This item is only available as the following downloads:
1 FACTORS AFFECTING ACCEPTABILITY AND IDENTIFICATION OF PURED FOODS By JAMILA REN LEPORE A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTE R OF SCIENCE UNIVERSITY OF FLORIDA 2013
2 2013 Jamila Ren Lepore
3 To my husband and parents
4 ACKNOWLEDGMENTS I would like to thank my parents and husband for all of their love, support, and enco uragement. I thank my supervisory committee: Dr. Wendy Dahl, Dr. Charles Sims, Dr. Rene Goodrich Schneider, and Dr. Kathy Light, whose knowledge and guidance has helped me succeed. I especially thank both of my major advisors, Dr. Dahl and Dr. Sims for p roviding me with the chance to attain my graduate degree. Thank you Dr. Dahl for entrusting in me with your project idea and for all of the expertise and advice you gave along the way. Thank you Dr. Sims for providing me with a workspace in your lab and f or teaching and assisting me with my statistical analyses. I also want to thank Nancy Gal for her help in coordi nating the older adult panels. This study could not have been possible without you I thank my food science lab mates: Elizabeth Gardner, Britta ny Martin, Betty Nastasha Coello, Rachel Glintz, Kelly Brown, Eric Dreyer, and Dr. Asli Odabasi for all of your help and support. I particularly thank Eric Dreyer for all of your advice and assistance in preparation of panels and in ensuring both on and of f campus panels ran smoothly. I also thank my nutrition lab mates: Amanda Ford, Abdullah Hanifi, Younis Salmean, Allyson Radford, and Lauren Khouri for helping me prepare samples and conduct sensory tests with the older adults. Thank you Amanda for assisti ng me with recruitment and coordination of the Ocala panels. I want to thank my volunteers and the taste panel workers for all of their hard work Thank you for labeling cups, preparing samples, and being kind and informative to all of the panelists. I th ank my panelists f or continued participation even when samples were not well liked. Thank you for your feedback and expressing interest in the study. I enjoyed working with each one of you.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF T ABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 11 2 LITERATURE REVIEW ................................ ................................ .......................... 15 Food Preferences and Acceptability ................................ ................................ ....... 15 Preferences and Acceptability: General ................................ ........................... 15 Acceptability in Texture Modified Foods ................................ ........................... 18 Food Identification ................................ ................................ ................................ .. 19 Identification: Introductory Background ................................ ............................ 20 Identification: Color Modification ................................ ................................ ....... 20 Identification: Tex ture Modification ................................ ................................ ... 21 Methods to Improve Identification and Acceptability ................................ ............... 24 Sensory Enhancement: Flavor ................................ ................................ ......... 25 Sensory Enhancement: Appearance and Texture ................................ ............ 26 Contradictions f or Sensory Enhancement ................................ ........................ 27 Acceptability w ith Age Related Sensory Decline ................................ .............. 28 Sensory Testing ................................ ................................ ................................ ...... 29 Summary ................................ ................................ ................................ ................ 31 3 METHODS AND MATERIALS ................................ ................................ ................ 33 Panelist Recruitment ................................ ................................ ............................... 33 Pureed Food Selection ................................ ................................ ........................... 33 Sensory Testing ................................ ................................ ................................ ...... 34 Questionnaire and Training ................................ ................................ ..................... 35 Study 1: Shaped Pures ................................ ................................ ......................... 36 Study 2: Combination v ersus Individual Pures ................................ ...................... 36 Study 3: Regular Sodium v ersus Reduced Sodium Bread Pures ......................... 38 Study 4: Conventional Pures ................................ ................................ ................ 39 Statistical Analysis ................................ ................................ ................................ .. 40 4 RESULTS ................................ ................................ ................................ ............... 44 Study 1: Shaped Pures ................................ ................................ ......................... 44
6 Study 2: Combination v ersus Individual Pures ................................ ...................... 44 Study 3: Regular Sodium v e rsus Low Sodium Bread Pures ................................ 45 Study 4: Conventional Pures ................................ ................................ ................ 45 Acceptability Ratings and Identification Accuracy ................................ ................... 46 5 DISCUSSION ................................ ................................ ................................ ......... 72 Study 1: Shaped Pures ................................ ................................ ......................... 73 Sample Ratings: General ................................ ................................ ................. 73 Sample Ratings and I dentification Accuracy: Young vs Older Panelists ......... 75 Study 2: Combination v ersus Individual Pures ................................ ...................... 76 Sample Ratings: General ................................ ................................ ................. 76 Sample Ratings and Id entification Accuracy: Young vs. Older Panelists ......... 78 Study 3: Regular Sodium v ersus Low Sodium Bread Pures ................................ 79 Sample Ratings: General ................................ ................................ ................. 79 Sample Ratings and Id entification Accuracy: Young vs. Older Panelists ......... 81 Study 4: Conventional Pures ................................ ................................ ................ 82 Sample Ratings and Identification Accuracy ................................ .................... 82 Acceptability Ratings and Identification Accuracy ................................ ................... 84 6 CONCLUSION ................................ ................................ ................................ ........ 85 APPENDIX A HgLMS AND DEMOGRAPHICS BALLOT ................................ .............................. 89 B SAMPLE COMPUSENSE TEST BALLOT ................................ .............................. 94 C SAMPLE PAPER TEST BALLOT ................................ ................................ ........... 97 D HgLMS TRAINING SHEET ................................ ................................ ................... 100 REFERENCE LIST ................................ ................................ ................................ ...... 102 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 106
7 LIST OF TABLES Table page 3 1 List of i dentification c ategories for s tudy 1: s haped p ures ................................ 42 3 2 List o f i dentification c ategories for s tudy 2: c ombination vs. i ndividual p ures ... 42 3 3 List of i dentification c ategories for s tudy 3: r egu lar s odium vs.r educed s odium b read p ures ................................ ................................ ................................ ...... 42 3 4 List of i dentification c ategories for s tudy 4: c onventional p ures ........................ 43 4 1 Sample r atings by y oung a dults for s pecific a ttributes of s hape d vs. u nshaped p ures ................................ ................................ ................................ 47 4 2 S ample r atings by o lder a dults for s pecific a ttributes of s haped vs. u nshaped p ures ................................ ................................ ................................ ................ 48 4 3 Overall a cceptability versus i dentification a ccuracy of y oung a dults ................... 7 0 4 4 Overall a cceptability versus i d entification a ccuracy of o lder a dults .................... 71
8 LIST OF FIGURES Figure page 4 1 Effect of a ge on m ean s ample r atings of s ha ped and u nshaped p ures ............ 49 4 2 Identification a ccuracy of c hicken p ur es. ................................ .......................... 53 4 3 I dentification a ccuracy of p ork p ur es. ................................ ............................... 54 4 4 Identification a ccurac y of b roccoli p ur es. ................................ .......................... 55 4 5 Identif ication a ccuracy of g reen b ean p ur es ................................ ..................... 56 4 6 Sample r atings for s pecific a ttribu tes of c ombination p ures for e ach a ge g roup. ................................ ................................ ................................ ................. 57 4 7 Sample r atings for s pecific a ttributes of i ndividual c hicken, c arrot, and p otato p ures. ................................ ................................ ................................ ............... 58 4 8 Sample r atings for s pecific a ttributes of i ndividual b eef, c orn, and p otato p ures ................................ ................................ ................................ ............... 59 4 9 Effect of a ge on m ean s ample r atings of c ombination versus i ndividual p ure c hicken m eal ................................ ................................ ................................ ....... 60 4 10 Effect of a ge on m ean s ample r atings of c ombinatio n versus i ndividual p ure b eef m eal ................................ ................................ ................................ ............ 61 4 11 Mean s ample r atings of c ombination versus i ndvidual p ure m eals .................. 62 4 12 Identification a ccuracy of c ombination and i ndividual c hicken, c arrot, p otato m eal p ures ................................ ................................ ................................ ........ 63 4 13 Identification a ccurac y of c ombination and i ndividual b eef, c orn, p otato m eal p ures ................................ ................................ ................................ ................ 64 4 14 Sample r atings for s pecific a ttributes of r egular s odium and r educed s odium b read p ur es ................................ ................................ ................................ ...... 65 4 15 Effect of a ge on m ean s ample r atings of r egular s odium and r educed s odium b read p ur es ................................ ................................ ................................ ...... 66 4 16 Identification a ccuracy of r egular s odium and r educed s od ium b read p ures .... 67 4 17 S ample r atin gs for s peci fic a ttributes of c onventional p ur es ............................. 68 4 18 Identification a ccuracy of c on ven tional p ures ................................ ................... 69
9 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science FACTORS AFFECTING ACCEPTABILITY AND IDENTIFICATION OF PURED FOODS By Jamila Ren Lepore August 2013 Chair: Wendy J. Dahl Cochair: Charles A. Sims Major: Food Science and Human Nutrition Pured foods are recommended for persons with dysphagia, mostly older adults. Age related changes in chemosensory perception combined with texture modification makes recognition of pured foods challenging. Few sensory studies have been done in this field, yet with advancements in both food processing and product development, sensory evaluation of commercial pures is imperative for continued improvement. Inability to identify pures is a common complaint for these individuals and may negatively impact consumption and quality of life. The objective of this study was to evaluate the influence of age on identification and acceptability of pures. Four sensory panels were conducted with young panelist s (18 35yr; n=90 100) and older panelist s (>60 yr; n=50 72) using the hedonic gLMS. Various pure qualities were evaluated, including shaping, sodium c ontent, and presentation. Our results are in agreement with previous studies on identification, that young panelist s were better at correctly identifying pures. Unformed pures were more acceptable than shaped pures, which may provide a cost advantage. Identification and acceptability improved when pures were presented separately rather than combined. Reduced sodium pures were
10 equally acceptable to the ir higher sodium counterparts. Conventional pures received greater acceptability scores when compare d to non conventional pures. These results have several implications in product development of commercial pures. Pures should be created as individual foods with optimal sensory qualities. They do not require shaping, and bread purees may be prepared wi th less sodium. Producing foods that are commonly consumed as pure s could be more nutritious and acceptable. These findings highlight the need for pured food product development.
11 CHAPTER 1 INTRODUCTION Dysphagia is a condition characterized by impaired or difficulty swallowing that occurs either at the oropharyngeal or esophageal phase. It is most commonly the result of cerebrovascular injury (stroke, head trauma) and neurodegenerative disease such as n highly associated with radiotherapy in the treatment of head and neck cancers. Management of dysphagia is generally a multidisciplinary approach including nutrition management and therapy to improve swallowing function. In some instances, surgery and/or other medical interventions are indicated (Lister 2006) The prevalence of dysphagia is highest in the elderly population, with approximately two thirds of dysphagia patients being age 65 or older (Ginocchio and others 2002) According to the U.S. Census Bureau the percentage of individ uals age 65 and older will rise from approximately 12.4% in 2000 to just over 20% in 2050 (U.S. Census Bureau 2004 ) With an ageing population, emphasis on dysphagia management becomes increasingly relevant, as dysphagia incidence is not only highest for elderly persons but is also prevalent within this population. Multiple studies demons trate that for persons over age 50, dysphagia occurrence is approximately 15 35% (Chen and others 2009; Roy and others 2007; Lindgren and Janzon 1991) The large variation indicates differences in sample demographic s, methods to obtain data, and possible disparity in definitions of dysphagia Nonetheless, these data suggest a noteworthy group of affected individuals. While abundant research exists in the field of dysphagia, most is focused on the clinical facets of t he condition such as diagnosis and medical intervention. Few papers
12 have been published in sensory evalu ation of the texture modified diet, the approach for nutritional managem ent of this condition, and its impact on quality of life (QOL). In many cases, the modified texture is pured, where all foods must be blended to a smooth consistent texture before consumed. Although limited, existing research clearly shows pured food s are poorly accepted and adversely impact QOL (Blaise 2009; Schiffman 1977; Schiffman and Warwick 1989; Schiffman and Warwick 1993) Given this is a concern for individuals with dysphagia currently, imagine the magnitude of the issue if research in this field remains stagnant while the number of dysphagia cases considerably increases as the populace continues to age. Specifically, this research will target identification and acceptability of pure d foods Few papers have been published on pure d food recognition; in fact, the most recent paper is from the mid 1980s (Schiffman 1977; Murphy 1985) Given the inability to recognize a pure d food has been associated with a negative mealtime experience for elders, this is an area which needs to be investigated further (Blaise 2009) Current studies on pured food recognition com pare young versus older subjects and highlight age as a factor influencing correct identification. However, acceptability and preference of these pure d foods have been neglected. These studies also remove d the visual element by no t allowing subjects to vi sually observe the pures. Not only is this an unnatural way of eating, but elders have been shown to rely more on visual cues to recognize foods (Philipsen and others 1995) There is also contradictory evidence surrounding the use of shaped versus unshaped pured food s an area which would also benefit if researched further.
