and lactate formation.63 Yet, the effect of fat on muscle glycogen and plasma lactate is not consistent in the literature,69 and the potential positive effects of supplemental fat on energy metabolism need further study.94 Additional benefits of fat supplementation include reduction in internal body heat when compared with highcarbohydrate or protein diets,90;192 a calming effect on the horse,73 and decreased weight of intestinal contents from reduced dry matter ingestion.90 Finally, fat is a useful source of energy for geriatric horses or thin hor ses that fail to reach a desirable weight;194 it also improves skin and hoof appearance, and reduces dustiness of the meal. Although fat supplementation appears to have many benefits, its use may be limited by palatability, excessive oxidation (rancidity ) during processing and storage, and the digestive and absorptive capacity of the equi ne gastrointestinal tract. The adaptation of the horse to dietary fat has been addressed in some studies. For instance, feeding an increased level of fat causes metabolic adapta tions that permit horses to preferentially use fat and spare glycogen during exercise. Fat supplementation is also associated with increased plasma total lipase activity (lipoprotein lipase and hepatic lipase),158 enhanced oxidative capacity of the muscle,158 lower glycemic and insulinemic responses after

PAGE 19

3 ingestion,231 and dose-dependent increas e in fat digestibility.23 Since excessive soluble carbohydrates are undoubtedly associated with clinical and someti mes life-threatening conditions, Williams et al.231 (p. 2199) suggests that the metabolic and health impacts are likely to be moderate for meals rich in fat and fiber, which may be more reflective of the nutritional heritage of the horse. The Effect of Diet on Gastric Physiology Little research has been conducted in the horse regarding the e ffects of dietary fat and dietary components in generalon gast ric function. In other species, dietary composition of the meal can affect the ra te of gastric emptying, proximal gastric accommodation and plasma concentration of regulatory hormones and peptides associated with meal ingestion. Modulation of gastric motility in response to consumption of a meal is the first stepe xcluding intake rateof a long and complex series of events aimed to maximize th e digestion and absorption of nutritional components of the diet. Dietary Regulation of Gastric Emptying The stomach stores ingested food in the pr oximal stomach, mixes it with secretions, discriminates solids and liquid, breaks solids down to small particles, and delivers food into the duodenum at a rate th at is compatible with effi cient digestion and absorption within the intestine.124 Far from being a simple hollow organ, the stomach is composed of three different functional components: the pr oximal stomach, the antrum and the pyloric sphincter. These parts function in a coordi nated way within each other and with the proximal duodenum to regulate gastric emptying according to the composition of the meal.173 Existence of neural and hormonal feedback reflexes originati ng at very distant

PAGE 20

4 sites, such as ileum and rectum, to modulat e gastric motility, gives further indication of the complexity of gastric emptying regulation.183 Meal consistency: emptying of liquids versus solids A first determinant of the way a meal is emptied from the stomach is its consistency. The stomach can discri minate between liquids and solids,124 with liquids emptying from the stomach much more rapidly than solids.30;119 In a solid-liquid mixed meal, the liquid component empties rapidly in an exponential manner, whereas the solid part remains in the proximal stomach until most of the liquid has emptied (period defined as lag phase).78 Unlike solids, a lag phase is no t observed in gastric emptying of liquids, unless they have a high caloric density.62 Scintigraphic studies have shown that both liquids and solids empty in an exponen tial manner, although emptying of liquids is more logarithmic in character.50 Finally, discrimination between emptying of the solid and the liquid phase of a mixed meal may be affected by the degree of homogenization of its components. That is, after the meal is mixed in the mouth and stomach, the solid and liquid components are not clearly separate d and gastric emptying may be delayed by increased viscosity.186 It is well recognized that gastric emptying of liquid follows an intraluminal pressure gradient between the proximal stomach and the duodenum. This pressure gradient is generated and m odulated by variations in toni city of the proximal gastric wall.124 Yet, additional research indicates that factors besides tone of the proximal stomach are crucial for normal liquid gastric em ptying, and that there is a role for antral and pyloric contractile activity.32;121;191 For example, it has been shown that pulsatile gastric emptying of liquids across the pylor us is correlated with antral waves of contraction in dog161 and pig,120 whereas the volume of the flows is correlated with

PAGE 21

5 proximal gastric and pyloric tone.161 Furthermore, liquid emptying is significantly more rapid in pylorus excised than in pylorus intact pigs during intraduodenal infusion of nutrients or hyperosmolar solutions.210 In contrast to liquids, em ptying of solids seems to depend primarily on antral contractions, which break up food into par ticles small enough to pass through the pyloric canal. Additionally, coordinated antropyloroduoden al contractions, acting as a peristaltic pump, are a major factor in emptying regulati on. Contraction of the pylorus serves as a barrier against the propulsive force of the antrum, causing retropropulsion and consequent trituration of large particles.67 Simultaneously, small pa rticles are propelled through the pylorus into the proximal duodenum.173 Finally, the proximal stomach assists the emptying of solids by delivering its contents down to the antrum,32 and by preventing reflux of antral contents into the prox imal stomach as the antrum contracts.173 Meal composition: nutrient modul ation of gastric emptying Volume, nutritional constituents, physical st ructure (i.e., viscosity and particle size), caloric density and osmolar ity of a meal are the principal factors affecting its rate of gastric emptying, and modulation of gastric em ptying can only be seen as the combined effect of all these factors on gastric motility. For example, gastric emptying of a liquid meal containing nutrients is influenced by the volume of fluid in the stomach and by its energy density.36 Increasing the volume of this liquid meal will speed emptying, but increasing the nutrient content will slow emptying.81 Presence of nutrients in the intestine is a potent stimulus for feedback regulation of gastric motor function and can be mimicked in experimental situations by direct infusion of nutrients into the gut.142 Intestinal feedback inhibition by nutrients involves relaxation of the proximal stomach, suppression of antr al motility, stimulation of isolated phasic

PAGE 22

6 pyloric contractions and increased pyloric a nd duodenal resistance, which together work to slow down the flow of gastric contents into the duodenum. As a result, the stomach delivers nutrients into the small intestine with rates that are compatible with digestion and absorption.173 Nutrient Sensing Initiates Nutrients Regulation of Gastric Emptying Receptors sensitive to nutrients play an im portant role in the control of gastric motility. They are present throughout the mucosa of the small intestine and show regional variation for nutrient sensitivity.8 For instance, gastric emptying of solids is approximately three times more potently inhi bited by glucose perfusion in the fourth quarter versus the first or second quarter of small bowel in the dog.107 Species differences may also exist since carbohydrat e infusion in the ileum slows gastric emptying of liquids and solids in the dog,209 but has no effect on gastric em ptying of solids in humans.227 In addition to regional distribution, parallel and sequential activation of receptors by transit of nutrients along the length of the intestine may occur.103;104 The complexity of dietary regulation is accentuated by st udies suggesting that gastri c emptying is influenced by patterns of previous nutrient intake, possibly through adap tation and sensitization of receptor mechanisms to their original stimu li. Accordingly, mainta ining a high-fat diet for two weeks, but not four days, results in acceleration in the gast ric emptying rate of high-fat meals in humans.39 These adaptive changes are nutr ient specific, since adaptation to a high-fat diet does not aff ect gastric emptying of carbohydrates.25 Similarly, shortterm supplementation of the diet with glucose leads to increased gastric emptying rate of a glucose test meal,76 whereas the emptying rate of a protein drink is unchanged.40

PAGE 23

7 Existence of nutrient-specific chemoreceptors Most mucosal receptors in the small intestine are polym odal (i.e., they respond to different mechanical and/or chemical stimuli),131;190 such as those activated by osmotic pressure variations and pr esence of acids and alkali.133 On the other hand, some receptors are specific for a single stimulus,131 and selective chemoreceptors for acid,49 glucose,49;106;190 fatty acids,135;163 and the amino acids tryptophan24;33;201 and phenylalanine22 have been described. Nutrients ac t independently of any osmotic or mechanical effects, and each macronutrient group acts via activation of separate and distinct mechanisms and pathways.165 In addition, mechanical stimulation by presence of contents in the small intestine inhibits gastric emptying.124 Accordingly, distension of a balloon within the duodenum i nduces inhibition of abomas al contractions in sheep19 and decreases fundic tone in the dog.41 The current scientific litera ture is abundant in studies on the effect of different nutrients on gastric emptying and the mechanis ms involved in it. Yet, it is difficult to draw definitive conclusions because of gr eat differences in the methodology of these studies. First, test meals are not standard ized among studies, which makes comparisons difficult. Furthermore, meals may differ in mo re than one characteristic and, therefore, it is difficult to conclude which specific ch aracteristic determined the outcome. Second, some studies used inaccurate methods to m easure gastric emptying, such as the intubation technique, which were likely to cause experimental errors. In addition, many studies describe the regulatory effect of nutrients by measuring the gastric emptying of a control meal in response to simultaneous infusion of nutrients into the small intestine. This experimental design is not as physiological as ingestion, and it does not take into account the additional influence of th e control meal on gastric emptying. As well, studies using

PAGE 24

8 meal preloads often yield very different re sults when compared with studies in which nutrients are infused directly into th e gut. Some studies performed in rats86;143 and humans26 address this issue, and suggest that or osensory factors play a role on gastric emptying of specific nutrients. Therefore, we cannot assume that gastric emptying is the same when a nutrient is directly deposited into the gastrointestinal tract as when it is fed orally. Finally, species to species variations may exist, so that what is true for one species may not apply to other species. Sensing of dietary fat It is known that fat is a potent inhibitor of gastric emptying.79 The intensity of inhibition is dependent on the concentration of fat, the leng th of intestine exposed to it104 and the type of fat. For example, studies in humans79 and cats136 show that free fatty acids, but not triglycerides, ar e effective stimuli for inhibiti on, with 12-carbon or greater chain-length fatty acids being the most effective. Similarly, instillation into the duodenum of a 12-carbon but not a 10-carbon long-chain fatty acid reduces antral contractile amplitude and reduces proximal gastric tone in humans.127 Fat is digested and absorbed mainly in the proximal small intestine by rapid diffusion across the brush border membrane of the enterocyte.211 Postabsorptive chylomicron formation is required for the abil ity of fatty acids to initiate feedback inhibition of gastric emptying, which suggests that activation of the fatty acid receptor does not occur within the lumen of the intestine.60 When a larger load is ingested, however, fat is also absorbed in the distal small intestine.232 The arrival of fat (unabsorbed lipids or fatty acids) into the distal small intestine causes further inhibition of gastric emptying in a dose-dependent manner.227 Further delay serves to increase the contact time between luminal contents and the absorptive epithelium. This feedback

PAGE 25

9 mechanism, referred to as ileal brake, is sp ecies and nutrient specific. For example, it is also induced by glucose and amino acids in dogs196 but not in man.227 In contrast, infusion of fat into the distal small intes tine of the dog has no effect on proximal gastric relaxation, whereas carbohydrate a nd protein reduce gastric tone.8 Since most of the dietary fat is almost entirely absorbed proximal to the ileum, feedback from the ileum may only occur after unusual meals or under a ltered physiological condi tions in order to enhance absorption by delaying the passage of food through the small intestine.172 The mechanisms determining the emptying rate of fat are not totally understood. On the one hand, one study64 in pigs suggested that fatty acids do not empty from the stomach with a constant caloric rate, in contrast to carbohydrates. On the other hand, studies in humans,81 monkeys,126 and pigs226 suggested that emptying of fat or any other macronutrient was mainly based on delivery of a constant rate of energy into the small intestine. Fats in the liquid phase empty at a rate different to the aqueous phase, but at a rate similar to that of an e quicaloric digestible solid meal48 following an initial lag phase.141 Also, the chemical structure of dietar y fats influences the rate of gastric emptying. For example, an equicaloric amount of olive oil, containing more unsaturated fatty acids, slows gastric emptying less than margarine, which contains more saturated fatty acids.119 Besides chemical composition, phys ical properties such as density, intragastric distribution, vi scosity and degree of emulsi fication determine gastric emptying of fat. Major differences in the intrag astric distribution of oil compared to solid and aqueous liquid meals exist, and may account for slower emptying of this component. In an oil/aqueous-based soup meal, gastric em ptying of oil was slower than soup, and it was associated with longer retention in th e proximal stomach and retrograde movement

PAGE 26

10 of oil from distal to proximal stomach. When a solid meal was added to the oil/soup mixture, more oil was retained in the proxi mal stomach and more solid was retained in the distal stomach.48 Layering of an oil phase above the water due to lower specific gravity may be in part responsible for this e ffect. This distribution may be consequent to initial passage of small am ounts of fat into the duodenum leading to relaxation and redistribution of fat into the proximal stomach.77 Finally, degree of meal homogenization influences gastric emptying of the oil phase. For example, butter emp ties slower than the aqueous contents, unless it is emulsified before ingestion.34 Sensing of dietary carbohydrates Similar to fat, the factors affecting th e gastric emptying rate of carbohydrates are poorly understood, but it seems to depend on the type of carbohydr ate (e.g., sucrose, fructose, galactose), its form (e.g., maltodextrins, starches)152 and the length of intestine to which it is exposed.103 For example, glucose empties linearly, whereas fructose empties exponentially and more rapidly th an xylose and glucose in the monkey.146 Caloric content, volume and osmolarity ha ve been regarded as potential factors controlling the gastric emptyi ng of carbohydrates. However, studies on glucose dose and volume-dependent inhibition of emptying are contradictory. On the one hand, Hunt et al.80 showed that, in humans, increases in ei ther energy density or meal volume of a polymer of glucose increased the rate of energy deliver y after 30, 60, or 120 min. Moran et al.148 showed that volume of a gl ucose beverage played a role only in the initial rapid rate of emptying, and the rate of gastric emptying (calories and volume) was identical after 20 minutes regardless of the initial volume. On the other hand, results of other studies64;81;126;226 suggest that emptying of carbohydr ates is mainly driven by the maintenance of a constant caloric flow. Finally, a study132 performed in anesthetized cats

PAGE 27

11 showed that the electrical activity of vaga l sensory neurons specifically responsive to glucose increased with increas ing doses of glucose infusions into the proximal small intestine. Yet, it is unknown to what extent the observ ed increased discharge of glucoreceptors may affect gastric motil ity and emptying. Similar to volume and concentration, it is also uncl ear whether osmolality contri butes to the glucose-induced modulation of gastric emptyi ng. Inhibition of emptying by hype rtonic glucose seems to depend mostly on chemospecific feedback,126 although osmotically sensitive pathways may also be involved.106 Nonfiber carbohydrates are digested and ab sorbed mainly in the proximal small intestine by active transport across the brush border membrane of the enterocyte.211 Activation of glucose receptors and consequent initiation of f eedback inhibition of gastric motility may be dependent either on rapid accu mulation of glucose within epithelial cells or on activation of the Na+-glucose co-transporter.171 With ingestion of a large load, the digestive and absorptive capacity of the sma ll ingestion may be exceeded, and arrival of carbohydrates into the distal sm all intestine causes further inhibition of gastric emptying. As mentioned previously, pres ence of carbohydrates within ileum slows gastric emptying of liquids and solids in the dog.209 Glucose exerts this inhi bition in a dose dependent way,196 and this effect is more potent when co mpared to exposure of the upper small intestine.107 Slowing of gastric emptying by ileal perfusion of carbohydrate has also been observed in humans, although perfusion was a ssociated with abdomi nal discomfort, and the authors of the study related the carbohydr ate-induced inhibition to activation of nociceptive pathways.82

PAGE 28

12 Fat versus carbohydrate on control of gastric emptying As has already been menti oned, it is generally assumed that fat is a more potent inhibitor of gastric emptying and motility than carbohydrate. However, there are a limited number of studies comparing the effect of fat and carbohydrate meals, irrespective of energy, volume or consistency of the meal. A fe w studies show that fat causes a greater inhibition of gastric emptying than carbohydrate of equivalent caloric density and volume in humans27 and in pig.64 Effect of Diet on the Meal-Induced Relaxation of the Proximal Stomach The proximal part of the stomach acts as a reservoir for food and participates in the creation of an intragastric pressure gradient which is in part responsible for gastric emptying of liquids. To accomm odate the ingested material without significant increases in intraluminal pressure, the proximal stomach is controlled by a va go-vagal reflex that enhances wall relaxation by decreasing gastri c tone. This relaxation decreases the fundoantral pressure gradie nt and slows delivery of fundic c ontents into the antrum. This, in turn, decreases the rate of gastric emptying.124;207 Previous work in our lab113 have revealed the existence of a meal-induced physiological relaxation in the proximal stom ach of the horse. The response, measured by an electronic barostat, cons isted of two components: a prompt, marked and defined relaxation phase during meal ingestion, fo llowed shortly by a period of sustained moderate relaxation lasting at least 90 minutes. It was concluded that these two consecutive components might correspond to receptive relaxation (primarily under pharyngeal and esophageal cont rol) and gastric accommodati on (primarily under gastric and duodenal control), respect ively. In addition, there was a significant, positive relationship between the amount of meal and the magnitude of the receptive relaxation

PAGE 29

13 component in the horse. This component is associated with stimulation of mechanoreceptors in the oropharynx and/ or esophagus, whereas accommodation is associated with stimulation arising from the st omach and/or small intestine. Other factors, besides amount of meal, may contribute to vary the relaxation response, thereby modulating the rate of gastric emptying. For example, orosensory stimulation produced by particular nutrients influences appe tite and gastrointestinal responses,27 and may also modulate the relaxation response of the proxima l stomach. All in all, the mechanisms by which different diets modulate gastric tone may be primarily controlled by duodenal receptors responsive to mechanical and chemical stimuli.218 That is, passage of food into the duodenum is detected by and activate speci fic receptors which, in turn, initiate a feedback inhibition pathway that results in a slower gastric emptying rate, in part through decrease in proximal gastric tone. Again, composition of the diet is a major factor in determining gastric emptying rate in other species and seems to modulate relaxation of the proximal stomach in a sitespecific way. As mentioned, perfusion of fat into the proximal, but not the distal, small intestine of the dog causes a strong gastric relaxation response. The opposite effect (no effect proximally and a potent relaxation di stally) is observed with perfusion of carbohydrate solutions.8 Meal fat content and osmolality, but not energy content, can also affect gastric relaxation in humans via a duodeno-g astric feedback mechanism. Dietary fat seems to be a stronger stimulus of gastric relaxation than carbohydrate in man.217 Liquid lipid meals of 2.5% or greater concentration induce a decrease in proximal gastric tone in a non-dose dependent manner, s uggesting a fat-mediated mechanism with threshold sensitivity (no effect for meals of less than 2.5% fat) and saturability (no further

PAGE 30

14 response above the 2.5% dose). Another potential factor in fluencing proximal gastric relaxation is osmolality, although in rats, dogs and primates it only affects gastric tone when meals of very high osmolality (at 2400 mmol/kg or greater) are ingested. Finally, caloric content of a meal per se does not seem to mediate relaxation, since ingestion of carbohydrate solutions, containing an energy load similar to that of relaxation-inducing fat meals, does not cause gastric tone to decrease.129 Although meal volume seems to affect the magnitude of proximal gastric relaxation in the horse,113 it is unknown whether the nutritional composition of the diet influences this response. Nutrients Regulate Gastric Emptying by Neural and Hormonal Pathways The molecular mechanisms underlying chemosensitivity to luminal contents are not completely understood. Interaction of recep tors with nutrients initiates different neural and humoral pathways involved in the comple x regulatory system of gastric motility. Neural pathways Vagal afferent fibers of the upper gut have been found to be sensitive to a range of chemical and physical meal-related properties, including pH, osmolar ity, nutrient content, and the mechanical distension produced by the presence of a load.190 With regard to nutrient content, sensory afferent terminal s innervating the small intestine play a significant role in inhibition of gastric emptying in response to lipid, protein, and glucose.74;164;167 Activation of these afferents initiates a series of neural reflexes that act through autonomic motor nerves to allow regu lation of gastric motili ty by the enteric and central nervous systems. Yet, it is not entirely clear whether vagal afferents are selectively and directly activated by nutrients.56 That is, vagal fi bers are not found between epithelial cells or making direct contact with them, so that direct neural sensing of luminal contents does not probably occu r. Instead, luminal contents may signal to

PAGE 31

15 these fibers via an indirect interaction with a specialized cell situated within the epithelium. A potential intermedia ry is the enteroendocrine ce ll, which releases peptides in response to changes in gut contents.165 These peptides may enter the bloodstream or act in a paracrine mode to stimulate afferent nerve terminals.56 Hormonal pathways The critical peptides involved in feedback si gnaling are gastrin, cholecystokinin (CCK), secretin, glucose-dependent insulinotropic polypeptide (GIP), glucagon-like peptide-1 (GLP-1), neurotensin and peptide YY.21 The ability of luminal nutrients to stimulate the release of these peptides has been extensively studied by monitoring peripheral plasma levels after ingestion of different nutrien ts. Thus, it has been observed that the profile of peptide secretion depends to a great extent on the composition of the meal and the intestinal segm ent exposed to nutrients. Since the majority of food absorption o ccurs in the upper sm all intestine, the endocrine cells located in this region are hi ghly sensitive to the presence of luminal nutrients.21 Fat is the most effective in stimul ating upper intestinal endocrine cells, increasing the secretion of CCK, GIP and secretin. This effect requires hydrolysis of fat with subsequent entry of fatty acids in to the epithelial cells and formation of chylomicrons. On the other hand, glucose infusi on into the upper intestine stimulates the release of GIP, serotonin (5-H T) and CCK, but not secretin.21;166 The release of GIP and CCK depends on the sodium-dependent glucose tr ansporter, suggesting that entry into the endocrine cell is requir ed in this response.171 When unabsorbed nutrients reach the distal sm all intestine, they also stimulate ileal endocrine cells to secr ete GLP-1, neurotensin200 and peptide YY.200;209 These peptides act as the aforementioned ileal brake, slowi ng gastric emptying. Glucose stimulates the

PAGE 32

16 release of the three peptides experimental ly, although it is believe d that this does not occur under physiological circumstances.47 With regard to fat, there are conflicting data about release of neurotensin and peptide YY. It has also been observed that the fatinduced ileal response may occur by direct stim ulation of cells by fat in the distal small intestine and colon or, indirec tly, by the presence of fat in the proximal intestine signaling to the distal gut.47;108 Cholecystokinin: A Key Hormone in th e Regulation of Gastric Emptying CCK plays an important role in the nutrie nt-induced feedback i nhibition of gastric emptying. The mechanism of action is unclear but it seems that its main effect is relaxation of the proximal stomach.170;235 CCK also delays gastric emptying by stimulating contractions of the pyloric sphincter and the proximal duodenum, thus increasing the resistance to gastroduodenal flow of chime.87;235 The relative potency of nutri ents in elevating plasma CCK concentration varies among species. Intact protein and fatty aci ds within the duodenu m are the major food stimulants of CCK release in rats, wher eas protein hydrolysates L-phenylalanine, Ltryptophan, intact fat, starch or glucose do not produce any effect.96;100 As well, mediumchain fatty acids seem to be more powerful stimulators of CCK secretion than long-chain fatty acids.46 In dogs and humans, duodenal infu sion of long-chain fatty acids54;127;204 and amino acids or digested pr otein, but not intact prot ein, are effective stimuli.140;208 On the other hand, carbohydrates in the form of starch and glucose have a weak70;99 or no significant effect in man.75 In pigs, intraduodenal starch hydrolysate, fatty acids and protein hydrolysate evoke CCK release.38 In cows, ingestion of a high concentration of long-chain fatty acids (90 g/kg DM), but not lower concentrations (30-60 g/kg) is associated with increased plasma CCK levels.28 In goats, duodenal infusion of

PAGE 33

17 phenylalanine and tryptophan causes an in crease in plasma CCK, but intraduodenal branched-chain amino acids such as le ucine and isoleucine fail to do so.57 In cats, plasma CCK increases in response to intragastric administration of long-chain triglycerides, intact protein or the amino acids residues, but not starch.10 Finally, intragastric administration of amino acids and medium-cha in, but not long-chai n, triglycerides induce release of CCK in chicks.115;236 Therefore, in all these spec ies, at least one nutrient has been recognized to induce the secret ion of CCK into the circulation. Fat is invariably a potent stimulus of CCK release in all the studied species In humans66 and rats,96 intact fat requires hydrolysis to be effective. Furthermore, in humans, only fatty acids with chain lengths greater than C11 are potent stimulants.127 These are the fatty acids that are absorbed via chylomicron formation, which is required for long-chain triglycerides to slow gast ric emptying and to increase circulating levels of CCK in man61;169 and in rat.61 In contrast to humans, medium-chain fatty acids are more powerful stimulators of CCK secretion than long-chain fatty acids in rat,46 a phenomenon that has also been observed in chicks.115 Therefore, a relationshi p between the pathways of absorption and the ability of fats to initiate feedback responses and release CCK may vary among species. Medium-chain triglycerides are hydrolyzed and absorbed faster and more completely because they simply diffuse across enterocytes, bypa ssing the packaging into chylomicrons and exocytosis reserved for long-chain fatty acids.212 Furthermore, unlike long chain triglycerides, medium-chain triglycerides do not require bile salts for digestion. Since rats and chicks have no gall bladder, bile salts ma y not be as readily available for fat digestion as in other sp ecies, and different mechanisms for inducing CCK secretion may reflect th is difference in physiology.

PAGE 34

18 CCK may be involved in the adaptive cha nge in gastric emptyi ng due to previous feeding with high-fat diets. One study37 demonstrated in rats th at the delay in gastric emptying of saline caused by bot h intestinal oleate infusion and intraperitoneal CCK is attenuated by prior consumption of a high-fat diet. Another study53 showed that plasma levels of CCK were raised after ingesti on of a fatty meal following high-fat diet adaptation in man. Because this raised CCK level was not associated with a reduction in the rate of gastric emptying, it has been suggested that subjects might become desensitized to the effects of CCK after a period of high-fat consumption.25 Finally, the increase in CCK levels that follows diet ary fat adaptation may be secondary to an increase in the capacity to digest and absorb fat,25 which is necessary for CCK release in humans61;169 and rats.96 Current evidence suggests that digested fat releases CCK from enteroendocrine cells, and activates extrinsic afferent nerve te rminals to stimulate a neural reflex that decreases gastric motility.60 CCK receptors are found in vagal terminals149;202 and CCK can activate vagal afferents th rough the CCK-A receptor subtype.61;189 Studies using the sensory neurotoxin capsaicin on vagal and spin al afferents have s hown that different nutrients act to inhibit gastric emptying thr ough distinct afferent pathways. For example, in rats, the inhibitory effect of lipid a nd peptone on gastric empt ying is attenuated by functional ablation of the vagal afferent, but not the spinal, pathway, and by blockade of CCK-A receptors.52;61;74 In contrast, inhibition of ga stric emptying in response to carbohydrates is attenuated by ablation of the vagal afferent pathway, and completely abolished by ablation of the spinal sensory pathway.167 Further evidence of the importance of neural pathways on CCK-mediate d effects is that inhibition of proximal

PAGE 35

19 gastric motility induced by administration of CCK is also abolished by vagotomy and is hexamethonium-sensitive.170 Activation of vagal afferent endings by CCK may occur locally (duodenal mucosa) or peripherally (by circulating CCK). Although the hypothesis that circulating CCK mediates gastric emptying by an endocrine mode of action has been formerly accepted, discrepancies between concentrations of plasma CCK and emptying inhibition by nutrients indicate a non-endocri ne source of CCK participa ting in this process. For example, intestinal perfusi on by carbohydrates inhibits ga stric emptying via a pathway involving vagal capsaicin-sensitive afferent s and CCK-A receptors, and yet does not increase circulating CCK.167 This suggests that the potency of nutrients to inhibit gastric emptying may not always be reflecte d in CCK peripheral plasma levels.168 That is, the CCK-mediated effect of glucose may be local, whereas that of fat, which is associated with increased plasma CCK levels, may be pe ripheral, as well. Finally, other mediators, such as 5-HT, may be more important than CCK in glucose-induced gastric emptying inhibition.166 Finally, there is scarcity of information on the role of CCK in the horse. Endocrine cells immunoreactive for CCK have b een identified with in the duodenal wall88 and in the female urethra.224 In a recent study, CCK-B receptors were identified in somatostatin cells of pancreatic islets.150 However, it is not known whether CCK-A receptors are present in equine non-endocrine pancreas intestinal mucosa or vagal nerves. Gastric Emptying in the Horse: Current Knowledge and Methodology The mechanisms controlling gastric emptyi ng of liquids and solids in the horse are poorly understood. Those done to date have ma inly assessed the effect of prokinetic drugs or clinical disord ers, but not nutrients.

PAGE 36

20 The repertoire of measuring techni ques for gastric emptying is large,187 but their application in the horse may be restricted due to anatomical or economical limitations. First, non-absorbable markers, such as phenol red,12;199 have been used in the horse to measure gastric emptying of liquid meals. In th is technique, ingestion of a labeled meal is followed by aspiration of gastric contents th rough a nasogastric tube, and the amount of marker recovered from the stomach will depe nd on the rate of gastric emptying of the meal. The shortcomings of this method are that the test meal is instilled directly into the stomach, which is not as physiological an appr oach as ingestion; the extraction procedure can be laborious and recovery may not be complete; and, finally, artificial mixing of gastric contents may occur.199 Another indirect measurem ent of liquid-phase gastric emptying involves assay of acetaminophen absorption.43-45;112;134;215;216 After oral administration, acetaminophen is absorbed almost exclusively in the proximal portion of the small intestine, and can be detected in blood. Gastric emptying is the rate-limiting step in acetaminophen absorption, provided that the intestinal mucosa is intact. Thus, both this and the phenol red technique pr ovide a safe and relatively inexpensive qualitative method for the measuremen t of liquid-phase gastric emptying.111 Additionally, use of solid, indi gestible markers has also been reported in the horse.3 Some disadvantages of this method are that par ticles need to be deliver ed directly into the stomach, their size will influence rate of empt ying and their recovery from a live horse is only possible via a gastric cannula. Another alte rnative is the use of radiopaque spheres13 or barium contrast,182 although the resolution of these me thods is adequate only in small subjects.

PAGE 37

21 In contrast to the previous methods, scintigraphy has the advantage of noninvasiveness, and has been a pplied to the horse before.177 Additional advantages of this technique are the possible differentiation be tween solid and liquidphase emptying by the use of specific markers, and the ability to m onitor the intragastric distribution of meal components.219 Although nuclear scintigraphy is the gold standard for measuring gastric emptying in humans, this method is expens ive and may be hazardous for technician health in case of exposure to excessive radioactive material.111 Ultrasonography, magnetic resonance imaging, applied tomogr aphy and changes in gastric impedance can accurately measure gastric emptying in human s, but they have lim ited application to equine medicine because of the animals body size. Finally, breath tests employing stable isotopically-labeled tracers offer a good alternative to all the previous methods, si nce they are non-invasi ve, non-radioactive and easy to perform. Moreover, samples can be stor ed at room temperature for a considerable period of time and be sent to a remote la boratory for automated analysis. For this technique, the test meal is labeled with a 13C-enriched marker. As it empties the stomach, the marker is readily absorbed in the proximal intestine and metabolized to CO2. Thus, breath 13CO2 enrichment reflects the rate of gast ric emptying of the labeled meal. The 13C-octanoate breath test has b een validated in horse and appears to be a promising tool for GE measurement.205 As mentioned earlier, limited studies have a ssessed the effect of nutrients in gastric emptying in the horse. In a study in foals, Baker et al.12 showed that ph enol red-labeled milk emptied slower than water and isotonic saline. However, addi tion of lipids to the saline meal, with a fat concentra tion comparable to that of milk (1.5% fat), did not lead to

PAGE 38

22 slowing of gastric emptying. The authors of th e study suggested that a different type of fatty acid, or the lactose and pr otein might be responsible fo r the slower gastric emptying rate of milk. Experimental factors such as lo w number of experimental units (four foals), the constant presence of a nasogastric tube for phenol red recovery and different degree of fat emulsification may have also influenced the rate of gastric emptying of both meals. Using the same phenol red technique, Sosa Len et al.199 studied gastric emptying of different 8-liter oral hydrati on solutions administered via a nasogastric tube. Isotonic, cold isotonic (5o C), isotonic + dextrose (34.7 g/l) or hypertonic solutions emptied with similar rates, suggesting that temperature, t onicity or glucose cont ent did not significantly affect gastric emptying. However, poor mixing of phenol red and interference with solid contents in some subjects occurred, causing inaccurate measurement. By use of scintigraphy, a preliminary study re vealed that one lite r of 25% dextrose solution emptied slower than the same volume of water or one pound of grain in three ponies,197 although intersubject variation was high. Finally, by use of the 13C-octanoic acid breath test, Wyse et al.234 observed that addition of 30 ml or 70 ml of soya oil to a concentrate meal of oats and bran caused a significant delay in gastric emptying in ponies However, breath test parameters were very similar between both soya-supplemen ted meals, and it was unclear which characteristic of the meal (energy density, viscosity) was responsible for delayed gastric emptying. In another study, Geor et al.58 evaluated the effect of adding corn oil (10%) to a sweet feed meal (2 g/kg bw t) on gastric emptying using the 13C-octanoic acid blood test. Addition of corn oil resulted in delaye d plasma appearance, and this effect was not affected by a 4-week or 8-week period of adaptation to a fat-supplemented diet.

PAGE 39

23 The Effect of Diet on Intragastric pH Squamous Ulceration of the Proximal Stomach Ulceration of the squamous portion of th e equine gastric mucosa is a common problem in ~80% of horses under intense traini ng programs irrespective of type of work or breed of horse.68;85;125;162;223 The degree of lesion severity can be quite variable, but common signs of serious affliction include relu ctance to finish grain meals, periods of low-grade abdominal discomfort especially after ingestion of a meal, and failure to perform up to expectations.221;222 Some of the current idea s concerning the pathogenesis of gastric squamous mucosal ulcer disease in horses have incl uded: 1) excessive intragastric production of vola tile fatty acids (VFA), which have corrosive potential within the stomach, secondary to ingestion of large grain meals;5 2) strict meal-feeding practices themselves; 3) stress-related pr oblems such as decreas ed gastric emptying rate or excessive bile reflux from the small intestine, especially during periods between meals when there is lit tle food in the stomach. Although the pathogenesis of squamous ulcer ation is most probably multifactorial, excessive exposure of the squamous mucosa of the proximal stomach to acid is believed to be the main reason of ulcer development in this gastric region. One study in our lab114 showed that acidic gastric contents are pushed upward during exercise, which may subsequently damage the mucosa. Therefore, modification of these contents may result in a lesser exposure of the mucosa and, thereby, a lesser risk of ulcer de velopment. Potential strategies to influence gastri c contents composition in the equine stomach during exercise could include changes in pre-exercise feedi ng times and formulation of the diet. Thus, it is very important to know how various dietary formulations affect intragastric pH, since this has strong implications regarding their ulce rogenic capacity. Fat-supplemented diets

PAGE 40

24 are becoming a common alternative to cover the high energetic demands of performance horses, and it is of great interest to see how such supplementation can affect intragastric pH. Studies on the Effect of Diet on Intragastric pH in the Horse Few studies have measured meal-induced ch anges in gastric pH in the horse, but buffering capacity of the meal seems to depend on composition of the diet. For example, ingestion of an alfalfa hay-gr ain diet is associated with a significantly higher pH of equine gastric contents, compar ed to a brome grass hay diet.156 This difference may be attributable to a higher calc ium and protein concentration of the former. In addition, ingestion of fermentable carbohydrates is cons idered to decrease intragastric pH by increasing production of lactate and VFA within the stomach. In vitro studies show that some forms of the latter are harmful fo r the equine gastric squamous mucosa.5;157 Study Objectives The main goal of this study was to eval uate the effect of ingestion of a highcarbohydrate versus a high-fat meal, both offe red in isocaloric and isovolumetric amounts, on gastric emptying, proximal gastri c relaxation, intragastric pH and plasma CCK in the fasted and resting horse. This was accomplished by measuring, simu ltaneously, the tone of the proximal stomach with an electronic barost at, and gastric emptying with the 13C-octanoate breath test to establish a novel a pproach for the study of e quine gastric physiology.

PAGE 41

25 CHAPTER 2 MATERIALS AND METHODS Animals Six adult horses (3 mares, 3 geldings) w ith previously insert ed gastric cannulas were used for the experiments. The animals weighed between 350 and 546 kg (average 479 kg) and aged 5 to 18 years. They were housed in paddocks and maintained on free choice of Coastal Bermuda hay, bahia grass pa sture and trace minerals. All studies were approved by the Institutional Animal Care and Use Committee of the University of Florida. Barostat A specially designed barostat (Isobar 3, G & J Electronics Inc., Willowdale, Ont.) was used to assess changes in volume/pressure of the proximal stomach (Fig. 2-1). The validity of an electronic barostat to measur e proximal gastric motility and tone was first demonstrated in dog by Azpiroz and Malagelada,8 and work in our lab113 has also proved its usefulness for the study of proximal gastri c tone in the horse. The basic principle of the barostat is to maintain a constant pressure within an air-filled intragastric bag that has infinite compliance and a volume greater than the range of volumes to be used by the barostat during the study. When the stomach c ontracts, the barostat removes air from the bag to maintain the intrabag pr essure constant and, conversely, when the stomach relaxes, air is injected into the bag. Thus, the changes in bag volume are a direct indication of changes in intragastric pressure induced by variations in wall tone.

PAGE 42

26 A)Decrease in Tone B) Increase in Tone PressurePressure A)Decrease in Tone B) Increase in Tone PressurePressure A)Decrease in Tone B) Increase in Tone PressurePressure Figure 2-1. Motility of the stomach measured by an electronic barostat. The barostat maintains a constant pressure within an intragastric bag. When the stomach relaxes, the system injects air into th e bag (A). Conversely, when the stomach contracts, air is aspirated from the bag (B). For this study, intrabag pressure was set at 2 0.5 mm Hg. A plastic bag (Commercial Mylar balloon, The New Garde n, Gainesville, FL) of ~1600 ml total capacity and 20-cm diameter (Fig. 2-2) wa s connected to the ba rostat through a 1.5-m long plastic catheter of 4 mm internal diam eter (inflation line). A separate catheter (pressure line) of 1.6 mm internal diameter c onnected the bag to the pressure transducer, so that pressure was monitored directly within the bag. The tip of both lines were attached to each other inside the bag and sealed to the bag 18 cm from the tips using dental floss. Finally, a plas tic probe of 35.5 cm was attached to the portion of the inflation line near the bag to increase the catheter rigidity inside the stomach. The

PAGE 43

27 polyester bag had infinite compliance, i.e., th e bag did not have any influence by itself on the internal pressure.229 Figure 2-2. Polyester bag connected to a pressure line and an inflation line. Analogue signals from the recording inst ruments were digitized and displayed by use of a computer program (WinDaq graphi cs package, DATAQ Instruments Inc., Akron, OH) at a sample rate of 50 Hz/channel. Changes in bag volume (ml) and pressure (mmHg) were also recorded in an additiona l computer program (Protocol Plus Deluxe 3.61 DH, G & J Electronics Inc ., Toronto, Ont.) provided by the manufacturer of the barostat. Following an experiment, the reco rded data were tran sferred to a CD for permanent record. 13C-Octanoic Acid Breath Test Rates of gastric emptying were assessed with the 13C-octanoic acid breath test validated by Sutton et al.205 This non-invasive, non-radioact ive technique is based on the

PAGE 44

28 detection of 13C enrichment in breath following the ingestion of a m eal labeled with 13Coctanoate. This marker is rapidly and totally absorbed in the small intestine and oxidized in the liver to produce CO2.233 For each test, approximately 1.5 mg/kg bwt of 13Coctanoic acid (Octanoic-1-13C acid, 99 atom % 13C, Sigma-Al drich, St Louis, MO) was added to egg yolk (1 yolk/250 mg of marker), baked in a microwave oven and thoroughly mixed into the test meal. This dose of octanoi c acid is higher than the one previously used by Sutton et al. (1 mg/kg).205 Sutton (personal commun ication, 2002) found that Bermuda grass is natu rally enriched with 13C, and suggested a higher dose to compensate for this source of exhaled 13CO2. Preparation of the octanoate -enriched yolk was done on the day prior to the experiment, and stored in the refrigerator until use in order to reduce the risk of feed rejection. The enriched yolk was added to the test meal and thoroughly mixed five minutes before it was offered to the horse. Intragastric pH Monitoring The pH of gastric contents was continua lly sampled by a self-referencing pH electrode (24-hour pH cath eter, Medtronic Functional Diagnostics A/S, Skovlunde, Denmark) that was inserted up through the gast ric cannula so that its tip protruded ~2 cm into the gastric lumen. The electrode wa s attached to a data collection device (Medtronics, Shoreview, MN) that sampled pH every 4 seconds for up to 24 hours, and stored the results (Digitrapper, Medtronics, Shoreview, MN). Recorded pH data were subsequently downloaded onto a computer th rough the use of software provided by the manufacturer (Esophagram MD, Medtronic, Shoreview, MN). Plasma Cholecystokinin Radioimmunoassay Plasma CCK-like activity was measured using a commercial, non-specific (human) radioimmunoassay (RIA) kit (Alpco Diagnostics, Windham, NH). Although this kit is

PAGE 45

29 commercialized for human CCK, this protein is well conserved in vertebrates, which suggests that there is a high stru ctural homology between both species.83 A different non-speci fic CCK RIA kit (Peninsula Laboratories Inc., San Carlos, CA), which had full cross-reactivity with ga strin, was used in pr eliminary studies to determine whether equine plasma or serum w ould be more useful to detect CCK. Since CCK and/or gastrin levels were obtained when using plasma, but not serum, the following assays were performed with plasma. Plasma was collected and the proteins ex tracted according to the manufacturers guidelines. To study which method of sample collection maximized CCK detection, we collected blood in EDTA tubes with and without aprotinin (Sigma, St. Louis, MO), i.e., 500 kallikrein inhibitory equivalents/ml of blood. Since CCK might be easily and readily degraded, we also compared the effect of centrifuging each sample immediately after collection, or after completion of sample collect ion (2.5 hours). In both cases, plasma was stored at -70C after centrifugation. Sample t ubes and syringes were constantly cooled in an ice-bath before and after use. Thawed samples were extracted with 96% ethanol and evaporated overnight using a Speed Vac Concentrator (Savant Instru ments, Inc., Holbrook, NY) at 37C. Dry extracts were dissolved to the original samp le volume with a diluent provided by the kit. A recovery control with CCK-8 (amino acids 26-33) sulphate was included to estimate the extraction recovery. A four-day assay wa s performed using rabbit antiserum raised against synthetic CCK-8 sulphate. Antibody-bound 125I-CCK sulphate was separated from the unbound fraction using double antibody solid phase.

PAGE 46

30 A separate experiment was performe d to collect plasma samples for in vitro bioassay measurement. The bioassay, developed by Liddle et al.,99 is based on the ability of bioactive CCK peptides extr acted from plasma to stimulate amylase release of isolated rat pancreatic acini, and is de scribed in detail elsewhere.99 Plasma samples were collected from one horse. After an overnight fast, th e horse was offered 226 g of sweet feed (0.5 g/kg) mixed with 30 ml of corn oil and 5 g of phenylalanine. D uplicate blood samples were obtained before the meal and 10, 20, 30 and 50 minutes after meal ingestion. Blood was collected in EDTA tubes with and without aprotinin, and constantly cooled in an ice bath until the end of the co llection. Plasma samples were extracted and brought to dryness as done for the CCK RIA. Dry samples of 1-ml and 5-ml original plasma volume were shipped to the University of Sher brook, Canada, for CCK bioassay. All samples were reconstituted with 1 ml of distilled wate r, so that samples with original volume and concentrated samples were assayed. Increasing concentrations of caerulein, an analog of CCK, were used as a positive control. Test Meals Composition The study consisted of two consecutive pha ses that differed in the test meals offered to the horses. Detailed composition of the different diets is shown in Tables 2-1 and 2-2. Phase I Studies In phase I studies, two isocaloric (1.5 kcal/kg bwt) and nearly isovolumetric pelleted meals (16% protein) were used. The high-carbohydrate pelleted meal was rich in starch (31%) and poor in fat (3%), whereas th e high-fat pelleted meal was rich in fat (8%), had no starch, and cont ained more fiber (43.5% NDF in the high-fat meal versus 28.4% NDF in the high-carbohydrate meal).

PAGE 47

31 Table 2-1. Approximate composition of the pelleted test meals used in the Phase I studies. Ingredients High-fat pelleted meal High-CHO pelleted meal (g/kg feed) Beet pulp 290 Soybean hulls 290 Corn meal 327 Oats Ground 320 48 Soybean meal 167 149 Wheat middling 71 70 Alfalfa meal 50 50 Cane molasses 50 50 Soybean oil 50 Biophosphate 14 9 Salt mixing 7 7 Calcium carbonate 6 14 Vitamin E 3.65 3.45 TM Premix 0.35 0.4 Vitamine Premix 6 0.25 0.25 Sodium Selenite 0.04 0.06 (relative composition) Dry Matter 88.20% 87.96% Starch 31.14% Fat 8% 3% Acid detergent fiber 28.40% 11.53% Neutral detergent fiber 43.50% 28.40% Protein 16.27% 16.18% Calcium 0.89% 0.91% CHO: carbohydrate. Phase II Studies Since diets used in the first part of the study differed in more than fat and carbohydrate composition, the phase II studies we re performed to determine the specific effect of these two components. Accordingl y, two new isocaloric (1.95 kcal/kg bwt) and isovolumetric diets were formul ated. A control meal consisting of 0.5 g/kg bwt of 10%

PAGE 48

32 protein sweet feed (Seminole Feed, Blue Ribbon 10) with 1.5 kcal/kg was used as the basis to prepare the two experi mental diets. A high-fat meal (12.3% fat) was prepared by adding 0.05 g/kg bwt of corn oil to the cont rol meal. A high-carbohydrate meal (2.9% fat) was prepared by adding 0.113 g/kg bwt of gluc ose to the control diet. To account for influence of the sweet feed in the results, th e control meal was included as an additional experimental meal. Table 2-2. Composition of the sweet feed meal (Seminole Feed, Blue Ribbon 10) used in the Phase II studies. A high-fat meal and a high-carbohydrate meal were prepared by adding 0.05 g/kg bwt of corn oil and 0.113 g/kg bwt of glucose, respectively, to the control sweet feed meal. Crude Protein (min) 10% Zinc (min) 120 ppm Crude Fat (min) 3.50% Copper (min) 40 ppm Crude Fiber (max) 6% Selenium (min) 0.3 ppm Calcium (min) 0.45% Vitamin A (min) 13200 IU/kg Calcium (max) 0.55% Vitamin D3 (min) 880 IU/kg Phosphorous (min) 0.35% Vitamin E (min) 92 IU/kg Study Design Phase I Studies In phase I, each horse participated in four experiments involving two sessions per diet (either with or without labeling the test meal with 13C-octanoic acid). The sequence of the experiments was based on a 2-period ra ndomized block design. In the first period, group A (3 horses) received the high-fat pellet ed diet and group B (3 horses) received the high-carbohydrate pelleted diet. Horses were gradually acclimated to the test diets during a 1-week period. At the end of the accommodation period, th ey received 5 g/kg bwt/day of the respective test diet, di vided in two feedings, until the completion of the first period. Horses were also fed free choice Be rmuda grass hay for the entire period. After the dietary accommodation period, horses completed two randomly assigned studies. In one study, baseline breath tests were performed with non-labeled meals to

PAGE 49

33 measure the presence of natural 13C. Some substrates like st arch are derived from C4 plants (corn in this case), wh ich, unlike most plants (i.e., C3 plants), have a CO2 incorporation mechanism th at results in a higher 13C content.206 The other study consisted of 13C-octanoate breath tests. Therefore, the only difference between both studies was that in basal studies no additional 13C-substrate was added in the test meals. After completion of both experiments, the diets were gradually switched between groups, and the two different experiments rep eated. All test meals were fed at 0.5 g/kg bwt. Phase II Studies In phase II, horses were maintained on 0.25 g/kg bwt/day of sweet feed (10% crude protein) twice a day, and free choice Bermuda grass hay. Each horse pa rticipated in three experiments using a random sequence. Unlike the first part of the study, no baseline breath tests were performed. The only difference among the three experiments was the meal offered to the horses (control, high-fa t or high-carbohydrate meal). All test meals were fed at 0.5 g/kg bwt. Experimental Procedure No horse participated in more than on e experiment per week. Food was withheld for 14 hours before each experiment. After this period, the horse was placed in the stocks and the gastric cannula cleansed. Status of th e squamous mucosa of the proximal stomach was evaluated with an endoscope introdu ced through the cannula (Fig. 2-3). The previously folded barostat bag, along with an attached pH probe, was introduced into the stomach through the cannula and inflated manually to ensure it became unfolded (Fig. 24). Position of the barostat bag within the proximal stomach was verified by the

PAGE 50

34 endoscope introduced nasogastrically (Fig. 25). Correct position wa s defined as being above the margo plicatus. Figure 2-3. Endoscopic view of the squa mous mucosa of the proximal stomach. Endoscopy was performed through the ga stric cannula at the beginning of every experiment to evaluate the status of the squamous mucosa. Figure 2-4. Positioning of the polyester bag used to measure intragastric pressure in the proximal stomach.113 The bag was introduced thr ough a previously inserted cannula.

PAGE 51

35 Figure 2-5. Endoscopic views s howing correct position of th e barostat bag within the proximal stomach. Once in the proximal stomach, the bag was emptied by a syringe and connected to the barostat by the catheter The data collection instrument for pH recording was attached to a harness placed around the thorax of the horse for the whole experiment (Fig. 2-6 and 2-7). The barostat was set to maintain a c onstant intrabag pr essure of 2 mmHg. Thereafter, activity of the proximal stomach was recorded during two hours by changes in volume of the isobarically controlled bag. The first 30 minutes of the experiment were recorded to obtain a baseline volume. The n, the horse was offered only one of the possible meals and recording continued for a total of 120 minutes. Duration of meal ingestion was also recorded. For plasma CCK measurement, a jugular ca theter was placed before starting the experiment and blood was withdrawn using ch illed syringes. Samp les were obtained 10 minutes before feeding, and then every 15 mi nutes for a total of 120 minutes. Collected blood was transferred into chilled EDTA t ubes (1 mg/ml of blood) with and without aprotinin, for later RIA comparison. Samples we re centrifuged and the plasma was stored at -70C until CCK determination.

PAGE 52

36 Figure 2-6. Motility of the proximal stomach was measured with an intragastric bag, inserted through the gastric cannula and connected to the electronic barostat by two separate catheters. One cathete r was used by the barostat for air injection and withdrawal, and the other catheter was connected to a pressure transducer. A pH probe was also insert ed through the cannu la. After correct position of the bag was confirmed by e ndoscopy, the cannula was clamped to avoid leaking of gastric conten t during the entire experiment. Breath samples were collected using a modified Aeromask (Trudell Medical International, London, Ont.) fitted with a 250-ml aluminum coated polyethylene bag (QuinTron Instrument Company, Milwakee, W I). The horse was allowed to breathe once through the mask before filling the bag, whic h was fitted with a unidirectional valve. Duplicate samples were transferred from this bag to 10-ml red cup tubes, conveniently sealed and stored until ready for stable isotop e analysis. Three basal breath samples were collected 60, 15 and 5 minutes before test meal ingestion, and th ereafter at 15-minute intervals for 3 hours, then 30-minute in tervals for a further 3 hours. The 13C:12C ratio of each breath sample was determined by automated continuous flow isotope ratio mass

PAGE 53

37 spectrometry (PDZ Europa ABCA analys is, PDZ Europa Ltd., Sandbach, UK) and expressed relative to an inte rnational standard. This rate was conve rted to parts per million (ppm) 13C, and expressed as ppm excess 13C, after subtraction of the average 13Cabundance of the three baseline breath sample s. The percentage dose recovery (PDR) of the administered isotope in the breath was also calculated, and plotted against time. Figure 2-7. The pH of gastric contents was measured using a pH electrode inserted through the gastric cannula and positione d in the most ventral part of the stomach. This electrode was connected to a data collection device (arrow) attached to a surcingle during the entire length of the experiment. Barostat recording and blood collecti on were entirely done with the horse in the stocks. Once these two components of the st udy were finished (2.5 h from beginning of the study), the horse was moved into a stall a nd the rest of breath samples collected there. Since it was very difficult to remove the intr agastric bag after feeding the horse, and to avoid loss of food through the cannula, remova l of the bag was done after completion of the study.

PAGE 54

38 Following each experiment, the recordi ng equipment was removed, the gastric cannula was plugged and the horse was returned to the paddock. Data Analysis All statistical analyses were performed us ing SAS version 8.2 (SAS Institute Inc., Cary, NC). All results are shown as mean SEM. Significance was set at p<0.05. Ingestion Time Time of ingestion among diets was compar ed by repeated measures analysis of variance (ANOVA). Data were previously tested for normality (SAS: proc univariate) and homogeneity of variance (SAS: proc gl m). A Friedmans two-way ANOVA test was used when there was inequality of variance. Proximal Gastric Compliance One bag-volume measurement per second was obtained throughout the experiments. For every diet, the data of the experiment s were grouped into 2-minute blocks and averaged for the six horses. The blocks comprising the first 30 minutes were used to obtain a baseline bag volume. The remaining blocks (90 minutes) corresponding to the postfeeding period were analyzed to study the relaxation response of the proximal stomach in relation to the baseline volume Accordingly, the average of the baseline blocks was subtracted from each postfeeding block, in order to account for baseline differences among diets. Blocks of different diets were then compared by repeated measures ANOVA using SAS Mixed Procedure. Time was set as a fixed effect, whereas horse was considered a random effect. Statis tical comparison of mean baseline volumes among different diets was also performed to measure reproducibility. Finally, mean baseline was also compared with postf eeding blocks within the same diet.

PAGE 55

39 Intragastric pH Similar to the analysis for intragastric volume, pH data were grouped in 5-minute blocks. The blocks comprising the first 30 minutes were used to obtain a baseline. Although pH was continuously monitored for 7.5 h, only the first 2 postprandial hours were used for comparison with baseline within diet and between diets. Mean values of the two diets were determined separately and compared by use of Proc Mixed ANOVA. Breath Samples Calculation of gastric emptying parameters All samples containing less than 0.5% CO2 were rejected to minimize analytical inaccuracies. Data were plotted against time as either ppm 13C-enrichment, or the percentage of the isotopic dose recovered per hour (PDR/hour). The 13CO2 excretion curve (PDR/h) was plotted against time using the formula: (i) y = atbe-ct where y is the cumulative percentage of 13C excretion in breath,195 t is the time in hours and m, k and are constants with m describing the total 13C recovery when time is infinite. This formula is derived from the finding that the inverse curve for cumulative dose recovery is empirically analogous to the scintigraphic curve of gastric emptying. The best fit curves, a nd hence the constants a,b,c,m,k and above, were calculated using least squares non-linear regres sion analysis, programmed into a Microsoft Excel Solver function (Microsoft Corporation, Redmond, WA). Using the above constants the following parameters of gastric emptying we re calculated: (a) the gastric emptying coefficient (GEC), equivalent to the natural logarithm of a and considered as a global index of the rate of gastric emptying.117 The GEC reflects the gradient of the emptying curve and is a universal index of gastric emptying rate; (b) the gastric half-emptying time

PAGE 56

40 (t1/2), equivalent to the time at wh ich the area under fitted cumulative 13C excretion curve demonstrates recovery of half the administered isotopic dose. T1/2 was calculated using both the Excel function Gammainv (0.5; b + 1; 1/ c ) (58) and Siegels method195 t1/2 = 1n[1 2-1/ ]/k; (c) the time to peak breath 13CO2 (tmax), calculated as b/c. Effect of diet on basal 13C output Effect of 13C-abundant components in the diet s was only studied by observational analysis during the Phase I of studies (highfat versus high-carbohydrate pelleted feed). Effect of the octanoic acid-loaded diets on 13C output The effect of diet on parameters of gastric emptying (GEC, t1/2 and tmax) was determined by use of paired difference t-test For unpaired data, a 2-sample t-test was performed. When the response variable was not normally distributed, a Friedmans twoway ANOVA test was performed.

PAGE 57

41 CHAPTER 3 RESULTS AND DISCUSSION TONE OF THE PROXIMAL STOMACH Ingestion of a meal induces relaxation of the proximal stomach to accommodate the meal without a significant rise in gastric pressure. This process has two components that are generated at different levels of the upper gastrointestinal tract. First, ingestion and swallowing of the meal stimulates recepto rs in the oropharynx and esophagus, and initiates a vago-vagal reflex that decreas es the tone of the proximal stomach.124 This first component is termed receptiv e relaxation because the pr oximal stomach relaxes in anticipation of the arrival of ingested material. Next, bot h arrival of meal into the stomach and its passage into the small intestine triggers an additional vago-vagal reflex that causes further relaxation. This second component is known as adaptive relaxation or accommodation.2;124 For the purposes of this chapter, only the term accommodation has been used to refer to this component. The magnitude of postprandial relaxation is related to the composition of the diet,8;129;217 and dietary fat appears to cause gr eater relaxation than dietary carbohydrate in humans.129;217 The hypothesis of this study was that in the horse, inge stion of a highfat meal would induce greater relaxation of the proximal stomach than ingestion of an isocaloric and isovolumetric high-carbohydrate meal To test that, changes in tone of the proximal stomach were measured with an elect ronic barostat after i ngestion of a high-fat or a high-carbohydrate meal.

PAGE 58

42 Effect of 13C-Octanoic Acid Labeling (Breath Test) Phase I Studies: Pelleted Diets In the first phase of the study, every horse participated in four experiments, which differed in the test meal offered to th e horse: a high-fat pelleted meal, a highcarbohydrate pelleted meal, or any of these meals enriched with 13C-octanoic acid (used as a marker for the breath test). Followi ng a randomized block design, horses were gradually adapted to the highfat or the high-carbohydrate diet for a minimum of 1 week before any experiment was perfor med using the respective diet. Responses in tone of the proximal stomach induced by ingestion of the unlabeled meals were compared to those induced by the labeled meals. Addition of the 13C-octanoic acid, which was mixed with egg yolk and used as the gastric emptying marker in the test meal, caused a significantly (p<0.05) greate r accommodation of th e proximal stomach after ingestion of the high-carbohydrate meal than after the same meal without the marker (Fig. 3-1). In contrast to the high -carbohydrate meal, no difference in accommodation was found between the high-fat meal and the same meal with the marker when mean bag volumes were compared within the same time intervals. An increment in fat content of the labe led high-carbohydrate meal may account for the longer accommodation observed after inge stion of this meal, compared to the unlabeled meal. For the breath test technique, a dose of 1.5 mg/kg of 13C-octanoic acid, a medium-chain fatty acid, was prepared in egg yolks (1 yolk/250 mg of marker) and mixed with the test meal. In a 500-kg horse, th is label would consist of 3 yolks and 750 mg of octanoic acid. Labeling of the high-carboh ydrate meal thus result ed in a change of 3% fat to 9% fat, whereas labeling of the hi gh-fat meal increased the amount of fat from

PAGE 59

43 8% to 13.6%. The magnitude of accommodation is related to the lipid content of the meal in humans, and concentrations of 2.5% or greater are needed to induce relaxation129;213. A threshold may exist in the horse as well, a nd may have been surpassed with labeling of the high-carbohydrate meal. Time after ingestion (min) 0306090Bag volume (mL) -200 0 200 400 600 800 Plain high-CHO pelleted meal Octanoic-enriched high-CHO pelleted meal * * * Figure 3-1. Changes in intragastric bag volum e after ingestion of the high-carbohydrate (CHO) pelleted meal with and without addition of 13C-octanoic acid. Data are expressed as mean bag volume SE M of 2-min blocks (n=6). Each postprandial volume represents values from which the baseline volume has been subtracted. Asterisks denote a si gnificant difference between pairs of blocks at the same postprandial tim e (p<0.05). Lines on top of the figure delimit the period where mean bag vo lume was significantly higher than baseline after ingestion of the meal with (dashed arrow) and without (continuous arrow) octanoic acid (p<0.05) Note: Blocks with negative values represent mean bag volumes that were lower than the baseline volume. In contrast to the high-carbohydrate meal labeling of the high-fat meal had no effect on the magnitude of accommodation desp ite the increase in fat content. One possible explanation is that, in the horse, fat induces relaxation in a dose-independent

PAGE 60

44 way. That is, beyond a threshold dose, incr easing doses of fat do not cause greater relaxation. This dose-independency has been reported in humans, although it is inconsistent in the literature. Two studie s have shown that in tragastric delivery129 or duodenal perfusion51 of liquid meals with different lipid concentra tion produce a nondose dependent reduction in ga stric tone, whereas one study213 states that the magnitude of relaxation increases with ingest ion of increasing amounts of fat. One goal of the present study was the simu ltaneous assessment of gastric emptying and proximal gastric relaxation, which are in terdependent, to examine the integrated effect of meal ingestion on gastric motilit y. This was achieved by using an electronic barostat and the 13C-octanoic acid breath test concurre ntly. The combined use of these two techniques is valid only if none of them interferes with the results of each other. Because the breath test procedure affected th e barostat response of the high-carbohydrate pelleted meal, but not that of the high-fat pelleted meal the combined use of both techniques was not considered valid in th ese Phase I studies. Bearing in mind that interference between the two techniques seems to occur depending on the test meal composition, the effect of labeling the test meals used in Phase II on the meal-induced relaxation response was investigated. Phase II Studies: Sweet feed Diets Test meals consisted of a 10% crude pr otein sweet feed meal (0.5 g/kg bwt) supplemented isocalorically with either corn o il or glucose. Horses received a daily sweet feed ration (0.25 g/kg bwt/day), along with fr ee-choice Bermuda hay, for the duration of the study. Changes in proximal gastric tone were evaluated using the test meals with and without labeling with 13C-octanoic acid.

