<%BANNER%>

Alterations in Metabolism During Ketogenic Therapy for Seizures

Permanent Link: http://ufdc.ufl.edu/UFE0024816/00001

Material Information

Title: Alterations in Metabolism During Ketogenic Therapy for Seizures
Physical Description: 1 online resource (211 p.)
Language: english
Creator: Jones, Lauren
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: brain, carnitine, diet, dyslipidemia, energy, epilepsy, growth, ketogenic, ketones, lipid, metabolism, metabolomics, nutrition, pediatrics, seizures, therapy
Food Science and Human Nutrition -- Dissertations, Academic -- UF
Genre: Food Science and Human Nutrition thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Ketogenic Therapy (KT) is a high fat, adequate protein, low carbohydrate diet used in the treatment of intractable epilepsy. Alterations in metabolism during KT result in seizure improvement. The literature is contradictory concerning possible adverse effects of KT on blood lipids and growth. During therapy, orally-fed patients usually receive a diet high in saturated fatty acids and tube-fed patients usually receive a diet high in omega 6 fatty acids. Preliminary analysis evaluating the incorporation of a balanced fatty acid profile found there were significant positive effects on blood lipids. A high fat diet could also alter the acylcarnitine profile due to carnitine's major role in fatty acid metabolism. During starvation and fat loading, conditions similar to KT, the proportion of carnitine that is acetylated significantly increases. Additionally, acetyl-l-carnitine (ALC) has been shown to be neuroprotective. Monitoring of the acylcarnitine profile, specifically ALC, in patients on KT may be important in understanding mechanism of action. Retrospective analysis was performed on the KT population at Shands to evaluate changes in growth and lipids. A prospective study was conducted in the General Clinical Research Center to more closely evaluate growth and blood lipids and to monitor changes in the acylcarnitine and metabolomic profile. A meal challenge test was performed to test changes in metabolites after ingestion of a keto meal. Results indicate type of fat to be more important than amount of fat for increasing the risk of dyslipidemia in patients receiving KT. Children treated with the KT show significant changes in height over time and ambulation status is important in growth and energy needs. Overall, beta-hydroxybutyrate and glucose are maintained relatively constant. Stabilization of energy metabolism and an increase in ALC may be important in mechanism of action against seizures. An intervention with a balanced fat blend to improve dyslipidemia, indirect calorimetry for optimization of calorie needs to improve growth, and ALC supplementation to enhance alterations in metabolism induced by KT was designed based on these data in order to improve the health and/or efficacy of therapy. A system for use of metabolomics in this population to further understand mechanism was developed.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Lauren Jones.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Borum, Peggy R.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024816:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024816/00001

Material Information

Title: Alterations in Metabolism During Ketogenic Therapy for Seizures
Physical Description: 1 online resource (211 p.)
Language: english
Creator: Jones, Lauren
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: brain, carnitine, diet, dyslipidemia, energy, epilepsy, growth, ketogenic, ketones, lipid, metabolism, metabolomics, nutrition, pediatrics, seizures, therapy
Food Science and Human Nutrition -- Dissertations, Academic -- UF
Genre: Food Science and Human Nutrition thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Ketogenic Therapy (KT) is a high fat, adequate protein, low carbohydrate diet used in the treatment of intractable epilepsy. Alterations in metabolism during KT result in seizure improvement. The literature is contradictory concerning possible adverse effects of KT on blood lipids and growth. During therapy, orally-fed patients usually receive a diet high in saturated fatty acids and tube-fed patients usually receive a diet high in omega 6 fatty acids. Preliminary analysis evaluating the incorporation of a balanced fatty acid profile found there were significant positive effects on blood lipids. A high fat diet could also alter the acylcarnitine profile due to carnitine's major role in fatty acid metabolism. During starvation and fat loading, conditions similar to KT, the proportion of carnitine that is acetylated significantly increases. Additionally, acetyl-l-carnitine (ALC) has been shown to be neuroprotective. Monitoring of the acylcarnitine profile, specifically ALC, in patients on KT may be important in understanding mechanism of action. Retrospective analysis was performed on the KT population at Shands to evaluate changes in growth and lipids. A prospective study was conducted in the General Clinical Research Center to more closely evaluate growth and blood lipids and to monitor changes in the acylcarnitine and metabolomic profile. A meal challenge test was performed to test changes in metabolites after ingestion of a keto meal. Results indicate type of fat to be more important than amount of fat for increasing the risk of dyslipidemia in patients receiving KT. Children treated with the KT show significant changes in height over time and ambulation status is important in growth and energy needs. Overall, beta-hydroxybutyrate and glucose are maintained relatively constant. Stabilization of energy metabolism and an increase in ALC may be important in mechanism of action against seizures. An intervention with a balanced fat blend to improve dyslipidemia, indirect calorimetry for optimization of calorie needs to improve growth, and ALC supplementation to enhance alterations in metabolism induced by KT was designed based on these data in order to improve the health and/or efficacy of therapy. A system for use of metabolomics in this population to further understand mechanism was developed.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Lauren Jones.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Borum, Peggy R.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024816:00001


This item has the following downloads:


Full Text

PAGE 1

1 ALTERATIONS IN METABOLISM DURING KETOGENIC THERAPY FOR SEIZURES By LAUREN LITTLE JONES A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGRE E OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Lauren Little Jones

PAGE 3

3 To my Mom and Da d, Dennis and Mary Julia Little

PAGE 4

4 ACKNOWLEDGMENTS First of all I would like to thank God for His constant love and countless blessings H e has given me I would like to sincerely thank my supervisory committee members Peggy R. Borum, PhD, Gail P.A. Kauwell, PhD, RD, LDN, Hord u r G. Kristinsson, PhD, and Paul R. Carney, MD for the unique contributions and valuable expertise that each of them provided to my project. Also I want to extend a huge thank you to both Natasha Singh and Claudia Ferri for their valuable input into the project. In particular, I would like to thank my committee chair, major advisor and ment or Dr. Peggy Borum, for her wisdom, encouragement, and guidance she gave me throughout my undergraduate and graduate experience. She is an extraordinary mentor and role model and has been an inspiration to me every day for 6 years. Without her expertise, s upport and confidence in me, this work would not have been possible. I am very grateful for the knowledge gained about both life and science from working with her and it will continue with me through all my endeavors. I would like to give a huge thank you to all of the InvestiGators for their valuable time and effort they put in to this research project, the KetoGators for their commitment to this patient population and for their support through the years. Additionally, I would like to thank the Carnitine team, specifically Jessica Parker and Kate Goodmon for their help in preparing samples and work on the metabolomics analyses, Dave McDonald for his time and effort on the literature review for acylcarnitines, and Soledad Cerutti in Chemistry for her hard work on the acylcarnitine isolation and analysis. My sincere appreciation goes to all of the clinical staff with the Ketogenic Therapy program at Shands and the GCRC staff for the valuable contributions to the patients and to our study. Lastly and most impor tantly I want to thank my family and friends for always encouraging me to fulfill this dream. Most of all, I would like to thank my parents, Dennis and Mary Julia

PAGE 5

5 Little for their lifelong commitment to my education. They have provided constant love, support, and guidance since day one. They have been my personal cheering squad and encouragement throughout my life. Finally I would like to thank my husband, Lance for his love and patience. He reminded me everyday how proud he was of my hard work and accompl ishments and his love and positive attitude gave me the strength to complete this program.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................................... 4 LIST OF TABLES .............................................................................................................................. 10 LIST OF FIGURES ............................................................................................................................ 13 ABSTRACT ........................................................................................................................................ 18 CHAPTER 1 INTRODUCTION ....................................................................................................................... 20 Research Hypotheses .................................................................................................................. 22 Significance ................................................................................................................................. 22 2 BACKGROUND AND LITERATURE REVIEW ................................................................... 25 Epilepsy and Ketogenic Therapy ............................................................................................... 25 Brain Energy Metabolism and Epilepsy ............................................................................. 25 Energy Metabolism on Ketogenic Therapy ....................................................................... 28 Growth ......................................................................................................................................... 33 Compounding Factors ......................................................................................................... 33 Calorie restriction ......................................................................................................... 34 Cerebral palsy ............................................................................................................... 34 Epilepsy ........................................................................................................................ 35 AEDs ............................................................................................................................. 35 Intractable Epilepsy ............................................................................................................. 36 Growth and KT .................................................................................................................... 37 Dysl ipidemia ................................................................................................................................ 41 Type of Fat and Efficacy of KT .......................................................................................... 44 Effect of KT on Blood Lipids ............................................................................................. 46 Effect of Modifying Fatty Acid Profile on Blood Lipids .................................................. 47 Preliminary Studies .............................................................................................................. 49 Evaluation of ketogenic eggnog an d KetoGatorNog treatment for seizures ............ 49 Emulsion project .......................................................................................................... 53 Acylcarnitine Profile ................................................................................................................... 54 Introduction .......................................................................................................................... 54 Acylcarnitine Profile and Clinical Applications ................................................................ 56 Changes in Carnitine and Acylcarnitine Levels ................................................................ 59 Carnitine and Acylcarnitines in Brain ................................................................................ 60 Acetyl L -carnitine (ALC) Role in Brain ............................................................................ 62 Energy metabolism and membranes ........................................................................... 62 Anti -oxidant and anti apoptotic functions .................................................................. 65

PAGE 7

7 Neuromodulatory ......................................................................................................... 66 Clinical applications ..................................................................................................... 67 Conclusions and Future Directions ..................................................................................... 70 M etabolomics .............................................................................................................................. 72 Methodology ........................................................................................................................ 73 Application in Nutritional Science ..................................................................................... 75 Prevention of disease ................................................................................................... 77 Treatment of disease .................................................................................................... 77 Use of Metabolomics in the KT Population....................................................................... 78 Untargeted metabolomics ............................................................................................ 78 Targeted metabolomics ................................................................................................ 79 3 METHODS .................................................................................................................................. 84 Retrospective Study .................................................................................................................... 84 Study Design ........................................................................................................................ 84 Calculations .......................................................................................................................... 85 Statistics ................................................................................................................................ 86 Limitations and Strengths .................................................................................................... 86 Prospective Study ........................................................................................................................ 86 Study Design ........................................................................................................................ 86 Measurements ...................................................................................................................... 87 Calculations .......................................................................................................................... 88 Blood Analysis ..................................................................................................................... 89 Statistics ................................................................................................................................ 89 Limitations and Strengths .................................................................................................... 89 Metabolomics .............................................................................................................................. 89 KetoMed Study Design ....................................................................................................... 91 Blood Analysis ..................................................................................................................... 92 Sample preparation ....................................................................................................... 92 Analytical methods ....................................................................................................... 93 Data processing ............................................................................................................ 94 Compound identification ............................................................................................. 96 Limitations and Strengths .................................................................................................... 97 4 RESULTS .................................................................................................................................... 98 Anthropometrics and Growth ..................................................................................................... 99 Retrospective Results .......................................................................................................... 99 Time on diet .................................................................................................................. 99 Age .............................................................................................................................. 105 Ambulation status ....................................................................................................... 107 Diet prescription ......................................................................................................... 107 Route of feeding ......................................................................................................... 110 -hydroxybutyrate (BHB) .......................................................................................... 111 Summary of results .................................................................................................... 111 Prospective Results ............................................................................................................ 112

PAGE 8

8 Experienced patients .................................................................................................. 114 Diet prescription ......................................................................................................... 114 Ambulation status ....................................................................................................... 119 Nave patients ............................................................................................................. 120 -hydroxybutyrate (BHB) .......................................................................................... 121 Summary of results .................................................................................................... 122 Blood Lipids and Dyslipidemia ................................................................................................ 125 Retrospective Results ........................................................................................................ 125 Time on diet ................................................................................................................ 125 Age and gender ........................................................................................................... 126 Dyslipidemia ............................................................................................................... 129 Diet prescription ......................................................................................................... 129 Summary of results .................................................................................................... 132 Prospective Results ............................................................................................................ 133 Dyslipidemia ............................................................................................................... 133 Time on diet ................................................................................................................ 135 Diet prescription ......................................................................................................... 1 36 Nave patients ............................................................................................................. 137 Summary of results .................................................................................................... 141 Meal Challenge .......................................................................................................................... 142 Prospective Results ............................................................................................................ 142 Demographics ............................................................................................................. 142 Time on diet ................................................................................................................ 143 Age .............................................................................................................................. 144 Route of feeding ......................................................................................................... 145 Diet Prescr iption ................................................................................................................ 145 Comparison of data .................................................................................................... 146 Summary of results .................................................................................................... 148 Metabol omics ............................................................................................................................ 149 Preliminary Results ............................................................................................................ 149 Untarged metabolomics ............................................................................................. 149 Targ eted metabolomics .............................................................................................. 153 Additional Results ............................................................................................................. 154 Summary of results .................................................................................................... 155 5 DISCUSSION AND CONCLUSIONS ................................................................................... 158 Conclusions ............................................................................................................................... 158 Growth ................................................................................................................................ 158 Lipids .................................................................................................................................. 160 Meal Challenge .................................................................................................................. 162 Metabolomics ..................................................................................................................... 163 Implications for Future Research ............................................................................................. 164 Design of future studies ..................................................................................................... 165 Research p lan and m ethods ............................................................................................... 166

PAGE 9

9 APPENDIX A ALC AND KT ........................................................................................................................... 168 B BLOOD ASSAY ....................................................................................................................... 172 C METABOANALYST TABLES AND FIGURES .................................................................. 177 LITERATURE CITED ..................................................................................................................... 187 BIOGRAPHICAL SKETCH ........................................................................................................... 211

PAGE 10

10 LIST OF TABLES Table page 2 1 Effects of insulin and glucagon in key enzymes in control of ketogenesis ........................ 29 2 2 Comparison of study results for a diet prescribing adequate calories for growth and a calorie restricted diet of 75% of total energy expenditure ................................................... 39 2 3 Growth and IGF 1 values (mean SD) before and after starting KT ................................ 40 2 4 Dietary interventions and improvement in patients on KT with hypercholesterolemia .... 48 2 5 Micronutrient adequacy for PO Ketogenic Eggnog (KE PO) ............................................. 51 2 6 Fatty acid profile of KGN -PO and KGN TF compared to KE PO and KE TF ................. 51 2 7 Comparison of blood lipid data for KE versus KGN ........................................................... 53 2 8 Inborn errors of metabolism detected by acylcarnitine profile analysis: disorders of fatty acid oxidation and organic acid metabolism. ............................................................... 81 2 9 Changes in plasma carnitine and acylcarnitine concentrations under dietary manipulations and alterations in metabolism. ...................................................................... 60 2 10 Summary of ALC neuroprotective mechanisms. ................................................................. 83 3 1 Tolerance ranges used for searching the Human Metabolome Database. .......................... 97 4 1 Comparison of height, weight, and BMI z scores for ambulatory versus nonambulat ory patients ov erall and at 0 6 months on diet ...................................................... 107 4 2 Comparison of ratio groups for height, weight and BMI z score ...................................... 108 4 3 Com parison of oral ly -fed versus tube -fed patients ............................................................ 110 4 4 Demographics at start of prospective study for experienced and nave patients .............. 113 4 5 Comparison of ratio groups for height, weight and BMI z score for experienced patients evaluated prospectively.......................................................................................... 118 4 6 Comparison of height, weight, and BMI z scores for ambulator y versus non ambulatory experienced patients evaluated prospectively ................................................ 119 4 7 Comparison of height, weight, and BMI z scores for nave pati ents before and after diet and comparison of growth ba sed on ambulation status .............................................. 121 4 8 Comparison of mean blood lipids for males versus females ............................................. 126

PAGE 11

11 4 9 Patients in KT population pres enting with dyslipidemia two or more times ................... 129 4 10 Comparison of mean blood lipids for KetoCal patients and the entire KT population ... 130 4 11 Comparison of mean blood lipids for patients on different KT ratios .............................. 131 4 12 Comparison of mean blood lipids for orally -fed and tube -fed patients on KT ................ 132 4 13 Patients in KT population presenting with dyslipidemia in the prospective study .......... 133 4 14 Comparison of dietary parameters between patients with abnormal TG and normal TG in prospective study ....................................................................................................... 133 4 15 Comparison of dietary parameters between patients with abnormal HDL and normal HDL in prospective study .................................................................................................... 134 4 16 Comparison of dietary parameters between patients with abnormal LDL and normal LDL in prospective study .................................................................................................... 134 4 17 Comparison of dietary parameters betwe en patients with abnormal Total Chol and normal Total Chol in prospective study .............................................................................. 135 4 18 Comparisons between orally-fed and tube -fed patients on KT in prospective study ...... 136 4 19 Comparison of mean blood lipids for patients on different KT ratios in prospective study ...................................................................................................................................... 137 4 20 Regression analyses for blood lipids and rat io and BHB evaluated prospectively .......... 137 4 21 Percent change in blood lipids in nave patie nts on KT and age in years ........................ 138 4 22 Mea n BHB and Glu in mg/dL at fasting and after ingestion of a ketogenic meal ........... 143 4 23 Mean percent changes of BHB and Glu after a ketogenic meal ....................................... 143 4 24 Effect of time after meal and age on BHB/Glu .................................................................. 144 4 25 Effect of time after a meal and route of feeding on BHB/Glu .......................................... 145 4 26 Effect of age and route of feeding on percent change of BHB and Glu ........................... 145 4 27 Number of significant masses identified by univariate analyses in Metaboanalyst ......... 151 4 28 Provisional acylcarnitines identified by searching significant compounds identified by fold change, t test, and volcano plot analyses ............................................................... 152 4 29 Provisional acylcarnitines identified by searching significant compounds identified by both volcano and fold change analyses ......................................................................... 152

PAGE 12

12 A 1 Acetyl l -carnitine (ALC) and Ketogenic Therapy (K T) similar neuroprotective mechanisms. ......................................................................................................................... 169 C1 Summary of data processing results in MetaboAnalyst for all 60 samples, before and after for female subjects. ...................................................................................................... 177 C2 Top 50 features identified by fold change analysis of all 60 samples comparing before and after for female subjects. ................................................................................... 178 C3 Top 50 features identified by t -test analysis of all 60 samples comparing before and after for female subjects. ...................................................................................................... 179 C4 Top 50 features identified by volcano plot analysis of all 60 samples comparing before and after for female subjects. ................................................................................... 181

PAGE 13

13 LIST OF FIGURES Figure page 1 1 Diagram of overall project of alterations in metabolism on KT. ........................................ 24 2 1 Compounding factors contributing to the evaluation of growth in the KT population. .... 41 2 2 Approximate fatty acid percentages in the various KT feedi ng options ............................. 43 2 3 Approximate fatty acid percentages in the AHA recommendations and the KGN fat blend. ....................................................................................................................................... 50 2 4 Mean values for tr iglycerides and total cholesterol in mg/dL at initiation (0 months) and end of study (3 months) for patients on traditional ketogenic therapy (KE) versus the new KetoGatorNog (KGN). ............................................................................................ 52 2 5 Mea n values for HDL cholesterol, Non HDL cholesterol, and LDL cholesterol in mg/dL at initiation (0 months) and end of study (3 months) for patients on traditional ketogenic therapy (KE) versus the new KetoGatorNog (KGN). ........................................ 52 2 6 ALC and energy metabolism. Integration of ketone and ALC metabolism. ...................... 55 2 7 L -carnitine structure. .............................................................................................................. 56 2 8 Example acylcarnitine structures. ......................................................................................... 56 2 9 KT and ALC synergistic and/or additive effects on neuroprotection. ................................ 71 2 10 Metabolomics in translational nutritional sciences. ............................................................. 76 2 11 Metabolomics workflow. ....................................................................................................... 78 3 1 Workflow of multidisciplinary approac h and continuity of metabolomics based analyses ................................................................................................................................... 90 3 2 Diagram of KetoMed study design. ...................................................................................... 91 4 1 Overall project and summary of specific aims included in data analysis and results. ....... 98 4 2 Mean height z score S D over time on diet in months in patients on KT.. ..................... 100 4 3 Mean weight z score SD over time on diet in months in patients on KT. ..................... 100 4 4 Mean BMI z score SD over time on diet in months in patients on KT. ........................ 1 01 4 5 Average h eight z score per patient versus time on diet in months. ................................... 101 4 6 Average weigh t z score per patient versus time on diet in months. .................................. 102

PAGE 14

14 4 7 Average BM I z score per patient versus time on diet in months. ..................................... 102 4 8 Height z score versus age in years for patients on KT at 0 6 months on diet. ................. 103 4 9 Weight z score versus age in years for patients on KT at 0 6 months on diet. ................ 103 4 10 BMI z score versus age in years for patients on KT at 0 6 months on diet. ..................... 104 4 11 Mean heig ht z score SD over time on diet in months in patients on KT for >60 months. .................................................................................................................................. 105 4 12 Mean height z score SD in each age group at 0 6 mo and >6 mo on diet .................... 106 4 13 Mean height z score SD in each age group. .................................................................... 106 4 14 Weight z score versus Calories per kilogram body weight per day (Cal/kg/day) for patients on KT. ..................................................................................................................... 108 4 15 Height z score versus -hydroxybutyrate (BHB ) in mg/dL in patients on KT. ............... 112 4 16 Summary of anthropometrics and growth results of retrospective analyses. ................... 113 4 17 Height z score versus time on diet in months for experienced patients on KT evaluated prospectively ....................................................................................................... 114 4 18 Height z score versus age in years for experienced patients on KT evaluated prospectively ........................................................................................................................ 115 4 19 H eight z score versus Calories per kilogram body weight per day (Cal/kg/day) for experienced patients on KT evaluated prospectively ........................................................ 115 4 20 H eight z score versus % recommended Calories (%EER) ................................................ 116 4 21 BMI z score versus % recommended Calories (%EER) .................................................... 116 4 22 W eight z score versus % recommended Calories (%EER) ............................................... 117 4 23 H eight z score versus % recommended Calories (%EER) as determined by equations from DRI for energy based on age, ge nder and activity level for experienced patients on KT with % EER less than 80%. ..................................................................................... 117 4 24 H eight z score versus % recommended Calories (%EER) as determined by equations from DRI for energy bas ed on age, gender and activity level for experienced patients on KT and stratified based on ambulation status. .............................................................. 120 4 25 H eight z score over time on diet in months for individual nave patients evaluated prospectively ........................................................................................................................ 121

PAGE 15

15 4 26 H eight z score versus % recommended Calories (%EER) as determined by equations from DRI for energy based on age, gender and activity level for nave patient s on KT. ..................................................................................................................... 122 4 27 H eight z score versus -hydroxybutyr ate (BHB) in mg/dL in patients on KT evaluated prospectively ....................................................................................................... 123 4 28 H ei ght z score versus Glucose (Glu) in mg/dL in patients on KT evaluated prospectively ........................................................................................................................ 124 4 29 Summary of anthropometric data including the retrospective and prospective study results. ................................................................................................................................... 124 4 30 Mean TG SD in mg/dL over time on diet in months. ..................................................... 126 4 31 Mean HDL SD in mg/dL over time on diet in months. .................................................. 127 4 32 Mean n onHDL SD in mg/dL over time on diet in months. ............................................ 127 4 33 Mean Total Chol SD in mg/dL over time on diet in months. ........................................ 128 4 34 Mean LDL SD in mg/dL over time on diet in months. .................................................. 128 4 35 Mean Total/HDL SD in mg/dL over time on diet in months. ........................................ 129 4 36 Average blood lipid levels in mg/dL before (B) and during (D) KetoCal for 5 patients on KT. ..................................................................................................................... 131 4 37 Summary of lipid results for the retrospective study ........................................................ 132 4 38 T otal cholesterol (mg/dL) versus time on diet in month s for experienced patients on KT evaluated prospectively. ................................................................................................ 135 4 39 LDL cholesterol (mg/dL) versus time on diet in months for experienced patients on KT evaluated prospectively. ................................................................................................ 136 4 40 Percent change in LDL cholesterol in nave pat ients on KT versus age in years evaluated prospectively ....................................................................................................... 138 4 41 Percent change in LDL cholesterol in nave patients on KT ve rsus Fat in grams per kilogram body weight per day (Fat g/kg/day) evaluated prospectively ........................... 139 4 42 T riglycerides (TG) in mg/dL over time on diet in months for individual nave patients stratified by route of feeding. ................................................................................ 139 4 43 T otal cholesterol (Total Chol) in mg/dL over time on diet in months for individual nave patients stratified by route of feeding ....................................................................... 140

PAGE 16

16 4 44 LDL cholesterol (LDL Chol) in mg/dL over time on diet in months for individual nave patients stratified by route of feeding. ...................................................................... 140 4 45 HDL cholesterol (HDL Chol) in mg/dL over time on diet in months for individual nave patients s tratified by route of feeding. ...................................................................... 141 4 46 Summary of lipid results of retrospective study with addition of prospective study results. ................................................................................................................................... 142 4 47 Percent change of -hydroxybutyrate (BHB) at 30 minutes versus time in diet in months in patients on KT. .................................................................................................... 143 4 48 Fasting -hydroxybutyrate (BHB)/glucose (Glu) ve rsus age in years in patients on KT. ........................................................................................................................................ 144 4 49 Percent change of glucose (Glu) at 2 hours versus diet prescription ratio (fat: protein +carbohydrate) in patients on KT. ...................................................................................... 146 4 50 Fasting -hydroxybutyrate (BHB)/glucose (Glu) versus carbohydrate (CHO) percentage of calories in patients on KT. ........................................................................... 147 4 51 Comparison of after meal blood levels of BHB and Glu in patient KG0040 at 1 year and 2 years on study. ............................................................................................................ 147 4 52 Summary of results for mechanism of action. .................................................................... 148 4 53 Diagram of plasma samples analyzed for preliminary metabolomics analyses using MetaboAnalyst and manual searching in excel. ................................................................. 150 4 54 Manual targeted identification of acylcarnitines. ............................................................... 154 4 55 Average After diet/Before diet (A/B) ratio response of the % total peak area counts versus free carnitine (C0) and acylcarnitines in female and male samples. ..................... 156 4 56 A verage After diet/Before diet (A/B) ratio response of the % total peak area counts for each female and male sample for C2 or acetyl l -carnitine (ALC). ............................. 157 C1 Box plots and kernel density pl ots produced in MetaboAnalyst before and after normalization of all before and after samples from female subjects. ............................... 182 C2 Fold change analysis of all 60 samples comparing before and after di et with a threshold of 2. ....................................................................................................................... 183 C3 Features identified by t -tests with a threshold of 0.1 comparing before and after diet for all samples from female subjects. .................................................................................. 183

PAGE 17

17 C4 Features identified by volcano plot analysis with fold change threshold (x) 2 and t tests threshold (y) 0.1 comparing before and after diet for all samples from female subjects. ................................................................................................................................. 184 C5 Score plot between the selected principal components (PC) in all 60 samples comparing before and after diet of female subjects. .......................................................... 185 C6 Partial least squares d iscriminant analysis (PLS DA) 3D score plot between selected principal components in all 60 samples comparing before and after diet of female subjects. ................................................................................................................................. 186

PAGE 18

18 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ALTERATIONS IN METABOLISM DURING KETOGENIC THERAPY FOR SEIZURES By Lauren Little Jones August 2009 Chair: Peggy R. Borum Major: Food Science and Human Nutrition K etogenic Therapy (KT) is a high fat, adequate protein, low carbohydrate diet used in the trea tment of intractable epilepsy. Alterations in metabolism during KT result in seizure improvement. The literature is contradictory co ncerning possible adverse effects of KT on blood lipids and growth. During therapy, oral ly -fed patients usually receive a diet high in saturated fatty acids and tube -fed patients usually receive a diet high in omega 6 fatty acids Preliminary analysis eval uating the incorporation of a balanced fatty acid profile found there were significant positive effects on blood lipids. A high fat diet could also alter the acylcarnitine profile due to carnitines major role in fatty acid metabolism. During starvation an d fat loading, conditions similar to KT, the proportion of carnitine that is acetylated significantly increases Additionally, a cetyl l -carnitine (ALC) has been shown to be neuroprotective Monitoring of the acylcarnitine profile, specifically ALC, in patients on KT may be important in understanding mechanism of action. Retrospective analysis was performed on the KT population at Shands to evaluate changes in growth and lipids. A prospective study was conducted in the General Clinical Research Center to more closely evaluate growth and blood lipids and to monitor changes in the acylcarnitine and

PAGE 19

19 metabolomic profile A meal challenge test was performed to test changes in metabolites after ingestion of a keto meal Results indicate type of fat to be more imp ortant than amount of fat for increasing the risk of dyslipidemia in patients receiving KT. Children treated with the KT show significant changes in height over time and ambulation status is important in growth and energy needs Overall, beta hydroxybutyra te and glucose are maintained relatively constant. Stabilization of energy metabolism and an increase in ALC may be important in mechanism of action against seizures. An intervention with a balanced fat blend to improve dyslipidemia, indirect calorimetry f or optimization of calorie needs to improve growth, and ALC supplementation to enhance alterations in metabolism induced by KT was designed based on these data in order to improve the health and/or efficacy of therapy. A system for use of metabolomics in t his population to further understand mechanism was developed .

PAGE 20

20 CHAPTER 1 INTRODUCTION Ketogenic therapy (KT) is a treatment for intractable childhood epilepsy that involves a rigidly calculated high fat, adequate protei n, and low carbohydrate diet. Although KT dates back to the 1920s, because of the development of antiepileptic drugs (AEDs) in the 1940s and 50s, i t became decreasingly popular. In more recent years there has been a re emergence of the ketogenic diet in treating patients with seizures; especia lly when AEDs have failed and/or patients are looking for an alternative to surgic al procedures. While this diet has been administered for more than 80 years and has been shown in numerous studies to have greater than a 50% decrease in seizure activity in 6075% of children, the mechanism of how it works is unknown (1 10) KT is unique from AED therapy in that it has both anticonvulsant (stopping a discrete seizure) and antiepileptic (stopping the propensity to develop recurrent seizures or underlying epile psy) activity (11) These disease modifying effects have been shown through follow up with patients who discontinued the diet and still saw long term seizure freedom (12, 13) While the diet causes alterations in metabolism that ultimately lead to these be neficial effects, these alterations may also cause changes in growth, blood lipids, and/or acylcarnitines tha t may be important to monitor. These potential alterations in metabolism need to be defined in this patient population. Calorie restriction is oft en used as part of the diet prescription; however there is conflicting evidence as to whether this is necessary for effectiveness of the therapy (3, 4) Most patients receiving this therapy are in the pediatric age range and literature reports are inconclu sive concerning whether or not KT results in an adverse effect on growth (14, 15) If there is a problem of growth in patients on KT, the question becomes is this due to restricted calories, the diet itself, or epi lepsy in general? Some studies have shown that epilepsy itself results in delayed

PAGE 21

21 growth in patients. El Khayat et al 2004 concluded that the longer duration of disease, the more negative the impact on stature in female patients with epilepsy compared to healthy controls (16) While KT has bee n shown to be effective against seizures, it requires a fat intake (80 90% of calories) much higher than recommended by the American Heart Association (AHA). The altered macronutrient ratio that forms the basis of KT also is a concern because it has been a ssociated with the development of dyslipidemia and chronic vascular disease (17, 18) There is conflicting evidence in the literature as to whether or not dyslipidemia is an adverse effect of KT. Currently there are no available feeding options that provi de a balanced fatty acid profile similar to what is recommended by the AHA. Carnitine is an important metabolite involved in fatty acid metabolism. In addition to induction of ketosis, a very high fat diet would be expected to alter the acylcarnitine pro file including an increase in acetyl -l -c arnitine (ALC) concentrations. ALC may be providing an activated acetyl group for the use of energy and may also be implicated in other neuroprotective pathways (19). Monitoring the plasma acylcarnitine profiles may be beneficial in determining alterations in metabolism during therapy and potentially provide some insight for the mechanism of action of KT. Major alterations in macronutrient composition in the diet would be expected to alter levels of metabolites other than acylcarnitines in serum. Understanding these alterations may lead to insight into mechanisms of action. Metabolomics is a global analysis of all small molecular weight compounds (<1500 Da) in a biological system under certain conditions (20). Metabol omics based analyses have not been published in the literature for this patient population. Due to the complexity and size of the data sets, processing and interpretation of data

PAGE 22

22 can be challenging. A system for analyzing untargeted metabolomics was developed to identify major changes in compounds in healthy subjects receiving a ketogenic diet. In the future, comparisons will be made between the metabolic changes in these health y subjects and patients receiving KT for seizures. Research Hypotheses Retrospe ctive and prospective analyses were implemented to assess alterations in metabolism of patients on KT with the following hypotheses: Potential adverse effects: The ketogenic therapy patient population at Shands at UF has an altered blood lipid profile cons istent with increased risk for atherosclerosis compared to pre therapy values or compared to a reference population. The ketogenic therapy patient population at Shands at UF has impaired growth compared to pre therapy values or compared to a reference po pulation. Mechanism of action: hydroxybutyrate after the ingestion of a ketogenic meal. Preliminary metabolomics analyses show ingestion of a ketogenic diet results in an altered acylcarnitine profile and alterations in the metabolome compared to pre diet values in healthy subjects. Significance Retrospectively and prospectively, blood lipids, -hydroxybutyrate (BHB) diet prescription and growth parameters were analy zed to evaluate the potential adverse effects of growth and dyslipidemia in this patient population. We also analyzed the change in BHB and glucose at 30 min, 1 hr, and 2 hr after the ingestion of a ketogenic meal once per year. These data were used to det ermine correlations and trends between parameters to help in defining the alteration s in metabolism in the KT population. At the end of the prospective study, an interim

PAGE 23

23 analysis was conducted and an intervention trial was designed to improve the therapy a nd further evaluate patients on KT for seizures In addition, a very high fat diet could alter the acylcarnitine profile including an increase in ALC concentrations; therefore, plasma from healthy subjects before and after ingestion of a ketogenic diet wa s evaluated for changes in the acylcarnitine profile and metabolome to lend insight into metabolic changes associated with KT. Ult imately these changes will be compared to samples obtained from patients receiving KT in order to provide insight into mechani sm of action. A summary of the overall project is found in Figure 1 1. Currently there are controversies in the literature as to whether or not growth and dyslipidemia are adverse effects of the ketoge nic diet. Clinical decisions with regard to changing th e diet prescription are often based on experience and on clinical discussion and often are not based on evidence. More research needs to be conducted in order to define these potential adverse effects in this population and to determine if they are necessa ry for efficacy. Finally, a metabolomics based approach to understand ing both changes in the acylcarnitine profile as well as the global plasma metabolome in response to a ketogenic diet has not yet been published. Ultimately understanding the alterations in metabolism associated with this therapy can provide knowledge for improvement in the treatment of seizures.

PAGE 24

24 Figure 1 1. Diagram of overall project of alterations in metabolism on KT.

PAGE 25

25 CHAPTER 2 BACKGROUND AND LITER ATURE REVIEW Epilepsy and Ketoge nic Therapy Epilepsy is a group of disorders involving recurrent seizures and its prevalence in the USA is estimated at 2.5 million or approximately 1% of the population (21). Roughly 1/3 of these patients have intractable or drug -resistant epilepsy (22) Seizures are defined as the clinical manifestation of abnormal excessive neuronal activity in the brain (23) Anti -epileptic medications (AEDs) either do not stop the seizures or result in unacceptable side effects in some patients. Additionally, som e pati ents are not candidates for neurosurgical intervention or the intervention has not been succe ssful. Ketogenic therapy (KT) is a treatment for intractable epilepsy that involves a rigidly calculated high fat, adequate protein, and very low carbohydrate diet Although KT dates back to the 1920s, because of the development of AEDs in the 1940s and 50s, i t became decreasingly popular. In more recent years there has been a re -emergence of the ketogenic diet in treating patients with seizures; especially when AED s have failed and/or patients are looking for an alternative to surgical procedures. While this diet has been administered for more than 80 years and has been shown in numerous studies to have greater than a 50% decrease in seizure activity in 6075% of ch ildren, the mechanism of how it works is unknown (1, 7 10, 24, 25) Brain E nergy M etabolism and Epilepsy The brain derives almost all of its energy from glucose oxidation under normal physiological conditions. Glucose is transported across the blood-brain barrier by facilitated diffusion through glucose transporters (particularly GLUT1). The majority of glucose is then metabolized to pyruvate which enters the mitochondria and is converted to acetyl CoA before entering the TCA cycle in both neurons and glia cells. Approximately 13% of pyruvate from

PAGE 26

26 glycolysis is converted to lactate in the normal brain (26) Energy homeostasis is maintained through glycolysis, the TCA cycle, electron transport chain and oxidative phosphorylation, similar to other tissues (26) However, due to metabolic and cellular compartmentalization involving neurotransmitter synthesis as well as different but critical roles of neurons and glia, energy metabolism in brain is more complex (27 30) During a seizure, glucose uptake and metabol ism increase (31, 32) and blood glucose positively correlates with flurothyl induced seizures in rats (33) Seizures cause a greater increase in cerebral metabolic rate in the area involved in the seizure propagation than under any other condition (31) I n addition, lactate significantly increases during seizure activity (31, 32, 34) This increase in lactate is due to a rapid increase in glycolytic rate compared to the rate of oxygen utilizing metabolism, with pyruvate dehydrogenase (PDH) being rate limiting (31) PDH is unable to metabolize pyruvate to acetyl -CoA as quickly as pyruvate is being produced, therefore pyruvate is converted to lactate through lactate dehydrogenase during periods of rapid glucose utilization and oxygen deprivation, both occurri ng in epileptic seizures. It is clear from the literature that neuronal excitability and epileptic seizures are correlated to rapid glucose utilization and glycolysis (26, 31, 33, 3538) While there is an increase in oxygen and glucose supply to the brain, this increase in demand of energy also results in a decrease in energy metabolites such as, glycogen, glucose, phosphocreatine, and ATP (31) In addition, epileptic foci in brain are associated with both hyper and hypometabolism. During an ictal, or se izure event, hypermetabolism has been detected in brain epileptic foci (39, 40) and hypometabolism during interictal, or in between seizures (31, 3941) suggesting altered brain energy homeostasis in epilepsy. Cornford et al have demonstrated that this decrease in metabolic rate of epileptic tissue may be due to the decreased expression of GLUT1 in the blood

PAGE 27

27 brain barrier (BBB) and decreased uptake of glucose (42, 43) Additionally, Fink et al. 1996 found a partial loss of the hexokinase reaction involved in the maintenance of glucose from blood to the brain using 18F deoxyglucose kinetics, which may also account for the hypometabolism observed in interictal tissue (44) In patients with temporal lobe epilepsy, the interictal hypometabolism was found to be due to impaired oxidative metabolism (41) The excitability of neurons is controlled by a complex interaction of excitatory and inhibitory potentials that are primarily regulated by voltage dependent activation of Na+ and Ca2+ channels. In addition, the c ontrol of extracellular potassium plays a major role in the regulation of neuronal firing (31). It appears that metabolic demand is increased in epileptic brain tissue in order to manage leakage of extracellular potassium across the BBB caused by abnormal neuronal firing and loss of homeostasis by voltage -dependent channels Cerebral blood flow does not match the increase in metabolic demand therefore an energy deficit results (40) The combination of altered ion homeostasis and an increase in metabolic dem and of hyperactive neurons may exacerbate the loss of glucose uptake by the BBB observed in epilepsy (40) Glucose uptake into the brain is unchanged during hypoglycemia due to starvation, while ketones in blood are increased and a rapid increase in BHB uptake are observed in rats (45) An inverse association between ketone concentration and glucose utilization has been observed (46) In conditions of starvation, or high fat feeding such as during KT, the increased transport of BHB at the BBB and the ava ilability as an alternate energy source for seizure foci in the brain may help under conditions of metabolic demand. In addition, a ketogenic diet resulted in an eightfold increase in the activity of the mono carboxylic transporter (MCT 1) and an increase i n GLUT1 in brain endothelial cells, indicating that KT may also promote ketone and glucose uptake (47) As a result, the energy deficit in epilepsy may be partially compensated by the alterations in

PAGE 28

28 metabolism observed during KT. Greene et al (2003) hypoth esizes that the disruption in brain energy homeostasis in epilepsy may be managed through metabolic control theory, in which moderate shifts in glucose and ketones can alter energy through glycolysis and the TCA cycle (48) Ultimately these shifts in metab olism result in adjustments in gene linked networks that manage or control epilepsy disorders (48) However, it is important to note that due to the complexity of epilepsy -related pathologies, it is difficult to establish a clear relationship between pheno typic abnormalities and progression toward abnormal neuronal ex citability because of the wide variation in seizure types and syndromes. Energy M etabolism on Ketogenic Therapy KT is based on a ratio of fat: protein + carbohydrate. Patients on KT generally a re initiated on a ratio of 3:1 to 4:1, in which approximately 8090% of their calori es are calculated as fat. The therapy is designed to mimic the physiological effects of fasting without starvation in which three ketone bodies are produced (acetoacetate, BHB, and acetone) and there is a radical shift in energy metabolism from primarily utilization of carbohydrate to fat. During a very high rate of fatty acid oxidation as in fasting or ingestion of a very high fat diet, large amounts of acetyl CoA are produ ced. Once levels exceed the capacity of the TCA cycle, ketones are produced by the liver from the excess acetyl -CoA. Ketones are elevated in the p lasma, wit h acetoacetate and BHB increasing approximately three -fold t o four -fold from basal levels (49) Both BHB and acetoacetate can be transported to the brain through the blood -brain barrier MCT 1 transporter converted to acetyl CoA and used as an alternat e energy source for the brain. BHB is the predominant ketone body in blood and is oxidized to acetoaceta te in mitochondria before entering the TCA cycle (50) During starvation, or KT, circulating levels of glucose are reduced and levels of ketones are increased (46, 51) The transport of BHB and acetoacetate into brain is regulated partly by

PAGE 29

29 circulating l evels of ketones and glucose (47) These changes are a result of an altered hormonal pattern consistent with decreased plasma insulin and increased glucagon (51, 52) This decrease in the plasma insulin/glucagon ratio leads to hepatic glycogen degradation, gluconeogenesis, oxidation and ketogenesis (53 55) These hormonal effects on enzyme activity are summarized in Table 2 1. Table 2 1. Effects of insulin and glucagon in key enzymes in control of ketogenesis Enzyme Location Action Result Effect of insulin ( ketogen esis) Effect of glucagon ( in ketogenesis) Hormone sensitive lipase Peripheral adipocytes Metabolize TG to FA Elevated serum FA inhibited stimulated Acetyl CoA carboxylase Hepatocytes Converts acetyl CoA to malonyl CoA Malonyl CoA inhibits FA transport into mitochondria stimulated inhibited HMG CoA synthase Hepatic mitochondria Converts acetoacetyl CoA to acetoacetate Rate limiting step in synthesis of ketones inhibited stimulated Adapted from Laffell et al 1999 (55) Ketogenesis is mediated by two e nzymes, mitochondrial HMG CoA synthase (mHS) and HMG CoA lyase (HL). The reaction involving mHS is the major regulatory site and the rate of the reaction depends on substrate availability and regulation of mHS mRNA and protein levels (56, 57). In fasting o r fat feeding in rats, activity of mHS is approximately doubled (56). Administration of glucagon or insulin can mimic conditions of KT and also alter mRNA levels (57 59). Additionally, hormonal changes affect transcription factors that control lipid metabolism and ketogenesis such as Foxa2 (60). Studies suggest that childrens ability to extract ketones is approxim ately 4 5 times greater than adults or adolescents (51, 61) The rate of production and utilization of glucose in infants

PAGE 30

30 and children after fa sting is almost 3 times that observed in overnight -fasted adults (51) Children have a higher brain -to -body ratio and lower glycogen stores causing an increased rate of glucose utilization per kilogram of weight, which accounts for the accelerated changes in plasma keton e and glucose during fasting compared to adults (62) Mechanisms of action regarding energy metabolism : A lterations in brain energy metabolism during KT result in alterations in the electrical properties of the brain causing decreased seizur es, or an increase in seizur e threshold in animal models. However, an elevation in ketones alone is unable to account fo r the mechanism of KT. There have been no direct associations found between seizure control and ketonemia (63 66) and ketones have not been found to directly modulate synaptic transmission (67) On the other hand, seizure response has been found to be correlated with reductions in blood glucose in calorie restricted mice (68) Some of the theories of mechanism of action of KT involve ener gy metabolism such as: limited glucose, diminished glycolysis, increase metabolic efficiency, and an increase in energy reserves. In animal models, calorie restriction alone has slowed seizure susceptibility (48, 68) in which blood glucose correlated with risk of seizures. Similarly, both KT and calorie restriction result in a moderate reduction in blood glucose concentrations. It is hypothesized that calorie restriction reduces energy production through glycolysis, which limits the ability of neurons to re ach high levels of synaptic activity necessary for the generation of seizures (48) There have also been reductions in glycolytic processes as sociated with treatment of KT. The concentration of fructose 1, 6 -b i sphosphate, the key regulatory enzyme of glycol ysis, is reduced after a ketogenic diet (46, 69, 70) In addition, glycolysis may be further diminished through elevations in ATP (46, 7173) and citrate (74, 75) observed on KT through feedback inhibition mechanisms.

