Short-terrm effects of altering the dietary carbohydrate to fat ratio on circulating leptin and satiety in women

Transcription

1 University of Wollongong Thesis Collections University of Wollongong Thesis Collection University of Wollongong Year 2004 Short-terrm effects of altering the dietary carbohydrate to fat ratio on circulating leptin and satiety in women Michelle A. Gordon University of Wollongong Gordon, Michelle A, Short-terrm effects of altering the dietary carbohydrate to fat ratio on circulating leptin and satiety in women, PhD thesis, Department of Biomedical Science, University of Wollongong, This paper is posted at Research Online.

2

3 SHORT-TERM EFFECTS OF ALTERING THE DIETARY CARBOHYDRATE TO FAT RATIO ON CIRCULATING LEPTIN AND SATIETY IN WOMEN A thesis submitted in fulfilment of the requirements for the degree of DOCTOR OF PHILOSOPHY from UNIVERSITY OF WOLLONGONG by MICHELLE A. GORDON, BSc (Honours), MSc (Nutrition & Dietetics) DEPARTMENT OF BIOMEDICAL SCIENCE 2004

4 Declaration I, Michelle A. Gordon, declare that this thesis, submitted in fulfilment of the requirements for the award of Doctor of Philosophy, in the Department of Biomedical Science, University of Wollongong, is wholly my own work unless otherwise referenced or acknowledged. All work in this thesis has not been submitted for qualifications at any other academic institution. Michelle A. Gordon 15 th December 2004 Wollongong, Australia i

5 Abstract Background: Overweight and obesity have reached epidemic proportions and appetite control may be important for its clinical management. Leptin is a plasma protein secreted from adipose tissue that is involved in body weight regulation. The role of leptin in regulating human appetite is not well established. Human feeding studies report that an increased dietary carbohydrate to fat ratio is associated with increased circulating leptin concentrations compared to a decreased dietary carbohydrate to fat ratio. These investigations have generally used diets with extreme variations in macronutrient intake that do not represent normal patterns of consumption. Whether less extreme variations in macronutrient intake have similar effects on circulating leptin and whether there is a relationship with satiety under these conditions is not established. The overall aim of this research was to determine the potential clinical relevance of the effects of altering the dietary carbohydrate to fat ratio on circulating leptin concentrations. Methods: Three short-term controlled human feeding studies were conducted involving 68 female subjects (age 37 ± 9 (SD) yrs, BMI 26.8 ± 4.1 kg/m 2 ). Study 1 and Study 2 were single-blind parallel design trials where realistic high carbohydrate (carbohydrate:fat= 60:20) or high fat (carbohydrate:fat= 40:40) isocaloric diets were provided. Study 3 was a double-blind cross-over trial where high carbohydrate (carbohydrate:fat= 60:20) and extreme high fat (carbohydrate:fat= 25:55) diets were provided. The primary outcomes were fasting leptin concentrations and subjective satiety measured on a multidimensional and single-dimension visual analogue scales. In Study 3 ad libitum intake was also assessed at a post-intervention buffet breakfast. Two-way repeated measures analysis of variance was used to analyse the effect of the intervention diets over time on the outcome measures. ii

6 Results: There was no between group difference in fasting leptin concentrations when realistic high carbohydrate and high fat controlled diets were consumed in Study 1 and Study 2 (p>0.4). This finding was similar when leptin concentrations were adjusted for body composition. Within subjects, changes in recent dietary carbohydrate and fat intake predicted a decrease in leptin concentrations during the intervention, but effects were small. A weak linear relationship between leptin concentrations and subjective satiety score was detected in Study 1 (p=0.06), but no relationship was detected in Study 2 and Study 3 (p>0.7). In Study 3 the more extreme high fat diet reduced leptin concentrations by 21% relative to the high carbohydrate diet (time*diet interaction, p<0.01). There was no influence of this difference in leptin concentration on ad libitum energy or macronutrient intake at the buffet breakfast (95% CI for difference in energy intake -411kJ to 190kJ). Conclusions: Circulating leptin concentrations are influenced by dietary carbohydrate to fat ratio such that decreasing the carbohydrate content of the diet results in decreased circulating leptin concentrations. However, the potential clinical relevance of this finding to the management of overweight and obesity is likely to be limited as i) extreme dietary patterns are necessary to detect this effect, ii) leptin concentrations were not related to subjective satiety score and iii) differences in leptin concentrations did not influence ad libitum food intake. Further research is necessary to confirm these findings over longer time frames, with different subject groups, twenty four hour blood sampling and ad libitum intake over the entire day. iii