PAGE 61

45 Adding octanoic acid to the corn oilor the glucose-enriched meals did not have any significant effect on the relaxation respons e compared to that r ecorded after ingestion of the unlabeled meals. Influence of Dietary Composition on the Effect of 13C-Octanoic Acid Labeling It is difficult to explain why addition of octanoic acid affected the accommodation induced by ingestion of the high-carbohydrate pelleted meal, but not that induced by ingestion of the sweet feed meals. The extra fat content of the brea th test marker may account for the longer accommodation obser ved after labeling the high-carbohydrate pelleted meal (fat increase from 3 to 9%). In contrast, although addition of the octanoic acid to the sweet feed meal supplemented w ith glucose increased the lipid content from 2.9% to 7.7%, the magnitude of accommoda tion was unaffected. Unlike the pelleted meals, bag volumes after any of the sweet feed meals remained above baseline (see section Effect of dietary composition: Phase II), and the additive effect of the octanoic acid may have been masked by this greater accommodation response. Therefore, because composition of the test meal may determine whether 13C-octanoic enrichment influences the meal-induced accommodation of the proximal stomach, the possible existence of such influence should be determined a priori in any study where the 13C-octanoic acid breath test is used in conjunction with the meas urement of proximal gastric tone. Because it existed in the Phase I studies, no integrat ed analyses of the gastric emptying and relaxation results were performed for this Phase.

PAGE 62

46 Effect of Dietary Composition Phase I. Pelleted Meals: Fat Versus Carbohydrate As mentioned previously, test meals consisted of a high-fat pelleted meal or a highcarbohydrate pelleted meal. None of the meals were labeled with 13C-octanoic acid for these experiments. Duration of meal ingestion. Mean SEM duration for complete meal ingestion (time to empty the food bucket) for the hi gh-fat meal was 228 30 seconds (range, 156 to 368 seconds), whereas mean duration for ingestion of the high-carbohydrate meal was 158 14 seconds (range, 120 to 208 seconds). Th us, horses spent a significantly (p=0.02) longer time ingesting the high-fat meal. Relaxation response. Baseline volume of the barostat bag did not differ significantly between diets. Re laxation of the proximal portio n of the stomach, indicated by an increase in bag volume, was observed in response to ingesti on of both meals (Fig. 3-2). Bag volume began to increase rapidly after ingestion, reached a peak volume, and then decreased sharply by the end of ingestion until returning to baseline volumes. This receptive relaxation episode lasted 6 minutes with ingestion of the high-fat meal and 10 minutes with ingestion of the highcarbohydrate meal. A second, less profound, significant increase in bag volume (accommodation) was observed one hour after ingestion of the high-fat meal with bag volumes remaining significantly (p<0.05) higher than baseline throughout most of the remainder of the recording period (Fig. 3-3). In contrast, this effect was not observed with the high-carbohydrate diet, except for discrete periods of time in which bag volume was signi ficantly (p<0.05) higher than baseline. Finally, there was no significant differen ce for any time interval between diets.

PAGE 63

47 Figure 3-2. Mean bag volume revealing the eff ect of intake of a high-fat (A) and a highcarbohydrate (B) pelleted diets (0.5 g/kg) on gastric tone of the proximal portion of the stomach in 6 horses. An increase in volume, indicating decreased gastric tone, was observed shor tly after beginning of ingestion for either diet. The initial peak response (receptive relaxation) was associated with the period of activ e food ingestion (box). Fo llowing this peak, bag volume returned to baseline levels. A second significant increase in bag volume (accommodation) was observed one hour after ingestion of the highfat pelleted meal, and bag volume remain ed significantly higher than baseline throughout most of the rest of the re cording period. Time 0 is start of recording period, and beginning of ingestion was at 30 minutes. A previous study113 in our lab demonstrated the ex istence of a solid meal-induced physiological relaxation in the proximal stomach of the horse, which could be divided in two components: a prompt, marked and defi ned relaxation phase dur ing ingestion and a second phase of sustained moderate relaxation lasting at least 90 mi nutes. The results of the present studies are similar to that previ ous one in that beginning of meal ingestion was followed by rapid onset of proximal gast ric relaxation, the duration of which was associated with duration of active ingestion. Ingestion of a meal in itiates a series of reflexes in order to receive and store ingesta within the st omach. Initially, stimulation of

PAGE 64

48 receptors within the oropharynx and esopha gus triggers receptive relaxation of the proximal stomach.1 Since the initial relaxation observed in the current study was coincident with time of active ingestion, th is relaxation episode is consistent with receptive relaxation, and it was probably induced by activation of oropharyngeal receptors. Time after ingestion (min) 0306090Bag volume (mL) -200 0 200 400 600 High-fat pelleted diet High-CHO pelleted diet * * * *** * * Figure 3-3. Changes in intragastric bag volume after ingestion of eith er the high-fat meal or the high-carbohydrate (CHO) pelleted meal. Data are expressed as mean bag volume SEM of 2-min blocks (n=6). Each postprandial volume represents values from which the baseline volume has been subtracted. Symbols (*) and () denote blocks th at were significantly different from baseline after ingestion of the highfat meal and high-carbohydrate meal, respectively. Neither meal contained 13C-octanoic acid. Note: Blocks with negative values represent mean bag volumes that were lower than the baseline volume. Data from previous work in our lab113 suggested that duration of receptive relaxation increases with durati on of meal ingestion in the ho rse. In particular, ingestion

PAGE 65

49 of a large hay meal took longer time to be finished and resulted in longer receptive relaxation than ingestion of a smaller hay meal, or two different sweet feed meals of weights identical to those of the hay meals. However, this relationship was not found in the present study, since ingestion of the high-fa t meal took longer time to be finished, but resulted in a shorter receptive relaxation pe riod, than ingestion of the high-carbohydrate meal (Fig. 3-3). It may be possible that di fferences in degree of receptive relaxation between diets are associated with duration of ingestion only when meals differ largely in the latter. For example, the difference in duration of ingestion between the large hay meal and the other meals of the previous study113 was considerably great er (448-889 sec) than the difference between the two pe lleted meals used in the pr esent study (70 sec) (Table 31). Table 3-1. Comparison of duration of meal ingestion between the present study (A) and that of Lorenzo-Figueras and Merritt113 (B). Data are expressed as mean (sec) SEM. In (A), ingestion of the high-carbohydrate (CHO) pelleted meal induced longer receptive relaxation than the high-fat pelleted meal. In (B), ingestion of hay (1 g/kg bwt) induced longer receptive relaxation than any of the other meals (p<0.05). A) Present study Duration of ingestion Difference between high-fat and high-CHO Pelleted meals (0.5 g/kg bwt) High-fat 228 30a High-CHO 158 14b 70 sec Sweet feed meals (0.5 g/kg bwt) High-fat 241 31 High-CHO 267 35 26 sec B) Lorenzo-Figueras and Merritt, 2002 Duration of ingestion Difference from Hay (1 g/kg) Large meal (1 g/kg bwt) Hay 1065 85a 0 sec Sweet feed 281 15b,d 784 sec Small meal (0.5 g/kg bwt) Hay 617 79b,c 448 sec Sweet feed 176 17b,d 889 sec Within the same study (A or B), values with different superscript letters differ significantly (p<0.05).

PAGE 66

50 Therefore, in the present study, the differe nce in consumption time between the pelleted meals, although significant, may have not b een large enough to produce differences in receptive relaxation. As indicated, receptive relaxation induced by ingestion of the high-carbohydrate meal lasted 4 minutes longer than that induced by ingest ion of the high-fat meal. If receptive relaxation were solely dependent on mechanical stimulation of the oropharynx, composition of the diet would have no in fluence on the magnitude of relaxation. However, the results of this study suggest that other mechanisms may play a role in the control of receptive relaxation in the hor se. One of the possible factors may be orosensory stimulation. That is, meals va rying in composition may provide different orosensory signals that may, in turn, i nduce different gastroin testinal responses.27 The importance of orosensory factors controlling other gastrointestinal parameters, such as hunger, satiation and gastric emptyi ng, has been previously shown.26;27 For example, in one study,27 oral administration of a high-fat soup slowed gastric emptying more than an isocaloric high-carbohydrate soup, but gastri c emptying rates were similar when the respective soups were infused into the stomach. Therefore, suppression of orosensory stimulation eliminated the effect of diet ary composition on the control of gastric emptying. Likewise, subtle differences in or osensory stimulation following ingestion of the high-fat meal and the hi gh-carbohydrate meal may be responsible for the observed difference in receptive relaxation. In conclusion, control of gastric receptive relaxation in the horse may depend on other factors besides mechanical stimulation. The initial relaxation episode, which has been defined as receptive relaxation, was followed by a period of baseline tone, until the onset of a second, less profound,

PAGE 67

51 accommodation phase, starting at about one hour of ingestion of the high-fat meal (Fig. 3-3). However, this was not observed fo llowing the high-carbohydrate meal. It is proposed that accommodation may be induced by activation of either of two vago-vagal reflexes. The first one begins when passage of feed into the stomach stimulates mechanoreceptors within the gastric wall. The second one is initiated when chyme delivered into the intestine stimulates intestinal mechano-41 and chemoreceptors.8 Activation of any of these reflex pathways will result in such accommodation or adaptive relaxation.2 In the present study, accommodation was observed long after the meal had entered the stomach and, therefore, it is more plausible that it was elicited by small intestinal feedback regulation. In the previous study in our lab,113 receptive relaxation after ingestion of a sweet feed meal, of equal weight (0.5 g/kg) and similar volume (450 ml) to these of the pelleted diets, was immediately followed by accommodation. Since the accommodation reflex is presumably triggered by stimulation of gastric mechanoreceptors, the same response pattern wo uld be expected when the pelleted meals of the present study, which had a simila r volume (400 ml), entered the stomach. However, in this case, the accommodation was not observed until one hour after ingestion, long after the receptiv e response (Fig. 3-3). In conc lusion, these results suggest that the accommodation response to the highfat meal was induced by passage of food into the small intestine. In contrast to the high-fat meal, no acco mmodation was observed after ingestion of the high-carbohydrate meal, except for short, discrete intervals towards the end of the recording session. Since barostat recording was limited to a postprandial period of 90 minutes, we cannot discount the possibility that accommodation occurred late after

PAGE 68

52 ingestion. A higher fat content of the high-fa t meal may explain the earlier occurrence of accommodation that was not seen after the hi gh-carbohydrate meal, i.e., the greater the amount of fat in the diet, the earlier the accommodation. As observed in humans,51;129;217 the magnitude of intestinal feedback on motility of the proximal stomach may be dependent on nutrient composition, where fat induces accommodation, and hypertonic, but not isosmotic, carbohydrate prepar ations elicit the same response.51;129 As well, the effect of fat on proximal gastri c tone appears to be independe nt of its energy content and osmolality,129 and be mediated by release of CCK from the upper intestine.138;203 In the dog,8 the effect of intraduodenal nutrients on accommodation is region-specific. Specifically, fat infused into the proximal intestine induces accommodation, whereas isocaloric and isosmotic carbohydrate has little effect. On the other hand, infusion of the same carbohydrate solution, but not fat, into the distal intestine reduces gastric tone. Therefore, as in dogs, the accommodation re sponse may be both nutrientand sitespecific in the horse. Finally, we should consider the fact that the test meals used in the Phase I series of studies differed in more than just thei r fat and carbohydrate co mposition. Any other factor related to differences in composition, such as the higher fiber content of the highfat meal, could also account for the di fference in the accommodation responses. Phase II. Sweet feed Meals: Corn Oil Versus Glucose To better determine the specific effect of fat and carbohydrate on the meal-induced relaxation of the proximal stomach, two test meals that only differed in their fat and carbohydrate contents were formulated. Test meals consisted of a 10% crude protein sweet feed meal (0.5 g/kg) supplemented isocal orically with either corn oil or glucose. Experiments were performed with and w ithout labeling of the test meal with 13C-octanoic

PAGE 69

53 acid. An additional labeled test meal, consisting of a sweet feed meal (0.5 g/kg) with no supplementation, was added as a control. Th e same base sweet f eed (Seminole Feed, Blue Ribbon 10), without enrichment, was used as the horses daily grain ration (0.25 g/kg/day). Unlabeled meals Duration of meal ingestion. Horses spent an average of 241 31 seconds (range, 180 to 376 seconds) and 267 35 seconds (range, 172 to 400 seconds) to finish the corn oil-enriched meal and the glucose-enriched meal, respectively. Duration of ingestion was not significantly different. Relaxation response. Basal bag volume did not diffe r significantly between diets. Similar to the pelleted meals, beginning of inge stion of either meal was associated with a rapid increase in bag volume (Fig. 3-4A,B). A peak volum e was observed soon after the end of ingestion, followed by a gradual decreas e in bag volume. This receptive relaxation response was more prolonged after the glucoseenriched meal than after the corn oilenriched meal, which was evidenced by signi ficantly (p<0.05) higher volumes at 12-16 min after ingestion of the former, compared to the latter (Fig. 3-5). Intragastric bag volume remained significantly (p<0.05) grea ter than baseline volume for the entire postprandial phase of the glucose-enriched meal experiments (90 minutes), whereas it returned to baseline volume at the end of the recording period in the oil-enriched meal experiments. Bag volume was significantly grea ter than baseline for all periods, except for the corn oil diet during the first postprandi al two minutes and the last four minutes of barostat recording.

PAGE 70

54 Figure 3-4. Mean volume trace (n=6) of the e ffect of ingestion of different sweet feed meals (0.5 g/kg) on baseline tone in the proximal stomach. A) Corn oilenriched sweet feed meal; B) Glucose-en riched sweet feed meal; C) Corn oilenriched sweet feed meal labeled with octanoic acid; D) Glucose-enriched sweet feed meal labeled with octanoic acid; E) control sw eet feed meal (no enrichment) labeled with octanoic acid. Be ginning of ingestion of either meal was associated with a rapid increase in bag volume, followed by a more gradual decrease in volume. Time 0 is st art of recording period, and beginning of ingestion was at 30 minutes. The box limits ingestion time. The relaxation response observed after inge stion of the sweet feed meals shares some characteristics with that observed afte r ingestion of the pelleted meals. First, ingestion of the sweet feed meals induced an initial reduction in tone of the proximal stomach, and the length of this relaxation ep isode was associated with time of active ingestion. Second, this receptive relaxation was four minutes longer after consumption of

PAGE 71

55 the glucose-enriched meal, compared to the co rn oil-enriched meal. The latter similarity reinforces the idea that dietary composition a ffects the magnitude of receptive relaxation in the horse, and that dietary carbohydrat es seem to prolong this response, when compared to dietary fat. The origin of th is time difference cannot be explained only by time of meal ingestion, since horses spent si milar times consuming any of the sweet feed meals. As discussed earlier, a combination of factors, including orosensory influences, may be implicated in the control of this receptive relaxation reflex. 0 200 400 600 800 1000 1200 102030405060708090Time after ingestion (min)Bag volume (mL) Glucose enrichment Corn oil enrichment* Figure 3-5. Changes in intragastric bag vol ume after ingestion of a 10% protein sweet feed meal (0.5 g/kg) enriched by either corn oil or glucose (n=6). Both meals were isocaloric (1.95 kcal/kg bwt) and isovolumetric (~ 400 ml). Data are expressed as mean bag volume SE M of 2-min blocks (n=6). Each postprandial volume represents values from which the baseline volume has been subtracted. Asterisks denote a significant difference between pairs of blocks at the same postprandial time (p<0.05). Few studies in the current literature a ddress the effect of dietary composition on proximal gastric relaxation, and differences in methodology make comparisons difficult. For example, the protocol of some studies invo lved the direct admini stration of test meals into the stomach or small intestine, t hus bypassing possible orosensory stimulation.8;51

PAGE 72

56 Another example is the use of methods othe r than the electronic barostat to measure meal-induced relaxation. One of these methods consists of stepwise inflation of an intragastric bag, and subsequent analysis of variations in intragastric pressure in relation to increasing intrabag volumes.129 This method does not provide a continuous trace showing changes in gastric tone over time, lik e that obtained with the barostat technique, and direct comparison with our study is not possible. Fina lly, consistency of the test meals (i.e., liquid versus solid), which varies among studies, may affect the outcome. For example, in one study217 in which subjects consumed a liquid carbohydrate meal (0 g fat) or the same meal supplemented with fat ( 28 g fat), addition of fat to the carbohydrate drink resulted in greater proxi mal gastric relaxation. Meals us ed in that study were liquid, and oropharyngeal stimulation wa s probably brief because they were rapidly swallowed. Therefore, the single relaxation observed af ter ingestion of these meals was probably accommodation, and not receptive relaxation. In contrast, the meals used in the present study were solid, and the longer period of ma stication and swallowing probably resulted in a prolonged time of oropharyngeal stimulatio n. Therefore, we cannot expect the same response when test meals of similar composition, but different consistency, are used. Contrary to the response to the pelleted meals, receptiv e relaxation in response to the sweet feed meals was followed imme diately by accommodation. The induction of accommodation in these studies was more likely originated in the intestine than in the stomach. Since the accommodation reflex is trig gered, in part, by stimulation of gastric mechanoreceptors, rather than chemoreceptors,123 arrival of the pelleted or the sweet feed meals of similar weight and volume into the stomach should result in a similar degree of mechanical stimulation and, therefore, similar onset of accommodation. However,

PAGE 73

57 accommodation after the sweet feed meals occurred immediately after receptive relaxation, whereas with the hi gh-fat pelleted meal it was observed long after the meal had entered the stomach. One possible explana tion for the absence of stomach-originated accommodation is that, in the horse, the cont ribution of the stomach in the relaxation response may not be as important as that of the oropharynx and the intestine. The idea that degree of accommodation depends on the site where the reflex originates has been suggested before. In particular, it has b een shown that the duodenum is a stronger triggering site for the accommodation reflex than the stomach in humans.218 Alternatively, larger meals may be needed in the horse to trigger a distinctive accommodation response arising from the stomach. The earlier onset of accommodation, which it was probably controlled by intestinal feedback, may be explained by the composition and physical characteristics of the sweet feed meals. Digestion and ab sorption of carbohydrates and fat are required to activate intestinal chemoreceptors involved in control of gastric emptying.60;171 The glucose contained in the glucose-enriched sweet feed meal was ready for rapid absorption by the intestine, whereas the car bohydrates of the high-carbohydrat e pelleted meal, which were in the form of starch, needed digestion before absorption. Similarly, fat of the corn oilenriched sweet feed meal was possibly more readily accessible for digestion because of the physical characteristics of the meal, i.e ., liquid fat was poured into sweet feed to prepare the fat-enriched sweet feed meal, whereas fat was homogenized and integrated into pellets in the high-fat pelleted meal Thus, the type of car bohydrate and the physical presentation of fat of the hi gh-carbohydrate and high-fat swee t feed meals, respectively, may have facilitated earlier digestion and absorption of these nutrients, causing earlier

PAGE 74

58 activation of intestinal chemoreceptors and, in turn, earlier onset of accommodation, compared to the pelleted meals. Thus, the results of this study suggest that carbohydrate seems to magnify the receptive relaxation, when compared to diet ary fat, whereas intestinal modulation of accommodation by fat and carbohydrate seems to be similar. These observations are contrary to the original h ypothesis, since, based on st udies performed in humans,129;213 it was expected that fat would induce greater relaxation of the proximal stomach than carbohydrate. Octanoic-acid labeled meals Based upon the findings that addition of 13C-octanoic acid to the corn oilor the glucose-enriched sweetfeed meals had no signi ficant effect on the relaxation responses, the outcome of these separate experiments was used to compare relaxation induced by the corn oiland glucose-enriched meals to that of the control meal. Duration of meal ingestion. Mean SEM duration for complete meal ingestion was 165 2 seconds (range, 160 to 172 sec onds) for the control meal, 280 48 seconds (range, 164 to 450 seconds) for the corn oilenriched meal, and 310 78 seconds (range, 164 to 672 seconds) for the glucose-enrich ed meal. A Friedmans two-way ANOVA, used for meal comparison because of inequality of variances, showed that time to finish the control meal was not significantly (p=0.0839) shorter than that of the other meals, and neither did it differ between enriched meals. Relaxation response. Baseline bag volume did not differ significantly among diets. Mean bag volume SEM of postprandial 2-minute blocks is presented in Table 32. Barostat bag volume increased significantly after ingestion of each meal (Fig. 3-4C-E), reached a peak and started to decline gradually after the end of ingestion. Bag volume

PAGE 75

59 remained above baseline for the entire experi ment (90 minutes) after ingestion of the corn oiland the glucose-enriched meals, whereas it only remained above baseline during the first 40 minutes after ingesti on of the control meal (Fig. 3-6). During the initial peak, mean bag volume after ingestion of the cont rol meal was significantly lower than after ingestion of the glucose-enriched meal (16-mi nute period) and the corn oil-enriched meal (6-minute period). For the rest of the record ing session, bag volume of the control group did not differ significantly from that followi ng the other meals except at 48-62 minutes, where it was significantly lower (p<0.05) than the glucose-enriched meal group. Finally, mean bag volume after ingestion of the glucos e-enriched meal was significantly (p<0.05) higher than after the corn oil-en riched meal only at 48-50 minutes. The receptive relaxation responses after the labeled meals (Fig. 3-6) were qualitatively similar to those of the unlabeled meals (Fig. 3-5). From the standpoint of statistical analysis, the recep tive relaxation after the gluc ose-enriched meal was not significantly longer than the corn oil-enriched one, possibly because of the effect of octanoic acid labeling. A tendency for a longe r time of meal ingestion may account for the greater relaxation of the enriched meals, compared with the control meal. Another possibility is that orosensory mechanisms ma y have come into play as a consequence of this enrichment. Supplementing the sweet feed meal with gl ucose or corn oil not only affected the receptive relaxation, but also the magnitude of accommodation. Meal supplementation resulted in a longer accommodation respons e that may have been caused by stronger feedback regulation from the intestine.

PAGE 76

60 Table 3-2. Postprandial varia tions in volume* of an intr agastric bag controlled by an electronic barostat. Min after feeding Control meal Control meal + corn oil Control meal + glucose 0-2 375 107 320 80 356 61 2-4 754 144 725 89 852 114 4-6 799 122 925 51 1001 124 6-8 667 145 917 59 990 97 8-10 546 112a 881 40b 965 93b 10-12 352 139a 711 34b 925 79b 12-14 281 136a 622 63b 853 144b 14-16 286 77a 595 96a,b 781 174b 16-18 434 92a 662 140a,b 839 176b 18-20 328 117a 627 145a,b 782 197b 20-22 438 146 557 138 735 209 22-24 302 110a 572 167a,b 729 187b 24-26 439 99 489 110 613 143 26-28 597 153 668 176 745 188 28-30 625 132 534 134 626 171 30-32 410 127 493 116 581 175 32-34 441 145 527 110 580 168 34-36 484 140 356 116 542 193 36-38 392 70 434 108 644 106 38-40 271 56 439 101 574 100 40-42 314 74 357 126 570 137 42-44 272 127 295 103 572 122 44-46 158 112 287 94 472 139 46-48 274 117 288 82 538 167 48-50 218 127a 278 100a 603 107b 50-52 219 130a 326 91a,b 639 130b 52-54 156 112a 265 109a,b 497 91b 54-56 138 117a 266 111a,b 513 124b 56-58 214 84 382 101 464 115 58-60 116 106 296 92 538 87 60-62 157 74 331 98 552 116 62-64 249 92 422 103 447 91 64-66 206 102 397 88 488 108 66-68 212 97 348 90 483 96 68-70 230 64 332 82 452 56 70-72 159 105 335 143 316 68 72-74 126 75 306 111 414 71 74-76 147 71 293 80 406 89 76-78 114 75a 313 162a,b 477 86b 78-80 73 80 392 142 360 77 80-82 118 83 419 159 394 130 82-84 171 93 333 125 325 89 84-86 131 90 317 98 431 62 86-88 123 93 367 144 319 102 88-90 119 78 290 105 327 114 *Volume is mean SEM number of milliliters (n=6). Each postprandial volume represents values from which the baseline volume has been subtracted. Time 0=onset of ingestion of each meal Periods associated completely or partially with active ingestion of a meal. a-bWithin a row, values with different superscript letters differ significantly (p<0.05).

PAGE 77

61 0 200 400 600 800 1000 1200 102030405060708090Time after ingestion (min)Mean bag volume (mL) Glucose-enriched meal Corn oil-enriched meal Control meal * * * * * * * Figure 3-6. Changes in intragastric bag volum e after ingestion of a control sweet feed meal (0.5 g/kg) with and without additi on of either corn oil or glucose. All meals were labeled with 13C-octanoic acid. Data are expressed as mean volume during 2-min blocks (n=6). E ach postprandial volume represents values from which the baseline volume has been subtracted. Bag volume was significantly higher (p <0.05) than baseline during the entire length of the experiment (90 minutes) after ingestion of either enriched meal, whereas it only remained significantly above baseline for 40 minutes after ingestion of the control meal. Symbols indicate signi ficant difference between diets within the same time period: (*) glucose-enriched meal versus control meal, () corn oil-enriched meal versus control meal and () glucose-enriched meal versus corn oil-enriched meal. SEM values are shown in Table 3-2. Conclusions Methodology Electronic barostat. An electronic barostat has been previously used in our lab to asses tone of the proximal stomach in the horse.113;114 The lack of signi ficant differences in baseline volumes among experiments suppor ts that this is a valid and reliable technique to study changes of proximal gastri c tone induced by meal ingestion in the horse.

PAGE 78

62 Simultaneous use of the barostat and 13C-octanoate breath test. The concurrent use of the 13C-octanoate breath test modified th e relaxation response measured by the electronic barostat. This effect was diet-speci fic, because it increased relaxation after the high-carbohydrate pelleted diet but not that after any of the other diets. One possible explanation of this interaction is that labeling with the breath test marker increased the fat content of this diet to a level that surpa ssed a fat threshold to induce relaxation. In contrast, the initial fat cont ent of the high-fat pelleted meal was possibly above this potential threshold, so that the extra fat c ontent of the breath test label had no further effect on relaxation. To confirm this theory two properties of the relaxation response in the equine stomach should be determined in further studies: first, the existence of a threshold and, second, the absence of dose-de pendency of fat-induced relaxation of the proximal stomach. Finally, the greater accommodation observed after ingestion of the sweet feed meals may have masked the possibl e influence of the breath test label on the magnitude of relaxation. In summary, the process of labe ling a test meal to carry out the 13C-octanoate breath test may influence rela xation of the proximal stomach induced by this meal. The probability that such interaction occurs may be high when test meals with originally low fat content are used. Relaxation Response Overall results. Ingestion of any test meal i nduced relaxation of the proximal stomach. However, the pattern of relaxation, i.e., number of relaxa tion episodes and their time of onset and duration, differed among diets. Receptive relaxation. Although the pattern varied am ong diets, every diet induced an initial receptive relaxation phase. That the length of this phase was associated with

PAGE 79

63 time of meal ingestion sugge sts that it was induced by activation of oropharyngeal receptors, and thus it corresponds to what could be termed receptive relaxation. Based on the results of the present study, dietary carbohydrate seems to be a more potent stimulus of receptive relaxation than fa t in the horse. This fact cannot be explained only by time of meal ingestion, because tim es of ingestion of the high-carbohydrate meals were not longer than those of the hi gh-fat meals. Thus, other factors besides mechanical stimulation of the oropharynx dur ing mastication and swallowing must be involved. One possible factor is the invol vement of nutrient-specific orosensory receptors. Although orosensory mechanisms seem to control some gastric functions, such as gastric emptying, in othe r species, nothing is known with regard to the meal-induced receptive relaxation of the proximal stomach. Th erefore, more studies are necessary to determine the existence and influence of oros ensory factors in th e horse, and in other species. Finally, enrichment of a sweet feed meal with corn o il or glucose increased the meal-induced receptive relaxation, compared to the control sweet feed meal, suggesting that orosensory stimulation seems to be important in the response to both fat and carbohydrate. Accommodation. A second component of meal-i nduced relaxation, which is consistent with accommodation, was observed wi th all meals, with the exception of the high-carbohydrate pellets. Time of onset of accommodation varied among meals. That induced by the sweet feed meals occurred immediat ely after receptive relaxation, whereas that induced by the high-fat pelleted meal was not seen until one hour after. Finally, barostat recording was

PAGE 80

64 limited to a postprandial period of 90 minutes and, therefore, we cannot discard the possibility that accommodation after the high-carbohydrate pelleted meal came about even later than that of the high-fat pelleted meal. Since all meals had similar weight and volume, it is probable that the intestine, a nd not the stomach, was the triggering site of this accommodation response. Therefore, th e intestine may be more important in controlling accommodation than the stomach in the horse. The importance of fat versus carbohydr ate on the magnitude of accommodation is inconsistent in this study. On the one hand, accommodation was observed with the highfat pelleted meal, but not with the high-carbohydrate meal. On the other hand, there was no difference in the magnitude of accommodati on between the corn oiland the glucoseenriched sweet feed meals. A higher fat conten t of the high-fat pelleted meal may explain the induction of accommodation that was not s een after the high-carbohydrate pelleted meal. However, both pelleted meals differed in other components besides fat and carbohydrate and, therefore, we cannot conclude that such difference was caused merely by their difference in fat and carbohydrate cont ent. In contrast, both sweet feed meals differed only in that they were supplemented w ith either corn oil or glucose. Therefore, the difference in accommodation between meals could be attr ibuted specifically to the corn oil and the glucose. The magnitude of accommodation was similar between the glucoseand the corn oil-enriched sweet f eed meals. Thus, intestinal modulation of accommodation by fat and carbohydrate seems to be similar in the horse. These observations are opposite to the original hypothesi s of the study, sin ce it stated that fat would induce greater relaxation of th e proximal stomach than carbohydrate.

PAGE 81

65 Finally, although both enriched sweet feed meals induced similar accommodation, the process of enrichment prolonged the ti me of accommodation. The origin of this longer accommodation may have been stronger feedback regulation from the intestine, and shows that both glucose and corn oil can modulate this response.

PAGE 82

66 CHAPTER 4 RESULTS AND DISCUSSION GASTRIC EMPTYING The passage of ingested food from the stom ach to the small intestine is achieved by the coordinated action of the proximal stomach the antrum, the pylorus and the proximal duodenum. The emptying rate of a meal may be influenced by its volume, nutritional constituents, physical structure, temperature, caloric density osmolarity and the amount of acid produced by the stomach in response to this meal.124 A major control of gastric emptying is accomplished by nutrient-induced small intestinal f eedback regulation.119 That is, lipids, certain amino acids, sugars, and nutrients of high osmolality trigger sensory mechanisms from the intestine that inhibit gastric emptying. Results of studies done to date indicate that, of these nutrients, the most potent inhibi tor of gastric emptying seems to be fat.25;173;180 Little is known about the factors control ling gastric emptying of liquids and solids in the horse, and to what ex tent fat and carbohydrate particip ate in intestinal feedback inhibition. Studies in other sp ecies show that fat causes a greater inhibition of gastric emptying than carbohydrate.25;27;180 Based on that, the hypothesis of this part of the study stated that, in the horse, a high-fat meal w ould have slower gastric emptying than an isocaloric and isovolumetric high-carbohydrate m eal. To test this hypothesis, rates of gastric emptying were measured following inge stion of a high-fat or a high-carbohydrate meal, by use of the 13C-octanoic acid breath test.

PAGE 83

67 Effect of Dietary Composit ion on Gastric Emptying Phase I. Pelleted Meals: Fat Versus Carbohydrate Effect of diet on basal 13C expiratory output As mentioned, every horse pa rticipated in two experiments, which differed in the test meal offered to the horse: a high-fat pelleted meal or a hi gh-carbohydrate pelleted meal. Experiments were performed after at l east one week of adapta tion to the respective diet. Basal metabolic 13C production was measured by fo llowing the protocol for the 13Coctanoic acid breath test but wit hout the addition of the isotop e. The results of both diets were compared to determine the influence of diet on basal 13C output. Data from four tests (two within each test meal group) were not included in the observational analysis for suspected errors in breath samples analysis. In three of these cases, ingestion of the unlabeled test meal was associated with unexpected increases in 13CO2 concentration in breath samples (as it would be expected with labeled meal consumption). In the fourth test, exhaled 13CO2 declined markedly over time compared to the rest of the tests within the same meal group, i.e., the high-fat diet. Mean trace of postprandial 13CO2 expiration was similar between both diets over time In the current study, horses were maintained on ad libitum Bermuda hay and 5 g/kg/day of one of the pelleted diets. The results of the basa l breath tests showed that the effect of any of the pelleted meals on basal 13C production was similar. With regard to Bermuda hay, it is known that this type of hay interferes with the outcome of the 13Coctanoic acid breath test. Because Be rmuda hay is rich in natural 13C, its ingestion increases the equine en dogenous production of 13C. This endogenous source of 13C could interfere with the signal produced by the metabolism of the 13C-octanoic acid tracer and, therefore, it is recommended that Bermuda gr ass should be avoided as a maintenance diet

PAGE 84

68 to minimize fluctuations in basal 13C production. Alternatively, a higher dose of isotope should be used in the test meal.205 Since changing Bermuda hay as the maintenance diet was not an option in the present study, the standard dose of the is otope, 1 mg/kg, was increased to 1.5 mg/kg. Effect of diet on gastric emptying Because one breath test of the high-fat group was excluded from the data analysis for suspected error in breath samples analysis a two-sample t-test was used to compare parameters of gastric emptying between diets. Mean % dose recovery/h of the 13C tracer and modeled dose recovery curves after each te st meal are shown in Fig. 4-1. The shape of the curves was very similar between diets. However, the high-fat meal group tended to reach a higher and more exponential peak for isotope recovery, compared to the highcarbohydrate meal group. This peak was unde restimated by the modeling function and led to slight underestimation of tmax. Mean values for t1/2, tmax and GEC are summarized in Table 4-1. There was no significant difference between both meals for any of the gastric emptying indices. In the 13C-octanoic acid breath test technique, th e pattern of gastric emptying of a meal is described, in part, by the rising slope of the dose re covery curve and the time to reach the maximal recovery of 13C. The slope of the curve gives information about the onset and speed of gastric emptying, which may be affected by the presence and duration of a lag phase and the degree of feedback i nhibition. Thus, a steeper rising slope will indicate lesser inhibition of gastric emptying, and may, addi tionally, indicate a shorter lag period. Finally, the peak of 13C recovery also describes the magnitude of gastric inhibition and, therefore, a higher peak indicates that ga stric emptying of a meal can achieve a faster rate over time.59;117

PAGE 85

69 Time (h) 01234567PDR/h 0 1 2 3 4 5 6 7 8 9 High-fat pelleted diet High-CHO pelleted diet Figure 4-1. Mean percentage dose recove ry (PDR/h) SEM and modeled curve of 13C in breath following ingestion of a highfat pelleted meal (n=5) or a highcarbohydrate (CHO) pelleted meal (n=6). Gastric emptying parameters did not differ significantly between diets. Table 4-1. Comparison of gastric em ptying parameters determined by the 13C-octanoate breath test. Within a phase study, no significant differences were found among diets. See Materials & Methods for description of composition of the diets. t1/2 (h) tmax (h) GEC Test meal Mean Range Mean Range Mean Range Phase I studies High-fat pelleted diet* 3.06 0.37 1.963.59 1.92 0.19 1.51-2. 63 2.24 0.39 1.29-3.42 High-CHO pelleted diet 3.55 0.82 1.65-7. 16 2.07 0.42 1.08-3.82 2.17 0.35 0.88-3.20 Phase II studies Control diet 2.30 0.15 1.84-2.74 1. 52 0.08 1.38-1.87 2.81 0.27 2.28-3.98 Corn oil diet 2.56 0.23 1.81-3.49 1.69 0.08 1.46-1.95 2.74 0.25 1.89-3.65 (without intragastric bag) 3.00 0.22 2.26-3.84 1.96 0. 12 1.50-2.28 2.38 0.14 1.92-2.97 Glucose diet 2.59 0.31 1.93-3.75 1.89 0. 22 1.34-2.69 2.57 0.27 1.56-3.47 (without intragastric bag) 3.08 0.41 1.86-4.67 2.19 0.20 1. 59-2.71 2.52 0.37 1.42-3.81 Data are reported in mean SEM; t1/2=half-dose recovery time; tmax=time to peak 13CO2 concentration; GEC=gastric emptying coefficient; CHO=carbohydrate; n=6, except for *n=5.

PAGE 86

70 Few studies have been conducted to eval uate the mechanisms controlling gastric emptying in the horse. A limitation of some of these studies is the use of the phenol red technique,12;199 which is known to be less accurate to assess gastric emptying than scintigraphy or the 13C-octanoic acid breath test. Additiona lly, some studies include a low sample size (3-4 horses),58;234 along with high interi ndividual variation.197 Based on this limited literature, both diet ary fat and carbohydrate appear to have the ability to modulate the rate of gastric emptying in the hors e. First, two studies showed that addition of fat, i.e., soybean234 or corn oil,58 to a concentrate meal resulted in delay of gastric emptying. This effect was not aff ected by previous adaptation to dietary fat,58 and seemed to be independent of meal volume and dose of fat.234 The increase in energy or viscosity that resulted from adding oil to th e concentrate meal might be responsible for this slowing effect. Anothe r possibility is that fat de layed emptying by activating intestinal receptors sp ecific for this nutrient. With regard to carbohydrates, one study showed that addition of dextro se to water emptied more slow ly than the same volume of water in three ponies.197 The authors of this study s uggested that caloric density, osmolality of the solution or activation of nut rient-specific receptors within the intestine might have been involved in this effect.80;106;148 In contrast to these previous studies, the high-fat pelleted diet of the present study had the same energy and protein content, volume and physical characteristics (size, consistency) as the high-carbohydrate pelleted diet. Should any of these factors be responsible for the rate of ga stric emptying of a meal, it woul d presumably be similar for both of the test diets, and would explain the lack of significant differences in emptying

PAGE 87

71 rate between the test meals. Another possibili ty is that the extent of nutrient-specific feedback inhibition of fat is similar to that of carbohydrate in the horse. Volume is thought to be a major factor c ontrolling gastric empt ying of a meal in humans,80;81 dogs105 and monkeys.148 Its effect is probably mediated by activation of gastric and intestinal receptors that are sensitive to mechanical distention.190 Whether volume affects solid emptying appears to be de pendent on the particle size of the meal. That is, an increase in meal volume accelerat es gastric emptying only when mechanical breakdown of solids into smalle r particles is not needed.31;105 In contrast to the species that are routinely used in gastric emptyi ng studies, the horse is an herbivore whose natural nutrition depends on a constant intake of high-fiber, low-energy food. Since the horse has a relatively small gast ric capacity, control of intake rate and gastric load may be more important than control of nutrient delivery rate.234 In other words, volume may be more important in controlling gastric emptyi ng than dietary composition. Nonetheless, this factor could not be ev aluated in the present study because test meals had similar volumes, which could be considered as small for an adult horse. The effect of energy content of a meal on gastric emptying rate is uncertain. On the one hand, studies in humans17 and monkeys148 indicated that gluc ose solutions emptied from the stomach at a constant caloric rate, i ndependent of the initial concentration. This observation was not limited to glucose, since isocaloric, isovolumetr ic liquids containing fat, protein or carbohydrate emptied from the stomach at an overall constant rate in man,81 in monkeys126 and in pigs.226 Thus, it was suggested that emptying rates of isocaloric meals were similar because of comparable effects of nutrients on small intestinal receptors.17;81 On the other hand, other studi es in humans showed that

PAGE 88

72 increasing volumes of a glucose drink80 or a mixed solid/liquid meal31 were associated with increased rate of energy delivery to th e duodenum, at least init ially. Finally, Gregory et al.64 showed that, in pigs, carbohydrates, bu t not fat, emptied from the stomach following a constant caloric rate. Nevertheless, since test meals of the present study had the same energy content, its effect on gastric emptying could not be determined in the horse. The high-fat and high-carbohydrate test meal s of this Phase I study had similar initial consistency. However, with ingestion, pellets were mixed w ithin the mouth and in the stomach to produce a viscous solution. The viscosity of such solutions may vary with the amount of salivary and gast ric secretions, and the rela tive presence of specific ingredients, such as fat and fiber.65 Thus, since the high-fat meal was richer in fiber and fat, it possibly became more viscous within th e stomach. While the effect of fat on gastric emptying does not seem to be mediated by its viscosity,35 fiber can influence gastric emptying rate by its capacity to retain water and to increase the viscosity of gastric contents. In other species, this effect depe nds on the type of fiber. For example, in pigs65 and humans,176 only soluble fibers appear to delay gastric emptying, whereas in the rat, insoluble fiber delays, and sol uble fiber accelerates, emptying.20 Both soluble and insoluble fiber was proportionally greater in the high-fat pellets and, thus, may have affected gastric emptying of this meal in comb ination with other factors, such as specific nutrient source. We do not know to what extent fiber affects rate of gastric emptying in the horse. Yet, it would be reas onable to presume that fiber per se does not inhibit gastric emptying. Fiber constitutes the base of herbivor es diet and, therefor e, a potential delay of

PAGE 89

73 emptying by fiber would result in earlier satiatio n and, in turn, a slower rate of an already low energy intake. Fat slows gastric emptying by a mechanism th at involves stimulation of intestinal receptors that are sensitive to its digestion products.135;163 This effect is determined primarily by its chemical composition more than its physical char acteristics (lower density and higher viscosity),35 dose of fat and degree of hom ogenization. In particular, the slowing effect of fat is increased with increasing amounts of fat and decreased homogenization.34 One reason that no difference in emptying rate was found between the high (8%) and low (3%) fat pellets is th at feedback mechanisms provoked by fat and carbohydrate may not be very discriminatory in the horse. An alternative reason is that previous adaptation of the horses to this di et may have eliminated the greater slowing effect of fat versus carbohydrate. In hu mans, high-fat feeding induces adaptation, ultimately reversing the slowing effect of a fatty meal and enhancing gastric emptying.39;53 Finally, since it is the lipolytic products that trigger intestinal receptors to induce slowing of gastric emptying, a limited capacity of the horse to digest fat could explain the inability of the high-fat meal to have a greater suppressive effect in these Phase I studies. Studies in humans and animals with pancreatic insufficiency indicate that diminished duodenal lipase speeds gastric emptying of oil.118 Yet, one of the most abundant components of the equine pancreatic juice appears to be lipase (Dr. Jean Morisset, personal communication), and it has been shown that the ability of the horse to digest fat increases with increasing amounts of dietary fat.139 Therefore, it is unlikely that a limitation in fat digestion explains the lack of difference in emptying rate of the two pelleted diets.

PAGE 90

74 Similar to fat, inhibition by carbohydra te mostly depends on chemospecific intestinal feedback,106;152 and, at least for glucose, the magnitude of feedback correlates with concentration.17;148 Carbohydrate solutions may also control emptying indirectly by stimulating osmotic receptor s within the duodenal mucosa.106 Another influential factor is adaptation to dietary carbohydrate, which leads to increased gastric emptying rate of carbohydrate in humans.76 Like fat, this phenomenon ma y have influenced gastric emptying of the high-carbohydrate diet in the present study. Most studies have evaluated the effect of fat by adding a lipid component to a control meal. In few studies, such as the present one, the caloric increase associated with fat supplementation was compensated for to can cel the possible effect of adding energy on emptying rate. By use of the 13C-octanoic acid breath test Robertson and Mathers180 showed in humans that a high-fat solid m eal emptied slower than a high-carbohydrate meal of identical energy and volume. Simila r results were observed by scintigraphy with liquid meals, but rates of gastric emptying for a high-fat and a hi gh-carbohydrate soups became similar when meals were infused directly into the stomach.27 Therefore, orosensory signals may also be contributory to control of emptyi ng rate. It has been suggested in the previous ch apter that, in the horse, nutrients may control the mealinduced relaxation of the proximal stomach through orosensory mechanisms, and this may also apply to gastric emptying. Finally, we should consider the fact that th e test meals used in this Phase I of the present study differed in more than fat and carbohydrate content, a nd that a combination of several factors regulati ng gastric emptying may be re sponsible for the present observations. In addition, it is difficult to determine whether labe ling with octanoic acid

PAGE 91

75 may have influenced gastric emptying of the test meals, as it happened with the mealinduced relaxation of the proximal stomach. Gastric emptying of a solid meal is dependent on the integrated function of the proximal stomach, the antrum and the pyloric sphincter and, therefore, changes in proxima l gastric function may influence overall rate of emptying. Nonetheless, the 13C-octanoic acid breath test tec hnique has been previously used in humans180 and the horse234 to evaluate the effect of fat and carbohydrate on gastric emptying, and labeling with 13C-octanoic acid did not ma sk the dampening effect of fat over carbohydrate. Phase II. Sweet feed Meals: Corn Oil Versus Glucose As mentioned, every horse pa rticipated in two experiments, which differed in the test meal offered to the horse: a sweet feed meal supplemented with either corn oil or an isocaloric amount of glucose. In contrast to the Phase I breath tests, these experiments were performed without, as well as with, simu ltaneous use of the intragastric barostat bag. Breath tests without presence of an intragastric barostat bag Mean % dose recovery/h of the 13C tracer and modeled dose recovery curves after each test meal are shown in Fig. 4-2. Shape of the curves was very similar between diets, except that the initial slope fo r the corn oil-enriched meal was slightly steeper and the maximal rate of gastric emptying was faster compared to the glucose-enriched meal. Similar to the high-fat pelleted meal, the peak of emptying of the corn oil-enriched meal was underestimated by the modeled curve, but wa s not significantly different from that of the glucose-enriched meal. Mean values for gastric emptying parameters are summarized in Table 4-1. None of the parameters were significantly different between the two diets.

PAGE 92

76 Like the pelleted meals, these results ar e opposite to the initia l hypothesis that the gastric emptying of a sweet feed meal supplem ented with corn oil w ould be slower than that supplemented with an isocaloric amount of glucose. However, both Phases of studies were consistent in that th ere was a tendency for the high-carbohydrate meals to have initially slower emptying rates than the high -fat meals, which sugge sts that carbohydrate may cause more profound feedback inhibiti on than fat on early phases of gastric emptying. A slower early phase of emptying of the glucose-enriched meal may have been compensated by faster emptying rate s at later stages, to yield a t1/2 value similar to that of the corn oil-enriched meal. It is possible that the breath test parameters do not have sufficient sensitivity to reflect differences in the early phase of gastric emptying of solid food. Finally, the lack of significant diffe rence between meals may also be due to the small number of animals included in th e study, combined with a high intersubject variability. The apparent tendency for the glucose-enrich ed meal to empty more slowly may be due to a difference in the nutrient compositi on of the meals, but also by their physical properties. Specifically, the relatively high volume of glucose (113 ml/500 kg bwt) added to the sweet feed meal resulted in the formation of a small liquid phase. The liquid phase of a meal is known to empty faster than the solid phase119 and, therefore, earlier passage of glucose to the duodenum may have led to earlier onset of fee dback inhibition. Finally, the sweet feed meal s of this second Phase of studies differed only in the glucose or corn oil content, in contrast to the pelleted meals of the Phase I studies. Therefore, any component that was simila r between the meals of Phase II may be responsible for the lack of differe nce in gastric emptying parameters.

PAGE 93

77 Time (h) 01234567PDR/h 0 1 2 3 4 5 6 7 8 9 Corn oil-enriched meal Glucose-enriched meal Figure 4-2. Mean percentage dose recove ry (PDR/h) SEM and modeled curve of 13C in breath following ingestion of a 10% crude protein sweet feed meal (Seminole Feed, Blue Ribbon 10) enriched with corn oil or glucose (n=6). Gastric emptying parameters were not signifi cantly different between diets. Breath tests with presence of an intragastric barostat bag A new series of breath tests were performe d with the sweet feed meals used in the previous section, but this time, with an in tragastric bag in place to measure proximal gastric tone simultaneously. An additional te st meal, consisting of the same amount of sweet feed without supplementa tion, was used as control. Mean % dose recovery/h of the 13C tracer and modeled dose recovery curves after each test meal are shown in Fig. 4-3. The shape of the curves was similar among diets, although emptying of the glucose-enriched meal appeared to be slower and to reach a

PAGE 94

78 lower maximal rate, compared to the other diets. Fitness of the modeled curves was higher for the corn oil-enriched di et than for the other two diets. Mean values for gastric emptying parameters are summarized in Table 4-1. Data of tmax under the control diet were not norm ally distributed and, therefore, tmax values were compared using a Friedmans two-way ANOVA. None of the parameters were significantly different among the test meals. The emptying curves of the enriched m eals are similar to those shown in the previous section. Surprisingly, addition of corn oil to the sweet feed meal did not modify the shape of gastric emptying of the control m eal. Furthermore, addition of corn oil or an isocaloric amount of glucose to the sweet feed meal resulte d in similar overall gastric emptying, compared to the sweet feed meal without enrichment. However, the previously observed tendency for the high-carbohydrate meal to empty more slowly was also observed with this new series of experiments. These results are opposite to other studies in th e horse. By use of the 13C-octanoic breath test, Wyse et al.234 reported that addition of soybean oil to a small meal of oats and bran caused a delay in gastric emptying. It is difficult to explain this discrepancy, but one possibility is that the differe nce in the relative energetic c ontribution of fat to the meal varied between studies. That is, in the study by Wyse et al., addition of the fat component to the original meal resulted in an incremen t of ~60% of energy, whereas addition of corn oil to the control meal of the present study increased the energy content in ~30%. Therefore, a greater increase in energy might be necessary to show a significant effect of fat on the breath test parameters. Another possi bility is that the small sample size (6 horses) used in this study, combined with poor fitness of the curve observed in the control

PAGE 95

79 group and, especially the glucose group, ma y be responsible, in part, for a lack of significant difference between meals. Fina lly, the physiology of gastric emptying in ponies may differ from that of horses. 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 01234567Time (h)PDR/hr Control meal Corn oil enriched meal Glucose enriched meal Figure 4-3. Modeled mean % dose recovery curves of the 13C label in the breath of 6 horses after ingestion of a sweet feed meal (contro l) or the same meal enriched with corn oil or glucose. Experiments were done without the presence of a barostat bag. Note: in c ontrast to the previous figures, symbols on the lines are descriptive points of the modeled curves, not the original mean % dose recovery values. Another study, presented as an abstract, repor ted that addition of corn oil to a sweet feed meal delayed gastric empt ying of the sweet feed meal.58 The size of the meal was four times larger, but the relative energetic c ontribution of fat was apparently similar to that of the present study. Ot her factors may be responsible for such inconsistency between studies, and more details of that study are needed for further comparison.

PAGE 96

80 Effect of the Barostat Bag on Gastric Emptying To assess whether the presence of the ba rostat bag within the proximal stomach would alter gastric emptying, brea th test results of the corn oiland the gl ucose-enriched sweet feed meals, measured w ith and without the pr esence of an intragastric bag, were compared. No significant differences were found for t1/2, tmax and GEC (Table 4-1). A potential disadvantage of th e barostat technique is its intrusiveness, since it requires positioning of a bag within the proximal stomach. Therefore, it has been suggested that this may affect intragastric distribution and emptying rate of a liquid meal.42;181 In contrast, other studies have failed to show an effect of the barostat bag on emptying of a liquid42;153 or a solid/liquid meal.144 In the present st udy, the presence of the barostat bag within the proximal stomach did not seem to affect gastric emptying of the test meals. Relation Between Proximal Gastric Relaxation and Gastric Emptying The results of this study have shown that supplementation of the sweet feed meal with glucose induced a more prolonged r eceptive relaxation than supplementation with corn oil, or no supplementation at all. In addition, meal supplementation with glucose showed a tendency to delay emptying at an early phase, compared to the other meals. As mentioned in the previ ous chapter, the more poten t effect of carbohydrate on the receptive relaxation response may be expl ained by orosensory stimulation of carbohydrate, which would lastly result in the modulation of proximal gastric tone. A similar orosensory mechanism may explain th e apparent tendency of carbohydrates to slow gastric emptying at initia l stages. The idea of orosenso ry factors modulating gastric emptying is not new, since Cecil et al.27 showed that the effect of oral feeding on gastric emptying differed from that of intragastric feeding.

PAGE 97

81 Conclusions For all studies, the high-fat meals and the high-carbohydrate meals emptied from the stomach at similar rates. These results are opposite to the prev ailing notion that fat has a more potent effect on regulation of gastric emptying than carbohydrates.27;64;180;234 Yet, the present study is differe nt from most studies in that test meals with identical caloric content were used. Therefore, the resu lts of this study suppor t the idea that meals with similar caloric content have simila r emptying rates, re gardless of nutrient composition.81;126;226 Although dietary adaptation to fat and carbohydrate appe ars to eliminate the slowing effect of these nutrie nts on gastric emptying in humans 39;40;76, this may not explain the results of this study, since horses were not previously adapted to supplementation with corn oil or glucose for the Phase II studies. Another unexpected finding was that supplem enting a sweet feed meal with either corn oil or glucose, which resulted in a 30% in crease in energy content, did not modify its rate of gastric emptying. It is difficult to explain why supplementation had no additive effect on inhibition of gastric emptying. Both the corn oiland glucose-enriched sweet feed meals had a high caloric density, which is far from the horses natural diet. Thus, it is possible that the mechanisms involved in regulation of gastric em ptying in the horse are a better reflection of the horses natural nutrition. In other words, the mechanisms controlling gastric emptying may be aimed to wards the continuous ingestion of a lowenergy diet, whereas they may be limited when the equine stomach is challenged with a high-energy meal. Therefore, volume and, to a much lesser extent, energy may control rate of gastric emptying.

PAGE 98

82 Although breath parameters of gastri c emptying obtained with the highcarbohydrate meals did not differ significantly from those of the other meals, there was a consistent tendency for the former to empt y slower at the ini tial phase of gastric emptying. The lack of significant differences between the high-carbohydrate meal and the others may have resulted from a low sample size, or poor fitness of the modeled curves for the high-carbohydrate diet, due to the higher variability ob served within this meal group. Alternatively, the paramete rs described by Ghoos et al.59 may not be sensitive enough to detect significant differences betw een meals in the different phases of the gastric emptying process. It would be very in teresting to further explore the possibility that carbohydrate may be more suppressive of gastric emptying than fat. Carbohydrates may induce more feedback inhibition by stimulation of orosensory signals, as it has been suggested for modulation of receptive relaxation in the previous chap ter. Alternatively, carbohydrates emptying into the duodenum ma y stimulate specific chemoreceptors or osmoreceptors that result in rapid onset of intestinal feedback inhibition of emptying. Finally, presence of an intragastric bag within the proximal stomach to measure tone by an electronic barostat did not influence gastric emptying rates of the test meals. This supports findings from studies42;144;153 in other species that show no effect of a barostat bag on gastric emptying.

PAGE 99

83 CHAPTER 5 RESULTS AND DISCUSSION pH OF GASTRIC CONTENTS Ingestion of food is associated with decreased gastric acidity in the horse,154 mainly due to the buffering effect of bicar bonate-rich salivary secretions.4 Few studies have measured meal-induced changes in gastric pH in the horse, but the e ffect of meal on pH may vary with meal composition and acidity, degree of stimulation of salivary and acid secretions, degree and products of meal ferm entation within the stomach, and the gastric emptying rate of the meal.4;6;110 Ulceration of the squamous portion of the e quine gastric mucosa is highly prevalent in adult performance horses,68;85;155;162;223 and excessive exposure to acid is believed to be the main cause.114 Therefore, a potential strategy to reduce the risk of ulcer development would consist of reducing the acidity of gastric contents. Dietary manipulation could be one way to approach such strategy, through se lection and use of equine rations with a large buffering power. It is thus of interest to know to what exte nt different dietary formulations affect the pH of gastric contents in the horse.156 This part of the study was aimed to determine to what degree a diet ri ch in fat and low in carbohydrate might affect intragastric pH, compared to a more tr aditional carbohydrate-based diet. Therefore, variations in intragastric pH in response to a high-fat and a high-carbohydrate meal were measured by use of a self-referencing electrode positioned in the most ventral portion of the stomach. This site was chosen because it corresponds to the most acidic site of the equine stomach. That is, layering of inge sted food creates a pr oximal-to-distal pH gradient within the equine stomach that resu lts in the formation of a high-pH upper part,

PAGE 100

84 rich in saliva and far from the glandular mucosa, and a low-pH bottom part, where acidic secretions are produced.11;137 Effect of Dietary Compos ition on Intragastric pH Phase I. Pelleted Meals: Fat Versus Carbohydrate Changes in intragastric pH were measured before, during and af ter ingestion of a high-fat pelleted meal, a high-carbohydrate pell eted meal, or any of these meals enriched with 13C-octanoic acid. Addition of the octanoic acid to the test meals did not have any significant effect on intragastric pH at any time point. Theref ore, results of experiments with and without octanoic acid were comb ined for each diet in the analyses. Mean baseline pH did not differ signifi cantly between diets. Mean pH increased significantly (p<0.05) from base line 15 minutes after ingestion of either meal (Fig. 5-1). Thereafter, it remained significantly highe r (p<0.05) than baseline for 60 minutes following the high-fat meal, and 80 minutes following the high-carbohydrate meal. Mean pH after the high-carbohydrate meal was significantly highe r (p<0.05) than after the high-fat meal during 75 minutes post-inges tion. Therefore, ingestion of the highcarbohydrate meal produced a significantly higher and more sustained increase in pH of the ventral part of the stomach, in comparison to the high-fat meal. It is unlikely that the larger buffering e ffect of the high-carbohydrate meal was the result of higher secretion of saliva, since the principal stimulus of salivary secretion is mastication,4 and horses did not spend more time eating the high-carbohydrate meal. In fact, time of ingestion of the high-fat meal was signi ficantly (p<0.05) longer. Alternatively, it is possible that other fact ors besides mastication may modulate salivary secretion in the horse, such as orosensory st imulation in response to different nutrients. Although there is no evidence in the literatur e that dietary composition or gustatory

PAGE 101

85 stimuli modulate the rate of salivary flow,71 an effect of nutrients cannot be ruled out. Finally, an alternative explanation is that the ingredie nts of the high-carbohydrate meal had more inherent buffering capacity than those of the high-fat meal. 0 1 2 3 4 5B153045607590105120Min after meal ingestionpH High-CHO pelleted meal High-fat pelleted meal* * * * * * * Figure 5-1. Changes in intragastric pH af ter ingestion of the high-fat or the highcarbohydrate (CHO) pelleted meal. Data ar e expressed as mean pH of 5-min blocks (n=6). The first block represen ts baseline pH (B), calculated as the mean pH of the 30-min peri od prior to meal ingestion. Postprandial mean pH increased significantly (p<0.05) from ba seline after 15 minutes, and remained significantly higher for 60 minutes af ter ingestion of the high-fat meal (continuous arrow) and 80 minutes af ter ingestion of the high-carbohydrate meal (dashed arrow). Asterisks indica te significant difference (p<0.05) between diets within the same time interval. The results of the present study are opposite to the current belief that diets rich in highly digestible carbohydrates, such as the one of this study, are associated with lowering of gastric pH. This effect has been attributed to the lactate and volatile fatty acids (VFA) that result from the fermentati on of carbohydrates by resident bacteria of the stomach.6 Two studies have measured VFA and l actate levels, with concomitant changes in gastric pH, in response to a hay-grain meal and a hay meal. Specifically, Nadeau et al.156 compared an alfalfa-grain meal versus a bromegrass hay meal, whereas Argenzio et

PAGE 102

86 al.6 compared a pelleted hay-grain meal vers us a pelleted high-cellulose, low-protein meal. In both studies, ingestion of the hay-grain meal was asso ciated with higher levels of VFA and lactate, but higher gastric pH, compared to the only-hay meal. The authors of those studies attributed the higher pH of th e hay-grain meal to the buffering capacity of the hay component, but did not consider any possible contribution of the grain to the change in pH. Hence, it would be of interest to measure the specific effect of grain on gastric pH with concomitant measurement of VFA and lactate produc tion. Since products of fermentation were not measured in the pr esent study, it is unknow n to what level the high-carbohydrate meal was fermented within the stomach, and whether this contributed to the changes observed after this meal. Fina lly, since test meals were relatively small (0.5 g/kg), the time that the high-carbohydrate meal remained in the stomach may have not been sufficient for the accumulation of a significant amount of fermentation products. Therefore, ingestion of larger meals may not have the same buffering effect as that observed with the small meal of this study. The lower pH after the high-fat meal may also be explained by greater stimulation of acid secretion by this meal. Volume and nut rient components of a meal are two main factors regulating acid secretion in other species.122 With regard to the horse, Sandin et al.185 suggested that distention by food entering th e stomach is a principal factor causing gastrin release. Based on this idea, both te st meals would induce similar acid secretion since they had similar volume. However, nutrient composition may also affect acid secretion in the horse. In othe r species, intragastric and in traduodenal protein is the most potent stimulus of acid secre tion, whereas intraduodenal li pid and carbohydrate have an inhibitory effect.95;110;142 The magnitude of inhibition by fat and carbohydrate appears to

PAGE 103

87 be similar in humans, since infusion of is osmolar, isocaloric lipid and carbohydrate solutions into the small intestine induced similar gastric acid output.160 In addition to the effect of intestinal nutrients, there is evidence that both fat and carbohydrate influence acid secretion during the circulatory phase of digestion. Specifically, intr avenous infusion of glucose or a lipid emulsion impairs acid secretion induced by intravenous amino acids in humans.92;220 Likewise, hyperglycemia induced by intravenous infusion of glucose impairs basal, mealand ga strin-induced acid secretion.91;91;93;116 It has been suggested that the inhibitory effect of hyperglycem ia is mediated by suppression of vagal cholinergic activity, the same mechanis m suggested for hyperglycemia-induced inhibition of gastric emptying.72;122 Regarding the horse, the effect of intestinal or circulating products of digested nutrient s on acid secretion is unknown, but may be different from other species. That is, in sulin-induced hypoglycemia appears to decrease acid secretion,184 whereas the opposite, i.e., stimulation of acid secretion, is observed in humans,7 rat 225 and dog.89 Nonetheless, the effect of fat and carbohydrate on intragastric pH was not measured in those studies. Finally, the test meals of the present study differed in more than the fat and carbohydrate density and, therefore, other ingredients, such as fiber, may be responsible for our results. However, it is unlikely that the high fiber content of the high-fat meal is responsible for the lower pH obs erved after ingestion of this meal, since some types of fiber, specially those rich in protein and calcium, seem to have a high buffering capacity.6 Moreover, the test meals of this study ha d similar calcium and protein contents. In conclusion, ingredients other than carbohydr ate and fat may account for the different buffering effect of the test meals, if bu ffering capacity was the final determinant.