PAGE 31

31 Other support for the theory of reduc ed glucose metabolism through glycoly sis stems from studies using 2 -deoxyglucose (2 DG), a glucose analogue that inhibits phosphoglucose isomerase and therefore inhibits glycolysis. Addition of 1mM 2 DG decreased epileptiform burst frequency in rat hippoca mpal slices exposed to extracellular potassium by 25 80% of baseline (76) In summary, both 2DG and KT elevate electrographic seizure threshold in vivo; also, both 2 DG and KT strongly reduce the progression of epileptogenesis in in vivo models of epileps y; and, finally, both 2 DG (in vitro) and KT (in vivo) have reduced measures of hippocampal hyperexcitability (76 78) Clinical experience has shown that cheating, or ingestion of carbohydrates can result in a rapid rise in blood glucose levels, a shift in the metabolism to glycolysis, and an increase in break through seizure activity (1, 61, 79, 80) In addition, fasting -induced seizure control has been associated with reduced blood glucose and increased blood ketones and when the fast was broken through f ood or glucose intake, seizure protection was lost in association with a rise in glucose and drop i n ketone levels (81) Similar l y, Huttenlocher PR (1976) showed that ketonemia and seizure control were rapidly reversed by intravenous infusion of g lucose (6 1) During experimental starvation in humans, consumption of relatively small amounts of glucose per day greatly reduced ketone body pr oduction (82) Therefore, maintaining low glucose levels appears to be important for seizure control in which the break i n metabolism resulting in seizures appears to be rapid, whereas the switch in metabolism to the use of ketone bodies appears to be more gradual. Clinically, researchers have tested an alternative to traditional KT in which a low glycemic -index diet was de signed to allow more liberal carbohydrate intake for incre ased tolerance to the therapy. This therapy requires restrictions in the type of food that is consumed in

PAGE 32

32 which only food that causes relatively small increa ses in blood glucose are used. One study found 50% of patients treated with this alternative therapy had a greater than 90% reduction in seizures (83) indicating that the low blood glucose may be important for mechanism of action against seizures. In addition to glucose metabolism, overall alte rations in energy metabolism and efficiency have been observed during KT. Studies using microarray expression demonstrated that a ketogenic diet induced a coordinated up regulation of several dozen metabolic genes associated with oxidative phosphorylation (73, 84) KT was found to stimulate mitochondrial biogenesis resulting in a 46% increase in the number of mitochondria in the hilus /dentate gyrus region of rat hippocampus (73) Several levels of energy metabolites are increased after KT, specifically glyc ogen and ATP concentrations were increased throughout the rodent brain (46) Also, the phosphcreatine:creatine energy -reserve ratio was found to be increased in both animals (73) and humans (72) Not only an increase in energy metabolites, but an increase in metabolic efficiency shown through a decreased respiratory quotient and maximal mitochondrial respiratory rate in rodents following the consumption of a ketogenic diet was observed. Similarly, ketones were suggested to be a more efficient source of ener gy per unit oxygen than glucose (85) Addition of either ketones or insulin markedly improved the energy status of the working perfused heart. It led to a 25% increase in hydraulic work with a significant decrease in oxygen consumption, increased free cyto solic ATP/ADP and phosphocreatine/creatine ratio, a 16-fold rise in acetyl CoA and increases in TCA cycle intermediates which were induced by not only changes in glycolysis but also changes in mitochondrial ATP production through the respiratory electron transport chain (85 88) Ultimately, the increase in metabolites involved in oxidative phosphorylation and energy reserves may be important in offsetting energy deficiencies observed

PAGE 33

33 in epileptic tissue, allowing brain tissue to be more resilient to irregu lar neuronal activity observed in seizures. Growth Although alterations in metabolism resulting from a high fat diet, or KT may be important for seizure control, the changes may also affect growth parameters. Compounding F actors While there is controversy in the literature as to whether or not abnormal growth is a direct adverse effect of KT, there are many compounding fa ctors influencing this debate. One in particular is restriction of calories being used in many protocols for administration of KT (1). A p otential risk of calorie restriction may be delayed growth and/o r suboptimal nutrition status. In addition to calorie restriction, patients on KT are generally complex and present with multiple pathologies that effect growth such as cerebral palsy (CP), spasticity, scoliosis, etc (23, 89) These disorders not only affect energy utilization, but also complicate length and weight measureme nts needed to evaluate growth. Also, the medications used to treat these pathologies, in particular AED therapy, a ffect en ergy balance and can interfere with bone health and development (90, 91) Additionally, malnutrition, growth abnormalities and micronutrient deficiencies are common in neurologically impaired children (92) Some of the factors that may contribute to this status are inadequate dietary intake, oral motor dysfunction, altered energy expenditure, and increased nutrient losses as a result of medication side effects or disease pathology (92) Patients may present with difficulty swallowing and feeding, vomiting, gastrointestinal reflux, or other gastrointestinal issues (92, 93) Overall health can be affected by inadequate nutritional status, in particular the immunity status (93) This is additionally important for patients with epilepsy in which illness can exa cerbate seizures (1) Consequently, the issue of growth and alterations in

PAGE 34

34 metabolism in the KT population is likely multifaceted and it is difficult to determine a central cause in a clinical setting. Calorie restriction Calorie restriction has been used in administering KT since its creation in 1921. It is often used as part of the diet prescription; however there is conflicting evidence as to whether this is necessary for effectiveness of the therapy (3, 4) Some literature suggests that energy i ntake ne eds to be restricted to up to 75% of the recommended energy allowance in order to induce ketosis (4, 24) In contrast, some studies have shown that seizure control may be achieved without calorie restriction (3, 14) One study in particular found between 6 2 and 71% of patients had >50% improvement in seizure control with 100% of recommended daily caloric intake (3). Studies conducted in animals have found that energy intake may be the key for seizure reduction, not ketosis resulting f rom a high fat ketogeni c diet. A study in rats fed a high carbohydrate diet that was restric ted to 90, 65, or 50% of energy requirements and found that they had seizure thresholds equal to thos e fed a ketogenic diet, but significantly lower blood ketone levels (94) C erebral pal sy In addition to calorie restriction, a large number of patients with neurological impairment are diagnosed with CP. Estimates are anywhere from 15 60% of children with CP also are diagnosed with epilepsy and it is more commonly seen in intractable epilepsy (95 97). Many studies have been conducted on energy expenditure and growth in patients with CP and have concluded CP results in lower resting energy expenditure (REE) and abnormal growth (89, 98). Reports have found altered nutritional status and signif icantly lower height-for age, weight -for height, triceps skinfold thickness and upper arm circumference in children with CP or disabled children compared to healthy children or to recommended values (99, 100).

PAGE 35

35 Epilepsy Another complicating factor is the issue of g rowth and epilepsy in general. Some studies have shown that epilepsy itself res ults in delayed linear growth. El Khayat et al (2004) concluded that the longer duration of disease, the more negative the impact on stature in female patients with e pilepsy compared to healthy controls (16) Sixty -six female patients were evaluated and found to have reduced mean height percentile compared with age -matched female controls and this was negatively correlated with duration of epilepsy (16) Additionally, E l Khayat et al (2003) studied male 8 18 y ea r olds with epilepsy and found they were significantly shorter than matched controls (16, 101) AEDs Finally, medi c ations may have a significant impact on growth and bone mineral density (91, 102104) One commo n criterion used for referring patients with epilepsy to KT is a failure of at least three AEDs (1, 105) Therefore most patients initiated on diet have been receiving multiple AEDs and continue these drugs until sei zure control is achieved. Guo et al (2 001) reported that long term AED therapy is associated with s hort stature (106) Furthermore some AEDs affect the regulation of energy balance and appetite (100, 107) Topiramate is associated with a loss of appetite and loss of body weight (107) whereas carbamazepine and valproic acid have resulted in weight gain (108, 109) Additionally, a number of the common AEDs have been shown to interfere with the metabolism of vitamin D and influence risk for osteopenia and o ste oporosis (93) Patients on AEDs have shown biochemical abnormalities in a number of metabolites involved in bone health such as decreased serum calcium, 25 hydroxycholecalciferol, and phosphorus and increased alkaline phosphatase (110113) Some of these changes have been attributed to incre ased microsomal P 450 oxidase activity leading to an

PAGE 36

36 increased conversion of vitamin D to inactive metabolites (114) direct inhibition of intestinal calcium transport (112) and increased bone calcium turnover (115) Intractable Epilepsy Although many s tudies on nutritional status and energy expenditure have been conducted in the CP population, three recent studies have focused on the intract able epilepsy (IE) population. The first of these studies was to evaluate nutritional status, energy, and dietary intake in 17 children with IE excluding patients enterally or parenterally fed (93) Mean daily energy intake was less than recommended and unbalanced with 18, 39, and 43% of total daily energy intake from pr otein, lipid and carbohydrate. Mean weight and height were lower compared to normal values for sex and age bu t not significantly different. However, 41.2% had a percentage of ideal body weight lower than 80% and were classified as malnourished and 24% were considered wasted as defined as a percentage of ideal we ight for height of less than 90%. Mean values for arm circumferen ce and subscapular thickness were 6 and 4% lower and tricep thickness was 20% higher than reference values. Additionally, the degree of impairment and presence of feeding prob lems correlated with percent ideal body weight a nd height. There was a large variation in REE ( 49 to 32%) between subjects and the mean was 5 % lower than predicted values. Patients also had lower intakes of micronutrients, in particular <60% of recommended for calcium, iron, and zinc (93) The second study was designed to evaluate the nutrient intake of children 1-8 years old with IE (n=43) compared to healthy children from the National Health and Nutrition Examination Survey (NHANES) 2001 2002 (n=1,718) and wi th the Dietary Reference Int akes (DRI) (116) The energy and macronutrient intake of children with IE was significantly lower compared to NHANES and more energy was consumed from f at and less from carbohydrate. They found that 70% were below the estimated energy requirement for low active children and

PAGE 37

37 49% were below the estimated energy needs for sedentary children. Children with IE had significantly lower intakes of vitamin A, vitamin E, vitamin B6, vitamin B12, riboflavin, niacin, folate, calcium, phosphorus, magnesium, zinc, copper and selenium compared to the NHAN ES data (116) Over 30% of IE children had intakes below the RDA for vitamin D, vitamin E, vitamin lino lenic acid (116) Intake of c alcium, vitamin D and vitamin K, key nutrients involved in bone health and metabolism was suboptimal in children with IE. Most recently, Bergqvist et al 2008 evaluated the REE usin g indirect calorimetry, energy intake from a 3 day weighed food record, growth status and body composition in 25 prepubertal children with IE compared to 75 healthy children of similar age a nd sex (117) They also stratified IE patients based on whether or n ot they had a diagnosis of CP. Children with IE had significantly lower percentage of predicted REE, weight z score, body mass index z -score, and fat free mass compared with healthy children. No significant difference was observed between IE without CP a nd the healthy children for weight, height, and BMI z score an d percentage of predicted REE. Weight, height, and BMI z score were significantly lower in IE children with CP c ompared with healthy controls. Therefore, CP was found to be a major factor in the suboptimal growth status and lowe r REE of the children with IE. Energy intake was found to be a significant predictor of REE in children with IE. They concluded that compared to healthy children, children with IE have suboptimal growth and lower REE due t o poor energy intake and not due to elevated energy expenditure (117) Growth and KT Currently it is controversial as to whether or not growth abnormalities are prese nt in the IE population on KT. Most patients receiving this therapy are in the pediatric age range and literature reports are inconclusive concerning whether or not KT results in an adverse effect on

PAGE 38

38 growth (14, 15) There have been 8 studies evaluating growth status in patients on KT (3, 14, 15, 118122) As previousl y reviewed, growth may be suboptimal in patients with IE regardless of administration of KT and energy restriction is som etimes used in patients on KT. Therefore it is difficult to interpret results from studies in th is complex patient population. One group found a significant decrease in height -for age z scores among subjects with high ketosis and following the diet for 12 months and no significant decrease in height -for age z scores for patients in moderate ketosis for whom the prescribed energy intake was based on a 3 day fo od record (14) Another study in which patients received adequate protein and energy intake s for normal growth (approximately 1g/kg body weight per day and 75kcal/kg per day) reported a significant increase in both height and weight after 6 months on the d iet. The mean percentile for height and weight, and the weight for height percent standard did not change from before the diet to 6 months o n therapy (15) Interestingly, a study that implemented calorie restriction by prescribing initial energy at 75% of estimated total energy expenditure, also found significant increases in height for children on both the classical diet and a modified medium chain triglyceride ( MCT ) diet from 0 to 4 months on therapy. Although there were significant increases in absolute height, there were not significant increases in height -forage percentiles and there was a significant decrease in the mean weight percentiles for the patients on the classical ketogenic diet from 0 to 4 months (118) Results of these two stud ies are outli ned in Table 2 2. This also indicates that although height and weight may be changing, if there is no comparison to a reference population, or the z score is not evaluated, some of the results in the l iterature may be insufficient for determining conclusions for growth bei ng a potential adverse effect. Whether or not calorie restriction is being used complicates the results and makes it difficult to distinguish

PAGE 39

39 whether growth abnormalities are a result of restricting energy and/or the composition of the die t involved with KT. Table 2 2. Comparison of study results for a diet prescribing adequate calories for growth and a calorie restricted diet of 75% of total energy expenditure Dietary Intake Absolute Percentiles Height Weight Height % Weight % Adeq uate protein and energy intake (15) Significant increase Significant increase No significant change No significant change Calorie restriction (118) Significant increase No significant change No significant increase Significant decrease More recently, Neal EG et al (2008) and Spulber G et al (2008) both evaluated growth parameters and examined relationships among dietary factors and blood parameters. Neal et al (2008) was the first to compare growth in children on both the classical and MCT oil keto genic diets and to evaluate relationships between growth and calorie and prot ein intakes (121) In 75 children on KT, they found weight z scores decreased significantly between baseline and 3, 6, and 12 months and height z scores decreased significantly b y 6 and 12 months on diet. Out of 40 children that completed 12 months of treatment, no significant difference was detected in mean energy intake of subjects on the classical compared to the MCT oil ketogenic diet however subjects on the MCT diet had a hi gher protein intake N o correlations were detected between calorie or protein intake and wei ght, height, or BMI. At baseline, significant differences in height, weight, and BMI z scores were not detected for ambulatory children compared to non ambulatory c hildren ; however at 3,6, and 12 months on diet, the decrease in weight z score from baseline was highly significant in the ambulatory group but not the nonambulatory group. Similarly, height z score decreased significantly in the ambulatory group at 6 mon ths but not in the nonambulatory; however at 12 months, both groups had a significant decrease in height z score. Using a paired t test, differences in weight z score between baseline and 3, 6, and 12

PAGE 40

40 months was significant in the 2 6 yr old children, but not in the older groups. Height z score s were not significantly different at 3 months in any age group; however at 6 and 12 months, the 2 6 yr old and 7 11 yr old group had a significant decrease in height z score compared to baseline but no changes were signific ant in the 12 16 yr old group. Overall, both weight and height z scores decreased during KT and at 12 months no difference was detected between classical and MCT diets even though the MCT diet resulted in a higher protein intake (121) Spulber G e t al (2008) were the first to examine the issue of growth in this population using height velocity and to compare growth to serum insulin-like growth factor 1 (IGF 1). In 22 patients on KT weight, height and BMI velocity decreased significantly after ini tiation of therapy. KT also resulted in a decrease in IGF 1 levels (Table 2 3). They found no correlations between seizures and growth; however height velocity correlated negatively with serum BHB (122). Table 2 3. Growth and IGF 1 values (mean S D) befo re and after starting KT Variable 1 year before KT Before KT After 1 year on KT Height SD score 0.65 2.06 0.77 2.27 1.14 2.09a Weight SD score 0.48 1.79 0.65 2.13 1.17 1.62a BMI SD score 0.12 1.37 0.21 1.10 0.71 1.04a Height ve locity SD score 0.60 2.22 4.10 2.17a IGF 1 SD score 1.04 0.98 3.25 0.83a,b Adapted from Spulber G et al (2008) (122) SD= standard deviation score similar to a z score. a Wilcoxon test, p<0.05. bAverage over the values measured during t he first year on KD. While the mechanism of therapy remains unknown, calorie prescription remains an important factor in both growth and possibly efficacy of therapy yet it remains a controversial issue. Currently clinicians have two options available wh en planning for growth of the patients: normal growth, providing recommended caloric intake (3) or delayed growth, restricting the amount of calories in the patients diet prescription (14, 24) There is con flicting evidence for support of both. Further re search should be conducted to determine whether or not the growth

PAGE 41

41 rates for children on KT are normal or abnormal with and without calorie restriction, and if calorie restriction is necessary for efficacy of therapy. As reviewed earlier, patients with IE o ften present with other underlying etiologies and are often on multiple AEDs, which both have been shown to effect growth and bone mineral density (90, 91, 102104, 123) Consequently, the issue of growth and alterations in metabolism in the KT population is likely multifaceted and it is difficult to determine a central cause in a clinical setting (Figure 2 1). Caution should be used in interpreting the literature and conclusions on growth in patients on KT. Fi gure 2 1. Compounding factors contributing to the evaluation of growth in the KT population. Dyslipidemia In addition to growth, the blood lipid profile may also be affected by changes in metabolism resulting from KT. While KT has been shown to be effective against seizures, it requires a fat inta ke (8090% of calories) much higher than recommended by the Ame rican Heart Association (AHA). The altered macronutrient ratio that forms the basis of KT is a concern because it is a potential r isk

PAGE 42

42 factor for dyslipidemia, which has been associated with chr onic vascular disease (17, 18) However, t here is conflicting evidence in the literature as to whether or not dyslipidemia is an adverse effect of KT (17, 124126) Currently none of the available feeding options for patients on KT provide a balanced fa tty acid profile similar to what is recommended by the AHA for healthy Americans. While some on a ketogenic diet are prescribed a 4:1 ratio of fat to protein and carbohydrate and consume 90% of their calories from fat, there is no evidence to suggest that their fatty acid profile should not be consistent wi th what is recommended by AHA. In general during therapy, orally -fed patients receive foods high in saturated fatty acids (SFA) such as heavy whipping cream, mayonnaise, and fat in meats and cheeses, wher eas tube -fed patients receive mostly polyunsaturated fatty acids (PUFAs), specif ically omega 6 fatty acids There are five main feeding options for patients undergoing therapy: 1 Solid food for oral ly -fed patients only, including heavy whipping cream, mayonnaise, oil, meat, cheese, etc., 2 Ketogenic eggnog for oral ly -fed patients made up of 36% heavy whipping cream plus egg substitutes, which may be used as the sole source of calories for the first few days of therapy 3 Ross Carbohydrate Free (RCF ) (Abbott Nu trition, Columbus, OH) or Beneprotein (N estl Nutrition, S.A., Vevey, Switzerland ) and Microlipid (N estl Nutrition, S.A., Vevey, Switzerland ), a safflower oil emulsion for tube -fed patients, 4 KetoCal ( Nutricia North America, Gaithersburg, MD ) a comme rcially available powdered product that provides a non -modular approach to KT and includes some vitamins and minerals for oral ly -fed and tube -fed patients Preparation only requires the addition of water and 5 Resource Benecalorie (N estl Nutrition, S.A. Vevey, Switzerland ), a low volume, calorie dense supplement with additional protein, vitamins, and minerals suitable for oral ly -fed and tube -fed patients. For patients who are fed soli d food only, the sources of fat consumed generally lead to a f at intak e of mostly SFA. For patients who consume ketogeni c eggnog via tube feeding, the

PAGE 43

43 primary source of fat comes from MicroLipid (Nestl Nutrition, S.A., Vevey, Switzerland), a safflower oil emulsion available by prescription only or from other vegetable oil s and resulting in an intake that provides primarily PUFAs. Additionally, although KetoCal is very convenient for caregivers to use, it is mostly monounsaturated fatty acids (MUFAs) and specifically uses hydrogenated soyb ean oil in the 4:1 product that le ads to a high trans fatty acid intake. Finally for patients using Benecalorie (Novartis), another commercially available product, the fatty acid composition is mainly MUFA due to the use of high oleic sunflower oil. None of these options provide a balanc ed fat profile that is similar to the recommended for healthy Americans (Figure 2 2), and intake of diets with these fatty acid profiles could ultimately lead to problems with dyslipidemia (17, 18). Figure 2 2. Approximate fatty acid percentages in the various KT feeding options. (SFA=saturated fatty acids, MUFA=monounsaturated fatty acids, PUFA=polyunsaturated fatty acids)

PAGE 44

44 Type of F at and Efficacy of KT Current research into the mechanism behind the anti -seizure effects of a high fat diet suggests tha t the beneficial effects may be related to elevated plasma levels of PUFAs especially arachidon ic acid (AA) and docosahexaeno ic acid (DHA) (127) KT has been found to produce elevations of both AA and DHA in serum (128, 129) and brain (130) of patients and animals; therefore many theories have developed in relation to PUFAs and mechanism of diet. Th ese elevations may be due in part to the idea that their precursors -linolenate an -linoleate are more easily oxidized, supported by more rapid transport of these PUFAs into the mitochondria (131) Another possible mechanism behind the clinical benefit was developed based on studies with labeled 14C and 13 -linole linoleate getting into newly synthesized lipids, such as cholesterol or a saturated or monounsaturat ed fatty acid (132) This pathway has been termed carbon recycling because the carbon starts out on the PUFA and then gets transferred t o cholest erol or a fatty acid. This recycling phenomenon has been well documented in animal models (133, 134) Carbon recycling is thought to be beneficial in that it allows the capture of -linolen ates end product DHA, is produced (127) linolenate induced more ketosis, a study using a rat model was performed. The researchers found that a high MCT diet raised BHB the most, followed by a flaxseed oil enriched diet linolenate), which was much more k etogenic than the butter diet. Although there was higher ketosis with the MCT diet, there was no difference in protection of seizures versus the flaxseed oil diet. linolenate are more ketogenic compared to diets with less unsaturated long chain fatty acids (64)

PAGE 45

45 Other studies have found that under fasting conditions, there is a differential, more rapid linolenate and linoleate from adipose tissue in humans and in rats compared to other long chain fatty acids (135, 136) A more recent study in rats showed that although the level of ketosis was not detectable on day 10 of the ketogenic diet, the diet reduced PUFAs in adipose by 44% and plasma up to 90%, respectively, increased PUFAs in liver triglycerides 25 -fold and inc reased the levels of AA and DHA in th e brain 15% (130) This led to the idea that the net fat ty acid flux on the ketogenic diet was from adipose tissue through the blood to other tissues including the brain a nd that redistribution of PUFA to the brain may have seizure suppress ing effects (130) Studies performed in humans have found that a 2 to 5 fold increase in AA and DHA in the plasma is accompanied by improved seizure control (128) The benefits of this redistribution of PUFA to the brain may be supported by studies conducted in a rat model in which exogenous -infusion of DHA produced direct s eizure -suppressing effects (137) Consequently, the increased ketogenic effect of these two fatty acids may be an add itive effect -oxidation (127) Research also suggest s that the mechanism responsible for the beneficial eff ects of KT involves the actions of PUFA on voltage gated Na+ channels neuronal excitability, activation of lipid -sensitive K2P channels, and induction of mitochondrial uncoupling proteins (138140) However, this hypothesis remains controversial due to differences in fatty acid content of the diet s consumed by patients, the route of feeding (i.e., oral ly -fed versus tube -fed), and among different medical centers, yet ef ficacy reports remain similar. In addition, a recent study by Dahlin et al 2007 measured the levels of 17 FAs in plasma phospholipids before initiating die t treatment and at 1,3,6, and 12 months after d iet initiation in 25 children. They also supplemented the diet at 1 month with omega 3 FAs as fluid fish oil with 40% omega 3 fatty acid s with 17.5%

PAGE 46

46 from eicosapentaenoic acid and 11% from DHA. While they foun d significant changes in several FAs, there was no correlation between s eizure response and FA levels (124) On the contrary, Fraser et al (2003) reported a positive correlation between total serum AA and a reduction in seizure frequency (128) These resu lts suggest that it may be possible to alter the composition of the diet to offer more cardiovascular protection on therapy without a chang e in efficacy This possibility needs to be tested. Based on these theories, new research into the use of dietary supplementation with PUFAs alone for the treatment of seizures has been conducted. Although early case reports suggested benefit (141) a recent randomized trial in adults with ep ilepsy did not show benefit of PUFA supplement ( EPA plus DHA, 2.2 mg/day in a 3: 2 ratio ) versus placebo (142) Yuen et al (2005) reported a decrease in seizure frequency at 6 weeks in patients supplemented with 1 g eicosapentaenoic acid and 0.7 g DHA daily compared to placebo, however the effect was not sustained (143) More researc h is needed to consider differences in amounts of EPA and DHA and the ratio between the two, duration of treatment, as well as age and seizure types. At this point it is still unknown whether differences exist for the optimal fatty acid profile for effecti veness against seizures versus overall cardiovascular health Effect of KT on Blood L ipids Several studies have been conducted on the effect of KT on blood lipids; however, dyslipidemia as a potential adverse effect of KT remains controversial Some resear chers have shown that the diet results in an atherogenic blood lipid profile (17, 144, 145) while others have not (126, 146) and some reports show mixed results depending on time on diet and type of ketogenic diet (18, 61, 118, 124, 147) A prospective s tudy in 141 children on traditional therapy found a significant increase in atherogenic apo B -containing lipoproteins and a significant decrease in anti atherogenic HDL cholesterol on diet as compared to before diet (17) The

PAGE 47

47 changes were less marked at 12 and 24 months on diet Similarly, Dahlin et al. 2007 showed that cholesterol and triglycerides showed maximal levels at 1 month after starting the diet but then decreased over time on diet to values within normal limits (124) A retrospective review of ea rly complications of the diet reported that 14.7% of patients had hypercholesterolemia, 27.1% had hypertriglyceridemia, and 3.9% had low HDL cholesterol within 4 weeks of initiation. However, of the 41 patients with hypercholesterolemia, 73.2% improved spo ntaneously, of the 46 patients with hypertriglyceridemia, 71.7% improved spontaneously and of the 6 with low HDL, 66.7% improved without interventions (18) Effect of Modifying F atty A cid P rofile on Blood Lipids In terms of modifying the therapy to alter blood lipids, several report s have been published. One study of healthy young adults in 2004 found that t he subjects who received a diet high in SFA had significantly higher LDL cholesterol and total cholesterol concentrations compared to subjects receivin g a diet high in PUFA (125). All diets contained 70% fat, 15% carbohydrate, and 15% protein. The PUFA enriched diet contained 15% SFA, 25% MUFA and 60% PUFA whereas the SFA fat diet contained 60% SFA, 25% MUFA, and 15% PUFA. The subjects who received a di et high in saturated fatty aci ds had significantly higher LDL cholesterol and total cholesterol concentrations compared to subjects receiving a diet high in polyunsaturated fatty acids ( 125). Another older study was designed to compare the effect of incorp orating a medium -chain triglyceride (MCT) diet and a modified MCT diet versus the classical ketogenic diet. The classical diet patients had a significantly higher blood ketone concentration compared to patients on the other two diets, and although choleste rol and LDL cholesterol were higher for patients on the traditional diet, they were not significantly different, and none of the three diets had significant effects on cholesterol and lipoprotein levels, despite the high fat intake (126). Huttenlocher PR (1976) found that children on the MCT oil diet did not

PAGE 48

48 have elevations in serum cholesterol, whereas children on the classical version did (61). Similarly, in a prospective study of 25 children, 14 on classical KT and 11 on MCT oil version, there was a 0.7 decrease in the ratio of total to HDL cholesterol at 4 months in the patients on MCT oil but not in the classical diet (118). The most recent study reported on the issue of dyslipidemia was a prospective study of 137 subjects designed to examine triglyceri de and total cholesterol at baseline and after diet initiation (147). Hypercholesterolemia was def ined as cholesterol >200 mg/dL. The percent of children with hypercholest er olemia at any time point increas ed from 25% at baseline to 60%. When patients were stratified into groups, the group with cholesterol <200 mg/dL were younger (1.5, 4.0, p<0.001) and there were significantly more patients fed solely formula through a gtube Children who were tube -fed were less likely to have high cholesterol than oral f eeders after adjusting for age and diet ratio. In patients receiving solid food, hypercholesterolemia occurred most often, however cholesterol improved in almost 50% even without interventions. Dietary interventions such as substitution with MCT or PUFA decreasing ratio, or adding carnitine were tried in patients with hypercholesterolemia There was only a slightly higher likelihood of a decrease in cholesterol in children who had an intervention compared to observati on alone (60% vs 41%; p=0.11). No chil dren discontinued the diet due to dyslipidemia as an adverse effect of the therapy however (147). Table 2 4 Dietary interventions and improvement in patients on KT with hypercholesterolemia Intervention Improvement No Improvement MCT oil substitution P UFA substitution Decrease in ratio Addition of carnitine 1 3 4 1 1 2 2 1 Adapted from Nizamuddin J, Turner Z, Rubenstein J, Pyzik P, Kossoff E. Management and risk factors for dyslipidemia with the ketogenic diet. J C hild Neurol. 2008 Jul;23:75861 (147). Improvement defined as and n=15.

PAGE 49

49 These data suggest that altering the fatty acid profile given to patients on KT may produce effects on blood lipid levels, but the long term effect on the KT population is unknown. Therefore we assessed the bl ood lipids of patients on KT retrospectively and prospectively to elucidate if dyslipidemia is prevalent in this population and if it is related to diet prescription, type of fat, and/or length of time on diet. Preliminary S tudies Evaluation of k etogenic e ggnog and KetoGatorNog treatment for seizures The rationale for improving KT was based on the variations in the fatty acid profile and the limited amounts of most vitamins and minerals present in the ketogenic diet In the late 1990s, a new dietary ketogen ic formula, call ed KetoGatorNog (KGN), was developed and tested with patients at the University of Florida to see if it improved nutrient intake. KGN was created using ProViMin a protein, vitamin, and mineral powder ( Ross laboratories, Columbus, OH), Micr olipid (a safflower oil emulsion by N estl Nutrition, S.A., Vevey, Switzerland ), and a unique blend of oils composed of canol a, olive and MCT oil. These were used in specific amounts to yield a fatty acid profile similar to what is recommended by the AHA for healthy Americans (Figure 2 3 and Tables 2 5 and 2 6 ). The main purposes of the study were to : Improve the vitamin and mineral composition of the traditional ketogenic eggnog (KE) Improve the fatty acid profile of the traditional therapy D ecrease t he nutrient content variability Increase the shelf life of the ingredients used in traditional KE E nsure palatability of the new formula This was a preliminary study that included 10 patients who completed the study at baseline and 3 months. Five subjec ts were randomly selected as standard KE patients and 5 were selected as KGN patients. There were no significant differences between the diet groups for height

PAGE 50

50 weight, age, gender, or race. After 3 months of treatment, serum triglycerides (TG) were signif icant ly higher in KE patients compared to KGN (p= 0.0019). T he increase in TG from baseline to 3 months was greater in KE than in KGN. A significant increase in total cholesterol was detected only in the subjects randomized to KE. HDL cholesterol decreased significantly (p=0.0079) in the KE patients after 3 months and remained roughly the same in the KGN patients (p=0.9371). Non -HDL cholesterol increased significantly after 3 months in KE (p=0.014) and not in KGN (p=0.6952). Additionally, non HDL was signif icantly higher at 3 mo in KE compared to KGN (p= 0.048). LDL cholesterol decreased in KGN during the 3 months of study, but was not statistically significant (p=0.5307) (Table 2 7) TG levels were too high to obtain LDL choles terol levels in the KE patients (Figures 2 4 and 2 5 ). Figure 2 3. Approximate fatty acid percentages in the AHA recommendations and the KGN fat blend. Calculations for the AHA were made for percentages of fatty acids based on recommendations in 2000 of total fat intake <30%, saturat ed 7 10%, and polyunsaturated up to 10% (148). Average of 8.5 was used for SFA, 10 for PUFA and 30 8.5 10=11.5 for MUFA, for a total of 30.5. SFA=8.5/30.5=28%, PUFA=10/30.5=33%, and MUFA=12/30.5=39%. (SFA=saturated fatty acids, MUFA=monounsaturated fatty a cids, PUFA=polyunsaturated fatty acids)

PAGE 51

51 Although preliminary, the serum lipid results for KGN patients suggest superiority over the traditional KE therapy. The ability of the new formula to keep the triglyceride level lower than with KE therapy, increase the levels of HDL cholesterol and to decrease the levels of LDL cholesterol, can suggest that it is possible to improve the overall nutrie nt intake of the patients on KT. Table 2 5 M icronutrient adequacy for PO Ketogenic Eggnog (KE -PO) Age vitamins <50 % minerals <50% 4 6 yr C (7%) thiamin (29%) niacin (3%) B6 (27%) D (0%) K (0%) biotin (0%) Ca (49%) Fe (28%) Mg (44%) P (48%) Zn (24%) Cu (5%) Mn (1%) I (0%) Se (0%) Adapted from Hang, T 2000 (149) PO Ketogenic Eggnog (KE PO) compared to current ketogenic eggnog used in traditional therapy versus the new dietary formula used in the study, KetoGatorNog (KGN). There were no vitamins and minerals less than 50% in the PO KetoGatorNog (KGN PO) formula. Analysis performed by Tran Hang and Jodi Passman. Tabl e 2 6 Fatty a cid p rofile of KGN PO and KGN TF compared to KE -PO and KE TF KGN PO KE PO KE TF SFA 28% 58% 10% short chain not detected 3% not detected medium chain 20% 17% not detected long chain 7% 37% 10% MUFA 39% 28% 12% PUFA 30% 8% 78% linol eic 26% 7% 78% linolenic 4% 1% not detected Adapted from Hang, T 2000 (149). The fatty acid profile compared to the traditional ketogenic eggnog (KE) used in current KT versus the new dietary formula used in the study, KetoGatorNog, KGN. Analysis performed by Tran H ang and Jodi Passman. Note that the sum of short, medium, and long chain FA did not always add up to total SFA because sum of short, medium, and long chain FA values as decimal form rather than whole numbers. KGN PO (PO KetoGatorNog) and KGN TF (Tube -fed K etoGatorNog) compared to KE -PO(PO ketogenic eggnog) and KE TF(Tube -fed ketogenic eggnog).

PAGE 52

52 Figure 2 4. Mean values for triglycerides and total cholesterol in mg/dL at initiation (0 months) and end of study (3 months) for patients on traditional ketogenic therapy (KE) versus the new KetoGatorNog (KGN). Figure 2 5 Mean values for HDL cholesterol, n on -HDL cholesterol, and LDL cholesterol in mg/dL at initiation (0 months) and end of study (3 months) for patients on traditional ketogenic therapy (KE) vers us the new KetoGatorNog (KGN).