7 Acknowledgements I would like to thank many people for their significant contribution to this thesis. Firstly, I extend my sincerest gratitude to my academic supervisors and mentors Associate Professor Arthur Jenkins and Professor Linda Tapsell. Thank you for being a constant source of wisdom, guidance and support. I am indebted to you for giving me your advice at all possible and impossible times. Without your trust in my abilities and support in the concurrent completion of my MSc and PhD, this thesis would not have been possible. To my colleagues and friends in the Department of Biomedical Science, thank you for providing a challenging yet welcoming environment for learning. My warmest thanks go to my fellow PhD students for sharing many moments in the office, laboratory, bar and coffee cart. To Nicole Smede and Nola Hurt thank you for your assistance in making things happen. Thank you to all members of my research teams. To Dr Brin Grenyer for his expert advice about the psychological aspects of the trials and Sr Sheena McGhee for her support and calm demeanour, thank you. A special thanks to all the women who participated in the studies. I gratefully acknowledge the financial support of the NH&MRC for providing the grant for this research. I thank the University of Wollongong for my scholarship, and the Smart Foods Centre, Metabolic Research Centre and Department of Biomedical Science for their assistance with conference attendance. Thank you to my supervisors for giving me their support in attending eight national and two international conferences. To my friends who have heard about my thesis many more times than is healthy, thank you for being there and for believing in me. A special thanks go to Jacqui, Vince, Nigel, Janelle and Trace for their encouragement and help with editing. The last word of thanks must go to my dear family who have been with me every step of the way and without whom the opportunities that lay before me would not be possible. iv

8 List of Abbreviations ANCOVA Analysis of covariance ANOVA Analysis of variance BIA Bioelectrical impedance analysis BMI Body mass index CCK Cholecystokinin CHO Carbohydrate CNS Central nervous system DH Diet history DR Dietary restraint DXA Dual energy x-ray absorptiometry FR Food record %E Percentage of energy GI Glycemic index GLP-1 Glucagon like peptide-1 HC High carbohydrate HF High fat kj Kilojoule MUFA Monounsaturated fatty acid NPY Neuropeptide Y PUFA Polyunsaturated fatty acid SD Standard deviation SEM Standard error of the mean SFA Saturated fatty acid TFEQ Three factor eating questionnaire VAS Visual analogue scale VMH Ventromedial hypothalamus v

9 Thesis publications and conference abstracts Gordon MA, Tapsell LC, and Jenkins AB (2002). Normal variation in dietary intake influences circulating leptin in women. Int J Obes 2002; 26: S42. Gordon MA, Tapsell LC, and Jenkins AB. Comparison of physiological and appetite related variables in non-obese women classified as restrained and nonrestrained eaters Australian Health Med Res Congress, Melbourne, Jenkins AB, Samaras K, Gordon MA, Snieder H, Spector T, and Campbell LV. Lack of heritability of circulating leptin concentration in humans after adjustment for body size and adiposity using a physiological approach. Int J Obes 2001;25: Gordon MA, Tapsell LC, and Jenkins AB. Relationship between macronutrient intake and serum leptin: implications for the clinical management of obesity. Dietitians Assoc Australia National Congress, Adelaide, Gordon MA, Grenyer B, Tapsell LC, Storlien LH and Jenkins AB. Relationships between leptin and satiety in humans: influence of dietary macronutrient composition. European Association for the Study of Diabetes, Israel, Gordon MA, Tapsell LC, Daniells S, Storlien LH and Jenkins AB. Contribution of dietary methodology in an intervention study examining diet composition and appetite regulation, International Congress of Dietetics, Scotland, Gordon MA, Tapsell LC, Johns S, Daniells S, Storlien LH and Jenkins AB Classification of habitual carbohydrate intake in a clinical study of diet, leptin and appetite regulation. Dietitians Assoc Australia National Congress, Canberra, vi

10 Table of Contents Declaration..i Abstract.. ii Acknowledgements..iv List of Abbreviations.v Thesis publications and conference abstracts....vi Table of Contents vii List of Tables..xiv List of Figures.....xvi Chapter 1: Introduction 1.1 Scope of thesis Overweight and obesity Definition Prevalence Implications Aetiology Clinical management Dietary carbohydrate to fat ratio Recommended intakes Dietary carbohydrate Dietary fat Appetite regulation Terminology Psychological appetite regulation Physiological appetite regulation Macronutrients and satiety Macronutrient preferences...14 vii

11 1.5 Circulating leptin Physiological role Role in human appetite Physiological regulation Macronutrient regulation Leptin administration Leptin resistance Summary Aims and Hypotheses Potential clinical relevance...31 Chapter 2: Methodology 2.1 Studying the diet and disease relationship Study population Subject compliance Blinding to intervention Intervention efficacy Dietary intake assessment Diet history interview Food records Dietary underreporting Body composition assessment Dual energy x-ray absorptiometry Bioelectric impedance analysis Satiety assessment Subjective satiety scales Ad libitum food intake Circulating leptin concentrations Blood collection Body composition adjustment...44 viii

12 2.6 Dietary restraint score Three factor eating questionnaire...45 Chapter 3: Human feeding study Rationale Aims & hypotheses Methods Subjects Experimental protocol Menstrual cycle data Anthropometric data Body composition data Dietary intake data Intervention diet Subjective satiety evaluation Biochemical analysis Statistical analysis Results Subject characteristics Menstrual cycle data Diet history intake Food record dietary intake Diet history and food record differences Intervention dietary intake Intervention and 24hr food record differences Leptin and body fat relationship Serum leptin concentration Serum leptin adjusted for body fat Dietary predictors of adjusted leptin Serum insulin concentration...64 ix

13 Serum leptin and insulin relationship Serum glucose concentration Subjective satiety score Serum leptin and satiety relationship Discussion Conclusions...76 Chapter 4: Human feeding study Rationale Aims & hypotheses Methods Subjects Experimental protocol Dietary restraint Anthropometric data Body composition data Dietary intake data Intervention diet Subjective satiety evaluation Biochemical analysis Statistical analysis Results Subject characteristics Menstrual cycle data Diet history intake Dietary restraint Food record intake Diet history and food record differences Intervention dietary intake Intervention and food record differences...89 x