13 The current research study will fill these gaps by not only examining identification of pured food s but asking more importantl y i s the pure acceptable? Is the additional pured food s better served individually or combined? How do low sodium products compare to regular? Should we focus more on conventional pured food s to increase acceptance and possibly consumption? Answers to these questions will be explored through sensory evaluation using comprehensive sensory panels. This research project is made up of four smaller individual studies. There are two primary objectives with several secondary aims specific to each of the four different studies. The first primary objective is to evaluate the effect of age on identification of pured foods. It is expe cted that younger persons will identify more correct items than older persons, especially when the visual element is absent. Second, this project examines the possibility of a relationship between identification and acceptability. It is expected that corre ct recognition of a pured food will be positively correlated with increased acceptability. The four studies within the project evaluate various characteristics of pured food s, including appearance, preparation styles, sodium content, and food type. Each of these will be explained in further detail in subsequent chapters. As with any food product, there is always room for improvement, especially if the product is consumed at every meal, as are pures for those on a texture modified diet. The results of t his research have the potential to substantially affect multiple fields, including but not limited to, health and wellbeing of dysphagia patients, product development, and nutr ition management of dysphagia. A better understanding of what is both accepted a nd preferred may lead to development of improved products, both
14 com mercial and in house prepared. Ultimately, the outcomes from this research will ideally enhance the meal time experience of individuals with dysphagia on a pured diet and improve their QOL
15 CHAPTER 2 LITERATURE REVIEW Food Preferences and Acceptability While there are many factors that affect what an individual eats including economic, cultural, familiarity, nutrition and convenience one of the most influential factors is personal food p reference or the selection of one food over another (Laureati and others 2006) While preferences may be shaped by these factors and o thers, it is the sensory attributes of a food product that are of utmost importance. However, why one product is selected over another may not always correlate with the sensory preference of that food item (i.e. the one that tastes, smells, and appears the best). consistently documented in the literature to have a significant impact on food preference. Another major factor influencing what an individual eats is acceptabili ty, or the level of satisfaction with a given food. Acceptance of a food product is closely linked with food preference in terms of sensory analyses and provides a more complete picture as to why certain foods are consumed. For instance, when comparing var ieties of apples, a panelist may prefer one apple over the other but provide low scores in acceptability because that individual does not like apples. Clearly, preference and acceptability are two distinct features. Both are required to understand fully th e rationale for food selection. Preferences and Acceptability: General In surveying focus groups of institutionalize d elderly, Laureati and others found that elderly ascribed sensory qualities as the most important consideration in food
16 preference and a cceptability (Laureati and others 2006) Specifically, in respective order of preference, elderly chose taste, aroma, appearance, a nd texture (Laureati and others 2006) Alternatively, Hall and Wendin (2008) found through focus groups that discernible food components consistency/ease of swallowing, appearance, powerful flavor, and nutritional quality were the most important factors in food acceptability (Hall and Wendin 2008) dysphagia who provided their opinion on what elderly would dee m important for food preference and acceptability. While the elderly group chose factors they deem relevant in food choices, it is possible that some of the criteria listed from the expert focus group (i.e. ease of swallowing, discernible food components) are inherent in the actual food choices and were simply not recognized by the elderly group. As taste and aroma seem to be recurring factors of importance, it is important to note that decline in gustatory and olfactory senses with age is well documented (Drewnows ki 1997; Fillion and Kilcast 2001) It would seem logical then, that with limited olfaction/gustation abilities, reliance on other sensory attributes such as appearance and texture as a compensation mechanism would assume a greater role in food prefe rence and acceptability. Forde and Delahunty (2004) found that when compared to connection between texture and acceptability (Forde and Delahunty 2004) As similar juices were liked by both the older an d younger group, the researchers stated that importance texture has on acceptance regardless of full or limited olfaction/gustation.
17 Nonetheless, it is also important to not e the older group was not able to distinguish between samples in r eference to their liking score (Forde and Dela hunty 2004) Such a finding implies the older subjects were reliant on fewer sensory attributes and therefore had a more limited perspective on all of the sensory qualities needed to discern samples. A similar finding was found in one of their earlier s tudies, where older subjects were less able to discriminate texture variation between samples than younger subjects. related to preference (Forde and Delahunty 2002) However, only one respective food from several textures (li quid, semi solid, and solid) was presented, and these results may be different for foods in which we would expect texture to play a primary role (i.e. meat products, dry/crunchy foods, etc.) in food acceptability. Based on these contradictory findings with regards to texture and food preference, more research n eeds to be done in this field. Regardless of the impact texture has on acceptability for those with reduced olfaction and gustation, aroma and taste qualities are still important to the eating experie nce. Aroma includes odorous substances while taste is made up of sweet, salty, sour, and bitter (Taylor and Roozen 1996) Taken together, these attributes make u p what is perceived as flavor. Flavor perception is therefore dependent on these variables and may be enhanced or diminished when aroma or taste is altered (Taylor and Roozen 1996) This may have implications for older adults who have decreased sense of aroma and taste and therefore of flavor perception (Schiffman 1993)
18 One of the most important nutrients, both in terms of flavor pe rception and health, is so dium. Reduced sodium diets are frequently recommended for older adults for various factors including increased age, medications, and medical conditions such as hypertension, CHF, renal disease, among others Nonetheless, salt is o ften crucial to optimize flavor perception. S alt is often added to food in order to increase volatile headspace which in turn enhances flavor perception (Guichard 2002) The threshold for salt (as NaCl) has been show n to be significantly greater for elder adults when compared with young (Mojet and others 2001) In general, the threshold for sodium salts in older persons i s nearly 12 times higher on average when compared to young persons (Schiffman 1993) Consequently, older adults require this increased sodium concentration for optimal flavor perception yet are often simultaneously on reduced sodium diets, which may have serious consequences. Older adults m ay end up consuming excess salt to compensate for what is perceived as reduced flavor. It is also possible that consumption will decrease as a result of foods seeming bland. Either way, this is a nutrient of concern, espe cially in the older population. Acc eptability in Texture Modified Foods As acceptability of foods are primarily evaluated on the basis of sensory parameters for the general population, acceptability of modified texture foods extends beyond sensory attributes for individuals with dysphagia. These foods often constitute the individuals diet in its entirety and as such, play a significant role in daily life. It has been shown that for those on texture modified diets, regardless of extent of modification (i.e. chopped, minced, pure d), there is a decreased quality of life (McHorney and others 2002) Perhaps this is due to personal embarrassment for needing such a
19 specified diet, the inconvenience associated with food preparation and/or procurement, or the disjo int between expected and actual texture of familiar foods. Therefore, increasing acceptability of specialty foods for individuals with dysphagia could be one w ay to improve quality of life. It has been well documented that noncompliance with a modified die t is a recurrent issue for patients with dysphagia (Colodny 2005) Further, in a survey of noncompliant individuals, it was found that second to outright denial of a swallowing problem, the most common claim for noncompliance was dissatisfaction of the food product further corroborating the need to improve the food products themselves as a step towards incre asing perceived quality o f life (Colodny 2005) Food Identification Most people do not consider food identification as part of the eating experience, as it is generally not a mystery what food item is being consumed and is therefore inherent in sensory assessment for preference and acceptability. However, in the absence or alteration of one attribute in an otherwise familiar product, such as with texture mod ified foods, identification becomes increasingly difficult. This is because the senses work in a complex system in recognition of a food product, which can ultimately potentially change preferences/acceptabili ty (Murphy 1985) Whether one conscientiously realizes it or not, memory of a food product is related to all of its sensory attributes and it is this combination that allows one to not only identify a product but assess its acceptability (Scott 2005) Therefore, acceptability may be altered when a product is not recognizable.
20 Identification: Introductory Ba ckground There are many tests to evaluate sensory cognitive interaction and accuracy. One such test utilizes the principle of oral shape recognition, where identification of a shape placed on the tongue is done without the use of tactitio n or visual influe nce. Fillion and Kilcast (2001) utilized this form of testing with alphabet letters made of sugar icing, where participants were asked to identify the letter using only oral manipulation. They found that adults age 65 and over were significantly less able to identify correct letters when compared to young adults (Fillion and Kilcast 2001) While this study examined sensory integrity based on non food items, several studies have looked at identific ation specifically with foods. Although taste and aroma may be perceived as the most that appearance and texture are crucial for correct identification of foods, which in turn can impact acceptability (Imram 1999) As a result, the influence of appearance and texture modifications on identification is examined here. Identification: Color M odification Color has been shown to have significant effects on correct identification of fruit flavored beverages (Dubose and others 1980) Correct scores increased as color went from inappropriate (i.e. green color, orange flavored beverage) to appropriate (i.e. orange color, orange flavored beverage). Moreo ver, in assessing the influence of color on flavor, intensity of the color (of the appropriate color) was correlated with increased overall acceptabili ty (Dubose and others 1980) Philipsen and others (1995) found similar results, though they also included the variable of age by using subj ects of a young and old cohort. Identification of a cherry flavored beverage declined as several samples with color of varying intensities went from red to yellow. Moreover, perceived
21 flavor intensity and overall acceptability increased as color progressiv ely became more representative of the actual flavor. This was true for both age groups, but to a much (Philipsen and others 1995) Thus, color affects not only identificat ion but acceptability as well. A product that is presented with its expected color (as obtained from previous experience with the food item) will be more likely identified and more acceptable. Whether or not identification and acceptability are merely correlated or dependent on one another is an area that needs further research. Identification: Texture M odification While appearance has been shown to play a significant role in recognition of foods, texture modifications have been extensively researched, especially in regards to those with dysphagia, of which this review is aimed. In a qualitative study of observations, patient chart reviews, and patient/staff interviews, it was found evident that difficulty in recognizing foods in pure d form was negatively associated with the meal time experience (Blaise 2009) having too similar taste and texture, and one signifying a disliking for the product. While there are likely additional variables contributing to the negative mealtime experience for individuals on pured food diets, inability to identify the food item was cited as the most important factor from the (Blaise 2009) It appears then, that identification of a food item is inherently ass ociated with its acceptability. Schiffman (1977) found significant differences between young and older subjects in correct identification. Subjects were blind folded, asked to smell and taste the product,
22 report its identity, and rate the extent of intensity of the aroma/flavor stimulation. Not only did young subjects report the actual food item correctly more often than older subjects, they were also more likel y to guess the correct food group when a misidentification was made. Further, older subjects reported significantly less intensity of both taste and aroma when compared to the young, though to a greater degree for aroma (Schiffman 1977) This supports the idea that gustation/olfaction decline with age and ultimately influence sensory perception. This may be especially true with modified texture foods, wher e reliance on taste and aroma p redominate in food recognition. food items and followed the same procedure with one exception. Rather than conducting one session of identificati on, Murphy used several sessions, where the first subjects regarding correctness in thei r response. The purpose was to assess cognitive ability in identification, as reduced ability of older subjects to correctly identify foods could be due to a decline in cognitive function associated with age rather than sensory losses (Murphy 1985) Essentially, poor identification could be related to poor memory retrieval of a food item. Further, while si gnificant differences in correct identification occurred between age cohorts when subjects were allowed to both taste and olfaction this was not the case when the factor of smell was taken away (Murphy 1985) Thus, these two studies agree that olfaction likely has a great er impact on identification than does taste.