PAGE 104

88 Phase II. Sweet Feed Meals: Corn Oil Versus Glucose Changes in intragastric pH were measured before, during and af ter ingestion of a sweet feed meal, or this meal enriched with either corn o il or an isocaloric amount of glucose. All meals were labeled with 13C-octanoic acid for simultaneous measurement of gastric emptying. Mean baseline pH was not different am ong test meals (Table 5-1). Mean pH increased significantly (p<0.05) from base line after 10 minutes of ingestion of the enriched sweet feed meals (Fig. 5-2), and remained significantly (p<0.05) elevated for 35 and 60 minutes with the glucose-enriched meal and the corn oil-enriched meal, respectively. When the cont rol (non-enriched) meal wa s ingested, pH increased significantly (p<0.05) after 25 minutes and re mained elevated for 45 minutes. Mean pH values dropped below baseline 85 minutes afte r ingestion of the glucose-enriched meal and remained significantly (p<0.05) low until the end of the recording period. Finally, when mean pH was compared among diets wi thin the same time period, pH after ingestion of the control meal remained si gnificantly (p<0.05) lower at 5-20 minutes compared to the glucose-enriched meal, and at 10-20 minutes, compared to the corn oilenriched meal. Therefore, ingestion of the corn oil-enriched meal produced a more sustained increase in pH, in comparison to the glucose-enriched and control meals. Thus, similar to the pelleted meals, inges tion of the enriched sweet feed meals was soon followed by a significant increase in in tragastric pH. However, this increase was delayed after ingestion of the control meal possibly because horses tended (p=0.0839) to spend less time ingesting this meal and, therefore, salivary production was lower compared to the enriched sweet feed meals.

PAGE 105

89 Table 5-1. Changes in intragastric pH after ingestion of a 0.5 g/kg sweet feed meal alone (control meal) or enriched with either corn oil or glucose. Period Control meal Control meal + corn oil Control meal + glucose Baseline* 1.78 0.10 1.95 0.25 2.37 0.37 Min. after feeding 0-5 1.58 0.08 1.83 0.23 1.78 0.20 5-10 1.65 0.05a 2.55 0.62a,b 2.77 0.56b 10-15 1.78 0.19a 3.06 0.56b 3.31 0.50b 15-20 1.99 0.26a 3.43 0.55b 3.56 0.53b 20-25 2.32 0.25a 3.79 0.51b 3.63 0.46b 25-30 2.98 0.31a 3.96 0.39b 3.53 0.39a 30-35 3.65 0.37 3.99 0.33 3.39 0.46 35-40 3.69 0.38 3.67 0.40 3.22 0.47 40-45 3.52 0.28 3.60 0.40 3.07 0.46 45-50 3.25 0.40 3.51 0.37 2.83 0.51 50-55 3.23 0.38 3.26 0.40 2.66 0.55 55-60 2.89 0.35 2.94 0.38 2.58 0.59 60-65 2.52 0.38 2.77 0.31 2.42 0.61 65-70 2.48 0.36 2.65 0.32 2.27 0.56 70-75 2.16 0.37 2.48 0.37 2.06 0.47 75-80 1.94 0.36 2.40 0.42 1.86 0.40 80-85 1.85 0.34 2.20 0.40 1.77 0.37 85-90 1.72 0.32 2.04 0.40 1.65 0.29 90-95 1.69 0.31 1.93 0.38 1.39 0.13 95-100 1.61 0.28 1.91 0.44 1.41 0.21 100-105 1.57 0.24 1.85 0.47 1.44 0.27 105-110 1.60 0.21 1.96 0.54 1.33 0.14 110-115 1.62 0.16 2.03 0.61 1.29 0.09 115-120 1.61 0.18 2.07 0.75 1.30 0.08 Values are expressed as mean SEM for 5-min periods. *Baseline is mean pH of 30 minutes prior to inges tion. Time 0 = Onset of ingestion of each meal. a,bWithin a row, values with different superscript letters differ significantly (p<0.05). In contrast to results in the Phase I studi es, the glucose-enriched sweet feed meal was associated with a shorter duration of pH increase than the corn oil-enriched meal (Fig. 5-2). Furthermore, mean pH dropped be low baseline only after ingestion of the glucose-enriched meal. The difference be tween the effects of both high-carbohydrate diets may be explained by a different degr ee of gastric bacterial fermentation. Since glucose is readily fermentable, produc tion of lactic acid and VFA by resident

PAGE 106

90 microorganisms might have started soon afte r ingestion. Gradual accumulation of these fermentation products within the stomach would result in a gradual decline in pH after ingestion of the glucose-enriched sweet f eed meal. In contrast, the high-carbohydrate pelleted meal was rich in starch, which is highly fermentable, but its fermentation requires longer time than glucose. Since both meals were small, the time they were retained in the stomach may have been long enough for significant fermentation of glucose, but not for that of starch. 0 1 2 3 4 5B153045607590105120Min after meal ingestionpH Corn oil-enriched meal Glucose-enriched meal Control meal Figure 5-2. Changes in intragastric pH after i ngestion of a control sw eet feed meal, or the same meal enriched with either corn oil or glucose. Data are expressed as mean pH of 5-min blocks (n=6). The fi rst block represents baseline pH (B), calculated as the mean pH of the 30min period prior to meal ingestion. Arrows with symbols delimit time period of significant (p<0.05) pH elevation above baseline for corn oil-enriched meal (), glucose-enriched meal (), and control (non-enriched) meal (). *Significant difference between the control diet and the glucoseand corn oi l enriched diets, respectively. * *

PAGE 107

91 In addition, the acidity of glucose may ha ve influenced the changes in pH observed with the glucose-enriched sweet feed meal, since the glucose solution that was added to the sweet feed had a pH of 3.6. In contrast, corn oil would not influence pH, since it does not mix with the aqueous phase of gastric contents. Finally, although duration of significant pH increase above baseline differed among the three test meals, only the control meal di ffered from the enriched meals when pH was compared within the same time period (Fig. 5-2). That is, mean pH was lower after the control meal during the early postprandial phase (25 minutes) and, as mentioned, this may be related to a shorter period of masti cation. A low sample size (6 horses) and the existence of high variability may account for the lack of significant difference between the enriched sweet feed meals. Conclusions Ingestion of any of the test meals was a ssociated with an in crease in the pH of contents in the most distal portion of the st omach, but the intensity and duration of this buffering effect varied among meals. Surprisingly, ingestion of the high-car bohydrate pelleted meal, which was rich in starch (31%), had the greatest buffering effect, and was the only one of all the test meals that increased mean pH above 4. This observation is opposite to the generalized assumption that diets rich in starch tend to increase gastric acidity. It is difficult to determine which factor of this meal was responsible for inducing a pH response that remained 75 minutes higher than that of th e high-fat pelleted meal. The small volume of the meal and its relatively fast emptying from the stomach, compared to the larger meals used as daily rations, minimized the chance for synthesis and accumulation of lactic acid and VFA in sufficient amounts to decrease pH. In addition, an inherent buffering capacity

PAGE 108

92 of the high-carbohydrate meal or a greater capacity to stimulate salivary secretion in response to orosensory signaling may account fo r the higher pH after this meal, compared to the high-fat pelleted meal. Alternativel y, the high-fat meal may stimulate acid secretion to a gr eater degree. In contrast to the high-c arbohydrate pelleted meal, the gl ucose-enriched sweet feed meal, which was also rich in highly fe rmentable carbohydrates, produced the lowest buffering effect among the sweet feed meals. The type of carbohydrates present in these meals may explain this disparity between the two Phases of the study. Both glucose and starch are highly fermentable carbohydrates, but glucose is readily available for synthesis of lactic acid and VFA, whereas starch has to be previously degrad ed to glucose. Thus, although both meals were retained in the stom ach for a relatively s hort period of time, it may have been long enough for significant fermentation of glucose, but not for that of starch. Another factor that may have cont ributed to the low pH observed after the glucose-enriched meal is the low pH (3.6) of the glucose solution us ed to supplement the sweet feed meal. Finally, the results of this study support th e idea that feeding results in buffering of the highly acidic contents of horses.154 With regard to the meals used in the present study, the small high-carbohydrate pelleted meal (0.5 g/kg) appeared to be the most effective in inducing low gastric acidity. Fu rther studies evaluating the buffering capacity of meals, but fed in larger volumes, need to be done.

PAGE 109

93 CHAPTER 6 RESULTS AND DISCUSSION CHOL ECYSTOKININ LIKE-ACTIVITY Cholecystokinin (CCK) is a major gastrointestinal hormone that is released from intestinal endocrine cells in response to food intake, and is known to regulate motility, pancreatic enzyme secretion, gastric emptying, and gastric acid secretion.98 However, it remains unknown whether CCK participates in the control of gastrointestinal functions in the horse. As yet, plasma CCK levels have not been measured in this species and, therefore, not hing is known about the ab ility of nutrients to stimulate CCK release from equine enteroendoc rine cells. In other species, the most potent stimulants of CCK secretion are di gested fat and protei n, whereas carbohydrate has a weak or no effect.99 In addition, it is well recognized that some of the effects of fat on gastrointestinal function are mediated, at le ast in part, by the release of CCK from the intestinal mucosa.21 Based on this, the hypothesis of this part of the study was that fat would be a greater stimulus fo r CCK release than car bohydrate in the horse. To test this hypothesis, plasma CCK-like activity was m easured by radioimmunoassay after ingestion of a high-fat and a high-carbohydrate pelleted meals of equal caloric density, volume and protein content. These meals corresponded to th e test meals used for the Phase I studies, and were labeled with 13C-octanoic acid for simultaneous measurement of gastric emptying. Radioimmunoassay All samples were extracted and assayed in duplicate. The standard curve ranged from 0 to 25 pmol/L and is illustrated in Fig. 6-1. The sensitivity of the assay, given by

PAGE 110

94 the commercial kit, was 0.3 pmol/L. CCK c oncentrations of the radioimmunoassay (RIA) controls (low dose and high dose controls) we re detected within the limits given by the kit. Standard CCK controls for the extract ion procedure provided a 73.5% recovery. For the recovery control, plasma samples obtained from a horse after an overnight of food deprivation were spik ed with known concentrations of CCK-8 standard (0, 0.78, 1.56, 3.12, 6.25, 12.5, and 25 pmol/L) provided by the commercial kit (Fig. 6-1). Recovery was good for all the concentrations (mean 69%, range 6896%), except for the sample with 3.12 pmol/L (32% recovery). Plasma samples from horses that were fed either the high-fat or the highcarbohydrate pelleted diet were also included in the assay. CCK was not detected in any of the samples, including those that were treated with aprotinin. Since CCK may be susceptible to degradation if plasma is not processed and frozen within two hours after collection, an additiona l experiment was performed to test whether early degradation of equine CCK was responsib le for the failure of the RIA to detect CCK. After an overnight of food deprivation, a horse was offered a sweet feed meal (0.5 g/kg bwt) mixed with 30 ml of corn oil. Blood samples were collect ed from the jugular vein before and 10, 20, 30 and 50 minutes after ingestion of the meal. Blood samples were centrifuged immediately after collec tion, and the plasma frozen at -20o C for subsequent RIA. Since the aim of the experiment was to detect CCK and not to measure it, the samples were assayed in combination with only two RIA controls of CCK-8 standard (0 and 6.25 pmol/L), and two recovery controls with known original concentrations. CCK was detected in the RIA a nd recovery controls, bu t not in the equine plasma samples.

PAGE 111

95 20 40 60 80 100 00.781.563.126.2512.525 pmol/L% B/B0 Standard curve for CCK Equine plasma spiked with CCK Figure 6-1. The standard curve shows the fraction of 125I-radiolabeled CCK-8 bound to antibodies (B/B0) at increasing levels of sta ndard CCK. Equine plasma from one horse was spiked with known quantities of standard CCK to study the effect of equine plasma on CCK degrad ation. Serial dilutions of two equine plasma samples modeled a curve similar to the standard curve. The value corresponding to the equine samples that were not spiked with standard CCK (0 pg/mL) fell out of the graph. The failure to detect CCK-like activity in equine plasma samples may be explained by the inability of the technique to measur e CCK. However, there are some reasons to question this possibility. Firs t of all, it is known that CCK is well conserved among mammals, and even non-mammalian vertebrates.83 Moreover, it is known that gastrin, which shares an identical C-terminal tetrap eptide sequence with CCK (minimal structure necessary for biological activity of both hormones), is highly conserved between the horse and other mammals.83;84 Hence, it is reasonable to assume that the high structural homology of CCK is also maintained in the horse, and that the antibody of the commercial kit for human CCK used in this st udy, which binds the C-terminal sulphated

PAGE 112

96 octapeptide, also has the capacity to bind equine CCK. Second, a rabbit antibody against sulphated CCK-8, with the same sequence as the one used in this study, has been successful in detecting CCK in neur oendocrine cells of equine urethra.224 Finally, the bioassay also failed to detect CCK bioact ivity in equine plasma samples. Although plasma samples from only one horse were us ed for this technique, they were collected after the horse had ingested a mixed meal containing corn oil and phenylalanine, which are known to be potent stimulants of plasma CCK release in other species.9;14;57;99;102;159;236 Furthermore, plasma samples were concentrated prior to bioassay to increase the likelihood of dete cting very low CCK concentrations. In conclusion, although we cannot totally exclude the possibil ity that the radioimmunoassay technique was unable to detect equine plasma CCK, it is reasonable to suggest that ingestion of a high-fat or a high-carbohydrate m eal did not cause the release of CCK into the plasma in the present study. Many studies have evaluated the effect of dietary fatwhether ingested, delivered into the stomach or infused into the duodenum on CCK secretion. Fat stimulates release of CCK into the circulation in humans,99;102;127 rats,96 dogs,54;102;204 pigs,38;102 cats,10 cows,28 and chicks.115 Hence, fat is invariably a potent stimulus of CCK release in all the studied species, and it is surp rising that in the present study neither basal nor postprandial plasma CCK was detected after inges tion of the high-fat pelleted diet. As also observed in other species, the chai n length of fatty acids may determine the CCK response. For example, in humans only fatty acids of chain lengths longer than C11 are potent stimulants,129 128 whereas medium-chain fatty ac ids (C10 or shorter) are more powerful stimulators of CCK secretion th an long-chain fatty acids in chicks115 and in

PAGE 113

97 rats.46;100 In the present study, the fat source of the high-fat meal (8 % fat) was soybean oil, which is rich in the l ong-chain triglycerides linoleic a nd oleic acid. Thus, similar to rats and chicks, medium-chain, rather than long-chain, fatty acids may be the major stimulant of CCK secretion in the horse. Mabayo et al.115 postulated that the absence of a gall bladder in chicks and rats may preclude a need for long-chain tr iglyceride-dependent CCK release, and this may also apply to the horse. Similarly, Cuber et al.38 suggested that species differences in nutrient potency may be the result of adaptation of CCK secreting cells to dietary composition. The effect of medium-chain fatty acids on plasma CCK concentration was not determined in the pres ent study. However, test meals labeled with 13C-octanoic acid, a medium-chain fatty acid, fo r the breath test, did not increase plasma CCK in the horse. In all species except pigs,38 carbohydrates have been found to be at best weak stimulants of CCK release, compared to fat and protein.96;99;228 Starch or glucose are ineffective in elevating plasma CCK levels in the rat96;100;193 and cat,10 whereas in humans, they have a weak70;99 to non-significant effect.75;101 Similar to most of these other species, carbohydrates may not be effec tive in releasing CCK from enteroendocrine cells in the horse. Finally, dietary proteins are known to be potent stimulants of CCK release in humans,14;99;208 rat,96;100 pig,38 goat,57 cat,10 and dog.151 Intact protein, but not amino acids, stimulates CCK release in the rat,96;100 whereas the opposite effect is observed in humans.14;208 Yet, in the present study, the prot ein content of the test meals did not increase plasma levels of CCK.

PAGE 114

98 Furthermore, not even basal CCK levels were detected in the horses, in spite of the high sensitivity of the radioimmunoassay a nd bioassay methods. One possible reason is that CCK may not be an important hormone in the regulation of gast rointestinal function in the horse. However, CCK immunoreactivity has been found in endocrine cells of the equine duodenum,88 suggesting that it has a physiological role. The other possibility is that CCK functions in a paracr ine fashion such that its rel ease from enteroendocrine cells in response to intraluminal nutrients is not reflected by increased plasma levels. In support of this idea, there is increasi ng evidence that nutrien ts in the intestine release CCK that then acts on neural pathways to produce reflex changes in gastric function.165 Different anatomical and functional studies support th is view of a paracrine mechanism involving neural pathways, as opposed to a true hormonal mechanism. 1) Nerve endings within the small intestinal mucosa are a likely target of CCK action because of their close proximity to enteroendocrine cells.15 This is consistent with the existence of CCK-A receptors on vagal termin als within the mucosa, myenteric plexus, muscle, and nodose ganglia cells of humans and rat.147;149;202;230 2) Electrophysiological studies show that vagal sens ory neurons are directly res ponsive to CCK, specifically through the CCK-A receptor subtype.16;61;175;189 3) Both inhibition of CCK-A receptors with selective antagonists and selective destru ction of vagal afferent c-fibers with the sensory neurotoxin capsaicin have shown that CCK mediates its effect, in part, via a neurally dependent fashion. For example, in the rat, inhibition of gastric emptying in response to intestinal perfusion with fat and glucose was reduced or abolished, respectively, by administration of a specific CCK-A antagonist, and also attenuated by intestinal capsaicin treatment.61;74;167 As well, the effect of peptone on gastric emptying

PAGE 115

99 inhibition was abolished by a CCK-antagonist and attenuated by systemic capsaicin treatment.52 Similar results have been found with re spect to other CCK-mediated effects, such as inhibition of acid secretion,109;110 pancreatic enzyme secretion in rat and humans,97;174 gall bladder contraction in dog198 and suppression of food intake in rat.178;179 4) Finally, the effect of CCK-A antagonists is not only limited to nutrients that increase circulating concentrations of the pe ptide. For instance, ingestion and intestinal perfusion of carbohydrates, which do not increas e plasma levels of CCK, inhibit gastric emptying55;145;167 and reduce food intake18 via CCK-A receptors, supporting the concept of a paracrine action. Similarly, infu sion of exogenous CCK doses that reproduce endogenous plasma CCK concentr ations fail to produce the same physiological response. For example, Lloyd et al.110 demonstrated that inhibi tion of acid output by duodenal infusion of lipid and dextrose is mediated by CCK-A receptors in rats. In that study, only lipid induced an increase in plasma CCK, but levels 50-fold higher were required to achieve similar acid inhibition by exogenous CCK-8. To summarize all these obs ervations, Raybould and Lloyd168 proposed that CCK may be released by nutrients and achieve local ly high concentrations that stimulate CCK receptors on afferent nerve endings within the duodenal mucosa. Therefore, circulating levels of CCK are not indicative of the impor tance of a physiological effect of peripheral CCK in gastric emptying. Because lipid increas es plasma levels of CCK in all species, this nutrient may also act by releasing CCK into the bloodstream, and stimulate gastric vagal mechanoreceptors to produce a vagovaga l reflex decrease in gastric motility.168 However, in the horse, fat and other nutrien ts may induce the release of high local CCK

PAGE 116

100 concentrations that act in a paracrine or ne urocrine fashion, and not reach the systemic circulation in amounts detectable by radioimmunoassay or bioassay. Conclusions By use of a commercial, non-specific radioimmunoassay, no plasma CCK-like activity was detected in the horses before and after ingestion of the high-fat and the highcarbohydrate pelleted meals. Likewise, the bi oassay technique failed to detect CCK activity of concentrated plas ma obtained from a horse before and after ingestion of a sweet feed meal supplemented with corn oil and phenylalanine. These results do not support the original hypot hesis that dietary fat releases CCK into the circ ulation in the horse. Although the possibility that both techniques were unabl e to detect plasma CCK cannot be excluded, there are some findings to the contrary. First, it is known that CCK has a high structural homology among mammals,83 and that the C-terminal tetrapeptide sequence of equine gastrin is identica l to that of CCK of other mammals.83;84 Additionally, the antibody of the commercial kit used in this study has been previously used successfully to detect equine CCK, although it was not used for plasma samples.224 If lack of sensitivity is excluded, the horse is a unique species in the sense that fat is not a stimulant of plasma CCK release.10;28;38;54;96;99;102;127;204 One possible explanation for the results of the present study is that, in the horse, nutrients may stimulate the release of high local CCK concentrations that act in a pa racrine way, while not necessarily elevating plasma CCK levels. Investigations with highly selective and potent antagonists for CCK-A receptors, such as loxiglumide and devazepide, should make it possible to define the physiological importance of CCK in the horse. A second appro ach for determining a role of CCK in the

PAGE 117

101 horse could be documenting the expression of CCK-A receptors in vagal afferent endings within the gastrointestinal mucosa.

PAGE 118

102 CHAPTER 7 SUMMARY AND CONCLUSIONS The main objective of this study was to co mpare the effect of ingestion of a highfat meal versus a high-carbohydrate meal on four distinct parameters of gastrointestinal function in the horse: relaxati on of the proximal portion of the stomach, gastric emptying, pH of gastric contents, and plasma CCK response. Test meals had identical energy content and weight, and similar volume, so that any difference observed between meals could be attributed specifically to the fat and carbohydrate contents. First, meal-induced changes in tone of the proximal stomach were assessed by use of an electronic barostat. Ingestion of any of the test meals was immediately followed by a defined period of receptive relaxation; this response was slightly longer with the highcarbohydrate meals compared to the high-fat m eals. These results suggest that, in the horse, dietary composition affects receptive relaxation, and that other mechanisms besides mechanical stimulation may participate in this reflex. In addition, the effect of supplementing a control meal of sweet feed with either corn oil or an isocaloric amount of glucose prolonged the receptive relaxation, compared to the sweet feed meal alone. Therefore, both nutrients seem to influence the magnitude of this response. One of the possible mechanisms by which fat and carbohyd rates modulate receptive relaxation may involve nutrient-specific orosensory stimulation. Receptive relaxation was followed by a mo re prolonged period of accommodation after all test meals, except w ith the high-carbohydrate pellet ed preparation, but time of onset varied among meals. Since volume wa s similar for all meals, the accommodation

PAGE 119

103 reflex was probably not elicited from the stom ach, but from the intestine. Addition of either corn oil or an isoc aloric amount of glucose to a sweet feed meal prolonged accommodation, compared to the sweet feed m eal alone. This suggests that both nutrients have the ability to cause str ong feedback regulation from the intestine. Moreover, both fat and carbohydrate, in isocaloric amounts, a ppear to modulate accommodation to a similar extent. Second, rates of gastric emptying were assessed by use of the 13C-octanoic acid breath test technique. The high-fat meals a nd the high-carbohydrate meals emptied at similar rates, which is in disagreement with the prevailing dogma that fat is a more potent inhibitor of gastric emptying than carbohydrate.27;64;180;234 One possible explanation for this discrepancy is that, in the present st udy, a meal with a high content of fat was compared with a meal rich in carbohydrate, but each had the same caloric density. Thus, based on this study, both nutrient classes, in similar caloric amounts, may induce similar intestinal feedback modulation of gastri c emptying in the hors e. Interestingly, supplementing a control meal of sweet feed w ith corn oil or glucose did not result in delayed emptying, compared to the sweet feed meal alone. Therefore, factors other than energy, such as volume, may be more important in controlling gastric emptying in this species. Although overall gastric emptying was statis tically similar between high-fat meals and high-carbohydrate meals, there was a cons istent tendency for the latter to empty slower at the initial phase of emptying. Thus, as mentioned for receptive relaxation, carbohydrates may be more potent modulators of gastric emptying than fat in the horse, at least initially. Ca rbohydrates may do this by activati on of oral and/or intestinal

PAGE 120

104 nutrient-specific receptors. Further studies in cluding a larger sample size or the use of alternative methods that are se nsitive to different phases of emptying may help define the apparent more suppressive role of carbohydrat es, compared to fat, on gastric emptying in the horse. Third, pH of gastric contents was measur ed by a self-referencing electrode at the most acidic site within the stomach. The eff ect of the high-fat meal versus the highcarbohydrate meal differed between the pellet ed (Phase I) and the sweet feed meals (Phase II). On the one hand, ingestion of the high-carbohydrate pelleted meal, which was rich in starch, had the greatest buffering effect, and was the only one of all test meals that increased mean pH above 4. This higher pH may be explained by a greater capacity of this meal to stimulate salivary secretion, w ith its inherent buffering capacity, or a lesser capacity to stimulate gastric acid secretion, compared to the high-fat pelleted meal. On the other hand, the high-carbohydr ate sweet feed meal, which wa s enriched with glucose, produced the least increase in pH among the sw eet feed meals. Intr agastric production of volatile fatty acids and lactic acid, or the ac idic nature of the glucose solution used for supplementation may account for this differen ce, when compared to the high-fat sweet feed meal. Finally, that th e high-carbohydrate pelleted meal appeared to be the most effective in increasing intragastric pH is contrary to the generalized assumption that diets rich in starch are acidogenic. Further, the sm all volume of this meal and its relatively fast gastric emptying may have limited carbohydr ate fermentation within the stomach. Lastly, plasma CCK-like activity was measured by a commercial, non-specific radioimmunoassay, and a bioassay. Both me thods failed to detect CCK activity in plasma. Although these results may be explaine d by inability of both techniques to detect

PAGE 121

105 equine CCK, the antibody sequence of the radioimmunoassay used for this study has been previously used successfully to measure CCK in the horse.224 Another possibility is that nutrients may stimulate the release of CCK to act in a paracrine fashion on afferent nerve endings within the inte stinal mucosa, without elevat ing plasma CCK levels. Should this be true, the horse is diffe rent from other species where ingestion of fat is followed by increased concentrations of plasma CCK.10;28;38;54;96;99;102;127;204 Further studies using CCK-A receptor antagonists, such as devazepid e, would be helpful in determining the role of CCK in equine ga strointestinal physiology. In the present study, proximal gastric t one and gastric emptying were measured simultaneously, an approach that has never be en reported for the study of equine gastric physiology. To test the validity of such approach, additiona l studies were performed to determine the existence of interactions between both techniques. First, the effect of the 13C-octanoic acid breath test technique was determined by comparing proximal gastric relaxation induced by the test meals with that induced by the same meals labeled with 13C-octanoic acid. Meal labeli ng modified the relaxation response measured by the electronic barostat but the effect was diet-specific. This interaction may be explained by the fat compon ent carried by the breath test label added to the meal, and suggests that the concurrent use of both technique s may yield erroneous results, especially when using low-fat meals. Second, the effect of the barostat tech nique was determined by comparing gastric emptying parameters of the test meals measured with and without the presence of the barostat bag within the proxi mal stomach. In the present st udy, the barostat bag did not significantly affect gastric emptying of the test meals.

PAGE 122

106 In conclusion, the combined use of both t echniques appears to be reliable, unless a formulation with low fat content is used as the test meal. Since an interaction between both techniques may occur with some types of meal composition, results obtained should be taken with caution unless the existen ce of an interaction is ruled out. Fat-supplemented diets are becoming a common alternative to traditional highstarch diets as a way to increase dietary ener gy content for athletic horses, but little is known about how such modification may affect gastrointestinal func tion. This study has attempted to give an insight into the effect of fat-suppl ementation on gastric function. The results were somewhat surprising since th ey suggest that the re sponse of the equine stomach to dietary fat may not be profoundly different than that to dietary carbohydrate, which is not consistent with what has been found in other species.

PAGE 123

107 APPENDIX A INDIVIDUAL ANIMAL DATA Table A-1. Individual body weight, test meal weight (Phase I and II), breath test label composition, and sweet feed supplementation (Phase II). Weight Breath test label Suppl sweet feed meal (mL) Animals Body (kg) Test meal (g) Octa noic (mg)Egg yolks Corn oil 50% Dextrose Nina 532 266 798 3 29 120 Seven 453 227 680 3 30 114 Mama 350 175 525 2 19 79 Gambler 508 254 762 3 26 123 Dusty 546 273 819 3 28 123 Iso 485 243 728 3 25 109 Table A-2. Barostat raw data (Phase I): bag volumes for baseline and 2-min postprandial blocks after ingestion of the unlab eled high-fat pelleted meal acid. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM (Baseline) 149 -36 12 40 58 304 88 50 0-2 81 593 76 43 34 352 196 93 2-4 194 674 200 179 116 582 324 98 4-6 5 273 170 171 121 591 222 82 6-8 -183 358 9 139 90 394 134 89 8-10 -108 153 5 63 64 290 78 55 10-12 -28 153 33 9 32 403 100 66 12-14 -134 181 36 24 74 382 94 71 14-16 -158 169 50 87 15 403 94 76 16-18 -150 86 72 -91 49 528 82 97 18-20 -153 17 4 -45 54 36 -15 31 20-22 -179 66 27 -54 29 168 9 48 22-24 -173 19 30 -8 54 -65 -24 34 24-26 -163 347 7 -12 48 259 81 77 26-28 -174 312 8 -24 47 -31 23 65 28-30 -166 266 -10 -4 110 57 42 59 30-32 -188 93 -7 146 83 369 82 75 32-34 -149 155 -4 0 78 65 24 42 34-36 -146 374 -3 102 127 37 82 70 36-38 -160 419 -7 73 134 221 113 81 38-40 -132 361 -7 -10 75 324 102 81 40-42 -133 286 56 -35 127 162 77 61 42-44 6 375 25 38 143 151 123 56

PAGE 124

108 Table A-2. Continued. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM 44-46 148 130 2 20 183 258 124 40 46-48 206 343 7 -51 158 345 168 68 48-50 356 279 5 -74 87 217 145 68 50-52 367 188 -1 13 172 232 162 57 52-54 117 229 5 -33 201 339 143 58 54-56 -39 302 14 -48 269 333 139 74 56-58 259 209 2 -25 236 387 178 65 58-60 256 140 -3 -14 301 406 181 69 60-62 78 104 2 17 308 418 155 69 62-64 55 160 8 49 280 315 144 53 64-66 52 248 0 49 306 400 176 67 66-68 274 196 4 14 231 462 197 70 68-70 213 202 5 101 340 349 202 55 70-72 88 172 0 29 218 373 147 57 72-74 111 501 -7 23 280 527 239 96 74-76 268 234 -4 39 322 645 251 95 76-78 35 786 -1 116 253 784 329 149 78-80 217 334 -3 81 330 573 255 84 80-82 101 371 -2 101 238 510 220 78 82-84 184 446 -3 239 425 604 316 89 84-86 20 287 -2 165 446 657 262 105 86-88 -2 290 0 173 357 772 265 118 88-90 3 283 -3 109 440 514 224 91 Table A-3. Barostat raw data (Phase I): bag volumes for baseline and 2-min postprandial blocks after ingestion of the labeled high-fat pelleted meal. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM (Baseline) 241 324 356 270 104 181 246 38 0-2 373 898 376 406 346 655 509 91 2-4 347 873 387 431 340 1008 564 121 4-6 218 892 369 340 437 1062 553 139 6-8 178 1022 364 318 633 763 546 129 8-10 160 952 363 185 573 297 421 122 10-12 164 781 365 262 463 498 422 88 12-14 178 656 365 95 365 326 331 79 14-16 160 551 367 -13 399 541 334 90 16-18 150 801 370 -19 274 368 324 113 18-20 144 620 369 16 137 388 279 90 20-22 154 598 373 24 169 342 277 83 22-24 159 655 372 -14 45 366 264 102 24-26 154 515 373 4 139 247 239 75 26-28 155 765 373 -22 92 183 258 114

PAGE 125

109 Table A-3. Continued. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM 28-30 146 665 373 7 215 379 298 93 30-32 150 655 374 -1 111 540 305 106 32-34 147 467 377 40 146 523 283 81 34-36 143 180 376 26 305 229 210 50 36-38 155 167 376 -61 277 397 219 69 38-40 158 186 379 -52 405 477 259 81 40-42 150 148 379 -52 326 253 201 63 42-44 149 71 379 -41 358 318 206 70 44-46 148 17 378 -40 342 646 249 105 46-48 163 403 379 -1 298 750 332 104 48-50 153 140 382 -2 392 425 248 71 50-52 182 236 381 -51 262 332 224 62 52-54 205 446 381 -36 362 221 263 71 54-56 180 373 384 -4 550 250 289 78 56-58 199 225 386 45 362 279 249 51 58-60 245 495 386 -10 265 164 257 72 60-62 332 473 386 -21 263 648 347 91 62-64 248 411 386 48 291 563 325 71 64-66 226 728 388 193 488 542 428 82 66-68 307 627 389 101 429 414 378 70 68-70 320 542 389 99 505 341 366 65 70-72 220 518 389 97 389 340 325 60 72-74 243 437 393 137 412 444 344 51 74-76 319 253 394 135 265 908 379 111 76-78 312 298 394 250 355 632 373 55 78-80 320 265 395 497 335 922 456 99 80-82 337 278 396 605 279 219 352 56 82-84 366 583 396 505 474 514 473 33 84-86 449 536 396 310 408 650 458 49 86-88 465 456 398 404 368 427 420 15 88-90 554 275 399 536 407 435 434 42 Table A-4. Barostat raw data (Phase I): bag volumes for baseline and 2-min postprandial blocks after ingestion of the unlabeled high-carbohydrate pelleted meal. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM (Baseline) 445 24 -6 499 150 352 244 89 0-2 368 55 88 770 230 413 321 107 2-4 543 119 77 1266 799 870 612 188 4-6 380 185 4 1205 1227 936 656 218 6-8 362 238 -15 1180 909 817 582 186 8-10 104 277 -7 950 513 742 430 152 10-12 88 214 -14 716 423 571 333 116

PAGE 126

110 Table A-4. Continued. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM 12-14 133 140 -15 539 234 275 218 76 14-16 62 159 -18 425 257 302 198 66 16-18 126 100 -18 692 239 454 266 107 18-20 83 126 -19 342 219 677 238 101 20-22 85 83 -19 215 335 685 231 104 22-24 53 34 -14 361 417 529 230 95 24-26 61 176 -18 391 325 392 221 72 26-28 47 110 -16 472 399 307 220 82 28-30 43 -50 -18 299 373 302 158 76 30-32 81 -17 734 498 631 385 150 32-34 111 -12 461 418 648 325 121 34-36 92 185 -17 731 506 436 322 116 36-38 120 71 -16 492 489 428 264 94 38-40 90 228 -23 572 414 384 277 90 40-42 105 88 -21 563 353 426 252 93 42-44 156 23 -19 537 403 310 235 90 44-46 157 270 -19 519 358 512 299 86 46-48 105 342 -19 586 414 518 324 97 48-50 218 343 -19 549 326 451 311 81 50-52 301 344 -19 584 400 522 356 87 52-54 452 546 -19 448 304 555 381 88 54-56 231 744 -18 534 381 495 394 108 56-58 329 724 -17 490 344 477 391 100 58-60 381 662 -18 632 416 637 452 106 60-62 454 438 -18 508 425 661 411 93 62-64 227 93 -19 673 310 671 326 119 64-66 373 109 -19 557 392 684 349 108 66-68 289 160 -18 736 387 644 366 117 68-70 305 231 -19 606 348 706 363 107 70-72 398 69 -17 696 254 658 343 121 72-74 329 243 -17 610 281 614 343 98 74-76 486 300 -11 346 293 872 381 119 76-78 247 203 -18 770 343 870 402 141 78-80 315 454 -21 556 360 735 400 104 80-82 352 -21 490 354 843 404 139 82-84 293 -18 436 483 944 428 156 84-86 412 -18 578 417 956 469 157 86-88 491 -19 464 342 889 433 146 88-90 455 555 559 984 638 118

PAGE 127

111 Table A-5. Barostat raw data (Phase I): bag volumes for baseline and 2-min postprandial blocks after ingestion of the labele d high-carbohydrate pelleted meal. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM (Baseline) 867 210 18 436 77 265 312 126 0-2 932 656 161 482 161 680 512 125 2-4 1255 543 302 1098 482 1234 819 173 4-6 1247 641 61 1126 879 1084 840 178 6-8 1034 445 8 954 976 895 719 166 8-10 625 312 28 940 1028 606 590 154 10-12 1012 340 21 711 623 531 540 138 12-14 852 411 50 533 633 349 471 111 14-16 985 282 24 808 341 284 454 149 16-18 701 206 17 955 527 413 470 138 18-20 1326 295 14 938 287 490 558 198 20-22 798 256 17 741 438 444 449 120 22-24 893 201 21 578 378 486 426 124 24-26 476 285 24 770 410 550 419 103 26-28 1190 273 25 971 392 492 557 180 28-30 1213 270 20 803 307 851 577 183 30-32 956 322 31 814 562 750 573 141 32-34 1093 571 38 881 367 437 564 154 34-36 1309 287 19 633 279 462 498 183 36-38 1210 577 115 483 472 644 584 146 38-40 1008 409 48 542 458 545 502 126 40-42 924 129 15 602 450 469 432 134 42-44 522 460 12 518 241 483 373 84 44-46 627 547 15 205 517 864 463 124 46-48 675 170 11 495 531 837 453 127 48-50 502 549 15 726 390 436 436 97 50-52 627 491 17 595 571 499 467 93 52-54 553 694 43 737 431 575 505 103 54-56 528 641 160 503 414 376 437 67 56-58 826 386 37 443 519 462 446 103 58-60 683 395 15 338 414 807 442 114 60-62 736 499 73 334 431 578 442 92 62-64 593 588 58 424 346 704 452 95 64-66 579 336 39 338 395 480 361 75 66-68 717 656 134 564 523 307 483 91 68-70 689 412 156 471 354 656 456 81 70-72 720 707 146 492 438 742 541 95 72-74 561 545 99 456 410 456 421 69 74-76 645 478 141 431 577 803 512 92 76-78 550 421 158 465 451 944 498 104 78-80 407 655 149 665 617 677 528 86

PAGE 128

112 Table A-5. Continued. Blocks Nina Seven Mama Gambler Dusty Iso Mean SEM 80-82 374 460 95 628 542 459 426 75 82-84 575 334 89 505 565 699 461 89 84-86 597 449 44 483 487 544 434 81 86-88 587 172 437 438 570 441 74 88-90 455 33 323 521 424 351 86 Table A-6. Barostat raw data (Phase I): ingestion time (sec) for the pelleted diets. Unlabeled Octanoic acid-labeled Animals High-fat mealHigh-CHO meal High-fat meal High-CHO meal Nina 200 178 228 182 Seven 240 154 284 124 Mama 200 124 208 166 Gambler 206 164 608 234 Dusty 368 208 408 276 Iso 156 120 134 150 Table A-7. Barostat raw data (Phase II): bag volumes for baseline and 2-min postprandial blocks after ingestion of the unlabeled corn oil-enriched sweet feed meal. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM (Baseline) 200 308 17 289 87 134 173 47 0-2 131 -172 113 103 -43 532 111 97 2-4 565 426 203 863 490 967 585 116 4-6 623 970 71 951 928 1208 792 163 6-8 491 981 3 835 732 1121 694 164 8-10 407 1017 0 734 695 1038 649 161 10-12 384 1009 6 582 763 860 601 149 12-14 95 1048 53 701 505 346 458 155 14-16 64 1085 59 654 551 146 427 168 16-18 255 1086 134 787 582 488 555 142 18-20 -48 1077 137 522 262 343 382 159 20-22 149 1081 197 516 326 666 489 143 22-24 59 1061 228 717 522 356 490 147 24-26 451 1091 104 514 192 551 484 142 26-28 327 1095 203 522 256 292 449 137 28-30 170 1090 80 729 223 509 467 159 30-32 336 1103 123 608 380 569 520 137 32-34 441 1103 114 532 112 833 522 161 34-36 532 1100 123 466 165 465 475 143 36-38 543 1105 48 415 17 751 480 171 38-40 864 1082 21 454 75 447 490 172 40-42 497 1052 21 359 122 659 451 154 42-44 43 1118 60 304 220 528 379 165

PAGE 129

113 Table A-7. Continued. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM 44-46 360 1114 48 300 190 680 449 158 46-48 32 1109 40 286 61 744 378 184 48-50 -37 1108 86 277 63 685 364 182 50-52 -11 1016 29 293 173 244 291 153 52-54 -63 1080 17 188 129 352 284 170 54-56 66 1068 18 279 326 392 358 154 56-58 -34 980 43 477 163 335 327 151 58-60 3 910 13 468 312 228 322 138 60-62 0 757 12 489 129 477 311 126 62-64 158 901 17 521 204 446 375 130 64-66 114 860 150 608 273 624 438 123 66-68 275 750 158 540 178 535 406 97 68-70 221 874 73 598 216 303 381 122 70-72 202 712 33 476 333 159 319 100 72-74 75 567 15 348 218 417 273 86 74-76 46 748 9 304 389 901 400 148 76-78 298 669 27 415 220 369 333 87 78-80 205 733 58 444 320 479 373 96 80-82 163 574 11 546 231 418 324 92 82-84 29 650 24 424 91 816 339 140 84-86 102 413 18 439 214 347 255 70 86-88 117 278 16 414 223 211 210 56 88-90 77 180 24 346 249 427 217 63 Table A-8. Barostat raw data (Phase II): bag volumes for baseline and 2-min postprandial blocks after ingestion of the labeled corn oil-enriched sweet feed meal. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM (Baseline) 43 442 24 255 54 190 168 66 0-2 365 297 254 275 65 666 320 80 2-4 691 695 947 854 329 832 725 89 4-6 776 874 989 1078 808 1028 925 51 6-8 1029 881 680 1089 863 964 917 59 8-10 936 843 722 999 849 935 881 40 10-12 761 707 560 681 774 785 711 34 12-14 641 402 644 882 578 583 622 63 14-16 536 554 755 976 350 400 595 96 16-18 489 604 1021 1098 189 568 662 140 18-20 396 704 1125 803 87 648 627 145 20-22 523 391 1013 793 32 589 557 138 22-24 606 255 1126 764 -39 721 572 167 24-26 753 379 625 653 2 521 489 110 26-28 1003 346 1098 801 -33 793 668 176

PAGE 130

114 Table A-8. Continued. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM 28-30 677 295 943 634 15 639 534 134 30-32 623 388 695 745 -25 533 493 116 32-34 890 548 550 558 52 566 527 110 34-36 761 169 375 590 -19 259 356 116 36-38 703 495 483 622 -40 340 434 108 38-40 657 310 303 730 81 551 439 101 40-42 662 305 153 789 -11 241 357 126 42-44 724 254 96 378 6 310 295 103 44-46 477 278 148 614 -29 229 286 94 46-48 528 90 361 475 44 230 288 82 48-50 359 -8 604 437 -8 284 278 100 50-52 513 54 449 552 55 335 326 91 52-54 495 -24 435 569 -18 130 265 109 54-56 351 -195 445 496 69 429 266 111 56-58 378 192 556 505 -1 660 382 101 58-60 321 -17 452 517 54 448 296 92 60-62 -2 278 496 597 119 500 331 98 62-64 148 514 762 573 114 423 422 103 64-66 200 424 632 578 79 469 397 88 66-68 229 71 536 610 172 471 348 90 68-70 410 23 355 509 165 532 332 82 70-72 199 -109 584 554 15 769 335 143 72-74 89 45 617 584 58 443 306 111 74-76 176 70 288 442 185 598 293 80 76-78 63 -252 700 652 94 621 313 162 78-80 184 -41 559 692 134 825 392 142 80-82 58 -12 446 747 289 984 419 159 82-84 185 -105 579 508 137 691 333 125 84-86 380 -35 344 507 109 597 317 98 86-88 431 -148 473 515 80 852 367 144 88-90 345 -116 442 520 65 484 290 105 Table A-9. Barostat raw data (Phase II): bag volumes for baseline and 2-min postprandial blocks after ingestion of the unlabeled glucose-enriched sweet feed meal. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM (Baseline) 183 41 35 415 210 157 174 57 0-2 148 301 215 456 255 639 336 74 2-4 501 627 443 805 1027 1220 771 125 4-6 380 1287 441 769 1204 1203 881 166 6-8 221 1322 427 702 1238 1303 869 198 8-10 315 1504 719 595 1244 1292 945 191 10-12 462 1508 866 477 1200 1239 959 176

PAGE 131

115 Table A-9. Continued. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM 12-14 546 1507 771 368 1031 1164 898 171 14-16 746 1502 514 224 1107 1109 867 189 16-18 630 1480 728 377 840 1121 863 159 18-20 795 1397 534 175 769 1103 795 174 20-22 889 1198 627 102 514 987 719 159 22-24 827 1224 533 232 480 1062 726 154 24-26 926 976 380 170 378 800 605 138 26-28 1018 1037 422 215 373 793 643 144 28-30 972 946 647 218 333 1008 687 141 30-32 894 982 389 226 238 957 614 150 32-34 948 979 331 173 305 678 569 142 34-36 955 936 296 113 520 564 564 138 36-38 916 758 286 -20 421 823 531 148 38-40 917 906 277 78 376 635 532 141 40-42 909 1053 199 145 556 728 598 151 42-44 906 1018 102 298 528 1073 654 165 44-46 860 1264 213 187 451 1020 666 183 46-48 828 1092 174 37 487 891 585 172 48-50 900 799 122 161 567 1013 594 155 50-52 834 703 74 156 461 908 523 144 52-54 905 811 45 47 453 911 529 167 54-56 899 1283 53 79 824 911 675 203 56-58 772 670 73 88 658 932 532 148 58-60 730 688 183 105 657 924 548 134 60-62 794 1004 108 71 743 963 614 171 62-64 805 1103 87 73 527 928 587 178 64-66 751 1160 74 131 748 1054 653 187 66-68 737 1062 35 159 903 994 648 181 68-70 790 935 72 197 785 954 622 158 70-72 743 756 79 98 945 1038 610 171 72-74 730 539 29 288 812 581 496 119 74-76 628 454 32 117 1060 681 495 156 76-78 772 353 31 -42 983 689 464 170 78-80 724 382 23 212 1023 809 529 157 80-82 604 359 20 161 971 708 471 146 82-84 552 1054 42 183 731 919 580 164 84-86 545 1066 25 172 900 937 607 177 86-88 596 988 32 14 962 758 558 179 88-90 579 1087 26 67 955 764 579 183

PAGE 132

116 Table A-10. Barostat raw data (Phase II): bag volumes for baseline and 2-min postprandial blocks after ingestion of th e labeled glucose-enriched sweet feed meal. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM (Baseline) 228 255 29 352 50 82 166 53 0-2 239 436 394 401 123 541 356 61 2-4 710 871 1001 852 425 1256 852 114 4-6 836 1097 1261 961 509 1340 1001 124 6-8 873 1048 1180 884 647 1307 990 97 8-10 755 1109 1194 719 809 1204 965 93 10-12 702 1047 1175 858 1041 725 925 79 12-14 687 982 1174 625 1282 367 853 144 14-16 554 825 1242 274 1319 475 781 174 16-18 446 831 1289 248 1306 912 839 176 18-20 266 835 1318 192 1285 796 782 197 20-22 254 514 1270 106 1257 1009 735 209 22-24 333 617 1270 96 1105 952 729 187 24-26 376 270 1138 312 862 721 613 143 26-28 392 687 1463 137 950 840 745 188 28-30 411 421 1285 237 1009 394 626 171 30-32 264 313 1177 67 827 838 581 175 32-34 358 106 1139 215 824 837 580 168 34-36 315 -65 1126 152 907 815 542 193 36-38 238 482 983 622 761 776 644 106 38-40 468 213 898 442 698 722 574 100 40-42 606 -63 937 566 726 648 570 137 42-44 445 106 783 439 927 736 572 122 44-46 427 58 1017 229 687 416 472 139 46-48 524 -120 1013 301 635 873 538 167 48-50 615 226 880 372 657 865 603 107 50-52 609 477 1267 388 482 613 639 130 52-54 545 378 877 465 509 207 497 91 54-56 366 523 1097 467 401 227 513 124 56-58 320 107 932 349 509 569 464 115 58-60 501 535 869 329 314 681 538 87 60-62 532 866 818 239 198 656 552 116 62-64 350 766 686 252 267 358 447 91 64-66 292 783 761 311 167 617 488 108 66-68 247 460 845 317 344 686 483 96 68-70 410 507 688 299 354 455 452 56 70-72 133 321 610 367 186 281 316 68 72-74 196 332 625 298 429 605 414 71 74-76 154 211 654 429 323 664 406 89 76-78 357 153 595 548 440 766 477 86

PAGE 133

117 Table A-10. Continued. Blocks Nina Seven Mama Gambler Dusty Iso Mean SEM 78-80 333 108 404 512 197 608 360 77 80-82 367 -102 358 650 275 817 394 130 82-84 374 31 378 311 177 678 325 89 84-86 577 521 320 494 173 501 431 62 86-88 263 -49 243 484 282 690 319 102 88-90 496 -94 240 369 222 732 327 114 Table A-11. Barostat raw data (Phase II): bag volumes for baseline and 2-min postprandial blocks after ingestion of the labeled control sweet feed meal. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM (Baseline) 259 314 51 280 78 135 186 46 0-2 336 659 177 720 63 297 375 107 2-4 930 718 502 1178 213 984 754 144 4-6 754 898 363 1130 563 1087 799 122 6-8 635 749 94 1019 475 1030 667 145 8-10 517 528 172 923 357 780 546 112 10-12 505 -28 115 929 206 383 352 139 12-14 172 -239 324 797 262 373 281 136 14-16 217 -18 545 402 298 269 286 77 16-18 360 363 612 686 58 527 434 92 18-20 320 -31 553 425 21 681 328 117 20-22 802 -6 514 412 66 840 438 146 22-24 456 -40 462 305 2 629 302 110 24-26 496 184 574 655 87 639 439 99 26-28 680 218 1151 599 141 796 597 153 28-30 406 450 865 1058 203 767 625 132 30-32 437 74 521 708 -1 722 410 127 32-34 555 22 543 797 -16 748 441 145 34-36 646 200 467 557 30 1003 484 140 36-38 578 383 440 563 174 212 392 70 38-40 307 234 325 458 43 258 271 56 40-42 240 106 279 552 186 521 314 74 42-44 90 18 118 727 72 608 272 127 44-46 105 -233 139 459 0 477 158 112 46-48 118 -67 295 643 73 583 274 117 48-50 -47 -145 543 465 13 483 218 127 50-52 33 -173 232 617 34 570 219 130 52-54 91 -220 155 421 -25 515 156 112 54-56 369 -331 104 358 -39 369 138 117 56-58 467 -52 74 369 86 344 214 84 58-60 -47 -190 47 313 49 525 116 106 60-62 214 -108 78 306 67 386 157 74

PAGE 134

118 Table A-11. Continued. Blocks Nina Seven Mama Gambler Dusty Iso Mean SEM 62-64 168 225 152 638 -32 346 249 92 64-66 202 -114 134 621 70 325 206 102 66-68 400 -125 114 525 71 288 212 97 68-70 350 163 79 429 44 316 230 64 70-72 308 -237 62 530 120 170 159 105 72-74 188 -12 25 448 -47 151 126 75 74-76 292 -23 15 299 -29 325 147 71 76-78 126 -149 43 365 33 267 114 75 78-80 130 -211 57 381 -37 116 73 80 80-82 217 -176 36 413 23 196 118 83 82-84 353 -184 134 480 140 105 171 93 84-86 170 -144 158 517 68 15 131 90 86-88 19 -24 81 571 -34 123 123 93 88-90 7 41 32 497 12 126 119 78 Table A-12. Barostat raw data (Phase II): i ngestion time (sec) for th e sweet feed diets. Unlabeled Octanoic acid-labeled Animals High-fat meal High-CHO meal High-fat meal High-CHO meal Control meal Nina 208 228 394 368 172 Seven 376 400 284 254 166 Mama 180 172 204 176 160 Gambler 284 208 450 228 164 Dusty 186 258 186 672 160 Iso 210 337 164 164 170 Table A-13. Gastric emptying ra w data (Phase I): paramete rs of the pelleted diets. Parameters Nina Seven Mama GamblerDusty Iso T1/2 High-fat meal 1.96 2.84 2.80 4.11 1.21* 3.59 High-CHO meal 7.16 1.65 3.88 3.95 2.67 1.98 Tmax High-fat meal 1.51 1.69 1.99 2.63 1.14* 1.77 High-CHO meal 3.82 1.37 2.13 2.59 1.41 1.08 GEC High-fat meal 3.42 1.87 2.87 1.29 12.30* 1.74 High-CHO meal 0.88 3.20 2.50 1.67 1.92 2.84 *Not included in data analysis.