14 4.4.9 Leptin and body fat relationship Serum leptin concentrations Leptin adjusted for body fat Dietary predictors of adjusted leptin Serum insulin concentrations Serum leptin and insulin relationship Serum glucose concentration Subjective satiety scores, multi-dimensional scale Subjective satiety scores, single dimension scale Dietary restraint, leptin and subjective satiety Serum leptin and satiety, multi-dimensional scale Serum leptin and satiety, single dimension scales Discussion Conclusions Chapter 5: Human feeding study Rationale Aims & hypotheses Methods Subjects Experimental protocol Anthropometric data Bioelectric impedance analysis Dietary intake data Intervention diet Sensory evaluation Buffet breakfast Subjective satiety evaluation Biochemical analysis Statistical analysis xi

15 5.4 Results Subject characteristics Menstrual cycle data Diet history intake Dietary restraint Food record intake Intervention dietary intake Intervention and food record differences Sensory evaluation Ad libitum breakfast intake Debriefing questionnaire Leptin and body fat relationship Serum leptin concentrations Dietary predictors of leptin Serum insulin concentrations Serum glucose concentrations Subjective satiety score, multi-dimensional scale Subjective satiety score, single dimension scales Subjective satiety and ad-libitum intake Dietary restraint, leptin and subjective satiety Serum leptin and satiety, multi-dimensional scale Serum leptin and satiety, single dimension scales Discussion Conclusions Chapter 6: Summary and conclusions 6.1 Potential clinical relevance of research findings Realistic dietary carbohydrate to fat ratio and circulating leptin Extreme dietary carbohydrate to fat ratio and circulating leptin Mechanisms of macronutrient effects on circulating leptin xii

16 6.1.4 Circulating leptin concentrations and subjective satiety Circulating leptin concentrations and ad libitum intake Limitations of research Conclusions Future research directions Chapter 7: References 7.1 Cited references 154 Appendix 1: Human feeding study 1 Appendix 2: Human feeding study 2 Appendix 3: Human feeding study 3 xiii

17 List of Tables 3.1 Study 1: Intervention diet composition Study 1: Intervention meal plan Study 1: Subject characteristics Study 1: Menstrual cycle data Study 1: Habitual dietary intake Study 1: Day 0 dietary intake Study 1: Day 0 and habitual intake difference Study 1: Intervention dietary intake Study 1: Intervention and Day 0 intake differences Study 1: Post-hoc dietary predictors of serum leptin Study 2: Intervention diet composition Study 2: Intervention drink composition Study 2: Subject characteristics Study 2: Habitual dietary intake Study 2: Restrained and unrestrained eaters characteristics Study 2: Restrained and unrestrained eaters habitual diet Study 2: Day 0 dietary intake Study 2: Day 0 and habitual intake difference Study 2: Part 1 and Part 2 dietary intake Study 2: Intervention dietary intake Study 2: Intervention and Day 0 intake difference Study 2: Post-hoc dietary predictors of serum leptin Study 2: Leptin and satiety, restrained and unrestrained eaters Study 3: Intervention diet composition Study 3: Intervention meal plan Study 3: Sensory properties of milkshake formulas Study 3: Buffet breakfast foods Study 3: Subject characteristics Study 3: Habitual dietary intake.121 xiv

18 5.7 Study 3: Restrained and unrestrained eaters characteristics Study 3: Restrained and unrestrained eaters habitual diet Study 3: Day 0 dietary intake Study 3: Intervention dietary intake Study 3: Sensory evaluation of milkshake formulas Study 3: Ad libitum breakfast intake Study 3: Post-buffet satiety ratings Study 3: Post-hoc dietary predictors of serum leptin Study 3: Relationship between satiety and dietary intake Study 3: Leptin and satiety, restrained and unrestrained eaters..134 xv

19 List of Figures 1.1 Hypothesised relationship between dietary carbohydrate, circulating leptin and satiety Harris-Benedict equation Study 1: Experimental timeline Study 1: Serum leptin and body fat relationship Study 1: Serum leptin concentrations over time Study 1: Serum leptin adjusted for body fat over time Study 1: Serum insulin concentrations over time Study 1: Serum insulin and leptin relationship Study 1: Serum glucose concentrations over time Study 1: Satiety score over time Study 1: Serum leptin and satiety relationship Study 2: Experimental timeline Study 2: Serum leptin and body fat relationship Study 2: Serum leptin concentrations over time, parts Study 2: Serum leptin concentrations over time, diets Study 2: Serum leptin adjusted for body fat over time Study 2: Serum insulin concentrations over time Study 2: Serum insulin and leptin relationship Study 2: Serum glucose concentrations over time Study 2: Satiety score over time Study 2: Hunger score over time Study 2: Fullness score over time Study 2: Desire to eat score over time Study 2: Prospective consumption score over time Study 2: Average satiety score over time, parts Study 2: Average satiety score over time, diets..100 xvi