23 Another key finding from was that upon subsequent sessions with feedback on correctness and repeated exposure to the same food items led to increased identification for both young and older subjec ts. However, young subjects improved significantly more than older subjects in subsequent sessions, suggesting a lack of cognitive ability in the older adults as an explanation f or the discrepancy (Murphy 1985) ebral recognition is innately part of identification, as it relies not only on sensory attributes but also on cognitive function Another study o n identification of texture modified foods found comparable sessing naturally prepared pure d food items (i.e. without added cornstarch or spices/seasonings), Bischmann and Witte (1996) found there was a significant difference in correct overall identifications for young adults when compared with older adults. Furt her, younger adults performed better than older adults in identifying nine of the ten items correctly. It is possible the older group identified one item better than the other due to familiarity of the product as an item served in the retirement home from which older subjects were recruited (Bischmann and Witte 1996) It is important to note that it is quite possible that confounding variables such as decreased dentition, declining health/cognitive status, a nd increased use of medications that are more prevalent in the older population could influence sensory perception and therefore ability to correctly identify food items (Mioche 2004) It is therefore critical to take these variables into account, as was so adeptly realized and exec uted in previous research (Bischmann and Witte 1996) Moreover, other variables
24 that could influence sensory perception but that are not necessarily associated with age (i.e. smoking) were also considered (Bischmann and Witte 1996) With a substantial group of possible confounding variables being eliminated via sta tistical analyses, the validity in their results is greatly augmented. While there are a limited number of studies examining food identification, it seems apparent this is a skill that declines with age, at least when one or more sensory characteristics is removed as a variable. While it has not been studied, one would assume that a young and older person comparing two food groups that contain all expected sensory attributes (taste, aroma, appearance, texture) would be able to equally identify the product. Nonetheless, the sensory attributes the older person may rely on more in identification may differ from the younger group who generally have a different sensory profile. In effect, an older person may use more visual and textural clues which have not been as affected by age in identifying a food product while a young person is more likely to use each of the senses more equally (Imram 1999; Forde and Delahunty 2004) Methods to Improve Identification and Acceptability As the dominant trend appears to be th at elder persons have a decreased ability to correctly identify foods and a decline in olfaction/gustation, several attempts have been made to improve or enhance foods in order to augment identification and acceptability of food products for the older popu lation, especi ally for those with dysphagia. If there truly is a compensation effect and the decline of one sensory attribute leads to reliance on other features, we would expect older subjects to rely primarily on appearance and textural cues. However, in pure d foods, texture and appearance are both modified and reliance on taste and aroma then becomes crucial in
25 yielding any form of familiarity f or recognition and acceptance. Thus, with limited olfactory and gustatory systems, elderly individuals on a pu re d diet may benefit from the enhancement or manipulation of taste and aroma. Moreover, as visual cues have been shown to improve elderly identification of food products, perhaps manipulation of appearance in a typical pured food may improve acceptabilit y as well. Sensory E nhancement: Flavor Flavor enhancement to improve acceptability has been widely researched. A flavor enhancer is any substance that increases the intensity of the flavor already present, for example, monosodium glutamate (MSG) (Schiffman and Warwick 1989) In comparing young versus elder subjects, one study found it took a greater percentage of flavor concentration to perceive the same level of intensity f o r the elder group. T he elder subjects preferred the higher intensity flavor, though to a varied degree dependent on the food product (deGraaf and others 1 996) These findings suggest that elderly would prefer flavor enhanced foods, though the exact flavor intensity would depend on the food item in question. Additionally, Laureati and others (2008) found elderly preferred foods that were fortified with flav or enhancers when compared with unfortified products and younger subjects (Laureati and o thers 2008) In another study examining the effects of flavor enhancement, several food items were prepared, each with a high and low concentration (and thus intensity) of flavor, and then rated and selected based on preference. The results show that the older cohort preferred the high flavor concentration for samples while the young preferred the low concentration (Griep and others 1997) While these findings support the idea that enhanced flavor is preferable among older individuals, the researchers also make the claim that flavor enhancement may (Griep and others 1997)
26 This claim should not be extrapolated from these results, as preference was only between samples and not for overall food preferences related to dietary intake. F lavor amplification has been shown to be preferable and lead to increased consumption by elderly (Schiffman and Warwick 1989) This could be explained by a compensation effect for decreased olfaction and gust ation (Schiffman and Warwick 1993) It appears from these results that flavor enhancement could indeed improve acceptability of food products designed for an older population. However, it is also should be noted that none of these products were pure d foods and it is possible that flavor enhancement woul d not produce the same effect. Therefore, research on flavor enhancement of pured food s would be valuable. Sensory E nhancement: Appearance and Texture In an effort to improve acceptability of texture modified foods, the use of shaping molds to give the food its characteristic shape have been employed. Shaped pure s are produced to emulate the same appearance qualities as the non pure d food counterpart. There are mixed re sults regarding the effect these molds have on acceptability. One study, which examined residents in a long term care facility, provided on two separate occasions both typically prepared and molded for ms of pure d foods. The foods were from the same menu cycle and given to the same residents to maintain pure d food However, they also note that staff may have indirectl perception and consumption by making positive comments on molded forms (Cassens and others 1996) Conversely, prel iminary data from Stahlman and others (2000) indicate sensory qualities (appearance, taste, texture) of molded fruits have significant ly lower ratings than unmolded (Stahlman and others 2000)
27 Contradictions f or S ensory E nhancement Contrary to the many studies described above s everal studies have found enhancement of food products to be negatively associated with acceptability. This could be explained by the idea that appreciation may not b e affected by sensory decline. As the decline in sensory acuity is a gradual process, elde rly are often unaware and simply adapt over time (Wysocki and Pelchat 1993) Thus, despite actual differences in perception of intensity of an aroma/flavor, an elderly person may have adjusted to the lower intensity and therefore appre ciate it better. For instance, enhancement of aroma in food products did not have a significant effect on overall pleasantness for the elderly group in a study on flavor and aroma enhancers (Koskinen and others 2003) Initial responses indicated unpleasantness with increased aroma; however, in a second trial the same sample was associated with increased pleasantness. The researche rs suggest this could be a result of adjustment to the aroma, given the fact elderly did not discern between samples to the same extent as the young group. The net result was that enhanced aroma did not increase acceptability as proposed in other studies (Koskinen and others 2003) Other research has found that despite documented decreases in taste sensitivity, there were no signi ficant differences in preference of flavor heightened food products of elderly versus young; rejecting the idea that enhanced products would be favored by the elderly with decreased sensory capabilities (Mojet and others 2005) Contradictions to enhancement of texture and appearance have been documented as well. Since molded fruits received negative feedback for taste and textur e in the preliminary study by Stahlman and others (2000), additional research was conducted to assess the thickener (added to molded forms to provide stability in
28 m aintaining shape) as a possible influence (Stahlman and others 2000; Stahlman and others 2001) In the subsequent study, a comparison between typical pure d food pure d food with added thickener, and molded pure d food were assessed to evaluate influence of appearance, taste, texture, and overall acceptability. The typical pure d food without thickener was rated highest in all of these categories, contradictory to other results (Stahlman and others 2001; Cassens and others 1996) A possible explanation for increased preference of the non thickened typical pured food is that it did not contain the thickening agent, which likely negatively impacted taste and viscosity of the pr oduct Additionally, typical pured food s (with and without thickening agent) were rated higher in appearance than the molds, which could conceivably be due to higher expectations imparted on the shap ed versus the un shap ed pu red food In essence, if one visually sees a pured food there is likely an expectation of a pure However, if one sees what appears to be a peach yet the taste and texture are different from a true peach, there may be more negative associations with th at molded peach, as it did not live up to its exp ectation (Stahlman and others 2001) Acceptability w ith Age Related Sensory Decline It appears then, the idea behind enhanced products whether for aroma, taste, or texture is a more complex dilemma and is inconclusive at t his time. Indeed, Laureati and others (2008) found that while the young subjects had high sensory capabilities in taste and odor identification, the elderly subjects varied in their sensory integrity (Laureati and others 2008) As a result, the older group was subdivided into low and high scoring groups. Nonetheless, the young significantly scored better than even the hig h scoring older subjects. The researchers propose that despite substantial sensory
29 decline of older persons when compared to young, there is extensive variability between this group of individuals and thus, elderly persons should not be considered as a sin gle that there is great individual variation in the extent of olfaction/gustation reduction with age (Forde and Delahunty 2004) This could perhaps offer one explanation as to why flavor enhanced products have been found preferable and non preferable in multiple studies; namely, that a group with higher sensory capability may prefer a more traditional product while those with lower sensory functioning may compensate by favoring enhanced products. It is also key to point out that several studies have found participants across age groups preferred si milar formulations of a food sample, despite reduced taste and aroma perception of elderly subjects (Booth and others 1989; Forde and Delahunty 2004) This finding may imply preference is not associated with a decli ne in sensory perception. Thus, preference may rely more on components of flavor/aroma rather than intensity, which would be a counterargument towards enhanced foods. The researchers also suggest that as a result of this finding, flavor amplification of fo ods may actually negatively impact acceptability for two reasons First, flavor enhancement may change the composition of flavor constituents and therefore lead to a product that is less preferred, and second, for any individual who has a lesser degree of sensory loss due to age, amplification of flavor in a food product may yield a flavor that is too intense and undesirable (Booth and others 1989) Sensory Testing The mo st traditional and widely used scale in sensory preference and acceptance testing is the 9 point hedonic scale. This relatively simple to use categorical
30 advantage of this scale is due to its simplicity, both for the researcher and participant, as well as consistently proving to be effective (Lim 2011) Nonetheless, there are several limitations to using the 9 point hedonic scale in preference and acceptability studies, es pecially across different groups of individuals. The largest limitation of a categorical scale such as the 9 point scale is that intensity responses are relative and thus may vary from per son to person (Bartoshuk and others 2005) With the discovery of differences in taste perception (including non tasters and supertasters), researchers tested and realized that perhaps categorical scales were not capable of captur ing relative differences in individual sensory experiences, including with eating. As a result, new scales were created to solve this dilemma, including the hedonic general l abeled magnitude scale (H gLMS). The most recent version of this scale provides a b ottom anchor ( the very center (Bartoshuk and others 2002) Person s using the scale select those experiences which scale w ill vary from person to person. As individuals, we all have very different sensory and life experiences, and a s such, the intensity to which a food experience is perceived will likely vary as well. For
31 food product would rank in ver y different places on the scales for these two individuals. Nonetheless, expert opinion suggests individuals rank favorite food sim ilarly, with an average rating of approximately 65 80 (Sims 2013) The scale is dependent on the hedonic experiences selected by the user, which are generally not food related (though they may be). As such, the scale allows for cross compari son between different groups of people who have differ ent experiences (Bartoshuk and others 2004) Food is just one area of hedonic acceptance and preference, and the relative intensity to which a food is ranked will ultima tely vary from person to person, depending on personal experiences. In using the H gLMS, it is expected to pick up on these variations. Summary Decline in chemosensory perception has been well documented but the relation of this decline to food acceptability and preference is still lacking (Drewnowski 1997) Further, decline in one sensory attribute affects the entire sensory perception of a food product, as sensory attributes are not single units functioning alone but rather work in unison as a complete system (Murphy 1993) Most frequently documented is the d ecline in olfaction/gustation sensory attributes. Mioche (2004) describes there is less of a decline in textural perception with age, despite a decline in chewing efficie ncy due to decreased dentition and saliva production, that are associated with aging. Further, while oral manipulation and mechanism of chewing differed for elderly and younger subjects, there were no significant differences in the end result of texture perception (Mioche 2004) Nonetheless, for elderly individuals with dysphagia, not only are olfactio n/gustation reduced due to age, but texture is modified in the products they consume as well. Consequently, we would expect this accumulative effect to
32 considerably reduce the overall perception of a food system. How this then affects food acce ptability is the next concern. Moreover, while identification studies have been extensive with regards to comparison between young and old and examination of the variables affecting ability to recognize a food system, it still remains to be evaluated whether or not id entification influences acceptability. Given lack of identification was cited as one of the primary reasons for a negative m ealtime experience it seems logical to look into acceptability as it relates to identification, spec ifically in pure food systems (Blaise 2009)
33 CHAPTER 3 METHODS AND MATERIAL S Panelist Recruitment The University of Florida Institutional Review Board (IRB) classified the study protocol as exempt and thus no further documentation was required. Two age groups were recruited, including a young cohort (age 18 35) and elder cohort (age 60 and older). Panelist screening was done to exclude individuals with specific food allergies and/or who did not consume beef, pork, or poultry products, as these would prevent panelists from tasting some of the samples. Young panelist s included Univ ersity of Florida students and staff who were recruited through a panelist listserv of past sensory testing participants. Elder panelist s were recruited from t he community in Ocala, Florida Four objectives were evaluated, which are subdivided into S tudie s 1 through 4 The same group of panelists was used for each study, though not all participants were able to make it to every sensory panel. Panelists were not required to attend every session, though participation was encouraged. This was not the case for Study 2, which was made up of two sessions. Compensation was provided for each panel attended, regardless of inclusion into the data Pureed Food Selection Pured foods to be studied were selected on the basis of the four objectives evaluated in each st udy. A variety of foods were chosen to account for potential preference differences. Samples were commercially produced ready made or ready to prepare pured foods. All samples were handled or prepared using appropriate food safety procedures. Each specif ic food is described in more detail under the
34 corresponding study section. Pure s were served at room temperature for optimal flavor perception (Ventanas and others 2010) Sensory Testing Sensory testing was conducted at different locations for each age group. Young panelist s evaluated samples in the sensory laboratory at th e University of Florida The lab is made up of individual booths with computer stations equipped with CompuSense five Sensor y Analysis Software for Windows (Compusense, Guelph, Canada). Copies of individual paper scales and a pencil were provid ed to panel ists for reference. Elder panelist s met in Ocala, Florida at two separate locations, as a result of participant availability and transportation. One group evaluated samples in a large banquet hall at a local church while the second group tested in a major conference room at the University extension office. Individual booths were not available, so panelists were seated an appropriate distance apart at long tables facing the same direction. Panelist s were also instructed not to converse during testing. Paper scales and ballots were provided for t esting, in addition to pencils. For both groups, unsalted crackers and deionized water were provided as a palate cleanser. Panelists were instructed to have a bite of cracker and sip of water before each sample. A spo on and napkin was a lso provided to each panelist. All samples were assigned a random three digit code. With the exception of the shaped pure s in Study 1, scoops of varying sizes were used to distribute samples into 4 ounce clear plastic cups. Samples that were transported to Ocala were covered with a lid. Shaped pure s were presented on a small paper plate in order to keep the shape intact. Panelists were instructed to consume a sufficient amount of the sample in order to answer questions.