PAGE 135

119 Table A-14. Gastric emptying raw data (Phase II): parameters of the sweet feed diets. Parameters Nina Seven Mama Gambler Dusty Iso T1/2 Control meal 2.67 2.74 1.84 1.93 2.17 2.44 Corn oil-enriched meal 3.49 2.85 2.49 2.50 2.20 1.81 (without intragastric bag)3.24 3.08 2.26 2.88 2.72 3.84 Glucose-enriched meal 1.94 2.70 1.99 3.75 3.22 1.93 (without intragastric bag)1.86 3.31 2.13 4.67 3.43 3.06 Tmax Control meal 1.47 1.87 1.42 1.38 1.57 1.38 Corn oil-enriched meal 1.95 1.87 1.51 1.70 1.66 1.46 (without intragastric bag)2.04 2.28 1.50 1.81 1.82 2.28 Glucose-enriched meal 1.50 2.02 1.45 2.69 2.32 1.34 (without intragastric bag)1.60 2.38 1.59 2.68 2.71 2.18 GEC Control meal 2.28 2.33 3.98 2.82 3.07 2.39 Corn oil-enriched meal 1.89 2.47 2.69 2.61 3.16 3.65 (without intragastric bag)2.26 2.30 2.97 2.29 2.52 1.92 Glucose-enriched meal 3.01 2.36 3.47 1.56 2.19 2.83 (without intragastric bag)3.81 2.20 3.43 1.42 1.93 2.31 Table A-15. Intragastric pH raw data (Phase I): mean pH for baseline and 5-min postprandial blocks after ingestion of the unlabeled high-fat pelleted meal. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM (Baseline) 1.14 2.7 1.41 1.97 1.55 3.44 2.04 0.36 0-5 1.08 3.94 1.4 1.56 1.36 3.98 2.22 0.55 5-10 1.06 2.55 2.28 1.55 1.34 4.06 2.14 0.45 10-15 1.16 2.77 2.45 1.75 4.4 4.14 2.78 0.52 15-20 1.54 3.83 3.86 2.17 5.38 4.24 3.50 0.58 20-25 1.59 3.63 4.47 2.22 4.26 4.48 3.44 0.51 25-30 1.14 3.01 4.76 2.14 3.11 4.56 3.12 0.57 30-35 1.03 2.25 4.43 1.98 1.1 4.44 2.54 0.63 35-40 1.24 2.16 3.81 1.9 1.09 4.35 2.43 0.55 40-45 1.85 2.33 2.94 1.78 1.19 4.29 2.40 0.45 45-50 2.76 2.24 3.28 1.74 1.23 4.38 2.61 0.46 50-55 2.91 1.93 3.59 1.73 1.52 4.24 2.65 0.45 55-60 3.54 1.52 4.29 1.73 1.41 4.16 2.78 0.56 60-65 3.75 1.31 3.81 1.69 1.86 4.04 2.74 0.51 65-70 3.73 1.25 3.86 1.69 1.56 3.85 2.66 0.52 70-75 3.6 1.22 4.07 1.66 1.47 3.63 2.61 0.53 75-80 3.64 1.21 3.39 1.66 1.44 3.31 2.44 0.46 80-85 3.47 1.22 2.91 1.69 1.34 3.09 2.29 0.40 85-90 3.08 1.24 3.38 1.73 1.26 2.81 2.25 0.39 90-95 2.58 1.27 3.29 1.75 1.17 2.39 2.08 0.34

PAGE 136

120 Table A-15. Continued. Blocks Nina Seven Mama Gambler Dusty Iso Mean SEM 95-100 1.88 1.3 3.67 1.74 1.22 2 1.97 0.36 100-105 1.39 1.3 2.62 1.73 1.3 1.75 1.68 0.21 105-110 1.32 1.29 2.33 1.8 1.31 1.63 1.61 0.17 110-115 1.14 1.3 2.5 1.81 1.28 1.64 1.61 0.20 115-120 1.05 1.32 2 1.82 1.32 1.63 1.52 0.15 Table A-16. Intragastric pH raw data (Phase I): mean pH for baseline and 5-min postprandial blocks after ingestion of the labeled high-fat pelleted meal. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM (Baseline) 1.06 1.18 1.67 1.48 1.46 1.31 1.36 0.09 0-5 0.92 1.3 1.61 1.41 1.4 1.26 1.32 0.09 5-10 1.05 1.42 1.63 1.78 1.48 1.33 1.45 0.10 10-15 1.34 2.57 1.66 2.85 2.33 1.36 2.02 0.27 15-20 1.44 3.6 1.61 3.86 3.02 1.4 2.49 0.46 20-25 1.54 4.74 1.67 2.4 3.11 1.47 2.49 0.52 25-30 1.94 5.16 1.59 3.3 3.19 1.91 2.85 0.55 30-35 2.13 5.1 1.46 3.77 3.3 2.94 3.12 0.52 35-40 2.69 5.02 1.41 3.7 3.45 3.48 3.29 0.49 40-45 3.44 4.9 1.47 3.33 3.43 4.63 3.53 0.50 45-50 3.6 4.42 1.43 4.03 3.51 3.47 3.41 0.42 50-55 3.49 3.92 1.39 4.3 3.5 1.39 3.00 0.52 55-60 3.33 2.77 1.44 4.64 3.57 1.45 2.87 0.51 60-65 3.44 1.95 1.28 3.9 3.57 1.43 2.60 0.48 65-70 3.45 1.72 1.36 3.92 3.46 1.21 2.52 0.50 70-75 3.4 1.68 1.41 4.09 3.29 1.24 2.52 0.50 75-80 3.19 1.67 1.43 3.8 3.11 1.2 2.40 0.45 80-85 2.8 1.66 1.48 4.33 2.68 1.21 2.36 0.47 85-90 2.68 1.66 1.5 4.5 2.35 1.26 2.33 0.49 90-95 2.63 1.69 1.51 3.32 2.3 1.34 1.89 0.22 95-100 2.24 1.62 1.49 2.76 2.12 1.42 1.94 0.21 100-105 1.71 1.56 1.54 2.64 1.94 1.49 1.81 0.18 105-110 1.23 1.58 1.52 2.02 1.88 1.56 1.63 0.11 110-115 1.13 1.6 1.51 1.81 1.73 1.56 1.56 0.10 115-120 1.07 1.64 1.47 1.54 1.63 1.57 1.49 0.09

PAGE 137

121 Table A-17. Intragastric pH raw data (Phase I): mean pH for baseline and 5-min postprandial blocks after ingestion of the unlabeled high-CHO pelleted meal. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM (Baseline) 1.71 2.08 1.97 2.28 1.89 1.81 1.96 0.08 0-5 0.85 1.61 1.97 1.62 1.27 2.06 1.56 0.18 5-10 0.82 1.59 3.74 1.7 1.36 2.33 1.92 0.41 10-15 0.98 4.39 4.42 2.18 1.59 2.87 2.74 0.59 15-20 1.05 5.88 4.25 2.56 1.71 3.78 3.21 0.73 20-25 1.12 5.71 4.12 2.39 1.8 4.34 3.25 0.72 25-30 1.36 6.16 3.7 2.83 2.01 4.83 3.48 0.73 30-35 1.82 6.74 4.02 3.29 2.13 5.09 3.85 0.76 35-40 3.49 6.58 4.06 3.49 2.67 5.09 4.23 0.57 40-45 3.68 6.55 3.89 3.63 3.95 5.03 4.46 0.47 45-50 3.67 6.21 4.51 3.71 3.85 5.17 4.52 0.41 50-55 3.74 5.81 4.71 3.8 3.79 5.14 4.50 0.35 55-60 3.81 5.32 4.54 3.84 3.83 5.09 4.41 0.28 60-65 3.9 6.26 3.9 3.84 3.64 5.1 4.44 0.42 65-70 4 6.21 3.06 3.84 3.4 5.06 4.26 0.48 70-75 4.07 5.85 3.22 3.81 3.15 4.99 4.18 0.43 75-80 4.02 6.06 2.43 3.73 2.9 4.8 3.99 0.54 80-85 3.87 4.73 1.99 3.68 2.37 4.65 3.55 0.47 85-90 3.82 2.54 1.89 3.41 1.96 4.39 3.00 0.42 90-95 3.79 2.49 1.74 3.21 1.86 4.04 2.86 0.40 95-100 3.71 2.62 1.74 3.01 1.81 3.79 2.78 0.36 100-105 3.58 2.97 1.69 2.58 1.72 3.53 2.68 0.34 105-110 3.41 2.35 1.67 2.3 1.75 3.32 2.30 0.28 110-115 3.28 2.17 1.67 2.22 1.8 3.14 2.38 0.28 115-120 3.11 1.83 1.72 2.25 1.67 3 2.26 0.26 Table A-18. Intragastric pH raw data (Phase I): mean pH for baseline and 5-min postprandial blocks after ingestion of the labeled high-CHO pelleted meal. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM (Baseline) 2.89 1.66 1.58 4.23 1.2 3.44 2.50 0.49 0-5 1.67 2.32 1.57 2.36 1.14 4.68 2.29 0.51 5-10 2.3 3.1 1.51 1.74 1.39 4.37 2.40 0.47 10-15 2.51 3.99 1.96 3.57 1.63 4.31 3.00 0.46 15-20 3.57 4.72 2.74 4.51 1.91 4.32 3.63 0.45 20-25 4.29 5.45 2.08 4.86 2.59 4.95 4.04 0.56 25-30 5.4 5.46 1.44 4.63 2.91 4.88 4.12 0.66 30-35 5.43 5.1 1.45 4.56 3.52 4.94 4.17 0.61 35-40 5.59 4.89 1.59 4.57 3.93 5.04 4.27 0.58 40-45 5.25 4.75 1.9 5.1 4.21 4.97 4.36 0.51 45-50 5.02 4.57 1.9 5.11 4.47 5.08 4.36 0.50 50-55 4.84 4.36 1.8 4.76 4.73 5.14 4.27 0.50

PAGE 138

122 Table A-18. Continued. Blocks Nina Seven Mama GamblerDusty Iso Mean SEM 55-60 4.7 4.25 1.86 4.38 4.54 5.07 4.13 0.47 60-65 4.59 4.08 1.74 4.26 4.54 5.22 4.07 0.49 65-70 4.51 3.85 1.72 4.1 4.48 4.86 3.92 0.46 70-75 4.44 3.67 1.87 3.78 4.47 5.06 3.88 0.45 75-80 4.36 3.5 2.24 3.39 4.36 5.07 3.82 0.41 80-85 4.22 3.27 2.1 2.57 4.34 5.1 3.60 0.47 85-90 4.07 3.01 2.16 2.28 4.37 5.2 3.52 0.50 90-95 3.89 2.73 2.19 1.96 4.34 5.08 3.37 0.51 95-100 3.59 2.1 2.35 1.78 4.29 4.62 3.12 0.49 100-105 3.17 1.81 2.12 1.46 4.28 4.06 2.82 0.49 105-110 2.87 2.12 1.95 1.27 4.19 3.74 2.69 0.46 110-115 2.46 2.44 1.83 1.12 3.79 3.67 2.55 0.42 115-120 2.26 1.72 1.7 1.24 3.47 3.22 2.27 0.37 Table A-19. Intragastric pH raw data (Phase II): mean pH for baseline and 5-min postprandial blocks after ingestion of the labeled corn oil-enriched meal. Blocks Nina Seven Mama GamblerDusty Iso (Baseline) 1.94 1.37 1.97 1.85 1.59 1.94 0-5 1.65 1.39 1.42 1.86 1.43 1.75 5-10 1.61 1.79 1.62 1.68 1.45 1.74 10-15 1.56 2.73 1.57 1.63 1.48 1.69 15-20 1.58 3.2 1.34 1.85 1.97 1.97 20-25 1.54 3.23 1.72 2.27 2.6 2.58 25-30 1.76 3.4 3.33 2.51 2.98 3.9 30-35 4.63 3.62 4.13 2.89 2.29 4.33 35-40 4.79 3.65 3.76 3.32 2.14 4.49 40-45 3.67 3.94 3.47 2.77 2.78 4.5 45-50 4.15 3.47 2.82 2.85 1.75 4.45 50-55 4.2 3.11 2.9 2.77 1.96 4.41 55-60 3.12 2.82 2.4 2.63 1.96 4.43 60-65 2.06 2.39 2.31 2.3 1.73 4.33 65-70 1.83 2.78 1.86 1.98 2.29 4.15 70-75 1.94 1.97 1.63 1.83 1.62 3.96 75-80 1.9 1.7 1.23 1.83 1.34 3.66 80-85 2.26 1.57 0.98 1.75 1.24 3.3 85-90 1.83 1.51 0.92 1.71 1.19 3.15 90-95 1.78 1.48 0.89 1.72 1.16 3.09 95-100 1.72 1.42 0.85 1.7 1.1 2.84 100-105 1.63 1.41 0.94 1.67 1.14 2.63 105-110 1.69 1.42 1.13 1.72 1.13 2.49 110-115 1.74 1.37 1.48 1.77 1.11 2.26 115-120 1.86 1.35 1.31 1.79 1.08 2.28

PAGE 139

123 Table A-20. Intragastric pH raw data (Phase II): mean pH for baseline and 5-min postprandial blocks after ingestion of the labeled corn oil-enriched meal. Blocks Nina Seven Mama GamblerDusty Iso (Baseline) 2.12 1.39 2.92 1.35 2.27 1.62 0-5 1.91 1.53 2.77 1.29 2.16 1.33 5-10 2.58 1.83 5.52 1.24 2.03 2.12 10-15 2.04 4.18 4.5 1.25 2.3 4.09 15-20 2.9 4.46 4.79 1.21 2.88 4.31 20-25 3.38 4.2 5.17 1.56 4.45 3.97 25-30 4.23 4 5.22 2.31 4.25 3.77 30-35 4.48 3.74 5.19 2.75 3.89 3.91 35-40 4.51 3.42 5.03 2.34 3.07 3.65 40-45 4.23 2.98 5.27 2.58 3.1 3.44 45-50 3.67 2.66 5.07 2.6 3.77 3.26 50-55 3.76 2.55 4.88 2.1 3.15 3.12 55-60 3.15 2.3 4.62 1.93 2.83 2.83 60-65 3.4 2.32 3.97 1.94 2.34 2.66 65-70 3.18 2.12 3.9 1.75 2.62 2.34 70-75 2.81 1.82 4.1 1.56 2.48 2.1 75-80 2.97 1.67 4.2 1.45 2.21 1.9 80-85 2.81 1.51 3.88 1.29 1.8 1.88 85-90 2.54 1.4 3.8 1.24 1.47 1.76 90-95 2.53 1.29 3.54 1.21 1.43 1.56 95-100 2.48 1.23 3.85 1.14 1.37 1.4 100-105 1.8 1.19 4.13 1.13 1.19 1.65 105-110 1.9 1.14 4.53 1.13 1.12 1.91 110-115 1.86 1.12 4.98 1.15 1.09 1.96 115-120 1.73 1.04 5.75 1.12 1.06 1.72 Table A-21. Intragastric pH raw data (Phase II): mean pH for baseline and 5-min postprandial blocks after ingestion of the labeled glucose-enriched meal. Blocks Nina Seven Mama GamblerDusty Iso (Baseline) 1.77 1.55 3.1 1.77 3.81 2.22 0-5 1.59 1.48 2.73 1.93 1.45 1.51 5-10 1.52 2.76 5.37 2.77 1.74 2.44 10-15 1.97 3.92 5.01 1.9 4.03 3.03 15-20 2.26 3.99 4.95 1.76 4.69 3.73 20-25 3.03 3.74 4.87 1.87 4.83 3.45 25-30 2.89 3.52 4.33 2.35 4.94 3.16 30-35 1.74 3.34 4.26 3.46 4.9 2.62 35-40 1.73 2.84 4.23 3.62 4.67 2.25 40-45 1.78 3.04 3.8 3.09 4.77 1.92 45-50 1.58 2.85 3.43 2.54 4.92 1.64 50-55 1.27 2.61 2.99 2.41 5.09 1.61

PAGE 140

124 Table A-21. Continued. Blocks Nina Seven Mama GamblerDusty Iso 55-60 1.26 2.47 2.88 2.15 5.24 1.48 60-65 1.13 2.3 2.69 1.86 5.22 1.33 65-70 1.26 2.12 2.02 1.88 4.95 1.36 70-75 1.2 1.91 1.77 1.7 4.38 1.42 75-80 1.19 1.59 1.47 1.61 3.83 1.45 80-85 1.18 1.43 1.36 1.5 3.62 1.52 85-90 1.24 1.35 1.22 1.46 3.06 1.54 90-95 1.16 1.22 1.07 1.39 1.95 1.52 95-100 1.08 1.15 1.04 1.31 2.41 1.46 100-105 1.09 1.06 1.02 1.34 2.75 1.39 105-110 1.05 1.1 1.18 1.33 1.96 1.36 110-115 1.04 1.09 1.22 1.38 1.66 1.36 115-120 1.1 1.12 1.19 1.45 1.53 1.42

PAGE 141

125 APPENDIX B STATISTICAL TESTS FOR BAROSTAT DATA Table B-1. ANOVA and CL for ingestion times of unlabeled pelleted meals (Phase I). Class Levels Values horse 6 DUSTY GAMBLER ISO MAMA NINA SEVEN diet 2 FAT CHO Source DF Sum of Squares Mean Square F Value Pr > F Model 6 41054.00000 6842.33333 5.42 0.0419 Error 5 6317.66667 1263.53333 Total 11 47371.66667 Source DF Anova SS Mean Square F Value Pr > F horse 5 26213.66667 5242.73333 4.15 0.0722 diet 1 14840.33333 14840.33333 11.75 0.0187 Variable Lower 95% CL Upper 95% CL FAT 151.2453980 305.4212687 CHO 123.0300078 192.9699922 Table B-2. Shapiro-Wilk test for normality of ingestion times of unlabeled pelleted meals (Phase I). Source Statistic (W) p-value Diets Difference 0.885335 0.2945 Table B-3. Bartlett's test for homogeneity of variance of unlabeled pelleted meals (Phase I). Source DF Chi-Square p-value Diet 1 2.5858 0.1078 Table B-4. ANOVA mixed procedure for mean bag volumes of pelleted meals (Phase I). A: labeled high-fat meal; B: labeled high-CHO meal; C: unlabeled high-fat meal; D: unlabeled high-CHO meal; 0: baseline; 1-45: 2-min postprandial blocks. Class Levels Values diet 4 A B C D horse 6 Dusty Gambler Iso Mama Nina Seven time 46 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

PAGE 142

126 Table B-4. Continued. Effect NumDF DenDF F Value Pr > F diet 3 20 1.77 0.1848 time 45 888 5.93 <.0001 diet*time 135 888 1.31 0.0161 Differences of Least Squares Means Within Diet (Relaxation Response) Effect diet time _diet _time Es timate SEM DF t Value Pr > |t| diet*time A 0 A 1 -303.51 100.18 888 -3.03 0.0025 diet*time A 0 A 2 -358.84 100.18 888 -3.58 0.0004 diet*time A 0 A 3 -347.51 100.18 888 -3.47 0.0005 diet*time A 0 A 4 -340.84 100.18 888 -3.40 0.0007 diet*time A 0 A 5 -216.18 100.18 888 -2.16 0.0312 diet*time A 0 A 6 -216.68 100.18 888 -2.16 0.0308 diet*time A 0 A 7 -125.34 100.18 888 -1.25 0.2112 diet*time A 0 A 8 -128.68 100.18 888 -1.28 0.1993 diet*time A 0 A 9 -118.51 100.18 888 -1.18 0.2372 diet*time A 0 A 10 -73.5090 100.18 888 -0.73 0.4633 diet*time A 0 A 11 -71.1757 100.18 888 -0.71 0.4776 diet*time A 0 A 12 -58.3424 100.18 888 -0.58 0.5605 diet*time A 0 A 13 -33.1757 100.18 888 -0.33 0.7406 diet*time A 0 A 14 -52.1757 100.18 888 -0.52 0.6026 diet*time A 0 A 15 -92.0090 100.18 888 -0.92 0.3587 diet*time A 0 A 16 -99.3424 100.18 888 -0.99 0.3217 diet*time A 0 A 17 -77.8424 100.18 888 -0.78 0.4374 diet*time A 0 A 18 -4.3424 100.18 888 -0.04 0.9654 diet*time A 0 A 19 -13.0090 100.18 888 -0.13 0.8967 diet*time A 0 A 20 -53.3424 100.18 888 -0.53 0.5946 diet*time A 0 A 21 4.8243 100.18 888 0.05 0.9616 diet*time A 0 A 22 -0.1757 100.18 888 -0.00 0.9986 diet*time A 0 A 23 -43.0090 100.18 888 -0.43 0.6678 diet*time A 0 A 24 -126.51 100.18 888 -1.26 0.2070 diet*time A 0 A 25 -42.8424 100.18 888 -0.43 0.6690 diet*time A 0 A 26 -18.1757 100.18 888 -0.18 0.8561 diet*time A 0 A 27 -57.6757 100.18 888 -0.58 0.5650 diet*time A 0 A 28 -83.3424 100.18 888 -0.83 0.4057 diet*time A 0 A 29 -43.8424 100.18 888 -0.44 0.6618 diet*time A 0 A 30 -52.0090 100.18 888 -0.52 0.6038 diet*time A 0 A 31 -141.34 100.18 888 -1.41 0.1586 diet*time A 0 A 32 -119.01 100.18 888 -1.19 0.2352 diet*time A 0 A 33 -222.01 100.18 888 -2.22 0.0269 diet*time A 0 A 34 -172.34 100.18 888 -1.72 0.0857 diet*time A 0 A 35 -160.51 100.18 888 -1.60 0.1095 diet*time A 0 A 36 -120.01 100.18 888 -1.20 0.2313 diet*time A 0 A 37 -138.84 100.18 888 -1.39 0.1661

PAGE 143

127 Table B-4. Continued. Effect diet time _diet _time Es timate SEM DF t Value Pr > |t| diet*time A 0 A 38 -173.51 100.18 888 -1.73 0.0836 diet*time A 0 A 39 -168.01 100.18 888 -1.68 0.0939 diet*time A 0 A 40 -250.18 100.18 888 -2.50 0.0127 diet*time A 0 A 41 -146.84 100.18 888 -1.47 0.1431 diet*time A 0 A 42 -267.51 100.18 888 -2.67 0.0077 diet*time A 0 A 43 -252.68 100.18 888 -2.52 0.0118 diet*time A 0 A 44 -214.18 100.18 888 -2.14 0.0328 diet*time A 0 A 45 -228.84 100.18 888 -2.28 0.0226 diet*time B 0 B 1 -183.67 89.4111 888 -2.05 0.0403 diet*time B 0 B 2 -517.33 89.4111 888 -5.79 <.0001 diet*time B 0 B 3 -569.17 89.4111 888 -6.37 <.0001 diet*time B 0 B 4 -452.67 89.4111 888 -5.06 <.0001 diet*time B 0 B 5 -321.83 89.4111 888 -3.60 0.0003 diet*time B 0 B 6 -230.67 89.4111 888 -2.58 0.0100 diet*time B 0 B 7 -323.50 89.4111 888 -3.62 0.0003 diet*time B 0 B 8 -338.50 89.4111 888 -3.79 0.0002 diet*time B 0 B 9 -270.50 89.4111 888 -3.03 0.0026 diet*time B 0 B 10 -320.67 89.4111 888 -3.59 0.0004 diet*time B 0 B 11 -261.67 89.4111 888 -2.93 0.0035 diet*time B 0 B 12 -220.83 89.4111 888 -2.47 0.0137 diet*time B 0 B 13 -241.17 89.4111 888 -2.70 0.0071 diet*time B 0 B 14 -301.33 89.4111 888 -3.37 0.0008 diet*time B 0 B 15 -243.00 89.4111 888 -2.72 0.0067 diet*time B 0 B 16 -334.00 89.4111 888 -3.74 0.0002 diet*time B 0 B 17 -405.17 89.4111 888 -4.53 <.0001 diet*time B 0 B 18 -264.33 89.4111 888 -2.96 0.0032 diet*time B 0 B 19 -319.50 89.4111 888 -3.57 0.0004 diet*time B 0 B 20 -235.67 89.4111 888 -2.64 0.0085 diet*time B 0 B 21 -174.50 89.4111 888 -1.95 0.0513 diet*time B 0 B 22 -128.33 89.4111 888 -1.44 0.1515 diet*time B 0 B 23 -117.83 89.4111 888 -1.32 0.1879 diet*time B 0 B 24 -105.17 89.4111 888 -1.18 0.2398 diet*time B 0 B 25 -183.83 89.4111 888 -2.06 0.0401 diet*time B 0 B 26 -205.17 89.4111 888 -2.29 0.0220 diet*time B 0 B 27 -196.17 89.4111 888 -2.19 0.0285 diet*time B 0 B 28 -137.83 89.4111 888 -1.54 0.1235 diet*time B 0 B 29 -179.50 89.4111 888 -2.01 0.0450 diet*time B 0 B 30 -138.50 89.4111 888 -1.55 0.1217 diet*time B 0 B 31 -198.17 89.4111 888 -2.22 0.0269 diet*time B 0 B 32 -160.00 89.4111 888 -1.79 0.0739 diet*time B 0 B 33 -132.33 89.4111 888 -1.48 0.1392 diet*time B 0 B 34 -252.33 89.4111 888 -2.82 0.0049

PAGE 144

128 Table B-4. Continued. Effect diet time _diet _time Es timate SEM DF t Value Pr > |t| diet*time B 0 B 35 -180.67 89.4111 888 -2.02 0.0436 diet*time B 0 B 36 -306.83 89.4111 888 -3.43 0.0006 diet*time B 0 B 37 -213.67 89.4111 888 -2.39 0.0171 diet*time B 0 B 38 -248.83 89.4111 888 -2.78 0.0055 diet*time B 0 B 39 -207.83 89.4111 888 -2.32 0.0203 diet*time B 0 B 40 -275.33 89.4111 888 -3.08 0.0021 diet*time B 0 B 41 -227.50 89.4111 888 -2.54 0.0111 diet*time B 0 B 42 -224.83 89.4111 888 -2.51 0.0121 diet*time B 0 B 43 -206.50 89.4111 888 -2.31 0.0211 diet*time B 0 B 44 -206.55 93.8709 888 -2.20 0.0280 diet*time B 0 B 45 -128.55 93.8709 888 -1.37 0.1712 diet*time C 0 C 1 -196.33 89.4111 888 -2.20 0.0284 diet*time C 0 C 2 -324.33 89.4111 888 -3.63 0.0003 diet*time C 0 C 3 -221.83 89.4111 888 -2.48 0.0133 diet*time C 0 C 4 -134.67 89.4111 888 -1.51 0.1324 diet*time C 0 C 5 -77.8333 89.4111 888 -0.87 0.3843 diet*time C 0 C 6 -100.33 89.4111 888 -1.12 0.2621 diet*time C 0 C 7 -94.1667 89.4111 888 -1.05 0.2925 diet*time C 0 C 8 -94.1667 89.4111 888 -1.05 0.2925 diet*time C 0 C 9 -82.3333 89.4111 888 -0.92 0.3574 diet*time C 0 C 10 14.6667 89.4111 888 0.16 0.8697 diet*time C 0 C 11 -9.3333 89.4111 888 -0.10 0.9169 diet*time C 0 C 12 24.0000 89.4111 888 0.27 0.7884 diet*time C 0 C 13 -80.8333 89.4111 888 -0.90 0.3662 diet*time C 0 C 14 -23.1667 89.4111 888 -0.26 0.7956 diet*time C 0 C 15 -42.0000 89.4111 888 -0.47 0.6387 diet*time C 0 C 16 -82.6667 89.4111 888 -0.92 0.3554 diet*time C 0 C 17 -24.1667 89.4111 888 -0.27 0.7870 diet*time C 0 C 18 -81.8333 89.4111 888 -0.92 0.3603 diet*time C 0 C 19 -113.17 89.4111 888 -1.27 0.2060 diet*time C 0 C 20 -101.83 89.4111 888 -1.14 0.2550 diet*time C 0 C 21 -77.5000 89.4111 888 -0.87 0.3863 diet*time C 0 C 22 -123.17 89.4111 888 -1.38 0.1687 diet*time C 0 C 23 -123.67 89.4111 888 -1.38 0.1670 diet*time C 0 C 24 -168.00 89.4111 888 -1.88 0.0606 diet*time C 0 C 25 -145.00 89.4111 888 -1.62 0.1052 diet*time C 0 C 26 -161.83 89.4111 888 -1.81 0.0706 diet*time C 0 C 27 -143.00 89.4111 888 -1.60 0.1101 diet*time C 0 C 28 -138.83 89.4111 888 -1.55 0.1208 diet*time C 0 C 29 -178.17 89.4111 888 -1.99 0.0466 diet*time C 0 C 30 -180.67 89.4111 888 -2.02 0.0436 diet*time C 0 C 31 -154.83 89.4111 888 -1.73 0.0837

PAGE 145

129 Table B-4. Continued. Effect diet time _diet _time Es timate SEM DF t Value Pr > |t| diet*time C 0 C 32 -144.33 89.4111 888 -1.61 0.1068 diet*time C 0 C 33 -176.00 89.4111 888 -1.97 0.0493 diet*time C 0 C 34 -196.67 89.4111 888 -2.20 0.0281 diet*time C 0 C 35 -201.50 89.4111 888 -2.25 0.0245 diet*time C 0 C 36 -146.67 89.4111 888 -1.64 0.1013 diet*time C 0 C 37 -239.33 89.4111 888 -2.68 0.0076 diet*time C 0 C 38 -251.00 89.4111 888 -2.81 0.0051 diet*time C 0 C 39 -329.00 89.4111 888 -3.68 0.0002 diet*time C 0 C 40 -255.17 89.4111 888 -2.85 0.0044 diet*time C 0 C 41 -219.83 89.4111 888 -2.46 0.0141 diet*time C 0 C 42 -316.00 89.4111 888 -3.53 0.0004 diet*time C 0 C 43 -262.17 89.4111 888 -2.93 0.0035 diet*time C 0 C 44 -265.17 89.4111 888 -2.97 0.0031 diet*time C 0 C 45 -224.33 89.4111 888 -2.51 0.0123 diet*time D 0 D 1 -76.6667 89.4111 888 -0.86 0.3914 diet*time D 0 D 2 -368.33 89.4111 888 -4.12 <.0001 diet*time D 0 D 3 -412.17 89.4111 888 -4.61 <.0001 diet*time D 0 D 4 -337.83 89.4111 888 -3.78 0.0002 diet*time D 0 D 5 -185.83 89.4111 888 -2.08 0.0380 diet*time D 0 D 6 -89.000 89.4111 888 -1.00 0.3198 diet*time D 0 D 7 26.333 89.4111 888 0.29 0.7684 diet*time D 0 D 8 46.1667 89.4111 888 0.52 0.6057 diet*time D 0 D 9 -21.500 89.4111 888 -0.24 0.8100 diet*time D 0 D 10 6.0000 89.4111 888 0.07 0.9465 diet*time D 0 D 11 13.3333 89.4111 888 0.15 0.8815 diet*time D 0 D 12 14.000 89.4111 888 0.16 0.8756 diet*time D 0 D 13 22.8333 89.4111 888 0.26 0.7985 diet*time D 0 D 14 24.1667 89.4111 888 0.27 0.7870 diet*time D 0 D 15 85.8333 89.4111 888 0.96 0.3373 diet*time D 0 D 16 -119.92 93.8830 888 -1.28 0.2018 diet*time D 0 D 17 -59.716 93.8830 888 -0.64 0.5249 diet*time D 0 D 18 -78.1667 89.4111 888 -0.87 0.3822 diet*time D 0 D 19 -20.0000 89.4111 888 -0.22 0.8231 diet*time D 0 D 20 -33.5000 89.4111 888 -0.37 0.7080 diet*time D 0 D 21 -8.3333 89.4111 888 -0.09 0.9258 diet*time D 0 D 22 9.0000 89.4111 888 0.10 0.9198 diet*time D 0 D 23 -55.5000 89.4111 888 -0.62 0.5349 diet*time D 0 D 24 -80.3333 89.4111 888 -0.90 0.3692 diet*time D 0 D 25 -67.3333 89.4111 888 -0.75 0.4516 diet*time D 0 D 26 -111.33 89.4111 888 -1.25 0.2134 diet*time D 0 D 27 -137.00 89.4111 888 -1.53 0.1258 diet*time D 0 D 28 -150.50 89.4111 888 -1.68 0.0927

PAGE 146

130 Table B-4. Continued. Effect diet time _diet _time Es timate SEM DF t Value Pr > |t| diet*time D 0 D 29 -147.17 89.4111 888 -1.65 0.1001 diet*time D 0 D 30 -207.67 89.4111 888 -2.32 0.0204 diet*time D 0 D 31 -167.33 89.4111 888 -1.87 0.0616 diet*time D 0 D 32 -81.8333 89.4111 888 -0.92 0.3603 diet*time D 0 D 33 -105.33 89.4111 888 -1.18 0.2391 diet*time D 0 D 34 -122.33 89.4111 888 -1.37 0.1716 diet*time D 0 D 35 -118.83 89.4111 888 -1.33 0.1842 diet*time D 0 D 36 -99.0000 89.4111 888 -1.11 0.2685 diet*time D 0 D 37 -99.3333 89.4111 888 -1.11 0.2669 diet*time D 0 D 38 -137.00 89.4111 888 -1.53 0.1258 diet*time D 0 D 39 -158.50 89.4111 888 -1.77 0.0766 diet*time D 0 D 40 -155.83 89.4111 888 -1.74 0.0817 diet*time D 0 D 41 -138.12 93.8830 888 -1.47 0.1416 diet*time D 0 D 42 -162.12 93.8830 888 -1.73 0.0846 diet*time D 0 D 43 -203.52 93.8830 888 -2.17 0.0304 diet*time D 0 D 44 -167.92 93.8830 888 -1.79 0.0740 diet*time D 0 D 45 -279.62 100.20 888 -2.79 0.0054 Differences of Least Squares Means Between Baselines Effect diet time _diet _time Estimat e SEM DF t Value Pr > |t| diet*time A 0 B 0 -77.6757 162.41 888 -0.48 0.6326 diet*time A 0 C 0 117.66 162.41 888 0.72 0.4690 diet*time A 0 D 0 -38.509 162.41 888 -0.24 0.8126 diet*time B 0 C 0 195.33 155.99 888 1.25 0.2108 diet*time B 0 D 0 39.166 155.99 888 0.25 0.8018 diet*time C 0 D 0 -156.17 155.99 888 -1.00 0.3170 Table B-5. ANOVA mixed procedure for mean bag volume minus baseline of pelleted meals (Phase I). 1-45: 2-min postprandial blocks. Class Levels Values diet 4 A B C D horse 6 Dusty Gambler Iso Mama Nina Seven time 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Effect NumDF DenDF F Value Pr > F diet 3 20 0.73 0.5445 time 44 864 5.71 <.0001 diet*time 132 864 1.28 0.0239 Differences of Least Squares Means Be tween Diets (Relaxation Comparison) Effect diet time _diet _time Estim ate SEM DF t Value Pr > |t| diet*time C 1 D 1 135.17 143.59 864 0.94 0.3468 diet*time C 2 D 2 -34.6667 143.59 864 -0.24 0.8093 diet*time C 3 D 3 -209.83 143.59 864 -1.46 0.1443

PAGE 147

131 Table B-5. Continued. Effect diet time _diet _time Estim ate SEM DF t Value Pr > |t| diet*time C 4 D 4 -215.00 143.59 864 -1.50 0.1347 diet*time C 5 D 5 -106.17 143.59 864 -0.74 0.4599 diet*time C 6 D 6 12.8333 143.59 864 0.09 0.9288 diet*time C 7 D 7 120.17 143.59 864 0.84 0.4029 diet*time C 8 D 8 138.67 143.59 864 0.97 0.3345 diet*time C 9 D 9 59.5000 143.59 864 0.41 0.6787 diet*time C 10 D 10 23.1507 146.34 864 0.16 0.8743 diet*time C 11 D 11 32.4507 146.34 864 0.22 0.8246 diet*time C 12 D 12 30.5000 143.59 864 0.21 0.8318 diet*time C 13 D 13 9.1667 143.59 864 0.06 0.9491 diet*time C 14 D 14 96.8333 143.59 864 0.67 0.5003 diet*time C 15 D 15 111.67 143.59 864 0.78 0.4370 diet*time C 16 D 16 -85.2047 146.35 864 -0.58 0.5606 diet*time C 17 D 17 10.4286 146.35 864 0.07 0.9432 diet*time C 18 D 18 10.3333 143.59 864 0.07 0.9426 diet*time C 19 D 19 64.0000 143.59 864 0.45 0.6559 diet*time C 20 D 20 46.1667 143.59 864 0.32 0.7479 diet*time C 21 D 21 101.00 143.59 864 0.70 0.4820 diet*time C 22 D 22 117.00 143.59 864 0.81 0.4154 diet*time C 23 D 23 54.0000 143.59 864 0.38 0.7070 diet*time C 24 D 24 74.2613 146.34 864 0.51 0.6120 diet*time C 25 D 25 108.06 146.34 864 0.74 0.4605 diet*time C 26 D 26 46.5000 143.59 864 0.32 0.7461 diet*time C 27 D 27 -11.1667 143.59 864 -0.08 0.9380 diet*time C 28 D 28 -10.5000 143.59 864 -0.07 0.9417 diet*time C 29 D 29 21.5000 143.59 864 0.15 0.8810 diet*time C 30 D 30 -29.5000 143.59 864 -0.21 0.8373 diet*time C 31 D 31 -13.5000 143.59 864 -0.09 0.9251 diet*time C 32 D 32 80.8333 143.59 864 0.56 0.5736 diet*time C 33 D 33 56.6667 143.59 864 0.39 0.6932 diet*time C 34 D 34 65.3333 143.59 864 0.45 0.6492 diet*time C 35 D 35 99.8333 143.59 864 0.70 0.4871 diet*time C 36 D 36 42.5000 143.59 864 0.30 0.7673 diet*time C 37 D 37 116.67 143.59 864 0.81 0.4167 diet*time C 38 D 38 98.0000 143.59 864 0.68 0.4951 diet*time C 39 D 39 148.00 143.59 864 1.03 0.3030 diet*time C 40 D 40 130.33 143.59 864 0.91 0.3643 diet*time C 41 D 41 90.0953 146.35 864 0.62 0.5383 diet*time C 42 D 42 137.53 146.35 864 0.94 0.3476 diet*time C 43 D 43 49.3286 146.35 864 0.34 0.7361 diet*time C 44 D 27 108.00 143.59 864 0.75 0.4522 diet*time C 45 D 45 22.8620 146.35 864 0.16 0.8759

PAGE 148

132 Table B-5. Continued. Differences of Least Squares Mean s Between Diets (Labeling effect) Effect diet time _diet _time Es timate SEM DF t Value Pr > |t| diet*time A 1 C 1 43.9813 146.34 864 0.30 0.7638 diet*time A 2 C 2 -5.3333 143.59 864 -0.04 0.9704 diet*time A 3 C 3 86.3333 143.59 864 0.60 0.5478 diet*time A 4 C 4 166.17 143.59 864 1.16 0.2475 diet*time A 5 C 5 97.3333 143.59 864 0.68 0.4981 diet*time A 6 C 6 73.6667 143.59 864 0.51 0.6081 diet*time A 7 C 7 -8.3333 143.59 864 -0.06 0.9537 diet*time A 8 C 8 -7.3333 143.59 864 -0.05 0.9593 diet*time A 9 C 9 -3.1667 143.59 864 -0.02 0.9824 diet*time A 10 C 10 16.6827 146.34 864 0.11 0.9093 diet*time A 11 C 11 11.5493 146.34 864 0.08 0.9371 diet*time A 12 C 12 2.5000 143.59 864 0.02 0.9861 diet*time A 13 C 13 -29.833 143.59 864 -0.21 0.8355 diet*time A 14 C 14 -60.8333 143.59 864 -0.42 0.6719 diet*time A 15 C 15 26.0000 143.59 864 0.18 0.8564 diet*time A 16 C 16 24.5000 143.59 864 0.17 0.8646 diet*time A 17 C 17 -33.5000 143.59 864 -0.23 0.8156 diet*time A 18 C 18 -124.17 143.59 864 -0.86 0.3874 diet*time A 19 C 19 -112.00 143.59 864 -0.78 0.4356 diet*time A 20 C 20 -69.1667 143.59 864 -0.48 0.6301 diet*time A 21 C 21 -154.67 143.59 864 -1.08 0.2817 diet*time A 22 C 22 -160.00 143.59 864 -1.11 0.2655 diet*time A 23 C 23 -106.00 143.59 864 -0.74 0.4606 diet*time A 24 C 24 -65.6667 143.59 864 -0.46 0.6476 diet*time A 25 C 25 -166.00 143.59 864 -1.16 0.2480 diet*time A 26 C 26 -180.83 143.59 864 -1.26 0.2082 diet*time A 27 C 27 -108.33 143.59 864 -0.75 0.4508 diet*time A 28 C 28 -98.500 143.59 864 -0.69 0.4929 diet*time A 29 C 29 -165.50 143.59 864 -1.15 0.2494 diet*time A 30 C 30 -165.50 143.59 864 -1.15 0.2494 diet*time A 31 C 31 -52.1667 143.59 864 -0.36 0.7165 diet*time A 32 C 32 -82.1667 143.59 864 -0.57 0.5673 diet*time A 33 C 33 18.5000 143.59 864 0.13 0.8975 diet*time A 34 C 34 -54.6667 143.59 864 -0.38 0.7035 diet*time A 35 C 35 -97.3333 143.59 864 -0.68 0.4981 diet*time A 36 C 36 -64.5000 143.59 864 -0.45 0.6534 diet*time A 37 C 37 -117.00 143.59 864 -0.81 0.4154 diet*time A 38 C 38 -99.6667 143.59 864 -0.69 0.4878 diet*time A 39 C 39 -179.00 143.59 864 -1.25 0.2129 diet*time A 40 C 40 -79.5000 143.59 864 -0.55 0.5800 diet*time A 41 C 41 -123.00 143.59 864 -0.86 0.3919

PAGE 149

133 Table B-5. Continued. Effect diet time _diet _time Es timate SEM DF t Value Pr > |t| diet*time A 42 C 42 -71.1667 143.59 864 -0.50 0.6203 diet*time A 43 C 43 -41.0000 143.59 864 -0.29 0.7753 diet*time A 44 C 44 -72.3333 143.59 864 -0.50 0.6146 diet*time A 45 C 45 -75.8333 143.59 864 -0.53 0.5976 diet*time B 1 D 1 123.67 143.59 864 0.86 0.3893 diet*time B 2 D 2 160.50 143.59 864 1.12 0.2640 diet*time B 3 D 3 138.00 143.59 864 0.96 0.3368 diet*time B 4 D 4 102.50 143.59 864 0.71 0.4755 diet*time B 5 D 5 137.33 143.59 864 0.96 0.3391 diet*time B 6 D 6 142.67 143.59 864 0.99 0.3207 diet*time B 7 D 7 349.83 143.59 864 2.44 0.0150 diet*time B 8 D 8 382.67 143.59 864 2.66 0.0078 diet*time B 9 D 9 248.17 143.59 864 1.73 0.0843 diet*time B 10 D 10 327.67 143.59 864 2.28 0.0227 diet*time B 11 D 11 274.00 143.59 864 1.91 0.0567 diet*time B 12 D 12 236.00 143.59 864 1.64 0.1006 diet*time B 13 D 13 248.83 143.59 864 1.73 0.0835 diet*time B 14 D 14 324.17 143.59 864 2.26 0.0242 diet*time B 15 D 15 328.83 143.59 864 2.29 0.0223 diet*time B 16 D 16 214.80 146.35 864 1.47 0.1425 diet*time B 17 D 17 346.26 146.35 864 2.37 0.0182 diet*time B 18 D 18 184.17 143.59 864 1.28 0.2000 diet*time B 19 D 19 300.33 143.59 864 2.09 0.0368 diet*time B 20 D 20 198.67 143.59 864 1.38 0.1669 diet*time B 21 D 21 167.00 143.59 864 1.16 0.2451 diet*time B 22 D 22 124.50 143.59 864 0.87 0.3862 diet*time B 23 D 23 62.500 143.59 864 0.44 0.6635 diet*time B 24 D 24 28.0946 146.34 864 0.19 0.8478 diet*time B 25 D 25 124.89 146.34 864 0.85 0.3936 diet*time B 26 D 26 91.8333 143.59 864 0.64 0.5226 diet*time B 27 D 27 55.3333 143.59 864 0.39 0.7001 diet*time B 28 D 28 -79.2652 146.34 864 -0.54 0.5882 diet*time B 29 D 29 24.8348 146.34 864 0.17 0.8653 diet*time B 30 D 30 -83.9985 146.34 864 -0.57 0.5661 diet*time B 31 D 31 5.3348 146.34 864 0.04 0.9709 diet*time B 32 D 32 49.0000 143.59 864 0.34 0.7330 diet*time B 33 D 33 37.0000 143.59 864 0.26 0.7967 diet*time B 34 D 34 80.6667 143.59 864 0.56 0.5744 diet*time B 35 D 35 114.50 143.59 864 0.80 0.4254 diet*time B 36 D 36 136.33 143.59 864 0.95 0.3427 diet*time B 37 D 37 121.83 143.59 864 0.85 0.3964 diet*time B 38 D 38 113.00 143.59 864 0.79 0.4315

PAGE 150

134 Table B-5. Continued. Effect diet time _diet _time Es timate SEM DF t Value Pr > |t| diet*time B 39 D 39 85.3333 143.59 864 0.59 0.5525 diet*time B 40 D 40 105.67 143.59 864 0.74 0.4620 diet*time B 41 D 41 75.9286 146.35 864 0.52 0.6040 diet*time B 42 D 42 110.03 146.35 864 0.75 0.4524 diet*time B 43 D 43 12.3286 146.35 864 0.08 0.9329 diet*time B 44 D 44 61.0286 146.35 864 0.42 0.6768 diet*time B 45 D 45 -90.8047 146.35 864 -0.62 0.5351 Table B-6. ANOVA and CL for i ngestion times of unlabeled sw eet feed meals (Phase II). Class Levels Values horse 6 DUSTY GAMBLER ISO MAMA NINA SEVEN diet 2 FAT CHO Source DF Sum of Squares Mean Square F Value Pr > F Model 6 55715.16667 9285.86111 3.88 0.0791 Error 5 11957.75000 2391.55000 Total 11 67672.91667 Source DF Anova SS Mean Square F Value Pr > F horse 5 53608.41667 10721.68333 4.48 0.0626 diet 1 2106.75000 2106.75000 0.88 0.3910 Variable Lower 95% CL Upper 95% CL FAT 160.9242883 320.4090450 CHO 177.2614066 357.0719268 Table B-7. Shapiro-Wilk test for normality of ingestion times of unlabeled sweet feed meals (Phase II). Source Statistic (W) p-value Diets Difference 0.983089 0.9658 Table B-8. Bartlett's test for homogeneity of variance of unlabeled sweet feed meals (Phase II). Source DF Chi-Square p-value Diet 1 0.00587 0.9389

PAGE 151

135 Table B-9. ANOVA mixed procedure for mean bag volumes of unlabeled sweet feed meals (Phase II). 0: baseline; 1-45: 2-min postprandial blocks. Class Levels Values diet 2 CHO FAT horse 6 Dusty Gambler Iso Mama Nina Seven time 46 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Effect NumDF DenDF F Value Pr > F diet 1 10 1.29 0.2827 time 45 450 5.82 <.0001 diet*time 45 450 0.74 0.8884 Differences of Least Squares Means Within Diet (Relaxation Response) Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time CHO 0 CHO 1 -335.67 120.91 450 -2.78 0.0057 diet*time CHO 0 CHO 2 -770.50 120.91 450 -6.37 <.0001 diet*time CHO 0 CHO 3 -880.67 120.91 450 -7.28 <.0001 diet*time CHO 0 CHO 4 -868.67 120.91 450 -7.18 <.0001 diet*time CHO 0 CHO 5 -944.83 120.91 450 -7.81 <.0001 diet*time CHO 0 CHO 6 -958.50 120.91 450 -7.93 <.0001 diet*time CHO 0 CHO 7 -897.67 120.91 450 -7.42 <.0001 diet*time CHO 0 CHO 8 -866.67 120.91 450 -7.17 <.0001 diet*time CHO 0 CHO 9 -862.67 120.91 450 -7.13 <.0001 diet*time CHO 0 CHO 10 -795.33 120.91 450 -6.58 <.0001 diet*time CHO 0 CHO 11 -719.50 120.91 450 -5.95 <.0001 diet*time CHO 0 CHO 12 -726.00 120.91 450 -6.00 <.0001 diet*time CHO 0 CHO 13 -605.00 120.91 450 -5.00 <.0001 diet*time CHO 0 CHO 14 -642.67 120.91 450 -5.32 <.0001 diet*time CHO 0 CHO 15 -687.00 120.91 450 -5.68 <.0001 diet*time CHO 0 CHO 16 -614.17 120.91 450 -5.08 <.0001 diet*time CHO 0 CHO 17 -568.83 120.91 450 -4.70 <.0001 diet*time CHO 0 CHO 18 -563.67 120.91 450 -4.66 <.0001 diet*time CHO 0 CHO 19 -530.50 120.91 450 -4.39 <.0001 diet*time CHO 0 CHO 20 -531.33 120.91 450 -4.39 <.0001 diet*time CHO 0 CHO 21 -598.17 120.91 450 -4.95 <.0001 diet*time CHO 0 CHO 22 -654.00 120.91 450 -5.41 <.0001 diet*time CHO 0 CHO 23 -665.67 120.91 450 -5.51 <.0001 diet*time CHO 0 CHO 24 -584.67 120.91 450 -4.84 <.0001 diet*time CHO 0 CHO 25 -593.50 120.91 450 -4.91 <.0001 diet*time CHO 0 CHO 26 -522.50 120.91 450 -4.32 <.0001 diet*time CHO 0 CHO 27 -528.50 120.91 450 -4.37 <.0001 diet*time CHO 0 CHO 28 -674.83 120.91 450 -5.58 <.0001 diet*time CHO 0 CHO 29 -532.17 120.91 450 -4.40 <.0001 diet*time CHO 0 CHO 30 -547.83 120.91 450 -4.53 <.0001 diet*time CHO 0 CHO 31 -613.67 120.91 450 -5.08 <.0001

PAGE 152

136 Table B-9. Continued. Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time CHO 0 CHO 32 -586.83 120.91 450 -4.85 <.0001 diet*time CHO 0 CHO 33 -652.83 120.91 450 -5.40 <.0001 diet*time CHO 0 CHO 34 -648.33 120.91 450 -5.36 <.0001 diet*time CHO 0 CHO 35 -622.17 120.91 450 -5.15 <.0001 diet*time CHO 0 CHO 36 -609.83 120.91 450 -5.04 <.0001 diet*time CHO 0 CHO 37 -496.50 120.91 450 -4.11 <.0001 diet*time CHO 0 CHO 38 -495.00 120.91 450 -4.09 <.0001 diet*time CHO 0 CHO 39 -464.33 120.91 450 -3.84 0.0001 diet*time CHO 0 CHO 40 -528.67 120.91 450 -4.37 <.0001 diet*time CHO 0 CHO 41 -470.33 120.91 450 -3.89 0.0001 diet*time CHO 0 CHO 42 -579.50 120.91 450 -4.79 <.0001 diet*time CHO 0 CHO 43 -607.33 120.91 450 -5.02 <.0001 diet*time CHO 0 CHO 44 -558.33 120.91 450 -4.62 <.0001 diet*time CHO 0 CHO 45 -579.17 120.91 450 -4.79 <.0001 diet*time FAT 0 FAT 1 -110.50 120.91 450 -0.91 0.3612 diet*time FAT 0 FAT 2 -585.33 120.91 450 -4.84 <.0001 diet*time FAT 0 FAT 3 -791.67 120.91 450 -6.55 <.0001 diet*time FAT 0 FAT 4 -693.83 120.91 450 -5.74 <.0001 diet*time FAT 0 FAT 5 -648.33 120.91 450 -5.36 <.0001 diet*time FAT 0 FAT 6 -600.50 120.91 450 -4.97 <.0001 diet*time FAT 0 FAT 7 -457.83 120.91 450 -3.79 0.0002 diet*time FAT 0 FAT 8 -426.50 120.91 450 -3.53 0.0005 diet*time FAT 0 FAT 9 -555.00 120.91 450 -4.59 <.0001 diet*time FAT 0 FA T 10 -382.00 120.91 450 -3.16 0.0017 diet*time FAT 0 FA T 11 -489.33 120.91 450 -4.05 <.0001 diet*time FAT 0 FA T 12 -490.33 120.91 450 -4.06 <.0001 diet*time FAT 0 FA T 13 -484.00 120.91 450 -4.00 <.0001 diet*time FAT 0 FA T 14 -449.00 120.91 450 -3.71 0.0002 diet*time FAT 0 FA T 15 -467.00 120.91 450 -3.86 0.0001 diet*time FAT 0 FA T 16 -519.67 120.91 450 -4.30 <.0001 diet*time FAT 0 FA T 17 -522.33 120.91 450 -4.32 <.0001 diet*time FAT 0 FA T 18 -475.00 120.91 450 -3.93 <.0001 diet*time FAT 0 FA T 19 -479.83 120.91 450 -3.97 <.0001 diet*time FAT 0 FA T 20 -490.50 120.91 450 -4.06 <.0001 diet*time FAT 0 FA T 21 -451.67 120.91 450 -3.74 0.0002 diet*time FAT 0 FA T 22 -378.67 120.91 450 -3.13 0.0019 diet*time FAT 0 FA T 23 -448.67 120.91 450 -3.71 0.0002 diet*time FAT 0 FA T 24 -378.33 120.91 450 -3.13 0.0019 diet*time FAT 0 FA T 25 -363.67 120.91 450 -3.01 0.0028 diet*time FAT 0 FA T 26 -290.50 120.91 450 -2.40 0.0167 diet*time FAT 0 FAT 27 -283.83 120.91 450 -2.35 0.0193 diet*time FAT 0 FAT 28 -358.17 120.91 450 -2.96 0.0032

PAGE 153

137 Table B-9. Continued. Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time FAT 0 FAT 29 -327.17 120.91 450 -2.71 0.0071 diet*time FAT 0 FAT 30 -322.00 120.91 450 -2.66 0.0080 diet*time FAT 0 FAT 31 -310.67 120.91 450 -2.57 0.0105 diet*time FAT 0 FAT 32 -374.33 120.91 450 -3.10 0.0021 diet*time FAT 0 FAT 33 -438.17 120.91 450 -3.62 0.0003 diet*time FAT 0 FAT 34 -406.00 120.91 450 -3.36 0.0009 diet*time FAT 0 FAT 35 -380.67 120.91 450 -3.15 0.0018 diet*time FAT 0 FAT 36 -318.83 120.91 450 -2.64 0.0087 diet*time FAT 0 FAT 37 -273.33 120.91 450 -2.26 0.0243 diet*time FAT 0 FAT 38 -399.67 120.91 450 -3.31 0.0010 diet*time FAT 0 FAT 39 -332.83 120.91 450 -2.75 0.0061 diet*time FAT 0 FAT 40 -373.33 120.91 450 -3.09 0.0021 diet*time FAT 0 FAT 41 -323.67 120.91 450 -2.68 0.0077 diet*time FAT 0 FAT 42 -339.17 120.91 450 -2.81 0.0052 diet*time FAT 0 FAT 43 -255.67 120.91 450 -2.11 0.0350 diet*time FAT 0 FAT 44 -209.67 120.91 450 -1.73 0.0836 diet*time FAT 0 FAT 45 -217.00 120.91 450 -1.79 0.0734 Differences of Least Squares Means Between Baselines Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time CHO 0 FAT 0 1.0000 228.28 450 0.00 0.9965 Table B-10. ANOVA mixed procedure for mean bag volume minus baseline of unlabelled sweet feed meals (Phase I I). 1-45: 2-min postprandial blocks. Class Levels Values diet 2 CHO FAT horse 6 Dusty Gambler Iso Mama Nina Seven time 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Effect NumDF DenDF F Value Pr > F diet 1 10 1.59 0.2358 time 44 440 4.37 <.0001 diet*time 44 440 0.70 0.9250 Differences of Least Squares Means Be tween Diets (Relaxation Comparison) Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time CHO 1 FAT 1 225.00 213.72 440 1.05 0.2930 diet*time CHO 2 FAT 2 184.83 213.72 440 0.86 0.3876 diet*time CHO 3 FAT 3 88.8333 213.72 440 0.42 0.6779 diet*time CHO 4 FAT 4 175.00 213.72 440 0.82 0.4133 diet*time CHO 5 FAT 5 296.33 213.72 440 1.39 0.1663 diet*time CHO 6 FAT 6 358.00 213.72 440 1.68 0.0946 diet*time CHO 7 FAT 7 439.83 213.72 440 2.06 0.0402 diet*time CHO 8 FAT 8 440.50 213.72 440 2.06 0.0399

PAGE 154

138 Table B-10. Continued. Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time CHO 9 FAT 9 307.33 213.72 440 1.44 0.1511 diet*time CHO 10 FAT 10 413.33 213.72 440 1.93 0.0538 diet*time CHO 11 FAT 11 230.33 213.72 440 1.08 0.2817 diet*time CHO 12 FAT 12 235.83 213.72 440 1.10 0.2704 diet*time CHO 13 FAT 13 121.17 213.72 440 0.57 0.5710 diet*time CHO 14 FAT 14 193.83 213.72 440 0.91 0.3649 diet*time CHO 15 FAT 15 220.50 213.72 440 1.03 0.3028 diet*time CHO 16 FAT 16 94.500 213.72 440 0.44 0.6586 diet*time CHO 17 FAT 17 46.5000 213.72 440 0.22 0.8279 diet*time CHO 18 FAT 18 88.833 213.72 440 0.42 0.6779 diet*time CHO 19 FAT 19 50.833 213.72 440 0.24 0.8121 diet*time CHO 20 FAT 20 41.000 213.72 440 0.19 0.8480 diet*time CHO 21 FAT 21 146.67 213.72 440 0.69 0.4929 diet*time CHO 22 FAT 22 275.33 213.72 440 1.29 0.1983 diet*time CHO 23 FAT 23 217.17 213.72 440 1.02 0.3101 diet*time CHO 24 FAT 24 206.17 213.72 440 0.96 0.3352 diet*time CHO 25 FAT 25 230.00 213.72 440 1.08 0.2824 diet*time CHO 26 FAT 26 232.00 213.72 440 1.09 0.2783 diet*time CHO 27 FAT 27 244.83 213.72 440 1.15 0.2526 diet*time CHO 28 FAT 28 316.67 213.72 440 1.48 0.1391 diet*time CHO 29 FAT 29 204.83 213.72 440 0.96 0.3384 diet*time CHO 30 FAT 30 225.50 213.72 440 1.06 0.2919 diet*time CHO 31 FAT 31 303.17 213.72 440 1.42 0.1567 diet*time CHO 32 FAT 32 212.67 213.72 440 1.00 0.3202 diet*time CHO 33 FAT 33 214.83 213.72 440 1.01 0.3153 diet*time CHO 34 FAT 34 242.33 213.72 440 1.13 0.2575 diet*time CHO 35 FAT 35 241.33 213.72 440 1.13 0.2594 diet*time CHO 36 FAT 36 290.67 213.72 440 1.36 0.1745 diet*time CHO 37 FAT 37 223.17 213.72 440 1.04 0.2970 diet*time CHO 38 FAT 38 95.833 213.72 440 0.45 0.6541 diet*time CHO 39 FAT 39 131.33 213.72 440 0.61 0.5392 diet*time CHO 40 FAT 40 155.67 213.72 440 0.73 0.4668 diet*time CHO 41 FAT 41 146.67 213.72 440 0.69 0.4929 diet*time CHO 42 FAT 42 241.17 213.72 440 1.13 0.2598 diet*time CHO 43 FAT 43 352.00 213.72 440 1.65 0.1003 diet*time CHO 44 FAT 44 348.50 213.72 440 1.63 0.1037 diet*time CHO 45 FAT 45 362.50 213.72 440 1.70 0.0906

PAGE 155

139 Table B-11. Bartlett's test for homogeneity of variance for ingestion times of labeled sweet feed meals (Phase II). Source DF Chi-Square p-value. Diet 2 27.2209 <.0001 Table B-12. Friedmans 2-way ANOVA for ingestion times of labeled sweet feed meals. Cochran-Mantel-Haenszel Sta tistics (Based on Rank Scores) Statistic Alternative H ypothesis DF Value Prob. 1 Nonzero Correlation 1 0.7826 0.3763 2 Row Mean Scores Differ 2 4.9565 0.0839 Table B-13. ANOVA mixed procedure for mean bag volumes of labeled sweet feed meals (Phase II). 0: baseline; 1-45: 2-min postprandial blocks. Class Levels Values diet 3 CHO CON FAT horse 6 Dusty Gambler Iso Mama Nina Seven time 46 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Effect NumDF DenDF F Value Pr > F diet 2 15 4.21 0.0355 time 45 675 14.24 <.0001 diet*time 90 675 1.00 0.4821 Differences of Least Squares Means Within Diet (Relaxation Response) Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time CHO 0 CHO 1 -355.67 117.69 675 -3.02 0.0026 diet*time CHO 0 CHO 2 -852.50 117.69 675 -7.24 <.0001 diet*time CHO 0 CHO 3 -1000.8 117.69 675 -8.50 <.0001 diet*time CHO 0 CHO 4 -989.83 117.69 675 -8.41 <.0001 diet*time CHO 0 CHO 5 -964.67 117.69 675 -8.20 <.0001 diet*time CHO 0 CHO 6 -924.67 117.69 675 -7.86 <.0001 diet*time CHO 0 CHO 7 -852.50 117.69 675 -7.24 <.0001 diet*time CHO 0 CHO 8 -781.50 117.69 675 -6.64 <.0001 diet*time CHO 0 CHO 9 -838.67 117.69 675 -7.13 <.0001 diet*time CHO 0 CHO 10 -782.00 117.69 675 -6.64 <.0001 diet*time CHO 0 CHO 11 -734.67 117.69 675 -6.24 <.0001 diet*time CHO 0 CHO 12 -728.83 117.69 675 -6.19 <.0001 diet*time CHO 0 CHO 13 -613.17 117.69 675 -5.21 <.0001 diet*time CHO 0 CHO 14 -744.67 117.69 675 -6.33 <.0001 diet*time CHO 0 CHO 15 -626.33 117.69 675 -5.32 <.0001 diet*time CHO 0 CHO 16 -580.83 117.69 675 -4.94 <.0001 diet*time CHO 0 CHO 17 -579.83 117.69 675 -4.93 <.0001 diet*time CHO 0 CHO 18 -541.67 117.69 675 -4.60 <.0001 diet*time CHO 0 CHO 19 -643.50 117.69 675 -5.47 <.0001 diet*time CHO 0 CHO 20 -573.50 117.69 675 -4.87 <.0001

PAGE 156

140 Table B-13. Continued. Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time CHO 0 CHO 21 -569.83 117.69 675 -4.84 <.0001 diet*time CHO 0 CHO 22 -572.33 117.69 675 -4.86 <.0001 diet*time CHO 0 CHO 23 -472.00 117.69 675 -4.01 <.0001 diet*time CHO 0 CHO 24 -537.33 117.69 675 -4.57 <.0001 diet*time CHO 0 CHO 25 -602.50 117.69 675 -5.12 <.0001 diet*time CHO 0 CHO 26 -639.17 117.69 675 -5.43 <.0001 diet*time CHO 0 CHO 27 -496.83 117.69 675 -4.22 <.0001 diet*time CHO 0 CHO 28 -513.00 117.69 675 -4.36 <.0001 diet*time CHO 0 CHO 29 -464.50 117.69 675 -3.95 <.0001 diet*time CHO 0 CHO 30 -538.17 117.69 675 -4.57 <.0001 diet*time CHO 0 CHO 31 -551.67 117.69 675 -4.69 <.0001 diet*time CHO 0 CHO 32 -446.50 117.69 675 -3.79 0.0002 diet*time CHO 0 CHO 33 -488.33 117.69 675 -4.15 <.0001 diet*time CHO 0 CHO 34 -483.00 117.69 675 -4.10 <.0001 diet*time CHO 0 CHO 35 -452.17 117.69 675 -3.84 0.0001 diet*time CHO 0 CHO 36 -316.17 117.69 675 -2.69 0.0074 diet*time CHO 0 CHO 37 -414.33 117.69 675 -3.52 0.0005 diet*time CHO 0 CHO 38 -406.00 117.69 675 -3.45 0.0006 diet*time CHO 0 CHO 39 -476.67 117.69 675 -4.05 <.0001 diet*time CHO 0 CHO 40 -360.17 117.69 675 -3.06 0.0023 diet*time CHO 0 CHO 41 -394.00 117.69 675 -3.35 0.0009 diet*time CHO 0 CHO 42 -325.00 117.69 675 -2.76 0.0059 diet*time CHO 0 CHO 43 -430.83 117.69 675 -3.66 0.0003 diet*time CHO 0 CHO 44 -319.00 117.69 675 -2.71 0.0069 diet*time CHO 0 CHO 45 -327.50 117.69 675 -2.78 0.0055 diet*time CON 0 CON 1 -328.83 117.69 675 -2.79 0.0054 diet*time CON 0 CON 2 -707.50 117.69 675 -6.01 <.0001 diet*time CON 0 CON 3 -752.33 117.69 675 -6.39 <.0001 diet*time CON 0 CON 4 -620.50 117.69 675 -5.27 <.0001 diet*time CON 0 CON 5 -499.33 117.69 675 -4.24 <.0001 diet*time CON 0 CON 6 -305.33 117.69 675 -2.59 0.0097 diet*time CON 0 CON 7 -234.83 117.69 675 -2.00 0.0464 diet*time CON 0 CON 8 -239.17 117.69 675 -2.03 0.0425 diet*time CON 0 CON 9 -387.67 117.69 675 -3.29 0.0010 diet*time CON 0 CON 10 -281.50 117.69 675 -2.39 0.0170 diet*time CON 0 CON 11 -391.50 117.69 675 -3.33 0.0009 diet*time CON 0 CON 12 -255.67 117.69 675 -2.17 0.0302 diet*time CON 0 CON 13 -392.67 117.69 675 -3.34 0.0009 diet*time CON 0 CON 14 -550.67 117.69 675 -4.68 <.0001 diet*time CON 0 CON 15 -578.17 117.69 675 -4.91 <.0001 diet*time CON 0 CON 16 -363.67 117.69 675 -3.09 0.0021 diet*time CON 0 CON 17 -394.83 117.69 675 -3.35 0.0008

PAGE 157

141 Table B-13. Continued. Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time CON 0 CON 18 -436.83 117.69 675 -3.71 0.0002 diet*time CON 0 CON 19 -345.00 117.69 675 -2.93 0.0035 diet*time CON 0 CON 20 -224.17 117.69 675 -1.90 0.0572 diet*time CON 0 CON 21 -267.33 117.69 675 -2.27 0.0234 diet*time CON 0 CON 22 -225.50 117.69 675 -1.92 0.0558 diet*time CON 0 CON 23 -111.33 117.69 675 -0.95 0.3445 diet*time CON 0 CON 24 -227.50 117.69 675 -1.93 0.0537 diet*time CON 0 CON 25 -171.83 117.69 675 -1.46 0.1448 diet*time CON 0 CON 26 -172.17 117.69 675 -1.46 0.1440 diet*time CON 0 CON 27 -109.50 117.69 675 -0.93 0.3525 diet*time CON 0 CON 28 -91.6667 117.69 675 -0.78 0.4363 diet*time CON 0 CON 29 -168.17 117.69 675 -1.43 0.1535 diet*time CON 0 CON 30 -69.8333 117.69 675 -0.59 0.5531 diet*time CON 0 CON 31 -110.50 117.69 675 -0.94 0.3481 diet*time CON 0 CON 32 -202.67 117.69 675 -1.72 0.0855 diet*time CON 0 CON 33 -159.50 117.69 675 -1.36 0.1758 diet*time CON 0 CON 34 -165.33 117.69 675 -1.40 0.1605 diet*time CON 0 CON 35 -183.67 117.69 675 -1.56 0.1191 diet*time CON 0 CON 36 -112.17 117.69 675 -0.95 0.3409 diet*time CON 0 CON 37 -78.8333 117.69 675 -0.67 0.5032 diet*time CON 0 CON 38 -100.00 117.69 675 -0.85 0.3958 diet*time CON 0 CON 39 -67.5000 117.69 675 -0.57 0.5665 diet*time CON 0 CON 40 -26.0000 117.69 675 -0.22 0.8252 diet*time CON 0 CON 41 -71.1667 117.69 675 -0.60 0.5456 diet*time CON 0 CON 42 -124.67 117.69 675 -1.06 0.2899 diet*time CON 0 CON 43 -84.1667 117.69 675 -0.72 0.4748 diet*time CON 0 CON 44 -76.1667 117.69 675 -0.65 0.5177 diet*time CON 0 CON 45 -72.5000 117.69 675 -0.62 0.5381 diet*time FAT 0 FAT 1 -320.50 117.69 675 -2.72 0.0066 diet*time FAT 0 FAT 2 -724.83 117.69 675 -6.16 <.0001 diet*time FAT 0 FAT 3 -925.33 117.69 675 -7.86 <.0001 diet*time FAT 0 FAT 4 -917.50 117.69 675 -7.80 <.0001 diet*time FAT 0 FAT 5 -880.83 117.69 675 -7.48 <.0001 diet*time FAT 0 FAT 6 -711.33 117.69 675 -6.04 <.0001 diet*time FAT 0 FAT 7 -621.33 117.69 675 -5.28 <.0001 diet*time FAT 0 FAT 8 -595.17 117.69 675 -5.06 <.0001 diet*time FAT 0 FAT 9 -661.67 117.69 675 -5.62 <.0001 diet*time FAT 0 FA T 10 -627.17 117.69 675 -5.33 <.0001 diet*time FAT 0 FA T 11 -557.00 117.69 675 -4.73 <.0001 diet*time FAT 0 FA T 12 -572.17 117.69 675 -4.86 <.0001 diet*time FAT 0 FA T 13 -488.50 117.69 675 -4.15 <.0001 diet*time FAT 0 FAT 14 -668.17 117.69 675 -5.68 <.0001