20 4.16 Study 2: Serum leptin and satiety relationship, multi-dimensional satiety scale Study 2: Serum leptin and satiety relationship, single dimension satiety scale Study 3: Experimental timeline Study 3: Serum leptin and body fat relationship Study 3: Serum leptin concentrations over time, diets Study 3: Serum insulin concentrations over time Study 3: Serum glucose concentrations over time Study 3: Average satiety score over time, multi-dimension Study 3: Average satiety score over time, single dimension Study 3: Serum leptin and satiety relationship, multi-dimensional satiety scale Study 3: Serum leptin and satiety relationship, single dimension satiety scale xvii

21 CHAPTER 1: INTRODUCTION 1.1 Scope of thesis This research is concerned with determining the potential relevance of macronutrient induced changes in circulating leptin concentrations to the clinical management of overweight and obesity. To answer this research question nutritional, physiological and psychological science domains were explored in three controlled human feeding studies. The findings indicate that leptin concentrations are influenced by dietary carbohydrate to fat ratio but that this effect is unlikely to play a major role in the clinical management of overweight and obesity. 1.2 Overweight and obesity Definition Obesity defined simply, is an excessive accumulation of body fat in proportion to body size. Overweight and obesity can also be expressed in terms of body mass index (BMI), the ratio an individual s weight (kg) to height (m 2 ). A normal weight BMI is defined as kg/m 2, overweight as kg/m 2 and obese as >30 kg/m 2 (316). Despite its common use in the clinical setting a major limitation of BMI is that it does not actually measure any aspect of adiposity such as total body fat mass or body fat distribution. This is important to consider as abdominal obesity carries a higher health risk than body fat deposited in peripheral areas (139) Prevalence Overweight and obesity have reached epidemic proportions (315). Obesity is the most prevalent nutritional disorder of the Western World with increasing trends in Chapter 1: Introduction Page 1

22 both children and adults. In Australia, the incidence is increasing at almost 1% per year with 67% of men and 52% of women overweight or obese in 2000 (79). The United States of America (US) have even more alarming prevalence rates with 67% of men and 62% of women overweight in 2000 (86). Further, the problem is not confined to the developed world with about 115 million people suffering from obesity-related problems in developing countries (316) Implications Economic Overweight and obesity threaten to financially overwhelm the health system of many nations. The National Health and Medical Research Council of Australia in estimated the financial cost of obesity to be $840 million per year of which 63% was directly borne by the health system (215). In the US direct costs have been estimated at 7%, and in Europe 1-5% of total health costs (314). These figures do not take into account the additional billions spent by consumers on weight control programs, diet clubs and special foods each year. Medical Obesity is a medical disease in its own right and is a key risk factor for at least forty five other diseases (149). It is a core component of the metabolic syndrome and is highly predictive of endpoint diseases such as coronary artery disease, stroke and type 2 diabetes mellitus (165, 208). Psychological Obesity is a highly stigmatized condition. Obese people are commonly subjected to prejudice and discrimination, suffering a lifetime of bullying and social castigation (250). Overweight individuals are likely to share common prejudices about themselves as being undisciplined and self-indulgent (298). High rates of depression and low self-esteem in this group are commonly reported (236). Chapter 1: Introduction Page 2

23 1.2.4 Aetiology The increasing prevalence and severe implications of obesity highlight the need to better understand the cause and clinical management of this condition. The cause is undoubtedly multifactorial in origin. Fundamentally, body fat accumulation is the result of an imbalance between energy intake and energy expenditure. The body is able to regulate energy balance to some extent, demonstrated by some individuals having a stable weight over multiple years (147). It is generally accepted that body weight is the result of interaction between an individual s genetic predisposition to weight gain and environmental influences. There is growing agreement that environmental influences are driving the current epidemic (129). Endogenous influences Endogenous influences on weight gain include genetic, neurochemical and hormonal factors, age and gender (200). Considerable research points to a role for circulating leptin in body weight regulation (see Section 1.5). The genetic regulation of body weight has been clarified in studies of twins and adopted children. Genetic susceptibility accounts for 20-60% of the variance in body weight (190). Few cases of obesity have a single gene origin and over three hundred genes, markers or chromosomal regions have been linked to the obese phenotype (55). Preference for high fat foods may also have a genetic component (234). Change in human genetics however can not explain the dramatic rise in obesity prevalence over recent years. Environmental influences The modern obesogenic environment creates positive energy balance by affecting both energy intake and energy expenditure. Some environmental influences include cultural, socioeconomic and demographic (214, 258), increased food variety (233, 274), and increased portion sizes (323). With Chapter 1: Introduction Page 3