35 Questionnaire an d Training A t raining session was provided during the first panel, provided by brief training at each subsequent panel. If a new panelist participated that had not done the panel before, that panelist was given individual instructions on how to use the sc ale and proceed with testing. During the first training session, participants were briefed on the primary objectives of the study but without any indication as to the types of food products that would be tested, other than that samples would be pured food s. Next, the scale was introduced, which was the hedonic gLMS. Each panelist was instructed to create his/her own personal scale by selecting anchors that best describe the strongest disliking and strongest liking ever experienced. These events represent the bottom ( individual scale. A value of 0 indicates neither liking nor disliking something. Examples and warm up questions (included in the Appendix) were provided to ensure participants understood the scale prior to rating samples. These warm up questions were briefly examined before sensory testing began Panelist s were encouraged to refer back to the personalized scale when tasting samples. The questionnaire was the same for all studies, though was presented electronically for young panelist s and as a paper ballot for the older group. The only demographics collected were age and gender. The attributes of overall liking, appearance, mouthfeel, and flavor were evaluated for each sample using the hedoni c gLMS. The characteristics of these attributes were described to participants. After rating attributes, panelists were asked to identify the food in the sample. If the exact identity was unknown, panelist s were to identify the food group ( i.e. vegetable, meat grain).
36 Study 1: Shaped Pure s The y oung cohort w as made up of 97 panelists 55 female and 42 male, ranging from 18 33 years of age with a median age of 23. The e lder cohort included 70 panelists 59 female and 11 males with an age range of 60 89 an d median age of 73. The shaped pure s in this study are commercially produced by Hormel Health Labs (Austin, MN) and include: Thick & Easy Pure d Shaped Chicken, Roast Pork Broccoli and Green Beans A total of eight samples were presented to panelists d uring one session, with the shaped and unshaped version of each of the foods. Samples were taken out of the freezer and put into refrigeration 48 hours prior to the panel. The samples were removed from refrigeration 2 3 hours before the panel began in or der to come to room temperature (within 19 25C). Half of the shaped pure s were taken out of the molds and placed onto plates. The remaining half were placed into a sheet pan and blended until the mixture resembled a standard pured food The unshaped pur e s were placed into 4 ounce plastic cups with an ivory #10 scoop to achieve volume comparable to the shaped pure s. The samples in this study were selected to provide two very different food groups but each with foods of similar color. This was done to re duce the chance of guessing correctly as a result of remembering a previous sample. Study 2: Combination v ersus Individual Pure s This study w as made up of two sections, combination pure s and individual pure s. In section one, young panelists were compr ised of 103 persons, 46 male and 57 female, with an age range of 18 32 and median age of 23 years Elder persons included 67 panelists, 53 female and 13 male, ranging in age 60 89 with median age 72 years. In section two, the young group consisted of 77 pa nelists, 34 males and 43
37 females with age range 18 32 and median age 23 years In the elder group, there were 58 panelists, 48 female and 10 male ranging in age from 60 88 with a median age of 72 years The combination pured food s were two samples made u p of three foods, including (Austin, MN) Thick & Easy Pure Roasted Chicken with Potatoes and Carrots and Thick & Easy Home style Beef with Potatoes and Corn Pure s were distributed using a b lue #16 scoop into 4 ounce plastic cups. The individual pure s correspond to the three foods present in the combination pure s, except that each food was presented as an individual food across six samples. These samples include garden pure (Toronto, Canada) Creamy Corn and Sweet Carro ts (St. Louis, MO) Thick It Salisbury Steak Pure and Chicken la King Pure and Publix (Lakeland, FL) Instant Potatoes (prepared according to package directions with Publix 2% milk, Publix margarine, Publix iodized salt ). Pured food s were d istributed using a red #24 sc oop into 4 ounce plastic cups. A smaller scoop size was used for individual pure s due to a limited supply of product The combination pure s and individual protein pure s were shelf stable and thus did not require further pre paration. The individual corn and carrot pure s came frozen and were thawed under refrigeration 48 hours in advance to the panel and then brought to room temperature (within 19 25C) approximately three hours before the beginning of the panel. The individu al pure d potatoes were prepared the night before the panel and kept under refrigeration until the next morning and then brought to room temperature. Section one comprised combination pure s which were evaluated first. These pure s were presented prior to the individual pure s to reduce the chance of food recall.
38 Combination pure s were randomized within each test group (up to 10 for young persons and up to 20 for elder persons) within each session. The six individual pure s in section two were presented o n two trays with three samples each. The two trays consisted of the three foods that corresponded to the combination pure Namely, one tray was made up of chicken, carrots, and potatoes, corresponding to the combination pure composed of the same three fo ods. Placement on the tray was fixed for each panelist but the two trays were randomized within each session simil arly to the combination pure s. In section one, panelists were asked to identify three foods present in the combination pure In section two panelists were asked to identify the three foods individually. Panelist s were not given any indication that these sam ples contained the same foods. Study 3: Regular S odium versus Reduced S odium Bread Pures Panelists in the young cohort were made up of 97 persons, 42 male and 55 female ranging from 19 33 years of age with a median age of 23 years. In the elder group there were 70 panelists 55 female and 14 male with an age range of 60 88 and median age of 73 years. A c omparison of regular sodium and red uced sodium prod ucts was done with a bread pure prepared from Darlington Farms Pure Bread & Bakery Mix (Noblesville, IN) Publix canola oil, and deionized water. Two samples were prepared one with sodium content comparable to an average slice of bread a nd the other wi th 50% reduced sodium content. To make the reduced sodium sample, the bread pure was prepared according to package directions with only the addition of oil and water, as the bread mix alone had only 65 mg sodium and thus could be considere d a lower sodium bread product. In
39 order to have a 50% reduced sodium product, the higher sodium food require d 130 mg sodium, comparable to an average slice of bread. For the regular sodium bread pure iodized salt was added at the appropriate concentrati on to achieve 130 mg sodium content per serving. According to the package, one serving of the bread pure is served with a r ed #24 scoop to achieve approximately a 1 ounce serving, equivalent to a slice of bread. This same scoop was used to distribute sam ples into 4 ounce plastic cups. As with the other studies, panelists were asked to rate attributes of overall liking, appearance, mouthfeel, and flavor. Panelist s were also asked to identify the food or food group of the bread pure s. Study 4: Conventional Pure s Only elder panelist s participated in S tudy 4 primarily as a preliminary investigation into the use of conventional pured food s. The group consisted of 60 panelists 48 female and 12 male, ranging in age from 60 88 with a median age of 72. The sa mples included three bean foods as well as a dessert like item, selected both for nutritional content and acceptance by the general population. The foods used were Publix deli style classic hummus (Lakeland, FL) Old el Paso traditional refried beans (Vict oria, Australia) (Chestnut Hill, TN) and Market organic pumpkin pie mix (Corvallis, OR) All samples were already in pure form, with the exception of baked beans which were placed into a food processor and complete ly homogenized. Samples were distributed into 4 ounce plastic cups using a r ed #24 scoop. As with other studies, panelists were asked to rate attributes and i dentify the food or food group.
40 Statistical Analysis The primary objective of this study was to e valuate acceptability ratings for various attributes across samples and to determine identification accuracy between age groups. Raw data for acceptability ratings among young adults was entered electronically and collected using the software program Compu Sense Data w ere analyzed using the program SAS 9.2 (SAS Institute Inc. Cary NC, USA). For each study, a nalysis of variance (ANOVA) to determine significance at p < 0.05 for each attribute across all sampl es. In Study 1, ANOVA was conducted across the eight unshaped and shaped samples for each attribute. In Study 2, Part 1, ANOVA was carried out across the two combination samples for each attribute In Study 2, Part 2, ANOVA was conducted a cross each of th e six individual samples for each attribute. In order to compare across Part 1 and 2 for Study 2, means from the individual pures in Part 2 were averaged for each meal type (the chicken or beef meal), thus condensing six samples into two. These averages w ere then compared with the means of the combination chicken meal and combination beef meal samples via ANOVA for every attribute For Study 3, each attribute was ev aluated using ANOVA across the two bread pure samples. In Study 4, ANOVA was carried out a cross four samples for each attribute. Data were also sorted base d on the two age groups within S tudies 1 3, where no additional sorting was needed for S tudy 4. Identification accurac y was analyzed by percentages. All identifications were entered as a fr ee form response by panelists. For S tudies 1, 3, and 4, panelists were asked to identify the exact food and instructed to enter the food group or make an educated guess if the exact food could not be identified. During training, food group
41 descriptions were briefly explained. For S tudy 2, identifications were classified based on the number of foods correctly identified. Since there were three foods present in both session 1 and 2, identifications were categorized as 0, 1, 2, or 3 correct foods. Because classi fying the degree of correctness in a response is rather subjective, the guidelines that were used are included below in Tables 3 1 through 3 4 Each table represents the standard used to categorize panelist responses. Once response s were classified into e ach identification category, these scores and overall acceptability ratings for each sample were compared by use of Spearman correlation coefficients. Overall likability was analyzed using the VAS ranging from 100 to 100 while identification of the pureed food fell into one of several possible categories depending on the study. For studies 1 and 4, identifications were categorized as 1= correct food, 2= correct food group, and 3= incorrect. For study 2, there were four possible categories: [ 1= 1 correct fo od, 2= 2 correct foods, 3= 3 correct foods, 4 = incorrect ]. For Study 3, identifications were classified as either 1= correct food/food group or 2= incorrect. These category assignments were compared to overall acceptability ratings to examine the relations hip between acceptability scores and identification accuracy.