PAGE 158

142 Table B-13. Continued. Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time FAT 0 FAT 15 -533.50 117.69 675 -4.53 <.0001 diet*time FAT 0 FAT 16 -493.17 117.69 675 -4.19 <.0001 diet*time FAT 0 FAT 17 -527.50 117.69 675 -4.48 <.0001 diet*time FAT 0 FAT 18 -355.83 117.69 675 -3.02 0.0026 diet*time FAT 0 FAT 19 -433.83 117.69 675 -3.69 0.0002 diet*time FAT 0 FAT 20 -439.00 117.69 675 -3.73 0.0002 diet*time FAT 0 FAT 21 -356.83 117.69 675 -3.03 0.0025 diet*time FAT 0 FAT 22 -294.67 117.69 675 -2.50 0.0125 diet*time FAT 0 FAT 23 -286.50 117.69 675 -2.43 0.0152 diet*time FAT 0 FAT 24 -288.00 117.69 675 -2.45 0.0147 diet*time FAT 0 FAT 25 -277.83 117.69 675 -2.36 0.0185 diet*time FAT 0 FAT 26 -326.33 117.69 675 -2.77 0.0057 diet*time FAT 0 FAT 27 -264.67 117.69 675 -2.25 0.0248 diet*time FAT 0 FAT 28 -265.83 117.69 675 -2.26 0.0242 diet*time FAT 0 FAT 29 -381.50 117.69 675 -3.24 0.0012 diet*time FAT 0 FAT 15 -533.50 117.69 675 -4.53 <.0001 diet*time FAT 0 FAT 16 -493.17 117.69 675 -4.19 <.0001 diet*time FAT 0 FAT 17 -527.50 117.69 675 -4.48 <.0001 diet*time FAT 0 FAT 18 -355.83 117.69 675 -3.02 0.0026 diet*time FAT 0 FAT 19 -433.83 117.69 675 -3.69 0.0002 diet*time FAT 0 FAT 20 -439.00 117.69 675 -3.73 0.0002 diet*time FAT 0 FAT 30 -295.67 117.69 675 -2.51 0.0122 diet*time FAT 0 FAT 31 -331.17 117.69 675 -2.81 0.0050 diet*time FAT 0 FAT 32 -422.33 117.69 675 -3.59 0.0004 diet*time FAT 0 FAT 33 -397.17 117.69 675 -3.37 0.0008 diet*time FAT 0 FAT 34 -348.00 117.69 675 -2.96 0.0032 diet*time FAT 0 FAT 35 -332.17 117.69 675 -2.82 0.0049 diet*time FAT 0 FAT 36 -335.33 117.69 675 -2.85 0.0045 diet*time FAT 0 FAT 37 -306.00 117.69 675 -2.60 0.0095 diet*time FAT 0 FAT 38 -293.33 117.69 675 -2.49 0.0129 diet*time FAT 0 FAT 39 -313.00 117.69 675 -2.66 0.0080 diet*time FAT 0 FAT 40 -392.17 117.69 675 -3.33 0.0009 diet*time FAT 0 FAT 41 -418.67 117.69 675 -3.56 0.0004 diet*time FAT 0 FAT 42 -332.67 117.69 675 -2.83 0.0048 diet*time FAT 0 FAT 43 -317.17 117.69 675 -2.69 0.0072 diet*time FAT 0 FAT 44 -367.00 117.69 675 -3.12 0.0019 diet*time FAT 0 FAT 45 -290.00 117.69 675 -2.46 0.0140 Differences of Least Squares Means Between Baselines Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time CHO 0 CON 0 -20.333 156.12 675 -0.13 0.8964 diet*time CHO 0 FAT 0 -2.1667 156.12 675 -0.01 0.9889 diet*time CON 0 FAT 0 18.1667 156.12 675 0.12 0.9074

PAGE 159

143 Table B-14. ANOVA mixed procedure for mean bag volume minus baseline of labelled sweet feed meals (Phase II). 1-45: 2-min postprandial blocks. Class Levels Values diet 3 CHO CON FAT horse 6 Dusty Gambler Iso Mama Nina Seven time 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Effect NumDF DenDF F Value Pr > F diet 2 15 2.94 0.0838 time 44 660 12.72 <.0001 diet*time 88 660 0.93 0.6508 Differences of Least Squares Means Be tween Diets (Relaxation Comparison) Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time CHO 1 CON 1 -19.6667 164.63 660 -0.12 0.9049 diet*time CHO 1 FAT 1 35.3333 164.63 660 0.21 0.8301 diet*time CHO 2 CON 2 98.3333 164.63 660 0.60 0.5505 diet*time CHO 2 FAT 2 127.83 164.63 660 0.78 0.4377 diet*time CHO 3 CON 3 201.50 164.63 660 1.22 0.2214 diet*time CHO 3 FAT 3 75.1667 164.63 660 0.46 0.6481 diet*time CHO 4 CON 4 322.83 164.63 660 1.96 0.0503 diet*time CHO 4 FAT 4 72.1667 164.63 660 0.44 0.6613 diet*time CHO 5 CON 5 418.83 164.63 660 2.54 0.0112 diet*time CHO 5 FAT 5 84.3333 164.63 660 0.51 0.6086 diet*time CHO 6 CON 6 573.00 164.63 660 3.48 0.0005 diet*time CHO 6 FAT 6 213.33 164.63 660 1.30 0.1955 diet*time CHO 7 CON 7 571.33 164.63 660 3.47 0.0006 diet*time CHO 7 FAT 7 231.17 164.63 660 1.40 0.1607 diet*time CHO 8 CON 8 496.00 164.63 660 3.01 0.0027 diet*time CHO 8 FAT 8 186.33 164.63 660 1.13 0.2581 diet*time CHO 9 CON 9 404.33 164.63 660 2.46 0.0143 diet*time CHO 9 FAT 9 177.17 164.63 660 1.08 0.2823 diet*time CHO 10 CON 10 453.83 164.63 660 2.76 0.0060 diet*time CHO 10 FAT 10 154.83 164.63 660 0.94 0.3473 diet*time CHO 14 FAT 14 76.8333 164.63 660 0.47 0.6409 diet*time CHO 15 CON 15 1.3333 164.63 660 0.01 0.9935 diet*time CHO 15 FAT 15 92.3333 164.63 660 0.56 0.5751 diet*time CHO 16 CON 16 170.83 164.63 660 1.04 0.2998 diet*time CHO 16 FAT 16 87.8333 164.63 660 0.53 0.5939 diet*time CHO 17 CON 17 138.33 164.63 660 0.84 0.4011 diet*time CHO 17 FAT 17 52.5000 164.63 660 0.32 0.7499 diet*time CHO 18 CON 18 57.8333 164.63 660 0.35 0.7255 diet*time CHO 18 FAT 18 185.83 164.63 660 1.13 0.2594 diet*time CHO 19 CON 19 252.00 164.63 660 1.53 0.1263 diet*time CHO 19 FAT 19 209.83 164.63 660 1.27 0.2029

PAGE 160

144 Table B-14. Continued. Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time CHO 20 CON 20 302.67 164.63 660 1.84 0.0664 diet*time CHO 20 FAT 20 134.83 164.63 660 0.82 0.4131 diet*time CHO 21 CON 21 256.00 164.63 660 1.55 0.1204 diet*time CHO 21 FAT 21 213.50 164.63 660 1.30 0.1951 diet*time CHO 22 CON 22 300.50 164.63 660 1.83 0.0684 diet*time CHO 22 FAT 22 278.00 164.63 660 1.69 0.0918 diet*time CHO 23 CON 23 314.50 164.63 660 1.91 0.0565 diet*time CHO 23 FAT 23 186.17 164.63 660 1.13 0.2585 diet*time CHO 24 CON 24 263.50 164.63 660 1.60 0.1100 diet*time CHO 24 FAT 24 249.67 164.63 660 1.52 0.1299 diet*time CHO 25 CON 25 383.83 164.63 660 2.33 0.0200 diet*time CHO 25 FAT 25 324.50 164.63 660 1.97 0.0491 diet*time CHO 26 CON 26 420.50 164.63 660 2.55 0.0109 diet*time CHO 26 FAT 26 313.00 164.63 660 1.90 0.0577 diet*time CHO 27 CON 27 340.67 164.63 660 2.07 0.0389 diet*time CHO 27 FAT 27 232.33 164.63 660 1.41 0.1586 diet*time CHO 28 CON 28 375.17 164.63 660 2.28 0.0230 diet*time CHO 28 FAT 28 247.67 164.63 660 1.50 0.1330 diet*time CHO 29 CON 29 249.67 164.63 660 1.52 0.1299 diet*time CHO 29 FAT 29 82.666 164.63 660 0.50 0.6157 diet*time CHO 30 CON 30 422.00 164.63 660 2.56 0.0106 diet*time CHO 30 FAT 30 242.33 164.63 660 1.47 0.1415 diet*time CHO 31 CON 31 394.33 164.63 660 2.40 0.0169 diet*time CHO 31 FAT 31 220.17 164.63 660 1.34 0.1816 diet*time CHO 32 CON 32 197.00 164.63 660 1.20 0.2319 diet*time CHO 32 FAT 32 24.166 164.63 660 0.15 0.8833 diet*time CHO 33 CON 33 282.17 164.63 660 1.71 0.0870 diet*time CHO 33 FAT 33 91.500 164.63 660 0.56 0.5785 diet*time CHO 34 CON 34 271.00 164.63 660 1.65 0.1002 diet*time CHO 34 FAT 34 135.00 164.63 660 0.82 0.4125 diet*time CHO 35 CON 35 222.00 164.63 660 1.35 0.1780 diet*time CHO 35 FAT 35 119.83 164.63 660 0.73 0.4669 diet*time CHO 36 CON 36 157.50 164.63 660 0.96 0.3391 diet*time CHO 36 FAT 36 -19.00 164.63 660 -0.12 0.9082 diet*time CHO 37 CON 37 288.67 164.63 660 1.75 0.0800 diet*time CHO 37 FAT 37 108.17 164.63 660 0.66 0.5114 diet*time CHO 38 CON 38 259.33 164.63 660 1.58 0.1157 diet*time CHO 38 FAT 38 112.67 164.63 660 0.68 0.4940 diet*time CHO 39 CON 39 362.33 164.63 660 2.20 0.0281 diet*time CHO 39 FAT 39 163.50 164.63 660 0.99 0.3210 diet*time CHO 40 CON 40 287.67 164.63 660 1.75 0.0810 diet*time CHO 40 FAT 40 -31.83 164.63 660 -0.19 0.8467

PAGE 161

145 Table B-14. Continued. Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time CHO 41 CON 41 276.00 164.63 660 1.68 0.0941 diet*time CHO 41 FAT 41 -24.500 164.63 660 -0.15 0.8817 diet*time CHO 42 CON 42 153.50 164.63 660 0.93 0.3515 diet*time CHO 42 FAT 42 -7.6667 164.63 660 -0.05 0.9629 diet*time CHO 43 CON 43 300.33 164.63 660 1.82 0.0686 diet*time CHO 43 FAT 43 114.00 164.63 660 0.69 0.4889 diet*time CHO 44 CON 44 196.17 164.63 660 1.19 0.2339 diet*time CHO 44 FAT 44 -48.333 164.63 660 -0.29 0.7692 diet*time CHO 45 CON 45 208.33 164.63 660 1.27 0.2062 diet*time CHO 45 FAT 45 37.5000 164.63 660 0.23 0.8199 diet*time CON 1 FA T 1 55.0000 164.63 660 0.33 0.7384 diet*time CON 2 FA T 2 29.5000 164.63 660 0.18 0.8578 diet*time CON 3 FA T 3 -126.33 164.63 660 -0.77 0.4431 diet*time CON 4 FA T 4 -250.67 164.63 660 -1.52 0.1283 diet*time CON 5 FA T 5 -334.50 164.63 660 -2.03 0.0426 diet*time CON 6 FA T 6 -359.67 164.63 660 -2.18 0.0293 diet*time CON 7 FA T 7 -340.17 164.63 660 -2.07 0.0392 diet*time CON 8 FA T 8 -309.67 164.63 660 -1.88 0.0604 diet*time CON 9 FA T 9 -227.17 164.63 660 -1.38 0.1681 diet*time CON 10 FAT 10 -299.00 164.63 660 -1.82 0.0698 diet*time CON 11 FAT 11 -118.83 164.63 660 -0.72 0.4707 diet*time CON 12 FAT 12 -269.83 164.63 660 -1.64 0.1017 diet*time CON 13 FAT 13 -49.666 164.63 660 -0.30 0.7630 diet*time CON 14 FAT 14 -70.5000 164.63 660 -0.43 0.6686 diet*time CON 15 FAT 15 91.0000 164.63 660 0.55 0.5806 diet*time CON 16 FAT 16 -83.000 164.63 660 -0.50 0.6143 diet*time CON 17 FAT 17 -85.833 164.63 660 -0.52 0.6023 diet*time CON 18 FAT 18 128.00 164.63 660 0.78 0.4371 diet*time CON 19 FAT 19 -42.166 164.63 660 -0.26 0.7979 diet*time CON 20 FAT 20 -167.83 164.63 660 -1.02 0.3084 diet*time CON 21 FAT 21 -42.500 164.63 660 -0.26 0.7964 diet*time CON 22 FAT 22 -22.5000 164.63 660 -0.14 0.8913 diet*time CON 23 FAT 23 -128.33 164.63 660 -0.78 0.4360 diet*time CON 24 FAT 24 -13.833 164.63 660 -0.08 0.9331 diet*time CON 25 FAT 25 -59.333 164.63 660 -0.36 0.7187 diet*time CON 26 FAT 26 -107.50 164.63 660 -0.65 0.5140 diet*time CON 31 FAT 31 -174.17 164.63 660 -1.06 0.2905 diet*time CON 32 FAT 32 -172.83 164.63 660 -1.05 0.2942 diet*time CON 33 FAT 33 -190.67 164.63 660 -1.16 0.2472 diet*time CON 34 FAT 34 -136.00 164.63 660 -0.83 0.4091 diet*time CON 35 FAT 35 -102.17 164.63 660 -0.62 0.5351 diet*time CON 36 FAT 36 -176.50 164.63 660 -1.07 0.2841

PAGE 162

146 Table B-14. Continued. Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time CON 37 FAT 37 -180.50 164.63 660 -1.10 0.2733 diet*time CON 38 FAT 38 -146.67 164.63 660 -0.89 0.3733 diet*time CON 39 FAT 39 -198.83 164.63 660 -1.21 0.2276 diet*time CON 40 FAT 40 -319.50 164.63 660 -1.94 0.0527 diet*time CON 41 FAT 41 -300.50 164.63 660 -1.83 0.0684 diet*time CON 42 FAT 42 -161.17 164.63 660 -0.98 0.3280 diet*time CON 43 FAT 43 -186.33 164.63 660 -1.13 0.2581 diet*time CON 44 FAT 44 -244.50 164.63 660 -1.49 0.1380 diet*time CON 45 FAT 45 -170.83 164.63 660 -1.04 0.2998 Table B-15. ANOVA mixed procedure for mean bag volume minus baseline of unlabeled (BFAT) and labeled (FFAT) corn oil-enriched sweet feed meals (Phase II). Values 1-45: 2-min postprandial blocks. Class Levels Values diet 2 BFAT FFAT horse 6 Dusty Gambler Iso Mama Nina Seven time 44 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Effect NumDF DenDF F Value Pr > F diet 2 10 0.06 0.8079 time 44 440 7.48 <.0001 diet*time 44 440 0.95 0.5741 Differences of Least Squares Mean s Between Diets (Labeling effect) Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time BFAT 1 FFAT 1 -209.67 178.88 440 -1.17 0.2418 diet*time BFAT 2 FFAT 2 -139.00 178.88 440 -0.78 0.4375 diet*time BFAT 3 FFAT 3 -133.67 178.88 440 -0.75 0.4553 diet*time BFAT 4 FFAT 4 -223.83 178.88 440 -1.25 0.2115 diet*time BFAT 5 FFAT 5 -232.17 178.88 440 -1.30 0.1950 diet*time BFAT 6 FFAT 6 -110.67 178.88 440 -0.62 0.5365 diet*time BFAT 7 FFAT 7 -163.67 178.88 440 -0.91 0.3607 diet*time BFAT 8 FFAT 8 -168.67 178.88 440 -0.94 0.3462 diet*time BFAT 9 FFAT 9 -106.17 178.88 440 -0.59 0.5531 diet*time BFAT 10 FFAT 10 -245.0 178.88 440 -1.37 0.1715 diet*time BFAT 11 FFAT 11 -67.6667 178.88 440 -0.38 0.7054 diet*time BFAT 12 FFAT 12 -81.6667 178.88 440 -0.46 0.6482 diet*time BFAT 13 FFAT 13 -5 .0000 178.88 440 -0.03 0.9777 diet*time BFAT 14 FFAT 14 218.83 178.88 440 -1.22 0.2218 diet*time BFAT 15 FFAT 15 -67.0000 178.88 440 -0.37 0.7082 diet*time BFAT 16 FFAT 16 26. 6667 178.88 440 0.15 0.8816 diet*time BFAT 17 FFAT 17 -4 .8333 178.88 440 -0.03 0.9785 diet*time BFAT 18 FFAT 18 119. 33 178.88 440 0.67 0.5050

PAGE 163

147 Table B-15. Continued. Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time BFAT 19 FFAT 19 46.0000 178.88 440 0.26 0.7972 diet*time BFAT 20 FFAT 20 51.8333 178.88 440 0.29 0.7721 diet*time BFAT 21 FFAT 21 95.1667 178.88 440 0.53 0.5950 diet*time BFAT 22 FFAT 22 84.1667 178.88 440 0.47 0.6382 diet*time BFAT 23 FFAT 23 162. 50 178.88 440 0.91 0.3641 diet*time BFAT 24 FFAT 24 90.6667 178.88 440 0.51 0.6125 diet*time BFAT 25 FFAT 25 85.6667 178.88 440 0.48 0.6322 diet*time BFAT 26 FFAT 26 35.6667 178.88 440 -0.20 0.8420 diet*time BFAT 27 FFAT 27 19. 333 178.88 440 0.11 0.9140 diet*time BFAT 28 FFAT 28 92.3333 178.88 440 0.52 0.6060 diet*time BFAT 29 FFAT 29 -54.3333 178.88 440 -0.30 0.7615 diet*time BFAT 30 FFAT 30 26.5000 178.88 440 0.15 0.8823 diet*time BFAT 31 FFAT 31 20.6667 178.88 440 -0.12 0.9081 diet*time BFAT 32 FFAT 32 47.8333 178.88 440 -0.27 0.7893 diet*time BFAT 33 FFAT 33 41.1667 178.88 440 0.23 0.8181 diet*time BFAT 34 FFAT 34 57.8333 178.88 440 0.32 0.7466 diet*time BFAT 35 FFAT 35 48.5000 178.88 440 0.27 0.7864 diet*time BFAT 36 FFAT 36 16.1667 178.88 440 -0.09 0.9280 diet*time BFAT 37 FFAT 37 32.6667 178.88 440 -0.18 0.8552 diet*time BFAT 38 FFAT 38 106.33 178.88 440 0.59 0.5525 diet*time BFAT 39 FFAT 39 20.0000 178.88 440 0.11 0.9110 diet*time BFAT 40 FFAT 40 19.0000 178.88 440 -0.11 0.9155 diet*time BFAT 41 FFAT 41 94.8333 178.88 440 -0.53 0.5963 diet*time BFAT 42 FFAT 42 6.5000 178.88 440 0.04 0.9710 diet*time BFAT 43 FFAT 43 61.5000 178.88 440 -0.34 0.7312 diet*time BFAT 44 FFAT 44 -157.33 178.88 440 -0.88 0.3796 diet*time BFAT 45 FFAT 45 72.8333 178.88 440 -0.41 0.6841 Table B-16. ANOVA mixed procedure for mean bag volume minus baseline of unlabeled (BCHO) and labeled (FCHO) glucose-enriched sweet feed meals (Phase II). Values 1-45: 2-min postprandial blocks. Class Levels Values diet 2 BCHO FCHO horse 6 Dusty Gambler Iso Mama Nina Seven time 44 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Effect NumDF DenDF F Value Pr > F diet 2 10 0.12 0.7319 time 44 440 5.36 <.0001 diet*time 44 440 0.61 0.9782

PAGE 164

148 Table B-16. Continued. Differences of Least Squares Mean s Between Diets (Labeling effect) Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time BCHO 1 FCHO 1 -20.0000 206.28 440 -0.10 0.9228 diet*time BCHO 2 FCHO 2 -82.0000 206.28 440 -0.40 0.6912 diet*time BCHO 3 FCHO 3 -120.00 206.28 440 -0.58 0.5611 diet*time BCHO 4 FCHO 4 -121.00 206.28 440 -0.59 0.5578 diet*time BCHO 5 FCHO 5 -20.1667 206.28 440 -0.10 0.9222 diet*time BCHO 6 FCHO 6 34.0000 206.28 440 0.16 0.8692 diet*time BCHO 7 FCHO 7 45.0000 206.28 440 0.22 0.8274 diet*time BCHO 8 FCHO 8 85.5000 206.28 440 0.41 0.6787 diet*time BCHO 9 FCHO 9 24.0000 206.28 440 0.12 0.9074 diet*time BCHO 10 FCHO 10 13.5000 206.28 440 0.07 0.9479 diet*time BCHO 11 FCHO 11 -15.5000 206.28 440 -0.08 0.9401 diet*time BCHO 12 FCHO 12 -2.5000 206.28 440 -0.01 0.9903 diet*time BCHO 13 FCHO 13 -8.1667 206.28 440 -0.04 0.9684 diet*time BCHO 14 FCHO 14 -101.83 206.28 440 -0.49 0.6218 diet*time BCHO 15 FCHO 15 61.1667 206.28 440 0.30 0.7670 diet*time BCHO 16 FCHO 16 33.3333 206.28 440 0.16 0.8717 diet*time BCHO 17 FCHO 17 -10.833 206.28 440 -0.05 0.9581 diet*time BCHO 18 FCHO 18 22.3333 206.28 440 0.11 0.9138 diet*time BCHO 19 FCHO 19 -113.00 206.28 440 -0.55 0.5841 diet*time BCHO 20 FCHO 20 -42.000 206.28 440 -0.20 0.8388 diet*time BCHO 21 FCHO 21 28.3333 206.28 440 0.14 0.8908 diet*time BCHO 22 FCHO 22 81.5000 206.28 440 0.40 0.6930 diet*time BCHO 23 FCHO 23 193.50 206.28 440 0.94 0.3487 diet*time BCHO 24 FCHO 24 47.1667 206.28 440 0.23 0.8192 diet*time BCHO 25 FCHO 25 -8.8333 206.28 440 -0.04 0.9659 diet*time BCHO 26 FCHO 26 -116.67 206.28 440 -0.57 0.5720 diet*time BCHO 27 FCHO 27 31.8333 206.28 440 0.15 0.8774 diet*time BCHO 28 FCHO 28 161.33 206.28 440 0.78 0.4346 diet*time BCHO 29 FCHO 29 67.8333 206.28 440 0.33 0.7424 diet*time BCHO 30 FCHO 30 9.6667 206.28 440 0.05 0.9626 diet*time BCHO 31 FCHO 31 62.3333 206.28 440 0.30 0.7627 diet*time BCHO 32 FCHO 32 140.67 206.28 440 0.68 0.4957 diet*time BCHO 33 FCHO 33 164.50 206.28 440 0.80 0.4256 diet*time BCHO 34 FCHO 34 165.17 206.28 440 0.80 0.4238 diet*time BCHO 35 FCHO 35 170.00 206.28 440 0.82 0.4103 diet*time BCHO 36 FCHO 36 293.50 206.28 440 1.42 0.1555 diet*time BCHO 37 FCHO 37 82.3333 206.28 440 0.40 0.6900 diet*time BCHO 38 FCHO 38 89.5000 206.28 440 0.43 0.6646 diet*time BCHO 39 FCHO 39 12.1667 206.28 440 -0.06 0.953 diet*time BCHO 40 FCHO 40 168.50 206.28 440 0.82 0.4145 diet*time BCHO 41 FCHO 41 76.333 206.28 440 0.37 0.7115

PAGE 165

149 Table B-16. Continued. Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time BCHO 42 FCHO 42 255.33 206.28 440 1.24 0.2165 diet*time BCHO 43 FCHO 43 176.50 206.28 440 0.86 0.3927 diet*time BCHO 44 FCHO 44 239.50 206.28 440 1.16 0.2463 diet*time BCHO 45 FCHO 45 252.17 206.28 440 1.22 0.2222

PAGE 166

150 APPENDIX C STATISTICAL TESTS FOR GASTRIC EMPTYING DATA Table C-1. Shapiro Wilk test for normality of Phase I parameters. Source T1/2 Tmax GEC W Sign. W Sign. W Sign. High-fat pelleted meal 0.96821 0.8636 0.88663 0.3404 0.925970.5691 High-CHO pelleted meal 0.87649 0.2533 0.89642 0.3532 0.974000.9182 Table C-2. Bartlett's test for homogeneity of variance of Phase I parameters. Parameter DF Chi Square Sign. t1/2 1 2.67640 0.1018 tmax 1 2.50590 0.1134 GEC 1 0.00463 0.9458 Table C-3. Two-sample t-test for Phase I parameters. Parameter Mean Diff. SEM 95% CL-lower95% CL-upperDF t Value Sign. T1/2 -0.488 0.965 -2.671 1.6948 9 -0.51 0.63 Tmax -0.149 0.494 -1.267 0.9693 9 -0.3 0.77 GEC 0.0697 0.521 -1.11 1.2493 9 0.13 0.9 Table C-4. Shapiro-Wilk test for normality of Phase II parameters without an intragastric bag. Source T1/2 Tmax GEC W Sign. W Sign. W Sign. Diets Difference 0.9879610.9836 0.88109 0.2741 0.9537920.7708 Table C-5. Bartlett's test for homogeneity of variance of Phase II parameters without an intragastric bag. Parameter DF Chi Square Sign. t1/2 1 1.7591 0.1847 tmax 1 1.0778 0.2992 GEC 1 3.7092 0.0541

PAGE 167

151 Table C-6. Paired-sample t-test for Phase II parameters without an intragastric bag. Parameter Mean Diff. SEM t Value Sign. t1/2 -0.07333 0.456747 -0.16 0.8787 tmax -0.235 0.21904 -1.07 0.3324 GEC -0.14 0.354194 -0.4 0.7089 Table C-7. Shapiro-Wilk test for normality of Phase II parameters with an intragastric bag. Source T1/2 Tmax GEC W Sign. W Sign. W Sign. Control meal 0.86952 0.2243 0.773291 0.0333 0.831644 0.111 Corn oil meal 0.953997 0.7725 0.970655 0.8968 Glucose meal 0.852777 0.1657 0.986584 0.9792 Table C-8. Bartlett's test for homogeneity of variance of Phase II parameters with an intragastric bag. Parameter DF Chi Square Sign. t1/2 2 1.90800 0.3852 tmax 2 GEC 2 0.0991 0.9517 Table C-9. Repeated meas ures ANOVA for Phase II t1/2 with an intragastric bag. Class Levels Values horse 6 DUSTY GAMBLER ISO MAMA NINA SEVEN diet 3 CHO CON .FAT Source DF Sum of Squares Mean Square F Value Pr > F Model 7 1.86470556 0.26638651 0.66 0.7037 Error 10 4.05618889 0.40561889 Total 17 5.92089444 Source DF Anova SS Mean Square F Value Pr > F horse 5 1.52522778 0.30504556 0.75 0.6033 diet 2 0.33947778 0.16973889 0.42 0.6691 Table C-10. Friedmans 2-way ANOVA for Phase II tmax with intragastric bag. Cochran-Mantel-Haenszel Sta tistics (Based on Rank Scores) Statistic Alternative H ypothesis DF Value Prob. 1 Nonzero Correlation 1 0.0000 1.0000 2 Row Mean Scores Differ 2 4.3636 0.1128

PAGE 168

152 Table C-11. Repeated measures ANOVA for Phase II GEC with an intragastric bag. Class Levels Values horse 6 DUSTY GAMBLER ISO MAMA NINA SEVEN diet 3 CHO CON .FAT Source DF Sum of Squares Mean Square F Value Pr > F Model 7 2.73777222 0.39111032 1.12 0.4228 Error 10 3.50505556 0.35050556 Total 17 6.24282778 Source DF Anova SS Mean Square F Value Pr > F horse 5 2.57462778 0.51492556 1.47 0.2823 diet 2 0.16314444 0.08157222 0.23 0.7965 Table C-12. Shapiro-Wilk test for normality of Phase II parameters: effect of an intragastric bag on gastric emptying. Source T1/2 Tmax GEC W Sign. W Sign. W Sign. Corn oil meals Difference 0.8141710.0785 0.8382710.1261 0.8960620.3512 Glucose meals Difference 0.9390680.6517 0.9013320.3819 0.8227220.0932 Table C-13. Bartlett's test for homogeneity of variance of Phase II parameters: effect of an intragastric bag on gastric emptying. Meal Parameter DF Chi Square Sign. Corn oil-enriched t1/2 1 0.0756 0.7833 tmax 1 0.3728 0.5415 GEC 1 1.1612 0.2812 Glucose-enriched t1/2 1 0.3259 0.5681 tmax 1 0.0335 0.8548 GEC 1 0.4167 0.5186 Table C-14. Paired-sample t-test for Phase II pa rameters: effect of an intragastric bag on gastric emptying. Source Parameter Mean Diff. SEM t Value Sign. Corn oil meals Diff. t1/2 0.4800000 0.3449444 1.39 0.2228 tmax 0.2066667 0.1457090 1.42 0.2153 GEC -0.3416667 0.3092724 -1.10 0.3196 Glucose meals Diff. t1/2 0.4883333 0.1944808 2.51 0.0538 tmax 0.3033333 0.1244901 2.44 0.0589 GEC -0.0533333 0.1832788 -0.29 0.7827

PAGE 169

153 APPENDIX D STATISTICAL TESTS FOR PH DATA Table D-1. ANOVA mixed procedure for mean intragastric pH of pelleted meals (Phase I). Diet A:high-fat meal, B:high-CHO meal. Exp. 1:labeled, 2:unlabeled. Time: 0:baseline; 1-24: 5min postprandial blocks. Class Levels Values horse 6 Dusty Gambler Iso Mama Nina Seven exp. 2 1 2 diet 2 A B time 25 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Effect NumDF DenDF F Value Pr > F diet 1 10 12.80 0.0050 exp 1 490 0.40 0.5270 time 24 490 10.01 <.0001 diet*time 24 490 1.06 0.3851 diet*exp 1 490 2.02 0.1557 diet*exp*time 48 490 0.64 0.9710 Differences of Least Squares Means Within Diet Effect diet time _diet _time Es timate SEM DF t Value Pr > |t| diet*time A 0 A 1 -0.07083 0.4123 490 -0.17 0.8637 diet*time A 0 A 2 -0.09667 0.4123 490 -0.23 0.8147 diet*time A 0 A 3 -0.7008 0.4123 490 -1.70 0.0898 diet*time A 0 A 4 -1.2983 0.4123 490 -3.15 0.0017 diet*time A 0 A 5 -1.2675 0.4123 490 -3.07 0.0022 diet*time A 0 A 6 -1.2867 0.4123 490 -3.12 0.0019 diet*time A 0 A 7 -1.1300 0.4123 490 -2.74 0.0064 diet*time A 0 A 8 -1.1608 0.4123 490 -2.82 0.0051 diet*time A 0 A 9 -1.2675 0.4123 490 -3.07 0.0022 diet*time A 0 A 10 -1.3100 0.4123 490 -3.18 0.0016 diet*time A 0 A 11 -1.1283 0.4123 490 -2.74 0.0064 diet*time A 0 A 12 -1.1233 0.4123 490 -2.72 0.0067 diet*time A 0 A 13 -0.9717 0.4123 490 -2.36 0.0188 diet*time A 0 A 14 -0.8908 0.4123 490 -2.16 0.0312 diet*time A 0 A 15 -0.8658 0.4123 490 -2.10 0.0362 diet*time A 0 A 16 -0.7233 0.4123 490 -1.75 0.0800 diet*time A 0 A 17 -0.6258 0.4123 490 -1.52 0.1297 diet*time A 0 A 18 -0.5900 0.4123 490 -1.43 0.1531 diet*time A 0 A 19 -0.4058 0.4123 490 -0.98 0.3255 diet*time A 0 A 20 -0.2575 0.4123 490 -0.62 0.5326

PAGE 170

154 Table D-1. Continued. Effect diet time _diet _time Es timate SEM DF t Value Pr > |t| diet*time B 0 B 1 0.3017 0.4123 490 0.73 0.4647 diet*time B 0 B 2 0.06583 0.4123 490 0.16 0.8732 diet*time B 0 B 3 -0.6383 0.4123 490 -1.55 0.1222 diet*time B 0 B 4 -1.1883 0.4123 490 -2.88 0.0041 diet*time B 0 B 5 -1.4133 0.4123 490 -3.43 0.0007 diet*time B 0 B 6 -1.5725 0.4123 490 -3.81 0.0002 diet*time B 0 B 7 -1.7792 0.4123 490 -4.31 <.0001 diet*time B 0 B 8 -2.0208 0.4123 490 -4.90 <.0001 diet*time B 0 B 9 -2.1808 0.4123 490 -5.29 <.0001 diet*time B 0 B 10 -2.2108 0.4123 490 -5.36 <.0001 diet*time B 0 B 11 -2.1567 0.4123 490 -5.23 <.0001 diet*time B 0 B 12 -2.0408 0.4123 490 -4.95 <.0001 diet*time B 0 B 13 -2.0275 0.4123 490 -4.92 <.0001 diet*time B 0 B 14 -1.8625 0.4123 490 -4.52 <.0001 diet*time B 0 B 15 -1.8033 0.4123 490 -4.37 <.0001 diet*time B 0 B 16 -1.6767 0.4123 490 -4.07 <.0001 diet*time B 0 B 17 -1.3458 0.4123 490 -3.26 0.0012 diet*time B 0 B 18 -1.0300 0.4123 490 -2.50 0.0128 diet*time B 0 B 19 -0.8817 0.4123 490 -2.14 0.0330 diet*time B 0 B 20 -0.7225 0.4123 490 -1.75 0.0804 diet*time B 0 B 21 -0.5192 0.4123 490 -1.26 0.2086 diet*time B 0 B 22 -0.3500 0.4123 490 -0.85 0.3964 diet*time B 0 B 23 -0.2375 0.4123 490 -0.58 0.5649 diet*time B 0 B 24 -0.0375 0.4123 490 -0.09 0.9276 Differences of Least Squares Means Between Diets Effect diet time _diet _time Es timate SEM DF t Value Pr > |t| diet*time A 0 B 0 -0.5308 0.4961 490 -1.07 0.2852 diet*time A 1 B 1 -0.1583 0.4961 490 -0.32 0.7498 diet*time A 2 B 2 -0.3683 0.4961 490 -0.74 0.4582 diet*time A 3 B 3 -0.4683 0.4961 490 -0.94 0.3457 diet*time A 4 B 4 -0.4208 0.4961 490 -0.85 0.3967 diet*time A 5 B 5 -0.6767 0.4961 490 -1.36 0.1732 diet*time A 6 B 6 -0.8167 0.4961 490 -1.65 0.1004 diet*time A 7 B 7 -1.1800 0.4961 490 -2.38 0.0178 diet*time A 8 B 8 -1.3908 0.4961 490 -2.80 0.0053 diet*time A 9 B 9 -1.4442 0.4961 490 -2.91 0.0038 diet*time A 10 B 10 -1.4317 0.4961 490 -2.89 0.0041 diet*time A 11 B 11 -1.5592 0.4961 490 -3.14 0.0018 diet*time A 12 B 12 -1.4483 0.4961 490 -2.92 0.0037 diet*time A 13 B 13 -1.5867 0.4961 490 -3.20 0.0015 diet*time A 15 B 15 -1.4683 0.4961 490 -2.96 0.0032 diet*time A 16 B 16 -1.4842 0.4961 490 -2.99 0.0029

PAGE 171

155 Table D-1. Continued. Effect diet time _diet _time Es timate SEM DF t Value Pr > |t| diet*time A 17 B 17 -1.2508 0.4961 490 -2.52 0.0120 diet*time A 18 B 18 -0.9708 0.4961 490 -1.96 0.0509 diet*time A 19 B 19 -1.0067 0.4961 490 -2.03 0.0430 diet*time A 20 B 20 -0.9958 0.4961 490 -2.01 0.0453 diet*time A 21 B 21 -1.0000 0.4961 490 -2.02 0.0444 diet*time A 22 B 22 -0.9558 0.4961 490 -1.93 0.0546 diet*time A 24 B 24 -0.7608 0.4961 490 -1.53 0.1258 Table D-2. ANOVA mixed procedure for mean intragastric pH of labeled sweetfeed meals (Phase II). 0:baseline; 1-24: 5-min postprandial blocks. Class Levels Values horse 6 Dusty Gambler Iso Mama Nina Seven diet 3 CHO CON FAT time 25 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Effect NumDF DenDF F Value Pr > F diet 2 15 0.56 0.5826 time 24 360 26.71 <.0001 diet*time 48 360 2.03 0.0002 Differences of Least Squares Means Within Diet Effect diet time _diet _time Estim ate SEM DF t Value Pr > |t| diet*time CHO 0 CHO 1 0.5883 0.3377 360 1.74 0.0823 diet*time CHO 0 CHO 2 -0.3967 0.3377 360 -1.17 0.2409 diet*time CHO 0 CHO 3 -0.9400 0.3377 360 -2.78 0.0057 diet*time CHO 0 CHO 4 -1.1933 0.3377 360 -3.53 0.0005 diet*time CHO 0 CHO 5 -1.2617 0.3377 360 -3.74 0.0002 diet*time CHO 0 CHO 6 -1.1617 0.3377 360 -3.44 0.0007 diet*time CHO 0 CHO 7 -1.0167 0.3377 360 -3.01 0.0028 diet*time CHO 0 CHO 8 -0.8533 0.3377 360 -2.53 0.0119 diet*time CHO 0 CHO 9 -0.6967 0.3377 360 -2.06 0.0398 diet*time CHO 0 CHO 10 -0.4567 0.3377 360 -1.35 0.1772 diet*time CHO 0 CHO 11 0.2933 0.3377 360 -0.87 0.3857 diet*time CHO 0 CHO 12 -0.2100 0.3377 360 -0.62 0.5345 diet*time CHO 0 CHO 13 -0.0516 0.3377 360 -0.15 0.8785 diet*time CHO 0 CHO 14 0.105 0.3377 360 0.31 0.7560 diet*time CHO 0 CHO 15 0.3067 0.3377 360 0.91 0.3645 diet*time CHO 0 CHO 16 0.5133 0.3377 360 1.52 0.1294 diet*time CHO 0 CHO 17 0.6017 0.3377 360 1.78 0.0757 diet*time CHO 0 CHO 18 0.7250 0.3377 360 2.15 0.0325 diet*time CHO 0 CHO 19 0.9850 0.3377 360 2.92 0.0038 diet*time CHO 0 CHO 20 0.9617 0.3377 360 2.85 0.0047 diet*time CHO 0 CHO 21 0.9283 0.3377 360 2.75 0.0063 diet*time CHO 0 CHO 22 1.0400 0.3377 360 3.08 0.0022

PAGE 172

156 Table D-2. Continued. Effect diet time _diet _time Estim ate SEM DF t Value Pr > |t| diet*time CHO 0 CHO 23 1.0783 0.3377 360 3.19 0.0015 diet*time CHO 0 CHO 24 1.0683 0.3377 360 3.16 0.0017 diet*time CON 0 CON 1 0.1933 0.3377 360 0.57 0.5674 diet*time CON 0 CON 2 0.1283 0.3377 360 0.38 0.7042 diet*time CON 0 CON 3 2.78E-17 0.3377 360 0.00 1.0000 diet*time CON 0 CON 4 -0.2083 0.3377 360 -0.62 0.5377 diet*time CON 0 CON 5 -0.5467 0.3377 360 -1.62 0.1064 diet*time CON 0 CON 6 -1.2033 0.3377 360 -3.56 0.0004 diet*time CON 0 CON 7 -1.8717 0.3377 360 -5.54 <.0001 diet*time CON 0 CON 8 -1.9150 0.3377 360 -5.67 <.0001 diet*time CON 0 CON 9 -1.7450 0.3377 360 -5.17 <.0001 diet*time CON 0 CON 10 -1.4717 0.3377 360 -4.36 <.0001 diet*time CON 0 CON 11 -1.4483 0.3377 360 -4.29 <.0001 diet*time CON 0 CON 12 -1.1167 0.3377 360 -3.31 0.0010 diet*time CON 0 CON 13 -0.7433 0.3377 360 -2.20 0.0284 diet*time CON 0 CON 14 -0.7050 0.3377 360 -2.09 0.0375 diet*time CON 0 CON 15 -0.3817 0.3377 360 -1.13 0.2592 diet*time CON 0 CON 16 -0.1667 0.3377 360 -0.49 0.6220 diet*time CON 0 CON 17 -0.07333 0.3377 360 -0.22 0.8282 diet*time CON 0 CON 18 0.05833 0.3377 360 0.17 0.8630 diet*time CON 0 CON 19 0.09000 0.3377 360 0.27 0.7900 diet*time CON 0 CON 20 0.1717 0.3377 360 0.51 0.6115 diet*time CON 0 CON 21 0.2067 0.3377 360 0.61 0.5410 diet*time CON 0 CON 22 0.1800 0.3377 360 0.53 0.5944 diet*time CON 0 CON 23 0.1550 0.3377 360 0.46 0.6465 diet*time CON 0 CON 24 0.1650 0.3377 360 0.49 0.6254 diet*time FAT 0 FAT 1 0.1133 0.3377 360 0.34 0.7374 diet*time FAT 0 FAT 2 -0.6083 0.3377 360 -1.80 0.0725 diet*time FAT 0 FAT 3 -1.1150 0.3377 360 -3.30 0.0011 diet*time FAT 0 FAT 4 -1.4800 0.3377 360 -4.38 <.0001 diet*time FAT 0 FAT 5 -1.8433 0.3377 360 -5.46 <.0001 diet*time FAT 0 FAT 6 -2.0183 0.3377 360 -5.98 <.0001 diet*time FAT 0 FAT 7 -2.0483 0.3377 360 -6.07 <.0001 diet*time FAT 0 FAT 8 -1.7250 0.3377 360 -5.11 <.0001 diet*time FAT 0 FAT 9 -1.6550 0.3377 360 -4.90 <.0001 diet*time FAT 0 FA T 10 -1.5600 0.3377 360 -4.62 <.0001 diet*time FAT 0 FA T 11 -1.3150 0.3377 360 -3.89 0.0001 diet*time FAT 0 FA T 12 -0.9983 0.3377 360 -2.96 0.0033 diet*time FAT 0 FA T 13 -0.8267 0.3377 360 -2.45 0.0148 diet*time FAT 0 FA T 14 -0.7067 0.3377 360 -2.09 0.0371 diet*time FAT 0 FA T 15 -0.5333 0.3377 360 -1.58 0.1152 diet*time FAT 0 FA T 16 -0.4550 0.3377 360 -1.35 0.1787

PAGE 173

157 Table D-2. Continued. Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time FAT 0 FA T 17 -0.2500 0.3377 360 -0.74 0.4596 diet*time FAT 0 FA T 18 -0.09000 0.3377 360 -0.27 0.7900 diet*time FAT 0 FA T 19 0.01833 0.3377 360 0.05 0.9567 diet*time FAT 0 FA T 20 0.03333 0.3377 360 0.10 0.9214 diet*time FAT 0 FA T 21 0.09667 0.3377 360 0.29 0.7749 diet*time FAT 0 FA T 22 -0.01000 0.3377 360 -0.03 0.9764 diet*time FAT 0 FA T 23 -0.08167 0.3377 360 -0.24 0.8091 diet*time FAT 0 FA T 24 -0.1250 0.3377 360 -0.37 0.7115 Differences of Least Squares Means Between Diets Effect diet time _diet _time Estim ate SEM DF t Value Pr > |t| diet*time CHO 0 CON 0 0.5933 0.5552 360 1.07 0.2859 diet*time CHO 0 FAT 0 0.4250 0.5552 360 0.77 0.4445 diet*time CHO 1 CON 1 0.1983 0.5552 360 0.36 0.7211 diet*time CHO 1 FAT 1 -0.0500 0.5552 360 -0.09 0.9283 diet*time CHO 2 CON 2 1.1183 0.5552 360 2.01 0.0447 diet*time CHO 2 FAT 2 0.2133 0.5552 360 0.38 0.7010 diet*time CHO 3 CON 3 1.5333 0.5552 360 2.76 0.0060 diet*time CHO 3 FAT 3 0.2500 0.5552 360 0.45 0.6528 diet*time CHO 4 CON 4 1.5783 0.5552 360 2.84 0.0047 diet*time CHO 4 FAT 4 0.1383 0.5552 360 0.25 0.8034 diet*time CHO 5 CON 5 1.3083 0.5552 360 2.36 0.0190 diet*time CHO 5 FAT 5 -0.1567 0.5552 360 -0.28 0.7780 diet*time CHO 6 CON 6 0.5517 0.5552 360 0.99 0.3211 diet*time CHO 6 FAT 6 -0.4317 0.5552 360 -0.78 0.4374 diet*time CHO 7 CON 7 -0.2617 0.5552 360 -0.47 0.6377 diet*time CHO 7 FAT 7 -0.6067 0.5552 360 -1.09 0.2753 diet*time CHO 8 CON 8 -0.4683 0.5552 360 -0.84 0.3995 diet*time CHO 8 FAT 8 -0.4467 0.5552 360 -0.80 0.4216 diet*time CHO 9 CON 9 -0.4550 0.5552 360 -0.82 0.4130 diet*time CHO 9 FAT 9 -0.5333 0.5552 360 -0.96 0.3374 diet*time CHO 10 CON 10 -0.4217 0.5552 360 -0.76 0.4481 diet*time CHO 10 FAT 10 -0.6783 0.5552 360 -1.22 0.2226 diet*time CHO 11 CON 11 -0.5617 0.5552 360 -1.01 0.3124 diet*time CHO 11 FAT 11 -0.5967 0.5552 360 -1.07 0.2832 diet*time CHO 12 CON 12 -0.3133 0.5552 360 -0.56 0.5729 diet*time CHO 12 FAT 12 -0.3633 0.5552 360 -0.65 0.5133 diet*time CHO 13 CON 13 -0.09833 0.5552 360 -0.18 0.8595 diet*time CHO 13 FAT 13 -0.3500 0.5552 360 -0.63 0.5288 diet*time CHO 14 CON 14 -0.2167 0.5552 360 -0.39 0.6966 diet*time CHO 14 FAT 14 -0.3867 0.5552 360 -0.70 0.4866 diet*time CHO 15 CON 15 -0.0950 0.5552 360 -0.17 0.8642 diet*time CHO 15 FAT 15 -0.4150 0.5552 360 -0.75 0.4553

PAGE 174

158 Table D-2. Continued. Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time CHO 16 CON 16 -0.0866 0.5552 360 -0.16 0.8760 diet*time CHO 16 FAT 16 -0.5433 0.5552 360 -0.98 0.3284 diet*time CHO 17 CON 17 -0.0816 0.5552 360 -0.15 0.8831 diet*time CHO 17 FAT 17 -0.4267 0.5552 360 -0.77 0.4427 diet*time CHO 18 CON 18 -0.0733 0.5552 360 -0.13 0.8950 diet*time CHO 18 FAT 18 -0.3900 0.5552 360 -0.70 0.4829 diet*time CHO 19 CON 19 -0.3017 0.5552 360 -0.54 0.5872 diet*time CHO 19 FAT 19 -0.5417 0.5552 360 -0.98 0.3299 diet*time CHO 20 CON 20 -0.1967 0.5552 360 -0.35 0.7234 diet*time CHO 20 FAT 20 -0.5033 0.5552 360 -0.91 0.3653 diet*time CHO 21 CON 21 -0.1283 0.5552 360 -0.23 0.8173 diet*time CHO 21 FAT 21 -0.4067 0.5552 360 -0.73 0.4644 diet*time CHO 22 CON 22 -0.2667 0.5552 360 -0.48 0.6313 diet*time CHO 22 FAT 22 -0.6250 0.5552 360 -1.13 0.2610 diet*time CHO 23 CON 23 -0.3300 0.5552 360 -0.59 0.5526 diet*time CHO 23 FAT 23 -0.7350 0.5552 360 -1.32 0.1864 diet*time CHO 24 CON 24 -0.3100 0.5552 360 -0.56 0.5770 diet*time CHO 24 FAT 24 -0.7683 0.5552 360 -1.38 0.1673 diet*time CON 0 FAT 0 -0.1683 0.5552 360 -0.30 0.7619 diet*time CON 1 FAT 1 -0.2483 0.5552 360 -0.45 0.6549 diet*time CON 2 FAT 2 -0.9050 0.5552 360 -1.63 0.1040 diet*time CON 3 FAT 3 -1.2833 0.5552 360 -2.31 0.0214 diet*time CON 4 FAT 4 -1.4400 0.5552 360 -2.59 0.0099 diet*time CON 5 FAT 5 -1.4650 0.5552 360 -2.64 0.0087 diet*time CON 6 FAT 6 -0.9833 0.5552 360 -1.77 0.0774 diet*time CON 7 FAT 7 -0.3450 0.5552 360 -0.62 0.5347 diet*time CON 8 FAT 8 0.0216 0.5552 360 0.04 0.9689 diet*time CON 9 FAT 9 -0.0783 0.5552 360 -0.14 0.8879 diet*time CON 10 FAT 10 -0.2567 0.5552 360 -0.46 0.6442 diet*time CON 11 FAT 11 -0.0350 0.5552 360 -0.06 0.9498 diet*time CON 12 FAT 12 -0.0500 0.5552 360 -0.09 0.9283 diet*time CON 13 FAT 13 -0.2517 0.5552 360 -0.45 0.6506 diet*time CON 14 FAT 14 -0.1700 0.5552 360 -0.31 0.7596 diet*time CON 15 FAT 15 -0.3200 0.5552 360 -0.58 0.5647 diet*time CON 16 FAT 16 -0.4567 0.5552 360 -0.82 0.4113 diet*time CON 17 FAT 17 -0.3450 0.5552 360 -0.62 0.5347 diet*time CON 18 FAT 18 -0.3167 0.5552 360 -0.57 0.5688 diet*time CON 19 FAT 19 -0.2400 0.5552 360 -0.43 0.6658 diet*time CON 20 FAT 20 -0.3067 0.5552 360 -0.55 0.5811 diet*time CON 21 FAT 21 -0.2783 0.5552 360 -0.50 0.6165 diet*time CON 22 FAT 22 -0.3583 0.5552 360 -0.65 0.5191 diet*time CON 23 FAT 23 -0.4050 0.5552 360 -0.73 0.4662

PAGE 175

159 Table D-2. Continued. Effect diet time _diet _time Estimate SEM DF t Value Pr > |t| diet*time CON 24 FAT 24 -0.4583 0.5552 360 -0.83 0.4096

PAGE 176

160 LIST OF REFERENCES 1. Abrahamsson H, Jansson G: Elicitation of reflex vagal relaxation of the stomach from pharynx and esophagus in the cat. Acta Physiol Scand 77:172-178, 1969. 2. Abrahamsson H, Jansson G: Vago-vagal ga stro-gastric relaxation in the cat. Acta Physiol Scand 88:289-295, 1973. 3. Adams SB, MacHarg MA: Neostigmine me thylsulfate delays gastric emptying of particulate markers in horses. Am J Vet Res 46:2498-2499, 1985. 4. Alexander F, Hickson JCD: The salivary and pancreatic secretions of the horse. In Phillipson AT (ed): Physiology of Dige stion and Metabolism in the Ruminant. Newcastle upon Tyne, England, Oriel, 1970, 375-389. 5. Argenzio RA, Eisemann J: Mechanisms of acid injury in porcine gastroesophageal mucosa. Am J Vet Res 57:564-573, 1996. 6. Argenzio RA, Southworth M, Stevens CE: Sites of organic acid production and absorption in the equine gastrointestin al tract. Am J Physiol 226:1043-1050, 1974. 7. Athow AC, Sewerniak AT, Barton TP, Clark CG, Lewin MR: Measurement of blood cortisol and acid output in patien ts with duodenal ulceration and normal subjects during insulin hypoglycaemia Clin Sci (Lond) 69:37-40, 1985. 8. Azpiroz F, Malagelada JR: Intestinal c ontrol of gastric tone. Am J Physiol 249:G501G509, 1985. 9. Backus RC, Howard KA, Rogers QR: The potency of dietary amino acids in elevating plasma cholecystokinin immunoreac tivity in cats is related to amino acid hydrophobicity. Regul Pept 72:31-40, 1997. 10. Backus RC, Rosenquist GL, Rogers QR, Calam J, Morris JG: Elevation of plasma cholecystokinin (CCK) immunoreactivity by fa t, protein, and amino acids in the cat, a carnivore. Regul Pept 57:123-131, 1995. 11. Baker SJ, Gerring EL: Technique for prol onged, minimally invasive monitoring of intragastric pH in ponies. Am J Vet Res 54:1725-1734, 1993. 12. Baker SJ, Gerring EL: Gastric emptying of four liquid meals in pony foals. Res Vet Sci 56:164-169, 1994.

PAGE 177

161 13. Baker SJ, Gerring EL: Gastric emptyi ng of solid, non-digestible, radiopaque markers in ponies. Res Vet Sci 56:386-388, 1994. 14. Ballinger AB, Clark ML: L-phenylalanine releases cholecystokinin (CCK) and is associated with reduced food intake in hu mans: evidence for a physiological role of CCK in control of eating. Metabolism 43:735-738, 1994. 15. Berthoud HR, Kressel M, Raybould HE, Ne uhuber WL: Vagal sensors in the rat duodenal mucosa: distribution and structur e as revealed by in vivo DiI-tracing. Anat Embryol (Berl) 191:203-212, 1995. 16. Blackshaw LA, Grundy D: Effects of c holecystokinin (CCK-8) on two classes of gastroduodenal vagal afferent fibre. J Auton Nerv Syst 31:191-201, 1990. 17. Brener W, Hendrix TR, McHugh PR: Re gulation of the gast ric emptying of glucose. Gastroenterology 85:76-82, 1983. 18. Brenner L, Yox DP, Ritter RC: Suppre ssion of sham feeding by intraintestinal nutrients is not correlated with plasma cholecystokinin elevation. Am J Physiol 264:R972-R976, 1993. 19. Brikas P, Kania BF, Fioramonti J, Bue no L: Central and peripheral serotonergic influences on viscerovisceral inhibitory reflex during duodenal distension in sheep. Dig Dis Sci 38:1079-1086, 1993. 20. Brown NJ, Greenburgh A, Tomlin J: The effects of pectin and wheat bran on the distribution of a meal in the gastrointestin al tract of the rat. Br J Nutr 72:289-297, 1994. 21. Buchan AM: Nutrient Tasting and Signali ng Mechanisms in the Gut III. Endocrine cell recognition of lumi nal nutrients. Am J P hysiol 277:G1103-G1107, 1999. 22. Bull JS, Grundy D, Scratcherd T: The effect of intraluminal tryptophan and phenylalanine on small intestinal motility in the conscious dog. J Physiol 367:353362, 1985. 23. Bush JA, Freeman DE, Kline KH, Merchen NR, Fahey GC: Dietary fat supplementation effects on in vitro nutrie nt disappearance and in vivo nutrient intake and total tract digestibility by horses. J Anim Sci 79:232-239, 2001. 24. Carney BI, Jones KL, Horowitz M, Sun WM, Hebbard G, Edelbroek MA: Stereospecific effects of tryptophan on gastric emptying and hunger in humans. J Gastroenterol Hepatol 9:557-563, 1994. 25. Castiglione KE, Read NW, French SJ: Adaptation to high-fat diet accelerates emptying of fat but not ca rbohydrate test meals in humans. Am J Physiol Regul Integr Comp Physiol 282:R366-R371, 2002.

PAGE 178

162 26. Cecil JE, Francis J, Read NW: Relative contributions of intestinal, gastric, orosensory influences and information to ch anges in appetite induced by the same liquid meal. Appetite 31:377-390, 1998. 27. Cecil JE, Francis J, Read NW: Compar ison of the effects of a high-fat and highcarbohydrate soup delivered orally and in tragastrically on gastric emptying, appetite, and eating behaviour. Physiol Behav 67:299-306, 1999. 28. Choi BR, Palmquist DL, Allen MS: Chol ecystokinin mediates depression of feed intake in dairy cattle fed high fat di ets. Domest Anim Endocrinol 19:159-175, 2000. 29. Clarke LL, Roberts MC, Argenzio RA: Feeding and digestive problems in horses. Physiologic responses to a concentrated meal. Vet Clin North Am Equine Pract 6:433-450, 1990. 30. Collins PJ, Horowitz M, Cook DJ, Hard ing PE, Shearman DJ: Gastric emptying in normal subjects-a reproducible technique using a single scintillation camera and computer system. Gut 24:1117-1125, 1983. 31. Collins PJ, Horowitz M, Maddox A, Myers JC, Chatterton BE: Effects of increasing solid component size of a mixe d solid/liquid meal on solid and liquid gastric emptying. Am J Physiol 271:G549-G554, 1996. 32. Collins PJ, Houghton LA, Read NW, Horow itz M, Chatterton BE, Heddle R, Dent J: Role of the proximal and distal st omach in mixed solid and liquid meal emptying. Gut 32:615-619, 1991. 33. Cooke AR, Ward WO: Effect of tryptopha n and its metabolites on gastric emptying of liquid meals in dogs. Proc Soc Exp Biol Med 152:656-658, 1976. 34. Cortot A, Phillips SF, Malagelada JR: Ga stric emptying of lipid s after ingestion of an homogenized meal. Gastroenterology 76:939-944, 1979. 35. Cortot A, Phillips SF, Malagelada JR: Parallel gastric emptyi ng of nonhydrolyzable fat and water after a solid-liquid meal in humans. Gastroenterology 82:877-881, 1982. 36. Costello M, Papasouliotis K, Barr FJ, Gruffydd-Jones TJ, Caney SM: Determination of solidand liquid-phase ga stric emptying half times in cats by use of nuclear scintigraphy. Am J Vet Res 60:1222-1226, 1999. 37. Covasa M, Ritter RC: Adaptation to high-fat diet reduces inhibition of gastric emptying by CCK and intestinal oleate. Am J Physiol Regul Integr Comp Physiol 278:R166-R170, 2000.

PAGE 179

163 38. Cuber JC, Bernard C, Levenez F, Chayvialle JA: Lipi ds, proteins and carbohydrates stimulate the secretion of in testinal cholecystokinin in the pig. Reprod Nutr Dev 30:267-275, 1990. 39. Cunningham KM, Daly J, Horowitz M, R ead MW: Gastrointest inal adaptation to diets of differing fat composition in human volunteers. Gut 32:483-486, 1991. 40. Cunningham KM, Horowitz M, Read NW : The effect of short-term dietary supplementation with glucose on gastric empt ying in humans. Br J Nutr 65:15-19, 1991. 41. De Ponti F, Azpiroz F, Malagelada JR : Relaxatory responses of canine proximal stomach to esophageal and duodenal distensi on. Importance of vagal pathways. Dig Dis Sci 34:873-881, 1989. 42. de Zwart IM, Mearadji B, Lamb HJ, Eile rs PH, Masclee AA, de Roos A, Kunz P: Gastric motility: comparison of assessment with real-time MR imaging or barostat measurement initial experien ce. Radiology 224:592-597, 2002. 43. Doherty TJ, Andrews FM, Abraha TW, Osborne D, Frazier DL: Metoclopramide ameliorates the effects of endotoxin on gastric emptying of acetaminophen in horses. Can J Vet Res 63:37-40, 1999. 44. Doherty TJ, Andrews FM, Provenza MK, Frazier DL: Acetaminophen as a marker of gastric emptying in ponies Equine Vet J 30:349-351, 1998. 45. Doherty TJ, Andrews FM, Provenza MK, Frazier DL: The effect of sedation on gastric emptying of a liquid marker in ponies. Vet Surg 28:375-379, 1999. 46. Douglas BR, Jansen JB, de Jong AJ, Lame rs CB: Effect of va rious triglycerides on plasma cholecystokinin levels in rats. J Nutr 120:686-690, 1990. 47. Dumoulin V, Moro F, Barcelo A, Dakka T, Cuber JC: Peptide YY, glucagon-like peptide-1, and neurotensin re sponses to luminal factors in the isolated vascularly perfused rat ileum. Endo crinology 139:3780-3786, 1998. 48. Edelbroek M, Horowitz M, Maddox A, Be llen J: Gastric emptyi ng and intragastric distribution of oil in the presence of a liquid or a solid meal. J Nucl Med 33:12831290, 1992. 49. El Ouazzani T, Mei N: Vagal acidoa nd glucoreceptors in the gastro-duodenal region. Exp Brain Res 42:442-452, 1981. 50. Elashoff JD, Reedy TJ, Meyer JH: Analysis of gastric emptying data. Gastroenterology 83:1306-1312, 1982.