24 regards to energy intake, high fat diets may induce hyperphagia by increasing palatability, decreasing volume or weight of food, increasing energy density and decreasing satiety (39, 239, 271). In support of this, an inverse relationship between dietary carbohydrate and body weight has been reported (14) and may be related to decreased energy density, increased volume, increased chewing time and satiety (31, 82, 276). With regards to energy expenditure, increasingly sedentary modern lifestyles can be attributed to technological advances and physical inactivity. Physical inactivity ranks second only to smoking in importance for the burden of disease and disability in Australia (10) Clinical management While research efforts have focused on the management of obesity it should be emphasised that obesity is preventable. Prevention of chronic disease is possible through lifestyle intervention as shown in large scale studies (160, 287). Prevention of weight gain is a good outcome of dietary intervention as body weight can rise 1-2kg per year in adults living modern day lifestyles (170). Research investigating interventions in the pre-obese state can address this. Any intervention that induces negative energy balance will result in weight loss. Numerous dietary strategies exist for achieving this state but calorie reduction is the key. Dietary intervention which creates a deficit of 2500kJ per day should achieve a weight loss of 0.5-1kg per week (67). A body weight reduction of 5-10% results in significant health benefits (215). Behavioural, pharmacological and surgical treatments can be considered on an individual basis. Dietary macronutrient composition may be important in weight management. There is a shortage of well-designed randomised clinical trials that assess the impact of diets differing in macronutrient composition on weight loss and maintenance of weight loss (128). Evidence has accumulated for the use of high carbohydrate low fat eating plans in weight management (12, 13, 15, 43, 286). Chapter 1: Introduction Page 4

25 This eating style induced a modest weight loss of ~3kg in some (223, 249) but not in all studies (96, 97, 171). The type of carbohydrate consumed (simple vs. complex) did not affect weight loss in one large multi-centre trial (249). Other approaches have been shown to be successful including a higher protein diet (83, 262, 311) a modified fat diet (103, 302, 325) and a low glycemic index diet (185). Although long-term data is required the success of multiple dietary patterns suggests the need for individualised weight management strategies (42). There is a central role for the nutrition professional in the process of individualisation. 1.3 Dietary carbohydrate to fat ratio The influence of macronutrient intake on the current epidemic of obesity is of concern to both scientists and lay persons. The proliferation of popular diet books on this topic in recent times is nothing short of phenomenal (for review see 89). Effects of different macronutrients in research may be separated but in reality they co-exist in food. Consumption of macronutrients alone is unlikely in the human diet. The focus of the current research is on dietary carbohydrate to fat ratio. These macronutrients are reciprocally related such that when the quantity of one is decreased in the diet the other usually increases. Together dietary fat and carbohydrate make up percent of dietary energy (%E) (276). Dietary patterns are often referred to as high carbohydrate or high fat but definitions of these vary and in most instances are simply relative comparisons Recommended intakes Dietary carbohydrate Population recommendations for carbohydrate are that an optimum diet for adults contains at least 55%E from a variety of food sources (82, 314). The minimum amount of carbohydrate needed to sustain the nervous system may be as low as Chapter 1: Introduction Page 5

26 fifty grams per day and a maximum of 75%E carbohydrate should not be exceeded (188). Australian s consume 46%E as carbohydrate according to the National Nutrition Survey (1). This is not different from a typical Western diet which contains approximately 45%E as carbohydrate (94). The major foods contributing to this intake are bread, breakfast cereal, rice, pasta, biscuits and pastry. Nutrition professionals advise carbohydrates intake to come from a variety of vegetables, fruits and whole grains. Dietary fat Reducing dietary fat consumption has been the primary focus of obesity prevention and management for years. Dietary fat intake of less than 30%E is recommended (314). Emerging evidence suggests that dietary fat type is also of importance to health. Dietary targets include reducing saturated fat to less than 10%E and increasing polyunsaturated fat intake to 6-8%E (314). The minimum dietary fat intake should be 15%E (146). Australian s consume 33%E as total fat with 13%E as saturated fat, 12%E as monounsaturated fat and 5%E as polyunsaturated fat (1). A typical Western diet contains 37%E as fat with the major contributors being dairy products, meat, poultry, biscuits and pastry (1). Dietary protein Dietary protein provides the least variable portion of energy intake. A minimum intake of 10%E or 0.8 grams per kilogram of body weight is necessary for human growth and repair (314). Australian s consume 16%E as protein and from dietary surveys, protein intake is approximately 15%E in Western populations (16, 187) Dietary carbohydrate Types of carbohydrate Carbohydrates include a wide range of compounds and can be classified according to degree of polymerisation into simple sugars, oligosaccharides and polysaccharides. Polysaccharides are commonly referred to as complex Chapter 1: Introduction Page 6

27 carbohydrate or starch. The dominant types of carbohydrate in the human diet are starch and sucrose (82). Conventional wisdom held starch to be nutritionally superior to other carbohydrate types in obesity management but at present little is known about which types of carbohydrate best protects against excess energy intake. Shorter-chain carbohydrates may have more of an intake promoting effect and longer-chain carbohydrates have an intake restraining effect (276). Type of starch in foods such as the amylose to amylopectin ratio may influence energy intake (25, 264). Fibre consumed at one meal has been shown to decrease hunger and energy intake at the next meal but effects are relatively modest (39). Soluble fibre in particular slows food digestion and hence nutrient absorption resulting in decreased postprandial glucose and insulin concentrations which may be partly responsible for the intake restraining effect of carbohydrate containing foods (82). Glycemic index of foods A newer method of classifying carbohydrate containing foods is according to its glycemic index. Glycemic index is defined as the incremental area under the blood glucose response curve to a fifty gram carbohydrate portion of test food expressed as a percentage of the response to the same amount of carbohydrate from a standard food taken by the same subject (82). The glycemic index of foods is affected by the nature of the carbohydrate, cooking method and the presence of other nutrients within a food. Glycemic load, which is the glycemic index of a food multiplied by the amount of carbohydrate consumed, has been positively associated with risk of coronary heart disease in large prospective cohort investigations (178, 179). Intervention studies have established the relevance of consuming low glycemic index carbohydrates in individuals with type 2 diabetes (140). Relevance of the glycemic index for other populations is still being discovered (2, 184). Chapter 1: Introduction Page 7