42 T able 3 1. List of i dentification c ategories for s tudy 1: s haped p ures Food Correct Food Correct Food Group Incorrect Food or Food Group Broccoli Broccoli Any vegetable, excluding potatoes/ beans Any other food choice or comment Green beans Green beans Any vegetable, excluding potatoes/beans Any other food choice or comment Chicken Poultry product chicken or turkey Any fish, beef, or pork item Any other food choice or comment Pork Pork a nd general pork processed foods (hot dog, ham, bacon, sausage, pepperoni) Any fish, beef, or poultry item Any other food choice or comment Table 3 2. List of i dentification c ategories for s tudy 2 : c ombination vs. i ndividual p ures Food Correct Food Incor rect Food Chicken Poultry product (chicken/turkey) Any other food choice or comment Carrot Carrot Any other food choice or comment Beef Any beef/meat product Any other food choice or comment Corn Corn Any other food choice or comment Potato Potato A ny other food choice or comment Table 3 3. List of i dentification c ategories for s tudy 3: r egular s odium v s r educed s odium b read p ures Food Correct Food /Food Group Incorrect Food or Food Group Bread Any grain product (e.g. bread cereal, pasta, rice) Any other food item or comment
43 Table 3 4. List of i dentification c ategories for s tudy 4 : c onventional p ures Food Correct Food Correct Food Group Incorrect Food or Food Group Baked Beans Baked beans or beans Vegetable or starch Any other food item or comment Refried beans Refried beans or beans Vegetable or starch Any other food item or comment Hummus Hummus or beans Vegetable or starch Any other food item or comment Pumpkin Pumpkin or sweet potato Vegetable Any other food item or comment
44 CH APTER 4 RESULTS Study 1: Shaped Pures All young persons were included in the analysis. Eleven persons in the elder group were excluded from data analysis due to misuse of the scale in rating samples. Tables 4 1 and 4 2 indicate acceptability ratings of y oung and older panelists of shaped and unshaped pures. Unshaped pures generally received higher acceptability ratings over shaped pures. Figure 4 1 illustrates the differences in acceptability ratings between age groups. Overall, these differences were not significant, signifying age as unrelated to shaping of pured foods. Figures 4 2 through 4 5 portray identification accuracy of each sample by both age groups. Shaping did not improve identification and in fact, some unshaped pures had greater correct identifications than the shaped counterpart. Generally, young persons were better at identifying the correct food while older persons were more likely to identify only the food group. Study 2: Combination versus Individual Pures Only panelists who participated in both sessions were included in data analysis As a result, data for 73 young persons and 51 elder persons were evaluated. Figure 4 6 compares the two combination meals for each age group, with beef pures significantly less liked for all attributes with young persons and flavor for older adults. Figures 4 7 and 4 8 demonstrate acceptance of individual pures for each age group, with the potato pure receiving the highest ratings for every attribute within both the beef and chicken meals fo r both age groups. In order to evaluate the relationship of individual with combination pured foods, an average of acceptability ratings for the three individual pures was calculated to create one value that could be compared. When
45 ratings of combination and individual foods was compared across age groups (Figures 4 9 and 4 10), it is clear that older adults rated individual pures quite differently than young persons while both groups rated combination pures similarly. Older adults tended to give higher ratings for individual pured food samples. In Figure 4 11, acceptability ratings of combination pures are compared to individual pures. In general, individual pures were given much higher acceptability ratings than the combination counterpart for the same meal type. Figures 4 12 and 4 13 illustrate identification accuracy for the beef and chicken meals, both when presented combined and individually. More foods were correctly identified when pured samp les were presented individually for both age groups as expected. Study 3: Regular Sodium versus Low Sodium Bread Pures All young and elder panelists were included in the analysis. Figure 4 14 demonstrates the comparison of acceptability ratings between regular sodium and reduced sodium bread pures f or both age groups. There were no significant differences between the samples for any attribute. Figure 4 15 portrays the relationship of age on acceptability ratings of the bread pure samples. It is apparent the older adults rated both the regular sodium and reduced sodium bread pures higher than the young adults, indicating a greater liking for this product. In Figure 4 16, identification accuracy is portrayed, with very few differences between bread pure samples. In general, most panelists were able t o identify the food or food group of this sample with young panelists slightly better at identifications than older panelists Study 4: Conventional Pures Only elder persons participated in this study and all panelists were included in analysis. Figure 4 17 illustrates acceptability ratings of the four conventional pured
46 food samples, with pumpkin receiving the highest ratings for every attribute versus hummus which scored the lowest ratings for every attribute. Baked beans and refried beans were simil arly liked across all attributes. Figure 4 18 conveys accuracy of identifications for conventional pured foods. A large number of panelists were able to correctly identify pumpkin, refried beans, and baked beans. Hummus had the lowest identification accur acy, though this was e xpected with a rather exotic food product. Acceptability Ratings and Identification Accuracy Comparison of acceptability ratings and identification accuracy are highlighted in Table 4 3 and 4 4. Student comparisons of identification and o verall lik e ability were significantly correlated at p value < 0.05 with th e following samples: shaped broccoli (p=0.046, r=0.20 3), shaped green beans (p=0.010, r=0.261 ) and regular bread (p=0.030 r=0.220 ) Elderly comparisons of ID and overall lik e a bility were significantly correlated at p value < 0.05 with the following samples : shap ed green beans (p= 0.015 r= 0.315 ), unshaped green beans (p= 0.0107, r= 0.3298) un shaped pork (p= 0.021, r= 0.299), and baked beans (p=0.011 r= 0.317 )
47 Table 4 1. S ample r atings by y oung a dults for s pecific a ttributes of s haped vs. u nshaped p ures Sample Overall Liking Appearance Mouthfeel Flavor Unshaped Chicken 5.2a 3.0a 5.2ab 5.6a Shaped Chicken 1.7a 6.1bc 5.1ab 5.2a Unshaped Pork 3.2a 2.9b 2.4a 4.6a Shap ed Pork 1.5a 9.7cd 3.3a 5.7a Unshaped Broccoli 2.4a 5.1bc 1.5a 2.7a Shaped Broccoli 0.43ab 6.3bc 1.5a 1.6ab Unshaped Green beans 0.90ab 2.7b 2.5a 0.83ab Shaped Green beans 5.8b 14d 10b 5.9b Note: Ratings were analyzed according to attrib ute. Ratings with the same letter within a specific attribute (column) are not significantly different at p < 0.05
48 Table 4 2. Sample r atings by o lder a dults for s pecific a ttributes of s haped vs. u nshaped p ures Sample Overall Liking Appearance Mouthfeel Flavor Unshaped Chicken 8.3a 15a 6.1abc 14a Shaped Chicken 6.6a 0.03b 6.6c 11c Unshaped Pork 2.8a 9.5ab 9.5ab 3.8abc Shaped Pork 6.7a 0.06b 7.7c 5.9bc Unshaped Broccoli 3.2a 3.7ab 18a 7.6ab Shaped Broccoli 1.3a 1.0b 15a 3.9abc Unshaped Green beans 1.2a 4.8ab 12a 0.98abc Shaped Green beans 8.7a 2.0b 3.3bc 6.1bc Note: Ratings were analyzed according to attribute. Ratings with the same letter within a specific attribute (column) are not significantly different at p < 0.05
49 A B F igure 4 1. Effect of a ge on m ean s ample r atings of s haped and u nshaped p ures. A) Shaped c hicken. B) Unshaped c hicken. C) Shaped p ork. D) Unshaped p ork. E) Shaped b roccoli. F) Unshaped b roccoli. G) Shaped g reen b eans. H) Unshaped g re en b eans (Note: indicates significance at p <0.05)
50 C D Figure 4 1 Continued
51 E F Figure 4 1. Continued
52 G H Figure 4 1 Continued
53 A B C D Figure 4 2. Identification a ccuracy of c hicken p ur es. A) Unshaped c hicken by y oung Adults. B) Shaped c hicke n by y oung a dults. C) Unshaped c hicken by o l der a dults D) Shaped c hicken by o lder a dults
54 A B C D Figure 4 3. Identification a ccuracy of p ork p ur es. A) Unshaped p ork by y oung a dults. B) Shaped p ork by y oung a dults. C) Unshaped p ork by o lder a dults. D) Shaped p ork by o lder a dults.
55 A B C D Figure 4 4. Identification a ccuracy of b roccoli p ur es A) Unshaped b roccoli by y oung a dults. B) Shaped b roc coli by y oung a dults. C) Unshaped b roccoli by o lder a dults. D) Shaped b roccoli by o lder a dults.
56 A B C D Figure 4 5. Identification a ccurac y of g reen b ean p ur es. A) Unshaped g reen b eans by y oung a dults. B) Shaped g reen b eans by y oung a dults. C) Unshaped g reen b eans by o lder a dults. D) Shaped g reen b eans by o lder a dults
57 A B Figure 4 6. Sample r atings for s pecific a ttributes of c ombination p ures for e ach a ge g roup A) Young a dults. B) Older a dults (Note: indicates significance at p <0.01)
58 A B Figure 4 7. Sample r atings for s pecific a ttributes of i ndividual c hicken, c arrot, and p otato p ures. A) Young a dults. B) Older a dults (Note: Ratings with different letters for a specific attribute indicate significance at p <0.01)
59 A B Figure 4 8. Sample r atings for s pecific a ttributes of i ndividual b eef, c orn, and p otato p ures A) Young a dults. B) Older a dults (Note: Ratings with different letters for a specific attribute indicate significance at p <0.01)
60 A B Figure 4 9. Effect of a ge on m ean s ample r atings of c ombination versus i ndividual p ure m eals. A) Combination c h icken, c arrot, p otato m eal. B) Individual c hicken, c arrot, p otato m eal ( Note: indicates significance at p <0.05)
61 A B Figure 4 10. Effect of a ge on m ean s ample r atings of c ombination versus i ndividual p ure m eals. A) Combination b eef, c o rn, p otato m eal. B) Individual b eef, c orn, p otato m eal ( Note: indicates significance at p <0.05)
62 A B Figure 4 11. Mean s ample r atings of c ombination versus i ndvidual p ure m eals. A) Chicken, c arrot, p otato p ure. B) Beef, c orn, p otato p ure. ( Note: indic ates significance at p <0.05).
63 A B C D Figure 4 12. Identification a ccuracy of c ombination and i ndividual c hicken, c arrot, p otato m ea l p u res A) Combination p ures by y oung a dults. B) Individual p ures by y oung a dults. C) Combination p ures by o lder a dults. D) Individual p ures by o lder a dults
64 A B C D Figure 4 13. Identification a ccuracy of c ombination and i ndividual b eef, c orn, p otato m eal p ures A) Combination p ures by y oung a dults. B) Individual p ures by y oung a dul ts. C) Combination p ures by o lder a dults. D) Individual p ures by o lder a dults
65 A B Figure 4 14. Sample r atings for s pecific a ttributes of r egular s odium and r educ e d s odium b read p ur es. A) Young a dults. B) Older a dults ( Note: indicates significance at p <0.05)
66 A B Figure 4 15. Effect of a ge on m e an s ample r atings of r egular s odium and r educed s odium b read p ur es. A) Regular s odium b read B) Reduced s odium b read ( Note: indicates significance at p <0.05)
67 A B C D Figure 4 16. Identification a ccuracy o f r egular s odium and r educed s odium ( RS) b read p ures A) Regular s odium b read p ures by y oung a dults. B) Reduced s odium b read p ures by y oung a dults. C) Regular s odium b read p ures by o lder a dults. D) Reduced s odium b read p ures by o lder a dults
68 Figure 4 17. Sample r atings for s pecific a ttr ibutes of c onventional p ur es by o lder a dults. (Note: Ratings with different letters for a specific attribute indicate significance at p <0.05).
69 A B C D Figure 4 18. Identification a ccuracy of c onventional p ures by o lder a dults A) Baked b eans. B) Refried b eans. C) Hummus. D) Pumpkin
70 Table 4 3 Overall a cceptability versus i dentification a ccuracy of y oung a dults Sample Spearman Coeffic ient P value Unshaped Chicken 0.01 0 0.9 27 Shaped Chicken 0.032 0.755 Unshaped Pork 0.0 07 0.95 0 Shaped Pork 0.0 09 0.93 2 Unshaped Broccoli 0.07 1 0.49 0 Shaped Broccoli 0.20 3 0.0 46* Unshaped Green beans 0.0 09 0.9 27 Shaped Green beans 0.26 1 0.01 0* C ombination Beef Meal 0.07 1 0.5 47 Combination Chicken Meal 0.01 5 0.90 1 Individual Beef Meal 0.10 3 0.38 3 Individual Chicken Meal 0.03 2 0.7 89 Regular Sodium Bread 0.09 4 0.36 1 Reduced Sodium Bread 0.18 0 0.0 78 Note: Ratings were compared to identificat ion classification via Spearman Correlation Coefficient test. *indicates significance at p value < 0.05.
71 Table 4 4. Overall a cceptability versus i dentification a ccuracy of o lder a dults Sample Spearman Coefficient P value Unshaped Chicken 0.058 0.660 S haped Chicken 0.189 0.156 Unshaped Pork 0.299 0.021* Shaped Pork 0.015 0.908 Unshaped Broccoli 0.060 0.658 Shaped Broccoli 0.155 0.242 Unshaped Green beans 0.330 0.011* Shaped Green beans 0.315 0.015* Combination Beef Meal 0.046 0.749 Combinati on Chicken Meal 0.024 0.865 Individual Beef Meal 0.170 0.237 Individual Chicken Meal 0.119 0.409 Regular Sodium Bread 0.084 0.481 Reduced Sodium Bread 0.087 0.472 Baked Bean 0.317 0.011* Refried Bean 0.019 0.883 Hummus 0.076 0.556 Pumpkin 0.2 02 0.112 Note: Ratings were compared to identification classification via Spearman Correlation Coefficient test. *indicates significance at p value < 0.05.
