POSTPRANDIAL METABOLIC RESPONSES TO
MACRONUTRIENT IN HEALTHY, HYPERINSULINEMIC AND
TYPE 2 DIABETIC SUBJECTS
by
Xiaomiao Lan-Pidhainy
A thesis submitted in conformity with requirements
for the degree of Doctor of Philosophy,
Graduate Department of Nutritional Sciences, Faculty of Medicine,
University of Toronto
© Copyright by Xiaomiao Lan-Pidhainy (2011)
ii
POSTPRANDIAL METABOLIC RESPONSES TO
MACRONUTRIENT IN HEALTHY, HYPERINSULINEMIC AND
TYPE 2 DIABETIC SUBJECTS
Xiaomiao Lan-Pidhainy
Doctor of Philosophy
Graduate Department of Nutritional Sciences, Faculty of Medicine
University of Toronto
2011
Abstract
The literature comparing macronutrient metabolism in healthy and diabetic subjects is
abundant; however, little data exists on how non-diabetic subjects with insulin resistance handle
macronutrient. We did two studies to investigate the postprandial responses to macronutrient in
healthy, hyperinsulinemic and type 2 diabetic (T2DM) subjects.
In the first study, twenty-five healthy, non-diabetic subjects [9 with fasting serum insulin
(FSI) <40pmol/L; 8 with 40 ≤ FSI < 70pmol/L; and 8 with FSI ≥ 70 pmol/L] were fed eleven test
meals (50g oral glucose with 0-30g doses of canola oil or whey protein) after an overnight fast.
There were no significant FSI × fat (p=0.19) or FSI × protein (p=0.08) interaction effects on
glucose response, suggesting that the effects of fat or protein on glycemia were independent of
FSI of the subjects. In addition, the changes in relative glucose response per gram of fat (r = -
0.05, p = 0.82) or protein (r = -0.08, p = 0.70) were not related to FSI of the subjects.
iii
In the second study, Healthy (FSI < 40pmol/L), Hyperinsulinemic (FSI ≥ 40pmol/L), and
T2DM were fed five foods with 50g available carbohydrate. Among the subject-groups, the
Glycemic Index (GI) values were not significantly different for each food, and the mean (±SEM)
GI values of all foods were not significantly different (p>0.05). However, the mean (±SEM)
Insulinemic Index of the foods was higher in T2DM (100±7, n=10) than those of Healthy (78±5,
n=9) and Hyperinsulinemic subjects (70±5, n=12) (p=0.05). The Insulinemic Index was
inversely associated with insulin sensitivity (r=-0.66, p<0.0001), positively related to fasting-
and postprandial-glucose (both r=0.68, p<0.0001) and hepatic insulin extraction (r=0.62,
p=0.0002).
The oral-glucose data were pooled from the two studies to investigate whether there was
any relationship between GLP-1 and insulin sensitivity, β-cell function and hepatic insulin
extraction. No significant correlation was observed (p>0.05).
The results suggest that the glucose-lowering effect of fat and protein is not affected by
insulin sensitivity. GI is independent of the metabolic status of the subjects; however, unlike GI,
Insulinemic Index is influenced by the metabolic status of the subjects, and thus may have
limited clinical utility.
Abstract Word Count: 349
iv
To My Baby, My Little Love
Taras (Dundun,敦敦)
献给妈妈的最爱
敦敦
v
ACKNOWLEDGEMENTS
As I draw near the completion of the Ph.D program, Qu Yuan‘s poem (ca. 340 BCE –
278 BCE) keeps resounding in my ears“the road ahead is long and rugged; I see no ending;
yet high and low I'll search with my will unbending (路曼曼其修远兮, 吾将上下而求索)‖. I
dare not compare the Ph.D journey to Qu Yuan‘s lofty pursuit of truth, beauty and ideal
political ambition, nor by no mean am I using the ancient poet to elevate the status of
successfully completing a Ph.D. At this moment of embarking on a new academic journey,
I could find no better way to describe the current state of my mind. Looking back at the
road travelled, I am in deep gratitude to so many people.
I would like to thank the many participants involved in the two clinical studies, the
nurses from the Risk Factor Modification Center at St.Michael‘s Hospital, the staff and
fellow students of Dr.Wolever‘s Lab, the staff at St.Michael‘s Core Laboratory and Banting
and Best Diabetes Core Laboratory. Special thanks also go to my three research assistants,
Michelle Liu, Jonathan Leung, and Cindy Huang. They not only helped me run the studies
but also offered joyful conversation and friendship.
I also would like to thank Dr. Deborah O‘Connor and Dr. Lindsay Robinson
(University of Guelph) whom gave insightful and critical appraisal of my thesis. My
committee members: Dr. Harvey Anderson, Dr. Anthony Hanley and Dr. Qinghua Wang
who gave me so many constructive suggestions and such good guidance on numerous
occasions.
My greatest gratitude goes to my supervisor Dr.Wolever, who has constantly offered
me support and guidance. His scientific ingenuity and passion for research I greatly admire.
vi
Family and friends have supported me in various ways. My mother-in-law and
sister-in-law helped to take care of my baby while I was in Toronto for various meetings.
My father-in-law encouraged me in my academic pursuit. The delightful friendship with
Derong, Suchang, Xuefei and Yixin give me much-needed escape from the perils of
graduate student life.
It is not possible to complete this thesis without the unconditional love of my parents
who are so proud of me. They have always been there for me even though I lived across the
ocean from them. My parents made a long hard trip from China to U.S to take care of me
and my new-born so I would be able to concentrate on my thesis.
I would not be what I am today without the devoted support of my beloved husband,
Ihor, a gentleman and scholar, without whom, there is nothing…
This thesis is dedicated to my baby boy Taras (Dundun), who I carried throughout
the laborious process of thesis writing, and has been the joy, hope and light of my life since
his birth.
vii
TABLE OF CONTENTS Page
Abstract…………………………………………………………………………………ii
Dedication………………………………………………………………………………iv
Acknowledgments…............................................................................................................v
Table of contents………………………………………………………………………..vii
List of tables….....................................................................................................................xi
List of figures…………………………………………………………………………...xiii
List of Appendices……………………………………………………………………...xv
Abbreviations …………………………………………….…………………………….xvii
Publications and presentations arising from dissertation……………………………….xix
1. INTRODUCTION…………………………………………………………………....1
2. LITERATURE REVIEW …………………………………………………………..8
2. 1 Effects of Macronutrient on Postprandial Glucose and Insulin Responses……9
2.1.1 Dietary Carbohydrate…………………………………………………….9
2.1.1.1 Glycemic Index…………………………………………………12
2.1.1.2 Insulinemic Index……………………………………………….17
2.1.2 Dietary Fat……………………………………………………………….18
2.1.3 Dietary Protein…………………………………………………………...19
2.2 Insulin Resistance and Compensatory Hyperinsulinemia………………………...21
2.2.1 Methods to measure insulin sensitivity/resistance………………………..23
2.3 Insulin Resistance/Hyperinsulinemia and the Risk of Chronic Diseases……26
2.3.1 Mechanisms of insulin resistance induced clinical syndromes……………..27
viii
2.4 Hepatic Insulin Extraction in the Regulation of Peripheral Insulin Concentration….....28
2.5 Effects of Macronutrient on Postprandial Responses in Hyperinsulinemic Subjects….33
2.5.1 Dietary Carbohydrate…………………………………………………………….33
2.5.2 Dietary Fat………………………………………………………………………..34
2.5.3 Dietary Protein……………………………………………………………………35
2.6 Mechanisms by Which Fat and Protein Modulating Postprandial Glycemia……………36
2.7 The Effects of Habitual Diet on Postprandial Responses………………………………...37
3. RATIONALE, HYPOTHESES, AND OBJECTIVES………………………………..38
3.1 Rationale…………………………………………………………………………….39
3.2 Overall Objective……………………………………………………………………39
3.3 Specific Hypotheses…………………………………………………………………40
4. THE HYPOGLYCEMIC EFFECT OF FAT AND PROTEIN IS NOT
ATTENUATED BY INSULIN RESISTANCE…………………………………………42
4.1 Abstract……………………………………………………………………………….43
4.2 Introduction…………………………………………………………………………...44
4.3 Subjects and Methods…………………………………………………………………45
4.3.1 Subjects and study design……………………………………………………...45
4.3.2 Test drinks……………………………………………………………………...47
4.3.3 Assessment of habitual dietary intake………………………………………….47
4.3.4 Assessment of physical activity………………………………………………..48
4.3.5 Blood analysis………………………………………………………………….48
4.3.6 Calculation and statistical analysis…………………………………………….49
4.4 Results…………………………………………………………………………………52
ix
4.5 Discussion and conclusions…………………………………………………………..67
5. ARE THE GLYCEMIC AND INSULINEMIC INDEX VALUES OF
CARBOHYDRATE FOODS SIMILAR IN HEALTHY CONTROL,
HYPERINSULINEMIC AND TYPE 2 DIABETIC PATIENTS?………………………71
5.1 Abstract…………………………………………………………………………………72
5.2 Introduction……………………………………………………………………………..73
5.3 Subjects and Methods…………………………………………………………………..74
5.3.1 Subjects and study design……………………………………………………….74
5.3.2 Test foods………………………………………………………………………..76
5.3.3 Blood analysis…………………………………………………………………...77
5.3.4 Calculations and statistical analysis……………………………………………..78
5.4 Results…………………………………………………………………………………..80
5.5 Discussion and conclusions…………………………………………………………….99
6. THE RELATIONSHIP BETWEEN PLASMA GLP-1 RESPONSE AND
INDIRECT MEASURES OF INSULIN SENSITIVITY, β-CELL FUNCTION AND
HEPATIC INSULIN EXTRACTION…………………………………………………...103
6.1 Abstract…………………………………………………………………………….....104
6.2 Introduction…………………………………………………………………………...105
6.3 Materials & Methods……………………………………………………………….....107
6.3.1 Subjects………………………………………………………………………...107
6.3.2 Study Design…………………………………………………………………...108
6.3.3 Blood analysis……………………………………………………………….....109
6.3.4 Calculations…………………………………………………………………….110
x
6.3.5 Statistical Analysis……………………………………………………………..111
6.4 Results………………………………………………………………………………….112
6.5 Discussion and Conclusion…………………………………………………………….126
7. GENERAL DISCUSSION ………………………………………………………………..131
7.1 General discussion……………………………………………………………………..132
7.2 Weakness………………………………………………………………………………136
7.3 Strengths……………………………………………………………………………….137
8. Future directions…………………………………………………………………………...139
9. Conclusions…………………………………………………………………………………142
10. References………………………………………………………………………………….145
11. Appendices…………………………………………………………………………………173
xi
LIST OF TABLES
Page
Table 2.1 Methods to measure insulin sensitivity/resistance……………………………..25
Table 4.1 Composition of the 9 test meals, their total energy content, and the
percentage of energy from carbohydrate, fat, and protein…………………….. 57
Table 4.2 Anthropometric and metabolic characteristics of subjects by FSI……………...58
Table 4.3 Mean areas under the curve (AUCs) for glucose, insulin, C-peptide,
and glucagon-like peptide 1 (GLP-1) responses after 3 separate
oral-glucose-tolerance tests and their inter- and intra-subject CVs
by fasting serum insulin (FSI)…………………………………………………..60
Table 4.4 P values for main and interaction effects of fasting serum insulin (FSI)
group, fat, and protein on glucose, insulin, C-peptide,
insulin secretion rate (ISR), and glucagon-like peptide 1 (GLP-1)
responses expressed as area under the curve (AUC) and hepatic
insulin extraction (HIE)….………………………………………………............61
Table 5.1 Nutrition composition of the test foods………………………………………….85
Table 5.2 Anthropometric and metabolic characteristics of the study groups……………...86
Table 5.3 The glucose, insulin, C-peptide, GLP-1 and insulin secretion rate (ISR)
expressed as area under the curve (AUC) and hepatic insulin extraction
(HIE) for each individual food and the mean of all carbohydrate foods
in the study groups……………………………………………………………….88
Table 5.4 Main and interaction effects of food, subject group on GI, II and
C-peptide index…………………………………………………………………...90
xii
Table 5.5 The GI, II and C-peptide index of different carbohydrate foods in
the study groups…………………………………………………………………..91
Table 6.1 Anthropometric and metabolic characteristics of the study groups………………115
Table 6.2 AUC of glucose, insulin, C-peptide, insulin secretion rate (ISR) and GLP-1 in the
study groups………………………………………………………………………..117
Table 6.3 Insulin sensitivity, β-cell function and HIE in the study groups…………………...118
xiii
LIST OF FIGURES
Page
Figure 1.1 Mechanisms regulating postprandial glycemia in healthy subjects....................6
Figure 1.2 The mechanisms regulating postprandial glycemia are disrupted in diabetes,
obesity and insulin resistance…………………………… ..................................7
Figure 2.1 Factors affecting peripheral insulin concentration…………………………......32
Figure 4.1 Mean (±SEM) 2-h postprandial plasma glucose, insulin, C-peptide,
and GLP-1 concentrations after 50 g oral glucose plus 0, 5,
and 30 g protein in nondiabetic humans with different
concentrations of FSI……………………..……………………………………62
Figure 4.2 Mean (±SEM) main effects of fat and protein on glucose, insulin,
C-peptide, ISR, and GLP-1 responses expressed as AUC and HIE
in nondiabetic humans by combined FSI groups……………………………….63
Figure 4.3 Mean (±SEM) effects of 50 g glucose plus 0, 5, or 30 g fat
or protein on glucose, insulin, C-peptide, ISR expressed as AUC,
HIE in healthy nondiabetic subjects with different levels of FSI……………….65
Figure 5.1 The 2-3hr postprandial glucose, insulin, C-peptide and GLP-1
responses (Mean±SEM) to different carbohydrate foods in Normal,
Hyper[I] and T2DM subjects……………………………………………………93
Figure 5.2 The linear correlation between GI and II and C-peptide index
respectively in Normal, Hyper[I] and T2DM……………………………………94
Figure 5.3 The linear correlation between GI or II or C-peptide index and
OGIS and HIE respectively in all subject-groups………………………………..95
xiv
Figure 5.4 The linear correlation between GI or II and fasting glucose and
mean postprandial glucose and glucose AUC respectively
in all subject-groups…………………………………………………………….97
Figure 5.5 The comparison of overall mean±SEM of GI, II and C-peptide index
for each food and all carbohydrate foods………………………………………98
Figure 6.1 The correlation between ISRbasal (adjusted for fasting glucose)
and postprandial insulin secretion rate (ISRauc)………………………………..119
Figure 6.2 Mean (±SEM) postprandial plasma GLP-1 response to oral glucose in type 2
diabetic patients and subjects with different levels of FSI……………………..120
Figure 6.3 The correlations between HIE and postprandial insulin response
and age respectively…………………………………………………………….121
Figure 6.4 The correlations between GLP-1 response after ingestion of