28 Metabolic effects of carbohydrate Carbohydrate is required as an immediate substrate source for energy metabolism particularly for the central nervous system. Mechanisms exist which keep plasma glucose within a narrow range. A 6-12% reduction in plasma glucose has been shown to initiate feeding in rodents and in humans (277). During food consumption and in the early post meal interval insulin is released to stabilise plasma glucose levels. Carbohydrate oxidation is closely matched to carbohydrate intake even under conditions of overfeeding (136). There is some evidence that carbohydrate oxidation is negatively related to subsequent energy intake (117) and that subjects eat to maintain carbohydrate balance (265). It seems that this relationship is not a powerful negative feedback (276). Under normal Western dietary conditions net de novo lipogenesis is an insignificant metabolic pathway (117). Diets with substantial amounts of carbohydrate suppress fat oxidation and foster fat storage making it possible to gain weight on a high carbohydrate diet (276). Carbohydrate is required to replenish glycogen stores for use between meals. Glycogen stores are limited in most adults to about five hundred grams. Hypotheses about the role of glycogen stores in energy balance have been proposed (226, 252). Over-consumption of carbohydrate Elevated plasma triglycerides and depressed high density lipoprotein have been associated with high carbohydrate diets (221). This effect has been demonstrated when carbohydrate was compared with monounsaturated fatty acids under iso-energetic conditions (103) and may not occur during negative energy balance (221). This plasma lipid response seems to be highly variable and dependent on carbohydrate source (320). Chapter 1: Introduction Page 8

29 1.3.3 Dietary fat Types of fat Dietary fatty acids can be grouped into saturated, monounsaturated and polyunsaturated according to the number of double bonds they contain. Fat has an important role in a range of metabolic and physiological processes including the delivery of essential fatty acids and fat soluble vitamins, cell membrane activity and gene regulation (269). In the dietary management of obesity focus has shifted from only total fat intake to also consider fatty acid subtypes. Saturated fat in particular has been associated with obesity through decreased mobilization and oxidation, reduced stimulation of the sympathetic nervous system and the creation of rigid cell membranes with reduced binding capacities (268). Polyunsaturated fatty acids may have opposite effects and monounsaturated fatty acids may have a neutral role (269). Metabolic effects of fat Dietary fat is the most energy dense metabolic fuel. Capacity to store fat in adipose tissue with the human body is essentially unlimited. Consumption of fat evokes limited oxidative response which occurs three to seven days following high fat feeding (254). Oxidation of saturated fats have been shown to be inversely related to carbon length while long-chain polyunsaturated fatty acids are readily mobilised and oxidised (71). Diet induced thermogenesis is lower following a high fat diet than other dietary patterns (312). Type of fat consumed have also been shown to influence insulin action in rodents and humans, with saturated fats associated with deleterious effects and long-chain polyunsaturated fats having a protective effect (268, 293). Chapter 1: Introduction Page 9

30 1.4 Appetite regulation Understanding the regulation of appetite may be important in reversing the accelerating trends in overweight and obesity. Eating behaviour bridges the gap between the nutritional environment and the biological mechanisms of body weight regulation (127). Food ingestion is a complex interaction of psychology, peripheral physiology and central physiology. Satiety is thus not based on a single event but rather multiple processes acting at different times with varying in intensity and duration (162). A model describing the time course events is the satiety cascade which includes sensory, cognitive, pre-absorptive and postabsorptive events (32) Terminology Hunger, satiety and appetite are terms used to describe aspects of eating behaviour in the scientific literature. There is no universally accepted definition of these terms and for the purpose of this research the following have been adopted. Hunger is the primary motivation to acquire and ingest food (310). Hunger has both cognitive and physical components but is not a necessary condition for eating to occur (34). Appetite is a disposition or orientation towards eating a food (310). Appetite controls the type and amount of food eaten and can be food specific (206). The term satiety may encompass the many constructs associated with eating including low hunger, high fullness, low desire to eat and low appetite. For methodology associated with the measurement of these appetite related constructs see Section Psychological appetite regulation Food choice and eating habits are influenced by sensory and cognitive factors. Sensory properties of food include palatability, smell, sight, taste, temperature Chapter 1: Introduction Page 10

31 and texture (77). Cognitive control of food intake includes the psychological construct of dietary restraint. Dietary restraint is practised in both obese and non-obese individuals and influences the reporting of dietary intake. Lower fat, higher protein intakes and lower energy intakes have been described by restrained eaters compared to unrestrained eaters (19, 288). A negative association between circulating leptin concentrations and restraint score has been reported in underweight subjects (295), obese binge eating women (3, 70) and obese non-binge eating women (3). The cognitive control of food intake is not well understood but should be considered in studies investigating appetite regulation as eating behaviour is not determined by physiological drive alone. For dietary restraint assessment methods see Section Physiological appetite regulation The physiological regulation of appetite is highly complex and includes a central feeding drive which interacts with peripheral signals. Peripheral signals are derived from adipose tissue as well as the endocrine, neurological and gastrointestinal systems. Physiological responses involved in the termination and inhibition of eating can be referred to as satiety signals. While some satiety signals are well understood, scientific knowledge is still accumulating in this field. Central satiety signals The historical dual-centre concept of the lateral hypothalamus as a feeding centre and ventromedial hypothalamus as a satiety centre emerged from early animal studies. Recently the arcuate and paraventricular nucleus of the hypothalamus and hindbrain centres have been shown to be involved in appetite responses (294). Neurons in the arcuate nucleus express receptors for neuropeptide Y, insulin and leptin and local blood-brain barrier permeability allows passage of these molecules (131). An extensive review of central satiety Chapter 1: Introduction Page 11