72 CHAPTER 5 DISCUSSION Given the hedonic gLMS is not as self explanatory as other scales used in sensory research, training was provided to make sure all panelists understood how to appropriately use the scale. Yet, several panelists inappropriately used the scale and wer e excluded from data analysis. Examples of misuse include making multiple markin gs on the scale, writing words rather than a single point on the line, and using only extreme values ( 100 0, 100) with every question. Additional details in this matter are discussed within results of each study section. Further, although a guideline was followed for identification analysis, classifications were at the discretion of the researchers. It is possible that these could be modified and standardized if future studies were to be carried out to make the process as objective as possible. Nonetheles s, identifications needed to be analyzed in a logical manner and this is the way it was done for these preliminary studies. It is also important to note that acceptability ratings for certain products may not reflect acceptance of these products in an a lternate setting, as every sample was presented at room temperature (within 19 25C). The decision to serve samples in this temperature range was made for two reasons. First, it allowed ease of preparation and ensured panelists received and consumed the pu red food samples within the same general temperature range. Second, and more importantly, samples at a neutral temperature provide for optimal flavor perception, which was necessary in identification of samples. However, in choosing to use this temperatur e range for serving samples, there are a few limitations. It is obvious that acceptability ratings may in fact be lower than if the samples were presented at recommended serving temperatures. This is
73 probably most pronounced for the meat/poultry products a nd the vegetables. The bean and bread pures are likely more appropriate served at room temperature, as these products would likely be consumed at these temperatures. Further, identification accuracy may actually be skewed in favor of panelists who had a b etter chance at identifying a product that w as served at this temperature. Nonetheless, pured foods are commonly served in hospitals and long term care homes and by the time the patient receives the meal and actually consumes it, the temperature has often dropped substantially. Thus, it may be argued that the choice to serve room temperature samples is actually more reflective of the way pured foods are often consumed, and would therefore be a better representation of acceptability and recognition accurac y. Study 1: Shaped Pure s Sample Ratings : General Samples were not completely randomized due to the large number of samples and limited labor. Instead, samples were randomized for each 10 person group (young cohort) and 20 person group (elder cohort). The differences in these groups are due to variations in testing site capabilities. Each set of randomized samples ensured that the same food (shaped and unshaped version) would never be subsequent to one another. For example, if shaped broccoli was presented in position one, than unshaped broccoli could not be presented in position two. Each sample was presented one at a time with sufficient time between samples to avoid sensory fatigue. Mean rating for favorite food was approx imately 65 for older adults and 60 for young adults. Overall liking averaged from 9 for shaped green beans to 8 for unshaped chicken pure with older adults. Young panelists gave average overall liking ratings ranging from 6 for shaped green
74 beans to 5 for unshaped ch icken. Overall the se samples were near neutral on the scale and far from the ratings given to favorite food. In general, the appearance of unshaped pure s was preferred over the shaped pure s. Among younger panelist s, shaped pork and shaped green beans were significantly rated lower than any other samples, while unshaped chicken received the highest appearance rating. Older panelist s also preferred the unshaped pure s and higher appearance ratings were given to those samples, though the differences between samples were no t as significant as with the younger group. With overall liking, there were not any differences in samples for older panelist s. Young panelist s preferred overall the pork and chicken pure s over the vegetable pure s, with the exception of unshaped broccoli which was rated similarly to the protein foods. Ratings for mouthfeel did not really show a trend in either age group though there were significant differences between samples within each group. For flavor, students generally gave higher ratings for the pork and chicken pure s when compared to the vegetable pure s, regardless of whether it was shaped or unshaped. While differences were not as significant between samples for older panelist s, this group seemed to care more about the shaping of the pured fo od with regards to flavor, where unshaped pure s were given higher flavor ratings. Since flavor is irrelevant to shape, it is possible the older panelist s shifted their feelings of overall liking into the flavor category as a way to signify greater liking of this product. Given the shaped and unshaped pures were identical in product formulation, it is also possible that in disrupting the gel structure of the shaped pures to prepare the unshaped pures, there was an alteration in mouthfeel which may have b een preferred. As such, the unshaped pures may indeed
75 have been perceived differently, simply from the manipulation of the molded structure in preparation of an unshaped sample. Sample Ratings and Identification Accuracy: Young vs. Older Panelists Overal l, young and older panelist s rated samples similarly and thus used the scale in a similar way. There were a few exceptions with each sample, most often with the attribute mouthfeel, followed by two instances with appearance and one with flavor. In each cas e, the older adults gave higher ratings when compared to young panelists. Since most of these differences were in relation to mouthfeel, it is possible the older panelist s gave higher ratings because softer foods are generally consumed in greater quantitie s for this population due to issues with chewing and swallowing and thus this texture may be better accepted. In terms of identification accuracy, young panelists were generally better at identifying foods correctly than older panelists. Both age groups h ad similar numbers of panelists were more likely to get the correct food and older panel ists were better at identifying the food group. This is consistent with previous studies and demonstra tes that young panelists have keener taste acuity than older panelists, likely as a result of more discriminating sensory perception. One of the objectiv es of this particular study was to evaluate the effect of shaped pures on identification accuracy. It was hypothesized that the shaped pure would be more easily identifiable since it would resemble the original food product in its natural form. It is cur ious to see that in fact, correct identifications were generally higher for the unshaped pures when compared to the shaped pures for both age groups. There are
76 several possible explanations for this finding. First, perhaps the shaped pures did not resem ble the actual food well enough to these panelists. It is also possible that the foods, as a molded pure, seemed like such a foreign item to the panelists that a food group could not be associated. This is quite possible, as the texture and color of foods are inherently part of how humans recognize food items. With pured food s, the texture is modified, the color may be slightly altered due to processing, and the shaping process may not be sufficient to make up for these modifications. Study 2: Combination versus Individual Pure s Sample Ratings : General Given the same food components, pured food s presented as separate individual components received higher sample ratings, on average, when compared to the same pured food s presented in a single combined pu re. Favorite foods were ranked as 68 for older adults and 63 for young adults. Overall liking for combination chicken meal was rated 2 for older adults and 8 for young adults, on average. Mean ratings for the attribute overall liking with individual chick en were 30 for older adults and only 10 for young persons. Overall liking for combination beef meal was 18 for older adults and 16 for young persons. With the individual beef meal, older adults gave a mean rating for overall liking a 3 while older adult s ranked this sample a 5. These values are far from the favorite food rating, particularly with the combination beef meal for both age groups. For older adults, the individual chicken meal was liked almost half as much as their favorite food, on average. Because the individual pures were averaged to be compared to the single value of a combined pure, it is quite possible that one of the individual pure components resulted in an inflated mean. In fact, the potato pure was rated significantly higher than
77 all of the individual pures. Thus, the average of the individual pures may reflect the greater acceptability of the potato pure rather than the greater acceptability of individual pures in general. Nonetheless, if an older person were to receive the s ame meal presented individually and then as combined, it is possible that intake would be greater for the individual component meal as a result of increased acceptability of at least one of the pured food s. Essentially, since the potato pure was more a cceptable, it would probably have a greater chance of being consumed. Even if the individual potatoes were the only pured food consumed, it would still equate to greater intake than if the potatoes that were hidden in the combination pure and not consume d at all. It should also be noted that while the individual and combined pures included the same food types, these were not manufactured by the same company nor processed in the same manner. Thus, it is very possible the discrepancy between samples is a r esult of variations in the food source, quality, and processing parameters rather than simply a matter of presentation (i.e. individual or combined). This matter clearly needs to be further tested with other foods and combinations as well as with oral inta ke studies. Regardless of whether the samples were presented as individual or combined pures, it was very clear that the beef meal was much less acceptable than the chicken meal. This was most evident with the combination pures, where both age groups ga ve significantly greater ratings for overall liking and flavor of the chicken meal over the beef meal. Upon closer inspection of the individual pures, specifically of the protein pures (chicken and beef), both age groups gave higher ratings for the chick en pure across all attributes (with the exception of appearance by young panelists who rated beef slightly
78 higher). This suggests that it is perhaps the protein source which is influencing acceptability ratings of both combination and individual pures. I n fact, animal muscle proteins such as beef, poultry, and pork do not lend to pureing as well as other high protein foods (beans, dairy, eggs), as things like gritty/sandy texture and poor flavor are common concerns. As with comparing individual versus co mbined pures, it is possible that differences in acceptability of beef versus chicken meals are a result of alterations in product formulation and not necessarily the food type itself. Sample Ratings and Identification Accuracy: Young vs. Older Panelists As mentioned above, the young and older adults were fairly similar in rating these samples. Both groups generally had higher ratings for the individual pures over the combination pures, the chicken meals over the beef meals, and the potato pure was the best liked pured food of all of the samples. The primary differences in this study were observed in identification accuracy. Unlike the other studies where panelists were categorized at correct, incorrect, or able to identify the correct food group, it w as more important to assess whether individual versus combined pures influenced identification accuracy. As such, panelists were categorized at correctly identifying zero (incorrect), one, two, or all three of the components. As expected, when pured foo d s were presented individually, correct identification of more foods occurred. This was especially evident for beef pures, where not one person in either age group could correctly identify the three foods in the combination pures whereas in the beef meal of individual pures, nearly half of older adults and over half of young panelists got at least two correct foods. The same was observed with the chicken meal, though a small percentage of both age groups were able to identify the three foods in the combi nation meals. Nonetheless, less than half
79 were able to correctly identify more than one correct food and many could not identify a single pured food With the chicken meal of individual pures, o ver half of each age group was a ble to identify at least two correct foods, similar to the beef meal. In terms of age, the young persons were again better at identifying the foods than the elder group. With the chicken meal, every young person was able to identify at least one food whereas 10% of older persons wer e still unable to identify a single item. Considering both the combined and individual presentations of the chicken meal, young persons were better at identifying all three foods in the sample. With the individual beef meal, young persons had a higher perc enta ge of persons getting at least one food item correct when compared to older panelists. These findings support the hypothesis that young people are better at correctly identifying pured food s than older persons. However, there was one exception to this trend, where the opposite was observed. For the combination beef meal, there were more incorrect identifications (41%) by young panelists than by older panelists (31%). This finding is rather inexplicable, especially given the young group was much better at identifying the individual beef meal pures. Perhaps the older persons were more familiar with the aroma and flavor of the specific combination beef meal used in this study. Study 3: Regular Sodium versus Low S odium Bread Pure s Sample Ratings : General Overall, the bread pures were given fairly good acceptance ratings. Both bread adults rated favorite food approximately 68 with overall ratings of regular and reduced so dium bread samples as 33 and 29, respectively. You ng persons had a mean rating of 64 for favorite food and gave overall acceptability ratings a 21 for regular bread and 23
80 for reduced sodium bread. Unlike many of the samples tested in the other studies, th ere were no significant differences for any attribute between the regular and reduced sodium bread pures. This rejects the original hypothesis that regular bread pures would be more acceptable than reduced sodium bread pures. It is possible that the sod ium range was not expansive enough to where panelists could pick out differences. Although 130 mg sodium is the general sodium content of an average bread slice, this gover products. Per haps if the study were done with hi gher sodium bread compared to a difference in acceptability would have been foun d It should be noted that although statistically there were no disparities in ratings, several panelists in each age cohort made comments either verbally or in the pref erable with this particular food matrix. Nonetheless, if the majority of panelists are unable to pick out substantial difference s between two products with a doubling of sodium content, the lower sodium product could be served without adverse sensory qual ities yet with a better nutrient profile. It is also possible that this finding could extend to o ther products, especially grain based products. Because many older adults on pured diets are also on sodium restricted diets for vario us reasons, this could h ave significant implications. Reduced sodium pured food s prepared in hospitals, long term care homes, and by personal caregivers are commonly prepared with very low sodium or salt free formulations, resulting in an unpalatable product. Rather than restric ting sodium completely, if pured
81 food s could simply be prepared with less sodium, as done in this study, perhaps acceptability would be enhanced and ultimately lead to increased intake. Sample Ratings and Identification Accuracy: Young vs. Older Panelists While there were no differences observed for the two bread samples within each age group, there were several significant differences between young and older persons, particularly with the regular bread pure. In general, older adults gave higher ratings across the board for every attribute in both samples. These ratings were significantly different for each attribute for the regular bread pure sample, though only significant for appearance and mouthfeel of t he reduced sodium bread pure. This can possibl y be explained by the larger discrepancy in acceptability of the samples by older adults, who appear to give slightly higher ratings for the regular bread compared to the reduced sodium bread and consequently have more significant differences for those att ributes. As expected, most panelists were able to identify the food or food group of this sample. It should be noted that the majority of these correct responses were a grain it em other than bread. This can be explained by the fact that g rain food product s are inherently recognized by textural qualities and once this element is removed, it becomes increasingly difficult to discern one grain from another. Further, once a grain is made into a pured food it may begin to resemble grain products traditionally consumed as a pured food or with a comparable texture, which is what was observed here. As such, responses were simply classified as correct (including all grain foods) or incorrect. In fact, most identifications were some type of cereal product such as oatmeal or cream of wheat rather than bread In terms of age groups, young adults had higher correct food and food group responses for both samples when compared to older adu lts as expected. When
82 comparing identification accuracy of regular versus reduce d sodium breads, it is apparent the change in sodium content did not influence young panelists who had nearly identical accuracy values However, for older adults there were much higher percent correct food/food group for the regular sodium bread pure com pared to the low sodium sample. It is possible the sodium difference was more pronounced for older adults who may require higher sodium levels to better perceive flavors Study 4: Conventional Pure s Sample Ratings and Identification Accuracy Conventional pured food s are those foods which are consumed as pures by the general population, either as a whole food or as an accompan iment (e.g. a dip or filling). Three bean pures and pumpkin filling were chosen as samples for this preliminary study. Although o nly the older adults were used in this study, the findings are still important. Ratings for pumpkin and baked beans were rather high, especially rite food. Overall liking for pumpkin was given a 63 and baked beans a 39 compared t o mean rating of 73 for favorite food. This means panelists liked pumpkin filling almost as much as their favorite food and baked beans a little over half as much. Refried beans and hummus received lower mean ratings, at 17 and 21, respectively. All sample s differ ed significantly on t beans, and hummus, r espectively. No significant differences were observed in or baked beans and refried beans, though significant differences were apparent for pumpkin and hummus with pumpkin again being the highest rating and hummus having the lowest ratings.