50g oral glucose and waist circumference and BMI respectively,
in type 2 diabetic patients and subjects with different levels of FSI……………122
Figure 6.5 The correlations between GLP-1 response after ingestion of 50g oral glucose
and OGIS and markers of β-cell function in type 2 diabetic patients and subjects
with different levels of FSI………………………………………………………123
Figure 6.6 The correlations between GLP-1 response after ingestion of 50g oral glucose
and OGIS and markers of β-cell function in non diabetic subjects with different
levels of FSI……………………………………………………………………..124
Figure 6.7 The correlations between GLP-1 response after ingestion of 50g oral glucose and
OGIS and markers of β-cell function in type 2 diabetic patients.........................125
xv
LIST OF APPENDICES
Page
Appendix 1 The correlations between fasting insulin and basal hepatic
insulin extraction, fasting insulin and postprandial hepatic
insulin extraction, and postprandial insulin response and
postprandial hepatic insulin extraction……………………………………….174
Appendix 2 The correlation between hepatic insulin extraction and
hepatic insulin resistance……………………………………………………..175
Appendix 3 Physical activity indices of the subjects by FSI………………………………176
Appendix 4 Mean (±SEM) 2-h postprandial plasma glucose, insulin, C-peptide,
GLP-1 concentrations after 50 g oral glucose plus 0, 5, and 30 g fat
in non-diabetic humans with different concentrations of FSI………………….177
Appendix 5 The correlations between relative glucose responses per g of fat
or protein and fasting insulin and between relative insulin responses
per g of fat or protein and fasting insulin………………………………………178
Appendix 6 Mean (±SEM) effects of 50 g glucose plus 0, 5, or 30 g fat or
protein on 2hr GLP-1 response in healthy nondiabetic subjects
with different levels of FSI……………………………………………………179
Appendix 7 Nutrient composition of whey protein…………………………………………180
Appendix 8 Screening form for fat and protein study………………………………………181
Appendix 9 3-day food records……………………………………………………………..182
Appendix 10 Consent form for the 1st study: Postprandial responses elicited by
fat and protein in normal and hyperinsulinaemic subjects……………………183
xvi
Appendix 11 Screening form for carbohydrate study……………………………………….191
Appendix 12 Concent form for the 2nd
study: postprandial response to
different carbohydrates in normal, hyperinsulinemic and
type 2 diabetic subjects………………………………………………………..192
xvii
ABBREVIATIONS USED IN THIS DISSERTATION
Analysis of covariance ANCOVA
Analysis of variance ANOVA
Area under the curve AUC
Blood pressure BP
BMI Body Mass Index
Carbohydrate CHO
Cholecystokinin CCK
Coefficient of variation CV
C-reactive protein CRP
Fasting serum insulin FSI
Glucose-dependent insulinotropic polypeptide GIP
Glucagon-like peptide-1 GLP-1
Glycemic Index GI
Glycemic load GL
Hemoglobin A1C HbA1c
High-density lipoprotein HDL
Homeostatic model assessment HOMA
Hyperinsulinemic Hyper[I]
Impaired glucose tolerance IGT
Insulin resistance IR
Insulin secretion rate ISR
xviii
Insulin sensitivity index ISI
Low-density lipoprotein LDL
National Health and Nutrition Examination Survey NHANES
Normal glucose tolerance NGT
Oral glucose tolerance test OGTT
Peptide YY PYY
Rate ratio RR
Relative glycemic response RGR
Relative insulin response RIR
Standard error of the mean SEM
Sex hormone binding globulin SHBG
Total area under the curve TAUC
Type 2 diabetes T2DM
Waist circumference WC
xix
PUBLICATIONS AND PRESENTATIONS ARISING FROM DISSERTATION
Peer reviewed publications
Lan-Pidhainy, XM, Wolever, TMS. The hypoglycemic effect of fat and protein is not attenuated
by insulin resistance. American Journal of Clinical Nutrition, 2010, 91(1), pp. 98-105.
Lan-Pidhainy, XM, Wolever, TMS. Are the glycemic and insulinemic index values of
carbohydrate foods similar in healthy control, hyperinsulinemic and type 2 diabetic patients?
European Journal of Clinical Nutrition, 2011, 28, pp1-8.
Presentations
28th
International Symposium on Diabetes and Nutrition of the Diabetes and
Nutrition Study Group (DNSG) of the EASD, July 1-4, 2010, Oslo, Norway. ―Are the glycemic
and insulinemic index values of carbohydrate foods the same in normal, hyperinsulinemic and
type 2 diabetic patients?‖
The 69th
Scientific Session of American Diabetes Association. New Orleans, LA, USA. June 5-
9, 2009. ―Attenuated GLP-1 response after oral glucose is associated with reduced β-cell
function in humans‖ and ―The postprandial insulin-raising effect of protein is not due to
increased insulin secretion but due to reduced hepatic insulin extraction‖.
1
CHAPTER 1
INTRODUCTION
2
1. Introduction
High postprandial blood glucose concentrations, even within the non-diabetic range [i.e.
impaired glucose tolerant (IGT) subjects with 2-hour postprandial glucose between 7.8 and 11.0
mmol/l in the 75g oral glucose tolerance test (1)], are associated with increased risk of
cardiovascular disease (2), diabetes (3) and cancer (4). In order to maintain the blood glucose
concentration within a narrow physiological range, an intricate system of neural, hormonal and
direct nutrient responses are initiated after a meal, which are all mediated in varying degrees by
macronutrient (carbohydrate, fat and protein) (Figure 1.1); however, in individuals with obesity,
diabetes and insulin resistance, the physiological mechanisms elicited by the macronutrient may
become abnormal, thereby disrupting substrate use and glucose homeostasis (Figure 1.2). Many
studies have compared the postprandial metabolic responses elicited by macronutrient in healthy
and diabetic subjects; however, less attention has been paid to non-diabetic subjects with insulin
resistance/ hyperinsulinemia.
Studies comparing the acute effects elicited by macronutrient in healthy and diabetic
subjects have demonstrated that in the latter, a series of metabolic responses to macronutrient is
adversely affected. For example, the glucose-lowering effects of fat and protein were attenuated
or absent in diabetic subjects (5, 6) and insulin secretory response following carbohydrate
ingestion or a mixture of carbohydrate and protein hydrolysate was substantially impaired in type
2 diabetes compared to the control subjects (6). Furthermore, the secretion and action of gut
hormones such as cholecystokinin (CCK), peptide YY (PYY), glucose-dependent insulinotropic
polypeptide (GIP) and glucagon-like peptide-1 (GLP-1) may also be attenuated as well in
diabetic subjects. For instance, the meal-induced increase in CCK was lower in type 2 diabetic
subjects than in the control (7) and the circulating PYY response to a meal was blunted in people
3
with type 2 diabetes compared to BMI-matched control subjects (8). These various
abnormalities of metabolic response to macronutrient in diabetic subjects suggest that the same
may occur in non-diabetic subjects with insulin resistance/hyperinsulinemia. Therefore, it is
important to understand the impact of insulin resistance/hyperinsulinemia on macronutrient
metabolism.
Insulin resistance is a pathologic state in which the target cells fail to respond or have
reduced sensitivity to metabolic actions of normal levels of circulating insulin (9). It is well
known that beside β-cell failure, insulin resistance is the major pathophysiological event
contributing to the development of type 2 diabetes mellitus (10-12). Further, insulin resistance is
tightly associated with major public health problems such as obesity, hypertension, coronary
artery disease, dyslipidemia and a cluster of metabolic and cardiovascular abnormalities that
define the metabolic syndrome (13, 14). Therefore, it is of great importance to understand the
impact of hyperinsulinemia/insulin resistance on macronutrient metabolism, specifically whether
the degree of insulin sensitivity of the subjects influences the effects of macronutrient
(carbohydrate, fat and protein) on acute postprandial glucose and insulin responses.
The blood glucose raising potential of carbohydrates in foods is numerically classified as
Glycemic Index (GI), which is calculated as the incremental area under the blood glucose
response curve (AUC) for each food expressed as a percentage of the area after taking the same
amount of carbohydrate as glucose (15,16). Many foods have been tested for their GI values;
however, this is usually done either in healthy or diabetic subjects (17). Though GI values of
foods did not differ significantly among a group of heterogeneous subjects with diabetes (18) and
were similar in healthy and diabetic subjects (19-21), the GI of foods has not been determined in
non-diabetic, insulin resistant/hyperinsulinemic subjects, a population highly susceptible to
4
diabetes. It is not known whether higher fasting insulin concentration and higher postprandial
insulin response of the subjects affect the Glycemic Index, thus its validity and clinical utility in
the prevention and management of diabetes.
Chronic hyperinsulinemia plays a significant role in the pathogenesis of insulin resistance
and associated chronic diseases (22-27); therefore, the extent of insulin response elicited by food
is as important as that of the glucose response. Some investigators have expressed concern that
the GI does not adequately address concurrent insulin response (i.e. lack of close association
between GI and Insulinemic Index), therefore, they have began to report values for Insulinemic
Index (28-30). Insulinemic Index is calculated similarly to GI so as to measure the extent to
which the available carbohydrate in food raises plasma insulin (31). The plasma insulin response
elicited by a meal is not only glucose-dependent but also tightly connected with the normal
function of gastrointestinal hormones (i.e. GLP-1, GIP, and CCK), the activity of the entero-
insulin axis, β-cell function and hepatic insulin extraction. All of these may be abnormal in
individuals with obesity, diabetes and insulin resistance/hyperinsulinemia.
Adding fat and/or protein to carbohydrate food reduces blood glucose response compared
to carbohydrate alone (32-34); The exact mechanisms associated with the hypoglycemic effect of
fat and protein are not clear, though they are suggested to be through similar mechanisms such as
delaying gastric emptying (35) and/or enhancing insulin secretion through augmented GIP and
GLP-1 secretion (36); however, a study in mice found that glucose-induced incretin hormone
release was differently modulated by fat and protein (37). In the same study, it was also found
that insulin response to glucose was 3-fold augmented by protein, and was associated with
enhanced oral glucose tolerance, whereas, the addition of fat resulted in only a 1.5 fold increase
in insulin with no accelerated glucose disposal (37). This is consistent with the result of a human
5
study that protein reduced glucose response 2 to 3 times more than fat, and there was no
significant fat × protein interaction (38). A recent study in rats found that upper small intestinal
lipids activate a gut-brain-liver neural axis to inhibit liver glucose production and decrease
plasma glucose level (39); however, it is not known whether protein also exerts similar effects.
Taken together, both animal and human studies suggest that fat and protein may modulate
postprandial glucose response through different mechanisms.
In summary, there is evidence that the acute effects of macronutrient (carbohydrate, fat
and protein) on glucose and insulin responses are adversely affected in patients with diabetes
compared to healthy subjects; however, it is not known whether the same occurs in non-diabetic
subjects with insulin resistance/hyperinsulinemia. In addition, the underlying mechanisms
associated with the hypoglycemic effects of fat and protein are unclear. In terms of the clinical
utility of GI and Insulinemic Index, it remains to be determined whether insulin
resistance/hyperinsulinemia modifies GI and Insulinemic Index, and whether GI and Insulinemic
Index are valid measures of the biological effects of carbohydrate foods in all conditions
regardless of subject‘s glucose tolerance status or degrees of insulin sensitivity.
6
Figure 1.1 The mechanisms regulating postprandial glycemia in healthy subjects. Many
substances in foods can affect postprandial glycaemic response. The major ones are
carbohydrate, fat and protein. Blood glucose rises after meal ingestion. In order to maintain
glucose homeostasis, the human body orchestrates both the peripheral and central nervous
system (CNS) to lower postprandial glucose levels. Three major pathways have been identified:
first, glucose stimulates pancreatic β-cell secretion of insulin. Insulin regulates glucose
concentration by acting upon the peripheral organs (i.e. insulin inhibits free fatty acids release
from adipose tissue, inhibits liver glucose production, and stimulates both skeletal muscle and
adipose tissue uptake of glucose). The second pathway is through gastrointestinal hormones
such as GLP-1, GIP, CCK and PYY, which work either by delaying gastric emptying, inhibiting
food intake, or through incretin mediated glucose-dependent insulin secretion. The third
pathway is that upper intestine lipids [long chain fatty acid (LCFA) is the main metabolite of
dietary fat digestion, and LCFA-CoA is the metabolite of LCFA] activate intestine-brain-liver
neural axis to increase liver insulin sensitivity and reduce hepatic glucose production (39). This
pathway, though well demonstrated in animal models, is extremely difficult to show in humans.
Nevertheless, through the concerted efforts of these different pathways, the glucose level goes
down.
7
Figure 1.2 The mechanisms regulating postprandial glycemia are disrupted in diabetes,
obesity and insulin resistance. For example, the subjects may have altered release pattern of
gut hormones or there is a reduced effect of gut hormones on gastric emptying, food intake and
insulin secretion, or they may have acquired defects in nutrient sensing (i.e. in diabetes and
obesity, nutrient-sensing mechanisms in both the gut and the brain are disrupted, leading to a
disregulation of glucose levels), or insulin loses its ability to carry out the functions such as
inhibition of free fatty acids release from adipose and suppression of hepatic glucose production.
Because of these disruptions, the blood glucose remains elevated.