PAGE 180

164 51. Feinle C, Grundy D, Read NW: Effects of duodenal nutrients on sensory and motor responses of the human stomach to distension. Am J Physiol 273:G721-G726, 1997. 52. Forster ER, Green T, Elliot M, Bremner A, Dockray GJ: Gastric emptying in rats: role of afferent neurons and cholecyst okinin. Am J Physiol 258:G552-G556, 1990. 53. French SJ, Murray B, Rumsey RD, Fadz lin R, Read NW: Adaptation to high-fat diets: effects on eating be haviour and plasma cholecy stokinin. Br J Nutr 73:179189, 1995. 54. Fried GM, Ogden WD, Sakamoto T, Greeley GH, Thompson JC: Experimental evidence for a vagally mediated and chol ecystokinin-independent enteropancreatic reflex. Ann Surg 202:69-74, 1985. 55. Fried M, Schwizer W, Beglinger C, Kelle r U, Jansen JB, Lamers CB: Physiological role of cholecystokinin on postprandial in sulin secretion and gastric meal emptying in man. Studies with the cholecystoki nin receptor antagonist loxiglumide. Diabetologia 34:721-726, 1991. 56. Furness JB, Kunze WA, Clerc N: Nutrient tasting and signaling mechanisms in the gut. II. The intestine as a sensory organ: neural, endocrine, and immune responses. Am J Physiol 277:G922-G928, 1999. 57. Furuse M, Kato M, Yang SI, Asakura K, Okumura J: Influence of dietary protein concentrations or of duodenal amino acid infusion on cholecystokinin release in goats. Comp Biochem Physiol Comp Physiol 101:635-638, 1992. 58. Geor RJ, Harris PA, Hoekstra KE, Paga n JD: Effect of corn oil on solid-phase gastric emptying in horses. Proc 19th ACVIM, 853, 2001. 59. Ghoos YF, Maes BD, Geypens BJ, Mys G, Hiele MI, Rutgeerts PJ, Vantrappen G: Measurement of gastric emptying rate of solids by means of a carbon-labeled octanoic acid breath test. Gastroenterology 104:1640-1647, 1993. 60. Glatzle J, Kalogeris TJ, Zittel TT, Gu errini S, Tso P, Raybould HE: Chylomicron components mediate intestinal lipid-induced inhibition of gastric motor function. Am J Physiol Gastrointest Li ver Physiol 282:G86-G91, 2002. 61. Glatzle J, Wang Y, Adelson DW, Kal ogeris TJ, Zittel TT, Tso P, Wei JY, Raybould HE: Chylomicron components ac tivate duodenal vagal afferents via a cholecystokinin A receptor-mediated pathway to inhibit gastric motor function in the rat. J Physiol 550:657-664, 2003. 62. Gonzalez A, Mugueta C, Parra D, Labaye n I, Martinez A, Varo N, Monreal I, Gil MJ: Characterisation with stable isotopes of the presence of a lag phase in the gastric emptying of liquids. Eur J Nutr 39:224-228, 2000.

PAGE 181

165 63. Graham-Thiers PM, Kronfeld DS, Kline KA, Sklan DJ: Dietary protein restriction and fat supplementation diminish the acidogeni c effect of exercise during repeated sprints in horses. J Nutr 131:1959-1964, 2001. 64. Gregory PC, McFadyen M, Rayner DV: C ontrol of gastric emptying in the pig: influence of duodenal infusions of glucos e and emulsified fat. Q J Exp Physiol 74:109-119, 1989. 65. Guerin S, Ramonet Y, LeCloarec J, Meunier-Salaun MC, Bourguet P, Malbert CH: Changes in intragastric meal distribution are better predictors of gastric emptying rate in conscious pigs than are meal visc osity or dietary fibre concentration. Br J Nutr 85:343-350, 2001. 66. Guimbaud R, Moreau JA, Bouisson M, Dura nd S, Escourrou J, Vaysse N, Frexinos J: Intraduodenal free fatty acids rather th an triglycerid es are responsible for the release of CCK in humans. Pancreas 14:76-82, 1997. 67. Haba T, Sarna SK: Regulation of gastroduodenal emptying of solids by gastropyloroduodenal cont ractions. Am J Physiol 264:G261-G271, 1993. 68. Hammond CJ, Mason DK, Williams CA: Gastric ulceration in mature Thoroughbred horses. Eq Vet J 18:284-287. 1986. 69. Harris P: Energy sources and requirement s of the exercising horse. Annu Rev Nutr 17:185-210, 1997. 70. Hasegawa H, Shirohara H, Okabayashi Y, Nakamura T, Fujii M, Koide M, Otsuki M: Oral glucose ingestion stimulates c holecystokinin release in normal subjects and patients with non-insulin-dependent diabetes mellitus. Metabolism 45:196-202, 1996. 71. Heintze U, Birkhed D: Influence of a single intake of various test meals on secretion rate, buffer effect, and electrol ytes of human stimulated whole saliva. Caries Res 18:265-268, 1984. 72. Hirano T, Niijima A: Effects of 2-deoxy-Dglucose, glucose and insulin on efferent activity in gastric vagus nerv e. Experientia 36:1197-1198, 1980. 73. Holland JL, Kronfeld DS, Meacham TN: Be havior of horses is affected by soy lecithin and corn oil in the diet. J Anim Sci 74:1252-1255, 1996. 74. Holzer HH, Turkelson CM, Solomon TE, Raybould HE: Intestinal lipid inhibits gastric emptying via CCK and a vagal capsaic in-sensitive afferent pathway in rats. Am J Physiol 267:G625-G629, 1994. 75. Hopman WP, Jansen JB, Lamers CB: Co mparative study of the effects of equal amounts of fat, protein, and starch on plasma cholecystokinin in man. Scand J Gastroenterol 20:843-847, 1985.

PAGE 182

166 76. Horowitz M, Cunningham KM, Wishart JM Jones KL, Read NW: The effect of short-term dietary supplementation with glucose on gastric emptying of glucose and fructose and oral glucose tolerance in normal subjects. Diabetologia 39:481486, 1996. 77. Houghton LA, Mangnall YF, Read NW: Eff ect of incorporating fat into a liquid test meal on the relation between intragas tric distribution and gastric emptying in human volunteers. Gut 31:1226-1229, 1990. 78. Houghton LA, Read NW, Heddle R, Horowitz M, Collins PJ, Chatterton B, Dent J: Relationship of the motor activity of th e antrum, pylorus, and duodenum to gastric emptying of a solid-liquid mixed me al. Gastroenterology 94:1285-1291, 1988. 79. Hunt JN, Knox MT: A relation between the chain length of fatty acids and the slowing of gastric emptying. J Physiol 194:327-336, 1968. 80. Hunt JN, Smith JL, Jiang CL: Effect of meal volume and energy density on the gastric emptying of carbohydrates Gastroenterology 89:1326-1330, 1985. 81. Hunt JN, Stubbs DF: The volume and ener gy content of meals as determinants of gastric emptying. J Physiol 245:209-225, 1975. 82. Jain NK, Boivin M, Zinsmeister AR Brown ML, Malagelada JR, DiMagno EP: Effect of ileal perfusion of carbohydrates and amylase i nhibitor on gastrointestinal hormones and emptying. Gastroenterology 96:377-387, 1989. 83. Johnsen AH: Phylogeny of the cholecys tokinin/gastrin fam ily. Neuroendocrinology 19:73-99, 1998. 84. Johnsen AH, Sandin A, Rourke IJ, Bundg aard JR, Nilsson G, Rehfeld JF: Unique progastrin processing in equine G-cells s uggests marginal tyrosyl sulfotransferase activity. Eur J Biochem 255:432-438, 1998. 85. Johnson B, Carlson GP, Vatistas N, Snyde r JR, Lloyd K, Koobs J: Investigation of the number and location of gastric ulcerati ons in horses in race training submiteed to the California racehorse postmortem program. Proc 40th Annual AAEP Convention, 123-124, 1994. 86. Kaplan JM, Spector AC, Grill HJ: Dynami cs of gastric emptying during and after stomach fill. Am J Physiol 263:R813-R819, 1992. 87. Katschinski M, Schirra J, Begliner C, Langbein S, Wank U, D'Amato M, Arnold R: Intestinal phase of human antro-pylo ro-duodenal motility: cholinergic and CCKmediated regulation. Eur J Clin Invest 26:574-583, 1996. 88. Kitamura N, Yamada J, Calingasa n NY, Yamashita T: Immunocytochemical distribution of endocrine cells in the gastrointest inal tract of the horse. Equine Vet J 16:103-107, 1984.

PAGE 183

167 89. Kovacs TO, Lloyd KC, Walsh JH: Gastrin partially mediates insulin-induced acid secretion in dogs. Pep tides 17:583-587, 1996. 90. Kronfeld DS: Dietary fat affects heat production and other va riables of equine performance, under hot and humid cond itions. Equine Vet J Suppl:24-34, 1996. 91. Lam WF, Masclee AA, De Boer SY, Lame rs CB: Hyperglycaemia reduces gastrinstimulated gastric acid secretion in humans. Eur J Clin Invest 28:826-830, 1998. 92. Lam WF, Masclee AA, Muller ES, Lamers CB: Effect of hyperglycemia on gastric acid secretion and gastrin re lease induced by intravenous amino acids. Am J Clin Nutr 61:1268-1272, 1995. 93. Lam WF, Masclee AA, Muller ES, Lamers CB: Effect of hyperglycemia on gastric acid secretion during the gastric phase of diges tion. Am J Physiol 272:G1116G1121, 1997. 94. Lawrence LM: Nutrition and fuel utilizati on in the athletic horse. Vet Clin North Am Equine Pract 6:393-418, 1990. 95. Levan VH, Ghazi A, Duong TN: Effect s of nutrients on acid secretion by innervated total gastric pouch in cons cious rats. Exp Pa thol 22:225-229, 1982. 96. Lewis LD, Williams JA: Regulation of cholecystokinin secretion by food, hormones, and neural pathways in th e rat. Am J Physiol 258:G512-G518, 1990. 97. Li Y, Owyang C: Vagal afferent pa thway mediates physiological action of cholecystokinin on pancreatic enzyme s ecretion. J Clin Invest 92:418-424, 1993. 98. Liddle RA: Integrated actions of cholecy stokinin on the gastrointestinal tract: use of the cholecystokinin bi oassay. Gastroenterol Clin North Am 18:735-756, 1989. 99. Liddle RA, Goldfine ID, Rosen MS, Taplitz RA, Williams JA: Cholecystokinin bioactivity in human plasma. Molecular forms, responses to feeding, and relationship to gallbladder contrac tion. J Clin Invest 75:1144-1152, 1985. 100. Liddle RA, Green GM, Conrad CK, Willia ms JA: Proteins but not amino acids, carbohydrates, or fats stimulate cholecystokin in secretion in the rat. Am J Physiol 251:G243-G248, 1986. 101. Liddle RA, Rushakoff RJ, Morita ET, Beccaria L, Carter JD, Goldfine ID: Physiological role for cholecystokinin in reducing postprandial hyperglycemia in humans. J Clin Invest 81:1675-1681, 1988. 102. Lilja P, Wiener I, Inoue K, Greeley GH, Thompson JC: Release of cholecystokinin in response to food and intraduodenal fat in pigs, dogs and man. Surg Gynecol Obstet 159:557-561, 1984.

PAGE 184

168 103. Lin HC, Doty JE, Reedy TJ, Meyer JH: Inhibition of gastric emptying by glucose depends on length of intestine exposed to nutrient. Am J Physiol 256:G404-G411, 1989. 104. Lin HC, Doty JE, Reedy TJ, Meyer JH: Inhibition of gastric emptying by sodium oleate depends on length of intestine expos ed to nutrient. Am J Physiol 259:G1031G1036, 1990. 105. Lin HC, Elashoff JD, Gu Y, Meyer JH: E ffect of meal volume on gastric emptying. J Gastrointest Motil 4:157-163, 1992. 106. Lin HC, Elashoff JD, Gu YG, Meyer JH: Nutrient feedback i nhibition of gastric emptying plays a larger role than osmoti cally dependent duodenal resistance. Am J Physiol 265:G672-G676, 1993. 107. Lin HC, Kim BH, Elashoff JD, Doty JE Gu YG, Meyer JH: Gastric emptying of solid food is most potently inhibited by car bohydrate in the canine distal ileum. Gastroenterology 102:793-801, 1992. 108. Lin HC, Taylor IL: Release of peptide YY by fat in the proximal but not distal gut depends on an atropine-sensitive choliner gic pathway. Regul Pept 117:73-76, 2004. 109. Lloyd KC, Holzer HH, Zittel TT, Rayboul d HE: Duodenal lipid inhibits gastric acid secretion by vagal, capsaicin-sensitive afferent pathways in rats. Am J Physiol 264:G659-G663, 1993. 110. Lloyd KC, Wang J, Solomon TE: Acid inhi bition by intestinal nutrients mediated by CCK-A receptors but not plasma CCK. Am J Physiol Gastrointest Liver Physiol 281:G924-G930, 2001. 111. Lohmann KL, Bahr A, Cohen ND, B oothe DM, Roussel AJ: Evaluation of acetaminophen absorption in horses with experimentally induced delayed gastric emptying. Am J Vet Res 63:170-174, 2002. 112. Lohmann KL, Roussel AJ, Cohen ND, Boothe DM, Rakestraw PC, Walker MA: Comparison of nuclear scintigraphy and acetaminophen absorption as a means of studying gastric emptying in horses. Am J Vet Res 61:310-315, 2000. 113. Lorenzo-Figueras M, Jones G, Merritt AM Effects of various diets on gastric tone in the proximal portion of the stomach of horses. Am J Vet Res 63:1275-1278, 2002. 114. Lorenzo-Figueras M, Merritt AM: Effects of exercise on gastric volume and pH in the proximal portion of the stomach of horses. Am J Vet Res 63:1481-1487, 2002. 115. Mabayo RT, Furuse M, Yang SI, Okumur a J: Medium-chain triacylglycerols enhance release of cholecystokinin in chicks. J Nutr 122:1702-1705, 1992.

PAGE 185

169 116. MacGregor IL, Deveney C, Way LW Meyer JH: The effect of acute hyperglycemia on meal-stimulated gastric, biliary, and pancreatic secretion, and serum gastrin. Gastroenterology 70:197-202, 1976. 117. Maes B. Measurement of gastric emptying using dynamic breath analysis. Dissertation, University of Leuven, 1994. 118. Maes BD, Ghoos YF, Geypens BJ, Hi ele MI, Rutgeerts PJ: Relation between gastric emptying rate and rate of intr aluminal lipolysis. Gut 38:23-27, 1996. 119. Maes BD, Hiele MI, Geypens BJ, Ghoos YF, Rutgeerts PJ: Gastric emptying of the liquid, solid and oil phase of a meal in normal volunteers and patients with Billroth II gastrojejunostomy. Eur J Clin Invest 28:197-204, 1998. 120. Malbert CH, Mathis C: Antropyloric modula tion of transpyloric flow of liquids in pigs. Gastroenterol ogy 107:37-46, 1994. 121. Malbert CH, Mathis C, Guerin S, La place JP: Influence of jejunal nutrients on transpyloric flow and pyl oric resistance in pigs. Reprod Nutr Dev 37:411-425, 1997. 122. Masclee AA, Gielkens HA, Lam WF, De Boer SY, Lamers CD: Effects of parenteral nutrients on gastrointestinal motility and secretion. Scand J Gastroenterol Suppl 218:50-55, 1996. 123. Mathis C, Moran TH, Schwartz GJ: Lo ad-sensitive rat gastric vagal afferents encode volume but not gastric nutrien ts. Am J Physiol 274:R280-R286, 1998. 124. Mayer EA: The physiology of gastric storage and emptying. In Johnson LR (ed): Physiology of the gastrointestinal tract, ed 3rd. New York, Raven Press, 1994, 929976. 125. McClure SR, Glickman LT, Glickman NW : Prevalence of gastric ulcers in show horses. J Am Vet Med Assoc 215:1130-1133, 1999. 126. McHugh PR, Moran TH: Calories and gast ric emptying: a regulatory capacity with implications for feeding. Am J Physiol 236:R254-R260, 1979. 127. McLaughlin J, Grazia LM, Jones MN, D'Amato M, Dockray GJ, Thompson DG: Fatty acid chain length determines chol ecystokinin secretion and effect on human gastric motility. Gastroen terology 116:46-53, 1999. 128. McLaughlin JT, Lomax RB, Hall L, Do ckray GJ, Thompson DG, Warhurst G: Fatty acids stimulate cholecystokinin secre tion via an acyl chain length-specific, Ca2+-dependent mechanism in the entero endocrine cell line ST C-1. J Physiol 513 ( Pt 1):11-18, 1998.

PAGE 186

170 129. McLaughlin JT, Troncon LE, Barlow J Heggie LJ, Thompson DG: Evidence for a lipid specific effect in nutrient induce d human proximal gastric relaxation. Gut 43:248-251, 1998. 130. Medina B, Girard ID, Jacotot E, Ju lliand V: Effect of a preparation of Saccharomyces cerevisiae on microbial prof iles and fermentation patterns in the large intestine of horses fed a high fiber or a high starch di et. J Anim Sci 80:26002609, 2002. 131. Mei N: Intestinal chemosensitivity. In Holle GE, Wood JD (eds ): Advances in the Innervation of the Gastrointestinal Tract Amsterdam, Elsevier Science, 1992, 273281. 132. Mei N: Vagal glucoreceptors in the sm all intestine of the cat. J Physiol 282:485506, 1978. 133. Mei N, Garnier L: Osmosensitive vagal recep tors in the small inte stine of the cat. J Auton Nerv Syst 16:159-170, 1986. 134. Meisler SD, Doherty TJ, Andrews FM Osborne D, Frazier DL: Yohimbine ameliorates the effects of endotoxin on gastric emptying of the liquid marker acetaminophen in horses. Can J Vet Res 64:208-211, 2000. 135. Melone J: Vagal receptors sensitive to lip ids in the small intestine of the cat. J Auton Nerv Syst 17:231-241, 1986. 136. Melone J, Mei N: Intestin al effects of the products of lipid digestion on gastric electrical activity in the cat. Possible i nvolvement of vagal intestinal receptors sensitive to lipids. Gast roenterology 100:380-387, 1991. 137. Merritt AM: The Equine Stomach: A Personal Perspective (1963-2003). Proc 49th Annual AAEP Convention, 49:75-102, 2003. 138. Mesquita MA, Thompson DG, Troncon LE, D'Amato M, Rovati LC, Barlow J: Effect of cholecystokinin-A receptor blockade on lipid-induced gastric relaxation in humans. Am J Physiol 273:G118-G123, 1997. 139. Meyer H, Flothow C, Radicke S: Preileal digestibility of co conut fat and soybean oil in horses and their influence on metabo lites of microbial origin of the proximal digestive tract. Arch Ti erernahr 50:63-74, 1997. 140. Meyer JH, Kelly GA: Canine pancreatic re sponses to intestinally perfused proteins and protein digests. Am J Physiol 231:682-691, 1976. 141. Meyer JH, Mayer EA, Jehn D, Gu Y, Fi nk AS, Fried M: Gastric processing and emptying of fat. Gastro enterology 90:1176-1187, 1986.

PAGE 187

171 142. Miller LJ, Malagelada JR, Taylor WF, Go VL: Inte stinal control of human postprandial gastric function: the role of components of jejunoileal chyme in regulating gastric secreti on and gastric emptying. Ga stroenterology 80:763-769, 1981. 143. Molina F, Thiel T, Deutsch JA, Puerto A: Comparison between some digestive processes after eating and ga stric loading in rats. Ph armacol Biochem Behav 7:347350, 1977. 144. Moragas G, Azpiroz F, Pavia J, Mala gelada JR: Relations among intragastric pressure, postcibal perception, and ga stric emptying. Am J Physiol 264:G1112G1117, 1993. 145. Moran TH, Field DG, Knipp S, Carrigan TS, Schwartz GJ: Endogenous CCK in the control of gastric emptying of glucose and maltose. Peptides 18:547-550, 1997. 146. Moran TH, McHugh PR: Distinctions among three sugars in their effects on gastric emptying and satiety. Am J Physiol 241:R25-R30, 1981. 147. Moran TH, Smith GP, Hostetler AM, Mc Hugh PR: Transport of cholecystokinin (CCK) binding sites in subdiaphragmatic vagal branches. Brain Res 415:149-152, 1987. 148. Moran TH, Wirth JB, Schwartz GJ, Mc Hugh PR: Interactions between gastric volume and duodenal nutrients in the cont rol of liquid gastri c emptying. Am J Physiol 276:R997-R1002, 1999. 149. Moriarty P, Dimaline R, Thompson DG, Dockray GJ: Characterization of cholecystokininA and cholecystokininB receptors expressed by vagal afferent neurons. Neuroscience 79:905-913, 1997. 150. Morisset J, Julien S, Laine J: Localiza tion of cholecystokinin receptor subtypes in the endocine pancreas. J Histochem Cytochem 51:1501-1513, 2003. 151. Moriyasu M, Lee YL, Lee KY, Chang TM, Chey WY: Effect of digested protein on pancreatic exocrine secretion and gut hormone release in the dog. Pancreas 9:129-133, 1994. 152. Moukarzel AA, Sabri MT: Ga stric physiology and function: effects of fruit juices. J Am Coll Nutr 15:18S-25S, 1996. 153. Mundt MW, Hausken T, Samsom M: Eff ect of intragastric barostat bag on proximal and distal gastric accommodati on in response to liquid meal. Am J Physiol Gastrointest Liver Physiol 283:G681-G686, 2002. 154. Murray MJ, Schusser GF: Measurement of 24-h gastric pH using an indwelling pH electrode in horses unfed, fed and treated with ranitidine. Equine Vet J 25:417-421, 1993.

PAGE 188

172 155. Murray MJ, Schusser GF, Pipers FS, Gro ss SJ: Factors associ ated with gastric lesions in thoroughbred racehorses. Equine Vet J 28:368-374, 1996. 156. Nadeau JA, Andrews FM, Mathew AG, Argenzio RA, Blackford JT, Sohtell M, Saxton AM: Evaluation of diet as a cause of gastric ulcers in horses. Am J Vet Res 61:784-790, 2000. 157. Nadeau JA, Andrews FM, Patton CS, Argenzio RA, Mathew AG, Saxton AM: Effects of hydrochloric, acetic, butyric, and propionic acids on pathogenesis of ulcers in the nonglandular portion of the stomach of horses. Am J Vet Res 64:404412, 2003. 158. Orme CE, Harris RC, Marlin DJ, Hu rley J: Metabolic adaptation to fatsupplemented diet by the thoroughbred horse. Br J Nutr 78:443-458, 1997. 159. Owyang C, Louie DS, Tatum D: Feedb ack regulation of pa ncreatic enzyme secretion. Suppression of cholecystokinin release by trypsin. J Clin Invest 77:20422047, 1986. 160. Owyang C, Miller LJ, Malagelada JR Go VL: Nutrient and bowel segment dependency of human intestinal control of gastric secretion. Am J Physiol 243:G372-G376, 1982. 161. Paterson CA, Anvari M, Tougas G, Huiz inga JD: Determinants of occurrence and volume of transpyloric flow during gastri c emptying of liquids in dogs: importance of vagal input. Dig Dis Sci 45:1509-1516, 2000. 162. Rabuffo TS, Orsini JA, Sullivan E, Eng iles J, Norman T, Boston R: Associations between age or sex and prevalence of gast ric ulceration in Sta ndardbred racehorses in training. J Am Vet Med Assoc 221:1156-1159, 2002. 163. Randich A, Tyler WJ, Cox JE, Meller ST, Kelm GR, Bharaj SS: Responses of celiac and cervical vagal afferents to infusi ons of lipids in the jejunum or ileum of the rat. Am J Physiol Regul Inte gr Comp Physiol 278:R34-R43, 2000. 164. Raybould HE: Capsaicin-sensitive vagal a fferents and CCK in i nhibition of gastric motor function induced by intestinal nutrients. Peptides 12:1279-1283, 1991. 165. Raybould HE: Nutrient tasting and signali ng mechanisms in the gut. I. Sensing of lipid by the intestinal mucosa. Am J Physiol 277:G751-G755, 1999. 166. Raybould HE, Glatzle J, Robin C, Meye r JH, Phan T, Wong H, Sternini C: Expression of 5-HT3 receptors by extrin sic duodenal afferents contribute to intestinal inhibition of gastric emptying. Am J Physiol Gastrointest Liver Physiol 284:G367-G372, 2003.

PAGE 189

173 167. Raybould HE, Holzer H: Dual capsaicin -sensitive afferent pathways mediate inhibition of gastric emptyi ng in rat induced by intest inal carbohydrate. Neurosci Lett 141:236-238, 1992. 168. Raybould HE, Lloyd KC: Integration of pos tprandial function in the proximal gastrointestinal tract. Role of CCK a nd sensory pathways. Ann N Y Acad Sci 713:143-156, 1994. 169. Raybould HE, Meyer JH, Tabrizi Y, Li ddle RA, Tso P: Inhibition of gastric emptying in response to intestinal lipid is dependent on chylomicron formation. Am J Physiol 274:R1834-R1838, 1998. 170. Raybould HE, Tache Y: Cholecystokinin in hibits gastric motility and emptying via a capsaicin-sensitive vagal pathway in rats. Am J Physiol 255:G242-G246, 1988. 171. Raybould HE, Zittel TT: Inhibition of gast ric motility induced by intestinal glucose in awake rats: role of Na(+)-glucose co-transporter. Neurogastroenterol Motil 7:914, 1995. 172. Read NW: Feedback regulation and se nsation. Dig Dis Sci 39:37S-40S, 1994. 173. Read NW, Houghton LA: P hysiology of gastric empt ying and pathophysiology of gastroparesis. Gastroenterol Clin North Am 18:359-373, 1989. 174. Reubi JC, Waser B, Gugger M, Friess H, Kleeff J, Kayed H, Buchler MW, Laissue JA: Distribution of CCK1 and CCK2 recep tors in normal and diseased human pancreatic tissue. Gastro enterology 125:98-106, 2003. 175. Richards W, Hillsley K, Eastwood C, Grundy D: Sensitivity of vagal mucosal afferents to cholecystokinin and its role in afferent signal trans duction in the rat. J Physiol 497 ( Pt 2):473-481, 1996. 176. Rigaud D, Paycha F, Meulemans A, Me rrouche M, Mignon M: Effect of psyllium on gastric emptying, hunger feeling and f ood intake in normal volunteers: a double blind study. Eur J Clin Nutr 52:239-245, 1998. 177. Ringger NC, Lester GD, Neuwirth L, Merr itt AM, Vetro T, Harrison J: Effect of bethanechol or erythromycin on gastric emptying in horses. Am J Vet Res 57:17711775, 1996. 178. Ritter RC, Covasa M, Matson CA: C holecystokinin: proofs and prospects for involvement in control of food intake and body weight. Neuropeptides 33:387-399, 1999. 179. Ritter RC, Ladenheim EE: Capsaicin pr etreatment attenuates suppression of food intake by cholecystokinin. Am J Physiol 248:R501-R504, 1985.

PAGE 190

174 180. Robertson MD, Mathers JC: Gastric empt ying rate of solids is reduced in a group of ileostomy patients. Dig Dis Sci 45:1285-1292, 2000. 181. Ropert A, des Varannes SB, Bizais Y, Roze C, Galmiche JP: Simultaneous assessment of liquid emptying and pr oximal gastric tone in humans. Gastroenterology 105:667-674, 1993. 182. Ross MW, Donawick WJ, Sellers AF, Lo we JE: Normal motility of the cecum and right ventral colon in ponies Am J Vet Res 47:1756-1762, 1986. 183. Ruppin H: Current aspects of intestinal motility and transport. Klin Wochenschr 63:679-688, 1985. 184. Sandin A, Andrews FM, Nadeau JA, Nil sson G: Effect of nervous excitation on acid secretion in horses. Acta Physiol Scand 168:437-442, 2000. 185. Sandin A, Girma K, Sjo holm B, Lindholm A, Nilsson G: Effects of differently composed feeds and physical stress on plasma gastrin concentrati on in horses. Acta Vet Scand. 39:265-272, 1998. 186. Santangelo A, Peracchi M, Conte D, Fraque lli M, Porrini M: Phys ical state of meal affects gastric emptying, cholecystokinin re lease and satiety. Br J Nutr 80:521-527, 1998. 187. Scarpignato C, Galmiche JP: Simultane ous measurement and recording of gastric potential difference and intragastric pH in man. Dig Dis 8 Suppl 1:60-70, 1990. 188. Schott HC, Hinchcliff KW: Treatments affe cting fluid and electrolyte status during exercise. Vet Clin North Am E quine Pract 14(1):175-204, 1998. 189. Schwartz GJ, McHugh PR, Moran TH: Ph armacological dissociation of responses to CCK and gastric loads in rat mechanosensitive vagal afferents. Am J Physiol 267:R303-R308, 1994. 190. Schwartz GJ, Moran TH: Duodenal nutrien t exposure elicits nutrient-specific gut motility and vagal afferent signals in rat. Am J Physiol 274:R1236-R1242, 1998. 191. Schwizer W, Fraser R, Bo rovicka J, Crelier G, Boesig er P, Fried M: Measurement of gastric emptying and gastric motility by magnetic resonance imaging (MRI). Dig Dis Sci 39:101S-103S, 1994. 192. Scott BD, Potter GD, Greene LW, Vogels ang MM, Anderson JG: Efficacy of fatsupplemented diet to reduce thermal stress in exercising Thoroughbred horses. Proc 13th Equine Nutrition and Phys iology Symposium, 66, 1993. 193. Sharara AI, Bouras EP, Misukonis MA, Liddle RA: Evidence for indirect dietary regulation of cholecystokinin release in rats. Am J Physiol 265:G107-G112, 1993.

PAGE 191

175 194. Siciliano PD: Nutrition and feeding of the geriatric horse. Vet Clin North Am Equine Pract 18:491-508, 2002. 195. Siegel JA, Urbain JLC, Alder LP, Char kes ND, Maurer AH, Krevsky B, Knight LC, Fisher RS, Malmud LS: Biphasic nature of gastric emptying. Gut 29:85-89, 1998. 196. Siegle ML, Schmid HR, Ehrlein HJ: Effect s of ileal infusions of nutrients on motor patterns of canine small intestine. Am J Physiol 259:G78-G85, 1990. 197. Sojka JE, Cantwell HD: The effect of di etary composition on gastric emptying rate in ponies. Proc 3rd Eq Colic Res Symposium, 24, 1988. 198. Sonobe K, Sakai T, Satoh M, Haga N, It oh Z: Control of gallbladder contractions by cholecystokinin through cholecystokininA receptors in the vagal pathway and gallbladder in the dog. Re gul Pept 60:33-46, 1995. 199. Sosa Leon LA, Hodgson DR, Rose RJ: Gastric emptying of oral rehydration solutions at rest and after exercise in horses. Res Vet Sci 63:183-187, 1997. 200. Spiller RC, Trotman IF, Adrian TE, Bl oom SR, Misiewicz JJ, Silk DB: Further characterisation of the 'ileal brake' reflex in man-effect of ileal infusion of partial digests of fat, protein, and starch on je junal motility and release of neurotensin, enteroglucagon, and peptide YY. Gut 29:1042-1051, 1988. 201. Stephens JR, Woolson RF, Cooke AR: Osmoltye and tryptophan receptors controlling gastric emptying in the dog. Am J Physiol 231:848-853, 1976. 202. Sternini C, Wong H, Pham T, De Giorgio R, Miller LJ, Kuntz, SM, Reeve JR, Walsh JH, Raybould HE: Expression of c holecystokinin A receptors in neurons innervating the rat stomach and intest ine. Gastroenterology 117:1136-1146, 1999. 203. Straathof JW, Mearadji B, Lamers CB, Masclee AA: Effect of CCK on proximal gastric motor function in humans Am J Physiol 274:G939-G944, 1998. 204. Sun G, Chang TM, Xue WJ, Wey JF, Lee KY, Chey WY: Release of cholecystokinin and secretin by sodium oleate in dogs: molecular form and bioactivity. Am J Physiol 262:G35-G43, 1992. 205. Sutton DG, Bahr A, Preston T, Christle y RM, Love S, Roussel AJ: Validation of the 13C-octanoic acid breath test for meas urement of equine gastric emptying rate of solids using radioscintigraphy. Equine Vet J 35:27-33, 2003. 206. Symonds E, Butler RN, Omari TI: Asse ssment of gastric emptying in the mouse using the [13C]-octanoic acid breath test Clin Exp Pharmacol Physiol 27:671-675, 2000. 207. Tack J: The physiology and the path ophysiology of the gastric accommodation reflex in man. Verh K Acad Geneeskd Belg 62:183-207, 2000.

PAGE 192

176 208. Thimister PW, Hopman WP, Sloots CE, Rosenbusch G, Willems HL, Trijbels FJ, Jansen JB: Role of intraduodenal prot eases in plasma cholecystokinin and pancreaticobiliary responses to protei n and amino acids. Gastroenterology 110:567575, 1996. 209. Tohno H, Sarr MG, DiMagno EP: Intr aileal carbohydrate regulates canine postprandial pancreaticobiliary secretion and upper gut motility. Gastroenterology 109:1977-1985, 1995. 210. Treacy PJ, Jamieson GG, Dent J: The importance of the pylorus as a regulator of solid and liquid emptying from the stom ach. J Gastroenterol Hepatol 10:639-645, 1995. 211. Tso P: Intestinal lipid absorption. In Johnson LR (ed): Physiology of the gastrointestinal tract, ed. 3rd. New York, Raven Press, 1996, 1867-1907. 212. Tso P, Balint JA: Formation and transpor t of chylomicrons by enterocytes to the lymphatics. Am J Physiol 250:G715-G726, 1986. 213. Undeland KA, Hausken T, Gilja OH, Rope rt R, Galmiche JP, Berstad A: Gastric relaxation in response to a soup meal in h ealthy subjects. A study using a barostat in the proximal stomach. Scand J Gastroenterol 30:1069-1076, 1995. 214. Valentine BA, Van Saun RJ, Thompson KN, Hintz HF: Role of dietary carbohydrate and fat in horses with equine polysaccharide storage myopathy. J Am Vet Med Assoc 219:1537-1544, 2001. 215. Valk N, Doherty TJ, Blackford JT, Abra ha TW, Frazier DL: Effect of cisapride on gastric emptying in horses following e ndotoxin treatment. Equine Vet J 30:344348, 1998. 216. Valk N, Doherty TJ, Blackford JT, Abraha TW, Frazier DL: Phenylbutazone prevents the endotoxin-induced delay in ga stric emptying in horses. Can J Vet Res 62:214-217, 1998. 217. van der Schaar PJ, Bremer Y, Lamers CB, Masclee AA: Role of cholecystokinin in relaxation of the proximal stomach. Scand J Gastroenterol 36:361-366, 2001. 218. Vanden Berghe PD, Vos R, Janssens J, Tack JF: Site of triggering of the accommodation reflex in man. Gast roenterology 118(4):AGA A185, 2000. 219. Vantrappen G: Methods to study gastric emptying. Dig Dis Sci 39:91S-94S, 1994. 220. Varner AA, Isenberg JI, Elashoff JD, Lamers CB, Maxwell V, Shulkes AA: Effect of intravenous lipid on gastric acid s ecretion stimulated by intravenous amino acids. Gastroenterology 79:873-876, 1980.

PAGE 193

177 221. Vatistas NJ, Sifferman RL, Holste J, C ox JL, Pinalto G, Schultz KT: Induction and maintenance of gastric ulceration in horses in simulated race training. Equine Vet J Suppl:40-44, 1999. 222. Vatistas NJ, Snyder JR, Carlson G, Johnson B, Arthur RM, Thurmond M, Zhou H, Lloyd KL: Cross-sectional study of gastri c ulcers of the squamous mucosa in thoroughbred racehorses. Equine Vet J Suppl:34-39, 1999. 223. Vatistas NJ, Snyder JR, Carlson G, Johnson P, Arthur RM, Thurmond M., Lloyd KCK: Epidemiological study of gastric ulceration in the Thoroughbred race horse. Proc 40th Annual AAEP Convention, 125-126, 1994. 224. Vittoria A, Cocca T, La Mura E, Cecio A: Immunocytochemistry of paraneurons in the female urethra of the horse, cattle sheep, and pig. Anat Rec 233:18-24, 1992. 225. Webber DE, Morrissey SM: Separate hypogl ycaemic thresholds for gastric acid and pepsin secretion following insulin or posterior hypothalamic stimulation. Can J Physiol Pharmacol 57:1283-1288, 1979. 226. Weber E, Ehrlein HJ: Relationships between gastric emptying and intestinal absorption of nutrients and energy in mini pigs. Dig Dis Sci 43:1141-1153, 1998. 227. Welch IM, Cunningham KM, Read NW: Re gulation of gastric emptying by ileal nutrients in humans. Gast roenterology 94:401-404, 1988. 228. Wells AS, Read NW, Uvnas-Moberg K, Alster P: Influences of fat and carbohydrate on postprandial sleepiness, mood, and hormones. Physiol Behav 61:679-686, 1997. 229. Whitehead W, Delvaux M: Standardizat ion of barostat procedures for testing smooth muscle tone and sensory thresholds in the gastrointestinal tract. Dig Dis Sci 42:223-241, 1997. 230. Widdop RE, Krstew E, Mercer LD, Carlberg M, Beart PM, Jarrott B: Electrophysiological and autoradiogra phical evidence for cholecystokinin A receptors on rat isolated nodose ganglia J Auton Nerv Syst 46:65-73, 1994. 231. Williams CA, Kronfeld DS Staniar WB, Harris PA: Plasma glucose and insulin responses of Thoroughbred mares fed a meal high in starch and sugar or fat and fiber. J Anim Sci 79(8):2196-2201, 2001. 232. Wollaeger EE: Editorial: Fat, feces, a nd the importance of the ileum. Mayo Clin Proc 48:833-835, 1973. 233. Wyse CA, Murphy DM, Preston T, Morrison DJ, Love S: Assessment of the rate of solid-phase gastric emptying in ponies by means of the 13C-octanoic acid breath test: a preliminary study. Eq Vet J 33:197-203, 2001.

PAGE 194

178 234. Wyse CA, Murphy DM, Preston T, Sutt on DG, Morrison DJ, Christley RM, Love S: The(13)C-octanoic acid breath test fo r detection of effect s of meal composition on the rate of solid-phase gastric emptyi ng in ponies. Res Vet Sci 71:81-83, 2001. 235. Yamagisi T, Debas HT: Cholecystokin in inhibits gastric emptying by acting on both proximal stomach and pylorus. Am J Physiol 234(4):E375-E378, 1978. 236. Yang SI, Furuse M, Muramatsu T, Okumura J: Enhanced release of cholecystokinin by dietary amino acids in chicks (Gallus domesticus). Comp Biochem Physiol A 92:319-322, 1989.

PAGE 195

179 BIOGRAPHICAL SKETCH Mireia Lorenzo-Figueras was born in Barcelona, Spain, on March 26th, 1973. She obtained her degree in Veterinary Medi cine at the Autonomous University of Barcelona (UAB) in 1996. After attending vete rinary school, Mireia spent a period of training at several equine hos pitals in Spain, Ireland and the United States. She then completed an internship in equine medici ne and surgery at the Veterinary Teaching Hospital of the UAB. In the process of a pplying for a Fulbright fellowship, Mireia worked in a small animal hospital for six m onths. Finally, she received a fellowship to spend one year as a visiting re searcher under the supervision of Dr. A.M. Merritt at the University of Florida in October of 1999. In 2001 Mireia became a graduate student at the College of Veterinary Medicine at the UF After graduation, she will start a residency in equine internal medicine at UC Davis.


Permanent Link: http://ufdc.ufl.edu/UFE0004394/00001

Material Information

Title: Effects of High-Fat vs. High-Carbohydrate Diets on Proximal Gastric Relaxation, Gastric Emptying, pH of Gastric Contents and Plasma Cholecystokinin in the Horse
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0004394:00001

Permanent Link: http://ufdc.ufl.edu/UFE0004394/00001

Material Information

Title: Effects of High-Fat vs. High-Carbohydrate Diets on Proximal Gastric Relaxation, Gastric Emptying, pH of Gastric Contents and Plasma Cholecystokinin in the Horse
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0004394:00001


This item has the following downloads:


Full Text











EFFECTS OF HIGH-FAT vs. HIGH-CARB OHYDRATE DIET S ON PROXIMAL
GASTRIC RELAXATION, GASTRIC EMPTYING, pH OF GASTRIC CONTENTS
AND PLASMA CHOLECYSTOKININ IN THE HORSE














By

MIREIA LORENZO-FIGUERAS


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2004
































Copyright 2004

by

Mireia Lorenzo-Figueras

































Dedicated to my father.
















ACKNOWLEDGMENTS

I would like to thank Dr. A. M. Merritt, my mentor and guide over these years, for

his patient support and scientific guidance. Not only has he given me uncountable

opportunities to grow professionally, but also has helped me feel comfortable since I

arrived in Gainesville, more than four years ago. Special gratitude is also extended to my

supervisory committee members, Drs. Colin Burrows, Richard Johnson, Robert MacKay,

and Edgar Ott, for willingly sharing their knowledge and for reviewing the dissertation. I

am particularly grateful to Dr. Burrows for his advice on the very first manuscript I

submitted for publication.

For their valuable assistance and guidance in the breath test studies, I would like to

thank Drs. Tom Preston and David Sutton. It is also a pleasure to acknowledge Dr. Jean

Morisset, who shared his expertise on cholecystokinin and kindly performed the bioassay

of this study. Acknowledgments are also extended to Drs. Chris Sanchez and Yong Bai,

Hilken Kuck and Jim Burrow for their valuable time and help. Additional thanks go to

Dr. Murray Brown, Kelly Merritt and Dr. Charles Woods for permitting me to use the

SpeedVac concentrator, Dr. Daniel C. Sharp for his assistance with the gamma counter,

and Mrs. Marty Johnson for guidance on the RIA technique.

I cannot forget to mention all the horses that have participated in this study, for

being patient with me, and for teaching me the importance of patience, care, and

gratitude.









I am deeply thankful to the Office of Research and Graduate Studies at the College

of Veterinary Medicine, Associate Dean Charles Courtney III, and Mrs. Sally O'Connell

for their support throughout my program. I would also like to thank all the staff of the

Department of Large Animal Clinical Sciences and the Deedie Wrigley-Hancock

Fellowship for Equine Colic Research for their Einancial support, and the Florida Pari-

mutuel Wagering Trust Fund and USA Equestrian Federation for providing funding for

this study. Also, I express my deepest gratitude to the Fulbright Commission of Spain,

which allowed me to take the first step of my American adventure.

Finally, I am extremely grateful to my family for their love and support, and for

understanding patiently all these years I have spent on the other side of the ocean. I am

also thankful to all the friends I have met in Gainesville. My life here would not have

been so enj oyable without them.


















TABLE OF CONTENTS


page

ACKNOWLEDGMENT S .............. .................... iv

LI ST OF T ABLE S ................. .............. ix...___ ....

LIST OF FIGURES ............_...... .__ ..............xiii...

AB S TRAC T ......_ ................. ............_........x

CHAPTER

1 LITERATURE REVIEW ................. ...............1...............


Fat Supplementation: A New Concept in Diet Formulation .............. ....................1
The Effect of Diet on Gastric Physiology ........_................. .. ......._...__.....3
Dietary Regulation of Gastric Emptying............_..._ ............ ........._...__......3
Meal consistency: emptying of liquids versus solids ..........._.... .........._..... .4
Meal composition: nutrient modulation of gastric emptying ........._..._............5
Nutrient Sensing Initiates Nutrients Regulation of Gastric Emptying ..................6
Existence of nutrient-specific chemoreceptors ................. ......................7
Sensing of dietary fat............... ...............8...
Sensing of dietary carbohydrates .............. .............. ........10
Fat versus carbohydrate on control of gastric emptying .............. ..... ..........12
Effect of Diet on the Meal-Induced Relaxation of the Proximal Stomach..........12
Nutrients Regulate Gastric Emptying by Neural and Hormonal Pathways ........14
Neural pathways ................. ...............14.................
Horm onal pathways............... ..... ... ...... ...................1
Cholecystokinin: A Key Hormone in the Regulation of Gastric Emptying........16
Gastric Emptying in the Horse: Current Knowledge and Methodology ....................19
The Effect of Diet on Intragastric pH ................. ........... ..... .......... ..... 2
Squamous Ulceration of the Proximal Stomach ................. ................. ...._23
Studies on the Effect of Diet on Intragastric pH in the Horse ................... ..........24
Study Objectives............... ...............2

2 MATERIALS AND METHODS .............. ...............25....

Animals............... ...............25
B arostat .........._... .. ... ... ..._.... ...............25.....
13C-Octanoic Acid Breath Test ........._...... ...............27..__._. .....
Intra gastric pH M monitoring ................. ...............28................












Plasma Cholecystokinin Radioimmunoassay ................. ..............................28
Test Meals Composition ................. ...............30................
Phase I Studies............... ...............3 0
Phase II Studies .............. ...............3 1....

Study Design............... ...............32.
Phase I Studies............... ...............32
Phase II Studies .............. ...............33....
Experimental Procedure............... ...............3
Data Analy sis............... ...............3
Ingestion Time ................. ...............3.. 8..............
Proximal Gastric Compliance............... ..............3
Intra gastric pH ................. ...............39.......... ......
Breath Samples ................. ............ ...............39 .....
Calculation of gastric emptying parameters ................. .......................39
Effect of diet on basal 13C Output ................. ............ ............. .....4
Effect of the octanoi c aci d-loaded di ets on 130 Output ............... ..............40


3 RESULT S AND DISCUS SION TONE OF THE PROXIMAL STOMACH.._.....41


Effect of 13C-Octanoic Acid Labeling (Breath Test) ........._.___..... .___ ..............42
Phase I Studies: Pelleted Diets .............. ...............42....
Phase II Studies: Sweet feed Diets ................. .... ............. ............4
Influence of Dietary Composition on the Effect of 13C-Octanoic Acid
Labeling .............. .. ...............45...
Effect of Dietary Composition................................. .......4
Phase I. Pelleted Meals: Fat Versus Carbohydrate ................. ......._._. ........46
Phase II. Sweet feed Meals: Corn Oil Versus Glucose .............. ...................52
Unlabeled meals .............. ...............53...
Octanoic-acid labeled meals............... ...............58.
Conclusions............... ..............6
Methodology ................. ...............61.................
Relaxation Response .............. ...............62....


4 RESULTS AND DISCUS SION GASTRIC EMPTYING ................. ................. 66


Effect of Dietary Composition on Gastric Emptying ................. .......................67
Phase I. Pelleted Meals: Fat Versus Carbohydrate ........._.._.. ... .._.._...........67
Effect of diet on basal 13C expiratory output ........._._ ..... ..._._...........67
Effect of diet on gastric emptying .........._... ........._. .........._.........6
Phase II. Sweet feed Meals: Corn Oil Versus Glucose .............. ....................75
Breath tests without presence of an intragastric barostat bag ......................75
Breath tests with presence of an intragastric barostat bag ...........................77
Effect of the Barostat Bag on Gastric Emptying .................. .. .......... ................. 80
Relation Between Proximal Gastric Relaxation and Gastric Emptying ................... ..80
Conclusions............... ..............8












5 RESULTS AND DISCUSSION -pH OF GASTRIC CONTENTS ................... ........83


Effect of Dietary Composition on Intragastric pH .......... ................ ...............84
Phase I. Pelleted Meals: Fat Versus Carbohydrate ................. ......................84
Phase II. Sweet Feed Meals: Corn Oil Versus Glucose ................... ...............88
Conclusions............... ..............9


6 RESULTS AND DISCUSSION CHOLECYSTOKININ LIKE-ACTIVITY.........93


Radioimmunoassay ................. ...............93.................
Conclusions............... ..............10


7 SUMMARY AND CONCLUSIONS ................ ...............102...............


APPENDIX


A INDIVIDUAL ANIMAL DATA .............. ...............107....


B STATISTICAL TESTS FOR BAROSTAT DATA ......____ ..... ... .._............125


C STATISTICAL TESTS FOR GASTRIC EMPTYING DATA. .............. ..... ..........150


D STATISTICAL TESTS FOR PH DATA ................. ...............153..............


LIST OF REFERENCES ................. ...............160................


BIOGRAPHICAL SKETCH ................. ...............179......... ......

















LIST OF TABLES


Table pg

2-1 Approximate composition of the pelleted test meals used in the Phase I
studies. ............. ...............3 1....

2-2 Composition of the sweet feed meal (Seminole Feed, Blue Ribbon 10) used in
the Phase II studies............... ...............32

3-1 Comparison of duration of meal ingestion between the present study (A) and
that of Lorenzo-Figueras and Merrittl13 (B).. ............ ...... ............... 4

3-2 Postprandial variations in volume* of an intragastric bag controlled by an
el ectronic barostat. .............. ...............60....

4-1 Comparison of gastric emptying parameters determined by the 13C-Octanoate
breath test ................. ...............69.................

5-1 Changes in intragastric pH after ingestion of a 0.5 g/kg sweet feed meal alone
(control meal) or enriched with either corn oil or glucose............... .................8

A-1 Individual body weight, test meal weight (Phase I and II), breath test label
composition, and sweet feed supplementation (Phase II). ........._.... ........._....107

A-2 Barostat raw data (Phase I): bag volumes for baseline and 2-min postprandial
blocks after ingestion of the unlabeled high-fat pelleted meal acid. ................... .107

A-3 Barostat raw data (Phase I): bag volumes for baseline and 2-min postprandial
blocks after ingestion of the labeled high-fat pelleted meal. ............. ..... ........._.108

A-4 Barostat raw data (Phase I): bag volumes for baseline and 2-min postprandial
blocks after ingestion of the unlabeled high-carbohydrate pelleted meal............109

A-5 Barostat raw data (Phase I): bag volumes for baseline and 2-min postprandial
blocks after ingestion of the labeled high-carbohydrate pelleted meal ................11 1

A-6 Barostat raw data (Phase I): ingestion time (sec) for the pelleted diets. ..............112

A-7 Barostat raw data (Phase II): bag volumes for baseline and 2-min postprandial
blocks after ingestion of the unlabeled corn oil-enriched sweet feed meal. ........1 12










A-8 Barostat raw data (Phase II): bag volumes for baseline and 2-min postprandial
blocks after ingestion of the labeled corn oil-enriched sweet feed meal. ............1 13

A-9 Barostat raw data (Phase II): bag volumes for baseline and 2-min postprandial
blocks after ingestion of the unlabeled glucose-enriched sweet feed meal. ........1 14

A-10 Barostat raw data (Phase II): bag volumes for baseline and 2-min postprandial
blocks after ingestion of the labeled glucose-enriched sweet feed meal. ............1 16

A-11 Barostat raw data (Phase II): bag volumes for baseline and 2-min postprandial
blocks after ingestion of the labeled control sweet feed meal. ................... .........1 17

A-12 Barostat raw data (Phase II): ingestion time (sec) for the sweet feed diets. ........118

A-13 Gastric emptying raw data (Phase I): parameters of the pelleted diets. ...............1 18

A-14 Gastric emptying raw data (Phase II): parameters of the sweet feed diets..........1 19

A-15 Intragastric pH raw data (Phase I): mean pH for baseline and 5-min
postprandial blocks after ingestion of the unlabeled high-fat pelleted meal. ......1 19

A-16 Intragastric pH raw data (Phase I): mean pH for baseline and 5-min
postprandial blocks after ingestion of the labeled high-fat pelleted meal. ..........120

A-17 Intragastric pH raw data (Phase I): mean pH for baseline and 5-min
postprandial blocks after ingestion of the unlabeled high-CHO pelleted meal. ..121

A-18 Intragastric pH raw data (Phase I): mean pH for baseline and 5-min
postprandial blocks after ingestion of the labeled high-CHO pelleted meal. ......121

A-19 Intragastric pH raw data (Phase II): mean pH for baseline and 5-min
postprandial blocks after ingestion of the labeled corn oil-enriched meal. .........122

A-20 Intragastric pH raw data (Phase II): mean pH for baseline and 5-min
postprandial blocks after ingestion of the labeled corn oil-enriched meal. .........123

A-21 Intragastric pH raw data (Phase II): mean pH for baseline and 5-min
postprandial blocks after ingestion of the labeled glucose-enriched meal. .........123

B-1 ANOVA and CL for ingestion times of unlabeled pelleted meals (Phase I).......125

B-2 Shapiro-Wilk test for normality of ingestion times of unlabeled pelleted
meals (Phase I). .......... ......__ ...............125..

B-3 Bartlett's test for homogeneity of variance of unlabeled pelleted meals
(Phase I). ............. ...............125....

B-4 ANOVA mixed procedure for mean bag volumes of pelleted meals (Phase I). .125










B-5 ANOVA mixed procedure for mean bag volume minus baseline of pelleted
meals (Phase I) ................. ...............130................

B-6 ANOVA and CL for ingestion times of unlabeled sweet feed meals
(Phase II). ................. ...............134_._._.......

B-7 Shapiro-Wilk test for normality of ingestion times of unlabeled sweet feed
meals (Phase II)............... ...............134..

B-8 Bartlett's test for homogeneity of variance of unlabeled sweet feed meals
(Phase II). ................. ...............134._.._._ ......

B-9 ANOVA mixed procedure for mean bag volumes of unlabeled sweet feed
meals (Phase II)............... ...............135..

B-10 ANOVA mixed procedure for mean bag volume minus baseline of unlabelled
sweet feed meals (Phase II) ................. ...............137........... ...

B-11 Bartlett's test for homogeneity of variance for ingestion times of labeled
sweet feed meals (Phase II) ................. ...............139........... ...

B-12 Friedman's 2-way ANOVA for ingestion times of labeled sweet feed meals.....139

B-13 ANOVA mixed procedure for mean bag volumes of labeled sweet feed meals
(Phase II) ................. ...............139................

B-14 ANOVA mixed procedure for mean bag volume minus baseline of labelled
sweet feed meals (Phase II) ................. ...............143........... ...

B-15 ANOVA mixed procedure for mean bag volume minus baseline of unlabeled
(BFAT) and labeled (FFAT) corn oil-enriched sweet feed meals (Phase II). .....146

B-16 ANOVA mixed procedure for mean bag volume minus baseline of unlabeled
(BCHO) and labeled (FCHO) glucose-enriched sweet feed meals (Phase II).....147

C-1 Shapiro-Wilk test for normality of Phase I parameters ................. ................. 150

C-2 Bartlett's test for homogeneity of variance of Phase I parameters. ................... 150

C-3 Two-sample t-test for Phase I parameters ................. .............................150

C-4 Shapiro-Wilk test for normality of Phase II parameters without an intragastric
bag ................. ...............150................

C-5 Bartlett's test for homogeneity of variance of Phase II parameters without an
intragastric bag ................. ...............150................

C-6 Paired-sample t-test for Phase II parameters without an intragastric bag............151










C-7 Shapiro-Wilk test for normality of Phase II parameters with an intragastric
bag ................. ...............151................

C-8 Bartlett's test for homogeneity of variance of Phase II parameters with an
intragastric bag ................. ...............151................

C-9 Repeated measures ANOVA for Phase II tl1/2 with an intragastric bag. ..............151

C-10 Friedman' s 2-way ANOVA for Phase II tmax with intragastric bag. ................... .151

C-11 Repeated measures ANOVA for Phase II GEC with an intragastric bag............152

C-12 Shapiro-Wilk test for normality of Phase II parameters: effect of an
intragastric bag on gastric emptying ................. ...............152........... ...

C-13 Bartlett's test for homogeneity of variance of Phase II parameters: effect of
an intragastric bag on gastric emptying. ............. ...............152....

C-14 Paired-sample t-test for Phase II parameters: effect of an intragastric bag on
gastric emptying ................. ...............152................

D-1 ANOVA mixed procedure for mean intragastric pH of pelleted meals
(Phase I). ............. ...............153....

D-2 ANOVA mixed procedure for mean intragastric pH of labeled sweetfeed
meal s (Phase II)............... ...............155..

















LIST OF FIGURES


Figure pg

2-1 Motility of the stomach measured by an electronic barostat ................. ................26

2-2 Polyester bag connected to a pressure line and an inflation line. ................... ..........27

2-3 Endoscopic view of the squamous mucosa of the proximal stomach. ................... ..34

2-4 Positioning of the polyester bag used to measure intragastric pressure in the
proximal stomach. ............. ...............34.....

2-5 Endoscopic views showing correct position of the barostat bag within the
proximal stomach. ............. ...............3 5....

2-6 Motility of the proximal stomach was measured with an intragastric bag,
inserted through the gastric cannula and connected to the electronic barostat by
two separate catheters............... ...............3

2-7 The pH of gastric contents was measured using a pH electrode inserted through
the gastric cannula and positioned in the most ventral part of the stomach. ............37

3-1 Changes in intragastric bag volume after ingestion of the high-carbohydrate
(CHO) pelleted meal with and without addition of 13C-Octanoic acid.. ...................43

3-2 Mean bag volume revealing the effect of intake of a high-fat (A) and a high-
carbohydrate (B) pelleted diets (0.5 g/kg) on gastric tone of the proximal
portion of the stomach in 6 horses. ............. ...............47.....

3-3 Changes in intragastric bag volume after ingestion of either the high-fat meal
or the high-carbohydrate (CHO) pelleted meal ................. ......... ................48

3-4 Mean volume trace (n=6) of the effect of ingestion of different sweet feed meals
(0.5 g/kg) on baseline tone in the proximal stomach. ............. .....................5

3-5 Changes in intragastric bag volume after ingestion of a 10% protein sweet feed
meal (0.5 g/kg) enriched by either corn oil or glucose (n=6) ................. ................55

3-6 Changes in intragastric bag volume after ingestion of a control sweet feed meal
(0.5 g/kg) with and without addition of either corn oil or glucose. ................... .......61










4-1 Mean percentage dose recovery (PDR/h) & SEM and modeled curve of 13C in
breath following ingestion of a high-fat pelleted meal (n=5) or a
high-carbohydrate (CHO) pelleted meal (n=6) .................... ...............6

4-2 Mean percentage dose recovery (PDR/h) & SEM and modeled curve of 13C in
breath following ingestion of a 10% crude protein sweet feed meal (Seminole
Feed@, Blue Ribbon 10) enriched with corn oil or glucose (n=6)...........................77

4-3 Modeled mean % dose recovery curves of the 13C label in the breath of 6 horses
after ingestion of a sweet feed meal (control) or the same meal enriched with
corn oil or glucose. ............. ...............79.....

5-1 Changes in intragastric pH after ingestion of the high-fat or the
high-carbohydrate (CHO) pelleted meal. ............. ...............85.....

5-2 Changes in intragastric pH after ingestion of a control sweet feed meal, or the
same meal enriched with either corn oil or glucose. ................ ..................9

6-1 The standard curve shows the fraction of 125I-radiolabeled CCK-8 bound to
antibodies (B/Bo) at increasing levels of standard CCK............_.._._ .............. ....95
















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

EFFECTS OF HIGH-FAT vs. HIGH-CARB OHYDRATE DIET S ON PROXIMAL
GASTRIC RELAXATION, GASTRIC EMPTYING, pH OF GASTRIC CONTENTS
AND PLASMA CHOLECYSTOKININ IN THE HORSE


By

Mireia Lorenzo-Figueras

May 2004

Chair: A.M. Merritt
Major Department: Veterinary Medicine

Addition of fat, rather than extra carbohydrate, to the diet has become a common

strategy to increase the energy density in rations for high performance horses. Little is

known, however, about the effect of fat on equine gastrointestinal function. Therefore,

the obj ective of this study was to evaluate and compare the effect of high-fat and high-

carbohydrate (CHO) diets on different parameters of gastrointestinal function in the

horse.

Six adult horses, each with a gastric cannula, were used in two different series of

studies. In Phase I, horses were offered 0.5 g/kg of a high-fat (8% fat) or a high-CHO

(3% fat) pelleted diets of identical volume, caloric density and protein content. In Phase

II, test meals consisted of 0.5 g/kg of a sweet feed meal, or this meal supplemented with

corn oil (12.3% fat) or an isocaloric amount of glucose (2.9% fat). Four parameters were

measured simultaneously: 1) proximal gastric tone by variations in the volume of an









intragastric bag, introduced through the gastric cannula and maintained with a constant

internal pressure by an electronic barostat; 2) rate of gastric emptying by the 13C-Octanoic

acid breath test; 3) pH of gastric contents by a self-referencing pH probe introduced

through the gastric cannula; and 4) plasma CCK-like activity by a commercial, non-

specific radioimmunoassay kit (Phase I only).