32 signals is beyond the scope of this thesis and has been published elsewhere (255, 294). Neuropeptide Y is the most abundant neuropeptide in the brain. Central administration of neuropeptide Y increases energy intake, decreases energy expenditure, increases lipogenesis and reduces sympathetic nervous system activity (255). Neuropeptide Y is elevated under conditions of negative energy balance and depressed on diets rich in saturated fat (303). Circulating leptin has been shown to inhibit the release and synthesis of neuropeptide Y (24). This interaction may be partly responsible for the physiological effects of leptin. Peripheral satiety signals A litany of candidates from the periphery may be involved in feeding behaviour. Comprehensive reviews of knowledge in this area have been published (41, 109, 256, 322). The emerging picture is that leptin may interact with numerous satiety signals in appetite regulation. Cholecystokinin (CCK) is the most researched peripheral satiety factor. This hormone is synthesised within the duodenal mucosa and released into the portal circulation in response to the presence of nutrients (particularly fat) in the gastrointestinal tract (47). CCK reduces meal size by inhibiting gastric emptying (131). CCK and leptin have been shown to act synergistically to control eating behaviour in rodents ( ). Glucagon-like peptide-1 (GLP-1) is secreted from the intestinal mucosa after mixed meal intake (87). GLP-1 delays gastric emptying and stimulates insulin release (33, 87). There is some evidence that circulating leptin stimulates GLP-1 release (8). Peptide YY 3-36 is released post prandially in proportion to the calorie content of a meal and reduces food intake in humans (21, 22). Ghrelin increases appetite and stimulates food intake in rats and humans when administered peripherally in doses that produce physiological increments in plasma leptin (308). Insulin is secreted in response to meals from pancreatic β-cells. Both fasting and postprandial insulin levels reflect body fat stores. For evidence illustrating a synergism between circulating insulin and leptin see Section Chapter 1: Introduction Page 12

33 Characteristics of the food consumed may be a satiety signal. Food volume, energy density and macronutrient proportions (see Section 1.4.4) have all been investigated. These factors tend to co-vary in diets but numerous experiments have attempted to determine their independent effects. Volume of food consumed independently influences satiety ratings in lean men (241, 242). Energy density of a meal affects both subjective satiety and spontaneous food intake (192, 225, 240, 273, 275). Absorbed nutrients and body reserves may be a satiety signal. Mayer s classic glucostatic concept states that the control of appetite is determined by the amount of available carbohydrate (205). Kennedy s lipostatic model links food intake to the state of adipose stores (155). Flatt proposed the glycogenostatic hypothesis suggesting that glycogen stores mediate between food intake and body weight control (85). High carbohydrate diets were hypothesised to generate more potent signals for controlling intake due to their limited storage capacity. From the available evidence large deviations in glycogen stores do not elicit measurable influences on human appetite (226) Macronutrients and satiety Whether the macronutrient composition of the diet can influence satiety is a key question in appetite research. It is generally accepted that a macronutrient hierarchy exists such that per kilojoule dietary protein is more satiating than dietary carbohydrate which is more satiating than dietary fat (271). Studies report finding evidence (37, 135, 312) and no evidence of this hierarchy (52, 56, 228, 296). Confounding influences of volume, energy density and palatability may influence outcomes. Ingestion of dietary carbohydrate may have a greater effect on short-term satiety than ingestion of dietary fat (14, 276). The mechanisms by which carbohydrate influences satiety are not fully understood. Decreased energy density is one Chapter 1: Introduction Page 13

34 potential mechanism (157). Endogenous glucose and insulin may also increase satiety (95, 114). Dietary fibre may further improve the satiating effect of carbohydrate-rich foods (36). Carbohydrate subtypes do not seem to affect appetite in the same manner (31). High fat foods may be passively over consumed due to decreased satiety. Limited research has investigated the effect of fat subtypes on satiety in humans but do not support different effects (6). High fat diet hyperphagia is expressed in rodents via increased meal size and decreased inter-meal interval (305). Dietary protein appears to be the macronutrient that induces satiety to the greatest extent (38). Mechanisms involved are unknown and different protein types do not seem to influence satiety uniquely (167). Habitual dietary protein intake can influence appetite response to protein consumption in humans (182) Macronutrient preferences Individuals display preferences for foods which vary in taste and composition together with a host of other attributes (38). Habitual consumption of a particular macronutrient leads to an adaption of physiological mechanisms activated by that nutrient. For example, habitual fat intake is associated with increased secretion of pancreatic lipase and habitual carbohydrate intake associated with increased amylase secretion (306). As described below, rodent and human research suggests that habitual macronutrient preference influences satiety. This may be important to consider in obesity management where individual diet regimes are prescribed. Rodent strains can differ in both their macronutrient preference (263) and susceptibility to high fat diet induced obesity (100). High fat diet consumption has been shown to enhance the dietary fat preference in rats (235). After three weeks rats fed a high fat diet displayed reduced sensitivity to the satiety effects Chapter 1: Introduction Page 14