8
CHAPTER 2
LITERATURE REVIEW
9
2. 1 Effects of Macronutrient on Postprandial Glucose and Insulin Responses
2.1.1 Dietary Carbohydrate
Dietary carbohydrate is the main dietary component affecting postprandial blood glucose
and insulin responses (40). In addition, carbohydrate foods provide energy, water-soluble
vitamins and minerals, and fiber. It is recommended that diets provide 45-65% of calories from
carbohydrate, with a minimum intake of 130g carbohydrate/day for adults (41).
The important dietary carbohydrates consist of monosaccharide (i.e. glucose, fructose [a
component of sucrose and also presents as a monosaccharide in fruits] and galactose [a
component of lactose or milk sugar]), disaccharides (pairs of monosaccharide linked together,
such as maltose, sucrose and lactose), oligosaccharides (chains of 3 to 10 glucose or fructose
polymers), and polysaccharides (i.e. starch and dietary fiber, which consists of long chains of
glucose or other monosaccharide molecules ).
Dietary carbohydrates influence glucose and insulin responses through at least 4 major
mechanisms: 1) the type of monosaccharide absorbed; 2) the amount of carbohydrate ingested;
3) the rate of carbohydrate digestion and absorption; and 4) the colonic fermentation (42). It is
recognized that both the type and amount of carbohydrate in a food influence the postprandial
glucose and insulin responses (43).
1) The type of monosaccharide absorbed
The major monosaccharides absorbed are glucose, galactose and fructose. However,
glucose is the only monosaccharide that appears in blood to any large extent (5-7 mmol/L)
whereas neither fructose (44) nor galactose (45) raises plasma glucose or insulin appreciably.
10
The differences in glucose and insulin raising potential of the monosaccharides (glucose,
fructose and galactose) resulted from their uniquely different metabolism. Take fructose for
example. Unlike glucose, fructose stimulates only modest insulin secretion and does not require
the presence of insulin to enter cells (46). In addition, fructose is rapidly utilized by the liver and
it bypasses the early, rate-limiting steps of glucose metabolism and is rapidly converted to
fructose-1-phosphate, which is subsequently converted to lactate, glucose, and glycogen (47).
Studies in both healthy and diabetic subjects demonstrated that fructose produces smaller
postprandial increases in plasma glucose and insulin responses than other common carbohydrates
(48-51); however, adding fructose (or sucrose) in large amounts to the diet is not recommended
because it adversely affects serum lipids and raises serum triglycerides (52, 53). These
undesirable effects may be due to lower insulin excursion after fructose, which results in less
activation of adipose tissue lipoprotein lipase and impaired triglyceride clearance (51).
2) The amount of carbohydrate ingested
Plasma glucose and insulin concentrations increase as the amount of carbohydrate
consumed increases. In healthy subjects, the AUC of plasma insulin increases linearly as the
dose of carbohydrate ingested increases from 0 to 100 g; however, the shape of the dose-
response curve for blood glucose is non-linear, with the curve flattening off as the dose of
carbohydrate increases, particularly over 50g (44, 54). The shape of dose-response curve for
glucose is very similar for healthy (54) and diabetic subjects (55).
3) The rate of carbohydrate digestion and absorption
11
The glycemic impact of carbohydrate foods is thought to be determined by the rate at
which they are digested and absorbed (15, 56). There are three lines of evidence to support this:
first, the rate of liberation of the carbohydrate products of digestion in vitro over 3-5 hrs is
closely associated with postprandial blood glucose response in vivo (57); second, a human
ileostomy model showed that carbohydrate malabsorption was insufficient to account for the
reduced glycemic response (58); third, consuming carbohydrate slowly mimics the metabolic
responses elicited by slowly digested carbohydrate foods (59, 60).
Numerous factors influence the digestion and absorption of carbohydrates in the small
intestine: type of starch (amylose vs. amylopection) (61), style of preparation (i.e. cooking
method and time, amount of heat and moisture used, degree of processing)(62-64), the physical
form of foods (i.e. juice vs. whole fruit, mashed potato vs. whole potato)(43), amount of fiber,
coingestion of fat and proteins (65, 66), presence of antinutrients such as α-amylase inhibitors
and phytic acid (67, 68), and transit time (rapidly digested starch vs. slowly digested starch )(69).
The rate and extent of starch digestion in vitro is commonly measured using the Englyst
method (69). Using controlled enzymic hydrolysis with pancreatin and amyloglucosidase, the
Englyst procedure measures rapidly digestible starch, slowly digestible starch and resistant starch
in carbohydrate foods. The released glucose is measured by colorimetry, using a glucose oxidase
kit (69). This technique also gives a value for rapidly available glucose, which includes rapidly
digestible starch, free glucose and the glucose moiety of sucrose (70). Using this technique,
Englyst et al measured the rapidly available glucose of 39 carbohydrate foods, and found that the
reported GI values were positively associated with the measured rapidly available glucose
(r=0.76, p<0.001) (70). This finding further confirms the concept that GI allows foods to be
ranked on the basis of the rate of digestion and absorption of the available carbohydrates.
12
4) Colonic fermentation
Carbohydrates resistant to digestion and those that escape absorption in the small
intestine enter the colon where they are fermented with the production of hydrogen, methane and
short chain fatty acids. The short chain fatty acids consist primarily of acetate, propionate and
butyrate (71, 72). The colonic fermentation of the undigested carbohydrates can be assessed by
measuring breath hydrogen and methane (73), or through the more quantitative method by
measuring acetate kinetics with 13
C-labelled acetate (73, 74).
The short chain fatty acids influence systemic glucose and lipid metabolism via direct or
indirect mechanisms. For example, though acetate has no direct effect on glucose metabolism
(75), it can reduce blood glucose in the long term by reducing serum free fatty acids
concentrations (76, 77); however, colonic propionate is a gluconeogenic substrate that raises
blood glucose (78) but inhibits the utilization of acetate for cholesterol synthesis (78).
2.1.1.1 Glycemic Index
The concept of Glycemic Index (GI) was founded on the notion that slowly digested
carbohydrates may reduce the metabolic risk factors associated with diabetes and coronary heart
diseases (79). The GI is a measure of carbohydrate quality. It ranks carbohydrate foods by
measuring the extent to which available carbohydrate in foods raises blood glucose relative to
that of oral glucose (15). The alternative measure of glycemic response to carbohydrate-
containing foods is glycemic load (GL), the product of dietary GI and the amount of available
dietary carbohydrate, which combines both the qualitative and quantitative measures of
carbohydrate and reflects the overall glycemic impact of the diet (80).
13
Since its conception decades ago, the GI has elicited much controversy and its clinical
relevance has been questioned and vigorously debated (34, 81-84). One of the major criticisms
of GI is that it does not apply in mixed meals because of the confounding effects of fat and
protein on glycemic response (34, 85). Studies of GI in mixed meals yield conflicting results -
some find the observed glycemic response to mixed meals were predicted by the GI and
carbohydrate contents of the foods (86, 87) whereas others do not (85, 88, 89). Venn et al
examined the studies on mixed meals, and concluded that summing the GI values of individual
components of a meal does not reliably predict the observed GI of the meal as a whole (90).
Furthermore, they argued that the association of low GI foods with a low glycemic response
alone does not necessarily justify a health claim (90). The rationale being that low GI foods are
usually naturally occurring and minimally processed carbohydrate containing cereals, vegetables
and fruit, and these foods have nutritional qualities (i.e. phytochemicals, antioxidants) other than
their immediate impact on postprandial glycemia as a basis to recommend their consumption
(90). Out of concern for these limitations, the 2007 FAO/WHO Scientific Update on
Carbohydrates in Human Nutrition cautioned that food selection should not be based on GI alone
and suggested that ―GI is perhaps most appropriately used to guide food choices when
considering similar carbohydrate-containing foods, for example bread with a low GI may be
preferable to a higher GI bread, with a resultant lower glycemic load (GL)‖(91).
Despite the controversies over the clinical utility of GI, evidence is accumulating that
low-GI or low-GL foods may play a role in the prevention and management of chronic diseases
such as diabetes, coronary heart disease, and certain types of cancer. The most recent 2011
American Diabetes Association position statement states that ―for individuals with diabetes, use
of GI and GL may provide a modest additional benefit for glycemic control over that observed
14
when total carbohydrate is considered alone‖(92). The following paragraphs illustrate the role of
low-GI or low-GL diet in the prevention and management of chronic diseases such as diabetes,
coronary heart disease, and certain types of cancer.
GI and diabetes mellitus
Several randomized clinical trials have reported that low GI diets were effective in
controlling glycemia in diabetic subjects (93-99), but others have not confirmed this effect (100-
102); therefore, to clarify the issue of the effect of low-GI diets in the management of diabetes,
Brand-Miller et al conducted a meta-analysis of 14 randomized controlled trials to compare low-
GI diets with conventional or high-GI diets in the management of type 1 and type 2 diabetes.
The meta-analysis showed that glycated proteins were reduced by 7.4% and hemoglobin A1C
(HbA1c) by 0.43% more on the low-GI diet than on the high-GI diet. This result is clinically
significant and is on par with that achieved by pharmacological agents (103). Recently, a meta-
analysis of 37 prospective cohort studies found that diets with a high GI or GL independently
increased the risk of type 2 diabetes (GI RR=1.40, 95% CI:1.23, 1.59; GL RR=1.27, 95%
CI:1.12, 1.45)(104).
In the most recent Cochrane review (11 randomized control trials involving 402
participants), the authors compared low- and high-GI diets for people with either type 1 or type 2
diabetes mellitus. They found a significant reduction in HbA1c (weighted mean difference of -
0.5%, 95% CI: -0.9, -0.1, P=0.02); moreover, the episodes of hypoglycemia were significantly
fewer in low- compared to high-GI diet (105). Recently, an updated version of a Cochrane
review including 12 randomized control trials and 612 participants found that low-GI diets
lowered % HbA1c levels by 0.4% (95% CI: -0.7, -0.20, P=0.001) compared with the comparison
15
diets (high-GI diet, carbohydrate exchange diet, and high-cereal fiber diet)(106). The reduction
of 0.4-0.5% HbA1C is clinically significant and is comparable to decreases achieved through
medications for newly diagnosed type 2 diabetes (107, 108).
Two plausible mechanisms linking the development of type 2 diabetes with high-GI diet
have been proposed (109). First, high-GI foods produce higher blood glucose concentrations and
a greater demand for insulin. Chronic hyperinsulinemia may eventually result in pancreatic β-
cell failure as the result of overstimulation, consequently, impaired glucose tolerance and frank
diabetes. Second, high-GI diet may result in insulin resistance because it directly leads to
postprandial hyperglycemia, counter-regulatory hormone secretion and increased late
postprandial serum free fatty acids.
GI and coronary heart disease
A low-GI diet also has implications in the prevention of coronary heart disease. Clinical,
metabolic and epidemiological studies have unanimously demonstrated that low-GI foods may
favorably affect metabolic predictors of coronary heart disease. In a recent 6-month randomized
parallel trial comparing the effect of low-GI diet vs. a high-cereal fiber diet on type 2 diabetes,
low-GI diet increased high-density lipoprotein (HDL) cholesterol by 1.7mg/dl, and dietary GI
was negatively related to HDL concentration (r=-0.19, p=0.009)(110). In a metabolic study,
higher GI scores were associated with higher fasting triglycerides and lower HDL-cholesterol
levels (111). In epidemiological studies, the British Adults Study demonstrated a significant
negative relation between serum HDL-cholesterol concentration and the GI of the diet for both
men and women (112), and the Nurse‘s Health Study showed a positive relation between GI and
fasting serum triglycerides (113).
16
In a recent systematic analysis of prospective cohort studies and randomized trials
investigating dietary exposures in relation to coronary heart disease, the authors implemented
Bradford Hill criteria to derive a causation score and identified strong evidence that there is a
causal association between coronary heart disease and the consumption of high-GI diet (114).
GI and cancer
The insulin/colon cancer hypothesis suggests that diets with a high glycemic response
and consequent hyperinsulinemia may play a role in colorectal carcinogenesis (115). The
biologically plausible mechanism being that the diets with high-GI and/or high-GL may
influence cancer risk via hyperinsulinemia and insulin-like growth factor axis, the major
determinants of proliferation and apoptosis (116). However, epidemiological studies
investigating the association between dietary GI and GL and the risk of cancers of digestive tract
have yielded inconsistent findings (117). In addition, a recent meta-analysis did not find any
association between dietary GI/GL and colorectal or pancreatic cancer risk (118).
For the role of GI/GL in the risks of breast cancer, the findings are conflicting as well. A
recent prospective cohort study of Swedish women (n=61433, mean follow-up of 17.4 years)
showed a significant positive relationship between both high-GI and high-GL diets and the risk
of developing breast cancer (119); however, in the Shanghai women‘s health study, a population-
based cohort study (n=74942, mean follow-up of 7.35 years), only diets with high-GL were
found associated with breast cancer risk in premenopausal women (<50 yr old). GI was not
found to be associated with breast cancer risk, which was ascribed to the narrow range and
centered distribution of the observed GI values (120). Even the meta-analyses show inconsistent
results: two meta-analyses did not find any significant association between GI/GL and the risk of
17
breast cancer (121, 122), whereas one meta-analysis of 37 prospective cohort studies found that
diets with a high GI independently increased the risk of breast cancer, the rate ratio (RR) for the
comparison of the highest with the lowest quartiles of GI is 1.08 (95% CI: 1.02, 1.16) (104).
The lack of significant association between GI/GL and cancer risks doesn‘t warrant the
rejection of the insulin/colon cancer hypothesis; in fact, the heterogeneity in epidemiological
studies investigating the association between GI/GL and cancer risks may be due to the lack of
specifically selected and validated dietary questionnaires to assess GI or GL, and thus may lead
to under- or over-estimation of GI or GL values. Indeed, a recent study has revealed several
methodological challenges in assigning GI values to food items using existing GI tables and
dietary questionnaires that were not designed specifically for the study purpose (123).