Meals with higher CHO content induced a significantly (p<0.05) more prolonged

receptive relaxation of the proximal stomach than those with higher fat content, but the

accommodation response was similar. Labeling of meals with the breath test marker

modified the receptive relaxation response. Gastric emptying rates were not significantly

different between meals, although those high in CHO tended to empty more slowly

initially. A significantly greater increase in intragastric pH was seen after ingestion of the

high-CHO pelleted meal. Radioimmunoassay and bioassay methods failed to detect

plasma CCK activity. This study suggests that in the horse, in contrast to most species,

dietary fat may not be more suppressive of gastric tone and emptying than CHO.














CHAPTER 1
LITERATURE REVIEW

Fat Supplementation: A New Concept in Diet Formulation

The horse has naturally evolved to digest and use high-fiber sources efficiently.

These sources are relatively low in energy content, but adequate to satisfy maintenance

requirement. Coevolution with humans, though, has resulted in increasing demands on

the horse to perform under circumstances that require energy intakes greater than that

provided by its more natural diet of fresh roughage.69 COnsequently, performance horses

are routinely supplemented with high-starch diets to reach their high energetic needs.

This common practice can represent a digestive and metabolic challenge for the horse,

and excessive high concentrate intakes are usually avoided. Concentrates are rich in

soluble carbohydrates, which may overwhelm the digestive capacity of the small

intestine, leading to rapid fermentation of the grain carbohydrates in the hindgut.29

Addition of fat to the diet has been receiving considerable attention as a way to

increase dietary energy content, and fat-supplemented diets are becoming a common

alternative to traditional high-starch diets. Multiple feed companies are introducing their

own commercial high-fat diets, which may contain as much as 12% of the total dry

matter. Although we refer to these diets as high-fat diets (6-12% of total diet or 20-40%

of the total calories), they are actually much closer to the normal diets used in other

species.94 Providing energy as fat rather than carbohydrate may avoid some

disadvantages of concentrate diets, including the production of potentially harmful

volatile fatty acids in the stomach" and in the large intestine,130 which may cause ulcers,









and colic and laminitis, respectively. Fat may also ameliorate some of the alterations in

fluid and electrolyte homeostasis associated with feeding a large, high carbohydrate,

concentrate meal,29;188 Or help in controlling development of rhabdomyolysis in

predisposed horses.214

Since fat is an efficient energy source for exercising muscle, additional benefits of

dietary fat on performance have been proposed.94 The glucose-sparing effect of fat-

supplemented diets seems to delay symptoms of fatigue94 associated with glycogen

depletion,188;231 acidosis and lactate formation.63 Yet, the effect of fat on muscle glycogen

and plasma lactate is not consistent in the literature,69 and the potential positive effects of

supplemental fat on energy metabolism need further study.94 Additional benefits of fat

supplementation include reduction in internal body heat when compared with high-

carbohydrate or protein diets,90;192 a calming effect on the horse,73 and decreased weight

of intestinal contents from reduced dry matter ingestion.90 Finally, fat is a useful source

of energy for geriatric horses or thin horses that fail to reach a desirable weight;194 it also

improves skin and hoof appearance, and reduces dustiness of the meal.

Although fat supplementation appears to have many benefits, its use may be limited

by palatability, excessive oxidation rancidityy) during processing and storage, and the

digestive and absorptive capacity of the equine gastrointestinal tract. The adaptation of

the horse to dietary fat has been addressed in some studies. For instance, feeding an

increased level of fat causes metabolic adaptations that permit horses to preferentially use

fat and spare glycogen during exercise. Fat supplementation is also associated with

increased plasma total lipase activity (lipoprotein lipase and hepatic lipase),15 enhanced

oxidative capacity of the muscle,"" lower glycemic and insulinemic responses after









ingestion,231 and dose-dependent increase in fat digestibility.23 Since excessive soluble

carbohydrates are undoubtedly associated with clinical and sometimes life-threatening

conditions, Williams et al.231 (p. 2199) suggests that "the metabolic and health impacts

are likely to be moderate for meals rich in fat and fiber, which may be more reflective of

the nutritional heritage of the horse."

The Effect of Diet on Gastric Physiology

Little research has been conducted in the horse regarding the effects of dietary fat-

and dietary components in general-on gastric function. In other species, dietary

composition of the meal can affect the rate of gastric emptying, proximal gastric

accommodation and plasma concentration of regulatory hormones and peptides

associated with meal ingestion. Modulation of gastric motility in response to

consumption of a meal is the first step-excluding intake rate-of a long and complex

series of events aimed to maximize the digestion and absorption of nutritional

components of the diet.

Dietary Regulation of Gastric Emptying

The stomach stores ingested food in the proximal stomach, mixes it with secretions,

discriminates solids and liquid, breaks solids down to small particles, and delivers food

into the duodenum at a rate that is compatible with efficient digestion and absorption

within the intestine.124 Far from being a simple hollow organ, the stomach is composed of

three different functional components: the proximal stomach, the antrum and the pyloric

sphincter. These parts function in a coordinated way within each other and with the

proximal duodenum to regulate gastric emptying according to the composition of the

meal.173 Existence of neural and hormonal feedback reflexes originating at very distant









sites, such as ileum and rectum, to modulate gastric motility, gives further indication of

the complexity of gastric emptying regulation.183

Meal consistency: emptying of liquids versus solids

A first determinant of the way a meal is emptied from the stomach is its

consistency. The stomach can discriminate between liquids and solids,124 with liquids

emptying from the stomach much more rapidly than solids.30;119 In a solid-liquid mixed

meal, the liquid component empties rapidly in an exponential manner, whereas the solid

part remains in the proximal stomach until most of the liquid has emptied (period defined

as "lag phase").' Unlike solids, a lag phase is not observed in gastric emptying of

liquids, unless they have a high caloric density.62 Scintigraphic studies have shown that

both liquids and solids empty in an exponential manner, although emptying of liquids is

more logarithmic in character.'o Finally, discrimination between emptying of the solid

and the liquid phase of a mixed meal may be affected by the degree of homogenization of

its components. That is, after the meal is mixed in the mouth and stomach, the solid and

liquid components are not clearly separated and gastric emptying may be delayed by

increased viscosity.186

It is well recognized that gastric emptying of liquid follows an intraluminal

pressure gradient between the proximal stomach and the duodenum. This pressure

gradient is generated and modulated by variations in tonicity of the proximal gastric

wall.124 Yet, additional research indicates that factors besides tone of the proximal

stomach are crucial for normal liquid gastric emptying, and that there is a role for antral

and pyloric contractile activity.32;121;191 For example, it has been shown that pulsatile

gastric emptying of liquids across the pylorus is correlated with antral waves of

contraction in dogl61 and pig,120 whereas the volume of the flows is correlated with









proximal gastric and pyloric tone.161 Furthermore, liquid emptying is significantly more

rapid in pylorus excised than in pylorus intact pigs during intraduodenal infusion of

nutrients or hyperosmolar solutions.210

In contrast to liquids, emptying of solids seems to depend primarily on antral

contractions, which break up food into particles small enough to pass through the pyloric

canal. Additionally, coordinated antropyloroduodenal contractions, acting as a peristaltic

pump, are a maj or factor in emptying regulation. Contraction of the pylorus serves as a

barrier against the propulsive force of the antrum, causing retropropulsion and

consequent trituration of large particles.67 Simultaneously, small particles are propelled

through the pylorus into the proximal duodenum.173 Finally, the proximal stomach assists

the emptying of solids by delivering its contents down to the antrum,32 and by preventing

reflux of antral contents into the proximal stomach as the antrum contracts.173

Meal composition: nutrient modulation of gastric emptying

Volume, nutritional constituents, physical structure (i.e., viscosity and particle

size), caloric density and osmolarity of a meal are the principal factors affecting its rate of

gastric emptying, and modulation of gastric emptying can only be seen as the combined

effect of all these factors on gastric motility. For example, gastric emptying of a liquid

meal containing nutrients is influenced by the volume of fluid in the stomach and by its

energy density.36 Increasing the volume of this liquid meal will speed emptying, but

increasing the nutrient content will slow emptying.8

Presence of nutrients in the intestine is a potent stimulus for feedback regulation of

gastric motor function and can be mimicked in experimental situations by direct infusion

of nutrients into the gut.142 Intestinal feedback inhibition by nutrients involves relaxation

of the proximal stomach, suppression of antral motility, stimulation of isolated phasic









pyloric contractions and increased pyloric and duodenal resistance, which together work

to slow down the flow of gastric contents into the duodenum. As a result, the stomach

delivers nutrients into the small intestine with rates that are compatible with digestion and

ab sorpti on. 173

Nutrient Sensing Initiates Nutrients Regulation of Gastric Emptying

Receptors sensitive to nutrients play an important role in the control of gastric

motility. They are present throughout the mucosa of the small intestine and show regional

variation for nutrient sensitivity.8 For instance, gastric emptying of solids is

approximately three times more potently inhibited by glucose perfusion in the fourth

quarter versus the first or second quarter of small bowel in the dog.'07 Species differences

may also exist since carbohydrate infusion in the ileum slows gastric emptying of liquids

and solids in the dog,209 but has no effect on gastric emptying of solids in humans.227 In

addition to regional distribution, parallel and sequential activation of receptors by transit

of nutrients along the length of the intestine may occur.103;104 The complexity of dietary

regulation is accentuated by studies suggesting that gastric emptying is influenced by

patterns of previous nutrient intake, possibly through adaptation and sensitization of

receptor mechanisms to their original stimuli. Accordingly, maintaining a high-fat diet

for two weeks, but not four days, results in acceleration in the gastric emptying rate of

high-fat meals in humans.39 These adaptive changes are nutrient specific, since adaptation

to a high-fat diet does not affect gastric emptying of carbohydrates.25 Similarly, short-

term supplementation of the diet with glucose leads to increased gastric emptying rate of a

glucose test meal,76 whereas the emptying rate of a protein drink is unchanged.40









Existence of nutrient-specific chemoreceptors

Most mucosal receptors in the small intestine are polymodal (i.e., they respond to

different mechanical and/or chemical stimuli),131;19 Such as those activated by osmotic

pressure variations and presence of acids and alkali.133 On the other hand, some receptors

are specific for a single stimulus,131 and selective chemoreceptors for acid,49

glucose,49;106;190 fatty acids,135;163 and the amino acids tryptophan24;33;201 and

phenylalanine22 have been described. Nutrients act independently of any osmotic or

mechanical effects, and each macronutrient group acts via activation of separate and

distinct mechanisms and pathways.165 In addition, mechanical stimulation by presence of

contents in the small intestine inhibits gastric emptying.124 Accordingly, distension of a

balloon within the duodenum induces inhibition of abomasal contractions in sheepl9 and

decreases fundic tone in the dog.41

The current scientific literature is abundant in studies on the effect of different

nutrients on gastric emptying and the mechanisms involved in it. Yet, it is difficult to

draw definitive conclusions because of great differences in the methodology of these

studies. First, test meals are not standardized among studies, which makes comparisons

difficult. Furthermore, meals may differ in more than one characteristic and, therefore, it

is difficult to conclude which specific characteristic determined the outcome. Second,

some studies used inaccurate methods to measure gastric emptying, such as the intubation

technique, which were likely to cause experimental errors. In addition, many studies

describe the regulatory effect of nutrients by measuring the gastric emptying of a control

meal in response to simultaneous infusion of nutrients into the small intestine. This

experimental design is not as physiological as ingestion, and it does not take into account

the additional influence of the control meal on gastric emptying. As well, studies using









meal preloads often yield very different results when compared with studies in which

nutrients are infused directly into the gut. Some studies performed in rats86;143 and

humans26 address this issue, and suggest that orosensory factors play a role on gastric

emptying of specific nutrients. Therefore, we cannot assume that gastric emptying is the

same when a nutrient is directly deposited into the gastrointestinal tract as when it is fed

orally. Finally, species to species variations may exist, so that what is true for one species

may not apply to other species.

Sensing of dietary fat

It is known that fat is a potent inhibitor of gastric emptying.79 The intensity of

inhibition is dependent on the concentration of fat, the length of intestine exposed to itl04

and the type of fat. For example, studies in humans79 and catsl36 Show that free fatty

acids, but not triglycerides, are effective stimuli for inhibition, with 12-carbon or greater

chain-length fatty acids being the most effective. Similarly, instillation into the

duodenum of a 12-carbon but not a 10-carbon long-chain fatty acid reduces antral

contractile amplitude and reduces proximal gastric tone in humans.127

Fat is digested and absorbed mainly in the proximal small intestine by rapid

diffusion across the brush border membrane of the enterocyte.211 Postabsorptive

chylomicron formation is required for the ability of fatty acids to initiate feedback

inhibition of gastric emptying, which suggests that activation of the fatty acid receptor

does not occur within the lumen of the intestine.60 When a larger load is ingested,

however, fat is also absorbed in the distal small intestine.232 The arrival of fat

unabsorbedd lipids or fatty acids) into the distal small intestine causes further inhibition

of gastric emptying in a dose-dependent manner.227 Further delay serves to increase the

contact time between luminal contents and the absorptive epithelium. This feedback









mechanism, referred to as ileall brake," is species and nutrient specific. For example, it is

also induced by glucose and amino acids in dogsl96 but not in man.227 In COntrast,

infusion of fat into the distal small intestine of the dog has no effect on proximal gastric

relaxation, whereas carbohydrate and protein reduce gastric tone.8 Since most of the

dietary fat is almost entirely absorbed proximal to the ileum, feedback from the ileum

may only occur after unusual meals or under altered physiological conditions in order to

enhance absorption by delaying the passage of food through the small intestine.172

The mechanisms determining the emptying rate of fat are not totally understood.

On the one hand, one study64 in pigS suggested that fatty acids do not empty from the

stomach with a constant caloric rate, in contrast to carbohydrates. On the other hand,

studies in humans,s monkeys,126 and pigs226 Suggested that emptying of fat or any other

macronutrient was mainly based on delivery of a constant rate of energy into the small

intestine. Fats in the liquid phase empty at a rate different to the aqueous phase, but at a

rate similar to that of an equicaloric digestible solid meal48 following an initial lag

phase.141 Also, the chemical structure of dietary fats influences the rate of gastric

emptying. For example, an equicaloric amount of olive oil, containing more unsaturated

fatty acids, slows gastric emptying less than margarine, which contains more saturated

fatty acids.119 Besides chemical composition, physical properties such as density,

intragastric distribution, viscosity and degree of emulsification determine gastric

emptying of fat. Maj or differences in the intragastric distribution of oil compared to solid

and aqueous liquid meals exist, and may account for slower emptying of this component.

In an oil/aqueous-based soup meal, gastric emptying of oil was slower than soup, and it

was associated with longer retention in the proximal stomach and retrograde movement









of oil from distal to proximal stomach. When a solid meal was added to the oil/soup

mixture, more oil was retained in the proximal stomach and more solid was retained in

the distal stomach.48 Layering of an oil phase above the water due to lower specific

gravity may be in part responsible for this effect. This distribution may be consequent to

initial passage of small amounts of fat into the duodenum, leading to relaxation and

redistribution of fat into the proximal stomach." Finally, degree of meal homogenization

influences gastric emptying of the oil phase. For example, butter empties slower than the

aqueous contents, unless it is emulsified before ingestion.34

Sensing of dietary carbohydrates

Similar to fat, the factors affecting the gastric emptying rate of carbohydrates are

poorly understood, but it seems to depend on the type of carbohydrate (e.g., sucrose,

fructose, galactose), its form (e.g., maltodextrins, starches)15 and the length of intestine

to which it is exposed.103 For example, glucose empties linearly, whereas fructose

empties exponentially and more rapidly than xylose and glucose in the monkey.146

Caloric content, volume and osmolarity have been regarded as potential factors

controlling the gastric emptying of carbohydrates. However, studies on glucose dose and

volume-dependent inhibition of emptying are contradictory. On the one hand, Hunt et

al.so showed that, in humans, increases in either energy density or meal volume of a

polymer of glucose increased the rate of energy delivery after 30, 60, or 120 min. Moran

et al.148 Showed that volume of a glucose beverage played a role only in the initial rapid

rate of emptying, and the rate of gastric emptying (calories and volume) was identical

after 20 minutes regardless of the initial volume. On the other hand, results of other

studieS64;81;126;226 Suggest that emptying of carbohydrates is mainly driven by the

maintenance of a constant caloric flow. Finally, a studyl32 perfOrmed in anesthetized cats









showed that the electrical activity of vagal sensory neurons specifically responsive to

glucose increased with increasing doses of glucose infusions into the proximal small

intestine. Yet, it is unknown to what extent the observed increased discharge of

glucoreceptors may affect gastric motility and emptying. Similar to volume and

concentration, it is also unclear whether osmolality contributes to the glucose-induced

modulation of gastric emptying. Inhibition of emptying by hypertonic glucose seems to

depend mostly on chemospecific feedback,126 although osmotically sensitive pathways

may also be involved.106

Nonfiber carbohydrates are digested and absorbed mainly in the proximal small

intestine by active transport across the brush border membrane of the enterocyte.211

Activation of glucose receptors and consequent initiation of feedback inhibition of gastric

motility may be dependent either on rapid accumulation of glucose within epithelial cells

or on activation of the Na -glucose co-transporter.l7 With ingestion of a large load, the

digestive and absorptive capacity of the small ingestion may be exceeded, and arrival of

carbohydrates into the distal small intestine causes further inhibition of gastric emptying.

As mentioned previously, presence of carbohydrates within ileum slows gastric emptying

of liquids and solids in the dog.209 Glucose exerts this inhibition in a dose dependent

way,196 and this effect is more potent when compared to exposure of the upper small

intestine.10 Slowing of gastric emptying by ileal perfusion of carbohydrate has also been

observed in humans, although perfusion was associated with abdominal discomfort, and

the authors of the study related the carbohydrate-induced inhibition to activation of

nociceptive pathways.82









Fat versus carbohydrate on control of gastric emptying

As has already been mentioned, it is generally assumed that fat is a more potent

inhibitor of gastric emptying and motility than carbohydrate. However, there are a limited

number of studies comparing the effect of fat and carbohydrate meals, irrespective of

energy, volume or consistency of the meal. A few studies show that fat causes a greater

inhibition of gastric emptying than carbohydrate of equivalent caloric density and volume

in humans and in pig.6

Effect of Diet on the Meal-Induced Relaxation of the Proximal Stomach

The proximal part of the stomach acts as a reservoir for food and participates in the

creation of an intragastric pressure gradient, which is in part responsible for gastric

emptying of liquids. To accommodate the ingested material without significant increases

in intraluminal pressure, the proximal stomach is controlled by a vago-vagal reflex that

enhances wall relaxation by decreasing gastric tone. This relaxation decreases the fundo-

antral pressure gradient and slows delivery of fundic contents into the antrum. This, in

turn, decreases the rate of gastric emptying.124;207

Previous work in our lab113 have revealed the existence of a meal-induced

physiological relaxation in the proximal stomach of the horse. The response, measured by

an electronic barostat, consisted of two components: a prompt, marked and defined

relaxation phase during meal ingestion, followed shortly by a period of sustained

moderate relaxation lasting at least 90 minutes. It was concluded that these two

consecutive components might correspond to receptive relaxation (primarily under

pharyngeal and esophageal control) and gastric accommodation (primarily under gastric

and duodenal control), respectively. In addition, there was a significant, positive

relationship between the amount of meal and the magnitude of the receptive relaxation









component in the horse. This component is associated with stimulation of

mechanoreceptors in the oropharynx and/or esophagus, whereas accommodation is

associated with stimulation arising from the stomach and/or small intestine. Other factors,

besides amount of meal, may contribute to vary the relaxation response, thereby

modulating the rate of gastric emptying. For example, orosensory stimulation produced

by particular nutrients influences appetite and gastrointestinal responses,27 and may also

modulate the relaxation response of the proximal stomach. All in all, the mechanisms by

which different diets modulate gastric tone may be primarily controlled by duodenal

receptors responsive to mechanical and chemical stimuli.218 That is, passage of food into

the duodenum is detected by and activate specific receptors which, in turn, initiate a

feedback inhibition pathway that results in a slower gastric emptying rate, in part through

decrease in proximal gastric tone.

Again, composition of the diet is a maj or factor in determining gastric emptying

rate in other species and seems to modulate relaxation of the proximal stomach in a site-

specific way. As mentioned, perfusion of fat into the proximal, but not the distal, small

intestine of the dog causes a strong gastric relaxation response. The opposite effect (no

effect proximally and a potent relaxation distally) is observed with perfusion of

carbohydrate solutions.8 Meal fat content and osmolality, but not energy content, can also

affect gastric relaxation in humans via a duodeno-gastric feedback mechanism. Dietary

fat seems to be a stronger stimulus of gastric relaxation than carbohydrate in man.217

Liquid lipid meals of 2.5% or greater concentration induce a decrease in proximal gastric

tone in a non-dose dependent manner, suggesting a fat-mediated mechanism with

threshold sensitivity (no effect for meals of less than 2.5% fat) and saturability (no further









response above the 2.5% dose). Another potential factor influencing proximal gastric

relaxation is osmolality, although in rats, dogs and primates it only affects gastric tone

when meals of very high osmolality (at 2400 mmol/kg or greater) are ingested. Finally,

caloric content of a meal per se does not seem to mediate relaxation, since ingestion of

carbohydrate solutions, containing an energy load similar to that of relaxation-inducing

fat meals, does not cause gastric tone to decrease.12 Although meal volume seems to

affect the magnitude of proximal gastric relaxation in the horse,113 it is unknown whether

the nutritional composition of the diet influences this response.

Nutrients Regulate Gastric Emptying by Neural and Hormonal Pathways

The molecular mechanisms underlying chemosensitivity to luminal contents are not

completely understood. Interaction of receptors with nutrients initiates different neural

and humoral pathways involved in the complex regulatory system of gastric motility.

Neural pathways

Vagal afferent fibers of the upper gut have been found to be sensitive to a range of

chemical and physical meal-related properties, including pH, osmolarity, nutrient content,

and the mechanical distension produced by the presence of a load.190 With regard to

nutrient content, sensory afferent terminals innervating the small intestine play a

significant role in inhibition of gastric emptying in response to lipid, protein, and

glucose.74;164;167 Activation of these afferents initiates a series of neural reflexes that act

through autonomic motor nerves to allow regulation of gastric motility by the enteric and

central nervous systems. Yet, it is not entirely clear whether vagal afferents are

selectively and directly activated by nutrients.56 That is, vagal fibers are not found

between epithelial cells or making direct contact with them, so that direct neural sensing

of luminal contents does not probably occur. Instead, luminal contents may signal to









these fibers via an indirect interaction with a specialized cell situated within the

epithelium. A potential intermediary is the enteroendocrine cell, which releases peptides

in response to changes in gut contents.16 These peptides may enter the bloodstream or

act in a paracrine mode to stimulate afferent nerve terminals.56

Hormonal pathways

The critical peptides involved in feedback signaling are gastrin, cholecystokinin

(CCK), secretin, glucose-dependent insulinotropic polypeptide (GIP), glucagon-like

peptide-1 (GLP-1), neurotensin and peptide YY.21 The ability of luminal nutrients to

stimulate the release of these peptides has been extensively studied by monitoring

peripheral plasma levels after ingestion of different nutrients. Thus, it has been observed

that the profile of peptide secretion depends to a great extent on the composition of the

meal and the intestinal segment exposed to nutrients.

Since the maj ority of food absorption occurs in the upper small intestine, the

endocrine cells located in this region are highly sensitive to the presence of luminal

nutrients.21 Fat is the most effective in stimulating upper intestinal endocrine cells,

increasing the secretion of CCK, GIP and secretin. This effect requires hydrolysis of fat

with subsequent entry of fatty acids into the epithelial cells and formation of

chylomicrons. On the other hand, glucose infusion into the upper intestine stimulates the

release of GIP, serotonin (5-HT) and CCK, but not secretin.21;166 The release of GIP and

CCK depends on the sodium-dependent glucose transporter, suggesting that entry into the

endocrine cell is required in this response."

When unabsorbed nutrients reach the distal small intestine, they also stimulate ileal

endocrine cells to secrete GLP-1, neurotensin200 and peptide YY.200;209 These peptides act

as the aforementioned ileal brake, slowing gastric emptying. Glucose stimulates the









release of the three peptides experimentally, although it is believed that this does not

occur under physiological circumstances.47 With regard to fat, there are conflicting data

about release of neurotensin and peptide YY. It has also been observed that the fat-

induced ileal response may occur by direct stimulation of cells by fat in the distal small

intestine and colon or, indirectly, by the presence of fat in the proximal intestine signaling

to the distal gut.47;108

Cholecystokinin: A Key Hormone in the Regulation of Gastric Emptying

CCK plays an important role in the nutrient-induced feedback inhibition of gastric

emptying. The mechanism of action is unclear but it seems that its main effect is

relaxation of the proximal stomach.170;235 CCK also delays gastric emptying by

stimulating contractions of the pyloric sphincter and the proximal duodenum, thus

increasing the resistance to gastroduodenal flow of chime.87;235

The relative potency of nutrients in elevating plasma CCK concentration varies

among species. Intact protein and fatty acids within the duodenum are the maj or food

stimulants of CCK release in rats, whereas protein hydrolysates, L-phenylalanine, L-

tryptophan, intact fat, starch or glucose do not produce any effect.96;100 As well, medium-

chain fatty acids seem to be more powerful stimulators of CCK secretion than long-chain

fatty acids.46 In dogs and humans, duodenal infusion of long-chain fatty acids54;127;204 and

amino acids or digested protein, but not intact protein, are effective stimuli.140;208 On the

other hand, carbohydrates in the form of starch and glucose have a weak70;99o nOfO

significant effect in man. In pigs, intraduodenal starch hydrolysate, fatty acids and

protein hydrolysate evoke CCK release.38 In COws, ingestion of a high concentration of

long-chain fatty acids (90 g/kg DM), but not lower concentrations (30-60 g/kg) is

associated with increased plasma CCK levels.28 In gOats, duodenal infusion of









phenylalanine and tryptophan causes an increase in plasma CCK, but intraduodenal

branched-chain amino acids such as leucine and isoleucine fail to do so.57 In cats, plasma

CCK increases in response to intragastric administration of long-chain triglycerides,

intact protein or the amino acids residues, but not starch.'0 Finally, intragastric

administration of amino acids and medium-chain, but not long-chain, triglycerides induce

release of CCK in chicks.115;236 Therefore, in all these species, at least one nutrient has

been recognized to induce the secretion of CCK into the circulation.

Fat is invariably a potent stimulus of CCK release in all the studied species. In

humans66 and rats,96 intact fat requires hydrolysis to be effective. Furthermore, in

humans, only fatty acids with chain lengths greater than C11 are potent stimulants.127

These are the fatty acids that are absorbed via chylomicron formation, which is required

for long-chain triglycerides to slow gastric emptying and to increase circulating levels of

CCK in man61;169 and in rat.61 In COntrast to humans, medium-chain fatty acids are more

powerful stimulators of CCK secretion than long-chain fatty acids in rat,46 a phenomenon

that has also been observed in chicks.ll Therefore, a relationship between the pathways

of absorption and the ability of fats to initiate feedback responses and release CCK may

vary among species. Medium-chain triglycerides are hydrolyzed and absorbed faster and

more completely because they simply diffuse across enterocytes, bypassing the packaging

into chylomicrons and exocytosis reserved for long-chain fatty acids.212 Furthermore,

unlike long chain triglycerides, medium-chain triglycerides do not require bile salts for

digestion. Since rats and chicks have no gall bladder, bile salts may not be as readily

available for fat digestion as in other species, and different mechanisms for inducing

CCK secretion may reflect this difference in physiology.









CCK may be involved in the adaptive change in gastric emptying due to previous

feeding with high-fat diets. One study37 demonstrated in rats that the delay in gastric

emptying of saline caused by both intestinal oleate infusion and intraperitoneal CCK is

attenuated by prior consumption of a high-fat diet. Another study53 Showed that plasma

levels of CCK were raised after ingestion of a fatty meal following high-fat diet

adaptation in man. Because this raised CCK level was not associated with a reduction in

the rate of gastric emptying, it has been suggested that subj ects might become

desensitized to the effects of CCK after a period of high-fat consumption.25 Finally, the

increase in CCK levels that follows dietary fat adaptation may be secondary to an

increase in the capacity to digest and absorb fat,25 which is necessary for CCK release in

humans61;169 and rats.96

Current evidence suggests that digested fat releases CCK from enteroendocrine

cells, and activates extrinsic afferent nerve terminals to stimulate a neural reflex that

decreases gastric motility.60 CCK receptors are found in vagal terminalsl49;202 and CCK

can activate vagal afferents through the CCK-A receptor subtype.61;189 Studies using the

sensory neurotoxin capsaicin on vagal and spinal afferents have shown that different

nutrients act to inhibit gastric emptying through distinct afferent pathways. For example,

in rats, the inhibitory effect of lipid and peptone on gastric emptying is attenuated by

functional ablation of the vagal afferent, but not the spinal, pathway, and by blockade of

CCK-A receptors.52;61;74 In COntrast, inhibition of gastric emptying in response to

carbohydrates is attenuated by ablation of the vagal afferent pathway, and completely

abolished by ablation of the spinal sensory pathway.167 Further evidence of the

importance of neural pathways on CCK-mediated effects is that inhibition of proximal









gastric motility induced by administration of CCK is also abolished by vagotomy and is

hexamethonium-sensitive.170

Activation of vagal afferent endings by CCK may occur locally (duodenal mucosa)

or peripherally (by circulating CCK). Although the hypothesis that circulating CCK

mediates gastric emptying by an endocrine mode of action has been formerly accepted,

discrepancies between concentrations of plasma CCK and emptying inhibition by

nutrients indicate a non-endocrine source of CCK participating in this process. For

example, intestinal perfusion by carbohydrates inhibits gastric emptying via a pathway

involving vagal capsaicin-sensitive afferents and CCK-A receptors, and yet does not

increase circulating CCK.167 This suggests that the potency of nutrients to inhibit gastric

emptying may not always be reflected in CCK peripheral plasma levels.168 That is, the

CCK-mediated effect of glucose may be local, whereas that of fat, which is associated

with increased plasma CCK levels, may be peripheral, as well. Finally, other mediators,

such as 5-HT, may be more important than CCK in glucose-induced gastric emptying

mnhibition.l6

Finally, there is scarcity of information on the role of CCK in the horse. Endocrine

cells immunoreactive for CCK have been identified within the duodenal walls and in the

female urethra.224 In a recent study, CCK-B receptors were identified in somatostatin

cells of pancreatic islets.lso However, it is not known whether CCK-A receptors are

present in equine non-endocrine pancreas, intestinal mucosa or vagal nerves.

Gastric Emptying in the Horse: Current Knowledge and Methodology

The mechanisms controlling gastric emptying of liquids and solids in the horse are

poorly understood. Those done to date have mainly assessed the effect of prokinetic

drugs or clinical disorders, but not nutrients.









The repertoire of measuring techniques for gastric emptying is large,'" but their

application in the horse may be restricted due to anatomical or economical limitations.

First, non-absorbable markers, such as phenol red,12;199 have been used in the horse to

measure gastric emptying of liquid meals. In this technique, ingestion of a labeled meal is

followed by aspiration of gastric contents through a nasogastric tube, and the amount of

marker recovered from the stomach will depend on the rate of gastric emptying of the

meal. The shortcomings of this method are that the test meal is instilled directly into the

stomach, which is not as physiological an approach as ingestion; the extraction procedure

can be laborious and recovery may not be complete; and, finally, artificial mixing of

gastric contents may occur.199 Another indirect measurement of liquid-phase gastric

emptying involves assay of acetaminophen absorption.43-45;112;134;215;216 After oral

administration, acetaminophen is absorbed almost exclusively in the proximal portion of

the small intestine, and can be detected in blood. Gastric emptying is the rate-limiting

step in acetaminophen absorption, provided that the intestinal mucosa is intact. Thus,

both this and the phenol red technique provide a safe and relatively inexpensive

qualitative method for the measurement of liquid-phase gastric emptying.'"

Additionally, use of solid, indigestible markers has also been reported in the horse.3

Some disadvantages of this method are that particles need to be delivered directly into the

stomach, their size will influence rate of emptying and their recovery from a live horse is

only possible via a gastric cannula. Another alternative is the use of radiopaque spheresl3

or barium contrast,182 although the resolution of these methods is adequate only in small

subj ects.









In contrast to the previous methods, scintigraphy has the advantage of non-

invasiveness, and has been applied to the horse before. Additional advantages of this

technique are the possible differentiation between solid and liquid-phase emptying by the

use of specific markers, and the ability to monitor the intragastric distribution of meal

components .219 Although nuclear scintigraphy is the gold standard for measuring gastric

emptying in humans, this method is expensive and may be hazardous for technician

health in case of exposure to excessive radioactive material."'l Ultrasonography,

magnetic resonance imaging, applied tomography and changes in gastric impedance can

accurately measure gastric emptying in humans, but they have limited application to

equine medicine because of the animal's body size.

Finally, breath tests employing stable isotopically-labeled tracers offer a good

alternative to all the previous methods, since they are non-invasive, non-radioactive and

easy to perform. Moreover, samples can be stored at room temperature for a considerable

period of time and be sent to a remote laboratory for automated analysis. For this

technique, the test meal is labeled with a 13C-enriched marker. As it empties the stomach,

the marker is readily absorbed in the proximal intestine and metabolized to CO2. Thus,

breath 13CO2 enrichment reflects the rate of gastric emptying of the labeled meal. The

13C-Octanoate breath test has been validated in horse and appears to be a promising tool

for GE measurement.205

As mentioned earlier, limited studies have assessed the effect of nutrients in gastric

emptying in the horse. In a study in foals, Baker et al.12 Showed that phenol red-labeled

milk emptied slower than water and isotonic saline. However, addition of lipids to the

saline meal, with a fat concentration comparable to that of milk (1.5% fat), did not lead to









slowing of gastric emptying. The authors of the study suggested that a different type of

fatty acid, or the lactose and protein might be responsible for the slower gastric emptying

rate of milk. Experimental factors such as low number of experimental units (four foals),

the constant presence of a nasogastric tube for phenol red recovery and different degree

of fat emulsification may have also influenced the rate of gastric emptying of both meals.

Using the same phenol red technique, Sosa LeC~n et al.199 Studied gastric emptying of

different 8-liter oral hydration solutions administered via a nasogastric tube. Isotonic,

cold isotonic (5o C), isotonic + dextrose (34.7 g/1) or hypertonic solutions emptied with

similar rates, suggesting that temperature, tonicity or glucose content did not significantly

affect gastric emptying. However, poor mixing of phenol red and interference with solid

contents in some subjects occurred, causing inaccurate measurement.

By use of scintigraphy, a preliminary study revealed that one liter of 25% dextrose

solution emptied slower than the same volume of water or one pound of grain in three

ponies,197 although intersubj ect variation was high.

Finally, by use of the 13C-Octanoic acid breath test, Wyse et al.234 Observed that

addition of 30 ml or 70 ml of soya oil to a concentrate meal of oats and bran caused a

significant delay in gastric emptying in ponies. However, breath test parameters were

very similar between both soya-supplemented meals, and it was unclear which

characteristic of the meal (energy density, viscosity) was responsible for delayed gastric

emptying. In another study, Geor et al.58 evaluated the effect of adding corn oil (10%) to

a sweet feed meal (2 g/kg bwt) on gastric emptying using the 13C-Octanoic acid blood

test. Addition of corn oil resulted in delayed plasma appearance, and this effect was not

affected by a 4-week or 8-week period of adaptation to a fat-supplemented diet.









The Effect of Diet on Intragastric pH

Squamous Ulceration of the Proximal Stomach

Ulceration of the squamous portion of the equine gastric mucosa is a common

problem in ~80% of horses under intense training programs irrespective of type of work

or breed of horse.68;85:125:162:223 The degree of lesion severity can be quite variable, but

common signs of serious affliction include reluctance to finish grain meals, periods of

low-grade abdominal discomfort, especially after ingestion of a meal, and failure to

perform up to expectations.221:222 Some of the current ideas concerning the pathogenesis

of gastric squamous mucosal ulcer disease in horses have included: 1) excessive

intragastric production of volatile fatty acids (VFA), which have corrosive potential

within the stomach, secondary to ingestion of large grain meals;' 2) strict meal-feeding

practices themselves; 3) "stress"-related problems such as decreased gastric emptying

rate or excessive bile reflux from the small intestine, especially during periods between

meals when there is little food in the stomach.

Although the pathogenesis of squamous ulceration is most probably multifactorial,

excessive exposure of the squamous mucosa of the proximal stomach to acid is believed

to be the main reason of ulcer development in this gastric region. One study in our lab114

showed that acidic gastric contents are pushed upward during exercise, which may

subsequently damage the mucosa. Therefore, modification of these contents may result in

a lesser exposure of the mucosa and, thereby, a lesser risk of ulcer development. Potential

strategies to influence gastric contents composition in the equine stomach during exercise

could include changes in pre-exercise feeding times and formulation of the diet. Thus, it

is very important to know how various dietary formulations affect intragastric pH, since

this has strong implications regarding their ulcerogenic capacity. Fat-supplemented diets









are becoming a common alternative to cover the high energetic demands of performance

horses, and it is of great interest to see how such supplementation can affect intragastric

pH.

Studies on the Effect of Diet on Intragastric pH in the Horse

Few studies have measured meal-induced changes in gastric pH in the horse, but

buffering capacity of the meal seems to depend on composition of the diet. For example,

ingestion of an alfalfa hay-grain diet is associated with a significantly higher pH of

equine gastric contents, compared to a brome grass hay diet.156 This difference may be

attributable to a higher calcium and protein concentration of the former. In addition,

ingestion of fermentable carbohydrates is considered to decrease intragastric pH by

increasing production of lactate and VFA within the stomach. In vitro studies show that

some forms of the latter are harmful for the equine gastric squamous mucosa.5;5

Study Objectives

The main goal of this study was to evaluate the effect of ingestion of a high-

carbohydrate versus a high-fat meal, both offered in isocaloric and isovolumetric

amounts, on gastric emptying, proximal gastric relaxation, intragastric pH and plasma

CCK in the fasted and resting horse.

This was accomplished by measuring, simultaneously, the tone of the proximal

stomach with an electronic barostat, and gastric emptying with the 13C-Octanoate breath

test to establish a novel approach for the study of equine gastric physiology.















CHAPTER 2
MATERIALS AND METHODS

Animals

Six adult horses (3 mares, 3 geldings) with previously inserted gastric cannulas

were used for the experiments. The animals weighed between 350 and 546 kg (average

479 kg) and aged 5 to 18 years. They were housed in paddocks and maintained on free

choice of Coastal Bermuda hay, bahia grass pasture and trace minerals. All studies were

approved by the Institutional Animal Care and Use Committee of the University of

Florida.

Barostat

A specially designed barostat (Isobar 3, G & J Electronics Inc., Willowdale, Ont.)

was used to assess changes in volume/pressure of the proximal stomach (Fig. 2-1). The

validity of an electronic barostat to measure proximal gastric motility and tone was first

demonstrated in dog by Azpiroz and Malagelada,s and work in our lab113 has also proved

its usefulness for the study of proximal gastric tone in the horse. The basic principle of

the barostat is to maintain a constant pressure within an air-filled intragastric bag that has

infinite compliance and a volume greater than the range of volumes to be used by the

barostat during the study. When the stomach contracts, the barostat removes air from the

bag to maintain the intrabag pressure constant and, conversely, when the stomach relaxes,

air is inj ected into the bag. Thus, the changes in bag volume are a direct indication of

changes in intragastric pressure induced by variations in wall tone.

































Figure 2-1. Motility of the stomach measured by an electronic barostat. The barostat
maintains a constant pressure within an intragastric bag. When the stomach
relaxes, the system inj ects air into the bag (A). Conversely, when the stomach
contracts, air is aspirated from the bag (B).

For this study, intrabag pressure was set at 2 + 0.5 mm Hg. A plastic bag

(Commercial Mylar balloon, The New Garden, Gainesville, FL) of ~1600 ml total

capacity and 20-cm diameter (Fig. 2-2) was connected to the barostat through a 1.5-m

long plastic catheter of 4 mm internal diameter (inflation line). A separate catheter

(pressure line) of 1.6 mm internal diameter connected the bag to the pressure transducer,

so that pressure was monitored directly within the bag. The tip of both lines were

attached to each other inside the bag and sealed to the bag 18 cm from the tips using

dental floss. Finally, a plastic probe of 3 5.5 cm was attached to the portion of the

inflation line near the bag to increase the catheter rigidity inside the stomach. The


A) Decrease in Tone


B) Increase in Tone










polyester bag had infinite compliance, i.e., the bag did not have any influence by itself on

the internal pressure."

















Figure~~~~~: 2-.Plete a onetdt resr ie n niflto ie

Anlge inl fo h rcrig ntuenswr dgtzd n iplydb

us o cmptr rorm Wi~q rahcspckge ATQ ntrmet Ic. krn


OH a smleT raeo 0H/hne.Cagsinbgvlm m)adpesr







bagrostat. Following an experientterded dat weresue ie trdansiferedtion a CDfo

per anent e record. h rcrig ntuenswr dgtzd n iplyd






Rates of, gasJ ltrics emting. wereno assesse wroith the 13-canoicacid rea fthtes






validated by Sutton et al.205 This non-invasive, non-radioactive technique is based on the









detection of 13C enrichment in breath following the ingestion of a meal labeled with 13C

octanoate. This marker is rapidly and totally absorbed in the small intestine and oxidized

in the liver to produce CO2.233 For each test, approximately 1.5 mg/kg bwt of 13C

octanoic acid (Octanoic-1-13C acid, 99 atom % 13C, Sigma-Aldrich, St Louis, MO) was

added to egg yolk (1 yolk/250 mg of marker), baked in a microwave oven and thoroughly

mixed into the test meal. This dose of octanoic acid is higher than the one previously

used by Sutton et al. (1 mg/kg).205 Sutton (personal communication, 2002) found that

Bermuda grass is naturally enriched with 13C, and suggested a higher dose to compensate

for this source of exhaled 13CO2. Preparation of the octanoate-enriched yolk was done on

the day prior to the experiment, and stored in the refrigerator until use in order to reduce

the risk of feed rejection. The enriched yolk was added to the test meal and thoroughly

mixed five minutes before it was offered to the horse.

Intragastric pH Monitoring

The pH of gastric contents was continually sampled by a self-referencing pH

electrode (24-hour pH catheter, Medtronic Functional Diagnostics A/S, Skovlunde,

Denmark) that was inserted up through the gastric cannula so that its tip protruded ~2 cm

into the gastric lumen. The electrode was attached to a data collection device

(Medtronics, Shoreview, MN) that sampled pH every 4 seconds for up to 24 hours, and

stored the results (Digitrapper, Medtronics, Shoreview, MN). Recorded pH data were

subsequently downloaded onto a computer through the use of software provided by the

manufacturer (Esophagram MD, Medtronic, Shoreview, MN).

Plasma Cholecystokinin Radioimmunoassay

Plasma CCK-like activity was measured using a commercial, non-specific (human)

radioimmunoassay (RIA) kit (Alpco Diagnostics, Windham, NH). Although this kit is









commercialized for human CCK, this protein is well conserved in vertebrates, which

suggests that there is a high structural homology between both species.83

A different non-specific CCK RIA kit (Peninsula Laboratories Inc., San Carlos,

CA), which had full cross-reactivity with gastrin, was used in preliminary studies to

determine whether equine plasma or serum would be more useful to detect CCK. Since

CCK and/or gastrin levels were obtained when using plasma, but not serum, the

following assays were performed with plasma.

Plasma was collected and the proteins extracted according to the manufacturer' s

guidelines. To study which method of sample collection maximized CCK detection, we

collected blood in EDTA tubes with and without aprotinin (Sigma, St. Louis, MO), i.e.,

500 kallikrein inhibitory equivalents/ml of blood. Since CCK might be easily and readily

degraded, we also compared the effect of centrifuging each sample immediately after

collection, or after completion of sample collection (2.5 hours). In both cases, plasma was

stored at -700C after centrifugation. Sample tubes and syringes were constantly cooled in

an ice-bath before and after use.

Thawed samples were extracted with 96% ethanol and evaporated overnight using

a Speed Vac Concentrator (Savant Instruments, Inc., Holbrook, NY) at 370C. Dry

extracts were dissolved to the original sample volume with a diluent provided by the kit.

A recovery control with CCK-8 (amino acids 26-33) sulphate was included to estimate

the extraction recovery. A four-day assay was performed using rabbit antiserum raised

against synthetic CCK-8 sulphate. Antibody-bound 125I-CCK sulphate was separated

from the unbound fraction using double antibody solid phase.









A separate experiment was performed to collect plasma samples for in vitro

bioassay measurement. The bioassay, developed by Liddle et al.,99 is based on the ability

of bioactive CCK peptides extracted from plasma to stimulate amylase release of isolated

rat pancreatic acini, and is described in detail elsewhere Plasma samples were collected

from one horse. After an overnight fast, the horse was offered 226 g of sweet feed (0.5

g/kg) mixed with 30 ml of corn oil and 5 g of phenylalanine. Duplicate blood samples

were obtained before the meal and 10, 20, 30 and 50 minutes after meal ingestion. Blood

was collected in EDTA tubes with and without aprotinin, and constantly cooled in an ice

bath until the end of the collection. Plasma samples were extracted and brought to

dryness as done for the CCK RIA. Dry samples of 1-ml and 5-ml original plasma volume

were shipped to the University of Sherbrook, Canada, for CCK bioassay. All samples

were reconstituted with 1 ml of distilled water, so that samples with original volume and

concentrated samples were assayed. Increasing concentrations of caerulein, an analog of

CCK, were used as a positive control.

Test Meals Composition

The study consisted of two consecutive phases that differed in the test meals

offered to the horses. Detailed composition of the different diets is shown in Tables 2-1

and 2-2.

Phase I Studies

In phase I studies, two isocaloric (1.5 kcal/kg bwt) and nearly isovolumetric

pelleted meals (16% protein) were used. The high-carbohydrate pelleted meal was rich in

starch (31%) and poor in fat (3%), whereas the high-fat pelleted meal was rich in fat

(8%), had no starch, and contained more fiber (43.5% NDF in the high-fat meal versus

28.4% NDF in the high-carbohydrate meal).










Table 2-1. Approximate composition of the pelleted test meals used in the Phase I
studies.


High-fat
pelleted meal


High-CHO
pelleted meal


Ingredients
(g/kg feed)
Beet pulp
Soybean hulls
Corn meal
Oats Ground
48 Soybean meal
Wheat middling
Alfalfa meal
Cane molasses
Soybean oil
Biophosphate
Salt mixing
Calcium carbonate
Vitamin E
TM Premix
Vitamine Premix 6
Sodium Selenite
(relative composition)
Dry Matter
Starch
Fat
Acid detergent fiber
Neutral detergent fiber
Protein
Calcium
CHO: carbohydrate.


290
290


327
320
149
70
50
50


9
7
14
3.45
0.4
0.25
0.06


167
71
50
50
50
14
7
6
3.65
0.35
0.25
0.04


88.20%


8%
28.40%
43.50%
16.27%
0.89%


87.96%
31.14%
3%
11.53%
28.40%
16.18%
0.91%


Phase II Studies

Since diets used in the first part of the study differed in more than fat and

carbohydrate composition, the phase II studies were performed to determine the specific

effect of these two components. Accordingly, two new isocaloric (1.95 kcal/kg bwt) and

isovolumetric diets were formulated. A control meal consisting of 0.5 g/kg bwt of 10%









protein sweet feed (Seminole Feed, Blue Ribbon 10) with 1.5 kcal/kg was used as the

basis to prepare the two experimental diets. A high-fat meal (12.3% fat) was prepared by

adding 0.05 g/kg bwt of corn oil to the control meal. A high-carbohydrate meal (2.9% fat)

was prepared by adding 0. 113 g/kg bwt of glucose to the control diet. To account for

influence of the sweet feed in the results, the control meal was included as an additional

experimental meal.

Table 2-2. Composition of the sweet feed meal (Seminole Feed, Blue Ribbon 10) used in
the Phase II studies. A high-fat meal and a high-carbohydrate meal were
prepared by adding 0.05 g/kg bwt of corn oil and 0. 113 g/kg bwt of glucose,
respectively, to the control sweet feed meal.
Crude Protein (min) 10% Zinc (min) 120 ppm
Crude Fat (min) 3.50% Copper (min) 40 ppm
Crude Fiber (max) 6% Selenium (min) 0.3 ppm
Calcium (min) 0.45% Vitamin A (min) 13200 IU/kg
Calcium (max) 0.55% Vitamin D3 (min) 880 IU/kg
Phosphorous (min) 0.35% Vitamin E (min) 92 IU/kg

Study Design

Phase I Studies

In phase I, each horse participated in four experiments involving two sessions per

diet (either with or without labeling the test meal with 13C-Octanoic acid). The sequence

of the experiments was based on a 2-period randomized block design. In the first period,

group A (3 horses) received the high-fat pelleted diet and group B (3 horses) received the

high-carbohydrate pelleted diet. Horses were gradually acclimated to the test diets during

a 1-week period. At the end of the accommodation period, they received 5 g/kg bwt/day

of the respective test diet, divided in two feedings, until the completion of the first period.

Horses were also fed free choice Bermuda grass hay for the entire period.

After the dietary accommodation period, horses completed two randomly assigned

studies. In one study, baseline breath tests were performed with non-labeled meals to









measure the presence of natural 13C. Some substrates like starch are derived from C4

plants (corn in this case), which, unlike most plants (i.e., C3 plants), have a CO2

incorporation mechanism that results in a higher 13C COntent.206 The other study consisted

of 13C-Octanoate breath tests. Therefore, the only difference between both studies was

that in basal studies no additional 13C-Substrate was added in the test meals.

After completion of both experiments, the diets were gradually switched between

groups, and the two different experiments repeated. All test meals were fed at 0.5 g/kg

bwt.

Phase II Studies

In phase II, horses were maintained on 0.25 g/kg bwt/day of sweet feed (10% crude

protein) twice a day, and free choice Bermuda grass hay. Each horse participated in three

experiments using a random sequence. Unlike the first part of the study, no baseline

breath tests were performed. The only difference among the three experiments was the

meal offered to the horses (control, high-fat or high-carbohydrate meal). All test meals

were fed at 0.5 g/kg bwt.

Experimental Procedure

No horse participated in more than one experiment per week. Food was withheld

for 14 hours before each experiment. After this period, the horse was placed in the stocks

and the gastric cannula cleansed. Status of the squamous mucosa of the proximal stomach

was evaluated with an endoscope introduced through the cannula (Fig. 2-3). The

previously folded barostat bag, along with an attached pH probe, was introduced into the

stomach through the cannula and inflated manually to ensure it became unfolded (Fig. 2-

4). Position of the barostat bag within the proximal stomach was verified by the










endoscope introduced nasogastrically (Fig. 2-5). Correct position was defined as being

above the margo plicatus.

















Figure 2-3. Endoscopic view of the squamous mucosa of the proximal stomach.
Endoscopy was performed through the gastric cannula at the beginning of
every experiment to evaluate the status of the squamous mucosa.


Figure 2-4. Positioning of the polyester bag used to measure intragastric pressure in the
proximal stomach.113 The bag was introduced through a previously inserted
cannula.




















Figure 25. EndocpicKw~ viw hwn orc oiino h aottbgwti h
proxial stmach
Once~~~~~~~~~~;. intepoia toah h a a epidb yigeadcnetdt
the barostat by the cathete~~~~~~~r. The daacletonisrmn orp eodngwsatce






Thgue bar Enostat wa vese t ho m intan constant poiiontraba presue baofa 2a mmtHg.


Theeafer aciviyo the proximal stomach, h a was recorded du sringe tw hurb changestdt

ine vaolume of the isoattrically controlleda.Then first 30minute fof the exerorimenatwhere







recorded to obtain a baseline volume. Then, the horse was offered only one of the

possible meals and recording continued for a total of 120 minutes. Duration of meal

ingestion was also recorded.

For plasma CCK measurement, a jugular catheter was placed before starting the

experiment and blood was withdrawn using chilled syringes. Samples were obtained 10

minutes before feeding, and then every 15 minutes for a total of 120 minutes. Collected

blood was transferred into chilled EDTA tubes (1 mg/ml of blood) with and without

aprotinin, for later RIA comparison. Samples were centrifuged and the plasma was stored

at -700C until CCK determination.

































Figure 2-6. Motility of the proximal stomach was measured with an intragastric bag,
inserted through the gastric cannula and connected to the electronic barostat
by two separate catheters. One catheter was used by the barostat for air
inj section and withdrawal, and the other catheter was connected to a pressure
transducer. A pH probe was also inserted through the cannula. After correct
position of the bag was confirmed by endoscopy, the cannula was clamped to
avoid leaking of gastric content during the entire experiment.

Breath samples were collected using a modified Aeromask (Trudell Medical

International, London, Ont.) fitted with a 250-ml aluminum coated polyethylene bag

(QuinTron Instrument Company, Milwakee, WI). The horse was allowed to breathe once

through the mask before filling the bag, which was fitted with a unidirectional valve.

Duplicate samples were transferred from this bag to 10-ml red cup tubes, conveniently

sealed and stored until ready for stable isotope analysis. Three basal breath samples were

collected 60, 15 and 5 minutes before test meal ingestion, and thereafter at 15-minute

intervals for 3 hours, then 30-minute intervals for a further 3 hours. The 13 :12C ratio of

each breath sample was determined by automated continuous flow isotope ratio mass









spectrometry (PDZ Europa ABCA analysis, PDZ Europa Ltd., Sandbach, UK) and

expressed relative to an international standard. This rate was converted to parts per

million (ppm) 13C, and expressed as ppm exceSS 13C, after subtraction of the average 13C

abundance of the three baseline breath samples. The percentage dose recovery (PDR) of

the administered isotope in the breath was also calculated, and plotted against time.

























Figure 2-7. The pH of gastric contents was measured using a pH electrode inserted
through the gastric cannula and positioned in the most ventral part of the
stomach. This electrode was connected to a data collection device (arrow)
attached to a surcingle during the entire length of the experiment.

Barostat recording and blood collection were entirely done with the horse in the

stocks. Once these two components of the study were finished (2.5 h from beginning of

the study), the horse was moved into a stall and the rest of breath samples collected there.

Since it was very difficult to remove the intragastric bag after feeding the horse, and to

avoid loss of food through the cannula, removal of the bag was done after completion of

the study.









Following each experiment, the recording equipment was removed, the gastric

cannula was plugged and the horse was returned to the paddock.

Data Analysis

All statistical analyses were performed using SAS version 8.2 (SAS Institute Inc.,

Cary, NC). All results are shown as mean + SEM. Significance was set at p<0.05.

Ingestion Time

Time of ingestion among diets was compared by repeated measures analysis of

variance (ANOVA). Data were previously tested for normality (SAS: proc univariate)

and homogeneity of variance (SAS: proc glm). A Friedman's two-way ANOVA test was

used when there was inequality of variance.

Proximal Gastric Compliance

One bag-volume measurement per second was obtained throughout the

experiments. For every diet, the data of the experiments were grouped into 2-minute

blocks and averaged for the six horses. The blocks comprising the first 30 minutes were

used to obtain a baseline bag volume. The remaining blocks (90 minutes) corresponding

to the postfeeding period were analyzed to study the relaxation response of the proximal

stomach in relation to the baseline volume. Accordingly, the average of the baseline

blocks was subtracted from each postfeeding block, in order to account for baseline

differences among diets. Blocks of different diets were then compared by repeated

measures ANOVA using SAS Mixed Procedure. Time was set as a fixed effect, whereas

horse was considered a random effect. Statistical comparison of mean baseline volumes

among different diets was also performed to measure reproducibility. Finally, mean

baseline was also compared with postfeeding blocks within the same diet.









Intragastric pH

Similar to the analysis for intragastric volume, pH data were grouped in 5-minute

blocks. The blocks comprising the first 30 minutes were used to obtain a baseline.

Although pH was continuously monitored for 7.5 h, only the first 2 postprandial hours

were used for comparison with baseline within diet and between diets. Mean values of

the two diets were determined separately and compared by use of Proc Mixed ANOVA.

Breath Samples

Calculation of gastric emptying parameters

All samples containing less than 0.5% COz were rej ected to minimize analytical

inaccuracies. Data were plotted against time as either ppm 13C-enrichment, or the

percentage of the isotopic dose recovered per hour (PDR/hour). The 13CO2 excretion

curve (PDR/h) was plotted against time using the formula:

(i) y = atb -ct

where y is the cumulative percentage of 13C excretion in breath,195 t is the time in hours

and m, k and p are constants with m describing the total 13C TOCOVery when time is

infinite. This formula is derived from the Einding that the inverse curve for cumulative

dose recovery is empirically analogous to the scintigraphic curve of gastric emptying.

The best fit curves, and hence the constants a~b, c,m,k and above, were calculated using

least squares non-linear regression analysis, programmed into a Microsoft Excel Solver

function (Microsoft Corporation, Redmond, WA). Using the above constants the

following parameters of gastric emptying were calculated: (a) the gastric emptying

coefficient (GEC), equivalent to the natural logarithm of a, and considered as a global

index of the rate of gastric emptying.'" The GEC reflects the gradient of the emptying

curve and is a universal index of gastric emptying rate; (b) the gastric half-emptying time









(tl/2), eqU1Valent to the time at which the area under fitted cumulative 13C excretion curve

demonstrates recovery of half the administered isotopic dose. T1/2 WAS calculated using

both the Excel function Gammainv (0.5;b + 1; 1/c) (58) and Siegel's methodl95 tl/2 -

In[1 2 1/]/k; (c) the time to peak breath 13CO2 (tmax), calculated as b/c.

Effect of diet on basal 1C output

Effect of 13C-abundant components in the diets was only studied by observational

analysis during the Phase I of studies (high-fat versus high-carbohydrate pelleted feed).

Effect of the octanoic acid-loaded diets on 1C output

The effect of diet on parameters of gastric emptying (GEC, tl/2 and tmax) was

determined by use of paired difference t-test. For unpaired data, a 2-sample t-test was

performed. When the response variable was not normally distributed, a Friedman's two-

way ANOVA test was performed.















CHAPTER 3
RESULTS AND DISCUSSION TONE OF THE PROXIMAL STOMACH

Ingestion of a meal induces relaxation of the proximal stomach to accommodate the

meal without a significant rise in gastric pressure. This process has two components that

are generated at different levels of the upper gastrointestinal tract. First, ingestion and

swallowing of the meal stimulates receptors in the oropharynx and esophagus, and

initiates a vago-vagal reflex that decreases the tone of the proximal stomach.124 This first

component is termed "receptive relaxation" because the proximal stomach relaxes in

anticipation of the arrival of ingested material. Next, both arrival of meal into the

stomach and its passage into the small intestine triggers an additional vago-vagal reflex

that causes further relaxation. This second component is known as "adaptive relaxation"

or "(accommodatiion."2;124 For the purposes of this chapter, only the term

"accommodation" has been used to refer to this component.

The magnitude of postprandial relaxation is related to the composition of the

diet,8;129;217 and dietary fat appears to cause greater relaxation than dietary carbohydrate

in humans.129;217 The hypothesis of this study was that, in the horse, ingestion of a high-

fat meal would induce greater relaxation of the proximal stomach than ingestion of an

isocaloric and isovolumetric high-carbohydrate meal. To test that, changes in tone of the

proximal stomach were measured with an electronic barostat after ingestion of a high-fat

or a high-carbohydrate meal.









Effect of 13C-Octanoic Acid Labeling (Breath Test)

Phase I Studies: Pelleted Diets

In the first phase of the study, every horse participated in four experiments, which

differed in the test meal offered to the horse: a high-fat pelleted meal, a high-

carbohydrate pelleted meal, or any of these meals enriched with 13C-Octanoic acid (used

as a marker for the breath test). Following a randomized block design, horses were

gradually adapted to the high-fat or the high-carbohydrate diet for a minimum of 1 week

before any experiment was performed using the respective diet.

Responses in tone of the proximal stomach induced by ingestion of the unlabeled

meals were compared to those induced by the labeled meals. Addition of the 13C-Octanoic

acid, which was mixed with egg yolk and used as the gastric emptying marker in the test

meal, caused a significantly (p<0.05) greater accommodation of the proximal stomach

after ingestion of the high-carbohydrate meal, than after the same meal without the

marker (Fig. 3-1).

In contrast to the high-carbohydrate meal, no difference in accommodation was

found between the high-fat meal and the same meal with the marker when mean bag

volumes were compared within the same time intervals.

An increment in fat content of the labeled high-carbohydrate meal may account for

the longer accommodation observed after ingestion of this meal, compared to the

unlabeled meal. For the breath test technique, a dose of 1.5 mg/kg of 13C-Octanoic acid, a

medium-chain fatty acid, was prepared in egg yolks (1 yolk/250 mg of marker) and

mixed with the test meal. In a 500-kg horse, this label would consist of 3 yolks and 750

mg of octanoic acid. Labeling of the high-carbohydrate meal thus resulted in a change of

3% fat to 9% fat, whereas labeling of the high-fat meal increased the amount of fat from
























br ...................... ......................>


43


8% to 13.6%. The magnitude of accommodation is related to the lipid content of the meal

in humans, and concentrations of 2.5% or greater are needed to induce relaxationl29;213. A

threshold may exist in the horse as well, and may have been surpassed with labeling of

the high-carbohydrate meal.