35 of intestinal oleate compared to rats fed a low fat diet (66). Feeding a high fat diet may also reduce sensitivity to the satiety effects of endogenous CCK (65). The notion of high fat and low fat phenotypes has been described in human studies of young lean men who are indistinguishable in terms of body mass index, body fat, age and general lifestyle. Appetite related characteristics of these phenotypes have been investigated. In response to fixed energy load, habitual high fat consumers ate a constant weight of food while habitual low fat consumers ate a constant energy level (63). High fat consumers had higher baseline hunger scores and a larger decline in hunger in response to food ingestion compared to low fat consumers. The high fat phenotype had a higher resting metabolic rate and fat oxidation (64), glucose and circulating leptin (62) than low fat consumers despite similar body fat levels. This research indicates that habitual intake may entrain different appetite control systems. This has implications for the design of studies investigating appetite control by highlighting the importance of assessing and controlling for background habitual diet. 1.5 Circulating leptin Over the past decade a new dimension in obesity research has developed. The landmark of this research was the identification of the gene responsible for obesity in the most intensively studied rodent model of obesity the ob/ob mouse (326). The ob gene was found to encode the plasma protein leptin (from the Greek word leptos meaning thin) in both mice and humans. Since this discovery a large and rapidly changing body of knowledge has been created. Circulating leptin is largely secreted by subcutaneous adipose tissue. Adipocyte size is an important determinant of leptin synthesis, the larger the adipocyte the more leptin it contains (91). Leptin has been detected in low concentrations in the heart, placenta, stomach, small intestine, skeletal muscle, mammary Chapter 1: Introduction Page 15

36 epithelium and ovaries (194). Leptin concentrations in the cerebrospinal fluid increase with body fat levels but are always lower than peripheral concentrations (51). The ratio of cerebrospinal to peripheral leptin is lower in obese subjects. Six variants of the leptin receptor have been identified (324). The long isoform or Ob-Rb is the only receptor with clearly demonstrated signalling capability (284). Leptin exerts its primary effect by acting on receptors in the hypothalamus in particular the arcuate nucleus. Leptin is transported to the brain through a unidirectional saturable system. It is possible that in obese individuals the blood brain transporter is near the top of the concentration response curve which may explain low cerebrospinal fluid concentrations (51). Leptin receptor isoforms are widely distributed throughout peripheral organs including the lymph, liver, uterus, adrenal gland, kidney, adipose tissue, ovary, skeletal muscle and gastrointestinal tract (17, 110). The Ob-Rb is expressed in pancreatic β-cells which may mediate the relationship between circulating leptin and insulin peripherally (156). Leptin circulates in the blood stream attached to binding proteins. Most leptin circulates in the bound form in lean individuals and free form in obese individuals (261). Soluble leptin receptor (sob-r) is the major leptin binding protein in the blood in humans (166). Soluble leptin receptor levels are inversely correlated with adiposity suggesting increased effectiveness in lean individuals (219). Compared with normal weight subjects sob-r levels are lower in obese subjects by approximately fifty percent (259) Physiological role Early rodent studies contributed much to the current knowledge about the physiological role of leptin. The ob/ob mouse deficient in leptin exhibits severe Chapter 1: Introduction Page 16

37 early onset obesity, hypometabolism, hyperphagia, extreme insulin resistance, infertility and hypothyroidism. Daily administration of recombinant leptin systemically and intracerebroventricularly corrects these metabolic parameters by reducing energy intake and increasing energy expenditure (49, 104, 222). The result is a selective reduction in adiposity hence weight reduction. Other genetic rodent models of obesity inhibit the action of leptin by a mutation in the leptin receptor (57). These observations led to the hypothesis that leptin is a satiety signal from the periphery to the central nervous system that is integral to the control of food intake and metabolism. The role of leptin in the pathogenesis of human obesity is still unclear. Mutations in the ob gene are rare. Only a handful of people worldwide have been identified for which congenital deficiency in leptin production or receptor result in morbid obesity (60, 84, 209, 270). Instead human obesity is characterised by elevated peripheral leptin concentrations and increased adipose tissue expression of leptin mrna (51). Plasma kinetics show this to be due to an increase in whole body leptin production not a defect in leptin clearance (158). Leptin has a diverse role in normal human physiology. In steady state leptin concentrations reflect adipose tissue mass and when there is an energy imbalance leptin acts as a sensor (108, 173). Leptin may not be involved in obesity prevention nor inhibit food overconsumption. Rather leptin may signal when body energy stores are depleting to guard against starvation (130). This is vital for activities such as reproduction when food is scarce. Leptin has been shown to be involved in reproduction, hematopoiesis, angiogenesis, immune function, blood pressure control and bone formulation (90). Leptin has also been shown to a modulator of sweet taste (153, 216) and related to palatability (232). Chapter 1: Introduction Page 17