83 If purely evaluating actual ratings rather than statistical significan ce it appears as though for each sample, acceptance ratings are in the same relative position across attributes. Essentially, for each attribute, pumpkin received the highest ratings followed by baked beans, refried beans, and finally hummus. It is possib le that panelists did not distinguish between attributes, leading to a effect, whereby one attribute influences the ratings of other attributes. For example, a panelist who really liked the flavor of pumpkin may have given high scores for all att ributes, even if that individual did not necessarily think the appe arance was equally acceptable. It is quite possible this effect was most pronounced for hummus, as many panelists commented that the sample was quite salty and that this feature negat ively impacted overall liking. This issue could be addressed in future studies, where panelists could receive further training on differentiating attributes or perhaps a different scale could be used that is more straightforward. While no direct comparison was made with non conventional pured food s, it was clear that the samples tested here were generally better received. After testing these particular samples, many panelists commented that the se conventional pured food s were much better overall than any of th e other samples tested in previous studies. This clearly marks the need for future research with additional samples and comparison of acceptability with conventional pures versus non conventional pured food s. A study comparing across age groups may be us eful as well, though this feature is most important when considering identification accuracy. In terms of accurate recognition, the older adults generally did much better at identifying the correct food with the conventional pured food s, most markedly wi th the
84 baked bea ns, refried beans and pumpkin. Hummus had a very high percentage of incorrect identifications. This can in part be explained by the fact that it is traditionally an ethnic food product that has gained popularity only recently in the United States. It is also possible that the saltiness and other flavor components including garlic and tahini masked the true flavor of the beans, making it difficult to identify. Pumpkin and refried beans had the highest correct identifications, likely as these were the most familiar and best fit the definition of a conventional pured food Acceptability Ratings and Identification Accuracy When exploring the relationship of identification accuracy and acceptability scores using Spearman correlation coefficient no identifiable trend was observed Although there were a few sign ificant values the highest correlation coefficient was only r=0.330 which is n ot considered a strong correlation. Thus we can conclude based on this data that identification had little im pact on overall likeability of a sample. It was originally hypothesized that co rrect identification would yield improved acceptability based on the logical assumption that recognition of a food being consumed would be better accepte d than eating something unknown. However, there was no consistent trend with identification accuracy and acceptability ratings across all attributes for all samples t ested
85 CHAPTER 6 CONCLUSION Pured foods are important for persons with severe swallowing problems, most commonl y older adults but may include young persons as well. For elderly individuals with dysphagia, decreased olfaction and gustation combined with texture modification may result in reduced sensory pe rception of a pured food particularly in terms of recogniti on. In general, both age groups had comparable incorrect responses with sample identification. However, young persons were better at identifying exact foods while older persons could typically only classify the food group. Young people appear to have a mor e discerning palate yet older persons are still able to identify the general food item. Identification ability does not seem to influence acceptability of pured foods. This is a key finding, as it suggests that perhaps food recognition is a superfluous component in pured food acceptability. It appears that sensory characteristics are more important in influencing overall acceptability ratings. Nonetheless, there is disconnect between these findings and what is reported in hospi tals and long term care h omes. Inability to identify pured food s is a common complaint among older persons, yet this study found no relation between recognition and acceptability. Perhaps the complaint is actually a manifestation of deeper emotions toward being on a pured diet, including but not limited to the stigma of such a diet (may such drastic dietary changes, or some other explanation. Because panelist s used in these studies were not individuals with swallowing problems it is also possible that dysphagia itself may interfere with the ability to identify foods and also influence acceptability. A portion of volatiles in a food are released after swallowing which directly impacts flavor perception. It is possible that
86 pers ons with swallowing difficulty do not have the same degree of flavor release in the testing with identification should be done using persons with dysphagia. Acceptability is affected by several factors, including specific attributes within a food as well as how it is prepared and presented. Of those factors evaluated here, there were several worthy findings. First, it appears that traditional scooped pured food s (unshaped) a re much more preferred than the shaped counterparts. This has implications for manufacturers of shaped pures and long term care facilities and hospitals purchasing these items. Because shaped pures may be more expensive and less acceptable as observed in this study, it seems logical that these pures should not be produced and served. One of the largest differences observed between the shaped versus unshaped samples for every food was in mouthfeel, with shaped pures typically receiving lower scores. Beca use the food items were identical other than presentation, it is possible that in disrupting the gel structure of the shaped pure led to an unshaped pure which provided a more pleasant mouthfeel. Shaping is not the only important quality of appearance i acceptability of samples. It appears that presenting pured food s as whole individual foods is more acceptable than combining pures into one. It is not known whether this is because of differences in appearance, flavor, texture, etc. but it is important in understanding how pured food s should be presented. Because individual pures had greater acceptability ratings, it is likely that this method of serving pured food s would result in greater consumption. This is imperative, as intak e is low for many older adults,
87 particularly those on a pured a diet. Future studies could be done to evaluate consumption patterns with these two forms of meal presentation. In addition to appearance, the issues of flavor and familiarity were addressed in the final two studies. In terms of flavor, salt is a key taste that allows flavors to be perceived optimally. Reduced sodium products generally do not have the same level of flavor as th e regular sodium counterparts. In testing the two bread pure sampl es of varied sodium content, there were essentially no differences between the samples but there were differences in how each age c ohort rated the samples. As there were no real differences between samples, it is possible that sodium content of other pure d food s could be reduced to the same level without adversely affected acceptability. Of course, additional sensory studies would need to be done to evaluate whether this pattern applies for foods other than the bread pure tested here. It is also important to note that the older adults gave generally higher ratings for both samples when compared to the young persons. This was especially true for the regular sodium bread pure, implying that older adults have a greater acceptance of the higher sodium product perhaps as a result of increased flavor perception. Nonetheless, since there were no real differences between the samples, there are potential implications for reducing sodium in other pured food s. Conventional pured food s which represent foods traditi onally served as pures were generally the most acceptable products of all samples tested, with the exception of hummus. It is not clear whether these products were more accepted due to increased familiarity with consuming the particular food as a pure or because of superior sensory qualities. Even so, future studies should evaluate additional types of conventional
88 pured food s as well as test direct relationships between conventional and non conventional samples. If conventional pured food s are more acce ptable, perhaps the types of pured food s currently being served at home and in long term care facilities or hospitals needs to be re evaluated. It would also be vital to test consumption patterns for conventional versus non conventional pures, as it is l ikely that intake will increased with a more accepted product. Serving conventional pured food s may also reduce some of the stigma associated with a pured diet, as even those without dysphagia will consume these types of pured food s as part of a regular diet. These studies highlight some important issues that need to be addressed in terms of product development and service of pured food s. They also brought forth several questions that could be addressed in future studies. If the recommendations listed in this paper are taken into consideration for further research, it may be possible to improve not only pure d foods as a product but potentially quality of life of many individuals obliged to follow this type of diet.
89 APPENDIX A HgLMS AND DEMOGRAPH ICS BALLOT Food Pur e Taste Panel Question # 1. Please indicate your gender. Male Female Question # 2. Please enter your age. __________
90 S C A L E 1. Now, please take a few minutes to identify the strongest LIKING (i.e. pleasure ) of any kind that you have ever experienced. 2. Once you have identified your strongest LIKING experienced, please write it down in the space provided below 3. Please remember to use the stronges t liking that you've identified and written down as the top of your scale (100). Please write the strongest LIKING OF ANY KIND YOU'VE EXPERIENCED in the space below, and remember that this will be 100 on your scale. ____________________________________________________________________________________ ____________________________________________________________________________________ ____________________________________________________________________________________
91 S C A L E 1. Now, please take a few minutes to identify the strongest DIS LIKING (i.e. displeasure) of any kind that you have ever experienced. 2. Once you have identified your strongest DIS LIKING experienced, please write it down in the space provided below 3. Please remember to use the strongest dis liking that you'v e identified and written down as the bottom of your scale ( 100). Please type the strongest DIS LIKING OF ANY KIND YOU'VE EXPERIENCED in the space below and remember that this will be 100 on your scale. ______________________________________________________ ______________________________ ____________________________________________________________________________________ ____________________________________________________________________________________
92 S C A L E PRACTICE QUESTIONS On the line scale, 100 indicate s the most intense liking (i.e. pleasure) you have ever experienced (no matter what the source). Similarly, 100 indicates the opposite: the most intense dis liking you have ever experienced. Neutral is indicated by 0. Please use your 100 and 100 (written on the paper provided) to answer the following questions. Eating your favorite food Strongest Neutral Strongest Dislike Like 100 0 100 Eating your least favorite food Strongest Neutral Strongest Dislike Like 100 0 100 Spending time wi th your loved ones Strongest Neutral Strongest Dislike Li ke 100 0 100
93 The most intense anger you've experienced Strongest Neutral Strongest Dislike Like 100 0 100 The shyest you've ever been Strongest Neutral Strongest Dislike Like 100 0 100 The m ost inspired you have ever been by a lecture Strongest Neutral Strongest Dislike Like 100 0 100
94 APPENDIX B SAMPLE COMPUSENSE T EST BALLOT Sample 213 Question # 1 Please use your S C A L E 1 to rate the following attributes for Sample 213 Overall Liking Strongest Neutral Strongest Dislike Like 100 0 100 Appearance Strongest Neutral Strongest Dislike Like 100 0 100 Mouthfeel Strongest Neutral Strongest Dislike Like 100 0 100 Flavor Strongest Neutral Strongest Dislike Like 100 0 100 Question # 2 Please try to ide ntify this food ( sample 213). If you cannot identify it exactly, please guess the food or food group. ______________________________________________________________________ __________________________________________________________________ ____ Please lif t the window to receive your next sample
95 Sample 213 Question # 1 Please use your S C A L E 1 to rate the following attributes for Sample 213 Overall Liking Strongest Neutral Strongest Dislike Like 100 0 100 Appearance Strongest Neutral Strongest Dislike Like 100 0 100 Mouthfeel Strongest Neutral Strongest Dislike L ike 100 0 100 Flavor Strongest Neutral Strongest Dislike Like 100 0 100 Question # 2 Please try to identify this food ( sample 213). If you cannot identify it exactly, please guess the food or food group. ______________________________________________________________________ __________________________________________________________________ ____ Please lift the window to receive your last sample.