PAGE 53

53 Table 2 7 Comparison of b lood lipid data for KE versus KGN TG mg/dL Total Chol mg/dL HDL mg/dL NonHDL mg/dL Baseline 3 mo Baseline 3 mo Baseline 3 mo Baseline 3 mo KE (mean SD) 11536 651217 21160 296153 587 39 11 15360 256146 KGN (mean SD) 8326 268121 17019 18125 4613 465 12420 13523 KE Baseline vs 3 m o (pvalue) 0.0003 0.0445 0.0079 0.0140 KGN Baseline vs 3 m o (pvalue) 0.0408 0.7216 0.9371 0.6952 KE vs KGN (p value) 0.6782 0.0019 0.4205 0.0690 0.0838 0.2990 0.5457 0.048 Emulsion project There are no commercial lipid emulsion s c urrently available that contain the fatty acid profile recommended by the AHA. In an effort to combine the recommended fat ty acid profile in KGN in a more patient friend ly manner, an emulsion proj ect was started. The purpose of the emulsion was to create a fat blend that would be convenient for parents and patients to use in stead of weighing and mixing all of the oils individually and to provide a healthier fat blend tha n what is currently ava ilable commercially. The blend was based on the KGN recipe but instead of using Microlipid, a fairly expensive commercial product, safflower o il and soy lecithin were used. A pressure homogenizer was used to blend the oil s, lecithin and water. For the purpose of optimizing the emulsion, 3 variables were examined : 1 Concentration of soy lecithin (within recommended use range) 2 Duration of homogenization 3 Concentration of water Emulsions were photographed immediately after preparation an d at varying peri ods of storage at 4 degrees C. Separation of the emulsion was measured based on separation

PAGE 54

54 percentages that were defined as the height of interface between aqueous and emulsified layers/ tot al height of the fluid volume. Quality of emulsion was defined by assigning a grade of the emulsion after mixing by 10 inversions. After storage for more than 100 days, an emulsion made with 0.74% soy lecithin, 45% water, and 40 homogenization cycles had no separation and good appearance following invers ion. These preliminary results need to be evaluated using a l arger volume and more sterile conditions. Acylcarnitine Profile While KT affects brain metabolism resulting in changes in seizure activity and may alter metabolism to e ffect both growth and blood lipids, ingestion of a very high fat diet may also be expected to affect the acylcarnitine profile. Carnitine is a major metabolite involved in fatty acid metabolism and therefore esters of carnitine, or acylcarnitines, may be altered from radical cha nges to dietary intake. Introduction L -Carnitine (trimethylamino -hydroxybutyrate ) (LC) is present in cells and tissues as both free carnitine and acylcarnitines, including acetyl L -carnitine (ALC). LC is a naturally occurring endogenous compound in all mammalian species and its most widely known functio n is as an important transporter of long -oxidation. Hu mans obtain carnitine pr edominately from meat and dairy foods and through endogenous biosynthesis. LC is synthesized in vivo from L lysine and L -methionine most ly in liver and kidney (150). Under normal conditions, c arnitine palmitoyltransferase 1 (CPT 1) uses carnitine and acyl Coenzyme A (acyl CoA) as substrate s and catalyzes one of the first highly regulated steps that is common to fatty acid oxidation and k etogenesis ( Figure 2 6) (151, 152). Therefore, the critical substrates carnitine and ac ylcarnitine s are important in understanding the metabolic pa thway

PAGE 55

55 associated with ketosis. While carnitine is an important me -oxidation of fatty acids, it may also be important in transporting other metabolites due to its ability to form esters with many carboxylic acids. The free hydroxyl group shown in Figure 2 7, can be enzymatically esterified to activ ated acetate groups or to an other activated carboxylic acid including fatty acids of all chain lengths, to form acylcarnitine and carry the acyl groups throughout the body (Figure 2 8) (151). Carnitine acetyltransferase (CAT) catalyzes the synthesis of sho rt -chain acylcarnitines, specifically ALC, and is located on the inner mitochondrial membrane as well as in microsomes and peroxisomes. ALC and other acylcarnitines can be transported across the inner mitochondrial membrane by carnitine acylcarnitine trans locase and transporte d out of the mitochondrial membrane into the cytosol (151, 152). ALC is an activated water -soluble molecule that is transportable from one organ to another providing an acetyl group for various functions. Figure 2 6 ALC and ene rgy metabolism. Integration of ketone and ALC metabolism.

PAGE 56

56 Figure 2 7. L carni tin e structure. Figure 2 8 Example acylcarnitine structures. Acylcarnitine Profile and Clinical Applications LC has an amphiphilic structure making it very mobile througho ut the cell and the free hydroxyl group allows for the potential of many different molecules to attach creating a wide array of possible acylcarnitines (Figure 2 7). The ability to esterify and transport metabolites throughout the body distinguishes LC as a unique metabolite and suggests the acylcarnitine

PAGE 57

57 profile as a useful indicator of metabolic changes, particularly related to disease states. In addition, this wide array of possibilities also leads to a broad range of structures that chemically and metab olically are very different. For instance, ALC is a small, water soluble molecule that is easily transportable and may be used to deliver acetyl groups to a variety of locations. Yet a long -chain fatty acylcarnitine, such as palmitoylcarnitine, needs a tra nsporter to cross the membrane and therefore may be more restrictive in its actions (Figure 2 8). As a result, changes in individual acylcarnitines may imply changes in specific metabolic pathways and monitoring of these should lead to a better understandi ng of mechanisms of disease and allow for better design of treatment regimens. The acylcarnitine profile has been shown to be useful in identifying inborn errors of metabolism in neonatal screening and in particular fatty acid oxidation defects such as lo ng-chain acyl CoA dehydrogenase -deficiency (VLCAD) disorders of organic acid metabolism such as propionyl CoA carboxylase deficiency using tandem mass spectrometry based analysis (MS/MS), (153 155) and recently using metabolomics methodology to identify p atients with methylmalonic and propionic acidemia (156). Inborn errors of metabolism can lead to a build up of toxic metabolites and can result in serious, even fatal diseases early in life. For this reason, early comprehensive neonatal screening is used t o detect abnormalities to avoid major physical and neurological effects (158). Mass spectrometry (MS) based analysis is used to detect and diagnose newborns by identifying and quantifying specific metabolites (157, 159). Newborn screening by MS/MS has identified oxidation, organic acidemias, disorders of the urea cycle, and rare disorders of metabolism such as short chain acyl CoA dehydrogenase (SCAD) deficiency before the o nset of symptoms. For example, a fter the application of tandem mass spectrometry was introduced clinically, infants with SCAD

PAGE 58

58 and isobutyryl CoA dehydrogenase deficiency have been detected based on elevated butyry l carnitine/isobutyryl -carnitine concentrat ions in newborn blood spots (160), even though two out of three infants remained asymptomatic at the time of diagnosis. Some agerelated variations in acylcarnitine and free carnitine concentrations have been observed and should be taken into consideration when diagnosing and managing inborn errors of metabolism (157, 161). Inborn errors of metabolism detected by acylcarnitine profile analysis adapted from Rinaldo et al 2008 and their corresponding acylcarnitine changes are listed in Table 2 8 at the end of the chapter (162). In addition to identifying inborn errors of metabolism, the acylcarnit i ne profile may also be useful in identifying other metabolic perturbations. In healthy neonates, cord blood concentrations of total acylcarnitines strongly correla ted to birth weight and lower umbilical artery pH caused accumulation of long -chain acylcarnitines (157). The acylcarnitine profile may be a useful parameter for identifying perinatal asphyxia and other metabolic disturbances in utero Outside of newborn s creening, other alterations in metabolism may be detected through monitoring of the acylcarnitine profile. Ulcerative colitis is a disorder involving chronic inflammation of the colonic mucosa in which the etiology and pathogenesis of the disease are unkno wn. Alterations in short chain fatty acid metabolism have been identified in pa tients with this disease (163). Celiac disease is an autoimmune disorder d erived from gluten intolerance. Bene J et al (2005, 2006) found a significant difference in acylcarni tine profile in plasma of adult patients with ulcerative colitis and patients with celiac disease compared with controls with no change to free carnitine levels (164, 165).

PAGE 59

59 More recently, metabolomics is being used to create a more comprehensive metaboli c profile of the plasma through nontargeted liquid chromatography MS that allows the investigator to view thousands of metabolites and identify significantly different metabolic features (156). The metabolome, including the acylcarnitine profile, is becomi ng an increasingly popular tool in the literature (more detail on metabolomics in the next section). Many labs have evaluated the utility of acylcarnitine profiles in the diagnosis of inborn errors of metabolism, some of which were revi ewed by Pasquali M e t al (2006) (166). Supplementation of specific acylcarnitines, furthermore, may lead to different effects for particular disease states. For example, propionyl carnitine may improve general fatigue (151) and may protect heart health (167), and palmitoyl c arnitine may regulate lipid esterification. Changes in C arnitine and A cylcarnitine Levels During starvation and after eating a high fat diet conditions similar to KT the proportion of carnitine that is acetylated in liver and kidney significantly increas es and, oppositely, a high carbohydrate diet causes very low levels of ALC in liver (168). In humans, there appears to be a delayed decrease in plasma LC and a rapid increase in both long and short chain acylcarnitines during fasting or diabetic ketosis ( 169 171). While plasma lev els do not directly correlate with cerebra l levels, in neonatal rats starvation led to a significant increase in mean brain acylcarnitine concentration compared to control rats, with almost all of the increase attributed to short -chain acylcarnit ines (172). The authors concluded carnitine and its relative esters may be redistributed to the brain during fa sting and the brain may use them for energy production or possibly for the delivery of acetyl groups. Plasma acylcarnitine levels were analyzed in 1 7 yr old children after fasting and ingestion of sunflower oil, which is composed largely of linoleic acid (66.2%), a PUFA, and oleic acid (21.3%), a MUFA. Under both conditions there was an increase in all plasma straight -chain

PAGE 60

60 acylcar nitines and ALC was the largest contributor to the increase in total esterified carnitine with an almost 4 -fold increase after fasting (173). A summary of the changes in carnitine concentrations due to alterations in metabolism are listed in Table 2 9. Ca rnitine and Acylcarnitines in Brain The importance of carnitine in brain is emphasized by carnitine deficiency symptoms, many of which involve major deleterious effects in the brain. Because the brain is highly reliant on oxidative metab olism, impaired fat ty acid metabolism and energy production due to lack of carnitine leads to metabolic encephalopathy (174). In addition, LC is taken up by neuronal cells through a Na+and energy-dependent mechanism, therefore conditions of metabolic disturbances, as seen in many neurological disorders, may exacerbate a carnitine deficiency. Table 2 9 Changes in plasma carnitine and acylcarnitine concentrations under dietary manipulations and alterations in metabolism. Model Diet Free L Carnitine (LC) Acylcarnitine/ ALC Re f. Rat Starvation/ Fasting (same in brain) Pearson, 1967 (liver and kidn ey, not heart); Murakami 1997 (brain); McGarry 1975; Brass 1980 High fat Pearson, 1967 (liver and kidn ey, not heart); Murakami 1997 (brain) High CHO Pearson, 1967 (liver and kidney, not heart) Human Starvation/ Fasting Frohlich et al 1978 (serum); Hoppel and Genuth 1980; Costa 1999 High fat/fat load Costa, 1999 (plasma); Hack 2006 (microdialysis) Diabetic ketosis Genuth and Hopp el 1979; Hoppel and Genuth 1980 While glucose is thought to be the primary energy source for the adult brain under normal conditions, fatty acids can be used by the brain as well (175) In a recent study, a lmost 20% of the total oxidative energy produced in brain was f rom the oxidation of 13C -octanoate in a dult male Sprague Dawley rats (175). In addition the majority of the brain is composed of fatty acids and fatty acids are n eeded for incorporation into structural lipids (176). Similarly, fatty acids

PAGE 61

61 become key energ y substrates for brain under metabolically compromised conditions such as fasting or starvation, states in which the levels of free acetyl CoA and ketone bodies also become important (50, 54). Carnitines and acylcarnitines functions in fatty acid metabol ism, ketosis and buffering of the acylCoA/CoA concentrations, therefore, are significant in brain metabolism, particularly metabolic disturbances present in neurological disease. Carnitine is found to accumulate to a lower extent in brain compared to perip heral tissues (177). The enzymes needed for the synthesis of carnitine are also present in brain tissue (178) as well as the acyltransferases necessary for the synthesis of acylcarnitines (179). Carnitine can also be synthesized and transported into brain. Although more work is needed on the details, it appears that carnitine can be transported through the blood -brain barrier through two transporters, OCTN2, a Na+ dependent transporter shown through RT PCR to be present in brain endothelial cells (180), and ATB0,+, a Na+-, Cl-dependent amino acid transporter expressed in the hippocampus (181, 182). Transport of carnitine in the brain has been previously reviewed (174). More recently, a detailed experiment localizing OCTN2 in cells forming the blood brain ba rrier suggests carnitine can also be transported out of the brain (183). Such data indicate carnitine and acylcarnitines may be playing a role in brain through the removal of certain acyl esters. LC accumulates in neurons of the cerebral cortex and forms a cylcarnitines. Isolated neurons of the adult brain contain approximately 80% free carnitine, 1015% ALC, and less than 10% long-chain acylcarnitines (184). In comparison, the ALC concentration is increased and free carnitine is decreased in suckling rats ( 184). Because carnitine and its acyl derivatives have chemical structures comparable to choline and acetylcholine, it has been proposed that they play a role in neurotransmission (185). Acylcarnitine and LC supplementation have shown beneficial

PAGE 62

62 effects in the treatment of aging, chronic degenerative diseases and slowing the progression of mental deterioration in Alzheimers disease (AD) (186, 187). Therefore, the roles of acylcarnitines and LC in brain and their potential therapeutic mechanisms need to be f urther elucidated. Acetyl -L-carnitine (ALC) Role in Brain ALC is present in relatively high levels in the brain ( 188) and it is highest in the hypothalamus ( 177) where the level of the ALC synthesizing enzyme, CAT, is also high. ALC can readily cross the b lood -brain barrier (189), so supplementing with this compound could affect brain metabolism Injection of ALC in rats led to reduced oxidation of glucose and increased glycogen syn thesis in brain (190). Changes in the activities of specific enzymes involve d in the tricarboxylic acid (TCA) cycle electron transport chain and amino acid metabolism have also been observed after treatment with ALC (191). Specifically, the activities of citrate synthase and glutamate dehydrogenase were decreased while the activi ketoglutarate dehydrogenase and cytochrome oxidase were increased. Some of ALCs proposed n europrotective benefits involve improved mitochondrial energetics and function, antioxidant activity, stabilization of membranes, protein and gene expressi on modulation, and enhancement of cholinergic neurotr ansmission (19, 174, 192194). This section will highlight these mechanisms of action for ALC in the brain and how these may be applicable in various clinical situations A summary can be found in Table 2 10. Energy metabolism and membranes Carnitines primary function in th e cell is in lipid metabolism. Animal studies have traced the fate of the acetyl group in ALC by using radiolabeled [1 -14C] ALC. Injection of this compound into mice showed a rapid incorporation of the acetyl groups ( that were not immediately metabolized to carbon dioxide ) into lipid biosynthesis pathways in the liver (195).

PAGE 63

63 When injected into the brain, the acetyl groups were mostly incorporated into SFAs (about 60% of the radioactivity present in the tissues), but they were also in MUFA and PUFAs (196). Interestingly, in this experiment labeled [U -14C]glucose was not incorporated into PUFAs unlike [1 -14C] ALC Due to ALC s role in overall energy metabolism, many researchers have invest igated how ALC can repair neurological damage through metabo lic pathways. Supplementation with ALC has been show n to normalize the levels of high-energy phosphate in brain of AD patients as measured by 31P magnetic resonance spectroscopy (197). In rats, AL C treatment increased the concentration of phosphocreatine and decreased the concentrations of lactate and inorganic phosphate in aging ( 198) and postischem ic brain models (199). ALC also seems to provide protection during metabolic stress, such as ischemi a, hypoxia, aging, alcohol, and brain injur y (193). For example, dogs treated with ALC showed significantly lower neurological deficit scores and more normal cerebral cortex lactate/pyruvate ratios than control animals after cerebral ischemia and reper fusi on (200). These results indicate that ALC improves neurological outcome potentially as a result of normalizing brain energy metabolites. Aging is associated with mitochondrial dysfunction and altered m etabol ic states (201). Many groups have researched ca rnitines utility in reversing the me tabolic side effects of aging. Feeding ALC to older rats (22 28 months of age) increases the cellular consumpt ion of oxygen, a parameter that decreases wit h increasing age (202). In this same study, ALC reversed the dec lines in mitochondrial membrane potential and cardiolipin (a phospholipid) concentrations t hat are associated with aging. Ames and Liu reviewed support for these results as well as a role for ALC in stabilizing the inner mitochondrial membrane ( 203). Sphin gomyelin and cholesterol tend to accumulate in the cerebra of rats with increasing age, and both of these increases are

PAGE 64

64 attenuated by long -term ALC supplementation (204). It is apparent that ALC can reverse alterations in membrane lipid content and function and it can improve age related changes in metabolism, either directly through supplying high energy acyl groups or indirectly through restoring membranes. Many studies have shown other beneficial effects of ALC in either protecting mitochondria against biochemical insult or reversing damage to mitochondria. ALC appears to have neuromodulatory action through increasing the synthesis of phospholipids that are required for membrane formation and integrity ( 193). When ALC was supplemented with -lipoic acid (LA) to rats, reversals in the age associated decline of mitochondrial membrane potential and the levels of ascorbate and malon yl dialdehyde (MDA, an indicator of lipid peroxidation) were observed (205). Another potential role for ALC was elucidated by Cass ano, et al (2006) who found an increase in transcripts related to mitochondrial biogenesis with ALC supplementation in a rat model for hindlimb muscle atrophy (206). ALC could, then, not only help preserve mitochondrial membrane integrity, but it could a lso help to generate more mitochondria under certain conditions helping to preserve overall metabolic function ALC supplementation prior to cyanide injection in rats had beneficial protective effects on preventing behavioral changes, but did not affect phosphoinositide metabolism ( 207). The authors propose carnitines beneficial effect may not be through supplying energy to the brain, but rather through providing activated acyl groups for the reacylat ion of membrane phospholipids. This role for carnitine was previously found in the deacylation reacylation process in the membrane phospholipids of human r ed blood cells (208) and has since been supported by other studies (209, 210). Researchers have reported ALC increases fluidity in rat brain microsomes and liposomes (208). ALC, as well as LC may affect membrane fluidity due to its amphiphi lic structure that

PAGE 65

65 may directly interact with the surface charges on cell me mbranes (211). The carboxylic group can interact with charges on membrane phospholipi ds, glycol ipids, and proteins. Neural membranes contain a large amount of phospholipids which can be degraded through the action of phospholipase A2 to important lipid mediators such as DHA and AA, which further form eicosanoids and docosanoids found to be importan t in inflammatory and oxidative stress re sponses (212). In addition, alterations to the composition of these phospholipids in membranes can alter membrane fluidity, permeability, and functioning of membrane proteins that act as important receptors and signals for multiple downstream reactions. Alterations in neural phospholipid composition and further effects on signal transduction pathways have been found to be characteristic of many neurological disorders ( 213216). Recent evidence suggests that ALC also plays a role in the elongation -desaturation of the n3 PUFAs to form 22:6n3, DHA, in mitochondr ia (217219). ALC is thought to provide an intra -mitochondrial source of acetyl groups and donate t hem in the elongation pathway. As stated earlier, labeled ALC is found to be incorporated into PUFAs (196). Changes in DHA content of membranes can markedly affect synaptic plasticity, inflammatory response, gene expression, ion channels, membrane bound proteins and neurotransmission ( 220). Anti -oxidant and anti -apo ptotic functions The formation of reactive oxygen species (ROS) and mitochondrial dysfunction leads to metabolic and oxidative stress that are underlying processes in many neurotoxic and neurodegenerative diseases, such as AD and Parkinsons disease (PD) ( 221). The oxidative stress affects the activities of the respiratory chain complexes I, II, III, IV, and V, which is integral i n many neurological disorders. Treatment with ALC has shown to provide protection from excitotoxicity induced by respiratory chai n complex inhibitors (210, 222224). Research has also shown that ALC increases the expression of heme -oxygenase 1 (HO 1), upregulates heat shock

PAGE 66

66 proteins (Hsp) superoxide dismutase, and the redox-sensitive transcription factor Nrf2 (225, 226). In additio n ALC appears to restore the GSH/GSSG (reduced/oxidized glutathione) and decrease the formation of HNE, protein carbonyls, and MDA, indicators of lipid peroxidation (202, 210, 211 226228). ALC treatment decreased cytochrome c release, caspase 3 levels, a nd modulated the mitochondrial membrane potential, indicating it is also involved in regulating the apoptotic pathway ( 202, 229231). ALC appears to provide protection against lipid peroxidation and membrane breakdown and seems to be involved in the mainte nance and repair systems i n brain. ALC may be protective against oxidative stress through a reduction in tissue lactic acidosis, which leads to formation of ROS, through shifts in both the mitochondrial and cytosolic redox state, and/or through the inducti on of antioxidant genes ( 19, 226). This could lead to an increase in reducing power available for detoxification through the glutathione system. Neuromodulatory Evidence suggests ALC to be involved in various other neuromodulatory actions that may contribu te to its observed protection in brain. ALC has been shown to increase nerve growth factor production and binding (232, 233), as well as enhance Na+/K+ ATPase activity, which is particularly important in excitable cells ( 234, 235). Supplementation with ALC alters gene expression, specifically levels of transcripts involved in energy and mitochondrial metabolism (206, 225, 226, 236, 237) and is involved in the acetylation of transcription factors ( 238). In addition, ALC appears to a ffect neurotransmitter me tabolism through increasing acetylcholine synthesis and release ( 239, 240), increasing the production of gamma butyric acid (GABA), a major inhibitory neurotransmitter in the brain ( 241), and increasing dopamine release and binding ( 242)

PAGE 67

67 Protein modulatio n : ALC may prevent the age a ssociated changes to proteins. Oxidative modification to proteins appears to increase with age, as evidenced by increased levels of protein carbonyls, 3 -nitrotyrosine and 4 -hydroxynonenal (227). Supplementing ALC to the older ra ts in this study decreased many of these parameters in most of the brain regions measured. Additionally, proteins are altered during and after translation, most notably through phosphorylation, ubiquitination, glycosylation, methylation, acetylation and p almitoylation. ALC may modulate protein function and concentra tions by directed acetylation or by merely increasing the pool of available activated ac e tyl groups. One area of the cell for which acetylation plays a major role is histones (243), proteins inv olved in the activation and silencing of DNA through chromatin remodeling. If ALC can affect the acetylation of histones, it can effectively augment gene transcription. Additionally, ALC has been used in the clinical setting for the treatment of neuropathi c pain. In rodents the mechanism of action has been suggested by up -regulating the expression of metabotropic glutamate receptor 2 (mGlu2) in dorsal root ganglia (DRG) of the spinal cord (244, 238). A recent study further investigating the mechanism, found this up-regulation involves transcriptional activation mediated by nuclear factor p65/RelA by ALC (238). Further studies should examine ALCs mechanism of action and if it involves the acetylation of histones and/or o ther proteins involved in g ene transcription Clinical applications All of these roles and effects of ALC have led researchers to investigate carnitines involvement in a variety of neurological disease states and treatments, including autism ( 245), PD (246 ), Down syndrome ( 247), hypoxia ( 248), Huntingtons disease (249), cerebellar ataxia (250252), age associated mental decline (253), and ammonia neurotoxicity (254). A summary of ALCs neuroprotective mechanisms is listed in Table 2 10.

PAGE 68

68 Brain pathologies : W hile neurodegenerative diseases such as AD involve mitochondrial dysfunction and oxidative stress, they also involve reduced levels of synaptic transmission ( 255). Pre -synaptic terminals have a high number of mitochondria ( 256) and neurons specifically r ely heavily on mitochondria becau se of their high energy needs. Mitochondria generate both ATP and ROS and may, in turn, be vital in regulating neu rotransmission (255). Because of ALC and LCs role in increasing mitochondrial energetics and function, these molecules may also be playing a role in the prevention of neurodegeneration and the regulation of neurotransmission. ognitive decline seen with AD. suppress ed l evels of acetyl CoA and the activity of choline acetyltransferase in cell culture and ALC reversed these effects, but it did not change the mortality of the undifferentiated cells (257) In other cases of ALC was able to at tenuate the oxidative stress, ATP depletion, and cell death (222, 258). ALC may be acting through buffering of oxidative stress and maintaining energy levels. As far as the extent to which ALC improves clinical symptoms of AD, results are variable: some st udies have observed significant improvements in biochemical assays and psychometric tests (259); others have not observed such a difference on a large scale (260). Many of the possible roles of ALC in treating AD have been reviewed by Pettegrew et al (2000) (194). KT treatment for seizures: An increase in the glucagon:insulin ratio (from starvation and high -fat diets) has been found to increase the total carnitine concentration of the liver and this change has been correlated to the concomitant increase i n ketone production (261). However, infusion of exogenous carnitine does not further enhance ketone production. Under these conditions, carnitine may increase fatty acid flux through the carnitine acyltransferase to produce

PAGE 69

69 ketones (262), but it may also a ccumulate in the liver to accept activated acyl groups from CoA, thereby producing free CoA and acylcarnitines such as ALC (Figure 2 6). In states of starvation, the concentration of ALC increases in liver ( 261, 263). Fasting increases total carnitine conc entrations not only in the cytosol, but also in mitochondria, due, at least in part, to an increase in the Vmax of carnitine transport (262). When children on KT were monitored by subcutaneous microdialysis for acylcarnitines, the concentration of ALC was found to be increased ( 264). Since the concentrations of ALC are increased under these conditions, it is likely the liver is releasing ALC potentially to provide high -energy substrates to other parts of the body including the brain. If this is true, the l iver would have competing systems for produ cing mobile energy substrates: ketones and ALC. Whereas ketones require ATP hydrolysis to regenerate acetyl CoA, ALC does not. A study on 11-year old boys found a normal increase in blood acetoacetate after 12 an d 17 -hour fasts (265). When these boys were injected with DLcarnitine after 12 hours of fasting, the increase in blood acetoacetate was attenuated These results may indicate the presence of additional exogenous carnitine would enable the liver to produce more ALC which then prevented the increase in blood ketones since the acetyl groups of acetyl CoA could be directed towards producing ALC instead of ketones. In the brain of starved monkeys, the uptake of carnitine was relatively slow but the uptake of ALC was faster with rapid metabolism of the acetyl group ( 266) implying that ALC can act as an alternative energy source for the brain. A microarray study on the hippocampus from rats treated with KT found significant changes in specific sets of transcript s (73). Metabolic genes were up regulated and synaptic transmiss ion genes were downregulated. The number of mitochondria in the hippocampus was also found to be increased on KT. Overall brain energy reserves were increased on KT, synaptic

PAGE 70

70 transmission was resistant to low glucose concentrations and the seizure threshold of the rats was elevated. The authors believe mitochondrial biogenesis to be a major factor in the efficacy of KT. As previously mentioned, ALC has been found to up regulate mitochondrial transcripts in soleus muscle (206) as well as to also improve mitochondrial morphology and function (267), so it could potentially bolster the effects of KT on mitochondria. KT can be effective in reducing seizures in many different types of epilepsy (8, 10), which implies this therapy may activate an endogenous mechanism for reducing seizures. The use of ketones and acylcarnitines to meet the brains energy needs could be the basis for such a mechanism. The KT and the presence of ALC have similar and often overlapping results: -oxidation, mitochondrial biogenesis and increase d energy reserves (Figure 29). Many of the previously detailed neuroprotective functions of ALC could also have beneficial effects for epileptic patients, namely modulation of neurotransmitters, upregulation of Hsps, and protection against excitotoxicity. Since free radical concentrations and apoptosis may be elevated in states of increased fat metabolism (268), a free radical quencher with anti apoptotic effects such as ALC may be particularly useful. More research is needed, however, to determine ALCs role in KT and to assess the benefits of ALC supplementation. Conclusions and Future Directions Much of the research on carnitine and its esterified derivatives, such as ALC, has centered on the role of these molecules in metabolism. There is a body of research suggesting that ALC could be used to support new and multifactorial roles in neuroprotection. ALC can provide highenergy acetyl groups to metabolic pathways to improve the overall energy status of the brain and to alter the biosynthesis patterns of some neurotransmitters. These acetyl groups also have the potential to be involved in modulating proteins and gene expression. Other work has found ALC to improve mitochondria l function through improvements in membrane lipid content and enzyme

PAGE 71

71 activities. Pre -treatment of cells or animals with ALC before an excitotoxic insult can protect neuronal cells and treatment after such an insult can improve the condition of neurons. Figure 2 9 KT and ALC synergistic and/or additive effects on neuroprotection. Many of the KT mechanisms summarized here are not discussed in this review and have been reviewed previously. All of the functions of ALC make it a potentially beneficial mol ecule in the treatment of various neurological diseases, most notably AD and epileptic patients on KT. The aforementioned roles of ALC, metabolic and otherwise, could be part of the currently unknown mechanism behind KT. More research on use of ALC in the treatment of these diseases, including a supplementation regimen, is necessary to determine if these effects would translate in to positive changes in a clinical setting. Ultimately, carnitine and acylcarn itines are not merely

PAGE 72

72 -oxidation, but rather involved in many known and unknown functions in physiology. Metabolomics Finally, although the alterations in metabolism resulting from KT may change the acylcarnitine profile they may also lead to changes in the entire metabolome. A global approach to understanding all the changes in metabolites may lead to insight into mechanisms of action against seizures. Metabolism i s a fundamental facet of phenotype because it is a result of both genetic and en vironmental factors; therefore attention has been focused on a more global approach to the study of metabolism through the analysis of the metabolome ( 20). Metabolomics is the study of all small molecules (<1500 Da) or metabolites found in a spec ific cell, organ, or organism. The ome or omics began with genomics, the holistic study of genes, and now the methodology has been adapted to a new genre of research including proteomics, transcriptomics, and recently, metabolomics (269). The metabolome more clo sely represents the functionality of the cell as compared to the genome or proteome, because it examines how metabolite s change in response to changes in environment Concentration, fluxes, and identities of m etabolites are a result of the interaction of t he systems genome and proteome with its environment Metabolites are part of an integrated regulatory system while proteins only represent the end product of gene expression (20). Various signals regulate transcription factors which then act in the nucl eus to regulate gene expression. Changes in the transcriptome lead to changes in the proteome, and finally in the metabolome of a cell or tissue. When a transient perturbation alters the rate of formation or consumption of a metabolite, compensatory change s in enzyme activities return the system to steady state (270). The changes in metabolites themselves represent the actual functional changes along metabolic pathways in the cell rather than changes in the machinery that produces those

PAGE 73

73 metabolites. Changes in gene expression and protein expression may represent compensatory mechanisms, without a change in metabolite level. Additionally, in contrast to some biochemical approaches that often focus on single reactions or metabolites, metabolomics involves the collection of data on a wide array of metabolites in order to understand overall metabolism under the conditions of interest (271). Until recently analyses have only allowed selective profiling of compounds or fingerprinting changes in metabolites due to l ack of the analytical methods for identification and quantitation of metabolite levels (272). In the future metabolomics will provide a comprehensive metabolic profiling of individuals linked to the biological understanding of human integrative metabolism (273) Diseases are known to disrupt metabolism and lead to changes in metabolic profiles. As a result, metabolomics can lead to a more comprehensive understanding of mechanism of disease, to new diagnostic biomarkers, and ultimately contribute to both bio medical research and clinical practice (271). Although metabolomics is a fairly ne w concept, initial research has already been reported identifying changes in metabolomic profiles of several disease states including Huntingtons disease ( 274), Alzheimers disease ( 275), motor neuron disease ( 276), depression ( 277), schizophrenia ( 278280), cardiovascular and coronary artery disease ( 281, 282), hypertension ( 283), type 2 diabetes ( 284, 285), and several types of cancer (286, 287). Methodology In order to stu dy a large profile of metabolites with various concentrations, mass spectrometry (MS) is used most frequently due to its sensitivity ( 273). Nuclear magnetic resonance (NMR) is the other common analytical tool used in metabolomic research, however the ability for MS to be used with chromatographic methods makes it a preferred method (156). NMR may be used for compound v erification after MS analysis. While MS provides the

PAGE 74

74 sensitivity needed, it alone can only provide qualitative data and not accurately repre sent quantitative values without the use of internal standards and stand ard curves for each metabolite. To obtain a large profile of diverse molecules with varying concentrations, MS run in parallel with liquid chromatography (LC) is preferable because of its sensitivity ( 273). The chromatography separates the metabolites into discriminate classes that allows easier metabolite elucidation and provides relative quantitative data via retention peak areas. This approach has been used in the identification of h undreds of metabolites. Quantitation is important for identifying metabolites in a clinical setting and determining abnormal and normal levels in both healthy and disease states and requires use of internal standards and standard curves. When using chro matography it is important to choose the type of column as well as the correct ion mode for anal ysis of the desired compounds. Metabolomic studies of complex samples cannot be accomplished with the use of a single separation method due to the chemical dive rsity of the mixture. Monolithic reverse -phase columns allow for separation of nonpolar compounds while newer HILIC (hydrophilic interaction liquid chromatography) columns separate more polar compounds (288). The two types can be used together to achieve a wide spectrum of data for analysis Additionally, for the most comprehensive profiling, it is important to acquire the data in both positive and negative ion modes in metabolomics because some metabolites are mostly detected as positive ions, such as acyl carnitines and amino acids, and some are detected as negative ions, such as fatty acids (288, 289). The most difficult challenge in metabolomics is the overwhelming amount of data and complex data processing and developing the technology to provide powerful data analysis capabilities ( 290). Several different software programs and statistical tools are required for the data processing and qualitative analysis in order to generate a list of masses from MS provided

PAGE 75

75 data. However standard program inputs must be developed for accurate comparison of results due to the ability of these programs to stratify and analyze data differently according to the settings. In addition because this approach is relatively new, using a comprehensive database for metabolite identi fication, similar to what is done for genomics and proteomics, has be en challenging for researchers. Currently, after the lists of masses or m/z have been created from MS, it is a manual process to search for the mass in a metabolite database, such as Huma n Metabolome Database (HMDB) or ME TLIN within a set error range. Also, because these databases ar e still fairly new, data are not available for every metabolite including relevant biological information on the metabolites or normal ranges in human tissues. The HMDB is currently the most comprehensive and contains the most endogenous metabolites with approximately 6000 ( 291) and METLIN provides the capability to identify metabolites based on known chemical and physical properties. Application in Nutritional Science Phenotype is influenced by both intrinsic and extrinsic factors such as diet, environment, genetics, behavior and stress, physical activity and medications All of these combined represent significant factors to consider in the study of metabolomic s. Nutritional status can affect nearly all body functions and therefore is important to consider in al l states of health and disease. Alterations in dietary intake can result in metabolic imbalances and modulate disease risk. Current research has focused on diseases such as cardiovascular disease, obesity, type 2 diabetes, inflammatory and gastrointestinal disorders; however, diet influences phenotype along the entire continuum from optimal function and health to disease and dysfunction ( 292). For example, metabolomics research using NMR technology and statistical analysis have identified changes in the metabolome in urine of healthy s ubjects on a vegetarian, low or high meat diet ( 293).

PAGE 76

76 Nutritional science is multidisciplinary and includes studying metabo lism in animals and humans in both health and disease and therefore provides an ideal opportunity to integrate the field of metabolomics Many disciplines are needed; however, the knowledge of biochemistry, physiology, and nutrition combined is necessary for acquiring and analyzing complex data sets and applying biologically relevant conclusions. A major focus of nutritional scientists in metabolomics currently is on ensuring the technology is developing and the databases are configured with the right too ls and organization for identifying the key metabolites and samples that are needed for analysis of dietary aspects of health and disease (294, 295). This involves collaboration across multiple disciplines. For example, both fasting and post -prandial effec ts on metabolites in the same individual consuming a controlled diet will be useful in metabolomic databases. Additionally, nutritional scientists are needed for the translation across disciplines and then for the translation into the clinical and public domain ( Figure 2 10). Figure 2 10. Metabolomics in translational nutritional sciences. Adapted from Zeisel SH et al. 2005. Metabolomics is a multidisciplinary field in which nutrition is important across all steps of metabolomics analyses, from collecti on and understanding to translation to both the clinical and public domains.

PAGE 77

77 Prevention of disease Understanding individual metabolomic profiles in relation to dietary intake as well as health and disease will change clinical practice Most of the resear ch on dietary requirements assumes all persons within an age range metabolize the nutrients the same. Human genetic a nd phenotypic variation is wide spread ; therefore optimal dietary recommendations for one individual may predispose another to disease ( 273). Through integration of knowledge and metabolomics technology, requirements can be based on individual genetic and/or metabolomic profiles. In addition, research into diet as it relates to health and disease often result in large vari ance and inconclusiv e results. Research into the metabolomic profiles can be used to differentiate the responders from nonresponders in order for recommendations to be more accurately targeted to individuals ( 295). It is important to both understand nutrition and metabolism a s well as disease pathology in order to interpret the metabolome and provide meaningful prevention and/or treatment recommendations. Treatment of disease In addition to nutrition related prevention of disease, modulations to dietary intake have also been u sed in the successful trea tment of a number of diseases. Some dietary modifications are made based on known metabolic pathways and mechanism of disease in the case of inbor n errors of metabolism such as g lycogen storage di sease and p henylketonuria. Others are used based on epidemiological evidence correlating nutrient intakes with disease outcomes for example, the use of omega 3 fatty acids in cardiovas cular and Alzheimers diseases. The use of metabolomics will open the field of nutritional science and all ow the study of mechanism of action in diet related treatments that are unknown. One example of the use of nutrition to alter energy metabolism in the brain to treat disease is KT.

PAGE 78

78 Use of Metabolomics in the KT Population Metabolomics can be grouped into t hree main analytical approaches. As defined by Fiehn in 2002, there is 1) targeted metabolomics involving quantifying specific metabolites that is hypothesis driven, 2) metabolic profiling to quantify metabolites belonging to a particular group or metaboli c pathway, or 3) untargeted metabolomics designed to view all of the metabolites in a sample, more of a hypothesis generating approach (272). Two of these approaches were used in determining alterations in compounds resulting from initiation of a ketogenic diet (Figure 2 11). Figure 2 11. Metabolomics workflow.

PAGE 79

79 Untargeted metabolomics Major alterations in macronutrient composition in the diet would be expected to alter levels of metabolites in serum. KT involves a radical shift in metabolism from the u se of primarily glucose to the use of fatty acids and ketones (more detail provided in the first section of Chapter 2). Understanding these alterations in metabolites may provide insight into mechanism of action against seizures. Untargeted metabolomics is a hypothesis generating approach in which computer software is used to identify major changes in metabolites in biological fluids or tissues in order to develop hypotheses concerning mechanism of action. To our knowledge, an untargeted metabolomics approa ch to identify major changes in compounds in subject s receiving a ketogenic diet has not been published in the literature. Targeted metabolomics In addition to untargeted metabolomics, a targeted approach is needed to identify quantitative changes in spe cific metabolites in order to test an existing hypothesis Based on an earlier discussion in this chapter of ALCs role in energy metabolism and neuroprotection, a hypothesis of increased concentrations of ALC in patients on KT was developed. Methods for detection and quantif ication of carnitines have varied. The determination of individual acylcarnitines is challenging due to the wide variation in structures from small polar molecules such as ALC, to large non -polar molecules in which a long chain fatty acid is esterified to carnitine. Three major limitations exist in methodology: 1) preparation of butyrl esters partially hydrolyzes acylcarnitines, 2) isobaric nonacylcarnitine compounds yield false positive results, and 3) constitutional isomers cannot be distinguished (296). These challenges have been addressed using a variety of schemes involving chromatography, derivatization, and radioisotopic exchange Development of electrospray ionization/ion trap MS coupled to HPLC has allowed a more sensitive and selective approach for detecting acylcarnitines with the ability

PAGE 80

80 to distinguish isomeric acylcarnitines and to allow for complex mixtures such as huma n plasma (297). The UF Metabolomics Core has developed methods for detection of acylcarnitines involving H PLC -ESI coupled to TOF MS using dual chromatographic methods.

PAGE 81

81 Table 2 8 Inborn errors of metabolism detected by acylcarnitine profile analysis: disorders of fatty acid oxidation and organic acid metabolism. Fatty acid oxidation disorders Acylcarnitine Changes Carnitine uptake defect Carnitine palmitoyltransferase I (CPTI) deficiency Carnitine acylcarnitine translocase (CACT) deficiency C16 with C18:2, C18:1, C18 Carnitine palmitoyltransferase II (CPTII) deficiency C1 6 with C18:2, C18:1, C18 Very long chain acyl CoA dehydrogenase (VLCAD) deficiency C14:1 with C14, C14:2 Long chain L 3 hydroxyacyl CoA dehydrogenase (LCHAD) deficiency C16 OH with C16:1 OH, C18:1 OH, C18 OH Trifunctional protein (TFP) deficiency C16 OH with C16:1 OH, C18:1 OH, C18 OH Medium chain acyl CoA dehydrogenase (MCAD) deficiency with C6, C10, C10:1 Medium/short chain L 3 hydroxyacyl CoA (M/SCHAD) dehydrogenase deficiency OH and Inc C10 OH Medium chain 3 ketoacyl CoA thiolase ( MCKAT) deficiency C10 OH Short chain acyl CoA dehydrogenase (SCAD) deficiency Multiple ETF, ETF QO deficiency) (Glutaric acidemia type II) (with C5, and other longer chain species) Dienoyl CoA re ductase deficiency Organic acid metabolism disorders Acylcarnitine Changes Propionyl CoA carboxylase deficiency (Propionic acidemia) Multiple carboxylase deficiency (Holocarboxylase synthetase deficiency and Biotinidase deficiency) C5 OH Methyl malonyl CoA mutase deficiency (Methylmalonic acidemia) Disorders of cobalamin metabolism (Cbl A/B/C/D/F deficiencies) Succinyl CoA synthetase deficiency (SUCLA2) C4DC Isobutyryl CoA dehydrogenase deficiency Ethylmalonic encephal opathy Ketothiolase (2 methylacetoacetyl CoA thiolase, or 3 oxothiolase) deficiency C5 OH with C5:1 Isovaleryl CoA dehydrogenase deficiency (Isovaleric acidemia)

PAGE 82

82 Table 2 8. Continued. Organic acid metabolism disorders Acylcarni tine Changes 3 Methylcrotonyl CoA carboxylase (3 MCC) deficiency C5 OH 3 Hydroxy 3 methylglutaryl CoA lyase (HMG CoA lyase) deficiency C5 OH with C6DC Malonyl CoA carboxylase deficiency C3DC Glutaryl CoA dehydrogenase deficiency (Glutaric acidemi a type I) C5DC Glutamate formiminotransferase deficiency (Formiminoglutamic aciduria) (with more prominent peak at m/z 287) Adapted from Rinaldo R et al. Acylcarnitine profile analysis. Genet Med 2008:10(2):151 156 (162) C0=free carnitine; C3= Pr opionyl -carnitine; C4= Butyryl carnitine, Isobutyryl -carnitine; C5= Isovaleryl -carnitine, Methylbutyryl -carnitine; C4 OH= 3 Hydroxybutyryl -carnitine; C5 OH= 3 Hydroxyisovaleryl -carnitine, 3 Hydroxy 2 -methylbutyryl -carnitine; C8= Octanoyl -carnitine; C3DC= M alonyl -carnitine; C4DC= Succinyl -carnitine/Methylmalonyl -carnitine; C5DC= Glutaryl -carnitine; C10 OH=3 Hydroxy decanoyl -carnitine; C14:1=Tetradecenoyl -carnitine; C16=Palmitoyl -carnitine; C16 OH=3 Hydroxypalmitoyl -carnitine.