2.1.1.2 Insulinemic Index
Because of the implication of raised plasma insulin in the pathogenesis of chronic
diseases (22-27) and the concern that GI fails to consider the insulin response (124), some
investigators have begun to report values for Insulinemic Index (28-30). Insulinemic Index is
calculated in similar way as GI [the area under the blood insulin response curve for test food
expressed as a percentage of the area after taking the same amount (50g) of carbohydrate as
glucose] to measure the extent to which the available carbohydrate in foods raises plasma insulin
(31). In general, there is a good correlation between the GI of the foods and their Insulinemic
Index (r=0.69, p<0.001) (125), with the exceptions of items such as milk (126) and protein-rich
foods (30), which elicited insulin responses that were disproportionately higher than their
glycemic responses. Although statistically significant, the GI only explains about 25-50% of the
variation in Insulinemic Index, which may be because protein or fat in a carbohydrate-rich meal
18
evokes additional or synergistic insulin secretion to contribute to the reduction in glycemia (127-
130). Nevertheless, this lack of close association between GI and Insulinemic Index is one of the
major criticisms of the GI (34) and the rationale for advocating using food Insulinemic Index to
classify foods (131).
A recent study in lean young healthy subjects found that the observed insulin responses to
mixed meals were strongly correlated with insulin demand predicted by the Insulinemic Index of
component foods (131). Hence, the authors suggested using the Insulinemic Index of composite
meals to estimate pre-prandial insulin doses for patient with type 1 diabetes (131). However, it is
recognized the worsening defects of insulin secretion in diabetic subjects may prevent the
detection of Insulinemic Index differences among foods (131). The bottom line is that for
Insulinemic Index to be a valid property of a food and be applicable in a broader population, its
values should be similar in different subjects regardless of their glucose tolerance status or
degree of insulin sensitivity.
2.1.2 Dietary Fat
Numerous studies have demonstrated that adding fat to carbohydrate reduces the blood
glucose response (129, 132, 133); however, adding a large amount of fat to diet is not advised
due to its detrimental effects on lipid metabolism. For example, adding a large amount of
sunflower oil (40g) to pasta initially delays postprandial glycemia, but 4 hours later, it leads to
high levels of triglycerides, non-esterified fatty acids and insulin resistance (134).
It is proposed that adding dietary fat to carbohydrate attenuates postprandial glycemic
response either through reduced rate of gastric emptying, resulting in a delay in carbohydrate
absorption (129, 135, 136) or the incretin hormone effect, mainly GLP-1 and GIP mediated
19
glucose-dependent insulin secretion (35, 137). Recently, a new pathway of the lipid-mediated
regulation of glucose homeostasis has been established. It was found that in rodents, upper
intestinal lipids activate a gut-brain-liver neural axis to inhibit liver glucose production and
decrease plasma glucose concentration (39).
The effects of fat on acute glycemic responses may differ in solid vs. liquid meals. For
example, adding 0, 5, 10, 20 and 40g solid fat (canola margarine) to white bread reduced
glycemic response in a dose-dependent, but non-linear manner (138). By contrast, adding 0, 5,
10 or 30g corn oil to oral glucose reduced glycemic responses in a linear manner (38).
Besides the physical form of fat, the fat source also differentially influences postprandial
glucose responses. Plasma glucose after butter-containing meals was higher than after meals
containing olive or corn oil (139). This may be due to differences in the ability of different types
of fat to stimulate insulin (140) and GLP-1 secretion (140). For example, monounsaturated fatty
acids stimulate much higher insulin secretion than saturated fatty acids (140), and fats rich in
monounsaturated fatty acids (i.e. olive oil) induced higher GLP-1 concentration than butter
(141). Another reason could be due to the differences in the proportion of fatty acids absorbed
via the portal vein as opposed to chylomicrons which, in turn, may influence hepatic glucose
output (134) and/or hepatic insulin extraction (142).
2.1.3 Dietary Protein
It is generally thought that adding protein to carbohydrate reduces glycemic responses by
stimulating insulin secretion (127) and/or delaying gastric emptying (143). However, the
reported effects of protein on postprandial glucose and insulin responses are inconsistent, and the
magnitude of the response varies tremendously (127, 144-146). These variations do not seem to
20
be explained by subject groups since a variety of effects are seen in both healthy and diabetic
subjects (147). Review of the literature suggests that the most likely explanations are the
differences in digestion kinetics and the source of protein.
Dietary proteins are heterogeneous with regards to their rate of digestion and absorption
(148), which affects insulin response to varying degrees. It has been demonstrated in both
healthy (149) and type 2 diabetic subjects (150) that ingestion of whey protein (a fast protein,
which is digested and absorbed quickly from the gut) with a medium calorie, high protein mixed
meal resulted in greater postprandial amino acid concentrations and in a 13-40% greater β-cell
response (evaluated as the rise in insulin, C-peptide, and proinsulin) than ingestion of casein (a
slow protein). Gannon et al found that in type 2 diabetic patients, ingestion of 50g of cottage
cheese resulted in significantly higher plasma insulin concentration compared to an equivalent
amount of egg white (127), which was attributed to the relatively poor digestibility of egg white
relative to cottage cheese (151).
Numerous studies have shown that different sources of protein resulted in different
glucose and insulin responses (127, 150, 152, 153). The discrepancies may be either due to
different amino acid composition of the proteins (153), different effect on gastric emptying (152)
or differential effects of varying protein sources on incretin hormone secretion (154). For
example, Van et al reported that in healthy subjects, a carbohydrate drink containing a mixture of
free leucine, phenylalanine, and arginine produced a much larger insulinotropic effect than pea,
whey hydrolysate, or intact casein, and the differences were attributed to the different amino acid
composition of the proteins (153). Lang et al found that in healthy, normal-weight men, a mixed
meal containing 50g soy protein elicited an earlier metabolic response (insulin, glucagon, and
glucose) than an equivalent amount of casein, with gelatin presenting intermediate results (152).
21
It was suggested that the differences may be explained by modifications in the rate of gastric
emptying (152) since it is well known that casein precipitates in the stomach and significantly
reduces the rate of gastric emptying (155). For the varying effects of different types of protein
on incretin hormone secretion, Nilsson et al showed that in healthy subjects, whey protein was a
much stronger GIP secretagogue than other food proteins such as cod, gluten, and cheese (156).
There is also evidence that whey protein induces higher GLP-1 response than casein (150).
2.2 Insulin Resistance and Compensatory Hyperinsulinemia
Insulin resistance refers to a reduced sensitivity to insulin-stimulated glucose uptake (9).
It results in inability of insulin to provide normal glucose and lipid homeostasis, and is
manifested as impaired suppression of hepatic glucose production (157), impaired glucose
uptake by skeletal muscle (158), and disinhibited lipolysis in adipose tissue (159). In order to
maintain glucose homeostasis, hyperinsulinemia occurs as an obligatory compensation for
insulin resistance and is the hallmark of insulin resistance (160). Chronic hyperinsulinemia can
also down-regulate insulin action, indeed, when hyperinsulinemia was induced experimentally
over a 48-72 hr period at normal-glycemic conditions, the healthy subjects became insulin
resistant (161).
Insulin resistance plays an important etiological role in the development of type 2
diabetes. In an individual with normal blood glucose concentrations, insulin resistance is usually
associated with a compensatory increase in plasma insulin levels (hyperinsulinemia) (162). As
long as pancreatic β-cells are able to augment their secretion of insulin sufficiently to
compensate for insulin resistance, glucose tolerance remains normal (163). However, over time,
22
pancreatic β-cell begins to fail and can no longer compensate for insulin resistance, impaired
glucose tolerance occurs, and overt type 2 diabetes develops (10-12).
Currently, there is disagreement over whether insulin resistance and hyperinsulinemia are
the cosegregated metabolic traits and whether they play a different role in the pathogenesis of
clinical phenotypes associated with the two abnormalities. Ferrannini et al measured insulin
resistance (by euglycemic-hyperinsulinemic clamp) and insulin response to oral glucose
tolerance test (OGTT) in a cohort of 1308 non-diabetic European subjects. They found that only
60% of the most insulin-resistant individuals (bottom quartile of insulin sensitivity) had the
highest insulin response (top quartile of insulin response) (164). Henceforth, they suggested that
insulin resistance and hyperinsulinemia may be two dis-associated entities and carry different
pathogenic potential. However, Kim & Reaven argued that Ferrannini et al‘s analysis did not
provide any information on the percentage of individuals in quartiles of insulin response by
quartiles of insulin sensitivity (165). Taking this issue into consideration, they found that insulin
resistance and hyperinsulinemia are well correlated (r=0.76, p<0.001) and rarely exist in
isolation from one another in a nondiabetic population; in addition, there were minimal
differences in CVD risk factors between individuals with different insulin responses but within
the same insulin sensitivity quartile (165). Therefore, Kim and Reaven suggested that from a
pathophysiological point of view, it is better to view insulin resistance and compensatory
hyperinsulinemia as one entity in nondiabetic individuals rather than using statistical method to
discern which abnormality is caused by insulin resistance and which is caused by
hyperinsulinemia (165).
23
2.2.1 Methods to measure insulin sensitivity/resistance
A number of direct, indirect and surrogate measurements of insulin sensitivity have been
developed (Table 2.1). The direct measures of insulin sensitivity include insulin suppression test
(166) and the ―gold-standard‖ hyperinsulinemic-euglycemic glucose clamp, in which insulin is
infused at a constant rate and glucose is held at basal levels by glucose infusion (167). The
indirect measurement of insulin sensitivity includes the minimal model analysis of frequently
sampled intravenous glucose tolerance test (168). These methods allow the measure of
maximum insulin action and minimize the effect of non-insulin-mediated glucose uptake.
However, the expense and technical difficulties associated with these methods render them rather
impractical for routine screening. Due to that reason, a number of surrogate indices of insulin
sensitivity/resistance based on fasting glucose and insulin such as homeostasis model assessment
(HOMA) (169) and quantitative insulin-sensitivity check index (QUICKI) (170) have been
developed. Though both methods have been widely used, their sensitivity and specificity in
detecting insulin resistance has been questioned. A study compared the insulin sensitivity
assessment indices (including HOMA and QUICKI) with euglycemic-hyperinsulinemic clamp
data after a dietary and exercise intervention in older adults, and found that these indices vary
substantially from the clamp method in their ability to assess insulin sensitivity (171). One
reason could be that surrogate indices based on fasting glucose and insulin concentrations reflect
primarily hepatic insulin sensitivity/resistance, and do not represent whole-body insulin
sensitivity (172). In response to this, a number of oral-glucose-tolerance-test (OGTT) based
indices of insulin sensitivity have been developed to evaluate whole-body insulin sensitivity.
The most frequently used ones are: 1) Matsuda‘s insulin sensitivity index calculated as
10,000/(fasting glucose × fasting insulin × mean glucose × mean insulin )0.5
, which takes into
24
account both hepatic and muscle insulin sensitivity, and is highly correlated (r = 0.73, p <
0.0001) with the rate of whole-body glucose disposal during the euglycemic-hyperinsulinemic
clamp (173); 2) Mari‘s oral glucose insulin sensitivity based on a physiological glucose-insulin
model, which is in good agreement with hyperinsulinemic-euglycemic glucose clamp (r=0.77,
p<0.0001) (174). To differentiate the relative contribution of liver and muscle to whole-body
insulin sensitivity, insulin resistance at the site of liver (hepatic insulin resistance) was calculated
as the product of total AUC of glucose and insulin during the first 30 minutes (glucose0-30[AUC]
x insulin0-30[AUC]) (172) and muscle insulin sensitivity was calculated as the rate of decay of
plasma glucose concentration from its peak value to its nadir divided by the mean insulin
concentration (172). Both indices of hepatic insulin resistance and muscle insulin sensitivity
derived from OGTT have been validated using hyperinsulinemic-euglycemic clamp (172).
Fasting insulin is often used as a surrogate measure of insulin resistance based on the
finding that among non-diabetic subjects, high fasting insulin level is moderately well correlated
with the euglycemic-hyperinsulinemic clamp (175) and minimal model techniques (176).
However, since fasting insulin levels are determined not only by the degree of insulin sensitivity
but also by pancreatic β-cell secretory function and hepatic insulin clearance, it is recognized that
fasting insulin is not an exact surrogate measure of insulin resistance. Nevertheless, it is
acknowledged that fasting insulin can be used as a simple, economical, and routine screening
tool for subjects in large clinical or population-based epidemiological studies (177).
25
Table 2.1 Methods to measure insulin sensitivity/resistance
Methods Primary Organ Formula
Direct measure
Hyperinsulinemic-euglycemic glucose clamp Whole-body For detailed procedure, see ref (167)
Insulin suppression test Whole-body For detailed procedure, see ref (166)
Indirect measure
Minimal model analysis of FSIVGTT Peripheral & Liver For detailed procedure, see ref (168)
OGTT-derived indices
Matsuda‘s insulin sensitivity index Liver & Muscle 10,000/(fasting glucose × fasting insulin × mean
glucose × mean insulin )0.5
(173)
Mari‘s oral glucose insulin sensitivity Liver & Muscle For this calculation, go to:
http://webmet.pd.cnr.it/ogis/index.php (174)
Hepatic insulin resistance Liver glucose0-30[AUC] x insulin0-30[AUC] (172)
Muscle insulin sensitivity Muscle dG/dt ÷ mean plasma insulin concentration
(172)
Fasting glucose/insulin based indices
Homeostasis model assessment (169) Liver For this calculation, go to:
www.dtu.ox.ac.uk/homacalculator/index.php
Quantitative insulin-sensitivity check index Liver 1/[log(I0µU/mL) + log(G0mg/dL)] (170)
FSIVGTT: Frequently sampled intravenous glucose tolerance test; G:glucose; I:insulin.
dG/dt: the rate of decay of plasma glucose concentration from its peak value to its nadir.
26
2.3 Insulin Resistance/Hyperinsulinemia and the Risk of Chronic Diseases
Recent years have witnessed rapidly increasing trends in insulin resistance and
compensatory hyperinsulinemia. Li et al found that the prevalence of hyperinsulinemia
increased by 35.1% overall (38.3% among men and 32.1% among women) in the past decade
among non-diabetic U.S adults ( age ≥ 20 years) based on the Third National Health and
Nutrition Examination Survey (NHANES III, 1988-1994) and NHANES 1999-2002 data (178).