-a- Plain hig h -CHO pelle te meal
*Octanoic-enriched high-CHO pelleted meal


* *+ *


30 60


Time after ingestion (min)



Figure 3-1. Changes in intragastric bag volume after ingestion of the high-carbohydrate
(CHO) pelleted meal with and without addition of 13C-Octanoic acid. Data are
expressed as mean bag volume + SEM of 2-min blocks (n=6). Each
postprandial volume represents values from which the baseline volume has
been subtracted. Asterisks denote a significant difference between pairs of
blocks at the same postprandial time (p<0.05). Lines on top of the figure
delimit the period where mean bag volume was significantly higher than
baseline after ingestion of the meal with (dashed arrow) and without
(continuous arrow) octanoic acid (p<0.05). Note: Blocks with negative values
represent mean bag volumes that were lower than the baseline volume.

In contrast to the high-carbohydrate meal, labeling of the high-fat meal had no

effect on the magnitude of accommodation despite the increase in fat content. One

possible explanation is that, in the horse, fat induces relaxation in a dose-independent









way. That is, beyond a threshold dose, increasing doses of fat do not cause greater

relaxation. This dose-independency has been reported in humans, although it is

inconsistent in the literature. Two studies have shown that intragastric deliveryl29 Or

duodenal perfusion" of liquid meals with different lipid concentration produce a non-

dose dependent reduction in gastric tone, whereas one study213 States that the magnitude

of relaxation increases with ingestion of increasing amounts of fat.

One goal of the present study was the simultaneous assessment of gastric emptying

and proximal gastric relaxation, which are interdependent, to examine the integrated

effect of meal ingestion on gastric motility. This was achieved by using an electronic

barostat and the 13C-Octanoic acid breath test concurrently. The combined use of these

two techniques is valid only if none of them interferes with the results of each other.

Because the breath test procedure affected the barostat response of the high-carbohydrate

pelleted meal, but not that of the high-fat pelleted meal, the combined use of both

techniques was not considered valid in these Phase I studies. Bearing in mind that

interference between the two techniques seems to occur depending on the test meal

composition, the effect of labeling the test meals used in Phase II on the meal-induced

relaxation response was investigated.

Phase II Studies: Sweet feed Diets

Test meals consisted of a 10% crude protein sweet feed meal (0.5 g/kg bwt)

supplemented isocalorically with either corn oil or glucose. Horses received a daily sweet

feed ration (0.25 g/kg bwt/day), along with free-choice Bermuda hay, for the duration of

the study. Changes in proximal gastric tone were evaluated using the test meals with and

without labeling with 13C-Octanoic acid.









Adding octanoic acid to the corn oil- or the glucose-enriched meals did not have

any significant effect on the relaxation response compared to that recorded after ingestion

of the unlabeled meals.

Influence of Dietary Composition on the Effect of 13C-Octanoic Acid Labeling

It is difficult to explain why addition of octanoic acid affected the accommodation

induced by ingestion of the high-carbohydrate pelleted meal, but not that induced by

ingestion of the sweet feed meals. The extra fat content of the breath test marker may

account for the longer accommodation observed after labeling the high-carbohydrate

pelleted meal (fat increase from 3 to 9%). In contrast, although addition of the octanoic

acid to the sweet feed meal supplemented with glucose increased the lipid content from

2.9% to 7.7%, the magnitude of accommodation was unaffected. Unlike the pelleted

meals, bag volumes after any of the sweet feed meals remained above baseline (see

section "Effect of dietary composition: Phase II"), and the additive effect of the octanoic

acid may have been masked by this greater accommodation response. Therefore, because

composition of the test meal may determine whether 13C-Octanoic enrichment influences

the meal-induced accommodation of the proximal stomach, the possible existence of such

influence should be determined a priori in any study where the 13C-Octanoic acid breath

test is used in conjunction with the measurement of proximal gastric tone. Because it

existed in the Phase I studies, no integrated analyses of the gastric emptying and

relaxation results were performed for this Phase.









Effect of Dietary Composition

Phase I. Pelleted Meals: Fat Versus Carbohydrate

As mentioned previously, test meals consisted of a high-fat pelleted meal or a high-

carbohydrate pelleted meal. None of the meals were labeled with 13C-Octanoic acid for

these experiments.

Duration of meal ingestion. Mean & SEM duration for complete meal ingestion

(time to empty the food bucket) for the high-fat meal was 228 & 30 seconds (range, 156

to 368 seconds), whereas mean duration for ingestion of the high-carbohydrate meal was

158 & 14 seconds (range, 120 to 208 seconds). Thus, horses spent a significantly (p=0.02)

longer time ingesting the high-fat meal.

Relaxation response. Baseline volume of the barostat bag did not differ

significantly between diets. Relaxation of the proximal portion of the stomach, indicated

by an increase in bag volume, was observed in response to ingestion of both meals (Fig.

3-2). Bag volume began to increase rapidly after ingestion, reached a peak volume, and

then decreased sharply by the end of ingestion until returning to baseline volumes. This

receptive relaxation episode lasted 6 minutes with ingestion of the high-fat meal and 10

minutes with ingestion of the high-carbohydrate meal. A second, less profound,

significant increase in bag volume (accommodation) was observed one hour after

ingestion of the high-fat meal, with bag volumes remaining significantly (p<0.05) higher

than baseline throughout most of the remainder of the recording period (Fig. 3-3). In

contrast, this effect was not observed with the high-carbohydrate diet, except for discrete

periods of time in which bag volume was significantly (p<0.05) higher than baseline.

Finally, there was no significant difference for any time interval between diets.












10001






0 15 30 45 60 75 90 105
Time (min)


1000
S800


S400
S200

0 15 30 45 60 75 90 105
Time (min)

Figure 3-2. Mean bag volume revealing the effect of intake of a high-fat (A) and a high-
carbohydrate (B) pelleted diets (0.5 g/kg) on gastric tone of the proximal
portion of the stomach in 6 horses. An increase in volume, indicating
decreased gastric tone, was observed shortly after beginning of ingestion for
either diet. The initial peak response (receptive relaxation) was associated
with the period of active food ingestion (box). Following this peak, bag
volume returned to baseline levels. A second significant increase in bag
volume (accommodation) was observed one hour after ingestion of the high-
fat pelleted meal, and bag volume remained significantly higher than baseline
throughout most of the rest of the recording period. Time 0 is start of
recording period, and beginning of ingestion was at 30 minutes.

A previous studyll3 in Our lab demonstrated the existence of a solid meal-induced

physiological relaxation in the proximal stomach of the horse, which could be divided in

two components: a prompt, marked and defined relaxation phase during ingestion and a

second phase of sustained moderate relaxation lasting at least 90 minutes. The results of


the present studies are similar to that previous one in that beginning of meal ingestion

was followed by rapid onset of proximal gastric relaxation, the duration of which was

associated with duration of active ingestion. Ingestion of a meal initiates a series of

reflexes in order to receive and store ingesta within the stomach. Initially, stimulation of







48


receptors within the oropharynx and esophagus triggers receptive relaxation of the

proximal stomach.l Since the initial relaxation observed in the current study was

coincident with time of active ingestion, this relaxation episode is consistent with

receptive relaxation, and it was probably induced by activation of oropharyngeal

receptors.




-* High-fat pelleted diet
-o High-CHO pelleted diet


T as as


30 60
Time after ingestion (min)


Figure 3-3. Changes in intragastric bag volume after ingestion of either the high-fat meal
or the high-carbohydrate (CHO) pelleted meal. Data are expressed as mean
bag volume + SEM of 2-min blocks (n=6). Each postprandial volume
represents values from which the baseline volume has been subtracted.
Symbols (*) and (Jf) denote blocks that were significantly different from
baseline after ingestion of the high-fat meal and high-carbohydrate meal,
respectively. Neither meal contained 13C-Octanoic acid. Note: Blocks with
negative values represent mean bag volumes that were lower than the baseline
volume .

Data from previous work in our lab113 Suggested that duration of receptive

relaxation increases with duration of meal ingestion in the horse. In particular, ingestion










of a large hay meal took longer time to be finished and resulted in longer receptive

relaxation than ingestion of a smaller hay meal, or two different sweet feed meals of

weights identical to those of the hay meals. However, this relationship was not found in

the present study, since ingestion of the high-fat meal took longer time to be finished, but

resulted in a shorter receptive relaxation period, than ingestion of the high-carbohydrate

meal (Fig. 3-3). It may be possible that differences in degree of receptive relaxation

between diets are associated with duration of ingestion only when meals differ largely in

the latter. For example, the difference in duration of ingestion between the large hay meal

and the other meals of the previous studyll3 was considerably greater (448-889 sec) than

the difference between the two pelleted meals used in the present study (70 sec) (Table 3-

1).

Table 3-1. Comparison of duration of meal ingestion between the present study (A) and
that of Lorenzo-Figueras and Merrittl13 (B). Data are expressed as mean (sec)
+ SEM. In (A), ingestion of the high-carbohydrate (CHO) pelleted meal
induced longer receptive relaxation than the high-fat pelleted meal. In (B),
ingestion of hay (1 g/kg bwt) induced longer receptive relaxation than any of
the other meals (p<0.05).
A) Present study Duration of Difference between
ingestion high-fat and high-CHO
Pelleted meals (0.5 g/kg bwt)
High-fat 228 & 30a
High-CHO 158 & 14b 70 sec
Sweet feed meals (0.5 g/kg bwt)
High-fat 241 &31
High-CHO 267 & 35 26 sec
B) Lorenzo-Figueras and Merritt, 2002 Duration of Difference from
ingestion Hay (1: A~ .~
Large meal (1 g/kg bwt)
Hay 1065 &85a 0 sec
Sweet feed 281 +15b~d 784 sec
Small meal (0.5 g/kg bwt)
Hay 617 &79b.c 448 sec
Sweet feed 176 &17b~d 889 sec
Within the same study (A or B), values with different superscript letters differ
significantly (p<0.05).









Therefore, in the present study, the difference in consumption time between the pelleted

meals, although significant, may have not been large enough to produce differences in

receptive relaxation.

As indicated, receptive relaxation induced by ingestion of the high-carbohydrate

meal lasted 4 minutes longer than that induced by ingestion of the high-fat meal. If

receptive relaxation were solely dependent on mechanical stimulation of the oropharynx,

composition of the diet would have no influence on the magnitude of relaxation.

However, the results of this study suggest that other mechanisms may play a role in the

control of receptive relaxation in the horse. One of the possible factors may be

orosensory stimulation. That is, meals varying in composition may provide different

orosensory signals that may, in turn, induce different gastrointestinal responses.27 The

importance of orosensory factors controlling other gastrointestinal parameters, such as

hunger, satiation and gastric emptying, has been previously shown.26;27 For example, in

one study,27 Oral administration of a high-fat soup slowed gastric emptying more than an

isocaloric high-carbohydrate soup, but gastric emptying rates were similar when the

respective soups were infused into the stomach. Therefore, suppression of orosensory

stimulation eliminated the effect of dietary composition on the control of gastric

emptying. Likewise, subtle differences in orosensory stimulation following ingestion of

the high-fat meal and the high-carbohydrate meal may be responsible for the observed

difference in receptive relaxation. In conclusion, control of gastric receptive relaxation in

the horse may depend on other factors besides mechanical stimulation.

The initial relaxation episode, which has been defined as receptive relaxation, was

followed by a period of baseline tone, until the onset of a second, less profound,









accommodation phase, starting at about one hour of ingestion of the high-fat meal (Fig.

3-3). However, this was not observed following the high-carbohydrate meal. It is

proposed that accommodation may be induced by activation of either of two vago-vagal

reflexes. The first one begins when passage of feed into the stomach stimulates

mechanoreceptors within the gastric wall. The second one is initiated when chyme

delivered into the intestine stimulates intestinal mechano-41 and chemoreceptors.8

Activation of any of these reflex pathways will result in such accommodation or adaptive

relaxation.2 In the present study, accommodation was observed long after the meal had

entered the stomach and, therefore, it is more plausible that it was elicited by small

intestinal feedback regulation. In the previous study in our lab,113 receptive relaxation

after ingestion of a sweet feed meal, of equal weight (0.5 g/kg) and similar volume (450

ml) to these of the pelleted diets, was immediately followed by accommodation. Since

the accommodation reflex is presumably triggered by stimulation of gastric

mechanoreceptors, the same response pattern would be expected when the pelleted meals

of the present study, which had a similar volume (400 ml), entered the stomach.

However, in this case, the accommodation was not observed until one hour after

ingestion, long after the receptive response (Fig. 3-3). In conclusion, these results suggest

that the accommodation response to the high-fat meal was induced by passage of food

into the small intestine.

In contrast to the high-fat meal, no accommodation was observed after ingestion of

the high-carbohydrate meal, except for short, discrete intervals towards the end of the

recording session. Since barostat recording was limited to a postprandial period of 90

minutes, we cannot discount the possibility that accommodation occurred late after










ingestion. A higher fat content of the high-fat meal may explain the earlier occurrence of

accommodation that was not seen after the high-carbohydrate meal, i.e., the greater the

amount of fat in the diet, the earlier the accommodation. As observed in humans,51;129;217

the magnitude of intestinal feedback on motility of the proximal stomach may be

dependent on nutrient composition, where fat induces accommodation, and hypertonic,

but not isosmotic, carbohydrate preparations elicit the same response.51;12 As well, the

effect of fat on proximal gastric tone appears to be independent of its energy content and

osmolality,129 and be mediated by release of CCK from the upper intestine.138;203 In the

dog,s the effect of intraduodenal nutrients on accommodation is region-specific.

Specifically, fat infused into the proximal intestine induces accommodation, whereas

isocaloric and isosmotic carbohydrate has little effect. On the other hand, infusion of the

same carbohydrate solution, but not fat, into the distal intestine reduces gastric tone.

Therefore, as in dogs, the accommodation response may be both nutrient- and site-

specific in the horse.

Finally, we should consider the fact that the test meals used in the Phase I series of

studies differed in more than just their fat and carbohydrate composition. Any other

factor related to differences in composition, such as the higher fiber content of the high-

fat meal, could also account for the difference in the accommodation responses.

Phase II. Sweet feed Meals: Corn Oil Versus Glucose

To better determine the specific effect of fat and carbohydrate on the meal-induced

relaxation of the proximal stomach, two test meals that only differed in their fat and

carbohydrate contents were formulated. Test meals consisted of a 10% crude protein

sweet feed meal (0.5 g/kg) supplemented isocalorically with either corn oil or glucose.

Experiments were performed with and without labeling of the test meal with 13C-Octanoic









acid. An additional labeled test meal, consisting of a sweet feed meal (0.5 g/kg) with no

supplementation, was added as a control. The same base sweet feed (Seminole Feed@,

Blue Ribbon 10), without enrichment, was used as the horses' daily grain ration (0.25

g/kg/day) .

Unlabeled meals

Duration of meal ingestion. Horses spent an average of 241 & 3 1 seconds (range,

180 to 376 seconds) and 267 & 35 seconds (range, 172 to 400 seconds) to finish the corn

oil-enriched meal and the glucose-enriched meal, respectively. Duration of ingestion was

not significantly different.

Relaxation response. Basal bag volume did not differ significantly between diets.

Similar to the pelleted meals, beginning of ingestion of either meal was associated with a

rapid increase in bag volume (Fig. 3-4A,B). A peak volume was observed soon after the

end of ingestion, followed by a gradual decrease in bag volume. This receptive relaxation

response was more prolonged after the glucose-enriched meal than after the corn oil-

enriched meal, which was evidenced by significantly (p<0.05) higher volumes at 12-16

min after ingestion of the former, compared to the latter (Fig. 3-5). Intragastric bag

volume remained significantly (p<0.05) greater than baseline volume for the entire

postprandial phase of the glucose-enriched meal experiments (90 minutes), whereas it

returned to baseline volume at the end of the recording period in the oil-enriched meal

experiments. Bag volume was significantly greater than baseline for all periods, except

for the corn oil diet during the first postprandial two minutes and the last four minutes of

barostat recording.








54




A) C)
1400, 1400
... 12001200
1000Ol L 1000
800 800
600 600o
400 400
S200-II~I~l 200

015 30 45 60 75 90 105 0 15 30 45 60 75 90 105
Time (min) Time (min)

B) D)
14001 1400
1200 1200


400o- 4000

2 00 200

015 30 45 60 75 90 105 0 15 30 45 60 75 90 105
Time (min) Time (min)

E)
1400
1200




S400
200,

015 30 45 60 75 90 105
Time (min)


Figure 3-4. Mean volume trace (n=6) of the effect of ingestion of different sweet feed
meals (0.5 g/kg) on baseline tone in the proximal stomach. A) Corn oil-
enriched sweet feed meal; B) Glucose-enriched sweet feed meal; C) Corn oil-
enriched sweet feed meal labeled with octanoic acid; D) Glucose-enriched
sweet feed meal labeled with octanoic acid; E) control sweet feed meal (no
enrichment) labeled with octanoic acid. Beginning of ingestion of either meal
was associated with a rapid increase in bag volume, followed by a more
gradual decrease in volume. Time 0 is start of recording period, and beginning
of ingestion was at 30 minutes. The box limits ingestion time.


The relaxation response observed after ingestion of the sweet feed meals shares


some characteristics with that observed after ingestion of the pelleted meals. First,


ingestion of the sweet feed meals induced an initial reduction in tone of the proximal


stomach, and the length of this relaxation episode was associated with time of active


ingestion. Second, this receptive relaxation was four minutes longer after consumption of










the glucose-enriched meal, compared to the corn oil-enriched meal. The latter similarity

reinforces the idea that dietary composition affects the magnitude of receptive relaxation

in the horse, and that dietary carbohydrates seem to prolong this response, when

compared to dietary fat. The origin of this time difference cannot be explained only by

time of meal ingestion, since horses spent similar times consuming any of the sweet feed

meals. As discussed earlier, a combination of factors, including orosensory influences,

may be implicated in the control of this receptive relaxation reflex.



-g Glucose enrichrrnt
1200 $ 4 $ Corn oil enrichrrnt

1000


E 600


0) 400

200


10 20 30 40 50 60 70 80 90
Time after ingestion (min)


Figure 3-5. Changes in intragastric bag volume after ingestion of a 10% protein sweet
feed meal (0.5 g/kg) enriched by either corn oil or glucose (n=6). Both meals
were isocaloric (1.95 kcal/kg bwt) and isovolumetric (~ 400 ml). Data are
expressed as mean bag volume + SEM of 2-min blocks (n=6). Each
postprandial volume represents values from which the baseline volume has
been subtracted. Asterisks denote a significant difference between pairs of
blocks at the same postprandial time (p<0.05).

Few studies in the current literature address the effect of dietary composition on

proximal gastric relaxation, and differences in methodology make comparisons difficult.

For example, the protocol of some studies involved the direct administration of test meals

into the stomach or small intestine, thus bypassing possible orosensory stimulation."'









Another example is the use of methods other than the electronic barostat to measure

meal-induced relaxation. One of these methods consists of stepwise inflation of an

intragastric bag, and subsequent analysis of variations in intragastric pressure in relation

to increasing intrabag volumes.129 This method does not provide a continuous trace

showing changes in gastric tone over time, like that obtained with the barostat technique,

and direct comparison with our study is not possible. Finally, consistency of the test

meals (i.e., liquid versus solid), which varies among studies, may affect the outcome. For

example, in one study217 in which subj ects consumed a liquid carbohydrate meal (0 g fat)

or the same meal supplemented with fat (28 g fat), addition of fat to the carbohydrate

drink resulted in greater proximal gastric relaxation. Meals used in that study were liquid,

and oropharyngeal stimulation was probably brief because they were rapidly swallowed.

Therefore, the single relaxation observed after ingestion of these meals was probably

accommodation, and not receptive relaxation. In contrast, the meals used in the present

study were solid, and the longer period of mastication and swallowing probably resulted

in a prolonged time of oropharyngeal stimulation. Therefore, we cannot expect the same

response when test meals of similar composition, but different consistency, are used.

Contrary to the response to the pelleted meals, receptive relaxation in response to

the sweet feed meals was followed immediately by accommodation. The induction of

accommodation in these studies was more likely originated in the intestine than in the

stomach. Since the accommodation reflex is triggered, in part, by stimulation of gastric

mechanoreceptors, rather than chemoreceptors,123 arrival of the pelleted or the sweet feed

meals of similar weight and volume into the stomach should result in a similar degree of

mechanical stimulation and, therefore, similar onset of accommodation. However,









accommodation after the sweet feed meals occurred immediately after receptive

relaxation, whereas with the high-fat pelleted meal it was observed long after the meal

had entered the stomach. One possible explanation for the absence of stomach-originated

accommodation is that, in the horse, the contribution of the stomach in the relaxation

response may not be as important as that of the oropharynx and the intestine. The idea

that degree of accommodation depends on the site where the reflex originates has been

suggested before. In particular, it has been shown that the duodenum is a stronger

triggering site for the accommodation reflex than the stomach in humans.218

Alternatively, larger meals may be needed in the horse to trigger a distinctive

accommodation response arising from the stomach.

The earlier onset of accommodation, which it was probably controlled by intestinal

feedback, may be explained by the composition and physical characteristics of the sweet

feed meals. Digestion and absorption of carbohydrates and fat are required to activate

intestinal chemoreceptors involved in control of gastric emptying.60;171 The glucose

contained in the glucose-enriched sweet feed meal was ready for rapid absorption by the

intestine, whereas the carbohydrates of the high-carbohydrate pelleted meal, which were

in the form of starch, needed digestion before absorption. Similarly, fat of the corn oil-

enriched sweet feed meal was possibly more readily accessible for digestion because of

the physical characteristics of the meal, i.e., liquid fat was poured into sweet feed to

prepare the fat-enriched sweet feed meal, whereas fat was homogenized and integrated

into pellets in the high-fat pelleted meal. Thus, the type of carbohydrate and the physical

presentation of fat of the high-carbohydrate and high-fat sweet feed meals, respectively,

may have facilitated earlier digestion and absorption of these nutrients, causing earlier









activation of intestinal chemoreceptors and, in turn, earlier onset of accommodation,

compared to the pelleted meals.

Thus, the results of this study suggest that carbohydrate seems to magnify the

receptive relaxation, when compared to dietary fat, whereas intestinal modulation of

accommodation by fat and carbohydrate seems to be similar. These observations are

contrary to the original hypothesis, since, based on studies performed in humans,129;213 it

was expected that fat would induce greater relaxation of the proximal stomach than

carbohydrate.

Octanoic-acid labeled meals

Based upon the findings that addition of 13C-Octanoic acid to the corn oil- or the

glucose-enriched sweetfeed meals had no significant effect on the relaxation responses,

the outcome of these separate experiments was used to compare relaxation induced by the

corn oil- and glucose-enriched meals to that of the control meal.

Duration of meal ingestion. Mean & SEM duration for complete meal ingestion

was 165 & 2 seconds (range, 160 to 172 seconds) for the control meal, 280 & 48 seconds

(range, 164 to 450 seconds) for the corn oil-enriched meal, and 310 & 78 seconds (range,

164 to 672 seconds) for the glucose-enriched meal. A Friedman's two-way ANOVA,

used for meal comparison because of inequality of variances, showed that time to finish

the control meal was not significantly (p=0.0839) shorter than that of the other meals, and

neither did it differ between enriched meals.

Relaxation response. Baseline bag volume did not differ significantly among

diets. Mean bag volume & SEM of postprandial 2-minute blocks is presented in Table 3-

2. Barostat bag volume increased significantly after ingestion of each meal (Fig. 3-4C-E),

reached a peak and started to decline gradually after the end of ingestion. Bag volume









remained above baseline for the entire experiment (90 minutes) after ingestion of the comn

oil- and the glucose-enriched meals, whereas it only remained above baseline during the

first 40 minutes after ingestion of the control meal (Fig. 3-6). During the initial peak,

mean bag volume after ingestion of the control meal was significantly lower than after

ingestion of the glucose-enriched meal (16-minute period) and the comn oil-enriched meal

(6-minute period). For the rest of the recording session, bag volume of the control group

did not differ significantly from that following the other meals except at 48-62 minutes,

where it was significantly lower (p<0.05) than the glucose-enriched meal group. Finally,

mean bag volume after ingestion of the glucose-enriched meal was significantly (p<0.05)

higher than after the corn oil-enriched meal only at 48-50 minutes.

The receptive relaxation responses after the labeled meals (Fig. 3-6) were

qualitatively similar to those of the unlabeled meals (Fig. 3-5). From the standpoint of

statistical analysis, the receptive relaxation after the glucose-enriched meal was not

significantly longer than the corn oil-enriched one, possibly because of the effect of

octanoic acid labeling. A tendency for a longer time of meal ingestion may account for

the greater relaxation of the enriched meals, compared with the control meal. Another

possibility is that orosensory mechanisms may have come into play as a consequence of

this enrichment.

Supplementing the sweet feed meal with glucose or comn oil not only affected the

receptive relaxation, but also the magnitude of accommodation. Meal supplementation

resulted in a longer accommodation response that may have been caused by stronger

feedback regulation from the intestine.








60



Table 3-2. Postprandial variations in volume* of an intragastric bag controlled by an
electronic barostat.

Min after feedingt COntrol meal Control meal +cornoil Control meal +glucose
0-2 375 + 1071 320 + 801 356 +611
2-4 754 + 1441 725 + 891 852 +1141
4-6 799 + 122 925 + 511 1001 + 1241
6-8 667+ 145 917+59 990+ 97
8-10 546 +112a 881 +40b 965 + 93b
10-12 352 + 139a 711 +34b 925 + 79b
12-14 281 +136a 622 + 63b 853 + 144b
14-16 286 + 77a 595 +96a.b 781 + 174b
16-18 434 + 92a 662 + 140ab 839 + 176b
18-20 328 +117a 627 + 145ab 782 + 197b
20-22 438+ 146 557+ 138 735+ 209
22-24 302 +110a 572 + 167a.b 729 + 187b
24-26 439+ 99 489+ 110 613+ 143
26-28 597+ 153 668+ 176 745+ 188
28-30 625+ 132 534+ 134 626+ 171
30-32 410+ 127 493+ 116 581+ 175
32-34 441+ 145 527+ 110 580+ 168
34-36 484+ 140 356+ 116 542+ 193
36-38 392+ 70 434+ 108 644+ 106
38-40 271+ 56 439+ 101 574+ 100
40-42 314+ 74 357+ 126 570+ 137
42-44 272+ 127 295+ 103 572+ 122
44-46 158+ 112 287+ 94 472+ 139
46-48 274+ 117 288+ 82 538+ 167
48-50 218 +127a 278 + 100a 603 + 107b
50-52 219 + 130a 326 + 91ab 639 + 130b
52-54 156 +112a 265 + 109a.b 497 + 91b
54-56 138 +117a 266~ + 111.b 513 +124b
56-58 214+84 382+ 101 464+ 115
58-60 116+ 106 296+ 92 538+ 87
60-62 157+ 74 331+98 552+ 116
62-64 249+ 92 422+ 103 447+91
64-66 206+ 102 397+ 88 488+ 108
66-68 212+97 348+ 90 483+ 96
68-70 230+ 64 332+ 82 452+ 56
70-72 159+ 105 335+ 143 316+68
72-74 126+ 75 306+ 111 414+71
74-76 147+ 71 293+ 80 406+ 89
76-78 114 +75a 313 +162a.b 477 + 86b
78-80 73+80 392+ 142 360+ 77
80-82 118+83 419+ 159 394+ 130
82-84 171+93 333+ 125 325+89
84-86 131+90 317+98 431+62
86-88 123+93 367+ 144 319+ 102
88-90 119+ 78 290+ 105 327+ 114
*Volume is mean + SEM number of milliliters (n=6). Each postprandial volume
represents values from which the baseline volume has been subtracted.
"fTime 0=onset of ingestion of each meal. (Periods associated completely or
partially with active ingestion of a meal. a-bWithin a row. values with different
superscript letters differ significantly (p<0.05).












+ Glucose-enriched rnal
-e Corn oil-enriched rnal
-A- Control rnal


10 20 30 40 50
Time after ingestion (min)


60 70 80 90


Figure 3-6. Changes in intragastric bag volume after ingestion of a control sweet feed
meal (0.5 g/kg) with and without addition of either corn oil or glucose. All
meals were labeled with 13C-Octanoic acid. Data are expressed as mean
volume during 2-min blocks (n=6). Each postprandial volume represents
values from which the baseline volume has been subtracted. Bag volume was
significantly higher (p<0.05) than baseline during the entire length of the
experiment (90 minutes) after ingestion of either enriched meal, whereas it
only remained significantly above baseline for 40 minutes after ingestion of
the control meal. Symbols indicate significant difference between diets within
the same time period: (*) glucose-enriched meal versus control meal, (Jf) corn
oil-enriched meal versus control meal and (8) glucose-enriched meal versus
corn oil-enriched meal. SEM values are shown in Table 3-2.

Conclusions

Methodology

Electronic barostat. An electronic barostat has been previously used in our lab to

asses tone of the proximal stomach in the horse.113;11 The lack of significant differences

in baseline volumes among experiments supports that this is a valid and reliable

technique to study changes of proximal gastric tone induced by meal ingestion in the


horse.









Simultaneous use of the barostat and 1C-octanoate breath test. The concurrent

use of the 13C-Octanoate breath test modified the relaxation response measured by the

electronic barostat. This effect was diet-specifie, because it increased relaxation after the

high-carbohydrate pelleted diet, but not that after any of the other diets. One possible

explanation of this interaction is that labeling with the breath test marker increased the fat

content of this diet to a level that surpassed a fat threshold to induce relaxation. In

contrast, the initial fat content of the high-fat pelleted meal was possibly above this

potential threshold, so that the extra fat content of the breath test label had no further

effect on relaxation. To confirm this theory, two properties of the relaxation response in

the equine stomach should be determined in further studies: first, the existence of a

threshold and, second, the absence of dose-dependency of fat-induced relaxation of the

proximal stomach. Finally, the greater accommodation observed after ingestion of the

sweet feed meals may have masked the possible influence of the breath test label on the

magnitude of relaxation.

In summary, the process of labeling a test meal to carry out the 13C-Octanoate

breath test may influence relaxation of the proximal stomach induced by this meal. The

probability that such interaction occurs may be high when test meals with originally low

fat content are used.

Relaxation Response

Overall results. Ingestion of any test meal induced relaxation of the proximal

stomach. However, the pattern of relaxation, i.e., number of relaxation episodes and their

time of onset and duration, differed among diets.

Receptive relaxation. Although the pattern varied among diets, every diet induced

an initial receptive relaxation phase. That the length of this phase was associated with










time of meal ingestion suggests that it was induced by activation of oropharyngeal

receptors, and thus it corresponds to what could be termed receptive relaxation.

Based on the results of the present study, dietary carbohydrate seems to be a more

potent stimulus of receptive relaxation than fat in the horse. This fact cannot be explained

only by time of meal ingestion, because times of ingestion of the high-carbohydrate

meals were not longer than those of the high-fat meals. Thus, other factors besides

mechanical stimulation of the oropharynx during mastication and swallowing must be

involved. One possible factor is the involvement of nutrient-specific orosensory

receptors. Although orosensory mechanisms seem to control some gastric functions, such

as gastric emptying, in other species, nothing is known with regard to the meal-induced

receptive relaxation of the proximal stomach. Therefore, more studies are necessary to

determine the existence and influence of orosensory factors in the horse, and in other

species.

Finally, enrichment of a sweet feed meal with corn oil or glucose increased the

meal-induced receptive relaxation, compared to the control sweet feed meal, suggesting

that orosensory stimulation seems to be important in the response to both fat and

carbohydrate.

Accommodation. A second component of meal-induced relaxation, which is

consistent with accommodation, was observed with all meals, with the exception of the

high-carbohydrate pellets.

Time of onset of accommodation varied among meals. That induced by the sweet

feed meals occurred immediately after receptive relaxation, whereas that induced by the

high-fat pelleted meal was not seen until one hour after. Finally, barostat recording was









limited to a postprandial period of 90 minutes and, therefore, we cannot discard the

possibility that accommodation after the high-carbohydrate pelleted meal came about

even later than that of the high-fat pelleted meal. Since all meals had similar weight and

volume, it is probable that the intestine, and not the stomach, was the triggering site of

this accommodation response. Therefore, the intestine may be more important in

controlling accommodation than the stomach in the horse.

The importance of fat versus carbohydrate on the magnitude of accommodation is

inconsistent in this study. On the one hand, accommodation was observed with the high-

fat pelleted meal, but not with the high-carbohydrate meal. On the other hand, there was

no difference in the magnitude of accommodation between the corn oil- and the glucose-

enriched sweet feed meals. A higher fat content of the high-fat pelleted meal may explain

the induction of accommodation that was not seen after the high-carbohydrate pelleted

meal. However, both pelleted meals differed in other components besides fat and

carbohydrate and, therefore, we cannot conclude that such difference was caused merely

by their difference in fat and carbohydrate content. In contrast, both sweet feed meals

differed only in that they were supplemented with either corn oil or glucose. Therefore,

the difference in accommodation between meals could be attributed specifically to the

corn oil and the glucose. The magnitude of accommodation was similar between the

glucose- and the corn oil-enriched sweet feed meals. Thus, intestinal modulation of

accommodation by fat and carbohydrate seems to be similar in the horse. These

observations are opposite to the original hypothesis of the study, since it stated that fat

would induce greater relaxation of the proximal stomach than carbohydrate.









Finally, although both enriched sweet feed meals induced similar accommodation,

the process of enrichment prolonged the time of accommodation. The origin of this

longer accommodation may have been stronger feedback regulation from the intestine,

and shows that both glucose and corn oil can modulate this response.














CHAPTER 4
RESULTS AND DISCUSSION GASTRIC EMPTYING

The passage of ingested food from the stomach to the small intestine is achieved by

the coordinated action of the proximal stomach, the antrum, the pylorus and the proximal

duodenum. The emptying rate of a meal may be influenced by its volume, nutritional

constituents, physical structure, temperature, caloric density, osmolarity and the amount

of acid produced by the stomach in response to this meal.124 A maj or control of gastric

emptying is accomplished by nutrient-induced small intestinal feedback regulation.119

That is, lipids, certain amino acids, sugars, and nutrients of high osmolality trigger

sensory mechanisms from the intestine that inhibit gastric emptying. Results of studies

done to date indicate that, of these nutrients, the most potent inhibitor of gastric emptying

seems to be fat.25;173;180

Little is known about the factors controlling gastric emptying of liquids and solids

in the horse, and to what extent fat and carbohydrate participate in intestinal feedback

inhibition. Studies in other species show that fat causes a greater inhibition of gastric

emptying than carbohydrate.25;27;180 Based on that, the hypothesis of this part of the study

stated that, in the horse, a high-fat meal would have slower gastric emptying than an

isocaloric and isovolumetric high-carbohydrate meal. To test this hypothesis, rates of

gastric emptying were measured following ingestion of a high-fat or a high-carbohydrate

meal, by use of the 13C-Octanoic acid breath test.









Effect of Dietary Composition on Gastric Emptying

Phase I. Pelleted Meals: Fat Versus Carbohydrate

Effect of diet on basal 1C expiratory output

As mentioned, every horse participated in two experiments, which differed in the

test meal offered to the horse: a high-fat pelleted meal or a high-carbohydrate pelleted

meal. Experiments were performed after at least one week of adaptation to the respective

diet. Basal metaboliC 13C prOduction was measured by following the protocol for the 13C

octanoic acid breath test but without the addition of the isotope. The results of both diets

were compared to determine the influence of diet on basal 13C Output.

Data from four tests (two within each test meal group) were not included in the

observational analysis for suspected errors in breath samples analysis. In three of these

cases, ingestion of the unlabeled test meal was associated with unexpected increases in

13CO2 COncentration in breath samples (as it would be expected with labeled meal

consumption). In the fourth test, exhaled 13CO2 declined markedly over time compared to

the rest of the tests within the same meal group, i.e., the high-fat diet. Mean trace of

postprandial 13CO2 expiration was similar between both diets over time.

In the current study, horses were maintained on a~d libitum Bermuda hay and 5

g/kg/day of one of the pelleted diets. The results of the basal breath tests showed that the

effect of any of the pelleted meals on basal 13C prOduction was similar. With regard to

Bermuda hay, it is known that this type of hay interferes with the outcome of the 13C

octanoic acid breath test. Because Bermuda hay is rich in natural 13C, its ingestion

increases the equine endogenous production of 13C. This endogenous source of 13C COuld

interfere with the signal produced by the metabolism of the 13C-Octanoic acid tracer and,

therefore, it is recommended that Bermuda grass should be avoided as a maintenance diet









to minimize fluctuations in basal 13C prOduction. Alternatively, a higher dose of isotope

should be used in the test meal.205 Since changing Bermuda hay as the maintenance diet

was not an option in the present study, the standard dose of the isotope, 1 mg/kg, was

increased to 1.5 mg/kg.

Effect of diet on gastric emptying

Because one breath test of the high-fat group was excluded from the data analysis

for suspected error in breath samples analysis, a two-sample t-test was used to compare

parameters of gastric emptying between diets. Mean % dose recovery/h of the 13C tracer

and modeled dose recovery curves after each test meal are shown in Fig. 4-1. The shape

of the curves was very similar between diets. However, the high-fat meal group tended to

reach a higher and more exponential peak for isotope recovery, compared to the high-

carbohydrate meal group. This peak was underestimated by the modeling function and

led to slight underestimation of tmax.

Mean values for tl/2, tmax and GEC are summarized in Table 4-1. There was no

significant difference between both meals for any of the gastric emptying indices.

In the 13C-Octanoic acid breath test technique, the pattern of gastric emptying of a

meal is described, in part, by the rising slope of the dose recovery curve and the time to

reach the maximal recovery of 13C. The slope of the curve gives information about the

onset and speed of gastric emptying, which may be affected by the presence and duration

of a lag phase and the degree of feedback inhibition. Thus, a steeper rising slope will

indicate lesser inhibition of gastric emptying, and may, additionally, indicate a shorter lag

period. Finally, the peak of 13C TOCOVery also describes the magnitude of gastric

inhibition and, therefore, a higher peak indicates that gastric emptying of a meal can

achieve a faster rate over time.59;11







69






+ High-fat pelleted diet
o High-CHO pelleted diet


8-I .





4-I

3-f






0 1 2 3 4 5 6 7
Time (h)



Figure 4-1. Mean percentage dose recovery (PDR/h) + SEM and modeled curve of 13C in
breath following ingestion of a high-fat pelleted meal (n=5) or a high-
carbohydrate (CHO) pelleted meal (n=6). Gastric emptying parameters did not
differ significantly between diets.

Table 4-1. Comparison of gastric emptying parameters determined by the 13C-Octanoate
breath test. Within a phase study, no significant differences were found among
diets. See Materials & Methods for description of composition of the diets.
tl/2 (h) tmax (h) GEC
Test meal Mean Range Mean Range Mean Range
Phase I studies
High-fat pelleted diet* 3.06 & 0.37 1.96-3.59 1.92 & 0.19 1.51-2.63 2.24 & 0.39 1.29-3.42
High-CHO pelleted diet 3.55 & 0.82 1.65-7.16 2.07 & 0.42 1.08-3.82 2.17 & 0.35 0.88-3.20
Phase II studies
Control diet 2.30 + 0.15 1.84-2.74 1.52 & 0.08 1.38-1.87 2.81 & 0.27 2.28-3.98
Corn oil diet 2.56 & 0.23 1.81-3.49 1.69 & 0.08 1.46-1.95 2.74 & 0.25 1.89-3.65
(without intragastric bag) 3.00 + 0.22 2.26-3.84 1.96 & 0.12 1.50-2.28 2.38 & 0.14 1.92-2.97
Glucose diet 2.59 & 0.31 1.93-3.75 1.89 & 0.22 1.34-2.69 2.57 & 0.27 1.56-3.47
(without intragastric bag) 3.08 & 0.41 1.86-4.67 2.19 & 0.20 1.59-2.71 2.52 & 0.37 1.42-3.81
Data are reported in mean & SEM; t,/,=half-dose recovery time; tmax=time to peak 13CO, concentration; GEC=gastric
emptying coefficient; CHO=carbohydrate; n=6, except for *n=5.









Few studies have been conducted to evaluate the mechanisms controlling gastric

emptying in the horse. A limitation of some of these studies is the use of the phenol red

technique,12;199 which is known to be less accurate to assess gastric emptying than

scintigraphy or the 13C-Octanoic acid breath test. Additionally, some studies include a low

sample size (3-4 horses),58;234 along with high interindividual variation.197

Based on this limited literature, both dietary fat and carbohydrate appear to have

the ability to modulate the rate of gastric emptying in the horse. First, two studies showed

that addition of fat, i.e., soybean234 Of COTH Oi1,58 to a concentrate meal resulted in delay of

gastric emptying. This effect was not affected by previous adaptation to dietary fat,'" and

seemed to be independent of meal volume and dose of fat.234 The increase in energy or

viscosity that resulted from adding oil to the concentrate meal might be responsible for

this slowing effect. Another possibility is that fat delayed emptying by activating

intestinal receptors specific for this nutrient. With regard to carbohydrates, one study

showed that addition of dextrose to water emptied more slowly than the same volume of

water in three ponies.197 The authors of this study suggested that caloric density,

osmolality of the solution or activation of nutrient-specific receptors within the intestine

might have been involved in this effect.80;106;148

In contrast to these previous studies, the high-fat pelleted diet of the present study

had the same energy and protein content, volume and physical characteristics (size,

consistency) as the high-carbohydrate pelleted diet. Should any of these factors be

responsible for the rate of gastric emptying of a meal, it would presumably be similar for

both of the test diets, and would explain the lack of significant differences in emptying









rate between the test meals. Another possibility is that the extent of nutrient-specific

feedback inhibition of fat is similar to that of carbohydrate in the horse.

Volume is thought to be a maj or factor controlling gastric emptying of a meal in

humans,so'si dogslos and monkeys.148 Its effect is probably mediated by activation of

gastric and intestinal receptors that are sensitive to mechanical distention.190 Whether

volume affects solid emptying appears to be dependent on the particle size of the meal.

That is, an increase in meal volume accelerates gastric emptying only when mechanical

breakdown of solids into smaller particles is not needed.31;105 In COntrast to the species

that are routinely used in gastric emptying studies, the horse is an herbivore whose

natural nutrition depends on a constant intake of high-fiber, low-energy food. Since the

horse has a relatively small gastric capacity, control of intake rate and gastric load may be

more important than control of nutrient delivery rate.234 In Other words, volume may be

more important in controlling gastric emptying than dietary composition. Nonetheless,

this factor could not be evaluated in the present study because test meals had similar

volumes, which could be considered as small for an adult horse.

The effect of energy content of a meal on gastric emptying rate is uncertain. On the

one hand, studies in humans" and monkeysl48 indicated that glucose solutions emptied

from the stomach at a constant caloric rate, independent of the initial concentration. This

observation was not limited to glucose, since isocaloric, isovolumetric liquids containing

fat, protein or carbohydrate emptied from the stomach at an overall constant rate in

man,s in monkeysl26 and in pigs.226 Thus, it was suggested that emptying rates of

isocaloric meals were similar because of comparable effects of nutrients on small

intestinal receptors.l7s On the other hand, other studies in humans showed that









increasing volumes of a glucose drinkso or a mixed solid/liquid meal31 were associated

with increased rate of energy delivery to the duodenum, at least initially. Finally, Gregory

et al.64 Showed that, in pigs, carbohydrates, but not fat, emptied from the stomach

following a constant caloric rate. Nevertheless, since test meals of the present study had

the same energy content, its effect on gastric emptying could not be determined in the

horse.

The high-fat and high-carbohydrate test meals of this Phase I study had similar

initial consistency. However, with ingestion, pellets were mixed within the mouth and in

the stomach to produce a viscous solution. The viscosity of such solutions may vary with

the amount of salivary and gastric secretions, and the relative presence of specific

ingredients, such as fat and fiber.65 Thus, since the high-fat meal was richer in fiber and

fat, it possibly became more viscous within the stomach. While the effect of fat on gastric

emptying does not seem to be mediated by its viscosity,35 fiber can influence gastric

emptying rate by its capacity to retain water and to increase the viscosity of gastric

contents. In other species, this effect depends on the type of fiber. For example, in pigs65

and humans,176 Only soluble fibers appear to delay gastric emptying, whereas in the rat,

insoluble fiber delays, and soluble fiber accelerates, emptying.20 Both soluble and

insoluble fiber was proportionally greater in the high-fat pellets and, thus, may have

affected gastric emptying of this meal in combination with other factors, such as specific

nutrient source. We do not know to what extent fiber affects rate of gastric emptying in

the horse. Yet, it would be reasonable to presume that fiber per se does not inhibit gastric

emptying. Fiber constitutes the base of herbivores diet and, therefore, a potential delay of









emptying by fiber would result in earlier satiation and, in turn, a slower rate of an already

low energy intake.

Fat slows gastric emptying by a mechanism that involves stimulation of intestinal

receptors that are sensitive to its digestion products.135:163 This effect is determined

primarily by its chemical composition more than its physical characteristics (lower

density and higher viscosity),35 dose of fat and degree of homogenization. In particular,

the slowing effect of fat is increased with increasing amounts of fat and decreased

homogenization.34 One reason that no difference in emptying rate was found between the

high (8%) and low (3%) fat pellets is that feedback mechanisms provoked by fat and

carbohydrate may not be very discriminatory in the horse. An alternative reason is that

previous adaptation of the horses to this diet may have eliminated the greater slowing

effect of fat versus carbohydrate. In humans, high-fat feeding induces adaptation,

ultimately reversing the slowing effect of a fatty meal and enhancing gastric

emptying.39:53 Finally, since it is the lipolytic products that trigger intestinal receptors to

induce slowing of gastric emptying, a limited capacity of the horse to digest fat could

explain the inability of the high-fat meal to have a greater suppressive effect in these

Phase I studies. Studies in humans and animals with pancreatic insufficiency indicate that

diminished duodenal lipase speeds gastric emptying of oil."" Yet, one of the most

abundant components of the equine pancreatic juice appears to be lipase (Dr. Jean

Morisset, personal communication), and it has been shown that the ability of the horse to

digest fat increases with increasing amounts of dietary fat.139 Therefore, it is unlikely that

a limitation in fat digestion explains the lack of difference in emptying rate of the two

pelleted diets.









Similar to fat, inhibition by carbohydrate mostly depends on chemospecific

intestinal feedback,106:152 and, at least for glucose, the magnitude of feedback correlates

with concentration.17:14 Carbohydrate solutions may also control emptying indirectly by

stimulating osmotic receptors within the duodenal mucosa.106 Another influential factor is

adaptation to dietary carbohydrate, which leads to increased gastric emptying rate of

carbohydrate in humans.76 Like fat, this phenomenon may have influenced gastric

emptying of the high-carbohydrate diet in the present study.

Most studies have evaluated the effect of fat by adding a lipid component to a

control meal. In few studies, such as the present one, the caloric increase associated with

fat supplementation was compensated for to cancel the possible effect of adding energy

on emptying rate. By use of the 13C-Octanoic acid breath test, Robertson and Mathersiso

showed in humans that a high-fat solid meal emptied slower than a high-carbohydrate

meal of identical energy and volume. Similar results were observed by scintigraphy with

liquid meals, but rates of gastric emptying for a high-fat and a high-carbohydrate soups

became similar when meals were infused directly into the stomach.27 Therefore,

orosensory signals may also be contributory to control of emptying rate. It has been

suggested in the previous chapter that, in the horse, nutrients may control the meal-

induced relaxation of the proximal stomach through orosensory mechanisms, and this

may also apply to gastric emptying.

Finally, we should consider the fact that the test meals used in this Phase I of the

present study differed in more than fat and carbohydrate content, and that a combination

of several factors regulating gastric emptying may be responsible for the present

observations. In addition, it is difficult to determine whether labeling with octanoic acid










may have influenced gastric emptying of the test meals, as it happened with the meal-

induced relaxation of the proximal stomach. Gastric emptying of a solid meal is

dependent on the integrated function of the proximal stomach, the antrum and the pyloric

sphincter and, therefore, changes in proximal gastric function may influence overall rate

of emptying. Nonetheless, the 13C-Octanoic acid breath test technique has been previously

used in humansiso and the horse234 to evaluate the effect of fat and carbohydrate on

gastric emptying, and labeling with 13C-Octanoic acid did not mask the dampening effect

of fat over carbohydrate.

Phase II. Sweet feed Meals: Corn Oil Versus Glucose

As mentioned, every horse participated in two experiments, which differed in the

test meal offered to the horse: a sweet feed meal supplemented with either comn oil or an

isocaloric amount of glucose. In contrast to the Phase I breath tests, these experiments

were performed without, as well as with, simultaneous use of the intragastric barostat

bag.

Breath tests without presence of an intragastric barostat bag

Mean % dose recovery/h of the 13C tracer and modeled dose recovery curves after

each test meal are shown in Fig. 4-2. Shape of the curves was very similar between diets,

except that the initial slope for the comn oil-enriched meal was slightly steeper and the

maximal rate of gastric emptying was faster, compared to the glucose-enriched meal.

Similar to the high-fat pelleted meal, the peak of emptying of the comn oil-enriched meal

was underestimated by the modeled curve, but was not significantly different from that of

the glucose-enriched meal.

Mean values for gastric emptying parameters are summarized in Table 4-1. None of

the parameters were significantly different between the two diets.









Like the pelleted meals, these results are opposite to the initial hypothesis that the

gastric emptying of a sweet feed meal supplemented with corn oil would be slower than

that supplemented with an isocaloric amount of glucose. However, both Phases of studies

were consistent in that there was a tendency for the high-carbohydrate meals to have

initially slower emptying rates than the high-fat meals, which suggests that carbohydrate

may cause more profound feedback inhibition than fat on early phases of gastric

emptying. A slower early phase of emptying of the glucose-enriched meal may have been

compensated by faster emptying rates at later stages, to yield a tl/2 Value Similar to that of

the corn oil-enriched meal. It is possible that the breath test parameters do not have

sufficient sensitivity to reflect differences in the early phase of gastric emptying of solid

food. Finally, the lack of significant difference between meals may also be due to the

small number of animals included in the study, combined with a high intersubj ect

variability.

The apparent tendency for the glucose-enriched meal to empty more slowly may be

due to a difference in the nutrient composition of the meals, but also by their physical

properties. Specifically, the relatively high volume of glucose (1 13 ml/500 kg bwt) added

to the sweet feed meal resulted in the formation of a small liquid phase. The liquid phase

of a meal is known to empty faster than the solid phasell9 and, therefore, earlier passage

of glucose to the duodenum may have led to earlier onset of feedback inhibition.

Finally, the sweet feed meals of this second Phase of studies differed only in the

glucose or corn oil content, in contrast to the pelleted meals of the Phase I studies.

Therefore, any component that was similar between the meals of Phase II may be

responsible for the lack of difference in gastric emptying parameters.






77







Corn oil-enriched meal
81 a Glucose-enriched meal


7-P
6-
5- 'E









0 1 2 3 4 5 6 7
Time (h)



Figure 4-2. Mean percentage dose recovery (PDR/h) 0 SEM and modeled curve of 13C in
breath following ingestion of a 10% crude protein sweet feed meal (Seminole
Feed@, Blue Ribbon 10) enriched with corn oil or glucose (n=6). Gastric
emptying parameters were not significantly different between diets.

Breath tests with presence of an intragastric barostat bag

A new series of breath tests were performed with the sweet feed meals used in the

previous section, but this time, with an intragastric bag in place to measure proximal

gastric tone simultaneously. An additional test meal, consisting of the same amount of

sweet feed without supplementation, was used as control.

Mean % dose recovery/h of the 13C tracer and modeled dose recovery curves after

each test meal are shown in Fig. 4-3. The shape of the curves was similar among diets,

although emptying of the glucose-enriched meal appeared to be slower and to reach a









lower maximal rate, compared to the other diets. Fitness of the modeled curves was

higher for the corn oil-enriched diet than for the other two diets.

Mean values for gastric emptying parameters are summarized in Table 4-1. Data of

tmax under the control diet were not normally distributed and, therefore, tmax values were

compared using a Friedman' s two-way ANOVA. None of the parameters were

significantly different among the test meals.

The emptying curves of the enriched meals are similar to those shown in the

previous section. Surprisingly, addition of corn oil to the sweet feed meal did not modify

the shape of gastric emptying of the control meal. Furthermore, addition of corn oil or an

isocaloric amount of glucose to the sweet feed meal resulted in similar overall gastric

emptying, compared to the sweet feed meal without enrichment. However, the previously

observed tendency for the high-carbohydrate meal to empty more slowly was also

observed with this new series of experiments.

These results are opposite to other studies in the horse. By use of the 13C-Octanoic

breath test, Wyse et al.234 TepOrted that addition of soybean oil to a small meal of oats and

bran caused a delay in gastric emptying. It is difficult to explain this discrepancy, but one

possibility is that the difference in the relative energetic contribution of fat to the meal

varied between studies. That is, in the study by Wyse et al., addition of the fat component

to the original meal resulted in an increment of ~60% of energy, whereas addition of corn

oil to the control meal of the present study increased the energy content in ~30%.

Therefore, a greater increase in energy might be necessary to show a significant effect of

fat on the breath test parameters. Another possibility is that the small sample size (6

horses) used in this study, combined with poor fitness of the curve observed in the control











group and, especially the glucose group, may be responsible, in part, for a lack of


significant difference between meals. Finally, the physiology of gastric emptying in


ponies may differ from that of horses.


SControl meal
9.0-
SCorn oil enriched meal
8.0 -I -E Glucose enriched meal

7.0-

6.0-

5.0-




3.0-

2.0-

1.0-

0.0
0 1 2 3 4 5 6 7
Time (h)

Figure 4-3. Modeled mean % dose recovery curves of the 13C label in the breath of 6
horses after ingestion of a sweet feed meal (control) or the same meal
enriched with corn oil or glucose. Experiments were done without the
presence of a barostat bag. Note: in contrast to the previous figures, symbols
on the lines are descriptive points of the modeled curves, not the original
mean % dose recovery values.

Another study, presented as an abstract, reported that addition of corn oil to a sweet


feed meal delayed gastric emptying of the sweet feed meal.'" The size of the meal was


four times larger, but the relative energetic contribution of fat was apparently similar to


that of the present study. Other factors may be responsible for such inconsistency


between studies, and more details of that study are needed for further comparison.









Effect of the Barostat Bag on Gastric Emptying

To assess whether the presence of the barostat bag within the proximal stomach

would alter gastric emptying, breath test results of the corn oil- and the glucose-enriched

sweet feed meals, measured with and without the presence of an intragastric bag, were

compared. No significant differences were found for t1 2, tmax and GEC (Table 4-1).

A potential disadvantage of the barostat technique is its intrusiveness, since it

requires positioning of a bag within the proximal stomach. Therefore, it has been

suggested that this may affect intragastric distribution and emptying rate of a liquid

meal.42:181 In COntrast, other studies have failed to show an effect of the barostat bag on

emptying of a liquid42:153 Or a solid/liquid meal.144 In the present study, the presence of

the barostat bag within the proximal stomach did not seem to affect gastric emptying of

the test meals.

Relation Between Proximal Gastric Relaxation and Gastric Emptying

The results of this study have shown that supplementation of the sweet feed meal

with glucose induced a more prolonged receptive relaxation than supplementation with

corn oil, or no supplementation at all. In addition, meal supplementation with glucose

showed a tendency to delay emptying at an early phase, compared to the other meals.

As mentioned in the previous chapter, the more potent effect of carbohydrate on the

receptive relaxation response may be explained by orosensory stimulation of

carbohydrate, which would lastly result in the modulation of proximal gastric tone. A

similar orosensory mechanism may explain the apparent tendency of carbohydrates to

slow gastric emptying at initial stages. The idea of orosensory factors modulating gastric

emptying is not new, since Cecil et al.27 Showed that the effect of oral feeding on gastric

emptying differed from that of intragastric feeding.









Conclusions

For all studies, the high-fat meals and the high-carbohydrate meals emptied from

the stomach at similar rates. These results are opposite to the prevailing notion that fat

has a more potent effect on regulation of gastric emptying than carbohydrates. 27;64;180;234

Yet, the present study is different from most studies in that test meals with identical

caloric content were used. Therefore, the results of this study support the idea that meals

with similar caloric content have similar emptying rates, regardless of nutrient

composition.81;126;226

Although dietary adaptation to fat and carbohydrate appears to eliminate the

slowing effect of these nutrients on gastric emptying in humans 39;40;76, this may not

explain the results of this study, since horses were not previously adapted to

supplementation with corn oil or glucose for the Phase II studies.

Another unexpected finding was that supplementing a sweet feed meal with either

corn oil or glucose, which resulted in a 30% increase in energy content, did not modify its

rate of gastric emptying. It is difficult to explain why supplementation had no additive

effect on inhibition of gastric emptying. Both the corn oil- and glucose-enriched sweet

feed meals had a high caloric density, which is far from the horse's natural diet. Thus, it

is possible that the mechanisms involved in regulation of gastric emptying in the horse

are a better reflection of the horse' s natural nutrition. In other words, the mechanisms

controlling gastric emptying may be aimed towards the continuous ingestion of a low-

energy diet, whereas they may be limited when the equine stomach is challenged with a

high-energy meal. Therefore, volume and, to a much lesser extent, energy may control

rate of gastric emptying.









Although breath parameters of gastric emptying obtained with the high-

carbohydrate meals did not differ significantly from those of the other meals, there was a

consistent tendency for the former to empty slower at the initial phase of gastric

emptying. The lack of significant differences between the high-carbohydrate meal and the

others may have resulted from a low sample size, or poor fitness of the modeled curves

for the high-carbohydrate diet, due to the higher variability observed within this meal

group. Alternatively, the parameters described by Ghoos et al.59 may not be sensitive

enough to detect significant differences between meals in the different phases of the

gastric emptying process. It would be very interesting to further explore the possibility

that carbohydrate may be more suppressive of gastric emptying than fat. Carbohydrates

may induce more feedback inhibition by stimulation of orosensory signals, as it has been

suggested for modulation of receptive relaxation in the previous chapter. Alternatively,

carbohydrates emptying into the duodenum may stimulate specific chemoreceptors or

osmoreceptors that result in rapid onset of intestinal feedback inhibition of emptying.

Finally, presence of an intragastric bag within the proximal stomach to measure

tone by an electronic barostat did not influence gastric emptying rates of the test meals.

This supports findings from studieS42;144;153 in Other species that show no effect of a

barostat bag on gastric emptying.














CHAPTER 5
RESULTS AND DISCUSSION -pH OF GASTRIC CONTENTS

Ingestion of food is associated with decreased gastric acidity in the horse,154 mainly

due to the buffering effect of bicarbonate-rich salivary secretions.4 Few studies have

measured meal-induced changes in gastric pH in the horse, but the effect of meal on pH

may vary with meal composition and acidity, degree of stimulation of salivary and acid

secretions, degree and products of meal fermentation within the stomach, and the gastric

emptying rate of the meal.4;6;110

Ulceration of the squamous portion of the equine gastric mucosa is highly prevalent

in adult performance horses,68;85;155;162;223 and excessive exposure to acid is believed to be

the main cause.114 Therefore, a potential strategy to reduce the risk of ulcer development

would consist of reducing the acidity of gastric contents. Dietary manipulation could be

one way to approach such strategy, through selection and use of equine rations with a

large buffering power. It is thus of interest to know to what extent different dietary

formulations affect the pH of gastric contents in the horse.156 This part of the study was

aimed to determine to what degree a diet rich in fat and low in carbohydrate might affect

intragastric pH, compared to a more traditional carbohydrate-based diet. Therefore,

variations in intragastric pH in response to a high-fat and a high-carbohydrate meal were

measured by use of a self-referencing electrode positioned in the most ventral portion of

the stomach. This site was chosen because it corresponds to the most acidic site of the

equine stomach. That is, layering of ingested food creates a proximal-to-distal pH

gradient within the equine stomach that results in the formation of a high-pH upper part,









rich in saliva and far from the glandular mucosa, and a low-pH bottom part, where acidic

secretions are produced.11:13

Effect of Dietary Composition on Intragastric pH

Phase I. Pelleted Meals: Fat Versus Carbohydrate

Changes in intragastric pH were measured before, during and after ingestion of a

high-fat pelleted meal, a high-carbohydrate pelleted meal, or any of these meals enriched

with 13C-Octanoic acid. Addition of the octanoic acid to the test meals did not have any

significant effect on intragastric pH at any time point. Therefore, results of experiments

with and without octanoic acid were combined for each diet in the analyses.

Mean baseline pH did not differ significantly between diets. Mean pH increased

significantly (p<0.05) from baseline 15 minutes after ingestion of either meal (Fig. 5-1).

Thereafter, it remained significantly higher (p<0.05) than baseline for 60 minutes

following the high-fat meal, and 80 minutes following the high-carbohydrate meal. Mean

pH after the high-carbohydrate meal was significantly higher (p<0.05) than after the

high-fat meal during 75 minutes post-ingestion. Therefore, ingestion of the high-

carbohydrate meal produced a significantly higher and more sustained increase in pH of

the ventral part of the stomach, in comparison to the high-fat meal.

It is unlikely that the larger buffering effect of the high-carbohydrate meal was the

result of higher secretion of saliva, since the principal stimulus of salivary secretion is

mastication,4 and horses did not spend more time eating the high-carbohydrate meal. In

fact, time of ingestion of the high-fat meal was significantly (p<0.05) longer.

Alternatively, it is possible that other factors besides mastication may modulate salivary

secretion in the horse, such as orosensory stimulation in response to different nutrients.

Although there is no evidence in the literature that dietary composition or gustatory