PAGE 83

83 Table 2 10. Summary of ALC ne uroprotective mechanisms. ALC Neuroprotective Effects Energy metabolism Incorporation into acetyl CoA and lipids (Farrell et al 1986;Ricciolini et al 1998) Normalize high energy phosphates (Pettegrew et al.1995 ) Normalize lactate and pyruvate (Aureli et al. 1990, 1994; Rosenthal et al 1992) Increase cellular consumption of oxygen (Hagen et al. 1998) Reduced glucose oxidation and increased glycogen synthesis (Aureli et al.1998) Changes in enzymes in TCA cycle ETC, and amino acid metabolism (Gorini et a l. 1998) Membrane Protection of mitochondrial membrane (Ames 2004 Review; Aureli et al. 2000; Virmani et al 1995, 2001) Restoration of cardiolipin (Hagen et al. 1998) Synthesis of n3 PUFA in membrane Deacylation/reacylation of membrane (Virmani and Bini enda 2004; Arduini et al. 1992,1994; Virmani et al 1995) Increase membrane fluidity (Arienti et al. 1992) Protection from excitotoxicity Protection from respiratory chain complex inhibitors (Virmani et al 1995; Virmani 2004; Bodis Wollner et al 1991; Ma zzio et al 2003) Anti oxidant Increase in expression of HO 1 (Calabrese et al 2005) Upregulation of Hsps, SOD, Nrf2 (Calabrese et al.2005;2006) Restored GSH/GSSG (Calabrese et al. 2005;2006) Decrease formation of HNE, protein carbonyls, and MDA (Calabrese, et al. 2006;Poon et al.2006; Hagen et al.1998); Virmani et al 1995, 2001; Lowitt et al 1995) Reduced levels of proinflammatory cytokines (Winter et al 1995) Anti apoptotic Decreased cytochrome c release and caspase 3 (Furono T et al 2001; Pillich et al. 2005; Ishii et al.2000) Changes in mitochondrial membrane potential (Hagen et al.1998) NGF and Nerve Regeneration Increase NGF production and NGF binding (Furlong et al. 1996; Foreman et al 1995) Effects on Na + /K + ATPase activity (Sima et al 1 996; Lombardo et al 2004) Protein modulation Acetylation of transcription factors (Chiechio S et al 2006) Gene modulation Up regulate VD 3 3 pr otein, Hsp72 (Calabrese et al.2004; Traina et al. 2004, 2006) Down regulate mitochondrial P3 of ATP synthase lipid binding protein (Traina et al. 2004) Increase in transcripts related to mitochondrial biogenesis in skeletal muscle (Cassano et al. 2006) NT modulation Increase in ACh synthesis and release (Ando et al 2001; Imperato et al 1989) Influences cholinergic neurotransmission (Furlong et al. 1996; Katarzyna et al 2004; Calvani et al.1992) Increase dopamine release and recep tor binding (Sershen et al 1991) Increase in GABA (Fariello et al 1988) Reduce GABAergic inhibitory postsynaptic currents (Bahring R et al 1994)

PAGE 84

84 CHAPTER 3 METHODS Retrospective Study Study Design A retrospective chart review was performed from records of patients started on KT at the University of F lorida from 1995 through 2006. A Microsoft Access database was designed and approved by the University of Florida (UF) Institutional Review Board 01 (IRB 01) to evaluate the KT population at Shands at UF. Sh adow charts from the KT patients were entered into E xcel for patient care purposes and the data w ere audited and then transferred into the da tabase and then audited again. Queries wer e designed to retrieve the data. The parameters analyzed for growth were height (Ht), weight (Wt), BMI, H t z score, Wt z score, and BMI z score. Diet prescription parameters analyzed were Calories (Cal), protein (Pro), Cal and Pro per kilogram body weight, % of recommended calories and protein, Fat, carbohydrate (CHO), and Pro % of calories, Fat and CHO per kilogram body weight, and ratio. Laboratory parameters analyzed were -hydroxybutyrate (BHB) in mg/dL and the following blood lipid parameters: triglycerides (TG), total cholesterol (Total Chol), HDL cholesterol (HDL), non HDL cholesterol (non HDL), LDL cholesterol (LDL), in mg/dL and Total/HDL cholesterol (Total/HDL) Patients were stratified according to time on diet (0 6, 6 12, 12 24, 2436, 3648, 4860, >60 months for growth and i nitiation, <4, 4 12, 12 24, 2436, 3648, 48 60, 6072, 7284, and 84100 months for lipids ) and the values were a veraged for each pat ient within each time period. Patients were also stratified according to ambulation status, route of feeding, age, gender, and diet prescription and groups were compared. For growth and lipid analysis with diet prescription only patients who were on diet f or at least 1 year were included.

PAGE 85

85 Calculations Z scores: Weight, height, and BMI z scores were calculated using a Macro developed in Excel based on data from reference populations. For males and females betwee n 2 and 20 years of age data were used from ref erence ranges for normal pediatric growth defined in the 2000 CDC growth charts (298). Additionally, length or height and weight z scores were calculated for 0 to 36 months of age based on reference ranges from the CDC growth curves. However, for BMI z sco res for males and females from 0 to 36 months calculations were made using the World Health Organization guidelines (299). For patients older than 20 years of age, z scores were calculated using the National Health and Nutrition Examination Survey I and II sample population ranges adjusted for age, gender and race (300). Recommended calories: The amounts of recommended calories were calculated based on the equations provided in the 2005 Dietary Reference Intake (DRI) by the Institute of Medicine (301) base d on age, gender and activity level. Due to the retrospective nature of the data, assumptions had to be made for activity level. Assumptions were made based on ambulation status as follows: n onambulatory patients were assumed to have a sedentary activity level and ambulatory patients were assumed to have a low active activity level. Calories were calculated using the DRI formulas and actual weight, height, age, gender, and activity level. Calories were then divided by weight to obtain Cal/kg/day. Actual Ca l/kg/day was divided by the recommended Cal/kg/day and multiplied by 100 to determine % of estimated energy requirements ( % EER). Z-tile: A blood lipid z tile was designed to compare lipid levels to the baseline of the population or to the average lipid l e vels at initiation. It is calculated based on Equation 3 1. Z -tile = (value average of population at initiation) (3 1) standard deviation of population at initiation

PAGE 86

86 Statistics The Student t test was used to compare groups assuming unequal variances and the Pearson correlation test was used to determine r elationships among parameters. All analyses were performed using StatTools (Palisade Coporation, Version 1.10). P values less than 0.05 were considered to be significant. The results are exp ressed as mean standard deviation. Limitations and Strengths This was a retrospective evaluation and therefore the data have limitations and should be interpreted accordingly. The blood parameters were obtained at different laboratories and we are unabl e to ensure fasting values. In particular, TG and LDL concentrations are affected the most by the ingestion of a meal ( 302) Additionally, height and weight were obtained at different centers and by different people. In this population, length measurements can be di fficult to obtain and therefore, consistency would be important for accuracy. Finally, the calculated recommended calories are based on actual height and weight measurements and are not taking into account catch up growth which may be needed in a large percentage of this population. The strengths of the retrospective study are the large number of patients an d the length of time. This allow ed for a better understanding of effect of time on diet. Prospective Study Study Design Informed consent was obtained and patients were e nrolled into an IRB 01 and Scientific Advisory Committee of the GCRC approved prospective study in the General Clinical Research Center (GCRC) at Shands at UF. This study was designed as a natura l history study with two strata. Stratum 1 patients were nave patients in which KT had not been initiated. These patients were referr ed for initiation onto the diet. Stratum 2 patients were experienced patients already receiving KT for varying periods of time. Stratum 1 attended a pre in itiation education session

PAGE 87

87 prior to initiation of KT in which baseline growth and blood parameters were obtained. Patients were admitted to the GCRC for initiation of the diet beginning at full calories and at a ratio of 1:1. The r atio was gradually increa sed over a 2 4 d ay period with the goal of a 4:1 ratio. During the first year on KT, patients were seen i n outpatient clinics at 1, 3, 6, 9 and 12 months post initiation. After the first year, they attended clinic visits every 4 months for the second year. Experienced patients attended outpatient clinic visits every 4 months for 2 years. Height and weight measurements were performed at pre -initiation, initiation, and at each subsequent clinic visit. Fasting blood was obtained by the GCRC study nurses at ini tiation and at each follow up clinic visit for analysis of blood lipids and BHB analysis. One time per year, patients were admitted for a 24 hour EEG and video monitoring. During this visit, a heparin lock was inserted for multiple blood draws for the anal ysis of BHB, glucose (Glu), and lactate (Lac) at fasting, 30 minutes, 1 hour, and 2 hours after the ingestion of a ketogenic meal. Data were entered in to Excel for patient care purposes and audited. Data were then entered into a Microsoft Access Database for population analysis and audite d again for copy/paste errors. Queries were designed to retrieve the data. The parameters analyzed for growth were Ht, Wt, BMI, H t z score, Wt z score, and BMI z score. Diet prescription parameters analyzed were Cal, Pro, Cal and Pro per kilogram body weight, % of recommended calories and protein, Fat, CHO, and Pro % of calories, Fat and CHO per kilogram body weight, and ratio. Laboratory parameters analyzed were serum BHB in mg/dL and the following blood lipid parameters: TG, Total Chol, HDL, non HDL, LDL and Total/HDL cholesterol in mg/dL. Measurements All measurements were obtained by the same person at the GCRC or Shands hospital. For ambulatory patients, height was obtained without shoes using a wall -mounted stadiometer measured to 0.1 cm and weight was measured to 0.1 kg using a digital scale in the GCRC. For

PAGE 88

88 non ambulatory patients, weight was obtained using a wheelchair scale at Shands hospital; then the wheelchair weight was subtracted, or children were weighed while being held by a caregiver; then the caregiver was reweighed on the same scale without the child. The two weights were subtracted to obtain the actual weight of the patient. Length was obtained for nonambulatory patients by positioning the patient on a fl at surface. Assistance was needed to hold the head, legs, and back straight with legs fully extended and toes pointed upward. A mark was made at the crown of the head and at the heel. Using a flexible tape, the length was obtained by measuring between the two marks. Segmental length using a flexible tape and a clip board situated at the head was used for a few patients with spastic quadriplegia or severe scoliosis. Four measurements were taken from: 1) top of head (cranium) to shoulder bone (acromion proces s), 2) shoulder bone to hip bone (iliac crest), 3) hip bone to knee bone (patella), 4) knee bone to the heel bone or sole of the foot (calcaneus), the sum of all the segments was used to obtain length. Calculations Z scores: Weight, height, and BMI z score s were calculated based on a reference population for age and gender using a Macro developed in Excel (see detail in Z scores section of Retrospective Study Methods). Recommended calories: Recommended calories were determined based on the DRI for age and sex Actual height and weight were used in the DRI formula based on age, gender and activity level for patients who had a height and weight above the 5th percentile (and weight below the 95th). If a patients height and/or weight fell below the 5th percent ile, recommended Cal/kg/day were determined based on age, gender and activity level at the 50th percentile. If a patients weight was above the 95th percentile, ideal weight was used. For patients 19 years or older, actual height and weight were used. Calo ries were then divided by the weight used to

PAGE 89

89 obtain Cal/kg/day. Actual Cal/kg/day was divided by the recommended Cal/kg/day and multiplied by 100 to determine % of estimated energy requirements ( % EER). Blood Analysis Serum BHB, Glu, and Lac w ere assayed in triplicate by GCRC lab personnel using whole blood. BHB was assayed using the CardioChek System (Polymer Technology Systems, Inc. Indianapolis, Indiana, USA). Glu and Lac were assayed using the YSI 2300 STAT PlusTM Glucose and Lactate Analyzer (YSI Life Sciences, Inc., Yellow Springs, Ohio, USA). Statistics The Student t test was used to compare groups assuming unequal variances and the Pearson correlation test was used to determine r elationships among parameters. All analyses were performed using StatTo ols (Palisade Coporation, Version 1.10). P values less than 0.05 were considered to be significant. The results are expressed as mean standard deviation Limitations and Strengths The major limitations for the prospective study were the small number of pa tients and the patients in the experienced Strata were on diet for varying lengths of time. Strengths include the prospective design and all measurements and lab parameters were taken consistently and at the same place. In addition, the nave strata of pat ients provided baseline data in order to calculate changes from baseline. Metabolomics Due to the new development of the Clinical and Translational Science Institute (CTSI) at the University of Florida, an interdisciplinary group of researchers collaborate d to accomplish the initial stages of identification of metabolites in response to a ketogenic diet. As discussed in Chapter 2, because metabolomics is so complex it requires many specialties to obtain, process and interpret the results. A workflow was est ablished for the nutritional scientists to be involved

PAGE 90

90 from the beginning of study design and sample preparation through data analysis and interpretation in order to develop the most accurate conclusions (Figure 3 1). Because each specialty is most interes ted in the separate sections of the process, it is important to have continuity throughout the study. It is vital to have an understanding of each specialty to interpret the results as accurately as possible. Plasma samples were collected and prepared by o ur lab, data was analyzed and processed in the Chemistry department and compound identification and analysis was a combined effort by both groups. Figure 3 1. Workflow of multidisciplinary approach and continuity of metabolomics based analyses To bett er understand the metabolome of patients receiving KT, metabolomics based analyses were performed on samples from healthy subjects before and after ingestion of a ketogenic diet. Future studies will be able to use the system that has been developed and com pare the results to samples obtained from patients with intractable epilepsy before and after initiation onto KT. The following are the detailed methods from the study of healthy subjects (KetoMed study) for preliminary metabolomics results.

PAGE 91

91 KetoMed Stud y Design This study was conducted in the GCRC at Shands at UF by first year medical student volunteers as part of a larger study evaluating dyslipidemia and insulin sensitivity in which results were previously published (125). The study was approved by the Scientific Advisory Committee of the GCRC and the IRB at Shands hospital and students served as volunteer subjects or investigators and all gave written informed consent before participation. Twenty young, healthy subjects (10 males/10 f emales) were screened for contraindications by history and physical examination. Subjects were excluded if they were pregnant, vegetarian or an endurance athlete, or if they had cardiovascular disease, diabetes mellitus, neurological diseases, kidney disease, or any food al lergies, or if on any medications. Subjects were matched for gender and BMI and ra ndomized by parallel design into two groups (Figure 3 2). Figure 3 2 Diagram of KetoMed study design.

PAGE 92

92 Subjects were initially admitted to the GCRC for baseline testin g. After an overnight fast, venous blood was obtained for analysis of the acylcarnitine and metabolomic profile. All blood samples were obtained in duplicate and stored at 80 degrees C until analyzed by the UF Metabolomics Core Laboratory Following base line measurements, subjects were administered either a SFA -enriched or PUFA -enriched ketogenic diet for 5 days. Both diets were 70% fat, 15% CHO, and 15% Pro which is equivalent to a 2.33:1 ratio of KT. The fat of the PUFA diet consisted of 60% PUFA, 15% SFA, and 25% MUFA and the fat of the SFA diet consisted of 60% SFA, 15% PUFA, and 25% MUFA. During the 5 days, subjects were limited to the food provided by the GCRC. The diets were individualized based on a nonconsecutive 4 day diet history obtained before enrollment and designed to be weight maintaining. Weight and diet were monitored daily and adjusted if needed to maintain consistent weight. The morning of the last day, subjects were admitted again to the GCRC to obtain fasting blood samples. Blood Ana lysis Fasting blood samples were obtained by GCRC nurses and placed on ice or in a freezer during all times through out sample preparation. Whole blood was centrifuged at 4750 XG and 4 6 degrees C from 15 minutes to isolate plasma and transported on ice and stored at 20 degrees C until further analysis. Sample preparation Samples were pr epared by the KT research staff. A solution of acetonitrile and methanol in a ratio of 3:1 was prepared for sample dependent extraction and protein precipitation. A detailed standard operating procedure for the human blood assay is located in Appendix C. Day 1, 100L of a 3:1 acetonitrile:methanol solution was add ed to the microcentrifuge tubes Plasma samples

PAGE 93

93 were thawed and vortexed before adding 1mL of plasma to each micro centrifuge tube containing the 3:1 mixture. Samples were kept on ice throughout the procedure and then stored overnight at 20 degrees C. On day 2, samples were centrifuged at 12,400 x g for 15 minutes and the supernatant was transferred to a new microcent rifug e tube with a Pasteur pipette. This step was repeated twice more to ensure particulates were not in the sample. Finally, samples were stored overnight at 20 degrees C and then evaporated under nitrogen the next day Analytical methods The analytical methods were conducted by the UF Metabolomics Core of the Clinical and Translational Science Institute (CTSI) by the Chemistry Department at the University of Florida in the Mass Spectrometry laboratory of David Powell, PhD. D ual monolithic and hydrophilic interaction liquid chromatographic approaches coupled to a high resolution time -of -flight mass spectrometer operating under positive ESI conditions were used (HPLC ESI TOF -MS) Specifically, the Agilent 1200 Series Rapid Resolution LC System (Agilent Tec hnologies, Inc., Santa Clara, California) the Agilent 6210 Time -of Flight mass spectrometer configured for ESI (Agilent Technologies, Inc., Santa Clara, California) was used. Due to the complex ity and chemical diversity of the samples, dual chromatographic techniques were used to separate polar and non -polar compounds. Traditional r everse -phase chromatography using two C18 monolithic columns (Phenomenex ONIX Monolithic Column-C18, 4.6 100 mm) in series allowed for the separation of larger hydrophobic comp ounds while the HILIC (hydrophilic interaction liquid chromatography) column (Phenomenex LUNA HILIC Column, 3 m, 2.0 150 mm) was used to separate the smaller, polar molecules. In the monolithic column, the mobile phase flowed at 1 mL min 1 and consist ed of water with 1% acetic acid and of acetonitrile with 1% acetic acid. The HILIC mobile phase had a flow rate of 0.3 mL min 1 and consisted of 7.5 mM ammonium

PAGE 94

94 formate in water and 7.5 mM ammonium formate in acetonitrile. All samples were separated into 3 replicates. Data processing Molecular feature extraction and MassHunter software ( Agilent Technologies, Inc., Santa Clara, California ) were used to generate retention time, mass pairs, and abundance sets called feat ures. Data were organized into E xcel sp read sheets for analysis in MetaboAnalyst, a freely accessible web server for metabolomics data. MetaboAnalyst was created by collaboration between the University of Alberta, The Alberta Ingenuity Fund and The Human Metabolome Project in order to provide a web based analytical pipeline for high throughput metabolomics data analyses (303). Data including m/z, retention time (rt) and abundance or intensity were imported to Metabo A nalyst for data alignment, normalization and statistical analyses. Hierarchical clustering was used for quality control to determine reproducibility of the biological replicates. Normalization of data was done to limit technical variation among replicates. ANOVA and fold -change filtering were used to determine features with statistica lly significant differences between experimental groups. Lists of provisional masses were created for searching in METLIN and HMDB. Metabo A nalyst : Data were converted from E xcel to comma separated values (csv) and then zipped before uploading. Within Meta boAnalyst, MS peak list, i ntensity and retention time were uploaded as two separate zipped files, before and after. In order to perform paired analyses, a text file including pairing information was also uploaded. Next peaks were matched and aligned based on their mz and rt and mz and rt of each peak was changed to the group median values. If a sample had more than one peak in a group, they were replaced by the sum. Peaks were excluded if they did not appear in at least half of both groups. Then a data inte grity check was performed to ensure all information is collected. If missing or zero values were found, they

PAGE 95

95 were replaced with a small value equal to the half of the minimum positive value in the original data. This is done based on the assumption that mo st missing values are caused by low abundance. Next the data were stored as one sample per row and one variable (peak o r metabolite) per column and were normalized per row and column. For row -wise normalization, data was normalized by sum; for col umn -wise normalization, data were normalized by unit scaling (mean -centered and divided by the standard deviation of each variable). Finally univariate analysis methods including fold change, t test, and volcano plot analysis were performed on the data to determine significant differences among groups. For fold change analysis, the data before column normalization was used to compare absolute value changes between the two group means. Results were plotted in log2scale so that both up and down regulated changes had t he same distance to the zero baseline. Volcano plot analysis combines the fold change and t -test analyses to determine the most significant features. Partial least squares discriminant analysis (PLS DA) was used to extract via linear combination of origina l variables (X) the information that can predict the class membership (Y). Additionally, both hierarchical clustering and partitional clustering were performed. MetabAnalyst is not currently designed to accommodate replicate data, therefore all data was u ploaded and then processed data w ere downloaded and replicates were averaged for each peak group formed. The averaged data were then re uploaded into MetaboAnalyst for further analyses. Additionally, subjects were separated into SFA and PUFA and uploaded s eparately, as well as the averaged data for both groups. Manual : Due to large amount of processing and normalization that takes place within metabolomics based softwares, targeted analysis was performed manually. Additionally, HMDB and METLIN are still in the initial stages and do not include all the possible acylcarnitines. A list

PAGE 96

96 was creat ed by our lab, which includes functional group locations for 77 documented and undocumented acylcarnitines. The fragmentation patterns wer e predicted by assuming the removal of the acyl group, quaternary ammonium group, carboxylic end, and water to leave the daughter ion m/z ratio of 85 characteristic of many acylcarnitines. After raw data were organized into E xcel files, a formula was writt en in excel to identify carnitines by m/z in the raw data based on a tolerance of 5 10 parts per million (ppm ) (Table 3 1). All provisional carnitines identified within each sample and replicate were compared across replicates and study groups, before and after, SFA and PUFA. Targeted analysis: Additional analyses were performed by Soledad Cerutti, a PhD student in the Chemistry department and within the UF Metabolomics Core using internal standards available for free carnitine a nd 10 acylcarnitines and q ua ntitative analysis software from Aglient (Agilent Technologies, Inc., Santa Clara, California). Comparisons were made for total peak area, or abundance between before and after diet and males and females. Compound identification Provisional masses were ent ered into METLIN and HMDB with a tolerance of 5 10 ppm (Table 3 1). The METLIN database provides information on metabolite mass, CAS #, some LC/MS spectra, and provides the ability to be interfaced directly with the programs that generate the feature table s for rapid and easy searching ( 290). HMDB on the other hand requires manual input for identification, but also provides important but limited structural and biochemical information regarding the compound ( 291). A tolerance limit of 5 10 ppm was set for se archin g HMDB based on the accuracy of the instrument used (Table 3 1). Compounds identified were copied into an E xcel spreadsheet along with HMDB ID, common name of molecule, chemical

PAGE 97

97 formula, molecular weight (Da), associated biochemical pathway, and location i n biological tissue or fluid. Table 3 1. Tolerance ranges used for searching the Human Metabolome Database. Masses were searched using the corresponding MW tolerance, which gives a tolerance of 5 10 ppm. Mass range MW tolerance for HMDB (Da) Averag e tolerance (ppm) 50 100 0.0005 5 10 100 200 0.001 5 10 200 300 0.002 5 10 300 400 0.003 5 10 400 500 0.004 5 10 500 600 0.005 5 10 600 700 0.006 5 10 700 800 0.007 5 10 800 900 0.008 5 10 900 1000 0.009 5 10 1000 1100 0.010 5 10 1100 1200 0.01 1 5 10 Limitations and Strengths The study design was a randomized, prospective study conducted in the GCRC, w h ere the food was provided and monitored and desi gned to be weight maintaining. However, these were healthy subjects, not patients with epilepsy and the diet provided was at a lower fat ratio compared to most patients on KT. Nevertheless, other investigators have published previously on these blood samples and the blood lipid response to the two ketogenic diets is similar to results in the KT pati ent population (125). Additionally, only ESI under positive conditions were analyzed for the healthy subjects on a ketogenic diet. The negative ions as well as the samples from patients with intractable epilepsy before and after diet initiation will be an alyzed at a later date.

PAGE 98

98 CHAPTER 4 RESULTS Figure 4 1 presents a project overview and serves as a guide for the results. KT has been used for decades and is an example of a nutritional therapy used in the treatment of disease Recently KT has been used to treat a variety of neurological disorders. This research focuses on the use of KT in a specific population, patients with intractable epilepsy. Whereas, neuroscientists think about KT in terms of its effects on seizures, within the nutritional sciences co mmunity it is important to also think about how these nutrients may cause potential adverse effects because not only are these nutrients used for seizure control, but they are needed for the health and growth of the child. Therefore, this dissertation rese arch addresses both potential adverse effects of KT and potential mechanisms of action and includes: preliminary studies testing dietary fatty acid profile and dyslipidemia; retrospective analysis to document growth and dyslipidemia parameters in this stud y population; design and implementation of a prospective Figure 4 1 Overall project and summary of specific aims included in data analysis and results.

PAGE 99

99 study to more accurately examine dyslipidemia and growth and to begin to also address mechanism of action; and design of an interventional study to test potential mechanism s of action and make improvements to therapy. Anthropometrics and Growth Retrospective R esults Retrospectively, 106 patients were on diet for at least 1 year and were initiated at Sh ands at UF from 19952006; 3 patients were started on the diet twice. The median age of patients started on KT was 7.05.3 years and there were 57 males and 49 females. Time on diet Height (Ht) z score became progressively worse over time on diet with the most significant difference compared to 0 6 mo found at >60 mo on diet (Figure 4 2). The only significant change from baseline for w eight (Wt) z score was seen at >60 mo on diet (Figure 4 3). There were no significant differences compared to baseline for BMI z score, however there was a trend for increasing BMI z scores with time on diet (Figure 4 4). Regression analys e s w ere performed on data averaged per patient for z scores versus time on diet. Ht z score showed a significant negative correlation with time on diet (r= 0.5 r2=0.2, p<0.0001) (Figure 4 -5 ), however Wt (Figure 4 6 ) and BMI z score (Figure 4 7 ) over time were not statistically significant. In order to determine if age, or potentially length of epilepsy had any effect, patients were stratif ied according to time on diet and age and regression analysis was performed for 0 6 mo on diet, assuming no dietary effect had yet occurred. A significant correlation was found with increasing age and decreasing Ht, Wt, and BMI z score (r= 0.3, p= 0.0 1 ; r = 0.2, p= 0.02 ; r= 0. 5 p<0.0001), indicating that epilepsy, or age may be a contributing factor for the delayed growth (Figures 4 8, 49, and 4 10)

PAGE 100

100 Figure 4 2. Mean height z score SD over time on diet in months in patients on KT. *p<0.05 using a t test assuming unequal variances. Figure 4 3 Mean weight z score SD over time on diet in months in patients on KT. *p<0.05 using a t test assuming unequal variances.

PAGE 101

101 Figure 4 4. Mean BMI z score SD over time on diet in months in patients on KT. *p<0.05 using a t test assuming unequal variances. Figure 4 5. Average height z score per patient versus time on diet in months.

PAGE 102

102 Figure 4 6. Average weight z score per patient versus time on diet in months. Figure 4 7. Average BMI z score per pa tient versus time on diet in months.

PAGE 103

103 Figure 4 8. Height z score versus age in years for patients on KT at 0 6 months on diet. Figure 4 9. Weight z score versus age in years for patients on KT at 0 6 months on diet.

PAGE 104

104 Figure 4 10. BMI z score versu s age in years for patients on KT at 0 6 months on diet. Patients who were on the diet for >60 mo were analyzed separately to determine if data was skewed by patients who discontinued the diet early. A similar trend was found with Ht z score becoming progr essively worse over time with the biggest difference found comparing 0 6 mo and >60 mo on diet (Figure 4 11). At 06 mo on diet, Ht z score and age were trending towards a significant correlation ( r=0. 4 r2=0.2, p= 0.0 85) for tho se patients on diet >60 mo. Data were available for 19 patients at both 0 6 mo and >60 mo on diet. Using a paired t -test, significant differences were found between 0 6 mo and >60 mo on diet for Ht z score ( 1.0, 2.7, p<0.0001). In addition, Cal/kg/day was also significantly lower at >60 mo on diet (59.7, 47.6, p=0.0004)

PAGE 105

105 Figure 4 11. Mean height z score SD over time on diet in months in patients on KT for >60 months. *p<0.05 using a t test assuming unequal variances. Age To determine if there we re any age specific effects on the growth, patients were stratified according to age (0 4, 4 9, 9 14, 1419, >19 yr) and their Ht z score from before diet (0 6 mo) and >6 mo on diet using a t test were compared assuming unequal variances. No significant differences were found in any group. A trend towards significance (p=0.06) was found for the 1419yr old group. Comparisons were then made using only data available when the same patients were in each group. A paired t -test was used to compare 0 6 mo to >6 mo on diet and significant d ifferences were found for the 0 4 yr old group ( 0.075, 0.47; p=0.004, n=25) and 4 9 yr old group ( 0.98, 1.5; p<0.0001, n=25) (Figure 4 12). Similarly, when Ht z scores were graphed over time on diet for each age group, it appeared that the younger patients were most affected by time on diet (data not shown). However, Ht z score also decreases with age (Figure 4 13). Ht z score was significantly lower in the 4 9 (p=0.02), 9 14 (p=0.001), and 14 19 (p=0.0001) yr old groups compared to the 1 4 yr old group.

PAGE 106

106 Figure 4 12. Mean height z s core SD in each age group at 0 6 mo and >6 mo on diet A paired t -test was used to compare 0 6 mo to >6 mo on diet and significant difference s were found for the 0 4yr old group ( 0.075, 0.47; p=0.004, n=25 ) and 4 9yr old g roup ( 0.98, 1.5; p<0.0001, n=25 ). Figure 4 13. Mean height z score SD in each age group. *p<0.05 using a t test assuming unequal variances comparing each mean height z score to the 1 4yr old group.

PAGE 107

107 Ambulation s tatus Since weight -bearing exercise, wal king, and physical movement can effect linear growth, we stratified the patient population based on ambulation status to examine the effect of diet on growth in each group. Overall Ht and Wt z score s were significantly lower for non ambulatory patients com pared to ambulatory, p<0.0 1 howeve r BMI z score did not differ. At 0 6 mo on diet, the same differences were seen (Table 4 1) There was also a significant differen ce in time on diet between groups ( 20. 6 14. 5 ; 13.6 9. 7 ; p=0.004 ), Pro% of calories (9.4 2.2 ; 8.3 2.1 ; p=0.0 2), CHO% of calories ( 3 7 2.6 ; 5 .5 2.2 ; p =0.001) and CHO/kg/day ( 0.6 0.5 ; 0.9 0.6 ; p=0.005 ) for non ambulatory versus ambulatory. There were no differences in % EER (93. 7 29. 0 ;9 5 3 2 3 7 ;p=0. 76), Cal/kg/day (53.7 16.5;61. 0 24.5; p=0. 13), ra tio (3.10.6;3.00.5; p=0.22 ), Fat% of calories (87 2;86 2 ; p= 0.20), Pro/kg/day (1. 2 0.2;1.2 0.4 ; p= 0.86) or % RDA for protein (122. 2 18. 8 ;120. 1 29.5; p =0.7 2 ). Table 4 1 Comparison of height, weight, and BMI z scores for ambulatory versus non ambulatory patients ov erall and at 0 6 months on diet Overall Mean SD Ambulatory Non ambulatory H t z score 0.6 1. 3 (n= 31 ) 1. 8 1. 6 (n= 67 ) W t z score 0. 4 1. 5 (n= 33 ) 1. 4 1.8* (n= 73 ) BMI z score 0.3 1. 9 (n= 30 ) 0. 4 1. 9 (n= 66 ) 0 6 months on diet Mean SD Ambulatory Non ambulatory H t z score 0.5 1. 2 (n= 25 ) 1.1 1. 3 (n= 54 ) W t z score 0. 3 1. 6 (n= 32 ) 1. 3 2.1 (n= 69 ) BMI z score 0.3 1. 8 (n= 24 ) 0. 6 2.1 (n= 45 ) *p<0.0 1 for Ambulatory versus Nonambulatory Diet p rescription To determine if diet prescription had an effect on growth, the ratio of fat: Pro + CHO Cal/kg/day and Pro/kg/day were examined. N o significant correlations were found between ratio and growth and no differences were found in Ht, Wt, or BMI z score for ratio groups (Table 4 2 ).

PAGE 108

108 There was no significant correlation between Cal/kg/day and Ht z score or BMI z score and a significant, but weak negat ive correlation for Wt z score (r= 0. 3 r2= 0.06, p =0.01) (Figure 4 14) Table 4 2 Comparison of ratio groups for height, weight and BMI z score Growth Parameter Ratio <3:1 Ratio 3:1 Ratio >3:1 Ht z score 1. 7 1.5 (n=5 6 ) 1.4 1. 8 (n=5 2 ) 1.5 1.6 (n= 49 ) Wt score 1.1 1.8 (n=7 2 ) 0.9 1.8 (n= 78 ) 1.2 1.7 (n=6 5 ) BMI z score 0. 2 2.0 (n=5 6 ) 0.1 1.9 (n= 48 ) 0.3 1. 6 (n=4 7 ) No significant difference was detected between ratios using t -test assuming unequal variances. Figure 4 14. Weight z score versus Calories per kilogram body weight per day (Cal/kg/day) for patients on KT. The p atients were stratified by age and gender in order to determine the Recommended Dietary Allowance (RDA) for protein and the Estimated Energy Requirement (EER) based on the DRI from the National Academy of Sciences, (2005) per kg of body weight (301) The p ercent age s of actual intake or prescribed intake of calories and protein from the recommended EER/kg/day for calories (% EER) and RDA/kg/day for protein (% RDA ) were the n calculated

PAGE 109

109 for each patient. The % EER was significantly correlated with Ht z score ( r= 0.3, r2=0.11, p=0.001), Wt z score (r= 0. 7 r2=0.42, p<0.0001) and BMI z score (r= 0. 4 r2=0.19, p<0.0001) After stratifying by ambulation status, the correlation for Ht and BMI z score and % EER was no longer significant for ambulatory patients, bu t was significant for non ambulatory patients (r= 0.4, r2=0.15, p=0.001; r= 0.5, r2=0.27, p<0.0001). Wt z score and % EER remained significant for both groups (ambulatory r= 0.7, r2=0.48, p<0.0001; nonambulatory r= 0.6, r2=0.39, p=0.0001). G iven that some patients had a % EER >100%, the data were also analyzed for patients with % EER A positive correlation was found for Ht z score and Cal/kg/day (r=0.3, r2=0.06, p=0.05), but the correlations for Ht and BMI z score and % EER w ere no longer significant. Wt z score and % EER remained significant, however the correlation was wea ker (r= 0.4, r2=0.17, p=0.0005). When the data were analyzed only for patients with % EER >100, Ht and Wt z score were significantly correlated to % EER (r= 0.5, r2=0.28, p=0.001; r= 0.5, r2=0.24, p=0.002). Wt z score w as significantly correlated with % RDA for protein (r= 0.4, r2=0.15, p<0.0001), however this was a weak correlation and Ht and BMI z score showed no correlation Protein intake does not seem to be a factor in this population. O nly 8.5% of the population had protein intakes <99.9% of t he RDA, and no patient s had an average < 70%. The average % RDA for protein of the population was 122%. Patients were then stratified according to age, time on diet, route of feeding, and ambulatory versus nonambulatory to determine if diet prescription h ad more o f an effect on subpopulations. The correlation coefficients for Ht z score versus Cal/kg/day, % EER, and %

PAGE 110

110 RDA for protein were different depending on time on diet and age in years, however no strong correlations were detecte d. Route of feeding U sing a t -test assuming unequal variances to compare tube -fed (TF) versus orally -fed (Oral) patients, significant difference s were observed for % EER, CHO/kg/day, Fat/kg/day, CHO and Pro% of calories, Cal/kg/day, and time on diet, but no significant differe nces were noted for age, ratio, Fat% of calories, Pro/kg/day, % RDA f or protein, Wt or BMI z score. A trend towards significance (p=0.08) was observed when comparing Ht z score for Oral versus TF patients ( 1.2 1.5 Oral, 1.7 1.6 TF). Results are liste d in the Table 4 3 Regression analyses were performed for both groups and n o significant differences in correlations were found between Oral and TF patients Table 4 3 Comparison of oral ly -fed versus tube -fed patients Parameter Oral (n=62 ) TF (n=52 ) Age 8. 8 5. 8 7. 7 4. 5 Time on diet (mo) 1 5 5 12.5 2 3 2 16.6 Cal/kg/day 6 0.9 21.7 50.3 14.3 Ratio 3.0 0.6 3.1 0. 6 %EER for calories 100.4 29.0 88.4 29.2 %RDA for protein 119. 8 28.1 12 4.0 15.9 CHO/kg/day 0.94 0.59 0. 42 0. 48 CHO% of cals 5.6 2.3 2. 9 2.4 Fat/kg/day 5.9 2.1 4. 9 1.4 Fat% of cals 86. 3 2. 4 87.0 2 .3 Pro/kg/day 1.2 0.3 1.2 0.2 Pro% of cals 8.1 1.9 10.1 2.2 Ht z score 1.2 1.5 (n=55) 1. 7 1. 6 (n= 49 ) Wt z score 1.1 2.0 1.2 1. 5 BMI z score 0.4 4 2. 1 (n=54) 0.0 8 1. 5 (n=48) *p<0.05; trend towards significance, p=0.0 8 Interestingly, there was a trend towards significance for Ht z score among groups, similar to the comparison for non ambulatory and ambulatory patients however there were a lso si gnificant differences in % EER (100% Oral vs 88% TF p=0.03) and in Cal/kg/day (61 vs 50,

PAGE 111

111 p=0.002) between the route of feeding groups but not between ambulation groups. These results indicate that the re may be a difference in use of calories and macronutrients between the subpopulations that was not apparent when looking at the KT population as a whole. Despite results indicating no strong correlations between calories, % EER and Ht z score, the average % EER in the KT population was 94%, minimum at 52% and maximum at 224%, with 65% of the population having a % EER < 100 and 23% having a % EER < 75. Approxim ately 36% of the patients had >100% EER. Results suggest that there is a wide range of calorie prescriptions based on recommendatio ns by the DRI and we may be unable to accurately determine energy needs based on formulas current ly available for the normal healthy population. -hydroxybutyrate (BHB) No correlations were found for BHB and Ht, Wt, or BMI z scores (Figure 4 -15). The a verage BHB per patient was stratified into tertiles and groups were compared (BHB range 4.5 81.4 mg/dL, n=30 in each group). Ratio, Wt z score, and B MI z score were not different among groups. Ht z score was significantly lower in group 2 compared to group 1 ( -1.1, 1.9, p=0.03), however not for group 1 and 3 ( 1.1, 1.8, p=0.07) or group 2 and 3 ( 1.8, 1.9, p=0.7). However, group 3 was significantly younger than group 1 (9.1, 6.7 yrs, p=0.04) and trending to be significantly younger for group 2 compared to group 3 (9.4, 6.7 yrs, p=0.06). Summary of results Overall, BMI and Wt z score were not affected by time on diet. Ht z score decreased over time o n KT, but age, or potentially length of epilepsy also had an effect. Ambulation status was a significant factor for Ht and Wt z score. Neither ratio nor protein intake affected growth parameters, but the analysis of calories and growth were unclear. While, calor ie intake in this population compared to the DRI for normal healthy children on average was significantly lower,

PAGE 112

112 no strong correlations were found. Additionally, stratifying by ambulation status made a difference in the calorie and Ht z score relatio nships, therefore ambulation is likely an important factor in determination of calorie needs. Unlike previously published, blood BHB was not found to be an important factor in growth. A summary of the anthropometrics and growth results for the retrospectiv e analyses is presented in Figure 4 16. Figure 4 15. Height z score versus -hydroxybutyrate (BHB) in mg/dL in patients on KT. Prospective Results For the prospective study, 29 experienced patients were enrolled mean age of 10.35.0 years and time on diet of 46.235.2 at start of the study. Fifteen nave patients were enrolled into the study and 12 patients were initiated in the GCRC, mean age at in itiation was 7.26.1 years. Demographics at the start of the study are described in Table 4 4. Three of the patients enrolled were not initiated, 2 because seizures improved and 1 bec ause of other health complications.