Recently, the same research group also found that hyperinsulinemia was independently
associated with pre-diabetes. In addition, the prevalence of pre-diabetes (impaired fasting
glucose and/or IGT) was four times higher among adolescents (12-19y) with hyperinsulinemia,
based on the most recent NHANES 2005-2006 data (179).
Insulin resistance/hyperinsulinemia is associated with increased risk of a cluster of
chronic diseases. In 1988, Reaven suggested that insulin resistance, characterized by
compensatory hyperinsulinemia, is the central element, and in large measure the cause of a
cluster of related abnormalities including glucose intolerance, dyslipidemia and hypertension
(180), which lead to a variety of clinical syndromes. For example, insulin
resistance/hyperinsulinemia has been associated with increased risk of type 2 diabetes (180,
181), coronary heart disease (26, 182, 183), essential hypertension (184), congestive heart failure
(185), polycystic ovarian syndrome (186), nonalcoholic fatty liver disease (187), Alzheimer‘s
disease (188) and cognitive decline (189) and cancers of certain sites such as prostate (190),
colon and rectum (191) and breast (192).
27
2.3.1 Mechanisms of insulin resistance induced clinical syndromes
The mechanisms through which insulin resistance/hyperinsulinemia increased the risks of
the aforementioned abnormalities and clinical syndromes are complicated and multifaceted. It is
proposed that the series of events that lead to the abnormalities and clinical syndromes are in
most instances, if not all, a result of differential tissue insulin sensitivity (14). The most
important organs involved are probably skeletal muscle, adipose tissue, kidney and liver. There
are a few lines of evidence to support this hypothesis: kidney retains normal insulin sensitivity in
the presence of muscle and adipose tissue insulin resistance (193), and the elevated plasma
insulin concentration enhances renal sodium reabsorption, thus contributing to the development
of hypertension (194). The second example is the etiology for nonalcoholic fatty liver. It was
suggested that peripheral insulin resistance (muscle and adipose tissue) led to increased lipolysis
and delivery of free fatty acid to the liver, which remains insulin sensitive to stimulate hepatic
triglyceride, thereby predisposing to the development of fatty liver (187). The third example is
the pathogenesis of breast cancer. Insulin has been shown to reduce sex hormone binding
globulin (SHBG) production in a hepatoma cell line (195), and in humans, insulin levels are
inversely related to SHBG and SHBG-bound estradiol; therefore, more circulating estradiol
becomes available with hyperinsulinemia (192), and estradiol is generally regarded as the most
important hormonal promoter of developing breast cancer (196). The fourth example is
polycystic ovary syndrome. High insulin acts through its own receptor (rather than the insulin-
like growth factor-1 receptor) on insulin sensitive ovary and adrenal to augment androgen
production (186). It is known that hyperandrogenemia is the hallmark of the polycystic ovary
syndrome (197).
28
Though the differential tissue insulin sensitivity hypothesis seems to explain the majority
of the insulin resistance induced clinical syndromes, certain diseases may not fall into this
category. Take Alzheimer‘s disease for example. It was suggested that brain hyperinsulinemia
(due to cerebral insulin resistance) may be associated with reduced amyloid-β clearance and
eventual accumulation of amyloid-β peptide, a pathological hallmark of Alzheimer‘s disease
(198). This is because both amyloid-β peptide and insulin compete for the insulin-degrading
enzyme for its own degradation; however, the affinity for the binding of insulin-degrading
enzyme to insulin is much greater than that for the amyloid-β peptide, as the result of which, less
amyloid-β peptides were cleared (199). It was even suggested that Alzheimer‘s disease could be
qualified as ―an insulin resistant brain state‖ due to the observation that activation of the insulin
receptor was impaired in brain autopsy sample of Alzheimer‘s disease patients (200).
2.4 Hepatic Insulin Extraction in the Regulation of Peripheral Insulin Concentration
Peripheral insulin concentration after meal ingestion is determined not only by pancreatic
insulin secretion, but also hepatic insulin clearance (Figure 2.1) (201, 202); therefore, the liver
plays a critical role in determining peripheral insulin levels. The liver is the primary site of
degradation of circulating insulin (203). In non-diabetic subjects, approximately 50–70% of the
insulin secreted into the portal system is removed by the liver, measured either by hepatic vein
catheterization techniques (204, 205) or by calculating the ratio of AUC of peripheral C-peptide
concentration to that of the insulin (206).
Reduced hepatic insulin clearance is usually observed after oral glucose or meal
ingestion, thereby increasing the systemic availability of insulin to aid glucose disposal (207-
209); however, the regulation of hepatic insulin clearance is not a static process, but rather
29
influenced by both physiological and pathophysiological states of individuals (Figure 2.1). For
example, increased hepatic insulin clearance had been reported in elderly men and women (210-
212) and patients with essential hypertension (213), whereas decreased hepatic insulin clearance
was observed in obese subjects (214, 215) and in patients with liver cirrhosis (216); in diabetic
patients, the results are conflicting as both normal (217, 218), decreased (219, 220), or even
increased (221) hepatic insulin clearance had been reported. The reasons for such divergent
findings are not clear. It is possible that the variations in liver fat content (222, 223) may
contribute to this. It is known that increased liver fat content is associated with impaired hepatic
insulin clearance (224).
Hepatic insulin clearance is decreased in obesity (214, 215), particularly in abdominal
obesity (225), which contributes to hyperinsulinemia in obesity. The cause for the reduction in
insulin clearance in obesity is not clear. It was suggested that pathological alterations in liver
metabolism may play a role (215). Evidence exist that elevated free fatty acids levels, which
commonly occur in obesity, impact on hepatic insulin extraction. For example, the in vitro
studies showed that free fatty acids impair insulin binding and degradation in isolated rat
hepatocytes (226-228). Furthermore, an in vivo study in dogs found that free fatty acids impair
hepatic insulin extraction (229). The study in non-diabetic humans found that increased liver fat
was associated with impaired liver insulin clearance (230). Fortunately, the diminishment of
hepatic insulin clearance in obesity seems to be reversible. In obese children and adolescents,
hepatic insulin clearance increased after weight loss, resulting in near normalization of insulin
levels (231).
Pharmacological agents also impact on hepatic insulin extraction, although their effects
vary tremendously. Take antidiabetic drugs for example, glyburide augments hepatic insulin
30
extraction whereas glipizide had no effect (232, 233); metformin enhances hepatic insulin
extraction in non-diabetic, first degree relatives of African Americans with type 2 diabetes (234);
and thiazolidinediones increase hepatic insulin clearance after oral glucose challenge in subjects
with impaired glucose tolerance and type 2 diabetics (235).
Hormones (sex or growth hormones) also affect hepatic insulin extraction. Monti et al
showed that the girls with Turner's syndrome [a chromosomal aberration caused by complete or
partial X chromosome monosomy in a phenotypic female (236)] have higher insulin clearance
than their healthy counterparts, and after growth hormone therapy, insulin clearance was
normalized, possibly to counteract hepatic insulin resistance (237). Krakover et al (238)
investigated the effects of female sex hormones on insulin binding and receptor-mediated insulin
degradation in hepatocytes from ovariectomized rats. They found that estradiol and progesterone
increased insulin binding and degradation, whereas testosterone decreased degradation. Thus,
they suggested that abnormalities in sex hormone levels could contribute to altered insulin
metabolism and peripheral hyperinsulinemia in androgenised women with abdominal obesity
(238).
There are a few methods to measure hepatic insulin extraction. The direct measurement
of hepatic insulin extraction involves hepatic vein catheterization techniques, with catheters
placed in artery and hepatic veins (204). The invasive nature of this technique prompted the
development of indirect, non-invasive approaches based on mathematical models to infer hepatic
insulin extraction from plasma measurements of C-peptide and insulin. Polonsky et al proposed
using the ratio of AUC of C-peptide to that of the insulin to reflect hepatic insulin extraction
(239). The rationale being that C-peptide is secreted from the pancreatic β-cells in equimolar
concentration with insulin, but not extracted by the liver (239). Toffolo et al assessed hepatic
31
insulin extraction during an insulin-modified intravenous glucose tolerance test (240). Hepatic
insulin extraction was determined by calculating insulin secretion rate (ISR) using plasma C-
peptide concentrations and the C-peptide minimal model and by calculating post-hepatic insulin
delivery rate using plasma insulin concentrations and the insulin minimal model (240). The
rationale for this method is that the simultaneous modeling of insulin secretion rate from C-
peptide concentration and insulin delivery rate in plasma after its passage through the liver from
insulin concentration provides an estimate of hepatic extraction profile and index (240).
32
Figure 2.1 Factors affecting peripheral insulin concentration. Peripheral insulin
concentration is determined by both pancreatic β-cell insulin secretion and hepatic insulin
extraction. Hepatic insulin extraction is influenced by multiple factors such as the physiological
and pathophysiological status of the individuals and use of pharmacological agents and
hormones. Various factors such as glucose, amino acids, free fatty acids, acetylcholine and
incretin hormones stimulate pancreatic insulin secretion.
33
2.5 Effects of Macronutrient on Postprandial Responses in Hyperinsulinemic Subjects
2.5.1 Dietary Carbohydrate
Glycemic Index (GI) and Insulinemic Index are quantitative measures of the blood
glucose and insulin raising potential of available carbohydrate in foods. The GI values of foods
are similar in healthy control and diabetic subjects (both type 1 and type 2 diabetes) (21, 241,
242). However, the GI values of foods have never been determined in non-diabetic, insulin
resistant/hyperinsulinemic subjects; thus we do not know whether they are similar in healthy
control, hyperinsulinemic, and diabetic subjects. It is important to clarify this issue because for
GI to have clinical utility, their values must be similar in different subjects regardless of glucose
tolerance status or degree of insulin sensitivity.
There are few data comparing the relative glucose and insulin responses to carbohydrate
test meals in different groups of subjects. A multicentre trial examining the relative glucose and
insulin response to a prototype standardized mixed meal (50g available CHO, 11g fat and 12g
protein) found that although the relative glucose responses were similar in lean normal, obese
normal, subject with impaired glucose tolerance (IGT) and subjects with type 2 diabetes, the
relative insulin responses were significantly different among obese normal, IGT and type 2
diabetic subjects (243). In another study comparing the acute plasma glucose and insulin
responses elicited by breakfast cereals with either high or low fiber content in 77 non-diabetic
subjects divided a priori into 35 with fasting plasma insulin (FPI) < 41pmol/L and 42 with FPI ≥
41pmol/L, it was found that the glycemic impact of a high fiber cereal, relative to that of a low
fiber cereal, did not differ in hyperinsulinemic subjects compared to healthy subjects. However,
relative insulin responses were inversely related to fasting insulin concentration (244). These
studies suggest that the relative insulin responses elicited by different carbohydrate meals differ
34
in hyperinsulinemic subjects compared to healthy subjects; however, these studies used meals
containing a mixture of foods and macronutrient. Thus no firm conclusion can be drawn.
Therefore, to see if different carbohydrates elicit similar relative glucose and insulin responses in
hyperinsulinemic and healthy subjects, it is important to study pure carbohydrates (i.e. sugars)
and carbohydrate foods fed alone and to control the amount of available carbohydrate in the
different test meals.
2.5.2 Dietary Fat
There is evidence that fat has no effect on postprandial glucose and insulin responses in
people with diabetes (5, 245). Simpson et al studied the glycemic and hormonal responses to
macronutrient in healthy subjects, patients with type 1 diabetes and patients with type 2 diabetes.
They found that fat markedly reduced the glycemic response to oral carbohydrate in non-
diabetics. However, in both type 1 and type 2 diabetes, the presence of fat had no significant
effect on the glycemic response (245). Since delaying gastric emptying is the presumed
mechanism for the reduced glycemic response in non-diabetic subjects, the authors suggest that
the absence of a fat effect in diabetic subjects may be due to the presence of mild gastric stasis
due to unrecognized autonomic neuropathy (245). Gannon et al found that adding 50g fat
(butter) to 50g carbohydrate reduced glycemic response in healthy, young subjects (246);
however, when the same study was done in older people with type 2 diabetes, there was
essentially no effect on blood glucose by ingestion of butter with carbohydrate meal (247). The
attenuated effect of fat on glucose-lowering is not a ―monopoly‖ of diabetes; a pilot study
comparing the effects of fat and protein on glycemic response in healthy control and
hyperinsulinemic subjects showed that adding fat to carbohydrate has a smaller effect in
35
reducing the glycemic response in hyperinsulinemic subjects than healthy subjects (38);
however, this result needs to be confirmed in larger studies.
The mechanism for the attenuated hypoglycemic effect of fat in diabetic subjects is not
clear. In rats with insulin resistance induced by high-fat diets, intraduodenal lipid infusion failed
to suppress glucose production, suggesting that lipid-induced gut-brain-liver neuronal network in
the regulation of glucose homeostasis may be impaired following high-fat diets (39). Therefore,
it is possible that the upper intestine of insulin resistant and diabetic subjects may have acquired
defect(s) in lipid sensing (through an intestine-brain-liver neurocircuitry), thus hindering glucose
homeostasis regulation, or they may have altered release pattern of insulin and gut hormones, or
there is a reduced effect of gut hormones on gastric emptying and insulin secretion. The exact
mechanisms still warrant investigation.
2.5.3 Dietary Protein
The stimulating effect of protein on plasma insulin concentration was found in both
healthy (248) and type 2 diabetic subjects (249). However, compared to the healthy subjects, the
insulinotropic effect of protein was attenuated in people with type 2 diabetes (6). A possible
explanation may be the impaired insulinotropic effect of GIP (250), and/or reduced postprandial
secretion of GLP-1 in type 2 diabetes (251). The insulinotropic capacity of protein has seldom
been investigated in non-diabetic subjects with insulin resistance/hyperinsulinemia.