96 Please lift the window to let the server know you h ave finished. Thank you!
97 APPENDIX C SAMPLE PAPER TEST BALLOT Sample 213 Question # 1 Please use your S C A L E 1 to rate the following attributes for Sample 213 Overall Liking Strongest Neutral Strongest Dislike Like 100 0 100 Appearance Str ongest Neutral Strongest Dislike Like 100 0 100 Mouthfeel Strongest Neutral Strongest Dislike Like 100 0 100 Flavor Strongest Neutral Strongest Dislike Like 100 0 100 Question # 2 Please try to identify this food ( sample 213). If you cannot identify it exactly, plea se guess the food or food group. ______________________________________________________________________ __________________________________________________________________ ____ Please wait for your next sample, then continue to the next page.
98 Sample 213 Qu estion # 1 Please use your S C A L E 1 to rate the following attributes for Sample 213 Overall Liking Strongest Neutral Strongest Dislike Like 100 0 100 Appearance Strongest Neutral Strongest Dislike Like 100 0 100 Mouthfeel Strongest Neutral Strongest Dislike Like 100 0 100 Flavor Strongest Neutral Strongest Dislike Li ke 100 0 100 Question # 2 Please try to identify this food ( sample 213). If you cannot identify it exactly, please guess the food or food group. ______________________________________________________ ________________ __________________________________________________________________ ____ Please complete this form then continue to the next page.
99 Please raise your hand to let the server know you have finished. Thank you!
100 APPENDIX D HgLMS TRAINING SHEET 1. Hello, today you will be tasting food purees and answering a few questions about the samples you taste. 2. Please answer the demographic questions appearing on the first page. 3. First, you will create your own personal scale t o use for answering questions. The anchors you choose are for your use and will not be shared or used directly in this research. The top anchor of your scale, +100 will be the strongest liking of an experience you have had. Some examples are falling in l ove, traveling, or spending time with your family and friends. o Please write your +100 on the paper in front of you. The bottom anchor of your scale, 100 will be the strongest disliking of any kind you have ever experienced. Some examples are the deat h of a loved one or being very ill. o Please write your 100 on the paper in front of you. 4. You will use this scale when answering questions. Keep in mind that the values you select should be appropriate of how you personally rank the item relative to your anchors. For instance, pretend your favorite food is chocolate. If you like chocolate more than the time you were most inspired by a lecture, chocolate should get a higher rating. The same is true for negative statements. Pretend your least favorite food is brussel sprouts. If you dislike your most angry experience more than you dislike eating brussel sprouts, give your most angry experience a more negative value on the scale. 5. Giving a statement a rating of 0 means you neither like nor dislike the ite m. 6. Do not agonize over choosing an answer for any single question. Pick a number which you feel is representative and natural. 7. Now you will answer several warm up questions based on the scale you have just created. These are meant to get you comfor table with using this scale. Keep your scale anchors in mind while you are answering these questions. You can refer to your anchors if you would like a reminder. 8. Now you will use this scale to rate samples of food purees. You will receive your first s ample and be asked a few questions on that sample. Each sample is on ONE page.
101 9. When you finish answering questions for a sample, look up to the front of the room so we know you have finished. Once everyone in the group finishes, we will bring your next sample. 10. Things to keep in mind when tasting samples: Take a bite of cracker and sip of water BEFORE EVERY SAMPLE to cleanse your palate. If you need more water or crackers, please ask one of the workers. Make sure the number on your sample cup matches the sample number on the page for the questions you are answering. Again, each sample gets ONE page of questions. Finally, and most importantly, do NOT talk to other panelists during testing. Does anyone have any questions?
102 REFERENCE LIST Bartoshuk LM, Duffy VB, Fast K, Green BG, Prutkin J, Snyder DJ. 2002. Labeled scales (e.g. category, Likert, VAS) and invalid across group comparisons: what we have learned from genetic variation in taste. Food Quality and Preference 14:125 38. Bartoshuk LM, Duffy VB, Green BG, Hoffman HJ, Ko CW, Lucchina LA, Marks LE, Snyder DJ, Weiffenbach JM. 2004. Valid across group comparisons with labeled scales: th e gLMS versus magnitude matching. Physiology & Behavior 82:109 14. Bartoshuk LM, Fast K, Snyder DJ. 2005. Differences in our sensory worlds: invalid comparisons with labeled scales. Current Directions in Psychological Science 14(3):122 5. Bischmann DA, Wit te KL. 1996. Food identification, taste complaints, and depression in younger and older adults. Experimental Aging Research 22(1):23 32. Blaise M. 2009. Mealtime experiences of hospitalized older patients requiring a puree consistency diet. University of M ontreal. Booth DA, Conner MT, Gibson EL. 1989. Measurement of food perception, food preference, and nutrient selection. Annals of the New York Academy of Sciences 561:226 42. Cassens D, Johnson E, Keelan S. 1996. Enhancing taste, texture, appearance, and p resentation of pureed food improved resident quality of life and weight status. Nutrition Reviews 54(1):S51 S4. Chen P H, Golub JS, Hapner ER, Johns MM, III. 2009. Prevalence of perceived dysphagia and quality of life impairment in a geriatric population. Dysphagia 24(1):1 6. Colodny N. 2005. Dysphagic independent feeders' justifications for noncompliance with recommendations by a speech language pathologist. American Journal of Speech Language Pathology 14(1):61 70. deGraaf C, vanStaveren W, Burema J. 1996 Psychophysical and psychohedonic functions of four common food flavours in elderly subjects. Chemical Senses 21(3):293 302. Drewnowski A. 1997. Taste preferences and food intake. Annual Review of Nutrition 17:237 53. Dubose CN, Cardello AV, Maller O. 198 0. Effects of colorants and flavorants on identification, perceived flavor intensity, and hedonic quality of fruit flavored beverages and cake. Journal of Food Science 45(5):1393 &.
103 Fillion L, Kilcast D. 2001. Food texture and eating difficulties in the el derly. Food Industry Journal 4(1):27 33. Forde CG, Delahunty CM. 2002. Examination of chemical irritation and textural influence on food preferences in two age cohorts using complex food systems. Food Quality and Preference 13(7 8):571 81. Forde CG, Delahu nty CM. 2004. Understanding the role cross modal sensory interactions play in food acceptability in younger and older consumers. Food Quality and Preference 15(7 8):715 27. Ginocchio D, Borghi E, Schindler A. 2002. Dysphagia assessment in the elderly. Nutr ition Therapy & Metabolism 27(1):9 15. Griep MI, Mets TF, Massart DL. 1997. Different effects of flavour amplification of nutrient dense foods on preference and consumption in young and elderly subjects. Food Quality and Preference 8(2):151 6. Guichard E. 2002. Interactions between flavor compounds and food ingredients and their influence on flavor perception. Food Reviews International 18(1):49 70. Hall G, Wendin K. 2008. Sensory design of foods for the elderly. Annals of Nutrition and Metabolism 52:25 8. Imram N. 1999. The role of visual cues in consumer perception and acceptance of a food product. Nutrition & Food Science (4/5):224 8. Koskinen S, Kalviainen N, Tuorila H. 2003. Flavor enhancement as a tool for increasing pleasantness and intake of a snack product among the elderly. Appetite 41(1):87 96. Laureati M, Pagliarini E, Calcinoni O. 2008. Does the enhancement of chemosensory stimuli improve the enjoyment of food in institutionalized elderly people? Journal of Sensory Studies 23(2):234 50. Laureati M, Pagliarini E, Calcinoni O, Bidoglio M. 2006. Sensory acceptability of traditional food preparations by elderly people. Food Quality and Preference 17(1 2):43 52. Lim J. 2011. Hedonic scaling: a review of methods and theory. Food Quality and Preference 2 2:733 47. Lindgren S, Janzon L. 1991. Prevalence of swallowing complaints and clinical findings among 50 79 year old men and women in an urban population. Dysphagia 6(4):187 92.
104 Lister R. 2006. Oropharyngeal dysphagia and nutritional management. Current Op inion in Gastroenterology 22(4):341 6. McHorney CA, Robbins J, Lomax K, Rosenbek JC, Chignell K, Kramer AE, Bricker DE. 2002. The SWAL QOL and SWAL CARE outcomes tool for oropharyngeal dysphagia in adults: III. Documentation of reliability and validity. Dy sphagia 17(2):97 114. Mioche L. 2004. Mastication and food texture perception: variation with age. Journal of Texture Studies 35(2):145 58. Mojet J, Christ Hazelhof E, Heidema J. 2001. Taste perception with age: generic or specific losses in threshold sens itivity to the five basic tastes? Chemical Senses 26:845 60. Mojet J, Christ Hazelhof E, Heidema J. 2005. Taste perception with age: pleasantness and its relationships with threshold sensitivity and supra threshold intensity of five taste qualities. Food Q uality and Preference 16(5):413 23. Murphy C. 1985. Cognitive and chemosensory influences on age related changes in the ability to identify blended foods. Journals of Gerontology 40(1):47 52. Murphy C. 1993. Nutrition and chemosensory perception in the eld erly. Critical Reviews in Food Science and Nutrition 33(1):3 15. Philipsen DH, Clydesdale FM, Griffin RW, Stern P. 1995. Consumer age affects response to sensory characteristics of a cherry flavored beverage. Journal of Food Science 60(2):364 8. Roy N, Ste mple J, Merrill RM, Thomas L. 2007. Dysphagia in the elderly: preliminary evidence of prevalence, risk factors, and socioemotional effects. Annals of Otology Rhinology and Laryngology 116(11):858 65. Schiffman S. 1977. Food recognition by elderly. Journals of Gerontology 32(5):586 92. Schiffman S. 1993. Perception of taste and smell in elderly persons. Critical Reviews in Food Science and Nutrition 33(1):17 26. Schiffman SS, Warwick ZS. 1989. Use of flavor amplified foods to improve nutritional status in el derly persons. Annals of the New York Academy of Sciences 561:267 76. Schiffman SS, Warwick ZS. 1993. Effect of flavor enhancement of foods for the elderly on nutritional status, food intake, biochemical indexes, and anthropometric measures. Physiology & B ehavior 53(2):395 402.
105 Scott K. 2005. Taste recognition: food for thought. Neuron 48(3):455 64. Sims CA. 2013. Personal Interview ed: University of Florida. Stahlman LB, Garcia JM, Chambers E, Smit AB, Hoag L, Chambers DH. 2001. Perceptual ratings for pure ed and molded peaches for individuals with and without impaired swallowing. Dysphagia 16(4):254 62. Stahlman LB, Garcia JM, Hakel M, Chambers E. 2000. Comparison ratings of pureed versus molded fruits: preliminary results. Dysphagia 15(1):2 5. Taylor AJ, R oozen JP. 1996. Volatile flavor release from foods during eating. Critical Reviews in Food Science and Nutrition 36(8):765 84. U.S. Census Burea u : Interim Projections by Age, Sex, Race, and Hispani c Origin [Internet] Washington, DC: U.S. Census Bureau ; c2004 [Accessed 2013 Mar 11] Available from: htt p:// www.census gov Ventanas S, Mustonen S, Puolanne E, Tuorila H. 2010. Odour and flavour perception in flavoured model systems: influence of sodium chloride, umami compounds and serving temperature. Food Quality and Preference 21:453 62. Wysocki CJ, Pelchat ML. 1993. The effects of aging on the human sense of smell and its relationship to food choice. Critical Reviews in Food Science and Nutrition 33 (1):63 82.
106 BIOGRAPHICAL SKETCH Jamila Ren Lepore was born in Dunedin, Florida and is the daughter to David and Jane Frazier. She has two sisters and one brother, whom she loves dearly. She started her undergraduate college career at New York Univers ity before she decided to pursue a degree in dietetics at the University of Florida, where she graduated in 2010. The following year, she completed a dietetic internship and passed the national registration exam to officially become a Registered Dietitian. Jamila continued her educational studies by returning to the University of Florida to complete a Master of Science degree in food science in two years. In her second ye ar, she married Dr. Ryan Lepore DMD a graduate of the University of Florida. She is pa ssionate about nutrition and food science and would like to pursue a career that incorporates either or both of these specialties.