PAGE 113

113 Two patients stopped the diet during the first week due to issues with caregiver administration of therapy ; therefore 10 nave patients were included in the analysis. Figure 4 16. Summary of anthropometrics and gro wth results of retrospective analyses. Table 4 4 Demographics at start of prospective study for experienced and nave patients Experienced (n=29) Nave (n=10) Time on diet (months) 46.2 35.2 0 Age (years) 10.3 5.0 7.2 6.4 Gender 17F/ 12M 1F/ 9 M Ambulation 2 2 Non/ 7 Amb 3 Non/ 7 Amb Route of feeding 12 Oral/ 5 Both/ 12 TF 7 Oral/ 2 Both/ 1 TF Race 19 White/4 African American/1 Biracial/5 Hispanic 9 White/ 1 Hispanic

PAGE 114

114 Experienced patients Si milar to the retrospective data, for experienced patients, Ht z score was negatively correlated with time on diet (r= 0.46, r2=0.21, p=0.01) (Figure 417); however Ht z score was also negatively correlated with age with a slightly stronger correlation (r= 0.48, r2=0.23, Figure 4 17. H eight z score versus time on diet in months for experienced patients on KT evaluated prospectively p=0.008) (Figure 4 18). Neither Wt nor BMI z score were correlated to time on diet (r= 0.3, r2=0.1, p=0.09; r= 0.07, r2=0.01, p=0.7), and both were ne gatively correlated with age (r= 0.4, r2=0.16, p=0.03; r= 0.5, r2=0.25, p=0.006). Diet prescription Ht z sc ore was positively correlated with Cal/kg/day (r= 0.4, r2=0.2, p=0.02) (Figure 4 19); no correlations were observed for Cal/kg/day and Wt and BMI z score. In contrast no correlation was observed for Ht and BMI z score and % EER (Figure 4 20 and 4 21), and Wt z score was negatively correlated to % EER (r= 0.4, r2=0.2, p=0.02) (Figure 4 22). However, a trend

PAGE 115

115 Figure 4 18. H eight z score versus age in years for experienced patients on KT evaluated prospectively Figure 4 19. H eight z score versus Calories per kilogram body weight per day (Cal/kg/day) for experienced patients on KT evaluated prospectively

PAGE 116

116 Figure 4 2 0. H eight z score versus % r ecommended Calories (%EER) as determined by equations from DRI for energy based on age, gender and activity level for experienced patients on KT evaluated prospectively Figure 4 21. BMI z score versus % recommended Calories (%EER) as determined by equa tions from DRI for energy based on age, gender and activity level for experienced patients on KT evaluated prospectively

PAGE 117

117 Figure 4 22. W eight z score versus % recommended Calories (%EER) as determined by equations from DRI for energy based on age, gender and activity level for experienced patients on KT evaluated prospectively Figure 4 23. H eight z score versus % recommended Calories (%EER) as determined by equations from DRI for energy based on age, gender and activity level for experienced patient s on KT with % EER less than 80%.

PAGE 118

118 towards a positive correlation with Ht z score and % EER were observed for patients with a % EER <80 (r= 0.5, r2=0.24, p=0.18) (Figure 4 23). Differences observed between Cal/kg/day and % EER for calories may be due to Cal /kg/day not taking into account age or activity level whereas the % EER does. No signi ficant difference in Ht, Wt, or BMI z score was observed between ratio groups using a t te st assuming unequal variances (Table 4 5) and no correlations with ratio and gr owth were observed. Table 4 5 C omparison of ratio groups for height, weight and BMI z score for experienced patients evaluated prospectively Growth Parameter Ratio <3:1 (n=9) Ratio 3:1 (n=10) Ratio >3:1 (n=10) Ht z score 2.4 1. 3 2.0 1. 5 1. 9 1 .6 Wt score 0.9 1. 3 1.2 1. 6 0.8 1. 2 BMI z score 0. 4 2. 6 0. 7 2.2 0.5 0.8 While no significant differences or correlations were observed with growth and ratio, other dietary parameters seem to be important. Notably, whereas the Fat % of calories (or essentially the ratio) only varies slightly within the population (81.8 90.4), the CHO% of calories and Pro% of calories vary widely across patients (0.6 8.6; 5.0 15.0). Ht z score was found to be positively correlated with CHO% of calori es (r= 0.5, r2=0.28, p=0.003), CHO g/kg/day (r= 0.5, r2=0.22, p=0.01) and Fat g/kg/day (r= 0.4, r2=0.19, p=0.02) and negatively correlated with Pro% of calories (r= 0.5, r2=0.24, p=0.007). CHO and Fat in g/kg/day follow ed a similar trend with Cal/kg/day and CHO% calories having the most significant correlation with Ht z score. Ht z score was also correlated to CHO:Pro (r=0.5, r2=0.2, p=0.006). Importantly, age was also negatively correlated to CHO% of calories (r= 0.5, r2=0.2, p=0.008), CHO:Pro (r= 0.5, r2=0.3, p=0.003) and strongly to Cal/kg/day (r= 0.7, r2=0.4, p<0.0001) and positively with Pro% of

PAGE 119

119 calories (r= 0.6, r2=0.3, p=0.001), suggesting again that this may be an effect of age. Although younger patients have higher protein requirements per k ilogram body weight they also have significantly higher energy needs based on total weight (Cal/kg/day) and therefore, this generally allow s for more CHO in their diet prescription compared to older patients. Additionally, both Wt and BMI z score were sig nificantly correlated to % RDA for recommended Pro/kg/day (r= 0.4, r2=0.15, p=0.04; r= 0.5, r2=0.23, p=0.009). Overall, the majority of experienced patients received adequate protein per kg body weight based on the DRI; on average, 4/29 (14%) patients ha d <100% of recommended Pro/kg/day and 1/29 (3%) had <80%. Based on the average % EER per patient, 20/29 (69%) had <100% of EER and 9/29 (31%) had <80% of EER. A ll 9 were non ambulatory patients. Ambulation status For prospective data, significant differenc es were found when comparing nonambulatory to ambulatory patients for Ht and Wt z score. However, there were also differences in time on diet, age, and Cal/kg/day. On average, nonambulatory patients were older, on diet longer, and receiv ed less Cal/kg/da y (Table 4 6). Additionally, ambulation status appears to be an important factor in the correlation between Ht z score and %EER (Figure 4 24). Table 4 6 C omparison of height, weight, and BMI z scores for ambulatory versus non ambulatory experienced patie nts evaluated prospectively Overall Mean SD Ambulatory (n=7) Non ambulatory (n=22) H t z score 0. 7 1. 2 2.5 1. 2 W t z score 0. 07 1. 0 1. 4 1. 3 BMI z score 0.5 1. 2 0. 4 2.2 Age (years) 8.4 3.4 11.9 5.2+ Time on diet (mon ths) 24.1 8.7 64.0 33.9* Cal/kg/day 67.7 21.3 44.3 18.9* *p<0.0 5 for Ambulatory versus Nonambulatory +p=0.06

PAGE 120

120 Figure 4 24. H eight z score versus % recommended Calories (%EER) as determined by equations from DRI for energy based on age, gender and activity level for experienced patients on KT and stratified based on ambulation status (NonAmb=non ambulatory; Amb=ambulatory). Nave patients Ht z score did not seem to be as affected in the nave patient group. As shown in Figure 4 25, some patie nts Ht z score impr oved over time on diet When stratified based on before and after diet and compared using a paired t -test, no significant differences were observed for Ht, Wt, or BMI z score. There was a trend for an increased Ht z score (p=0.09) on di et compared to before. Ambulation status was only important for Ht z score and not Wt or BMI z score before diet. Additionally, the significance was not observed comparing nonambulatory to ambulatory Ht z scores on diet (p=0.1) (Table 4 7). Ht z score was negatively correlated with % EER (r= 0.7, r2=0.4, p=0.04), however, when the 3 patients with % EER>100% were removed, the correlation was not significant (Figure 4 26).

PAGE 121

121 Figure 4 25. H eight z score over time on diet in months for individual nave pati ents evaluated prospectively Table 4 7 Comparison of height, weight, and BMI z scores for nave patients before and after diet and comparison of growth based on ambulation status Mean SD Before Diet (n=10) On diet (n=10) H t z score 1.1 2.0 0. 9 1.8 + W t z score 1.2 1.6 1.0 1.5 BMI z score 0.4 1.1 0.7 1.4 Before Diet (n=10) On diet (n=10) Ambulatory (n=7) Non ambulatory (n=3) Ambulatory (n=7) Non ambulatory (n=3) H t z score 0.4 2.0 2.9 0.7* 0.4 2.0 1.9 0.5 ^ W t z score 0.9 1.6 1.8 1.5 0.8 1.7 1.4 0.9 BMI z score 0.6 1.2 0.04 0.9 0.9 1.6 0.1 0.7 +p= 0.0 9 using a paired t test, *p<0.05, ^p=0.1 using an unpaired t test assuming unequal variances. -hydroxybutyrate (BHB) Similar to the retrospective data, no correlations were found for BHB (mg/dL) and Ht, Wt, or BMI z scores (Figure 4 27). However, Ht z score was significantly correlated to Glu (mg/dL)

PAGE 122

122 Figure 4 26. H eight z score versus % recomme nded Calories (%EE R) as determined by equations from DRI for energy based on age, gender and activity level for nave patients on KT. (r= 0.4, r2=0.13, p=0.02) (Figure 4 28). When stratified by nave or experienced, the correlation was no longer signific ant for either group. Summary of results The demographics of the experi enced and nave patient groups we re different. Whereas the experienced group was mostly nonambulatory and tube -fed, there was only 1 solely tube -fed patient and 3 nonambulatory patients in the nave group. Similar to the data from the retrospective study in the exp erienced group, Ht z score was a ffected by time on diet whereas Wt and BMI z score were not. All three growth parameters were again correlated with age and BHB did not show any correlations. However, no affect of time on diet on Ht, Wt, or BMI z score was found in the nave group. Actually, some patients Ht z scores improved over time on

PAGE 123

123 diet (Figure 4 25). Nevertheless, this was over a shorter period of time, and the di et prescription was more closely monitored in the prospective study compared to the retrospective study. Again, energy intake compared to the DRI showed no clear correlations with growth and differences were found based on ambulation status. Interestingly, patients receiving too many calories according the DRI (%EER >100) and patients receiving too few (%EER <100) had z scores in the negative range, and therefore, were not growing optimally for their age (Figure 4 24). Results indicate that we are unable to accurately assess energy needs in this patient population with current equations for healthy children. While ratio and protein intake again were not correlated with growth, other dietary parameters such as CHO in g/kg/day and percent of calories need to b e further evaluated. A summary of the anthropometric data for the retrospective and prospective studies is presented in Figure 4 29. Figure 4 27. H eight z score versus -hydroxybutyrate (BHB) in mg/dL in patients on KT evaluated prospectively

PAGE 124

124 Figure 4 28. H eight z score versus Glucose (Glu) in mg/dL in patients on KT evaluated prospectively Figure 4 29. Summary of anthropometrics data including the retrospectiv e and prospective results.

PAGE 125

125 Blood Lipids and Dyslipidemia Retrospective Results One hundred and eight -seven patients from our medical center were initiated on KT from 19952006. Of those patients blood lipid data were available for 164 patients and 3 of t hese patients were started on the diet twice. The mean age at initiation was 7.4 5.1 years and there were 88 males and 76 females. For diet prescription analysis, blood lipids and diet prescription data were available for 76 pati ents, 42 males and 34 fema les. The mean age at initiation for this subgroup was 7.3 5. 6 years 7.9 6.1 for males and 6.6 5. 1 for females. Two females were started on the diet twice. Time on diet Blood lipid levels appear to go up initially during the first year of KT, and return to levels similar to baseline the long er the patient is on KT. Each time period on diet was compared to 0 months or baseline levels at initiation. For all time points TG concentrations were significantly increased from baseline except at 84 100 mo on diet ( Figure 4 30). For HDL, all were significantly decreased after 4 mo except 36 48 and >60 mo on diet (Figure 4 31); however none of the mean levels for each time period were below 40 mg/dL. For nonHDL, there were no significant increases detected and mean nonHDL levels trended to decrease over time on diet (Figure 4 32). Mean Total Chol and LDL levels were not significantly higher at any time periods compared to baseline and at 48 60 mo on diet both Total Chol and LDL levels were significantly lower than bas eline levels (Figure 4 33 and 4 34). Total/HDL was s ignificantly higher at 4 12 (p=0.008), 1224 (p=0.02), and 2436 mo (p=0.03) on diet compared to 0 mo on diet (Figure 4 35) Similarly, the z tiles appear to normalize f or all lipids with time on KT. To determine if the data were skewed by the patients who had high lipids in the beginning of therapy and then were

PAGE 126

126 taken off diet, 22 patients who were on diet for 4860 months were graphed with their lipids versus time on diet and the same effect was seen ( results not shown). Age and gender No definitive effect of age on blood lipids was observe d. The only differences found were 0 3yr olds had the highe st mean TG, however they were also on diet significantly less time; and Total Chol was found to be signifi cantly higher in > 15yr olds compared to 1215yr olds (158.928, 183.342, p=0.0 5). There were no significant differences when comparing mean levels for males versus females for any of the blood lipids or time on diet (Table 4 8). Table 4 8 Comparison of mean blood lipids for males versus females Mean SD Parameter Males Females p value TG (mg/dL) 136.5 95.7 (n=88) 141.4 84.9 (n=76) 0.73 HDL (mg/dL) 53.3 12.9 (n=87) 52.6 12.8 (n=75) 0.72 nonHDL (mg/dL) 123.1 43.4 (n=87) 123.3 42.1 (n=7 5) 0.97 LDL (mg/dL) 98.0 35.7 (n=80) 93.7 35.5 (n=67) 0.47 Total Chol (mg/dL) 176.3 44.9 (n=88) 175.3 46.2 (n=76) 0.93 Total/HDL 3.5 1.2 (n=87) 3.6 1.2 (n=75) 0.97 Time on diet (months) 10.3 14.8 (n=88) 11.2 15.1 (n=76) 0.70 Figure 4 30. Mean TG SD in mg/dL over time on diet in months. Dashed line indicates normal levels based on AHA recommendations. *p<0.05 using a t test assuming unequal variances.

PAGE 127

127 Figure 4 31. Mean HDL SD in mg/dL over time on diet in months. Dashed l ine indicates normal levels based on AHA recommendations. *p<0.05 using a t test assuming unequal variances. Figure 4 32. Mean nonHDL SD in mg/dL over time on diet in months. Dashed line indicates normal levels based on AHA recommendations. *p<0.05 using a t test assuming unequal variances.

PAGE 128

128 Fi gure 4 33. Mean Total Chol SD in mg/dL over time on diet in months. Dashed line indicates normal levels based on AHA recommendations. *p<0.05 using a t test assuming unequal variances. Figure 4 34. Me an LDL SD in mg/dL over time on diet in months. Dashed line indicates normal levels based on AHA recommendations. *p<0.05 using a t test assuming unequal variances.

PAGE 129

129 F igure 4 35. Mean Total/HDL SD in mg/dL over time on diet in months. Dashed line in dicates normal levels based on AHA recommendations. *p<0.05 using a t -test assuming unequal variances. Dyslipidemia Dyslipidemia was defin ed as a T G >150, HDL <40, n onHDL >190, LDL >110, Total/ HDL >4.4 and/or Total Chol >170. To circumvent the issue of non-fasting levels, we considered values to be abnormal if the patient had dyslipidemia 2 or more times. The number of patients and percentage of the KT population with dyslipidemia are listed in T able 4 -9 Table 4 9 Patients in KT population presenting w ith dyslipidemia two or more times Dyslipidemia # of patients (% of population) Males Females Total TG >150 (n=1 6 4) 3 2 2 6 5 8 (35. 4 ) HDL <40 (n=1 62 ) 17 1 0 2 7 ( 16.7 ) nonHDL >190 (n=1 62 ) 7 5 1 2 ( 7.4 ) LDL >110 (n=1 47 ) 19 1 4 33 (2 2 4 ) Total Chol >170 (n=1 64 ) 35 2 8 63 (38. 4 ) Total/HDL >4.4 (n=1 62 ) 16 1 4 30 ( 18.5 ) Diet prescription KetoCal 4:1 (Nutricia North America, Gaithersburg, MD) (KC), a commercially available product that represents one dietary option for patients on KT is very high i n

PAGE 130

130 monounsaturated fat and trans fatty acids due to the use of hydrogenated soybean oil P reliminary data suggest ed that the blood lipid levels are more abnormal for patients on KC compa red to all other patients on KT; however, statistical significance was not observed perhaps because there were only 8 patients on KC (Table 4 10). Table 4 10. Comparison of mean blood lipids for KetoCal patients and the entire KT population Mean SD Blood Lipid Ketocal patients (n=8) All patients (n=164) TG (mg/dL) 145. 8 43.7 138.8 90.6 HDL (mg/dL) 47.4 11.2 53.0 12.8 n onHDL Chol (mg/dL) 146.5 43.3 123.2 42.7 LDL (mg/dL) 117.5 40.1 96.1 35.5 Total Chol (mg/dL) 193.9 50.2 176.1 43.0 Total/HDL 4.2 0.9 3.5 1.2 Because we had a limited number o f patients on KC and th ey were all on KC for different lengths of time, multiple comparisons were made. The lipid levels were averaged before KC and compared to lipids for the patient while they wer e on KC using a paired t test. The same number of dates be fore and after KC was averaged for each patient. A trend towards a significant incr ease in LDL after KC compared to before (p=0.07) was observed in all 5 patients When one patient who appeared to be an outlier was removed from the data set significant in creases in n on HDL, LDL, and Total Chol before versus during KC administration were detected (Figure 4 36). We then compared the patients on KC to a matched control at a similar age, time on diet, ratio on diet and route of feeding. A significant increase i n Total/HDL (p=0.02) for patients on KC compared to control patients was the only difference detected using a paired t -test (data not shown) Seventy-six patients who had been on KT for at least 1 year were stratified according to fat ratio (fat: Pro + CHO ) and groups were defined as a ratio <3:1, 3:1, and >3:1. Each group was

PAGE 131

131 Figure 4 36. Average blood lipid levels in mg/dL before (B) and during (D) KetoCal for 5 patients on KT. compared to each another using a t test assuming unequal variances and no effect of ratio on blood lipids was observed (Table 4 11). Regression analysis was also examined for ratio, Fat%, CHO%, Fat (g/day), CHO (g/day), Fat/kg/day, and CHO/kg/day and each blood lipid parameter N o significant correlations were detected (data no t shown). P atients were stratified between the high SF A orally -fed diet and the high omega 6 PUFA tube -fed diet and blood lipids were compared between groups. Orally -fed patients had significantly higher nonHDL (p=0.003), LDL (p=0.001) and Total Chol (p=0. 002) compared to tube -fed patients (Table 4 12). Table 4 11. Comparison of mean blood lipids for patients on different KT ratios Blood Lipid Ratio <3:1 (n=41) Ratio 3:1 (n=29) Ratio >3:1 (n=47) TG (mg/dL) 160.9 95.2 203.9 198.3 213.5 155.9 HDL (mg /dL) 55.4 14.9 51.0 16.2 49.0 15.1 nonHDL (mg/dL) 126.3 45.3 129.8 47.1 118.3 45.6 LDL (mg/dL) 93.3 40.3 93.2 44.4 80.2 40.6 Total Chol (mg/dL) 181.8 46.6 180.9 49.8 168.8 47.2 Total/HDL 3.5 1.2 3.8 1.5 3.6 1.3

PAGE 132

132 Table 4 12. Comparison of mean blood lipids for orally -fed and tube -fed patients on KT Blood Lipid Orally fed (n=38) Tube fed (n=37) TG (mg/dL) 174.1 97.9 189.4 142.7 HDL (mg/dL) 52.3 15.1 50.5 13.2 nonHDL (mg/dL) 142.1 46.6 111.9 39.1* LDL (mg/dL) 106.9 47.5 76.5 26.9** Total Chol (mg/dL) 195.6 49.5 162.5 37.2** Total/HDL 4.0 1.6 3.5 1.3 *p<0.05, **p<0.001 Summary of results Overall, dyslipidemia was present in this population and time on diet was a significant factor in the blood li pids examined. Additionally, ratio or amount of fat was not as important as the type of fat for effects on blood lipids based on the route of feeding comparison. A summary of the blood lipids and dyslipidemia results for the retrospective study is presente d in Figure 4 37. Figure 4 37. Summary of lipid results for the retrospective study

PAGE 133

133 Prospective Results Refer to the Prospective Results in Anthropometrics and Growth Results section for demographic data of prospective study. Similarly to retrospectiv e data no effect of age and gender on experienced patients blood lipids was observed Dyslipidemia Using the same definitions for dyslipidemia used in the retrospective study some differences were documented between the retrospective and prospective study populations. In the prospective study population, there was a lower percentage of patients with an abnormal TG two or more times and a higher percentage of abnormal HDL t wo or more times. This result was most likely due to the issue of non-fasting which retrospectively, was unable to be controlled for, and TG is most affected by non -fasting (Table 4 13). Table 4 13. Patients in KT population presenting with dyslipidemia in the prospective study TG HDL LDL Total Chol # % # % # % # % Values 1 0/160 6.3 49/160 30.6 42/159 26.4 49/160 30.6 Patients 7/38 18.4 23/38 60.5 16/38 42.1 21/38 55.3 Patients 2 or more times 1/38 2.6 15/38 39.5 12/38 31.6 15/38 39.5 Patients were stratified based on whether or not they had abnorm al lipid values two or more times and compared to the other patients with normal leve ls within each lipid parameter. Due to only one patient having an abnormal TG two or more times, the seven patients with at least one abnormal value were compared. No sign ificant differences in dietary parameters were detected ; results are represented in Table 4 14, 4 15, 4 16 and 417. Table 4 14. Comparison of dietary parameters between patients with abnormal TG and normal TG in prospective study Diet parameter Abnormal patients (n =7) Normal patients (n=3 1) TG (mg/dL) 138.5 73.8* Age (yr) 8.1 10.6 Time on diet (mo) 38.3 39.9

PAGE 134

134 Table 4 14. Continued Ratio 3.1 3.1 Fat g/kg/day 6.7 5.3 CHO g/kg/day 1.0 0.62 Fat (g/day) 132.2 144.3 CHO (g/da y) 16.8 14.6 Fat % of cals 87.4 87.4 Protein % of cals 7.8 8.9 CHO % of cals 4.8 3.7 Cal/kg/day 69.1 54.8 Pro/kg/day 1.2 1.1 *p<0.05 Table 4 15. Comparison of dietary parameters between patients with abnormal HDL and normal HDL in prospective study Diet parameter Abnormal patients (n =15) Normal patients (n=2 3) HDL (mg/dL) 36.1 54.9* Age (yr) 11.5 9.3 Time on diet (mo) 47.5 37.0 Ratio 3.2 3.1 Fat g/kg/day 5.2 5.8 CHO g/kg/day 0.58 0.76 Fat (g/day) 149.6 137. 1 CHO (g/day) 14.2 15.6 Fat % of cals 87.6 87.2 Protein % of cals 8.8 8.6 CHO % of cals 3.5 4.2 Cal/kg/day 53.9 59.7 Pro/kg/day 1.1 1.1 *p<0.05 Table 4 16. Comparison of dietary parameters between patients with abnormal LDL and normal LDL in prospective study Diet parameter Abnormal patients (n =12) Normal patients (n=2 6) LDL (mg/dL) 136.0 87.7* Age (yr) 9.1 10.6 Time on diet (mo) 28.6 46.4+ Ratio 3.2 3.1 Fat g/kg/day 5.8 5.5 CHO g/kg/day 0.74 0.67 Fat (g/da y) 149.5 138.6 CHO (g/day) 16.6 14.3 Fat % of cals 87.6 87.3 Protein % of cals 8.1 9.0 CHO % of cals 4.3 3.8 Cal/kg/day 59.8 56.3 P ro/kg/day 1.1 1.1 *p<0.05, +p=0.11

PAGE 135

135 Table 4 17. Comparison of dietary parameters between patient s with abnormal Total Chol and normal Total Chol in prospective study Diet parameter Abnormal patients (n =15) Normal patients (n=2 3) Total Chol (mg/dL) 191.5 151.2* Age (yr) 10.5 10.0 Time on diet (mo) 35.0 44.6 Ratio 3.2 3.1 Fat g/kg/day 5.9 5.4 CHO g/kg/day 0.76 0.64 Fat (g/day) 149.7 137.1 CHO (g/day) 16.7 14.0 Fat % of cals 87.6 87.2 Protein % of cals 8.1 9.1 CHO % of cals 4.3 3.7 Cal/kg/day 60.6 55.3 Pro/kg/day 1.1 1.1 *p<0.05 Time on diet Although no significant differences were observed in time ond diet when comparing abnormal patients to normal patients, it appeared for Total and LDL patients with abnormal levels were on diet less time. Additionally, Total and LDL Chol were significantly negatively correlated with time on diet, whereas TG and HDL were not, indicating the longer on diet, the lower the Total and LDL cholesterol (Figure 4 38 and 4 39). Figure 4 38. T otal cholesterol (mg/dL) versus time on diet in months for experienced patients on K T evaluated prospectively.

PAGE 136

136 Figure 4 39. LDL cholesterol (mg/dL) versus time on diet in months for experienced patients on KT evaluated prospectively. Diet prescription Similar to the retrospective study nonHDL, LDL, and Total cholesterols were all sign ificantly higher in the orally -fed patients compared to the tube -fed patients (Table 4 18). Additionally, Pro% of calories was found to be significantly lower in the orally-fed group. However, it is important to note, that time on diet was trending to be l ower (p=0.053) in the orally -fed group as well. Table 4 18. Comparisons between orally -fed and tube -fed patients on KT in prospective study Diet parameter Orally fed (n=17) Tube fed (n=21) p value Age (yr) 9.3 5.8 10.8 5.6 0.44 Time on diet (mo) 28. 1 32.1 51.1 38.7 0.053 Ratio 3.2 0.4 3.1 0.4 0.76 TG (mg/dL) 87.5 34.0 84.3 31.1 0.96 HDL (mg/dL) 49.3 16.0 46.0 12.2 0.71 nonHDL (mg/dL) 139.2 62.7 103.8 26.8 0.04 LDL (mg/dL) 122.7 61.3 87.0 24.6 0.04 Total Chol (mg/dL) 188. 5 70.2 149.8 26.6 0.04 Cal/kg/day 64.5 25.6 51.7 24.8 0.13 Fat g/kg/day 6.3 2.5 5.0 2.4 0.25 CHO g/kg/day 0.9 0.6 0.6 0.6 0.21 CHO% of calories 4.7 2.3 3.3 2.3 0.06 Pro% of calories 7.8 2.1 9.4 2.8 0.04

PAGE 137

137 No effect of ratio on blood lipids was observed when comparing ratio groups using an unpaired t -test assuming unequal variances (Table 4 19), and no correlations were observed with BHB or ratio and any of the blood lipid parameters (Table 4 20). Table 4 19. Comparison of mean blood lipids for patients on different KT ratios in prospective study Blood lipid Ratio <3:1 (n=11) Ratio 3:1 (n=14) Ratio >3:1 (n=16) TG (mg/dL ) 83.4 43.5 88.4 27.7 83.2 36.3 HDL (mg/dL ) 46.6 9.7 41.3 8.2 45.8 10.4 LDL (mg/dL ) 85.3 33.8 96.1 34.7 100.6 47.0 Total C hol (mg /dL ) 148.4 36.3 155.2 36.0 163.0 47.5 Table 4 20. Regression analyses for blood lipids and ratio and BHB evaluated prospectively Ratio BHB (mg/dL ) Blood lipid r p value r p value TG (mg/dL ) 0.09 0.6 0 0.07 0.67 HDL (mg/dL ) 0.07 0.66 0.01 0.95 nonHDL (mg/dL ) 0.10 0.55 0.10 0.55 LDL (mg/dL ) 0.10 0.56 0.11 0.52 Total Chol (mg/dL ) 0.11 0.51 0.09 0.58 Nave patients Preliminary evidence suggests age and other dietary factors beside s ratio may be important to monitor in terms of percent change in blood lipids after starting KT When examining all nave blood lipids, age was significantly correlated to percent change in nonHDL (r=0.5, r2=0.2, p=0.02), LDL (r= 0.6, r2=0.3, p=0.002) (Fi gure 4 40) and Total Chol (r= 0.5, r2=0 .2, p=0.01). When patients were stratified based on age, significant differences were only observed when comparing all data, not when averaged per patient (Table 4 21), most likely due to the small numbers within the groups. Additionally, althou gh ratio or Fat % of calories was not correlated with blood lipids, a trend towards significance was observed for Fat g /kg/day and percent change in LDL cholesterol (Figure 4 41).

PAGE 138

138 Table 4 21. Percent change in blood lipids in nave patients on KT and age in years (yrs) in prospective study Averaged per patient <10 yrs (n=7) >10 yrs (n=3) p value % change in TG 45.1 16.7 0.47 % change in HDL 2.2 3.3 0.95 % change in nonHDL 1.1 73.1 0.16 % change in LDL 4.8 82.2 0.11 % change in T otal 3.8 48.8 0.11 All data <10 yrs (#=18) >10 yrs (#=9) p value % change in TG 63.3 10.1 0.07 % change in HDL 4.1 11.8 0.37 % change in nonHDL 7.7 60.8* 0.02 % change in LDL 2.4 70.9* 0.01 % change in T otal 1.3 38.3* 0.01 Despite the fact that there were a small number of nave patients to compare groups, graphs of nave patients blood lipids over time on diet were examined. Graphs were stratified by route of feeding, and the one sol ely tube -fed patient had reduced TG, Total and LDL cholesterol and increased HDL cholesterol over time (Figure 4 42, Figure 4 43, Figure 4 44, and Figure 4 45). Figure 4 40. Percent change in LDL cholesterol in nave patients on KT versus age in years evaluated prospectively

PAGE 139

139 Figure 4 41. Percent change in LDL cholesterol in nave patients on KT versus Fat in grams per kilogram body weight per day (Fat g/kg/day) evaluated prospectively. Figure 4 42. Triglycerides (TG) in mg/dL over time on diet i n months for individual nave patients stratified by route of feeding and evaluated prospectively. Orange=Orally fed; Green= Tube -fed; Blue=Both orally -fed and tube -fed.

PAGE 140

140 Figure 4 43. T otal cholesterol (Total Chol) in mg/dL over time on diet in months for individual n ave patients stratified by route of feeding and evaluated prospectively Orange=Orally -fed; Green= Tube -fed; Blue=Both orally -fed and tube -fed. Figure 4 44. LDL cholesterol (LDL Chol) in mg/dL over time on diet in months fo r individual n ave patients stratified by route of feeding and evaluated prospectively Orange=Orally -fed; Green= Tube -fed; Blue=Both orally -fed and tube -fed.

PAGE 141

141 Figure 4 45. HDL cholesterol (HDL Chol) in mg/dL over time on diet in months for individual nave patients stratified by route of feeding and evaluated prospectively Orange=Orally -fed; Green= Tube -fed; Blue=Both orally -fed and tube -fed. Summary of results Results from the prospective study were similar to those in the retrospective study for blood lipids. Time on diet still appears to be important, but more so for Total and LDL Chol than for TG and HDL. While type of fat remains important and ratio or amount of fat still does not appear to effect blood lipid levels, in examining the percent change in lipids in the nave group, age as well as other dietary factors, such as Fat g/kg/day, need to be further evaluated in a larger sample. Other dietary factors besides the ratio, which is routinely evaluated clinically, may help in determining risk of dyslipidemia. A summary of the blood lipids and dyslipidemia results for the retrospective and prospective studies is presented in Figure 4 46.

PAGE 142

142 Figure 4 46. Summary of lipid results of retrospective study with addition of prospective study results. Meal Challenge Pr ospective Results Demographics Data were available for 23 patients with mean age of 10.25 years and t ime on diet of 51.438 months. There were 7 males and 16 females. Eight patients were solely tube -fed (TF), 5 were fed orally and via a g -tube (Both), and 10 were orally -fed (Oral). Mean and standard deviations for BHB and Glu at all time points are listed in Table 4 22. Mean percent changes at 30 min, 1 hr, and 2 hr for both BHB and Glu were all less than 10% (Table 4 23)

PAGE 143

143 Table 4 22. Mean BHB and Glu in mg/dL at fasting and after ingestion of a ketogenic meal Mean (mg/dL ) SD Mean (mg/dL ) SD Fasting BHB (n=23) 34.2 11.4 Fasting G lu (n=22) 72.4 6.6 BHB at 30 min (n=22) 32.9 9.7 Glu at 30 min (n=21) 75.3 11.1 BHB at 1 hr (n=23) 31.8 10. 5 Glu at 1 hr (n=21) 74.9 7.9 BHB at 2 hr (n=23) 34.9 12.4 Glu at 2 hr (n=22) 74.6 8.5 Table 4 23. Mean percent changes of BHB and Glu after a ketogenic meal Mean perce nt change SD (n=23) BHB (%) Glu (%) 30 minutes 1.719.6 3.014.1 1 hou r 3.732.8 2.310.3 2 hour 6.546.8 2.110.2 Time on diet Patients were stratified by time on diet in months which ranged from 11118 mo. Patients on the diet for 11 40 mo (n=12) were compared to patients on the diet >40 mo (n=11). There were no signi ficant differences for percent changes in glucose. Percent change of BHB at 30min was significantly lower for 11 40 mo group compared to >40 mo ( 9.4; 7.5; p=0.05). In addition, t ime on diet correlated to percent change of BHB at 30 min (r=0.5, r2=0.3, p=0.01) (Figure 4 47). Figure 4 47. Percent change of -hydroxybutyrate (BHB) at 30 minutes versus time in diet in months in patients on KT evaluated prospectively

PAGE 144

144 Age Patients were stratif ied by age in years (range 2.5 19.2 yrs ). Patients <10 yr (n=11) were compared to patients >10 yr old (n=12). Younger pa tients had higher fasting, 30min, and 2hr BHB/Glu (Table 4 24) Fasting BHB/Glu was negatively correlated with age (r= 0.5, r2=0.2, p=0.03) (Figure 448). Table 4 24. Effect of time after meal and age on BHB/Glu BHB/G lu All (n=23) <10yr (n=11) >10yr (n=12) Fasting 0.490.2 0.600.2 0.380.1* 30 minutes 0.460.2 0.550.2 0.380.2* 1 hour 0.450.2 0.510.2 0.380.2# 2 hour 0.490.2 0.600.2 0.390.2* *p<0.05, #p=0.09 comparing <10yr and >10yr Figure 4 48. Fasting hydroxybutyrate (BHB)/glucose (Glu) versus age in years in patients on KT.

PAGE 145

145 Route of feeding Patients were stratified by route of feeding. Orally -fed (Oral) patients (n=10) were compared to tube -fed (TF) patients (n=13). Patients who were fed both orally and via G tube were grouped with the TF patients for comparison. Route of feeding did not affect BHB/Glu at any time point (Table 4 25). Percent changes appear to be different for route of feeding comparisons, but there is large variability in the data w hich may be why no statistical significance was detecte d. However, when patients were stratified according to age and route of feeding, there was a significant difference in the younger group (<10yrs) when comparing orallyfed versus tube -fed patients for percent change of Glu at 30 min (14.28.2%; 7.814.0%, p=0.01) (Table 426). Table 4 25. Effect of time after a meal and route of feeding on BHB/Glu BHB/G lu All (n=23) Oral (n=10) TF (n=8) TF/B oth (n=13) Fasting 0.490.2 0.480.1 0.500.2 0.430.2 30 minutes 0.460.2 0.430.2 0.480.2 0.400.2 1 hour 0.450.2 0.410.1 0.470.2 0.410.2 2 hour 0.490.2 0.490.2 0.480.3 0.460.3 Table 4 26. Effect of age and route of feeding on percent change of BHB and Glu Age and route of feeding <10 yr >10 y r Oral (n=5) TF (n=6) Oral (n=5) TF (n=7) % change of BHB at 30 min 5.823.2 11.87.1 0.813.9 11.025.3 % change of BHB at 1 hr 16.414.0 12.35.6 0.411.7 10.557.1 % change of BHB at 2 hr 12.741.4 6.112.0 4.710.5 14.280.2 % chang e of G lu at 30 min 14.28.2 7.814.0* 3.713.8 6.412.0 % change of G lu at 1 hr 4.76.7 2.48.6 2.06.6 6.616.6 % change of G lu at 2 hr 1.25.4 2.48.3 5.15.9 4.816.2 *p=0.01 Diet P rescription Regression analyses were performed for BHB/Glu and percent changes and ratio, Cal/kg/day, fat, carbohydrate, and protein % of calories, fat and carbohydrate in g/kg/day. A ll correlations (r) were found to be less than 0.5 and r2 less than 0.3. Ratio and fat% of calories

PAGE 146

146 seem relatively important for perce nt change of Glu, but not percent changes of BHB. Correlations for ratio and percent change of Glu we re as follows: at 30min (r= 0.3, r2=0.1, p=0.16) at 1 hr (r= 0.3, r2=0.1, p=0.16) and at 2 hr (r= 0.4, r2=0.14, p=0.09) (Figure 4 49). In addition a we ak, but significant correlation was found for CHO% of calories and fasting BHB/Glu (r=0.4, r2=0.2, p=0.04) (Figure 4 50). There was a trend for the CHO% of calories to be higher in younger (<10 yr, 5.05%) compared to older patients (>10 yr, 2.87%) (p=0.07) Comparison of data Data were available for 15 patients to compare after meal response during a second inpatient EEG visit at the GCRC to their first visit An example of a patients BHB and Glu response at their first visit compared to their second visit 1 year later is represented in Figure 4 51. There was both inter and intra -pati ent variability within the data. Figure 4 49. Percent change of glucose (Glu) at 2 hours versus diet prescription ratio (fat: protein +carbohydrate) in patients on KT.

PAGE 147

147 Fi gure 4 50. Fasting hydroxybutyrate (BHB)/glucose (Glu) versus carbohydrate (CHO) percentage of calories in patients on KT. Figure 4 51. Comparison of after meal blood levels of BHB and Glu in patient KG0040 at 1 year and 2 years on study. F=Fasting, 30m=30 minutes, 1hr =1 hour, 2hr= 2 hour, BHB= -hydroxybutyrate during first year, BHB2= -hydroxybutyrate during second year, Glu= Glucose during first year, Glu2=Glucose during second year.

PAGE 148

148 Summary of results Overall, while there was much variability in the data, small perc ent changes in BHB and glucose were observed after a ketogenic meal (Table 4 23). On average, Glu was 3% higher at 30 minutes after ingestion of a ketogenic meal, a much lower percentage than typically observed after a meal. Results suggest that after a me al, patients on KT may not be stimulating the normal insulin response and are not experiencing a rapid rise in blood glucose. This stabilization of metabolism may be important in mechanism of action. Age contributed to BHB but not Glu and route of feedin g was not a significant factor in BHB or Glu. A summary of the prospective study meal challenge results is presented in Figure 4 52. Figure 4 52. Summary of results for mechanism of action. Dashed lines represent data that is not confirmed and is stil l in progress of validation and identification.