A recent study found that adding soy protein to glucose reduced glycemic response in
both healthy and hyperinsulinemic subjects, but 5g protein tended to have lesser effect on
reducing glycemic response in hyperinsulinemic than it had on healthy subjects (38), suggesting
that in hyperinsulinemic subjects, the insulinotropic effect of protein may be attenuated. Brand-
36
Miller et al studied the effect of protein (lean beef steak) in lean healthy Caucasian subjects, and
found a significant inverse relationship between insulin sensitivity and the extent of the decline
in plasma glucose (33). These results must be interpreted with caution because both had wide
scatter of data around the line of best fit (33, 38).
2.6 Mechanisms by Which Fat and Protein Modulate Postprandial Glycemia
It is generally thought that adding fat and protein to carbohydrate reduces glycemic
response through similar mechanisms such as delaying gastric emptying (35) and/or enhancing
insulin secretion through augmented GIP and GLP-1 secretion (36). However, there is emerging
evidence that fat and protein may modulate glucose homeostasis through different mechanisms.
A recent study in rats found that upper intestine lipids activate gut-brain-liver neuronal axis to
down-regulate hepatic glucose production and reduce plasma glucose concentration (39);
however, it is not known if protein has similar effect. In a mouse model, it was found that
glucose-induced incretin hormone release was differently modulated by fat (oleic acid) and
protein (whey protein) (37). Whey protein resulted in more GLP-1 secretion than oleic acid and
significantly increased GIP; however, oleic acid had no effect on GIP secretion. In addition,
whey protein deactivated the dipeptidyl peptidase-4 (the enzyme that inactivates GLP-1) activity,
whereas oleic acid doesn‘t have such an effect (37). The same study also found that insulin
response to oral glucose was 3-fold augmented by protein, and was associated with enhanced
oral glucose tolerance, whereas, the addition of fat resulted in only 1.5 fold increase in insulin
and there was no accelerated glucose disposal (37). This is consistent with the human study that
protein reduced glucose response 2-3 times more than fat, and there was no significant fat ×
protein interaction (38). Both the animal and human studies suggest that the effects of fat and
37
protein on postprandial metabolic responses are independent of each other, and the mechanisms
by which protein reduces glycemic responses may differ from those for fat.
2.7 The Effects of Habitual Diet on Postprandial Responses
Besides the macronutrient, the habitual dietary intake might also influence postprandial
hormonal and metabolic responses. There is evidence that in humans, a high fat diet potentiates
the effect of oral glucose on GIP secretion (252) and influences gastrointestinal transit (253,
254), suggesting that the acute glucose-lowering effects and other metabolic effects of fat may be
influenced by habitual fat intake.
Animal studies found that habitual dietary fiber intake up-regulates proglucagon
messenger RNA, increases GLP-1 concentration and reduces glucose level (255, 256). However,
in humans, the addition of the soluble fibre pysllium to a pasta meal did not alter gastric
emptying or GLP-1 concentrations (257). A low dose of 1.7g psyllium was used in this study,
which might explain the lack of effects (257).
A pilot study from our lab found that higher habitual dietary fiber intake amplifies the
effects of protein in lowering glycemic response (38). The mechanism that could explain such
effect remains unclear though it was suggested that fiber intake may be associated with increased
GLP-1 secretion (38). This hypothesis is consistent with the results of animal studies that higher
fiber diet up-regulates GLP-1 secretion, possible via short chain fatty acids, the products of the
fermentation of fiber in the colon (255). Recently, this effect was confirmed in humans. A study
in hyperinsulinemic subjects found that wheat fiber intake increased plasma butyrate and GLP-1
concentrations, although it takes 9-12 month for these changes to occur (258).
38
CHAPTER 3
RATIONALE, HYPOTHESES, AND OBJECTIVES
39
3.1 Rationale
To maintain glucose homeostasis, the normal functioning of the gastrointestinal tract
(liver, gut and pancreas), β-cell function, central nervous system, and entero-insulin axis is
essential. However, in people with diabetes, obesity, and insulin resistance, these physiological
systems may become abnormal, and the mechanisms by which macronutrient (carbohydrate, fat
and protein) modulate postprandial blood glucose and insulin responses may be altered as well.
Previous studies suggest that the effects of macronutrient on glucose and insulin responses differ
in healthy control and hyperinsulinemic subjects, and this discrepancy may be due to differences
in gut hormone secretion, pancreatic β-cell insulin secretion and/or hepatic insulin extraction.
Clarification of the mechanisms is important in lieu of the fact that insulin resistance/
hyperinsulinemia is at the early stage of the pathogenic pathway towards type 2 diabetes, and the
complex interactions between macronutrient and hyperinsulinemia at the cellular level remain
unclear. In addition, by testing GI and Insulinemic Index in ―pure‖ carbohydrate and
carbohydrate foods, this thesis may clarify the controversial issues surrounding the clinical utility
of GI and Insulinemic Index.
3.2 Overall Objective
The overall objective of this thesis is to determine whether macronutrient (carbohydrate,
fat and protein) elicit different effects on acute postprandial metabolic responses in healthy
control, hyperinsulinemic and type 2 diabetic subjects, and to explore potential mechanisms that
may explain any differences observed.
40
3.3 Specific Hypotheses
1. Adding fat and protein to carbohydrate results in less attenuation of glucose response in
non-diabetic, hyperinsulinemic subjects than in the healthy control subjects.
2. The effects of macronutrient on postprandial responses are affected by habitual intakes of fat
and dietary fibre.
3. The GI values of carbohydrate foods depend on differences in their relative rates of digestion
and absorption which do not differ in healthy control, hyperinsulinemic and diabetic subjects;
therefore, the GI values of carbohydrate foods will not differ in these subject groups. The GI
value of sucrose depends on the hepatic handling of fructose which might differ in these
subject groups.
4. The Insulinemic Index of carbohydrate foods depends not only on their relative glycemic
impact but also on β-cell function, hepatic insulin extraction and incretin hormone effects,
which differ in healthy control, hyperinsulinemic and type 2 diabetic subjects. Hence
Insulinemic Index of carbohydrate foods will differ in these subject groups.
5. Insulin sensitivity, β-cell function and hepatic insulin extraction are interconnected. GLP-1 as
an incretin hormone may play a role in these parameters. Therefore, the endogenous plasma
GLP-1 response to oral glucose is associated with insulin sensitivity, β-cell function and
hepatic insulin extraction, a triad important for glucose homeostasis.
In order to address the research questions above, two studies were carried out. The aim of
the first study was to determine the postprandial metabolic effects of adding fat (canola oil) and
protein (whey protein) to carbohydrate (50g oral glucose drink) in healthy control and
hyperinsulinemic subjects (Chapter 4); The aim of the second study was to determine the effects
of different sources of carbohydrate foods on postprandial glucose and insulin responses in
41
healthy control, hyperinsulinemic and type 2 diabetic subjects (Chapter 5). Since in both studies,
the oral glucose drink was repeated 3 times for each subject, data from the oral glucose drink was
pooled together to analyze if there were any associations between plasma active GLP-1
concentration and the indirect measures of insulin sensitivity, β-cell function and hepatic insulin
extraction (Chapter 6).
42
CHAPTER 4
THE HYPOGLYCEMIC EFFECT OF FAT AND PROTEIN IS NOT ATTENUATED BY
INSULIN RESISTANCE
A partial content of this chapter is published in American Journal of Clinical Nutrition (354)
43
4.1 Abstract
Background: The glucose-lowering effects of fat and protein are attenuated or absent in diabetic
patients, suggesting the same may occur in insulin resistant subjects without diabetes.
Objective: To determine whether the postprandial metabolic responses elicited by fat and protein
were influenced by insulin sensitivity of the subjects and to see whether fat and protein modulate
glucose response through different mechanisms.
Design: Overnight fasted, healthy, non-diabetic subjects aged 18-45 y, 9 with fasting serum
insulin (FSI) < 40pmol/L; 8 with 40 ≤ FSI < 70pmol/L; and 8 with FSI ≥ 70 pmol/L took 50g
oral glucose with 0-30g doses of canola oil and whey protein on 11 separate mornings. The
relative glycemic response (RGR) was expressed as incremental area under the curve (AUC)
after each test meal divided by the mean AUC of the 3 glucose controls in each subject.
Results: Protein significantly reduced glucose (p<0.0001), hepatic insulin extraction (p<0.0001),
and increased insulin (p<0.0001) and GLP-1 (p=0.004); however, protein had no significant
effect on C-peptide (p=0.69) and insulin secretion rate (p=0.13). Fat (30g) only significantly
increased GLP-1 response (p=0.025) and had a tendency to increase insulin (p=0.05). There
were no significant FSI × fat (p=0.19) and FSI × protein (p=0.08) interaction effects on glucose
AUC. In addition, the changes in RGR per gram of fat (r = -0.05, p = 0.82) or protein (r = -0.08,
p = 0.70) were not related to FSI.
Conclusions: The hypoglycemic effect of fat and protein was not blunted by insulin resistance.
Protein increased insulin but had no effect on C-peptide and insulin secretion rate, suggesting
reduced hepatic insulin extraction or increased C-peptide clearance. However, the exact
mechanism for whey protein-induced reduction in hepatic insulin extraction remains to be
examined.
44
4.2 Introduction
It is generally accepted that adding fat and protein to carbohydrate reduces blood glucose
compared with carbohydrate alone (32-34, 259); however, the glucose-lowering effect of fat is
attenuated or absent in people with diabetes (5, 147, 245). Unlike the effect of fat, the variations
in the effect of protein on glucose and insulin responses do not seem to be explained by subject
groups, because a variety of effects are seen in both healthy and diabetic subjects (147). The
most likely explanations seem to be the different protein source (152-154) and digestion kinetics
(149-151). Most of the studies on the metabolic effects of fat and protein focus on healthy and
diabetic subjects; very few studies have investigated how non-diabetic, insulin-resistant subjects
handle fat and protein.
In a pilot study, we investigated the hypoglycemic effect of fat (corn oil) and protein (soy
protein) in healthy and hyperinsulinemic humans, and found that higher fasting insulin was
associated with a lower glucose-lowering effect of fat (38). Brand-Miller et al studied the effect
of protein (lean beef steak) in lean healthy white subjects, and found a significant inverse
relationship between insulin sensitivity and the extent of the decline in plasma glucose (33).
These results must be interpreted with caution because they had a wide scatter of data around the
line of best fit.
The exact mechanisms associated with the hypoglycemic effect of fat and protein are not
clear, although they are suggested to occur through similar mechanisms, such as delayed gastric
emptying (35) and/or enhanced insulin secretion through augmented GLP-1 and GIP secretion
(36); however, a study in mice found that glucose-induced incretin hormone release was
differently modulated by fat and protein (37). The same study also found that the insulin
response to glucose was augmented 3-fold by protein and was associated with enhanced oral
glucose tolerance, whereas, the addition of fat resulted in only a 1.5-fold increase in insulin with
45
no accelerated glucose disposal (37). This is consistent with a human study that showed that
protein reduced glucose response 2 to 3 times more than fat, and there was no significant fat ×
protein interaction (38). Taken together, this evidence suggests that fat and protein may
modulate postprandial glucose response through different mechanisms.
Therefore, we investigated whether acute postprandial metabolic responses elicited by fat
and protein were influenced by the degree of insulin sensitivity of the subjects and examined
whether fat and protein modulate postprandial metabolic responses through different
mechanisms.
4.3 Subjects and Methods
4.3.1 Subjects and study design
Forty-six healthy, non-diabetic men and women (18-45 y old) were screened by
measuring height, weight, waist circumference (WC), hip circumference, blood pressure and
fasting serum glucose, insulin, C-reactive protein (CRP), liver enzymes, and creatinine (initial
recruitment date: April 30, 2007). The aim was to recruit 8 to 9 non-diabetic subjects in each of
the following categories of fasting serum insulin (FSI): low (FSI < 40 pmol/L), medium (40 ≤
FSI < 70 pmol/L) and high (FSI ≥ 70 pmol/L). The rationale of this method of recruitment was
based on the fact that FSI is strongly correlated with insulin resistance measured by euglycemic-
hyperinsulinemic clamp (260), and the 40 pmol/L cut-off point was chosen because this
represents approximately the 67th
percentile for non-diabetic subjects in our laboratory (244). In
our preliminary study, in which FSI>40 pmol/L was used to define hyperinsulinemia (38),
subjects with FSI>40 pmol/L had significantly greater WC, BMI, HOMA insulin resistance
index, total and LDL cholesterol and triglycerides, and lower HDL cholesterol than those with
46
FSI<40pmol/L (38). We divided subjects with FSI>40pmol/L into medium (FSI<70 pmol/L)
and high (FSI ≥ 70 pmol/L) groups because people with newly diagnosed diabetes have mean
FSI>70 pmol/L (261, 262). Two subjects from the high-FSI group had impaired glucose
tolerance (IGT) (2-h postprandial glucose: 7.8-11.1mmol/L) (1). Data from the 2 IGT subjects
was still used for analysis because even after excluding the 2 IGT subjects, the results did not
change. Subjects were not eligible if they had BMI>35kg/m², type 2 diabetes (fasting glucose ≥
7.0mmol/L), serum triglycerides ≥ 10.0mmol/L, impaired renal function (serum creatinine >1.2
times upper limit of normal), or liver function (aspartate transaminase, alanine transaminase, or
gamma-glutamyl transpeptidase >2 times upper limit of normal). Other exclusion criteria were:
smoking, use of diuretics, corticosteroids or β-blockers, or presence of medical events or chronic
conditions influencing gastrointestinal function or glucose metabolism.
Twenty-five eligible subjects were enrolled in the study; n=9 with a low FSI, n=8 with a
medium FSI, and n=8 with a high FSI (the screened subjects were excluded because we only
needed certain number of people with low, medium or high FSI). The research protocol was
reviewed and approved by Research Ethics Boards at the University of Toronto and St.Michael‘s
Hospital. All subjects gave written informed consent.
The subjects were instructed to maintain their usual daily routine between study days and
refrain from exercise on the morning of the test. After an overnight fast (10-14h), they came to
the Risk Factor Modification Center at St. Michael‘s hospital on 11 separate mornings (over a
three month period) between 7:30 and 9:30. Venous blood samples were drawn before and at 15,
30, 45, 60, 90 and 120min after starting to ingest test meals.