PAGE 149

149 Metabolomics A system was developed and utilized for preliminary metabolomics based analyses. Through analyzing the data both manually and with multiple software systems, it is clear how many different concl u sions can be drawn depending on factors such as data processing normalization techniques and setting tolerance limi ts for mass, retention time, and other parameters Without understanding the data fully, it is difficult to make decisions regarding these factors. Therefore, this issue is being handled within our system by having continuity in some of the investigators from beginning to the end. Plasma from subjects on a ketogenic diet w as prepared for analysis by our r esearch staff and analyzed by the UF M etabolomics Core Identifying changes in metabolites after ingestion of a ketogenic diet may lead to a better understanding of changes in specific metabolic pathways. Additionally, in the future, ch anges in the metabolome will be compared to samples obtai ned from patients receiving KT before and after diet. Ultimately these results may lead to questions that can be tested regarding the mechanism of action of KT. Preliminary Results Proces sed data were available for 10 female subjects before and after diet, 5 PUFA and 5 SFA and 3 replicates each for a total of 60 samples for further me tabolomics analyses, Figure 4 53. Untarged metabolomics After uploading all 60 samples into MetaboAnalyst, a total of 243, 515 peaks were identified with an average of 4 059 p er sample. Following peak matching and alignment 1, 812 peak groups were detected Processed data were downloaded, and then replicates were averaged per subject and re uploaded into MetaboAnalyst. Of the 1, 812 peaks per sample, 1, 796 peak groups were formed

PAGE 150

150 Figure 4 53. Diagram of plasma samples analyzed for preliminary metabolomics analyses using MetaboAnalyst and manual searching in excel. Set I, II, and III indicates when the samples were run. A total of 20 biological samples and 3 technical replicat es each were run through the analyses for a total of 60 samples. T o further limit variation, subjects were stratified based on diet group into PUFA (n=5) and SFA groups (n=5) and uploaded into MetaboAnalyst. Of the 30 PUFA samples, 123, 239 peaks were ident ified with an average of 4 108 per sample. A total of 1, 903 peak groups were formed. For the averaged data, 1 884 peak groups were formed. Of the 30 SFA samples, 120, 276 peaks were identified with an average of 4 009 per group. A total of 1 882 peak groups were formed. For the averaged data, 1 854 peak groups were formed. Univariate analyses, including fold change, t test, and volcano plot, were performed to compare before and after diet peaks among all groups. Paired analyses were performed for the averag ed data. The top 50 peaks (mz/rt) or potential compounds identified by fold change, t -test, and volcano plot analyses for before/after comparison of all data are listed in tables in Appendix C. Additionally, multivariate analysis was performed using PCA and PLS DA. Detailed figures for each MetabAnalyst comparative analysis can also be found in Appendix C. The number of

PAGE 151

151 significant masses identified by each comparison and in each group is listed in Table 4 27. Clearly, reducing the variation in the samples by analyzing the diet groups separately produced more significant masses with the fold change and volcano plot Table 4 27. Number of significant masses identified by univariate analyses in Metaboanalyst Fold change t test Volcano plot All data Average d data All data Averaged data All data Averaged data Paired Before/after (n=10, 3 replicates, 60 samples) 273 269 426 163 181 84 14 P UFA (n=5, 3 replicates, 30 samples) 426 418 340 104 340 59 37 SFA (n=5, 3 replicates, 30 samples) 394 388 352 65 223 4 0 25 Volcano plot analysis takes into account both the fold change and the t -test, therefore it represents the most significant provisional compounds. These are defined as provisional because further validation is needed to confirm identification of the compound. Significant masses resulting from the volcano plot in the paired analyses were searched in HMDB in positive mode with 5 10 ppm tolerance to identify provisional compounds. In the averaged data before/after comparison, 58 provisional compounds, i ncluding isomers were identified in HMDB. One hundred and fifty six p rovisional compounds were found in HMDB for the comparison of the averaged data before/after in the PUFA group and 97 for the SFA group. Specifically for acylcarnitines, Table 4 28 repres ents the provisional acylcarnitines identified with the HMDB search including possible adducts. As expected, different acylcarnitines were observed in the before/after comparisons for the SFA group compared to the PUFA group. Additionally, more

PAGE 152

152 provisional acylcarnitines were detecte d when diets were analyzed as a separate group compared to uploading all data for before/after comparison (Table 4 29). Table 4 28. Provisional acylcarnitines identified by searching significant compounds identified by fold c hange, t test, and volcano plot analyses comparing all data before and after values in HMDB with a 5 10 ppm tolerance range. Name Acyl chain length M W m z Hexanoylcarnitine C6 259.1778 260.1848 L octanoylcarnitine C8 287.2091 288.2161 Decanoylcarnitine C 10 315.2404 316.2474 Dodecanoylcarnitine C12 343.2717 344.2787 Cis 5 tetradecenoylcarnitine C14 371.303 372.31 Trans hexadec 2 enoyl carnitine, palmitoleoylcarnitine C16 397.3187 398.32565 Stearidonyl carnitine C18 433.3187 434.32565 Arachidyl carnitine C20 455.3969 456.40389 Clupanodonylcarnitine, cis docsapentaenoyl carnitine C22 473.3499 474.35694 Trans docosa 4,7,10,13,16 pentaenoyl carnitine C22 473.3499 474.35694 Adrenyl carnitine C22 475.3656 476.37259 Table 4 29. Provisional acylcarnitines ide ntified by searching significant compounds identified by both volcano and fold change analyses comparing before and after values in HMDB with a 5 10 ppm tolerance range. Data compared Statistical test Possible acylcarnitines and direction of change Befo re after avg data Paired, volcano Decanoylcarnitine (up) PUFA avg data Paired, volcano 3 methylglutarylcarnitine (down) Decanoylcarnitine (up) PUFA avg data Paired, FC 3 methylglutarylcarnitine (down) Decanoylcarnitine (up) Stearidonyl carnitine (up) SF A avg data Paired, volcano L octanoylcarnitine (up) 4,8 dimethylnonanoyl carnitine (up) SFA avg data Paired, FC Arachidonyl carnitine (up) L -octanoylcarnitine (up) 4,8 dimethylnonanoyl carnitine (up) Hexacosanoyl carnitine (up) Furthermore, other provis ional compounds identified in the search were different among the SFA and PUFA groups. Some expected provisional compounds in each dietary group such as palmitate in the SFA group and a number of oxidized lipids such as h ydroperoxylinoleic acid in the PUF A group were observed. Phosphatidylcholine (PC) was the only provisional compound

PAGE 153

153 identified in all 3 comparison groups. Some of the differences observed in provisional compounds between dietary groups were that the significant compounds in the SFA group were mostly PC and diglycerides, whereas the significant compounds in the PUFA group were mostly phosphotidylinositol (PI), phosphatidlyethanolamine (PE) and monoglycerides. In addition, the potential acylcarnitines identified were different with L -octanoy l carnitine and 4,8 d imethylnonanoyl carnitine in the SFA group and 3 m ethylglutarylcarnitine in the PUFA group. Provisionally, d ecanoylcarnitine was only detecte d in the before/after comparison of all the data. Targeted metabolomics After searching the raw data for provisional carnitines in all 60 samples, a ll isomers with the same mass were considered the same carnitine, therefore there was a total of 68 possible carnitine structures. Out of 68, 44 were identified as provisional carnitines that occurred in at least 1 sample replicate. Twenty -five were found in at le ast 2 replicates of one sample (Figure 4 5 4 ). Only dimethylnonanoyl carnitine was identified in at least 2 replicates of all females both before and after diet. Dodecenoylcarnitine was identified i n at least 1 replicate of all samples both before and after diet and dodecanoylcarnitine in at least 1 replicate of all samples both before and after diet except for 1 subject before diet. Aspartylcarnitine, hexadecanedioylcarnitine, and arachidonoylcarnit ine were only found before diet. Citrylcarnitine, palmitoylcarnitine, stearidonoylcarnitine were only found after diet. D ocosahexaenoyl L -carnitine or cervonylcarnitine was identified in 4 out of 10 females after diet and only 1 before diet. Stearoylcarnit ine was identified in 3 out of 10 females before diet and only 1 after diet.

PAGE 154

154 Overall, more acylcarnitines were po ssibly identified after diet compared to before; 61 provisional acylcarnitines were identified in at least 2 replicates before whereas 79 prov isional acylcarnitines were identified in at least 2 replicates after diet Figure 4 54. Manual targeted identification of acylcarnitines. X acylcarnitine was found in at least 2 out of 3 replicates; MW=molecular isotopic weight; F1 F10=Female 1 Female 10; orange=before diet, acylcarnitine was only found before diet; blue=after diet, acylcarnitine was only found after diet. Additional R esults Samples from both females and males were prepared by our lab and analyzed using Agilent Quantitative analysis s oftware ( Agilent Technologies, Inc., Santa Clara, California) and internal standards for free carnitine and 10 acylcarnitines by Soledad Cerutti, a PhD student in the Chemistry department and within the UF Metabolomics Core. The following results were obta ined by her and adapted with her permission. The % of total peak area counts was obtained for each sample and represented as a ratio of after diet to before diet (A/B) in which values above 1.0 represent an i ncreased response after diet compared to before diet. Likewise, all values below 1.0 represent a decreased response after

PAGE 155

155 diet. Values were averaged for the 3 replicates within each sample and represented in Figure 4 55 and Figure 4 5 6 Overall, the male samples showed larger changes in these specific acylcarnitines compared to the females after diet. In the males, C6 (hexano yl carnitine) and C18 (stearoylcarnitine) on average had the largest increase after diet and C3 (propionylcarnitine) on average was the only acylcarnitine that showed a decrease af ter diet. In females, C8 (octanoylcarnitine), C10 (decanoylcarnitine), C14 (myristoylcarnitine), and C16 (palmitoylcarnitine) on average decreased after diet, and the largest increase after diet was observed in C4 (butyrylcarnitine), C6, and C18 (Figure 4 55). For C2, or ALC, on average in both males and females, there was an increase after diet (Figure 4 56). Seven out of 10 females and 6 out of 10 males had an increase in C2. Summary of results In general acylcarnitines increased after diet. The acylcarn itines as well as significant provisional compounds identified were different among diet groups using both an untargeted and targeted approach In both males and females on average, ALC was increased after diet. A summary of the results regarding mechanism of action of KT, including the meal challenge test and metabolomics based approach is presented in Figure 4 52. V alidation and identification of provisional compounds are ongoing Additionally, samples have been collected from patients receiving KT and will be used for comparison against samples from healthy subjects The system for the use of metabolomics in this patient population was developed and preliminary analyses have been conducted in healthy subjects receiving the ketogenic diet.

PAGE 156

156 Figure 4 55. Average After diet/B efore diet (A/B) ratio response of the % total peak area counts versus free carnitine (C0) and acylcarnitines in female and male samples All values above the dashed line at 1.0 indicate higher response after diet compared to befor e. [Adapted with permission from Cerutti S, Johnson J, Borum P, Yost R, Powell D. Metabolic Profiling in Plasma samples of Patients Before and After Ketogenic Diet Therapy by Monolithic C18LC/(+)ESI MS. 57th ASMS Conference on Mass Spectrometry and Allied Topics ThP 065 (2009) (304) ]

PAGE 157

157 Figure 4 56. Average After diet/B efore diet (A/B) ratio response of the % total peak area counts for each female and male sample for C2 or acetyl l -carnitine (ALC) All values above the dashed line at 1.0 indicate higher response after diet compared to before. [Adapted with permission from Cerutti S, Johnson J, Borum P, Yost R, Powell D. Metabolic Profiling in Plasma samples of Patients Before and After Ketogenic Diet Therapy by Monolithic C18LC/(+)ESI MS. 57th ASMS Confer ence on Mass Spectrometry and Allied Topics ThP 065 (2009) (304) ]

PAGE 158

158 CHAPTER 5 DISCUSSION AND CONCL USIONS Conclusions Growth Children treated with KT for seizures show significant changes in height over time on diet as wel l as age or possibly length of epil epsy. In general, the age of onset of epilepsy in this patient population is less than 1 year old; therefore, correlations with growth and age may be related to length of epilepsy or time of disease. Consequently, these changes in height may not be directl y due to diet composition. S trong correlations were not detected for calorie or protein intake or percentage of the DRI or RDA and growth parameters similar findings to a recent study by Neal et al. (2008). D ue to the wide variation in percent of recommen ded calories and lack of significant correlations between growth parameters, these data suggest we are unable to accurately assess energy needs in this population using current equations for healthy children. Additiona lly, we found ambulation status to be important in growth parameters and calorie prescription and the youngest age groups had the most significant impact of KT on growth Similarly, Neal et al (2008) found ambulation status and age to be contributing factors to poor growth in patients on KT Likewise, other severely disabled children in the youngest age group have been found to be at greatest risk for poor nutritional and growth status (99).With more data on calorie needs and energy expenditure in this patient population, calories could be bet ter optimized for growth. Currently, there are no data on the energy expenditure caused by seizure activity. Estimating the o ptimal calorie prescription in this patient population may need to include age, ambulation status, disease and activity factors, as well as seizure frequency and intensity.

PAGE 159

159 W hile diet ratio is a common factor considered to be important in this patient population, neither ratio of fat: protein + carbohydrate, nor blood BHB had significant effects on growth parameters In contrast, a re cent paper reported weight, height, BMI, and height velocity were affected by KT and height velocity was negatively correlated to BHB (122). Our data indicate closer attention should be made to the diet prescription, and that other dietary factors calculat ed from the diet prescription, such as CHO and Fat in g/kg/day, need to be evaluated further for their role in KT and growth. M alnutrition, growth abnormalities and micronutrient deficiencies are common in patients with intractable epilepsy not on KT (92, 93). Subsequently these patients are recommended to start KT, a nutritional treatment for their seizures. This therapy then needs to be designed to provide t he nutrients needed to grow, stop seizures and correct existing malnutrition. While patients are o n KT, their growth and nutritional status are more intensely monitored. In addition, if seizures improve, the need for AEDs, which also affect growth and bone density (90, 91) is decreased. Therefore the diet could actually be part of the solution to the g rowth issues and improving the nutritional status in this population instead of the cause. Data from our nave strata, patients in which baseline data were available, indicate growth parameters may actually improve over time on KT. Prospectively, diet and calorie prescription were more closely controlled and monitored as compared to the restrospective study. One complicating factor is the majority of the nave group were ambulatory. Additionally, one example in the literature is improved vitamin D status i n patients after initiation of KT due to supplementation and closer monitoring (305). Overall, more attention needs to be given by the nutrition community to this issue of growth and nutrition in this population.

PAGE 160

160 Given that the patient population with intr actable epilepsy on KT is very diverse and has numerous underlying pathologies the issue of growth in this population m ost likely is multi factorial. The refore, the potential effect of KT on growth calorie restriction, epilepsy, and AEDs/medications as well as the potential interaction of the se factors need to be elucidated. Controlled, randomized clinical trials are needed to investigate th e issue of growth retardation in epilep sy patients with and without KT and to determine the direct role of KT in gr owth with and without calorie restriction Spulber et al (2008) found no correlations between growth and seizure response (122). Furthermore, no studies have found growth retardation to be necessary for effectiveness of therapy. Ultimately, in KT the same nutrients used to treat seizures are neede d for the growth of the patient and therefore the therapy should be designed to optimally facilitate both roles. Lipids Patients on KT in general have some problems with dyslipidemia and t ime on therapy does pl ay a role in the blood lipid profile of patients receiving KT Abnormal lipids appear to be more of a factor initially, yet data suggest levels normalize as the length of time on KT increases There may be an adaptation of metabolism involved with consumin g such a high fat intake for extended periods of time. Additionally, results suggest that orally -fed patients have higher non HDL, LDL, and total cholesterols as compared to tube -fed patients Importantly, the ratio, or amount of fat, was not found to have an impact on blood lipid levels. This may be due in part to the fact that the total fat percent is not very different between ratios, e.g. 3:1 is approximately 87% fat versus a 4:1 is approximately 90% fat. In addition, a 4:1 ratio has been shown to be mo re effective against seizures compared to a 3:1 ratio in a randomized trial (9). However, decreasing the ratio is often a first line of therapy for abnormal lipids in patients receiving KT. A recent study reported that use of lowering the ratio as an inter vention for hypercholesterolemia in

PAGE 161

161 a child on KT resulted in increased seizu res (147). They also reported 60% of patients on KT presen ted with hypercholesterolemia, defined as >200mg/dL which is higher than our data; however they calculated it based on a patients single highest follow up value. Therefore, the type of fat may be more important than the amount of fat for increasing the risk of dyslipidemia i n this patient population. Consequently, dyslipidemia may not be a necessary side effect of the the rapy and t he fatty acid profile should be optimized for KT. I n clinical practice, modifying the type of fat may be a more useful option for alleviating dyslipidemia in individual patients on KT than changing the ratio. The blood lipid data from our prelim inary study evaluating KetogGatorNog, a new dietary formula with a balanced fatty acid profile, compared to traditional therapy are consistent with the hypothesis that a ltering the fatty acid blend results in less severe dysli pidemia. There was no indicati on that a balanced fatty acid blend altered clinical tolerance to the therapy or altered seizure reduct ion in response to the therapy. Additionally, there have been conflicting results in the literature in terms of blood fatty acid levels and seizure response (124, 128). The results of this very preliminary study provide data needed to calculate sample sizes for a clinical trial to compare these two approaches to KT for seizures. The ability of our new formula to maintain lower TG levels compared to traditional therapy to increase the levels of HDL cholesterol and to decrease the levels of LDL cholesterol, could be the beginning of improving the overall nutrient intake of the patients on KT. In addition, a stable emulsion was created using the fat blend tha t may provide a less expensive and more user -friendly application for KT and potentially increase compliance on therapy. Despite the fact that atherogenic lipids are present in this population, there are some patients who present with normal lipids throughout KT; therefore other factors aside from diet,

PAGE 162

162 such as genetics, may also be playing a major role. The issue of fasting as well as diet composition and genetic variation may account for the large standard deviations seen with mean blood lipid values in this population. Further studies should be designed to compare fatty acid intake of KT patients and their effect on seizure control to determine if there is an optimal fatty acid profile for efficacy and tolerability while minimizing potential adverse e ffects. Meal Challenge Overall, BHB and Glu levels are maintained relatively constant after the ingestion of a ketogenic meal. An average response of glucose after a glucose tolerance test in normal children is 53% at 30 min, 30% at 1 hr, and 14% at 2 hr ( 306). Even though the meal was low in carbohydrate and would not be expected to increase the blood glucose to the same extent as a glucose tolerance test, the lack of effect on blood glucose of the KT meal may be an important characteristic of KT. Similarly, clinical experience has shown that cheating, or ingestion of carbohydrates can result in a rapid rise in blood glucose levels, a shift in the metabolism to glycolysis, and an increase in break through seizure activity (1, 61, 79, 80). Therefore, maintai ning low glucose levels throughout the day may be important for seizure control. Average fasting glucose was 72.46 mg/dL in this population, compared to approximately 7926 mg/dL in normal children ( 306) and on average increased by 3% at 30 minutes after ingestion of the ketogenic meal This stabilization of energy metabolism may be a factor in the mechanism of action against seizures. Additionally, o ur data also suggested younger patients had higher BHB compared to older patients, which has been illustrat ed before ( 51, 61, 62). Research suggests that childrens ability to extract ketones is approx imately 4 5 times greater than adults as reviewed in Chapter 2 (51). While there is variability among patients on KT in their response to a ketogenic meal, over all the changes in BHB and Glu are relatively small. It is unknown what the ideal metabolic

PAGE 163

163 response after a meal should be for effectiveness of therapy. Future studies should be designed to test the changes in metabolites after a meal before and after ini tiation of ketogenic therapy and in comparison to se izure response. In addition to the variability across patients in this group, the meal consumed during the meal challenge was specific to each patient and their individual diet prescription. Therefore, th is added to the variability within and across patients. Further analyses are currently underway in our lab to examine the meal consumed during the visit using the Nutrient Data Systems for Research (NDSR) ( Minnesota database ), to determine if specific diet ary components can account for the variability. Metabolomics As seen in the results in Chapter 4, many conclusions could have been drawn depending on the stratification of samples, statistical test used, or the type of software. While many significant mass es were observed in the over 240,000 peaks identified, there is a lot more to understanding these differences and validating identification. Ultimately, in order to define biologically relevant conclusions specific understanding of the whole process is nec essary. It is still early in the process but this project has started the optimization o f the system of metabolomics. Using both manual methods and specific software programs it is clear there are differences in metabolites before and after diet and betw een different dietary fat groups. However, at this point it is difficult to make definitive conclusions regarding the differences. As expected, acylcarnitines were found to be increased after diet as compared to before. Additionally, provisional compounds identified for the SFA group were different than the PUFA group, indicating that type of fat is important in overall alterations in metabolism during KT as well as dyslipidemia. Diseases, as well as dietary interventions, are known to disrupt metabolism and lead to changes in metabolic profiles. As a result, metabolomics can lead to a more comprehensive understanding of mechanism of disease, to new diagnostic biomarkers, and ultimately contribute

PAGE 164

164 to both biomedical research and clinical practice ( 271). Af ter viewing metabolomics data both manually and through multiple software programs, it is clear that results and conclusions published in this field should be reviewed carefully. This project established a system for analysis of metabolomics data and estab lished the importance of continuity throughout the workflow : formulate hypothesis design study collect samples prepare and analyze samples to identify compounds analyze data evaluate hypothesis and generate hypo theses apply to patient care. Fur ther research is underway to identify and validate significant compounds. Additionally, samples have been collected and stored for analysis of patients with intractable epilepsy before and after initiation of KT. The mechanism of KT is still unknown. Metab olomics is a new tool that is ideal for answering questions regarding alterations in the metabolism before and after a ketogenic diet and will lead to further hypothesis generation and hopefully lead to insight into mechanism of action. Implications for Fu ture Research Due to the wide variation in per centage of recommended calories and lack of clear correlations with calorie and growth parameters, traditional formulas used for calculating calorie needs in the healthy population are not adequate for the intr actable epilepsy population on KT. Future research should be designed to understand energy expenditure before and after initiation of KT. Use of indirect calorimetry to predict resting energy expenditure has been shown to be more accurate than formulas and should be used in this population (307). It is also unclear how much energy is expended during seizure activity. In the future, calorie adjustments may also need to be made based on seizure activity changes. Additionally, prospective, controlled trials ne ed to be designed to further understand the issue of growth in the patient populations with and without KT and with adequate calorie prescriptions.

PAGE 165

165 Furthermore, since the results indicate that type of fat is an important risk factor in the blood lipid pro file, future research should focus on improved fatty acid blends for both health and efficacy of therapy. Based on the preliminary results, larger intervention trials should be designed to investigate the incorporation of a balanced fat blend on both the f atty acid profile and on seizure control. Although this study examined the after meal labs in patients already on KT and stabilization of Glu and BHB was observed, it is not clear if the stabilization was directly due to the diet and if the changes in Glu and BHB are related to seizure response. More research is needed to test the after meal labs in patients before and after initiation of KT and compare the Glu and BHB response to seizure frequency and severity. Based on the literature review on ALCs role in neuroprotection and the preliminary results from the metabolomics indicating an increase in ALC in healthy subjects receiving a ketogenic diet, more research is needed to investigate ALCs role in patients on KT. Further research is already underway to compare the changes in the metabolome of patients with intractable epilepsy before and after diet initiation to these healthy subjects. Additionally, future studies should be designed to investigate the use of ALC for improved seizure control in this pati ent population. Design of future studies Based on these results, a change to the IRB/GCRC protocol was made to add the use of indirect calorimetry in this patient population once per year and to extend the length of time in order to gather the energy data and to increase the number of nave patients enrolled. In addition, a balanced fat emulsion similar to the one used in the preliminary study is in the process of being developed. An intervention trial to be conducted in the GCRC has been designed to test t he

PAGE 166

166 incorporation of this emulsion on blood lipids and the use of ALC supplementation on seizure response. The research plan and methods are discussed below. Research p lan and m ethods The following are a research plan and methods for a randomized, contro lled trial involving two patient strata. Stratum 1 patients have been on KT at UF for varying periods of time and stratum 2 patients are referred for initiation to KT E fficacy of KT will be evaluated by assessing seizures and requirements for AEDs. Advers e effects of KT will be evaluated by assessing dyslipidemia and growth. Both strata will attend a baseline clinic visit to determine blood levels and energy needs from indirect calorimetry before randomization. Stratum 1 will be randomized to receive a bal anced fat emulsion incorporated into their ketogenic meals with and without acetyl l -carnitine supplementation. For stratum 2, the baseline visit will be used to determine energy expenditure for assessing calorie needs. They will be randomized to be initia ted to either standard therapy, or therapy with the new balanced fat emulsion. The specific aims and their implementation are briefly described below. Specific Aim : Use the GCRC Sample Processing Laborator y to obtain fasting blood and the sublicense for the Nutrition Data System for Research ( NDSR ) of the GCRC Metabolic Research Unit to analyze daily food diaries to determine the effect of balanced fat blend on blood lipid profile. o Specific Aim A -Comparision of change in fatty acid profiles based on food diaries and change in blood lipids from baseline will be conducted in Stratum 1 o Specific Aim B -Comparison of blood lipids among groups in Stratum 2 Specific Aim : Use of a newly designed seizure score and AED medication score to determine the effect of balanced fat blend on efficacy of KT. o Specific Aim A Comparison of seizure score and AED score among groups in Stratum 1 o Specific Aim B Comparison of seizure score and AED score for among groups in Stratum 2

PAGE 167

167 Specific Aim : To determine the effect of bala nced fat blend on tolerability by evaluating gastrointestinal disturbances (diarrhea, vomiting, constipation, nausea) and taste preferences from questionnaires obtained at follow up clinic visits. o Specific Aim A -Compare tolerability issues among groups in Stratum 1 o Specific Aim B Compare tolerability issues among groups in Stratum 2 Specific Aim : Use of a newly designed seizure score and AED medication score to determine the effect of acetyl -l -carnitine on efficacy of KT. o Specific Aim Comparison of sei zure score and AED score among groups in Stratum 1 Specific Ai m : To determine the effect of acetyl l -carnitine on tolerability by evaluating gastrointestinal disturbances (diarrhea, vomiting, constipation, nausea) and taste preferences from questionnaires obtains at follow up clinic visits. o Specific Aim A -Compare tolerability issues among groups in Stratum 1 o Specific Aim B Compare tolerability issues among groups in Stratum 2 Specific Aim : Use of the GCRC TrueOne 2400 Metabolic Measurement System and G CRC nursing staff to perform indirect calorimetry to improve estimation of energy needs and compare to growth parameters. o Specific Aim A -Compare current calorie prescription to predicted REE and monitor growth parameters (height, weight, BMI, skinfolds, B IA, waist circumference) among patients in Stratum 1 based on calorie changes. o Specific Aim B -Evaluate REE and 3 day food record at baseline for determination of calorie needs at initiation in Stratum 2. Monitor growth parameters (height, weight, BMI, sk infolds, BIA, waist circumference) among patients over time on diet.

PAGE 168

168 APPENDIX A ALC AND KT

PAGE 169

169 Table A 1. Acetyl l -carnitine ( ALC ) and Ketogenic Therapy (KT) similar neuroprotective mechanisms. ALC KT Energy Metabolism Lactate and inorganic phosphate levels were significantly reduced and ATP and creatinephosphate were elevated after ALC treatment. (Aureli et al 1994, 1990). KD increases total brain ATP (DeVivo et al., 1978) and PCr/Cr or PCr/ATP energy reserve (Pan et al., 1999; Bough et al., 2006). PCr/ATP ratio correlated with the recovery of the membrane potential following a stimulus train, which was inversely correlated with granule cell bursting in human temporal lobe epilepsy. (Williamson et al., 2005) I ncrease in the levels of ATP, PC, an d normalized PME levels and high energy phosphate levels were observed in 7 Alzheimers disease patients administered ALC (Pettegrew 1995) K etones may be a more efficient source of energy per unit oxygen than glucose. (Veech 2001) An elevation in PCr:C r energy reserve ratio in both animals and humans (Bough et al., 2006) animals,humans (Pan et al., 1999). ALC feeding to unloaded rats upregulated expression of mitochondrial transcripts (COX I, ATP6, NDP6, 16 S rRNA) and resulted in an increase of trans cript level for factors involved in mitochondrial biogenesis (PGC NRF 1, TFAM) in SM. (Cassano et al 2006) Microarray expression studies demonstrated that KD induces a coordinated upregulation of several dozen metabolic genes associated with ox phos. (Noh et a l., 2004) KD treatment stimulated mitochondrial biogenesis, resulting in a 46% increase in the number of mitochondria in rat hippocampus.(Bough et al., 2006) ALC reduction in cerebral glycolytic flux (Benzi et al 1984) ALC could restore mitochondri al function and/or improve use of energy from glycolysis (Virmani and Binienda et al 2004) ALC is providing an acetyl group for metabolism in the brain, promoting oxidative energy production and minimizing glycolysis and lactic acidosis (Zanelli et al 2005) Reduction in glycolysis (DeVivo et al., 1978; Puchowicz et al., 2005; Melo et al., 2006; Stafstrom et al.2005;Garriga Canut et al.,2006) KD induces UCP expression, stimulates mitochondrial biogenesis, and enhances energy production. (Sullivan et a l., 2004)

PAGE 170

170 Table A 1. Continued. ALC KT Anti oxidant ALC also seems to provide protection during metabolic stress, such as ischemia, hypoxia, aging, alcohol, and brain injury (Virmani and Binienda 2004). K etones reduce glutamate induced free radical formation by increasing the NAD + /NADH ratio and enhancing mitochondrial respiration in neocortical neurons. (Maalouf M et al 2007) LC inhibited the decrease in mitochondrial membrane potential and generation of ROS in hippocampal neuronal cells cultured in glucose deprived medium (Hino et al 2005). ALC led to the up regulation of Hsp60,72, SOD, and a high expression of the redoxsensitive transcription factor Nrf2 (Calabrese et al. 2005;2006). Restored GSH/GSSG ratio and reversed the inhibition of complex IV. ALC decreases both HNE formation and I protein carbonyls. (Calabrese et al. 2006) Ketones potently decrea se ROS generation (Veech et al. 2001; Veech et al. 2004) T reatment with ALC improved nerve conduction velocities and reduced MDA conte nt in diabetic rats. (Lowitt et al 1995) KD induces GPX activity in the r at hippocampus. (Ziegler et al. 2003) Anti apoptotic Toxicity and cell damage, as assessed by LDH release was significantly reduced in cultured rat cortical neurons by both L C and ALC. (Virmani et al 1995) KD was protective against apoptosis in neurons of the hippoc ampus of male mice. (Noh et al. 2003) Protection against mitochondrial alterations and cell death from cytokines along with an increased expression of HO 1 was found in primary rat cortical astrocytes treated with ALC (Calabrese et al 2005). Calbindin is increased in mice on the KD (McIntosh et al. 1998; Noh et al. 2005 ) neuroprotective activity through its capacity to buffer intracellular calcium, which is a mediator of cell d eath (Mattson et al. 1995; Bellido et al. 2000). ALC and LC have been shown to reduce apoptosis through the mitochondrial pathway. (Pillich et al. 2005; Ishii et al. 2000) KD prevented kainic acid induced accumulation of the prot ein clusterin ( Noh et al., 2005 ), clusterin can act as a prodeath signal (Jones and Jomary, 2002).

PAGE 171

171 Table A 1. Continued. ALC KT Neuro modulatory/ Neurotrophic Hyperpolarization and a significant increase in the afterhyperpolarization in the neurons injected with 2 mM ALC a few days before. They also indicate that ALCs action is most likely due to the sustained effect on Na+/K+ ATPase activity. (Lombardo et al 2004) G lucose restriction during KT activates ATP sensitive potassium (KATP) channels (Schwartzkroin et al. 1999; Vamecq et al. 2005) D iabetic rats, ALC for 4mo corrected the Na+/K+ ATPase defect, 63% prevention of the nerve conduction defect. (Sima A et al 1996) ACA and acetone may activate K 2P channels (Vamecq et al. 2005) PUFAs may activate a lipid sensitive class of K2P potassium channels Gene for mitochondrial VDAC is positively mod ulated by treatment with ALC. (Traina et al 2006) PUFAs may enhance the activity of the Na + /K + ATPase (Wu et al. 2004) Microarray study of gene expression in the hipp ocampus of rats fed KD change in the transcripts of a total of 39 genes. Most transcripts were reduced, including the VDAC subunits (Bough et al 2006) Abbreviations: Ketogenic therapy (KT), ketogenic diet (KD), acetylcarnitine (ALC); L -carnitine (LC); p olyunsaturated fatty acids (PUFA); acetoacetate (ACA), beta -hydroxybutyrate (BHB), reactive oxygen species (ROS), phosphocreatine to -creatine (PCr:Cr); Uncoupling proteins (UCP), GPX -glutathione peroxidase; reduced/oxidized glutathione (GSH/GSSG); oxidativ e phosphorylation (ox phos); voltage -dependent anion channel (VDAC); lactate dehydrogenase (LDH); malonyldialdehyde (MDA); skeletal muscle (SM); phosphomonoester (PME); Heat shock protein (Hsp); superoxide dismutase (SOD); 4 -hydroxytransnonenal (HNE) ; Heme -oxygenase 1 (HO 1)

PAGE 172

172 APPENDIX B BLOOD ASSAY Purpose The purpose of the Standard Operating Procedure (SOP) is to provide standard procedures that are used in the Metabolic Assessment Laboratory (MAL) as required by Section 58.81 of the Good Laboratory Practice for Non -clinical Laboratory. Scope This procedure provides instructions for the human blood assay procedure Reagents and Materials 1. Methanol 2. Acetonitrile 3. 5 inch Pasteur pipets 4. Glass beakers 5. P1000 Pipette 6. P200 Pipette 7. 1000 uL Disposable pipette tips 8. 200 uL disposable pipette tips 9 Deionized water 10. Microcentrifuge tubes 11. Transfer pipettes 12. Glass green cap vial 13. Glass black cap vial s Equipment 1. Laminar Flow Hood 2. Microcentrifuge 3. Mixer 4. Evaporating M anifold 5. Nitrogen Tank 6. Scale 7. Face mask 8. Gloves 9. Long Sleeve Lab Coat Safety precautions Members of the MAL have been trained extensively in the procedures described in this SOP. Definitions Standard Operating Procedure (SOP) Standard Operat ing Procedure is a document that provides instructions for completing a specific task in the lab. Metabolic Assessment Laboratory (MAL) The Metabolic Assessment Laboratory is the laboratory that will use this SOP. Day 0 Preparation of RBC hemolysate. Da y 1 Preparation of RBC hemolysate/plasma carnitine extracts and the Hemoglobin hemolysate standard. Day 2 Protein separation and evaporation.

PAGE 173

173 Procedure A. Setup A.1. Prepare labels in the laminar flow hood according to the label template for hemolysa te samples. A.1.a. Be sure to label the hemolysate plastic tubes with the date prepared, and sample ID number A.1.b. Remember to use gloves, a long sleeve lab coat and a facemask when handling tubes and vials containing sample material B. Blood separat ion B.1. Prepare plasma and RBC from blood sample according to SOP describing blood preparation on day of collection. B.2 Make sure all blood separation is also performed in the laminar flow hood C. Day 0 Preparation of RBC Hemolysate Prepared unde r a laminar flow hood C.1. Fill black ice buckets with ice from pilot plant C.1.a. Obtain a plastic beaker and fill with water and ice for thawing C.2 Bring all materials needed to the laminar flow hood C.2.a. Obtain appropriat e RBC/Plasma tubes from freezer C.3. Add approximately an equal amount of deionized water to the tubes of RBCs C. 3 .a. Remember to use gloves, lab coat and a facemask when handling tubes and vials containing sample material C.4. Thaw the RBCs slowly and completely until the mixt ure is homogeneous C. 4 .a. Rub the tube between hands quickly, and swirl in the beaker of ice water. However, do not allow the tube to become too warm. There should be a piece of ice in the tube for as long as it is out of contact with ice and the liquid should always be in constant movement. C.5. Vortex, and freeze overnight C. 5 .a. Tubes can be combined after the addition of water with a transfer pipette to make more appropriate volumes about 2/3 full. However, do not fill the tubes too much as the li quid will expand in the freezer. C. 5 b Remember to use gloves, a lab coat and a facemask when handling tubes and vials containing sample material. Also, remember that all procedures should be performed under the laminar flow hood D. Day 1 Preparatio n of RBC hemolysate/plasma carnitine extracts and the Hemoglobin hemolysate standard D.1. Thaw hemolysate/plasma samples according to procedures described in B D.2. Prepare a plastic tube of Hgb corresponding to each hemolysate tube used. D.1.a. Label a HgB plastic test tube for each hemolysate tube. Include the organism ID, sample ID, HgB, and date the HgB was made D.1.b. Pipet 450L deinozied water into the HgB tube D.1.c Pipet 50L of a hemolysate sample into the HgB tube D.1.d. Vortex the HgB t ube for 10 seconds and freeze overnight D.3. Prepare a 100L sample of RBC hemolysate and/or plasma for carnitine analysis D.3.a. Prepare and label 1.6mL microcentrifuge tubes for each sample. D.3.a.i Label Organism ID + Sample # + Assay ID + B atch

PAGE 174

174 Number D.3.a.i i. Remember to perform the procedure in a laminar flow hood and use gloves and a facemask when handling tubes and vials containing sample material D.3.b. Prepare the A cetonitrile/Methanol solution D.3.c Pipet 1mL of 3:1 A/M so lution to each microcentrifuge tube D.3.d. Under the laminar flow hood, vortex the hemolysate/plasma tube, and pipet 100L of hemolysate/plasma into the microcentrifuge tube once thawing is complete D.3.d.i. In order to make things go smoothly, one person should pipet the sample, while the other vortexes and opens the test tubes D.3.d.ii. Change pipet tips in between each volume of blood pipette D.3.d.iii. Remember to use gloves, a lab coat and a facemask when handling tubes and vials containing sample material. D.3.e. Vortex thoroughly to mix, and freeze overnight D.3.f Once samples have been added to the A/M solution, all possibly infectious agents will be deactivated E. Day 2 Protein separation and evaporation E.1. Remov e the s amples from the freezer E.2. Prepare the samples for the first centrifugation E.2.a Place samples on centrifuging rack and balance on the beam balance E.2.a.i. Use empty microcentrifuge tubes filled with water in each rack to balance properly E.2.b Place the racks into the centrifuge making sure that the samples are arranged properly in the rack so that they balance in the centrifuge E.2.b.i. The tubes should be positioned so that when they are placed in the centrifuge, the racks are 180 degrees apart and the tubes are on the same row in the rack but in the position exactly alternate the one in the other rack. E.2.c. Set the centrifuge to speed 12 (12,000rpm), at 15 minutes. Turn on the centrifuge, push the time button, and flip the start butt on E.2.E. While the centrifuge is running, prepare the microcentrifuge tubes for the second centrifugation E.2.E.i. Label accordingly E.2.E.ii. Prepare Pasteur pipets with bulbs for each sample E.2.E.iii. Remember to use gloves and a facemask when handling tubes and vials containing sample material E.2.e. When the first centrifugation is finished, remove the samples carefully from the centrifuge without jostling the pellets E.2.f. Remove the supernatant from the sample and transfer to the microce ntrifuge tubes for the second centrifugation E.2.f.i. Uncap the empty tube and while holding both tubes in one hand, transfer the supernatant from the first tube into the second tube E.2.f.ii. Make sure to obtain the entire liquid sample while

PAGE 175

175 avoiding t he pellet E.2.f.iii. Slant the tube with the supernatant to one side so that the liquid can be drawn out of the tube without disturbing the supernatant. Also make sure the pipette is touching the side of the tube without residue to avoid pulling in the pe llet E.2.f.iv. Be sure to record the appearance of the pellet and supernatant in the research notebook page E.2.f.v. Remember to use gloves and a facemask when handling tubes and vials containing sample material E.2.g. Repeat for all samples and replace them in the centrifuge rack to balance on the balance beam. E.2.h. Centrifuge a second time for 15 minutes at 12,000rpm. E.2.i. Prepare microcentrifuge tubes labeled with initials and sample number for the third centrifugation E.2.i.i. Remember to use glov es and a facemask when handling tubes and vials containing sample material E.2.j. Remove samples from the centrifuge and repeat steps E.2.e E.2.g E.2.j.ii. Record whether or not there was pellet in the sample tube E.2.k. Centrifuge a third time for 15 min utes at 12,000rpm. E.2.l. Remove the supernatant from the sample and use a plastic transfer pipette to place the supernatant in the black glass vial labeled with the date, sample ID and assay ID. E.2.m. Place samples in the freezer until evaporation. F. Evaporation of samples F.1. Setup: F.1.a. Prepare Nitrogen tank for use F.1.a.i. Make sure nitrogen tank is full or more than 100 lb/psi before use; any less may not be enough for the evaporation. F.1.a.ii. Record the start time and starting pressure o n the nitrogen tank tracking sheet. F.1.a.iii. Tips for Using Nitrogen Tank F.1.a.iii.a. Any tank with air should be bolted down to the ground using an appropriate harness. F.1.a.iii.b. The main valve is used to turn the tank on/off. F.1.a.iii.c. The r egulator increases/decreases the nitrogen pressure as needed. F.1.a.iii.d. The gauge shows the pressure of the air being released. F.1.a.iii.e. At 1000 lbs/psi, reorder nitrogen F.1.b. Prepare the block manifold for use: F.1.b.i. Attach Nitrogen tank to the manifold using tubing. The tubing is attached to a spigot on the back of the block manifold F.1.b.ii. Determine number of samples that will be evaporated and arrange needles on the nitrogen air releasing apparatus. Space them evenly and plug the re maining unused air releasers. F.1.b.iii. Place the sample vials in the block holder

PAGE 176

176 F.1.b.iv. Lower the air releasing apparatus into the sample vials so that the bottom of the needle is not lower than the neck of the vial. F.1.b.iv.a. The needle sho uld not come in contact with the liquid. F.1.b.iv.b. Use a technique of taping the appropriate height for the needle to be placed into each vial. F.1.b.v. Turn on the nitrogen tank F.1.b.v.a. Turn the main valve in the open direction. F.1.b.v.b. Adjus t the pressure to between 2,000 and 2,500 lb/psi using the regulator. F.1.b.v.c. Adjust the knob on the manifold so that the nitrogen is being released through the needles and into the sample. The sample should not be bubbling under the nitrogen air. A puddle should be formed where nitrogen is hitting the sample uniformly. F.1.b.vi. Allow the sample to evaporate with the fume hood completely down, checking the samples every 15 minutes until the sample is dry. F.1.b.vi.a. Lower the needles so that the sa mples are constantly under efficient evaporation but not to the point of bubbling. F.1.c. Turn off nitrogen tank. F.1.c.i. Decrease the pressure by turning the regulator. F.1.c.ii. Turn off the main valve by rotating the valve to the close position. F.1 .c.iii. Increase the pressure by turning the regulator to aspirate any nitrogen in the tank through the opening of the manifold by bleeding the line. F.1.d. Record the stop pressure and time on the nitrogen tank tracking sheet.