47
4.3.2 Test drinks
Nine test drinks, consisting of 50g anhydrous glucose (Grain Processing Enterprising
LTD, Scarborough, On, Canada), 250ml water, 0g, 5g or 30g fat (Canola oil, Sardo Foods,
Brampton, Ontario, Canada), and 0g, 5g or 30g protein (whey protein concentrate 392, Fonterra
Inc. Camp Hill, PA, USA), were blended to smooth drink prior to consumption. The
composition and caloric content of the test meals are shown in Table 4.1. The test meals are not
calorically equal meals, with 50g glucose + 50g fat + 50g protein containing the highest calorie
(Table 4.1). The test meals were administered in a random order. The choices of 5g and 30g of
fat or protein are based on the results of preliminary study (38) that the difference in the effect of
fat on glycemic responses between healthy and hyperinsulinemic subjects (FSI>40pmol/L) was
greatest at the highest level of fat (30g). In contrast, a difference in the effect of protein was
evident only at the lowest dose (5g). The control drink was 50g glucose alone, and was tested on
each subject three times at the beginning, middle and end of the study. This is because relative
glycemic response (RGR) and relative insulin response (RIR) to the test meals were calculated as
AUC after each test meal divided by the mean AUC of the 3 glucose controls in each subject.
The use of the average AUC of the 3 oral glucose tests in each subject results in a greater
probability that the glucose is representative of the subject‘s true response to oral glucose than a
single determination of AUC. In addition, use of the mean of the 3 glucose control tests reduces
within-subject variability and increases the power of the study (263).
4.3.3 Assessment of habitual dietary intake
Habitual nutrient intakes were estimated using two 3-day food records (at the beginning
and end of the study period), which will correctly classify 70-80% of subjects in the top or
bottom tertile of intake for percentage (%) energy from fat, carbohydrate and dietary fibre (264),
48
the main nutrients of interest. Nutrient analysis was done using Food Processor SQL edition,
version 9.3.1m (ESHA Research, Salem, OR).
4.3.4 Assessment of physical activity
Because exercise improves insulin sensitivity (265, 266), the subjects‘ physical activities
including work-, sports-, and leisure-time were assessed using Baecke Physical Activity
Questionnaire (267), which was administered at the beginning and the end of the study.
4.3.5 Blood analysis
Venous blood samples for the measures of glucose, insulin, and C-peptide were collected
in BD VacutainerTM
SSTTM
tubes (BD, Franklin Lakes, NJ, USA). Serum glucose was measured
by a glucose oxidase method (SYNCHRON LX Systems, Beckman Coulter, Brea, California,
USA), with inter-assay coefficient of variation (CV) of 1.9%. Insulin was measured using one-
step immunoenzymatic (―sandwich‖) assay (Beckman Access Ultrasensitive Insulin Assay,
Beckman Coulter, Brea, California, USA), with inter-assay CV of 2.5 to 4.3 %. Insulin has no
cross-reactivity with proinsulin. C-peptide was measured using double antibody competitive
radioimmunoassay (Siemens Medical Solutions Diagnostics, Los Angeles, CA), with inter-assay
precision of 10% or less.
Venous blood samples for GLP-1 were collected in BD VacutainerTM
EDTATM
tubes
(BD, Franklin Lakes, NJ, USA). The dipeptidyl peptidase-4 inhibitor (Linco Research, St.
Charles, Missouri) was added immediately (less than 30 seconds) after collection. Plasma GLP-
1 was measured by capturing the active GLP-1 from the sample by a monoclonal antibody
(specific binding to the N-terminal region) using GLP-1 (active) ELISA Kit (Linco Research, St.
49
Charles, Missouri, USA), with inter-assay CV ranges from 1-13%. All the samples were stored
at -70°C before analysis.
4.3.6 Calculation and statistical analysis
Incremental areas under the curve (AUC), ignoring area below baseline, for glucose,
insulin, C-peptide and GLP-1 were calculated using the trapezoid rule (31). The net AUC,
including the areas falling below the baseline was also calculated using the trapezoid rule (31).
The endpoints analyzed from the AUC and net AUC were similar; therefore, only results from
AUC were presented. Total AUCins/glu was calculated using the trapezoidal rule applied to the
insulin and glucose curves from 0 to 120 mins. The same method was used to calculate total
AUC of insulin secretion rate (ISR). Basal hepatic insulin extraction (HIEbasal) was calculated as
fasting C-peptide divided by fasting insulin and postprandial hepatic insulin extraction (HIEauc)
was determined by the AUC of C-peptide divided by that of insulin (268). Insulin secretion rate
was calculated from deconvolution of the plasma C-peptide concentration using ISEC software
package developed by Hovorka et al (269). The relative glycemic response (RGR) and relative
insulin response (RIR) to the test meals were calculated as AUC after each test meal divided by
the mean AUC of the 3 glucose controls in each subject.
The means of 3 OGTT were used to calculate insulin sensitivity index. The validated
insulin sensitivity index of Matsuda and DeFronzo (173) was used as an index of whole-body
insulin sensitivity; it was calculated as 10,000/(fasting glucose × fasting insulin × mean glucose
× mean insulin )0.5
(173). Mari‘s oral glucose insulin sensitivity index, a glucose-insulin model
constructed on established principles of glucose kinetics and insulin action was also calculated
(174). The oral glucose insulin sensitivity has been validated against the euglycemic-
hyperinsulinemic clamp in healthy, obese and type 2 diabetic subjects (174).
50
The hepatic insulin resistance was calculated as the product of total AUC of glucose and
insulin during the first 30 minutes (glucose0-30[AUC] x insulin0-30[AUC]) (172) and muscle
insulin sensitivity was calculated as the rate of decay of plasma glucose concentration from its
peak value to its nadir divided by the mean insulin concentration (172). Both indices of hepatic
insulin resistance and muscle insulin sensitivity derived from OGTT have been validated using
hyperinsulinemic-euglycemic clamp (172).
The β-cell compensation for insulin resistance was estimated using the insulin
secretion/insulin resistance (disposition) index derived from OGTT (270), which was shown to
be the best predictor of future development of type 2 diabetes in subjects with normal glucose
tolerance (NGT) compared with other predictive models such as San Antonio Diabetes
Prediction Model (271) (including age, sex, ethnicity, BMI, blood pressure, fasting plasma
glucose, triglycerides, and HDL) and 2-hour plasma glucose concentration. To calculate insulin
secretion/insulin resistance (disposition) index, first, insulin secretion was expressed as the
changes in AUC of insulin secretion rate (∆ISRAUC) relative to changes in AUC of plasma
glucose response (∆ISRAUC/∆GAUC ), then ∆ISRAUC/∆GAUC was divided by the severity of insulin
resistance (∆ISRAUC/∆GAUC÷ IR), as measured by the inverse of Matsuda‘s insulin sensitivity
index.
Data were expressed as mean ± SEM for normally distributed variables or median
(interquartile range) for non-normally distributed variables. Normality was assessed using the
Shapiro and Wilk statistic and the normality plots (PROC UNIVARIATE procedure of SAS).
Skewed variables were log-transformed prior to analysis. The AUCs of the biomarkers (glucose,
insulin, C-peptide, insulin secretion rate, GLP-1) and hepatic insulin extraction were subjected to
the repeated measures of analysis of variance (ANOVA) (PROC MIXED) to test for the main
51
effects of FSI (low vs. medium vs. high), fat dose, protein dose and fat × protein, FSI × fat, FSI ×
protein, and FSI × fat × protein interactions. Since the design is not balanced and the sample
size is relatively small, the correlation of FSI× fat ×protein term with FSI ×fat and FSI × protein
terms might obscure the significance of an effect; therefore, the FSI × fat × protein term was
omitted when it was not significant, and the reduced model was reanalyzed to see if there were
any significant FSI × fat or FSI × protein interaction effects. Age and BMI were included in the
model as covariates to control for the differences in these variables between subjects. Tukey‘s
post hoc test was performed to compare treatments if the treatment effects were statistically
significant.
The effect of fat or protein on relative glycemic response and relative insulin response in
each subject was defined as the slope of the regression line of relative glycemic response or
relative insulin response on doses of fat (or protein). The regression equations were based on 9
points in each subject (3 × 3 levels of fat and protein). Before pooling the data for each subject
to calculate the slopes, the relative glycemic response and relative insulin response were
subjected to the repeated measures ANOVA to test whether there were significant fat × protein
and FSI × fat × protein interaction. This is because it is only valid to pool data in this way if
there was no significant fat × protein or FSI × fat × protein interaction and no significant
evidence of a nonlinear dose-response relation. The results showed no significant fat × protein
(relative glycemic response, p=0.72; relative insulin response, p=0.55) or FSI × fat × protein
interaction (relative glycemic response, p=0.47; relative insulin response, p=0.20). The
correlations between changes in relative glycemic response and relative insulin response per g of
fat (or protein) and the variables (FSI, habitual dietary fat and fiber intakes) were determined by
52
simple linear regressions. All analyses were done using SAS 9.1, (SAS Institute Inc, Cary).
Differences were considered significant if 2-tailed p<0.05.
4.4 Results
The anthropometric and metabolic characteristics of subjects with a low, medium or high
FSI are shown in Table 4.2. Compared with those with a low- or medium-FSI, subjects with a
high-FSI were older and had a significantly higher waist circumference, waist-to-hip ratio,
systolic & diastolic blood pressure and total cholesterol, triglycerides and LDL cholesterol
(Table 4.2). Subjects with a high-FSI also had a significantly higher BMI and total:HDL
cholesterol ratio than did those with a low-FSI. Fasting glucose, HDL cholesterol and C-reactive
protein were not significantly different between the 3 FSI groups.
Both HIEbasal and postprandial HIEauc in those with a medium- or high-FSI were
significantly lower than those of the low-FSI (Table 4.2). In fact, a significant inverse relation
was observed between FSI and basal (r=-0.70, p<0.0001) and postprandial hepatic insulin
extraction (r=-0.72, p<0.0001) (Appendix 1). In addition, postprandial insulin response was also
inversely associated with postprandial HIEauc (r=-0.80, p<0.0001) (Appendix 1). These results
are consistent with Meier et al‘s observation that plasma insulin concentration is inversely
related to liver insulin clearance (208). Higher insulin concentrations at both basal and
postprandial were associated with lower liver insulin clearance, which conserves insulin and
contributes to elevated insulin concentrations.
Mari‘s oral glucose insulin sensitivity, Matsuda‘s whole body insulin sensitivity indices
(ISI) and muscle insulin sensitivity index decreased in a step-wise fashion across the groups;
subjects with a high-FSI were the least insulin sensitive (Table 4.2). Subjects with a medium- or
high-FSI had a significantly higher hepatic insulin resistance than did those with a low-FSI
53
(Table 4.2). In fact, an inverse association was observed between hepatic insulin resistance and
hepatic insulin extraction (r=-0.61, p=0.001) (Appendix 2), showing that insulin clearance is
inextricably linked to hepatic insulin action. The more insulin resistant the liver becomes, the
lower the liver insulin clearance.
Insulin secretion/insulin resistance (disposition) index (∆ISRAUC/∆GAUC ÷ IR) was
significantly lower in subjects with a high-FSI than those with a low- or medium-FSI (Table
4.2). Subjects with a medium-FSI had lower disposition index than those with a low-FSI, albeit
it was not significant (Table 4.2).
Neither the individual physical activity indices (work-, sports-, and leisure-time) nor were
the mean physical activity indices differ significantly among the subject groups (Appendix 3).
This lack of significant difference may be attributable to the fact that all the subjects were
relatively young (18-45y) and healthy.
The mean AUC of glucose, insulin, C-peptide, and GLP-1 responses after 3 separate
OGTTs and their inter- and intra-subject coefficients of variations (CV) are shown in Table 4.3.
As the subjects‘ fasting insulin increased, the AUC of glucose and insulin increased. There was
also a tendency towards an increased C-peptide AUC with increasing fasting insulin, although it
was not significant (p=0.06). The AUC of GLP-1 was not significantly different among the three
subject groups (p=0.995). In general, for all the biomarkers, the inter-subject variation was
higher than intra-subject variation in the 3 FSI groups (Table 4.3).
P values for the main and interaction effects of FSI group, fat, and protein on the AUCs
of the biomarkers (glucose, insulin, C-peptide, insulin secretion rate, and GLP-1) and hepatic
insulin extraction are shown in Table 4.4. Fat had a significant main effect only on the GLP-1
response. Protein significantly influenced glucose, insulin, GLP-1 responses, and hepatic insulin
54
extraction. There were significant FSI group effects on glucose, insulin, the insulin secretion
rate, GLP-1 and hepatic insulin extraction. There was no fat × protein interaction effect on any
of the biomarkers. The FSI × fat interaction effect was observed only for the insulin response,
which suggested that the ability of fat to influence the insulin response depended on FSI of the
subjects. There were significant FSI × protein interaction effects on insulin, GLP-1, and hepatic
insulin extraction, which suggested that the effects of protein on these biomarkers depended on
FSI of the subjects. The FSI × fat × protein interaction effects were only observed on insulin and
hepatic insulin extraction, which suggested that the effect of fat and protein on insulin and
hepatic insulin extraction depended on FSI of the subjects.
The 2-hour responses of glucose, insulin, C-peptide, and GLP-1 after 50g oral glucose
plus 0, 5g or 30g protein in non-diabetic humans with a low, medium or high FSI are shown in
Figure 4.1. In the 3 FSI groups, 30g protein significantly reduced glucose between 30 to 90 min
compared to 0 or 5g protein. There was no protein dose effect on insulin in low-FSI group;
however, 30g protein significantly increased insulin at 30 min in the medium-FSI group and at
30 and 45 min in the high-FSI group. Protein had no significant effect on C-peptide and GLP-1
at any of the time points or in any of the FSI groups. The 2-hour postprandial responses
(glucose, insulin, C-peptide, and GLP-1) after 50g oral glucose plus 0, 5g or 30g fat showed a
trend similar to that of protein, albeit not significant (Appendix 4).