PAGE 177

177 APPENDIX C METABOANALYS T TABLES AND FIGURES Before/After diet comparison of all data Table C 1. Summary of data processing results in MetaboAnalyst for all 60 samples, before and after for female subjects. Samples Peaks (raw) Missing/Zero Peaks (processed) AfterF001 FK 1 402 5 0 1812 AfterF001 FK 2 3971 0 1812 AfterF001 FK 3 3751 0 1812 AfterF002 FL 1 4307 0 1812 AfterF002 FL 2 4692 0 1812 AfterF002 FL 3 3726 0 1812 AfterF003 FM 1 4096 0 1812 AfterF003 FM 2 4462 0 1812 AfterF003 FM 3 3818 0 1812 AfterF 004 FN 1 4138 0 1812 AfterF004 FN 2 3945 0 1812 AfterF004 FN 3 4137 0 1812 AfterF005 FO 1 3964 0 1812 AfterF005 FO 2 3820 0 1812 AfterF005 FO 3 3803 0 1812 AfterF006 FP 1 4086 0 1812 AfterF006 FP 2 3802 0 1812 AfterF006 FP 3 4479 0 1812 AfterF007 FQ 1 4008 0 1812 AfterF007 FQ 2 4047 0 1812 AfterF007 FQ 3 3794 0 1812 AfterF008 FR 1 4319 0 1812 AfterF008 FR 2 3683 0 1812 AfterF008 FR 3 3693 0 1812 AfterF009 FS 1 4112 0 1812 AfterF009 FS 2 3455 0 1812 AfterF00 9 FS 3 4740 0 1812 AfterF010 FT 1 4044 0 1812 AfterF010 FT 2 4004 0 1812 AfterF010 FT 3 3886 0 1812 BeforeF001 FA 1 4093 0 1812 BeforeF001 FA 2 3969 0 1812 BeforeF001 FA 3 3952 0 1812 BeforeF002 FB 1 3674 0 1812 BeforeF002 FB 2 4525 0 1812 BeforeF002 FB 3 3866 0 1812 BeforeF003 FC 1 3697 0 1812 BeforeF003 FC 2 4239 0 1812

PAGE 178

178 Table C 1. Continued. Samples Peaks (raw) Missing/Zero Peaks (processed) BeforeF003 FC 3 3741 0 1812 BeforeF004 FD 1 4280 0 1812 BeforeF004 FD 2 4 070 0 1812 BeforeF004 FD 3 3739 0 1812 BeforeF005 FE 1 4195 0 1812 BeforeF005 FE 2 4127 0 1812 BeforeF005 FE 3 3957 0 1812 BeforeF006 FF 1 4246 0 1812 BeforeF006 FF 2 4240 0 1812 BeforeF006 FF 3 3953 0 1812 BeforeF007 FG 1 3965 0 1 812 BeforeF007 FG 2 3879 0 1812 BeforeF007 FG 3 3769 0 1812 BeforeF008 FH 1 4921 0 1812 BeforeF008 FH 2 5039 0 1812 BeforeF008 FH 3 3933 0 1812 BeforeF009 FI 1 4410 0 1812 BeforeF009 FI 2 4580 0 1812 BeforeF009 FI 3 3478 0 1812 Befo reF010 FJ 1 4170 0 1812 BeforeF010 FJ 2 4010 0 1812 BeforeF010 FJ 3 3991 0 1812 Table C 2. Top 50 features identified by fold change analysis of all 60 samples comparing before and after for female subjects. Peaks (mz/rt) Fold Change log2(FC) 5 42.076/5 0.022 5.5035 391.109/29 0.0349 4.8411 729/37 0.037 4.7549 540.209/31 0.0467 4.421 572.46/28 0.0472 4.4039 481.29/29 0.0524 4.2544 595.922/29 0.0789 3.6638 530.213/29 0.0827 3.596 683.976/4 0.0938 3.415 845.451/26 10.5421 3.3981 551.447/35 10.2173 3.3529 546.266/32 0.0988 3.339 671.573/40 10.0715 3.3322 490.192/33 0.1001 3.3205 583.902/32 0.1014 3.3021 348.995/11 0.1124 3.1528 936.669/31 8.6513 3.1129

PAGE 179

179 Table C 2. Continued. Peaks (mz/rt) Fold Change log2(FC) 705.595/3 5 8.5473 3.0955 586.505/29 0.1219 3.0357 634.224/32 0.1294 2.9499 521.109/28 0.1296 2.9478 494.235/33 0.1333 2.9068 478.014/5 0.1374 2.8633 882.603/37 6.7341 2.7515 569.889/5 0.1488 2.7486 695.569/35 6.6719 2.7381 997.643/27 6.3228 2.6606 6 53.476/36 6.2761 2.6499 630.487/29 0.1645 2.6042 734.49/32 0.1649 2.6004 674.438/29 0.1658 2.5927 683.559/42 5.6722 2.5039 480.003/5 0.1768 2.4995 327.358/26 0.1802 2.4721 704.964/4 5.4058 2.4345 760.458/33 0.1979 2.3375 996.685/30 0.1996 2 .3245 664.461/28 4.9925 2.3197 839.551/31 4.8741 2.2851 855.225/40 4.8357 2.2737 500.351/28 0.21 2.2514 780.613/30 4.7497 2.2478 747.587/34 4.7268 2.2409 549.439/36 4.6896 2.2295 289.667/45 0.214 2.2245 745.599/42 0.2142 2.2232 669.47/33 4.642 2.2147 916.561/33 0.218 2.1974 647.488/37 4.3785 2.1304 130.178/16 0.2285 2.1299 Table C 3. Top 50 features identified by t test analysis of all 60 samples comparing before and after for female subjects. Peaks (mz/rt) p value log10(p) 583.90 2/32 0 6.07847 191.181/26 0 5.76796 729/37 1.00E 05 5.01904 286.203/25 2.00E 05 4.71271 403.24/28 2.00E 05 4.65552 615.333/33 8.00E 05 4.08118 209.191/26 2.30E 04 3.6387

PAGE 180

180 Table C 3. Continued. Peaks (mz/rt) p value log10(p) 373.235/28 2.60E 04 3.5 8048 575.366/32 3.50E 04 3.45391 724.929/4 4.30E 04 3.36215 183.102/21 5.30E 04 3.27912 236.223/27 5.30E 04 3.27472 389.232/27 5.60E 04 3.254 312.228/27 5.90E 04 3.23158 144.089/6 5.90E 04 3.23155 330.26/26 6.30E 04 3.19753 1095.652/44 7.30E 04 3. 13655 587.86/5 7.50E 04 3.12708 649.451/30 8.00E 04 3.0981 693.478/30 0.00104 2.9824 569.889/5 0.00111 2.95647 471.941/5 0.00132 2.88017 629.476/37 0.00133 2.87571 424.018/4 0.00136 2.86709 351.217/22 0.00139 2.85844 511.936/5 0.00141 2.85097 396 .01/5 0.00156 2.80552 615.469/34 0.00157 2.80493 396.237/34 0.0016 2.7961 910.589/32 0.00164 2.78459 653.476/36 0.00168 2.77574 839.551/31 0.00173 2.7624 381.232/30 0.00176 2.75435 352.254/30 0.00177 2.75166 480.003/5 0.00189 2.72443 321.236/30 0. 0019 2.72149 109.1/24 0.0019 2.72139 641.46/33 0.00196 2.70865 434.22/30 0.00198 2.70331 595.922/29 0.00199 2.70061 1045.659/27 0.00207 2.68435 89.065/17 0.00212 2.67442 996.685/30 0.00215 2.66849 95.084/22 0.00221 2.65536 359.5/31 0.00232 2.63534 258.113/16 0.0025 2.60203 319.231/28 0.00283 2.54763 130.178/16 0.00284 2.54704 539.06/5 0.00286 2.54417 280.094/6 0.00287 2.54178

PAGE 181

181 Table C 4. Top 50 features identified by volcano plot analysis of all 60 samples comparing before and after for fema le subjects. Peaks (mz/rt) FC log2(FC) p value log10(p) 583.902/32 0.101 3.302 0 6.078 729/37 0.037 4.755 0 5.019 286.203/25 0.35 1.513 0 4.713 575.366/32 0.488 1.036 0 3.454 724.929/4 0.436 1.198 0 3.362 236.223/27 0.271 1.886 0.001 3.275 38 9.232/27 0.38 1.396 0.001 3.254 144.089/6 0.369 1.437 0.001 3.232 1095.652/44 3.865 1.951 0.001 3.137 587.86/5 0.383 1.386 0.001 3.127 649.451/30 0.496 1.011 0.001 3.098 693.478/30 0.488 1.036 0.001 2.982 569.889/5 0.149 2.749 0.001 2.956 471. 941/5 0.365 1.456 0.001 2.88 629.476/37 2.522 1.335 0.001 2.876 424.018/4 0.237 2.08 0.001 2.867 351.217/22 2.725 1.446 0.001 2.858 511.936/5 0.389 1.362 0.001 2.851 396.237/34 0.44 1.183 0.002 2.796 910.589/32 2.476 1.308 0.002 2.785 653.476/36 6.276 2.65 0.002 2.776 839.551/31 4.874 2.285 0.002 2.762 480.003/5 0.177 2.499 0.002 2.724 641.46/33 3.481 1.799 0.002 2.709 595.922/29 0.079 3.664 0.002 2.701 1045.659/27 0.277 1.854 0.002 2.684 89.065/17 0.369 1.437 0.002 2.674 996.685/30 0. 2 2.324 0.002 2.668 359.5/31 0.371 1.43 0.002 2.635 130.178/16 0.228 2.13 0.003 2.547 539.06/5 0.362 1.466 0.003 2.544 135.108/25 0.434 1.205 0.003 2.522 882.603/37 6.734 2.751 0.003 2.516 671.482/32 3.902 1.964 0.004 2.449 289.366/28 0.326 1. 617 0.004 2.385 648.327/27 2.219 1.15 0.004 2.349 356.022/5 0.269 1.892 0.005 2.272 846.577/27 3.135 1.649 0.005 2.269 442.363/34 2.174 1.12 0.006 2.243 600.482/31 2.002 1.001 0.006 2.24 333.147/20 0.407 1.298 0.006 2.216 338.573/46 2.497 1.32 0.0 06 2.204 445.311/31 0.494 1.019 0.006 2.198

PAGE 182

182 Table C 4. Continued. Peaks (mz/rt) FC log2(FC) p value log10(p) 836.558/29 3.23 1.692 0.007 2.166 466.236/26 0.462 1.116 0.007 2.153 452.233/33 0.47 1.09 0.007 2.149 456.28/31 0.492 1.025 0.008 2.111 869.586/30 0.361 1.469 0.008 2.106 660.964/5 0.382 1.39 0.008 2.086 1167.658/27 3.235 1.694 0.008 2.086 Figure C 1 Box plots and kernel density plots produced in MetaboAnalyst before and after normalization of all before and after samples fro m female subjects.

PAGE 183

183 Figure C 2 Fold change analysis of all 60 samples comparing before and after diet with a threshold of 2. Red circles indicate features above the threshold. Figure C 3 Features identified by t tests with a threshold of 0.1 compar ing before and after diet for all samples from female subjects Red circles represent features above the threshold. P values were transformed by -log10 so that the more significant features w ere be plotted higher on the graph.

PAGE 184

184 Figure C 4 Features iden tified by volcano plot analysis with fold change threshold (x) 2 and t tests threshold (y) 0.1 comparing before and after diet for all samples from female subjects R ed circles represent features above the threshold. Note both fold changes and p values are log transformed.

PAGE 185

185 Figure C 5 Score plot between the selected principal components ( PC ) in all 60 samples comparing before and after diet of female subjects The explained variances are shown in brackets. Red triangles represent after diet samples and green + represent before diet samples.

PAGE 186

186 Figure C 6 Partial least squares discriminant analysis (PLS DA) 3D score plot between selected principal components in all 60 samples comparing before and after diet of female subjects. The explained variances ar e shown in parentheses. Red triangles represent after diet samples and green + represent before diet samples.

PAGE 187

187 LITERATURE CITED 1. JM F, JB F, MT K. The Ketogenic Diet. A Treatment for Epilepsy. Third ed. New York, New York: Demos Medical Publishing, Inc.; 2000. 2. Sankar R, Sotero de Menezes M. Metabolic and endocrine aspects of the ketogenic diet. Epilepsy Res. 1999 Dec;37:191201. 3. Vaisleib I, Buchhalter J, Zupanc M. Ketogenic diet: outpatient initiation, without fluid, or caloric restrictions. Pediatr Neurol. 2004 Sep;31:198202. 4. Sinha S, Kossoff E. The ketogenic diet. Neurologist. 2005 May;11:16170. 5. Freeman J, Vining E, Pillas D, Pyzik P, Casey J, Kelly L. The efficacy of the ketogenic diet 1998: a prospective evaluation of intervention in 150 children. Pediatrics. 1998 Dec;102:135863. 6. Groesbeck D, Bluml R, Kossoff E. Long-term use of the ketogenic diet in the treatment of epilepsy. Dev Med Child Neurol. 2006 Dec;48:97881. 7. Henderson C, Filloux F, Alder S, Lyon J, Caplin D. Efficacy of th e ketogenic diet as a treatment option for epilepsy: meta analysis. J Child Neurol. 2006 Mar;21:1938. 8. Hartman A, Gasior M, Vining E, Rogawski M. The neuropharmacology of the ketogenic diet. Pediatr Neurol. 2007 May;36:28192. 9. Seo J, Lee Y, Lee J, Ka ng H, Kim H. Efficacy and tolerability of the ketogenic diet according to lipid:nonlipid ratios --comparison of 3:1 with 4:1 diet. Epilepsia. 2007 Apr;48:8015. 10. Bough K, Rho J. Anticonvulsant mechanisms of the ketogenic diet. Epilepsia. 2007 Jan;48:435 8. 11. Freeman J, Kossoff E, Hartman A. The ketogenic diet: one decade later. Pediatrics. 2007 Mar;119:53543. 12. Hemingway C, Freeman J, Pillas D, Pyzik P. The ketogenic diet: a 3 to 6 -year follow up of 150 children enrolled prospectively. Pediatrics. 2 001 Oct;108:898905. 13. Marsh E, Freeman J, Kossoff E, Vining E, Rubenstein J, Pyzik P, Hemingway C. The outcome of children with intractable seizures: a 3 to 6 -year follow up of 67 children who remained on the ketogenic diet less than one year. Epilepsi a. 2006 Feb;47:42530. 14. Peterson S, Tangney C, Pimentel Zablah E, Hjelmgren B, Booth G, Berry -Kravis E. Changes in growth and seizure reduction in children on the ketogenic diet as a treatment for intractable epilepsy. J Am Diet Assoc. 2005 May;105:71825.

PAGE 188

188 15. Couch S, Schwarzman F, Carroll J, Koenigsberger D, Nordli D, Deckelbaum R, DeFelice A. Growth and nutritional outcomes of children treated with the ketogenic diet. J Am Diet Assoc. 1999 Dec;99:15735. 16. El -Khayat H, Abd El -Basset F, Tomoum H, Toh amy S, Zaky A, Mohamed M, Hakky S, El Barbary N, Nassef N. Physical growth and endocrinal disorders during pubertal maturation in girls with epilepsy. Epilepsia. 2004 Sep;45:110615. 17. Kwiterovich PJ, Vining E, Pyzik P, Skolasky RJ, Freeman J. Effect of a high -fat ketogenic diet on plasma levels of lipids, lipoproteins, and apolipoproteins in children. JAMA. 2003 Aug;290:91220. 18. Kang H, Chung D, Kim D, Kim H. Early and late -onset complications of the ketogenic diet for intractable epilepsy. Epilepsia 2004 Sep;45:111623. 19. Zanelli S, Solenski N, Rosenthal R, Fiskum G. Mechanisms of ischemic neuroprotection by acetyl -L -carnitine. Ann N Y Acad Sci. 2005 Aug;1053:15361. 20. Rochfort S. Metabolomics reviewed: a new "omics" platform technology for sy stems biology and implications for natural products research. J Nat Prod. 2005 Dec;68:181320. 21. Centers for Disease Control and Prevention. National Center for Chronic Disease Prevention and Health Promotion. Epilepsy: frequently asked questions. Availa ble at: http://www.cdc.gov/epilepsy/faqs.htm Accessed: December 11, 2008. 2 2 Lefevre F, Aronson N. Ketogenic diet for the treatment of refractory epilepsy in children: A systematic review of efficacy. Pediatrics. 2000 Apr;105:E46. 2 3 Fisher R, van Emde Boas W, Blume W, Elger C, Genton P, Lee P, Engel JJ. Epileptic seizures and epilepsy: definitions proposed by the International League Against Epilepsy (ILAE) and the International Bureau for Epilepsy (IBE). Epilepsia. 2005 Apr;46:4702. 2 4 Freeman J, Vin ing E. Ketogenic diet: a time tested, effective, and safe method for treatment of intractable childhood epilepsy. Epilepsia. 1998 Apr;39:4501. 2 5 Vining E, Freeman J, Ballaban Gil K, Camfield C, Camfield P, Holmes G, Shinnar S, Shuman R, Trevathan E, Whe less J. A multicenter study of the efficacy of the ketogenic diet. Arch Neurol. 1998 Nov;55:14337. 2 6 D. CD, L S. Circulation and energy metabolism. In: J. SG, W. AB, W. AR, K. FS, D UM, editors. Basic Neurochemistry: Molecular, Cellular and Medical Aspe cts. 6th Edn ed. New York: Lippincott -Raven; 1999. p. 637 69. 2 7 Pellerin L, Bouzier Sore A, Aubert A, Serres S, Merle M, Costalat R, Magistretti P. Activity -dependent regulation of energy metabolism by astrocytes: an update. Glia. 2007 Sep;55:125162.

PAGE 189

189 2 8 McKenna M, Stevenson J, Huang X, Tildon J, Zielke C, Hopkins I. Mitochondrial malic enzyme activity is much higher in mitochondria from cortical synaptic terminals compared with mitochondria from primary cultures of cortical neurons or cerebellar granule cells. Neurochem Int. 2000 Apr;36:4519. 2 9 Waagepetersen H, Sonnewald U, Larsson O, Schousboe A. Multiple compartments with different metabolic characteristics are involved in biosynthesis of intracellular and released glutamine and citrate in astrocyte s. Glia. 2001 Sep;35:24652. 30. Laughton J, Bittar P, Charnay Y, Pellerin L, Kovari E, Magistretti P, Bouras C. Metabolic compartmentalization in the human cortex and hippocampus: evidence for a cell and region -specific localization of lactate dehydrogen ase 5 and pyruvate dehydrogenase. BMC Neurosci. 2007;8:35. 31. Meldrum B, Chapman A. Epileptic seizures and epilepsy. In: J. SG, W. AB, W. AR, K. FS, D. UM, editors. Basic Neurochemistry: Molecular, Cellular and Medical Aspects. 6th Edn ed. New York; 1999. p. 755 68. 3 2 Cornford E, Shamsa K, Zeitzer J, Enriquez C, Wilson C, Behnke E, Fried I, Engel J. Regional analyses of CNS microdialysate glucose and lactate in seizure patients. Epilepsia. 2002 Nov;43:136071. 3 3 Schwechter E, Velskov J, Velsek L. Co rrelation between extracellular glucose and seizure susceptibility in adult rats. Ann Neurol. 2003 Jan;53:91101. 3 4 During M, Fried I, Leone P, Katz A, Spencer D. Direct measurement of extracellular lactate in the human hippocampus during spontaneous sei zures. J Neurochem. 1994 Jun;62:235661. 3 5 Meric P, Barrere B, Peres M, Gillet B, Berenger G, Beloeil J, Seylaz J. Effects of kainateinduced seizures on cerebral metabolism: a combined 1H and 31P NMR study in rat. Brain Res. 1994 Feb;638:5360. 3 6 Corn ford E, Nguyen E, Landaw E. Acute upregulation of blood-brain barrier glucose transporter activity in seizures. Am J Physiol Heart Circ Physiol. 2000 Sep;279:H134654. 3 7 Knowlton R, Abou -Khalil B, Sawrie S, Martin R, Faught R, Kuzniecky R. In vivo hippoc ampal metabolic dysfunction in human temporal lobe epilepsy. Arch Neurol. 2002 Dec;59:18826. 3 8 Ikemoto A, Bole D, Ueda T. Glycolysis and glutamate accumulation into synaptic vesicles. Role of glyceraldehyde phosphate dehydrogenase and 3 -phosphoglycerate kinase. J Biol Chem. 2003 Feb;278:592940. 3 9 Kuhl D, Engel JJ, Phelps M, Selin C. Epileptic patterns of local cerebral metabolism and perfusion in humans determined by emission computed tomography of 18FDG and 13NH3. Ann Neurol. 1980 Oct;8:34860.

PAGE 190

190 40. J anigro D. Blood brain barrier, ion homeostatis and epilepsy: possible implications towards the understanding of ketogenic diet mechanisms. Epilepsy Res. 1999 Dec;37:22332. 41. Vielhaber S, Von Oertzen J, Kudin A, Schoenfeld A, Menzel C, Biersack H, Kral T Elger C, Kunz W. Correlation of hippocampal glucose oxidation capacity and interictal FDG PET in temporal lobe epilepsy. Epilepsia. 2003 Feb;44:1939. 42. Cornford E, Gee M, Swartz B, Mandelkern M, Blahd W, Landaw E, Delgado Escueta A. Dynamic [18F]fluor odeoxyglucose positron emission tomography and hypometabolic zones in seizures: reduced capillary influx. Ann Neurol. 1998 Jun;43:8018. 43. Cornford E, Hyman S, Cornford M, Landaw E, Delgado Escueta A. Interictal seizure resections show two configurations of endothelial Glut1 glucose transporter in the human blood -brain barrier. J Cereb Blood Flow Metab. 1998 Jan;18:2642. 44. Fink G, Pawlik G, Stefan H, Pietrzyk U, Wienhard K, Heiss W. Temporal lobe epilepsy: evidence for interictal uncoupling of blood fl ow and glucose metabolism in temporomesial structures. J Neurol Sci. 1996 Apr;137:2834. 45. A G, C C. Induction processes in blood -brain transfer of ketone bodies during starvation. American Journal of Physiology. 1975;229:11659. 46. DeVivo D, Leckie M, Ferrendelli J, McDougal DJ. Chronic ketosis and cerebral metabolism. Ann Neurol. 1978 Apr;3:33137. 47. Leino R, Gerhart D, Duelli R, Enerson B, Drewes L. Diet -induced ketosis increases monocarboxylate transporter (MCT1) levels in rat brain. Neurochem Int. 2001 May;38:51927. 48. Greene A, Todorova M, Seyfried T. Perspectives on the metabolic management of epilepsy through dietary reduction of glucose and elevation of ketone bodies. J Neurochem. 2003 Aug;86:52937. 49. Musa -Veloso K, Rarama E, Comeau F, Cur tis R, Cunnane S. Epilepsy and the ketogenic diet: assessment of ketosis in children using breath acetone. Pediatr Res. 2002 Sep;52:4438. 50. Morris A. Cerebral ketone body metabolism. J Inherit Metab Dis. 2005;28:10921. 51. Haymond M, Howard C, Ben-Gali m E, DeVivo D. Effects of ketosis on glucose flux in children and adults. Am J Physiol. 1983 Oct;245:E3738. 52. Fery F, Bourdoux P, Christophe J, Balasse E. Hormonal and metabolic changes induced by an isocaloric isoproteinic ketogenic diet in healthy sub jects. Diabete Metab. 1982 Dec;8:299305.

PAGE 191

191 53. Barthel A, Schmoll D. Novel concepts in insulin regulation of hepatic gluconeogenesis. Am J Physiol Endocrinol Metab. 2003 Oct;285:E68592. 54. Fukao T, Lopaschuk G, Mitchell G. Pathways and control of ketone body metabolism: on the fringe of lipid biochemistry. Prostaglandins Leukot Essent Fatty Acids. 2004 Mar;70:24351. 55. Laffel L. Ketone bodies: a review of physiology, pathophysiology and application of monitoring to diabetes. Diabetes Metab Res Rev.15:41226. 56. Williamson D, Bates M, Krebs H. Activity and intracellular distribution of enzymes of ketone -body metabolism in rat liver. Biochem J. 1968 Jul;108:35361. 57. Casals N, Roca N, Guerrero M, Gil Gmez G, Ayt J, Ciudad C, Hegardt F. Regulation of th e expression of the mitochondrial 3 -hydroxy 3 -methylglutaryl -CoA synthase gene. Its role in the control of ketogenesis. Biochem J. 1992 Apr;283 ( Pt 1):2614. 58. Thumelin S, Forestier M, Girard J, Pegorier J. Developmental changes in mitochondrial 3 hydro xy3 -methylglutaryl CoA synthase gene expression in rat liver, intestine and kidney. Biochem J. 1993 Jun;292 ( Pt 2):4936. 59. Arias G, Asins G, Hegardt F, Serra D. The effect of fasting/refeeding and insulin treatment on the expression of the regulatory genes of ketogenesis in intestine and liver of suckling rats. Arch Biochem Biophys. 1997 Apr;340:28798. 60. Wolfrum C, Asilmaz E, Luca E, Friedman J, Stoffel M. Foxa2 regulates lipid metabolism and ketogenesis in the liver during fasting and in diabetes. Nature. 2004 Dec;432:102732. 61. Huttenlocher P. Ketonemia and seizures: metabolic and anticonvulsant effects of two ketogenic diets in childhood epilepsy. Pediatr Res. 1976 May;10:53640. 62. Bonnefont J, Specola N, Vassault A, Lombes A, Ogier H, de Kler k J, Munnich A, Coude M, Paturneau Jouas M, Saudubray J. The fasting test in paediatrics: application to the diagnosis of pathological hypoand hyperketotic states. Eur J Pediatr. 1990 Dec;150:805. 63. Bough K, Chen R, Eagles D. Path analysis shows that increasing ketogenic ratio, but not beta hydroxybutyrate, elevates seizure threshold in the Rat. Dev Neurosci. 1999 Nov;21:4006. 64. Likhodii S, Musa K, Mendonca A, Dell C, Burnham W, Cunnane S. Dietary fat, ketosis, and seizure resistance in rats on the ketogenic diet. Epilepsia. 2000 Nov;41:140010. 65. Todorova M, Tandon P, Madore R, Stafstrom C, Seyfried T. The ketogenic diet inhibits epileptogenesis in EL mice: a genetic model for idiopathic epilepsy. Epilepsia. 2000 Aug;41:93340.

PAGE 192

192 66. Appleton D, DeV ivo D. An animal model for the ketogenic diet. Epilepsia. 1974 Jun;15:21127. 67. Thio L, Wong M, Yamada K. Ketone bodies do not directly alter excitatory or inhibitory hippocampal synaptic transmission. Neurology. 2000 Jan;54:32531. 68. Greene A, Todorov a M, McGowan R, Seyfried T. Caloric restriction inhibits seizure susceptibility in epileptic EL mice by reducing blood glucose. Epilepsia. 2001 Nov;42:13718. 69. Puchowicz M, Emancipator D, Xu K, Magness D, Ndubuizu O, Lust W, LaManna J. Adaptation to chr onic hypoxia during diet induced ketosis. Adv Exp Med Biol. 2005;566:517. 70. Mel T, Nehlig A, Sonnewald U. Neuronal glial interactions in rats fed a ketogenic diet. Neurochem Int. 2006 2006 MayJun;48:498507. 71. Otani K, Yamatodani A, Wada H, Mimaki T Yabuuchi H. [Effect of ketogenic diet on the convulsive threshold and brain amino acid and monoamine levels in young mice]. No To Hattatsu. 1984;16:196204. 72. Pan J, Bebin E, Chu W, Hetherington H. Ketosis and epilepsy: 31P spectroscopic imaging at 4.1 T. Epilepsia. 1999 Jun;40:7037. 73. Bough K, Wetherington J, Hassel B, Pare J, Gawryluk J, Greene J, Shaw R, Smith Y, Geiger J, Dingledine R. Mitochondrial biogenesis in the anticonvulsant mechanism of the ketogenic diet. Ann Neurol. 2006 Aug;60:22335. 74. Yudkoff M, Daikhin Y, Nissim I, Lazarow A. Brain amino acid metabolism and ketosis. J Neurosci Res. 2001 Oct;66:27281. 75. Yudkoff M, Daikhin Y, Nissim I, Lazarow A. Ketogenic diet, amino acid metabolism, and seizure control. J Neurosci Res. 2001 Dec; 66:93140. 76. Stafstrom C, Roopra A, Sutula T. Seizure suppression via glycolysis inhibition with 2 deoxy -D glucose (2DG). Epilepsia. 2008 Nov;49 Suppl 8:97100. 77. Garriga Canut M, Schoenike B, Qazi R, Bergendahl K, Daley T, Pfender R, Morrison J, Ockul y J, Stafstrom C, et al. 2 DeoxyD -glucose reduces epilepsy progression by NRSF CtBP -dependent metabolic regulation of chromatin structure. Nat Neurosci. 2006 Nov;9:13827. 78. Stafstrom C, Ockuly J, Murphree L, Valley M, Roopra A, Sutula T. Anticonvulsant and antiepileptic actions of 2 deoxyD -glucose in epilepsy models. Ann Neurol. 2009 Apr;65:43547. 79. VanItallie T, Nufert T. Ketones: metabolism's ugly duckling. Nutr Rev. 2003 Oct;61:32741.

PAGE 193

193 80. McGhee B, Katyal N. Avoid unnecessary drug related carboh ydrates for patients consuming the ketogenic diet. J Am Diet Assoc. 2001 Jan;101:87101. 81. G. LW. Epilepsy and Related Disorders. Boston: Little, Brown; 1960. 82. Aoki T, Mller W, Brennan M, Cahill GJ. Metabolic effects of glucose in brief and prolonged fasted man. Am J Clin Nutr. 1975 May;28:50711. 83. Pfeifer H, Thiele E. Low -glycemicindex treatment: a liberalized ketogenic diet for treatment of intractable epilepsy. Neurology. 2005 Dec;65:18102. 84. Noh H, Lee H, Kim D, Kang S, Cho G, Rho J, Choi W A cDNA microarray analysis of gene expression profiles in rat hippocampus following a ketogenic diet. Brain Res Mol Brain Res. 2004 Oct;129:807. 85. Veech R, Chance B, Kashiwaya Y, Lardy H, Cahill GJ. Ketone bodies, potential therapeutic uses. IUBMB Lif e. 2001 Apr;51:2417. 86. Kashiwaya Y, Sato K, Tsuchiya N, Thomas S, Fell D, Veech R, Passonneau J. Control of glucose utilization in working perfused rat heart. J Biol Chem. 1994 Oct;269:2550214. 87. Sato K, Kashiwaya Y, Keon C, Tsuchiya N, King M, Radda G, Chance B, Clarke K, Veech R. Insulin, ketone bodies, and mitochondrial energy transduction. FASEB J. 1995 May;9:6518. 88. Kashiwaya Y, King M, Veech R. Substrate signaling by insulin: a ketone bodies ratio mimics insulin action in heart. Am J Cardiol. 1997 Aug;80:50A 64A. 89. Stallings V, Cronk C, Zemel B, Charney E. Body composition in children with spastic quadriplegic cerebral palsy. J Pediatr. 1995 May;126:8339. 90. Tsukahara H, Kimura K, Todoroki Y, Ohshima Y, Hiraoka M, Shigematsu Y, Tsukahara Y Miura M, Mayumi M. Bone mineral status in ambulatory pediatric patients on long term anti -epileptic drug therapy. Pediatr Int. 2002 Jun;44:24753. 91. Nissen Meyer L, Svalheim S, Taubll E, Reppe S, Lekva T, Solberg L, Melhus G, Reinholt F, Gjerstad L, J emtland R. Levetiracetam, phenytoin, and valproate act differently on rat bone mass, structure, and metabolism. Epilepsia. 2007 Oct;48:185060. 92. Marchand V, Motil K. Nutrition support for neurologically impaired children: a clinical report of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition. J Pediatr Gastroenterol Nutr. 2006 Jul;43:12335. 93. Bertoli S, Cardinali S, Veggiotti P, Trentani C, Testolin G, Tagliabue A. Evaluation of nutritional status in children with re fractory epilepsy. Nutr J. 2006;5:14.

PAGE 194

194 94. Eagles D, Boyd S, Kotak A, Allan F. Calorie restriction of a high -carbohydrate diet elevates the threshold of PTZ induced seizures to values equal to those seen with a ketogenic diet. Epilepsy Res. 2003 Apr;54:415 2. 95. Singhi P, Jagirdar S, Khandelwal N, Malhi P. Epilepsy in children with cerebral palsy. J Child Neurol. 2003 Mar;18:1749. 96. palsy in north -eastern Poland. Brain Dev. 2003 Oct;25:499506. 97. Kwong K, Wong S, So K. Epilepsy in children with cerebral palsy. Pediatr Neurol. 1998 Jul;19:316. 98. Azcue M, Zello G, Levy L, Pencharz P. Energy expenditure and body composition in children with spastic quadriplegic cerebra l palsy. J Pediatr. 1996 Dec;129:8706. 99. Dahl M, Thommessen M, Rasmussen M, Selberg T. Feeding and nutritional characteristics in children with moderate or severe cerebral palsy. Acta Paediatr. 1996 Jun;85:697701. 100. Thommessen M, Kase B, Riis G, Hei berg A. The impact of feeding problems on growth and energy intake in children with cerebral palsy. Eur J Clin Nutr. 1991 Oct;45:47987. 101. El -Khayat H, Shatla H, Ali G, Abdulgani M, Tomoum H, Attya H. Physical and hormonal profile of male sexual develop ment in epilepsy. Epilepsia. 2003 Mar;44:44752. 102. Ali I, Schuh L, Barkley G, Gates J. Antiepileptic drugs and reduced bone mineral density. Epilepsy Behav. 2004 Jun;5:296300. 103. Babayigit A, Dirik E, Bober E, Cakmakci H. Adverse effects of antiepile ptic drugs on bone mineral density. Pediatr Neurol. 2006 Sep;35:17781. 104. Tekgul H, Serdaroglu G, Huseyinov A, Gkben S. Bone mineral status in pediatric outpatients on antiepileptic drug monotherapy. J Child Neurol. 2006 May;21:4114. 105. Kossoff E, Z upec -Kania B, Amark P, Ballaban Gil K, Bergqvist A, Blackford R, Buchhalter J, Caraballo R, Cross J, et al. Optimal clinical management of children receiving the ketogenic diet: Recommendations of the International Ketogenic Diet Study Group. Epilepsia. 20 08 Sep. 106. Guo C, Ronen G, Atkinson S. Longterm valproate and lamotrigine treatment may be a marker for reduced growth and bone mass in children with epilepsy. Epilepsia. 2001 Sep;42:11417. 107. Richard D, Ferland J, Lalonde J, Samson P, Deshaies Y. In fluence of topiramate in the regulation of energy balance. Nutrition. 2000 Oct;16:9616.

PAGE 195

195 108. Lampl Y, Eshel Y, Rapaport A, Sarova -Pinhas I. Weight gain, increased appetite, and excessive food intake induced by carbamazepine. Clin Neuropharmacol. 1991 Jun; 14:2515. 109. El -Khatib F, Rauchenzauner M, Lechleitner M, Hoppichler F, Naser A, Waldmann M, Trinka E, Unterberger I, Bauer G, Luef G. Valproate, weight gain and carbohydrate craving: a gender study. Seizure. 2007 Apr;16:22632. 110. Winnacker J, Yeager H, Saunders J, Russell B, Anast C. Rickets in children receiving anticonvulsant drugs. Biochemical and hormonal markers. Am J Dis Child. 1977 Mar;131:28690. 111. Lifshitz F, Maclaren N. Vitamin D dependent rickets in institutionalized, mentally retarded c hildren receiving long term anticonvulsant therapy. I. A survey of 288 patients. J Pediatr. 1973 Oct;83:61220. 112. Keck E, Gollnick B, Reinhardt D, Karch D, Peerenboom H, Krskemper H. Calcium metabolism and vitamin D metabolite levels in children receiv ing anticonvulsant drugs. Eur J Pediatr. 1982 Sep;139:525. 113. Morijiri Y, Sato T. Factors causing rickets in institutionalised handicapped children on anticonvulsant therapy. Arch Dis Child. 1981 Jun;56:4469. 114. Kruse K, Kracht U, Gpfert G. Response of kidney and bone to parathyroid hormone in children receiving anticonvulsant drugs. Neuropediatrics. 1982 Feb;13:39. 115. Weinstein R, Bryce G, Sappington L, King D, Gallagher B. Decreased serum ionized calcium and normal vitamin D metabolite levels with anticonvulsant drug treatment. J Clin Endocrinol Metab. 1984 Jun;58:10039. 116. Volpe S, Schall J, Gallagher P, Stallings V, Bergqvist A. Nutrient intake of children with intractable epilepsy compared with healthy children. J Am Diet Assoc. 2007 Jun;107:10148. 117. Bergqvist A, Trabulsi J, Schall J, Stallings V. Growth failure in children with intractable epilepsy is not due to increased resting energy expenditure. Dev Med Child Neurol. 2008 Jun;50:43944. 118. Liu Y, Williams S, Basualdo Hammond C, St ephens D, Curtis R. A prospective study: growth and nutritional status of children treated with the ketogenic diet. J Am Diet Assoc. 2003 Jun;103:70712. 119. Williams S, Basualdo Hammond C, Curtis R, Schuller R. Growth retardation in children with epileps y on the ketogenic diet: a retrospective chart review. J Am Diet Assoc. 2002 Mar;102:4057. 120. Vining E, Pyzik P, McGrogan J, Hladky H, Anand A, Kriegler S, Freeman J. Growth of children on the ketogenic diet. Dev Med Child Neurol. 2002 Dec;44:796802.

PAGE 196

196 1 21. Neal E, Chaffe H, Edwards N, Lawson M, Schwartz R, Cross J. Growth of children on classical and medium -chain triglyceride ketogenic diets. Pediatrics. 2008 Aug;122:e33440. 122. Spulber G, Spulber S, Hagens L, Amark P, Dahlin M. Growth dependence on i nsulin like growth factor 1 during the ketogenic diet. Epilepsia. 2008 Aug. 123. bone mineral density. Pediatr Neurol. 2004 Oct;31:27982. 124. Dahlin M, Hjelte L, Nilsson S, Amark P. Plasma phospholipid fatty acids are influenced by a ketogenic diet enriched with n3 fatty acids in children with epilepsy. Epilepsy Res. 2007 Feb;73:199207. 125. Fuehrlein B, Rutenberg M, Silver J, Warren M, Theriaque D, Duncan G, Stacpoole P Brantly M. Differential metabolic effects of saturated versus polyunsaturated fats in ketogenic diets. J Clin Endocrinol Metab. 2004 Apr;89:16415. 126. Schwartz R, Boyes S, AynsleyGreen A. Metabolic effects of three ketogenic diets in the treatment of severe epilepsy. Dev Med Child Neurol. 1989 Apr;31:15260. 127. Cunnane S. Metabolism of polyunsaturated fatty acids and ketogenesis: an emerging connection. Prostaglandins Leukot Essent Fatty Acids. 2004 Mar;70:23741. 128. Fraser D, Whiting S, Andrew R, Macdonald E, Musa -Veloso K, Cunnane S. Elevated polyunsaturated fatty acids in blood serum obtained from children on the ketogenic diet. Neurology. 2003 Mar;60:10269. 129. Cunnane S, Musa K, Ryan M, Whiting S, Fraser D. Potential role of polyunsaturates i n seizure protection achieved with the ketogenic diet. Prostaglandins Leukot Essent Fatty Acids.67:1315. 130. Taha A, Ryan M, Cunnane S. Despite transient ketosis, the classic high-fat ketogenic diet induces marked changes in fatty acid metabolism in rats Metabolism. 2005 Sep;54:112732. 131. Gavino G, Gavino V. Rat liver outer mitochondrial carnitine palmitoyltransferase activity towards long-chain polyunsaturated fatty acids and their CoA esters. Lipids. 1991 Apr;26:26670. 132. Cunnane S, Ryan M, Nadea u C, Bazinet R, Musa -Veloso K, McCloy U. Why is carbon from some polyunsaturates extensively recycled into lipid synthesis? Lipids. 2003 Apr;38:47784. 133. Sinclair A. Incorporation of radioactive polyunsaturated fatty acids into liver and brain of develo ping rat. Lipids. 1975 Mar;10:17584.

PAGE 197

197 134. Hassam A, Sinclair A, Crawford M. The incorporation of orally fed radioactive gamma linolenic acid and linoleic acid into the liver and brain lipids of suckling rats. Lipids. 1975 Jul;10:41720. 135. Raclot T, Gro scolas R. Differential mobilization of white adipose tissue fatty acids according to chain length, unsaturation, and positional isomerism. J Lipid Res. 1993 Sep;34:151526. 136. Raclot T, Groscolas R. Selective mobilization of adipose tissue fatty acids during energy depletion in the rat. J Lipid Res. 1995 Oct;36:216473. 137. Voskuyl R, Vreugdenhil M, Kang J, Leaf A. Anticonvulsant effect of polyunsaturated fatty acids in rats, using the cortical stimulation model. Eur J Pharmacol. 1998 Jan;341:14552. 138. Xiao Y, Li X. Polyunsaturated fatty acids modify mouse hippocampal neuronal excitability during excitotoxic or convulsant stimulation. Brain Res. 1999 Oct;846:11221. 139. Sullivan P, Rippy N, Dorenbos K, Concepcion R, Agarwal A, Rho J. The ketogenic die t increases mitochondrial uncoupling protein levels and activity. Ann Neurol. 2004 Apr;55:57680. 140. Vamecq J, Valle L, Lesage F, Gressens P, Stables J. Antiepileptic popular ketogenic diet: emerging twists in an ancient story. Prog Neurobiol. 2005 Jan; 75:128. 141. Schlanger S, Shinitzky M, Yam D. Diet enriched with omega 3 fatty acids alleviates convulsion symptoms in epilepsy patients. Epilepsia. 2002 Jan;43:1034. 142. Bromfield E, Dworetzky B, Hurwitz S, Eluri Z, Lane L, Replansky S, Mostofsky D. A randomized trial of polyunsaturated fatty acids for refractory epilepsy. Epilepsy Behav. 2008 Jan;12:18790. 143. Yuen A, Sander J, Fluegel D, Patsalos P, Bell G, Johnson T, Koepp M. Omega 3 fatty acid supplementation in patients with chronic epilepsy: a r andomized trial. Epilepsy Behav. 2005 Sep;7:2538. 144. Chesney D, Brouhard B, Wyllie E, Powaski K. Biochemical abnormalities of the ketogenic diet in children. Clin Pediatr (Phila). 1999 Feb;38:1079. 145. Dekaban A. Plasma lipids in epileptic children tr eated with the high fat diet. Arch Neurol. 1966 Aug;15:17784. 146. Katyal N, Koehler A, McGhee B, Foley C, Crumrine P. The ketogenic diet in refractory epilepsy: the experience of Children's Hospital of Pittsburgh. Clin Pediatr (Phila). 2000 Mar;39:1539.

PAGE 198

198 147. Nizamuddin J, Turner Z, Rubenstein J, Pyzik P, Kossoff E. Management and risk factors for dyslipidemia with the ketogenic diet. J Child Neurol. 2008 Jul;23:75861. 148. Krauss R, Eckel R, Howard B, Appel L, Daniels S, Deckelbaum R, Erdman JJ, Kris Etherton P, Goldberg I, et al. AHA Dietary Guidelines: revision 2000: A statement for healthcare professionals from the Nutrition Committee of the American Heart Association. Circulation. 2000 Oct;102:228499. 149. Hang T. Clinical trial food database and caregiver educational module for ketogenic therapy. Gainesville, FL: University of Florida; 2000. 150. Hoppel C, Davis A. Inter tissue relationships in the synthesis and distribution of carnitine. Biochem Soc Trans. 1986 Aug;14:6734. 151. Ramsay R, Zam mit V. Carnitine acyltransferases and their influence on CoA pools in health and disease. Mol Aspects Med. 2004 2004 Oct Dec;25:47593. 152. Foster D. The role of the carnitine system in human metabolism. Ann N Y Acad Sci. 2004 Nov;1033:116. 153. Spiekerkoetter U, Tokunaga C, Wendel U, Mayatepek E, Exil V, Duran M, Wijburg F, Wanders R, Strauss A. Changes in blood carnitine and acylcarnitine profiles of very longchain acyl CoA dehydrogenase -deficient mice subjected to stress. Eur J Clin Invest. 2004 Mar ;34:1916. 154. Meyburg J, Schulze A, Kohlmueller D, Pschl J, Linderkamp O, Hoffmann G, Mayatepek E. Acylcarnitine profiles of preterm infants over the first four weeks of life. Pediatr Res. 2002 Nov;52:7203. 155. Wood J, Magera M, Rinaldo P, Seashore M Strauss A, Friedman A. Diagnosis of very long chain acyl dehydrogenase deficiency from an infant's newborn screening card. Pediatrics. 2001 Jul;108:E19. 156. Wikoff W, Gangoiti J, Barshop B, Siuzdak G. Metabolomics identifies perturbations in human disor ders of propionate metabolism. Clin Chem. 2007 Dec;53:216976. 157. Meyburg J, Schulze A, Kohlmueller D, Linderkamp O, Mayatepek E. Postnatal changes in neonatal acylcarnitine profile. Pediatr Res. 2001 Jan;49:1259. 158. Clague A, Thomas A. Neonatal bioch emical screening for disease. Clin Chim Acta. 2002 Jan;315:99110. 159. Chace D, Kalas T, Naylor E. The application of tandem mass spectrometry to neonatal screening for inherited disorders of intermediary metabolism. Annu Rev Genomics Hum Genet. 2002;3:17 45.

PAGE 199

199 160. Koeberl D, Young S, Gregersen N, Vockley J, Smith W, Benjamin DJ, An Y, Weavil S, Chaing S, et al. Rare disorders of metabolism with elevated butyryl and isobutyryl carnitine detected by tandem mass spectrometry newborn screening. Pediatr Res. 2003 Aug;54:21923. 161. Cavedon C, Bourdoux P, Mertens K, Van Thi H, Herremans N, de Laet C, Goyens P. Age related variations in acylcarnitine and free carnitine concentrations measured by tandem mass spectrometry. Clin Chem. 2005 Apr;51:74552. 162. Rinaldo P, Cowan T, Matern D. Acylcarnitine profile analysis. Genet Med. 2008 Feb;10:1516. 163. Roediger W, Heyworth M, Willoughby P, Piris J, Moore A, Truelove S. Luminal ions and short chain fatty acids as markers of functional activity of the mucosa in ulcerative colitis. J Clin Pathol. 1982 Mar;35:3236. 164. Bene J, Komlsi K, Gasztonyi B, Juhsz M, Tulassay Z, Melegh B. Plasma carnitine ester profile in adult celiac disease patients maintained on long term gluten free diet. World J Gastroenterol. 20 05 Nov;11:66715. 165. Bene J, Komlsi K, Havasi V, Talin G, Gasztonyi B, Horvth K, Mzsik G, Hunyady B, Melegh B, Figler M. Changes of plasma fasting carnitine ester profile in patients with ulcerative colitis. World J Gastroenterol. 2006 Jan;12:1103. 166. Pasquali M, Monsen G, Richardson L, Alston M, Longo N. Biochemical findings in common inborn errors of metabolism. Am J Med Genet C Semin Med Genet. 2006 May;142C:6476. 167. Orlandi A, Francesconi A, Marcellini M, Di Lascio A, Spagnoli L. Propionyl -L -carnitine reduces proliferation and potentiates Bax related apoptosis of aortic intimal smooth muscle cells by modulating nuclear factor kappaB activity. J Biol Chem. 2007 Feb;282:493242. 168. Pearson D, Tubbs P. Carnitine and derivatives in rat tissu es. Biochem J. 1967 Dec;105:95363. 169. Frohlich J, Seccombe D, Hahn P, Dodek P, Hynie I. Effect of fasting on free and esterified carnitine levels in human serum and urine: correlation with serum levels of free fatty acids and beta -hydroxybutyrate. Meta bolism. 1978 May;27:55561. 170. Genuth S, Hoppel C. Plasma and urine carnitine in diabetic ketosis. Diabetes. 1979 Dec;28:10837. 171. Hoppel C, Genuth S. Carnitine metabolism in normal -weight and obese human subjects during fasting. Am J Physiol. 1980 May;238:E40915. 172. Murakami R, Tanaka A, Nakamura H. The effect of starvation on brain carnitine concentration in neonatal rats. J Pediatr Gastroenterol Nutr. 1997 Oct;25:3857.

PAGE 200

200 173. Costa C, de Almeida I, Jakobs C, Poll The B, Duran M. Dynamic change s of plasma acylcarnitine levels induced by fasting and sunflower oil challenge test in children. Pediatr Res. 1999 Oct;46:4404. 174. functions in the brain. Mol Aspects Med. 2004 Oct -Dec;25:55167. 175. Ebert D, Haller R, Walton M. Energy contribution of octanoate to intact rat brain metabolism measure d by 13C nuclear magnetic resonance spectroscopy. J Neurosci. 2003 Jul;23:592835. 176. -a known compound, a novel function in neural cells. Acta Neurobiol Exp (Wars). 1996;56:597609. 177. Bresolin N, Freddo L, Vergani L, A ngelini C. Carnitine, carnitine acyltransferases, and rat brain function. Exp Neurol. 1982 Nov;78:28592. 178. Rebouche C, Engel A. Tissue distribution of carnitine biosynthetic enzymes in man. Biochim Biophys Acta. 1980 Jun;630:229. 179. Bird M, Munday L, Saggerson E, Clark J. Carnitine acyltransferase activities in rat brain mitochondria. Bimodal distribution, kinetic constants, regulation by malonyl CoA and developmental pattern. Biochem J. 1985 Feb;226:32330. 180. Kido Y, Tamai I, Ohnari A, Sai Y, Kagami T, Nezu J, Nikaido H, Hashimoto N, Asano M, Tsuji A. Functional relevance of carnitine transporter OCTN2 to brain distribution of L -carnitine and acetyl -L -carnitine across the blood-brain barrier. J Neurochem. 2001 Dec;79:95969. 181. Nakanishi T, Hatanaka T, Huang W, Prasad P, Leibach F, Ganapathy M, Ganapathy V. Na+ and Cl --coupled active transport of carnitine by the amino acid transporter ATB(0,+) from mouse colon expressed in HRPE cells and Xenopus oocytes. J Physiol. 2001 Apr;532:297304. 182. Sloan J, Mager S. Cloning and functional expression of a human Na(+) and Cl( ) dependent neutral and cationic amino acid transporter B(0+). J Biol Chem. 1999 Aug;274:237405. 183. Miecz D, Januszewicz E, Czeredys M, Hinton B, Berezowski V, Cecchelli R, Localization of organic cation/carnitine transporter (OCTN2) in cells forming the bloodbrain barrier. J Neurochem. 2008 Jan;104:11323. 184. synthesis of acetylchol ine in rat cerebral cortex cells. Neurochem Int. 1995 Jun;26:63541. 185. Falchetto S, Kato G, Provini L. The action of carnitines on cortical neurons. Can J Physiol Pharmacol. 1971 Jan;49:17.

PAGE 201

2 01 186. Forloni G, Angeretti N, Smiroldo S. Neuroprotective activity of acetyl L -carnitine: studies in vitro. J Neurosci Res. 1994 Jan;37:926. 187. Spagnoli A, Lucca U, Menasce G, Bandera L, Cizza G, Forloni G, Tettamanti M, Frattura L, Tiraboschi P, Comelli M. Long -term acetyl L -carnitine treatment in Alzheimer's di sease. Neurology. 1991 Nov;41:172632. 188. Shug A, Schmidt M, Golden G, Fariello R. The distribution and role of carnitine in the mammalian brain. Life Sci. 1982 Dec;31:286974. 189. Parnetti L, Gaiti A, Mecocci P, Cadini D, Senin U. Pharmacokinetics of IV and oral acetyl L -carnitine in a multiple dose regimen in patients with senile dementia of Alzheimer type. Eur J Clin Pharmacol. 1992;42:8993. 190. Aureli T, Di Cocco M, Puccetti C, Ricciolini R, Scalibastri M, Miccheli A, Manetti C, Conti F. Acetyl -L -carnitine modulates glucose metabolism and stimulates glycogen synthesis in rat brain. Brain Res. 1998 Jun;796:7581. 191. Gorini A, D'Angelo A, Villa R. Action of L acetylcarnitine on different cerebral mitochondrial populations from cerebral cortex. N eurochem Res. 1998 Dec;23:148591. 192. Calabrese V, Giuffrida Stella A, Calvani M, Butterfield D. Acetylcarnitine and cellular stress response: roles in nutritional redox homeostasis and regulation of longevity genes. J Nutr Biochem. 2006 Feb;17:7388. 1 93. Virmani A, Binienda Z. Role of carnitine esters in brain neuropathol ogy. Mol Aspects Med. 2004 Oct Dec;25:53349 194. Pettegrew J, Levine J, McClure R. Acetyl -L -carnitine physical -chemical, metabolic, and therapeutic properties: relevance for its mode of action in Alzheimer's disease and geriatric depression. Mol Psychiatry. 2000 Nov;5:61632. 195. Farrell S, Vogel J, Bieber L. Entry of acetyl L -carnitine into biosynthetic pathways. Biochim Biophys Acta. 1986 Mar;876:1757. s 196. Ricciolini R, Scalib astri M, Kelleher J, Carminati P, Calvani M, Arduini A. Role of acetyl L -carnitine in rat brain lipogenesis: implications for polyunsaturated fatty acid biosynthesis. J Neurochem. 1998 Dec;71:25107. 197. Pettegrew J, Klunk W, Kanal E, Panchalingam K, McC lure R. Changes in brain membrane phospholipid and high -energy phosphate metabolism precede demen tia. Neurobiol Aging. 1995 Nov Dec;16:9735. 198. Aureli T, Miccheli A, Ricciolini R, Di Cocco M, Ramacci M, Angelucci L, Ghirardi O, Conti F. Aging brain: ef fect of acetyl -L -carnitine treatment on rat brain energy and phospholipid metabolism. A study by 31P and 1H NMR spectroscopy. Brain Res. 1990 Aug;526:10812.

PAGE 202

202 199. Aureli T, Miccheli A, Di Cocco M, Ghirardi O, Giuliani A, Ramacci M, Conti F. Effect of acet yl L -carnitine on recovery of brain phosphorus metabolites and lactic acid level during reperfusion after cerebral ischemia in the rat --study by 13P and 1H -NMR spectroscopy. Brain Res. 1994 Apr;643:929. 200. Rosenthal R, Williams R, Bogaert Y, Getson P, Fiskum G. Prevention of postischemic canine neurological injury through potentiation of brain energy metabolism by acetyl -L carnitine. Stroke. 1992 Sep;23:13127; discussion 7 8. 201. Conley K, Marcinek D, Villarin J. Mitochondrial dysfunction and age. C urr Opin Clin Nutr Metab Care. 2007 Nov;10:68892. 202. Hagen T, Ingersoll R, Wehr C, Lykkesfeldt J, Vinarsky V, Bartholomew J, Song M, Ames B. Acetyl L -carnitine fed to old rats partially restores mitochondrial function and ambulatory activity. Proc Natl Acad Sci U S A. 1998 Aug;95:95626. 203. Ames B, Liu J. Delaying the mitochondrial decay of aging with acetylcarnitine. Ann N Y Acad Sci. 2004 Nov;1033:10816. 204. Aureli T, Di Cocco M, Capuani G, Ricciolini R, Manetti C, Miccheli A, Conti F. Effect of long -term feeding with acetyl -L -carnitine on the age related changes in rat brain lipid composition: a study by 31P NMR spectroscopy. Neurochem Res. 2000 Mar;25:3959. 205. Hagen T, Liu J, Lykkesfeldt J, Wehr C, Ingersoll R, Vinarsky V, Bartholomew J, Ame s B. Feeding acetyl -L -carnitine and lipoic acid to old rats significantly improves metabolic function while decreasing oxidative stress. Proc Natl Acad Sci U S A. 2002 Feb;99:18705. 206. Cassano P, Sciancalepore A, Pesce V, Flck M, Hoppeler H, Calvani M Mosconi L, Cantatore P, Gadaleta M. Acetyl L -carnitine feeding to unloaded rats triggers in soleus muscle the coordinated expression of genes involved in mitochondrial biogenesis. Biochim Biophys Acta. 2006 Sep Oct;1757:14218. 207. Blokland A, Bothmer J Honig W, Jolles J. Behavioural and biochemical effects of acute central metabolic inhibition: effects of acetyl l -carnitine. Eur J Pharmacol. 1993 Apr;235:27581. 208. Arduini A, Mancinelli G, Radatti G, Dottori S, Molajoni F, Ramsay R. Role of carnitin e and carnitine palmitoyltransferase as integral components of the pathway for membrane phospholipid fatty acid turnover in intact human erythrocytes. J Biol Chem. 1992 Jun;267:1267381. 209. Arduini A, Denisova N, Virmani A, Avrova N, Federici G, Arrigon i -Martelli E. Evidence for the involvement of carnitine -dependent long -chain acyltransferases in neuronal triglyceride and phospholipid fatty acid turnover. J Neurochem. 1994 Apr;62:15308.

PAGE 203

203 210. Virmani M, Biselli R, Spadoni A, Rossi S, Corsico N, Calvani M, Fattorossi A, De Simone C, Arrigoni -Martelli E. Protective actions of L -carnitine and acetyl -L -carnitine on the neurotoxicity evoked by mitochondrial uncoupling or inhibitors. Pharmacol Res. 1995 Dec;32:3839. 211. Virmani M, Caso V, Spadoni A, Rossi S, Russo F, Gaetani F. The action of acetyl L carnitine on the neurotoxicity evoked by amyloid fragments and peroxide on primary rat cortical neurones. Ann N Y Acad Sci. 2001 Jun;939:16278. 212. Farooqui A, Horrocks L, Farooqui T. Modulation of inflammat ion in brain: a matter of fat. J Neurochem. 2007 May;101:57799. 213. Sderberg M, Edlund C, Kristensson K, Dallner G. Fatty acid composition of brain phospholipids in aging and in Alzheimer's disease. Lipids. 1991 Jun;26:421-5. 214. Farooqui A, Rapoport S, Horrocks L. Membrane phospholipid alterations in Alzheimer's disease: deficiency of ethanolamine plasmalogens. Neurochem Res. 1997 Apr;22:5237. 215. Murphy E, Schapiro M, Rapoport S, Shetty H. Phospholipid composition and levels are altered in Down s yndrome brain. Brain Res. 2000 Jun;867:918. 216. Puri B. Impaired phospholipid -related signal transduction in advanced Huntington's disease. Exp Physiol. 2001 Sep;86:6835. 217. Infante J, Huszagh V. On the molecular etiology of decreased arachidonic (2 0:4n 6), docosapentaenoic (22:5n6) and docosahexaenoic (22:6n3) acids in Zellweger syndrome and other peroxisomal disorders. Mol Cell Biochem. 1997 Mar;168:10115. 218. Infante J. A function for the vitamin E metabolite alpha -tocopherol quinone as an es sential enzyme cofactor for the mitochondrial fatty acid desaturases. FEBS Lett. 1999 Mar;446:15. 219. Infante J, Huszagh V. Secondary carnitine deficiency and impaired docosahexaenoic (22:6n 3) acid synthesis: a common denominator in the pathophysiology of diseases of oxidative phosphorylation and beta -oxidation. FEBS Lett. 2000 Feb;468:15. 220. Horrocks L, Farooqui A. Docosahexaenoic acid in the diet: its importance in maintenance and restoration of neural membrane function. Prostaglandins Leukot Essent Fatty Acids. 2004 Apr;70:36172. 221. Hinerfeld D, Traini M, Weinberger R, Cochran B, Doctrow S, Harry J, Melov S. Endogenous mitochondrial oxidative stress: neurodegeneration, proteomic analysis, specific respiratory chain defects, and efficacious antio xidant therapy in superoxide dismutase 2 null mice. J Neurochem. 2004 Feb;88:65767. 222. Virmani A, Gaetani F, Binienda Z, Xu A, Duhart H, Ali S. Role of mitochondrial dysfunction in neurotoxicity of MPP+: partial protection of PC12 cells by acetyl L car nitine. Ann N Y Acad Sci. 2004 Oct;1025:26773.

PAGE 204

204 223. Bodis Wollner I, Chung E, Ghilardi M, Glover A, Onofrj M, Pasik P, Samson Y. Acetyl levo -carnitine protects against MPTP induced parkinsonism in primates. J Neural Transm Park Dis Dement Sect. 1991;3:6372. 224. Mazzio E, Yoon K, Soliman K. Acetyl -L -carnitine cytoprotection against 1 -methyl 4 phenylpyridinium toxicity in neuroblastoma cells. Biochem Pharmacol. 2003 Jul;66:297306. 225. Calabrese V, Ravagna A, Colombrita C, Scapagnini G, Guagliano E, Ca lvani M, Butterfield D, Giuffrida Stella A. Acetylcarnitine induces heme oxygenase in rat astrocytes and protects against oxidative stress: involvement of the transcription factor Nrf2. J Neurosci Res. 2005 Feb;79:50921. 226. Calabrese V, Colombrita C, Su ltana R, Scapagnini G, Calvani M, Butterfield D, Stella A. Redox modulation of heat shock protein expression by acetylcarnitine in aging brain: relationship to antioxidant status and mitochondrial functi on. Antioxid Redox Signal. 2006 Mar -Apr;8:40416. 227. Poon H, Calabrese V, Calvani M, Butterfield D. Proteomics analyses of specific protein oxidation and protein expression in aged rat brain and its modulation by L acetylcarnitine: insights into the mechanisms of action of this proposed therapeutic agent for CNS disorders associated with oxidative stre ss. Antioxid Redox Signal. 2006 Mar Apr;8:38194. 228. Lowitt S, Malone J, Salem A, Korthals J, Benford S. Acetyl L -carnitine corrects the altered peripheral nerve function of experimental diabetes. Metabolis m. 1995 May;44:67780. 229. Furuno T, Kanno T, Arita K, Asami M, Utsumi T, Doi Y, Inoue M, Utsumi K. Roles of long chain fatty acids and carnitine in mitochondrial membrane permeability transition. Biochem Pharmacol. 2001 Oct;62:103746. 230. Pillich R, Sc arsella G, Risuleo G. Reduction of apoptosis through the mitochondrial pathway by the administration of acetyl -L -carnitine to mouse fibroblasts in culture. Exp Cell Res. 2005 May;306:18. 231. Ishii T, Shimpo Y, Matsuoka Y, Kinoshita K. Anti apoptotic effe ct of acetyl l -carnitine and I carnitine in primary cultured neurons. Jpn J Pharmacol. 2000 Jun;83:11924. 232. Furlong JH. Acetyl -L Carnitine: Metabolism and Applications in Clinical Practice. Alternative Medicine Review. 1996;1:8593. 233. Foreman P, Per ez-Polo J, Angelucci L, Ramacci M, Taglialatela G. Effects of acetyl L carnitine treatment and stress exposure on the nerve growth factor receptor (p75NGFR) mRNA level in the central nervous system of aged rats. Prog Neuropsychopharmacol Biol Psychiatry. 1 995 Jan;19:11733.

PAGE 205

205 234. Sima A, Ristic H, Merry A, Kamijo M, Lattimer S, Stevens M, Greene D. Primary preventive and secondary interventionary effects of acetyl L -carnitine on diabetic neuropathy in the bio-breeding Worcester rat. J Clin Invest. 1996 Apr;97:19007. 235. Lombardo P, Scuri R, Cataldo E, Calvani M, Nicolai R, Mosconi L, Brunelli M. Acetyl L -carnitine induces a sustained potentiation of the afterhyperpolarization. Neuroscience. 2004;128:293303. 236. Traina G, Valleggi S, Bernardi R, Rizzo M, Calvani M, Nicolai R, Mosconi L, Durante M, Brunelli M. Identification of differentially expressed genes induced in the rat brain by acetyl L -carnitine as evidenced by suppression subtractive hybridisation. Brain Res Mol Brain Res. 2004 Dec;132:5763. 237. Traina G, Bernardi R, Rizzo M, Calvani M, Durante M, Brunelli M. Acetyl -L -carnitine up -regulates expression of voltage-dependent anion channel in the rat brain. Neurochem Int. 2006 Jun;48:6738. 238. Chiechio S, Copani A, De Petris L, Morales M, Nicoletti F, Gereau Rt. Transcriptional regulation of metabotropic glutamate receptor 2/3 expression by the NF kappaB pathway in primary dorsal root ganglia neurons: a possible mechanism for the analgesic effect of L acetylcarnitine. Mol Pain. 2006;2:20. 239. Ando S, Tadenuma T, Tanaka Y, Fukui F, Kobayashi S, Ohashi Y, Kawabata T. Enhancement of learning capacity and cholinergic synaptic function by carnitine in aging rats. J Neurosci Res. 2001 Oct;66:26671. 240. Imperato A, Ramacci M, Angelucci L. Acetyl L -carnit ine enhances acetylcholine release in the striatum and hippocampus of awake freely moving rats. Neurosci Lett. 1989 Dec;107:2515. 241. Fariello R, Ferraro T, Golden G, DeMattei M. Systemic acetyl -L carnitine elevates nigral levels of glutathione and GABA. Life Sci. 1988;43:28992. 242. Sershen H, Harsing LJ, Banay Schwartz M, Hashim A, Ramacci M, Lajtha A. Effect of acetyl L -carnitine on the dopaminergic system in aging brain. J Neurosci Res. 1991 Nov;30:5559. 243. Bhaumik S, Smith E, Shilatifard A. Coval ent modifications of histones during development and disease pathogenesis. Nat Struct Mol Biol. 2007 Nov;14:100816. 244. Chiechio S, Caricasole A, Barletta E, Storto M, Catania M, Copani A, Vertechy M, Nicolai R, Calvani M, et al. L -Acetylcarnitine induce s analgesia by selectively up regulating mGlu2 metabotropic glutamate receptors. Mol Pharmacol. 2002 May;61:98996. 245. Filipek P, Juranek J, Nguyen M, Cummings C, Gargus J. Relative carnitine deficiency in autism. J Autism Dev Disord. 2004 Dec;34:61523.

PAGE 206

206 246. Beal M. Bioenergetic approaches for neuroprotection in Parkinson's disease. Ann Neurol. 2003;53 Suppl 3:S39 47; discussion S 8. 247. Pueschel S. The effect of acetyl L -carnitine administration on persons with Down syndrome. Res Dev Disabil. 2006 N ovDec;27:599604. 248. Wainwright M, Mannix M, Brown J, Stumpf D. L -carnitine reduces brain injury after hypoxia -ischemia in newborn rats. Pediatr Res. 2003 Nov;54:68895. 249. Goety C, Tanner C, Cohen J, Thelen J, Carroll V, Klawans H, Fariello R. L -ace tyl carnitine in Huntington's disease: double -blind placebo controlled crossover study of drug effects on movement disorder and dementia. Mov Disord. 1990;5:263 5. 250. Sorbi S, Forleo P, Fani C, Piacentini S. Double blind, crossover, placebo-controlled c linical trial with L acetylcarnitine in patients with degenerative cerebellar ataxia. Clin Neuropharmacol. 2000 Mar -Apr;23:1148. 251. Salvioli G, Neri M. L acetylcarnitine treatment of mental decline in the elderly. Drugs Exp Clin Res. 1994;20:16976. 252. Genazzani E. [A controlled clinical study on the efficacy of L acetylcarnitine in the treatment of mild -to -moderate mental deterioration in the aged. Conclusions]. Clin Ter. 1990 Mar;132:5112. 253. Herrmann W, Dietrich B, Hiersemenzel R. Pharmaco -electroencephalographic and clinical effects of the cholinergic substance -acetyl -L -carnitine --in patients with organic brain syndrome. Int J Clin Pharmacol Res. 1990;10:814. 254. Matsuoka M, Igisu H. Comparison of the effects of L -carnitine, D -carnitine an d acetyl L carnitine on the neurotoxicity of ammonia. Biochem Pharmacol. 1993 Jul;46:15964. 255. Keating D. Mitochondrial dysfunction, oxidative stress, regulation of exocytosis and their relevance to neurodegenerative diseases. J Neurochem. 2008 Jan;104:298305. 256. Palay S. Synapses in the central nervous system. J Biophys Biochem Cytol. 1956 Jul;2:193202. 257. growth factor and acetyl L -carnitine evoked shifts in acety l -CoA and cholinergic SN56 cell vulnerability to neurotoxic inputs. J Neurosci Res. 2005 Jan 115;79:18592. 258. Dhitavat S, Ortiz D, Shea T, Rivera E. Acetyl L -carnitine protects against amyloid -beta neurotoxicity: roles of oxidative buffering and ATP l evels. Neurochem Res. 2002 Jun;27:5015.

PAGE 207

207 259. Montgomery S, Thal L, Amrein R. Meta analysis of double blind randomized controlled clinical trials of acetyl L -carnitine versus placebo in the treatment of mild cognitive impairment and mild Alzheimer's disea se. Int Clin Psychopharmacol. 2003 Mar;18:6171. 260. Thal L, Calvani M, Amato A, Carta A. A 1 -year controlled trial of acetyl l -carnitine in early onset AD. Neurology. 2000 Sep;55:80510. 261. McGarry J, Robles -Valdes C, Foster D. Role of carnitine in he patic ketogenesis. Proc Natl Acad Sci U S A. 1975 Nov;72:43858. 262. Parvin R, Pande S. Enhancement of mitochondrial carnitine and carnitine acylcarnitine translocase -mediated transport of fatty acids into liver mitochondria under ketogenic conditions. J Biol Chem. 1979 Jun;254:54239. 263. Brass E, Hoppel C. Relationship between acid-soluble carnitine and coenzyme A pools in vivo. Biochem J. 1980 Sep;190:495504. 264. Hack A, Busch V, Pascher B, Busch R, Bieger I, Gempel K, Baumeister F. Monitoring of ket ogenic diet for carnitine metabolites by subcutaneous microdialysis. Pediatr Res. 2006 Jul;60:936. 265. Gravina E, Gravina -Sanvitale G. Effect of carnitine on blood acetoacetate in fasting children. Clin Chim Acta. 1969 Feb;23:3767. 266. Kuratsune H, Wat anabe Y, Yamaguti K, Jacobsson G, Takahashi M, Machii T, Onoe H, Onoe K, Matsumura K, et al. High uptake of [211C]acetyl L -carnitine into the brain: a PET study. Biochem Biophys Res Commun. 1997 Feb;231:48893. 267. Liu J, Head E, Gharib A, Yuan W, Ingers oll R, Hagen T, Cotman C, Ames B. Memory loss in old rats is associated with brain mitochondrial decay and RNA/DNA oxidation: partial reversal by feeding acetyl L -carnitine and/or R alpha lipoic acid. Proc Natl Acad Sci U S A. 2002 Feb;99:235661. 268. Li S, Liu Y, Sigmon V, McCort A, Ren J. High -fat diet enhances visceral advanced glycation end products, nuclear O Glc Nac modification, p38 mitogenactivated protein kinase activation and apoptosis. Diabetes Obes Metab. 2005 Jul;7:44854. 269. Kussmann M, R aymond F, Affolter M. OMICS -driven biomarker discovery in nutrition and health. J Biotechnol. 2006 Aug;124:75887. 270. Nelson D and Cox M. Principles of Biochemistry. Fifth ed New Youl, New York: W.H. Freeman & Co.; 2007. 271. Kaddurah -Daouk R, Kristal B, Weinshilboum R. Metabolomics: a global biochemical approach to drug response and disease. Annu Rev Pharmacol Toxicol. 2008;48:653 83.

PAGE 208

208 272. Fiehn O. Metabolomics -the link between genotypes and phenotypes. Plant Mol Biol. 2002 Jan;48:15571. 273. German J, Hammock B, Watkins S. Metabolomics: building on a century of biochemistry to guide human health. Metabolomics. 2005 Mar;1:3 9. 274. Underwood B, Broadhurst D, Dunn W, Ellis D, Michell A, Vacher C, Mosedale D, Kell D, Barker R, et al. Huntington disease patients and transgenic mice have similar pro catabolic serum metabolite profiles. Brain. 2006 Apr;129:87786. 275. Han X, M Holtzman D, McKeel DJ, Kelley J, Morris J. Substantial sulfatide deficiency and ceramide elevation in very early Alzheimer's dis ease: potential role in disease pathogenesis. J Neurochem. 2002 Aug;82:80918. 276. Rozen S, Cudkowicz M, Bogdanov M, Matson W, Kristal B, Beecher C, Harrison S, Vouros P, Flarakos J, et al. Metabolomic analysis and signatures in motor neuron disease. Met abolomics. 2005;1:1018. 277. Paige L, Mitchell M, Krishnan K, Kaddurah Daouk R, Steffens D. A preliminary metabolomic analysis of older adults with and without depression. Int J Geriatr Psychiatry. 2007 May;22:41823. 278. Holmes E, Tsang T, Huang J, Le weke F, Koethe D, Gerth C, Nolden B, Gross S, Schreiber D, et al. Metabolic profiling of CSF: evidence that early intervention may impact on disease progression and outcome in schizophrenia. PLoS Med. 2006 Aug;3:e327. 279. Tsang T, Huang J, Holmes E, Bah n S. Metabolic profiling of plasma from discordant schizophrenia twins: correlation between lipid signals and global functioning in female schizophrenia patients. J Proteome Res. 2006 Apr;5:75660. 280. Kaddurah -Daouk R. Metabolic profiling of patients wi th schizophrenia. PLoS Med. 2006 Aug;3:e363. 281. Sabatine M, Liu E, Morrow D, Heller E, McCarroll R, Wiegand R, Berriz G, Roth F, Gerszten R. Metabolomic identification of novel biomarkers of myocardial ischemia. Circulation. 2005 Dec;112:386875. 282. Brindle J, Antti H, Holmes E, Tranter G, Nicholson J, Bethell H, Clarke S, Schofield P, McKilligin E, et al. Rapid and noninvasive diagnosis of the presence and severity of coronary heart disease using 1H NMR based metabonomics. Nat Med. 2002 Dec;8:143944. 283. Brindle J, Nicholson J, Schofield P, Grainger D, Holmes E. Application of chemometrics to 1H NMR spectroscopic data to investigate a relationship between human serum metabolic profiles and hypertension. Analyst. 2003 Jan;128:326.

PAGE 209

209 284. Wang C, Kong H, Guan Y, Yang J, Gu J, Yang S, Xu G. Plasma phospholipid metabolic profiling and biomarkers of type 2 diabetes mellitus based on high-performance liquid chromatography/electrospray mass spectrometry and multivariate statistical analysis. Anal Chem. 2005 Jul;77:410816. 285. Yuan K, Kong H, Guan Y, Yang J, Xu G. A GC -based metabonomics investigation of type 2 diabetes by organic acids metabolic profile. J Chromatogr B Analyt Technol Biomed Life Sci. 2007 May;850:23640. 286. Odunsi K, Wollman R, Ambrosone C, Hutson A, McCann S, Tammela J, Geisler J, Miller G, Sellers T, et al. Detection of epithelial ovarian cancer using 1H NMR -based metabonomics. Int J Cancer. 2005 Feb;113:7828. 287. Fan X, Bai J, Shen P. Diagnosis of breast cancer using HPLC metabonomi cs fingerprints coupled with computational methods. Conf Proc IEEE Eng Med Biol Soc. 2005;6:60814. 288. Tolstikov V, Fiehn O, and Tanaka N. Application of Liquid Chromatography Mass Spectrometry Analysis in Metabolomics Reversed -Phase Monolithic Capillar y Chromatography and Hydrophilic Chromatography Coupled to Electrospray Ionization Mass Spectrometry From: Methods in Molecular Biology, vol. 358: Metabolomics: Methods and Protocols Edited by: W. Weckwerth, Humana Press Inc., Totowa, NJ. 2006; 141155. 289. Want E, Nordstrm A, Morita H, Siuzdak G. From exogenous to endogenous: the inevitable imprint of mass spectrometry in metabolomics. J Proteome Res. 2007 Feb;6:45968. 290. Smith M, Vermeulen R, Li G, Zhang L, Lan Q, Hubbard A, Forrest M, McHale C, Z hao X, et al. Use of 'Omic' technologies to study humans exposed to benzene. Chem Biol Interact. 2005 May;153154:1237. 291. Wishart D. Human Metabolome Database: completing the 'human parts list'. Pharmacogenomics. 2007 Jul;8:6836. 292. Zeisel S, Freak e H, Bauman D, Bier D, Burrin D, German J, Klein S, Marquis G, Milner J, et al. The nutritional phenotype in the age of metabolomics. J Nutr. 2005 Jul;135:16136. 293. Stella C, Beckwith Hall B, Cloarec O, Holmes E, Lindon J, Powell J, van der Ouderaa F, Bingham S, Cross A, Nicholson J. Susceptibility of human metabolic phenotypes to dietary modulation. J P roteome Res. 2006 Oct;5:27808. 294. German J, Watkins S, Fay L. Metabolomics in practice: emerging knowledge to guide future dietetic advice toward indi vidualized health. J Am Diet Assoc. 2005 Sep;105:142532.

PAGE 210

210 295. Zeisel S. Nutrigenomics and metabolomics will change clinical nutrition and public health practice: insights from studies on dietary requirements for choline. Am J Clin Nutr. 2007 Sep;86:5428. 296. Minkler P, Stoll M, Ingalls S, Yang S, Kerner J, Hoppel C. Quantification of carnitine and acylcarnitines in biological matrices by HPLC electrospray ionization -mass spectrometry. Clin Chem. 2008 Sep;54:145162. 297. Minkler P, Ingalls S, Hoppel C. Strategy for the isolation, derivatization, chromatographic separation, and detection of carnitine and acylcarnitines. Anal Chem. 2005 Mar;77:144857. 298. Kuczmarski R, Ogden C, Guo S, Grummer Strawn L, Flegal K, Mei Z, Wei R, Curtin L, Roche A, Johnson C 2000 CDC Growth Charts for the United States: methods and development. Vital Health Stat 11. 2002 May:1 190. 299. Group. WMGRS. WHO Child Growth Standards: Length/height -for age, weight -for age, weight -for length, weight -for -height and body mass index-f or age: Methods and development.: Geneva: World Health Organization; 2006. p. 312. 300. Frisancho A. Anthropometric Standards for the Assessment of Growth and Nutritional Status. Ann Arbor, MI: University of Michigan Press; 1990. 301. Dietary Reference Int akes For Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids. Washington, D.C.: THE NATIONAL ACADEMIES PRESS; 2005. 302. Third report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluatio n, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) Final Report. National Heart, Lung, and Blood Institute NIH Publication No. 02 5215, September 2002. 303. Xia J, Psychogios N, Young N, Wishart D. MetaboAnalyst: a web server for metabolomic data analysis and interpretation. Nucleic Acids Res. 2009 May. 304. Cerutti S, Johnson J, Borum P, Yost R, Powell D. Metabolic Profiling in Plasma samples of Patients Before and After Ketogenic Diet Therapy by Monolithic -C18LC/(+)ESI -MS. 57th ASMS Conference on Mass Spectrometry and Allied Topics ThP 065 (2009) 305. Bergqvist A, Schall J, Stallings V. Vitamin D status in children with intractable epilepsy, and impact of the ketogenic diet. Epilepsia. 2007 Jan;48:6671. 306. Knopf C, Cres to J, Dujovne I, Ramos O, de Majo S. Oral glucose tolerance test in 100 normal children. Acta Diabetol Lat.14:95 103. 307. da Rocha E, Alves V, da Fonseca R. Indirect calorimetry: methodology, instruments and clinical application. Curr Opin Clin Nutr Meta b Care. 2006 May;9:24756.

PAGE 211

211 BIOGRAPHICAL SKETCH Mrs. Lauren Little Jones was born and raised in Fernandina Beach, Florida and received her high school diploma in 2001 from Fernandina Beach High S chool. She graduated summa cum laude from the University of F lorida and received her Bachelor of Science in food s cience and h uman n utrition with a specialization in n utritional s ciences in May 2005. In summer 2005, Lauren began her graduate studies in nutritional sciences. Throughout graduate school, Lauren receive d the Checkers Scholarship in 2008, the James Davidson Travel Grant Scholarship, Delta Zeta Graduate Scholarship, FSHN Graduate Student Travel Award s Graduate Student Council travel grants and IFAS t ravel grants. Lauren has presented her research at Exper imental Biology in 2005, 2006, 2007, 2008, and 2009 and also presented an oral presentation at A merican Society for P arenteral and E nteral N utrition Clinical Nutrition Week in 2005. Lauren was the departmental graduate student representative and worked wit h the Graduate Student Council from Fall 2006 until Fall 2008. She has also been a student member of the American So ciety of Nutrition since 2007. Lauren received her PhD from the University of Florida in August 2009. Her academic interests include clini c al and translational research involving topics such as nutrition and disease; brain energy metabolism; chronic illness and energy expenditure; childhood malnutrition; growth and disease; lipid metabolism; public heath nutrition; and metabolomics. Lauren an d her husband, Lance currently reside in Gainesville, FL with their dog Mikey.