The main and interaction effects of FSI group, fat, and protein on the AUCs of the
biomarkers (glucose, insulin, C-peptide, insulin secretion rate, and GLP-1) and hepatic insulin
extraction are illustrated in Figure 4.2 and Figure 4.3. The main effects of fat or protein on
glucose, insulin, C-peptide, insulin secretion rate, and GLP-1 responses expressed as AUC and
hepatic insulin extraction are shown in Figure 4.2. Protein significantly decreased glucose
55
(p<0.0001), increased insulin (p<0.0001), and had no significant effect on C-peptide (p=0.69)
and insulin secretion rate (p=0.13); however, protein significantly decreased hepatic insulin
extraction (p<0.0001) and increased GLP-1 (p=0.004). Fat (30g) significantly increased only the
GLP-1 response (p=0.025) and had a tendency to increase insulin (p=0.05).
The main and interaction effects of FSI group, fat, and protein on glucose, insulin, C-
peptide, the insulin secretion rate, and hepatic insulin extraction shown in Figure 4.3 are
consistent with those shown in Table 4.4. Fat had no significant effect on glucose, insulin, C-
peptide, insulin secretion rate, or hepatic insulin extraction in any of the FSI groups; however,
there was a significant FSI × fat interaction effect on the insulin response (Figure 4.3). For
instance, insulin responses to 30g fat intake were significantly higher in the high-FSI group than
in the low- and medium- FSI groups. Unlike fat, 30 g protein significantly decreased the glucose
response in all FSI groups, to a similar magnitude. Protein at 30g significantly increased the
insulin responses in the medium- and high-FSI groups, but exerted no effect on C-peptide or the
insulin secretion rate; however, 30g protein significantly decreased hepatic insulin extraction in
the low- and medium-FSI groups, but had no effect on the high-FSI group. This may have been
because the high-FSI group already had a reduced hepatic insulin extraction (Table 4.2).
Significant FSI ×protein interaction effects were shown for the insulin response and hepatic
insulin extraction: with the 30g protein intake, the insulin response was significantly higher in
the high-FSI group than in the low-FSI group, and hepatic insulin extraction was significantly
lower in the medium- and high-FSI groups than in the low-FSI.
The changes in relative glycemic response per gram of fat (r = -0.05, p = 0.82) or protein
(r = -0.08, p = 0.70) were not related to FSI (Appendix 5). Similarly, the changes in relative
56
insulin response per gram of fat (r = -0.06, p = 0.77) or protein (r =0.04, p =0.86) were also not
related to FSI (Appendix 5).
For habitual dietary fat and fiber intake, the changes in relative glycemic response per
gram of fat (r = -0.12, p = 0.58) or protein (r = 0.06, p = 0.76) were not related to habitual dietary
fat intake. Similarly, the changes in relative insulin response per gram of fat (r = -0.03, p = 0.89)
or protein (r =-0.12, p =0.58) were also not related to habitual dietary fat intake. Contrary to
previous finding (38), the changes in relative glycemic response per gram of protein (r = 0.29, p
= 0.16) was not related to habitual dietary fiber intake.
57
Table 4.1 Composition of the 9 test meals, their total energy content, and the percentage of
energy from carbohydrate, fat, and protein1.
Test meals Total energy Carbohydrate
Fat Protein
(Calories) % of energy % of energy % of energy
50g glucose + 0g fat + 0g protein (control) 200 100 0.0 0.0
50g glucose + 0g fat + 5g protein 220 90.9 0.0 9.1
50g glucose + 0g fat + 30g protein 320 62.5 0.0 37.5
50g glucose + 5g fat + 0g protein 245 81.6 18.4 0.0
50g glucose + 5g fat + 5g protein 265 75.5 17.0 7.5
50g glucose + 5g fat + 30g protein 365 54.8 12.3 32.9
50g glucose + 30g fat + 0g protein 470 42.6 57.4 0.0
50g glucose + 30g fat + 5g protein 490 40.8 55.1 4.1
50g glucose +30g fat +30g protein 590 33.9 45.8 20.3
1The test meals were randomly administered. The control glucose drink was tested at the
beginning, middle, and end of the study.
58
Table 4.2 Anthropometric and metabolic characteristics of subjects by fasting serum insulin1
Low
(FSI<40pmol/L)
Medium
(40 ≤ FSI <70pmol/L)
High
(FSI ≥ 70 pmol/L)
P2
Age (y) 27±3a,3
26±2a 38±2
b 0.003
Ethnicity (n) (C:SA:EA:Af:ME:Mix)4 6:0:2:0:0:1 2:2:2:1:1:0 3:4:0:1:0:0
BMI (kg/m2) 24±1.1
a 27±1.5
ab 31±1.0
b 0.001
Waist circumference (cm) 82±3.3a 85±5.1
a 102±3.4
b 0.003
Waist-to-hip ratio 0.79±0.02a 0.8±0.03
a 0.91±0.03
b 0.01
Fasting insulin (pmol/L) 22(21-30)a,5
53(50-54)b 88(76-123)
c <0.0001
Fasting glucose (mmol/L) 4.6±0.1 5.0±0.1 5.0±0.2 0.05
HIEbasal6 16.2±0.6
a 12.6±0.5
b 11.8±0.5
b <0.0001
HIEauc6 6.6±0.4
a 4.3±0.5
b 2.8±0.3
b <0.0001
Mari OGIS (ml/min/m2) 6 507±7
a 458±10
b 377±10
c <0.0001
Matsuda Whole-body ISI 6 36.7(25.1-41.8)
a 18.8(15.5-20.5)
b 9.5(8.3-10.9)
c <0.0001
Muscle ISI6 3.7(3.6-6.3)
a 1.7(1.4-2.0)
b 0.8(0.6-1.8)
b 0.003
Hepatic insulin resistance index6 2.9(2.3-4.2)
a 7.4(4.4-9.9)
b 6.7(5.2-11.5)
b 0.01
∆ISRAUC/∆GAUC ÷ IR6 102±8
a 81±12
a 27±3
b 0.008
Systolic BP (mm Hg) 107±3a 102±4
a 127±5
b 0.001
Diastolic BP (mm Hg) 64±2a 62±3
a 85±3
b <0.0001
Total cholesterol (mmol/L) 4.1±0.2a 4±0.3
a 5.5±0.2
b 0.001
Triglycerides (mmol/L) 0.77(0.5-1)a 0.78(0.5-1.23)
a 1.72(1.59-1.93)
b 0.0004
HDL cholesterol (mmol/L) 1.3±0.1 1.0±0.1 1.1±0.2 0.18
LDL cholesterol (mmol/L) 2.4±0.2a 2.6±0.2
a 3.5±0.3
b 0.01
59
Table 4.2 (Continued)
Low
(FSI<40pmol/L)
Medium
(40 ≤ FSI <70pmol/L)
High
(FSI ≥ 70 pmol/L)
P2
Total: HDL cholesterol 3.2±0.2a 4.2±0.5
ab 5.4±0.6
b 0.01
C-reactive protein (mg/L) 0.33(0.19-1.21) 0.4(0.19-4.1) 2.91(1.08-5.81) 0.13
1 n = 25 except where otherwise noted. FSI, fasting serum insulin; HIE, hepatic insulin
extraction; OGIS, oral glucose insulin sensitivity; ISI, insulin sensitivity index; AUC, area under
the curve; IR, insulin resistance; BP, blood pressure. Values in the same row with different
superscript letters are significantly different, P<0.05 (Tukey‘s post hoc test).
2 P represents overall significant differences across groups by one-factor ANOVA. The p-values
for HIEbasal, HIEauc, OGIS, ISI and disposition index were derived from repeated measures
ANOVA.
3 Mean±SEM (all such values).
4 C. Caucasian; SA, South Asian; EA, East Asian; Af, African; ME, Middle Eastern.
5 Median; interquartile range in parentheses (all such values).
6 Calculated by using the mean of 3 oral-glucose-tolerance test results. PROC MIXED repeated
measures ANOVA, adjusted for age, BMI and WC.
60
Table 4.3 Mean areas under the curve (AUCs) for glucose, insulin, C-peptide, and glucagon-
like peptide 1 (GLP-1) responses after 3 separate oral-glucose-tolerance tests and their inter- and
intra-subject CVs by fasting serum insulin (FSI)1
Low
(FSI<40pmol/L)
Medium
(40 ≤ FSI < 70pmol/L)
High
(FSI ≥ 70 pmol/L)
P2
Glucose AUC (mmol×min/L) 206(184-228)a,3
197(137-236)a
322(257-377)b
0.004
Intersubject CV 35 45 39
Intrasubject CV 35 32 21
Insulin AUC (nmol×min/L) 13(11-17)a 34(27-39)
b 45(35-79)
b 0.001
Intersubject CV 60 47 50
Intrasubject CV 29 27 30
C-peptide AUC (nmol×min /L) 99±114 131±16 150±16 0.06
Intersubject CV 46 47 63
Intrasubject CV 37 40 64
GLP-1 AUC (pmol×min /L) 182(83-285) 177(75-352) 156(79-235) 0.995
Intersubject CV 99 83 116
Intrasubject CV 66 51 55
1 n = 25. Values in the same row with different superscript letters are significantly different, P <
0.05 (Tukey‘s post hoc test).
2 P represents overall significant differences across groups by one-factor ANOVA.
3 Median; interquartile range in parentheses (all such values).
4 Mean ± SEM (all such values).
61
Table 4.4 P values for main and interaction effects of fasting serum insulin (FSI) group, fat, and
protein on glucose, insulin, C-peptide, insulin secretion rate (ISR), and
glucagon-like peptide 1 (GLP-1) responses expressed as area under the curve (AUC) and hepatic
insulin extraction (HIE)1
Glucose
Insulin
C-peptide
ISR
GLP-1
HIE
Main effects
Fat 0.53 0.05 0.33 0.16 0.025 0.21
Protein <0.0001 <0.0001 0.69 0.13 0.004 <0.0001
FSI2 0.02 <0.0001 0.08 0.03 0.009 0.001
Interactions
Fat × Protein 0.50 0.06 0.49 0.18 0.18 0.28
FSI × Fat 0.19 0.02 0.61 0.53 0.12 0.32
FSI × Protein 0.08 0.02 0.69 0.48 0.03 0.04
FSI × Fat × Protein 0.04 0.01
1 P values were derived from repeated-measures ANOVA (PROC MIXED; SAS Institute, Cary,
NC); P < 0.05 indicates significance. Age and BMI were included in the model as covariates.
2 Low compared with medium compared with high.
62
Low FSI
2
4
6
8
10
* *
0g protein
5g protein
30g protein
Glu
cose
(mm
ol/L
)Medium FSI
* * *
High FSI
* * *
0
400
800
1200
Insu
lin
(pm
ol/L
) * **
1000
2000
3000
4000
C-p
eptid
e
(pm
ol/L
)
0 30 60 90 120
8
12
16
GL
P-1
(pm
ol/L
)
0 30 60 90 120 0 30 60 90 120
Time (min)
Figure 4.1 Mean (±SEM) 2-h postprandial plasma glucose, insulin, C-peptide, and glucagon-like
peptide 1 (GLP-1) concentrations after 50 g oral glucose plus 0, 5, and 30 g protein in
nondiabetic humans with different concentrations of fasting serum insulin (FSI): Low (FSI < 40
pmol/L), Medium (40 ≤ FSI < 70 pmol/L), and High (FSI ≥ 70 pmol/L). *Significantly different
from 0 and 5 g protein, P < 0.05 (2-factor ANOVA). n = 25. The error bars are not shown if they
overlap or are smaller than the symbol.
63
80
130
180
230
280
330
aa
a
AA
B
Fat
Protein
Glu
cose A
UC
(mm
ol x m
in/L
)
0 5 10 15 20 25 30
2
3
4
5 A
BB
aa a
HIE
30
40
50
60
A
B
C
aa
aInsulin
AU
C
(nm
ol x m
in/L
)
500
600
700
800 A
A A
a a
a
ISR
AU
C
(pm
ol/kg)
100
110
120
130A
AAa
a
a
C-p
eptide A
UC
(nm
ol x m
in/L
)
0 5 10 15 20 25 304.0
4.5
5.0
5.5 B
AA
b
aa
Log(G
LP
-1 A
UC
)
(pm
ol x m
in/L
)
Grams of Fat or Protein Added
Figure 4.2 Mean (±SEM) main effects of fat and protein on glucose, insulin, C-peptide, insulin
secretion rate (ISR), and glucagon-like peptide 1 (GLP-1) responses expressed as area under the
curve (AUC) and hepatic insulin extraction (HIE) in nondiabetic humans by combined fasting
serum insulin (FSI) groups: low (FSI < 40 pmol/L), medium (40 ≤ FSI <70 pmol/L), and high
(FSI ≥ 70 pmol/L). Means not sharing a common letter (a and b are for the comparisons of
different levels of fat and A and B are for the comparison of different levels of protein) are
significantly different [repeated-measures ANOVA, PROC MIXED procedure (SAS Institute,
64
Cary, NC)]. Age and BMI were included in the model as covariates (Tukey‘s post hoc test, P <
0.05). n = 25. The error bars are not shown if they overlap or are smaller than the symbol.
65
0 5 30 0 5 30 0 5 300
100
200
300
ababab
a abab
abab
b
Glu
co
se
AU
C
(mm
ol x m
in/L
)
0 5 30 0 5 30 0 5 30
aac a ac
bb
c c
ab
0 5 30 0 5 30 0 5 300
20
40
60
80
100
a a a
ab abbd
bccd
cIn
su
lin A
UC
(nm
ol x m
in/L
)
0 5 30 0 5 30 0 5 30
a aa
aebe
bcd be
c
d
0 5 30 0 5 30 0 5 300
50
100
150
200
C-p
ep
tid
e A
UC
(mm
ol x m
in/L
)
0 5 30 0 5 30 0 5 30
0 5 30 0 5 30 0 5 300
2
4
6
8
ababab
c bcbcc
cc
HIE
0 5 30 0 5 30 0 5 30
a
bd
ca
bdcd
bd
c
d
g of fat added g of protein added
0 5 30 0 5 30 0 5 300
200
400
600
800
1000
ab
b
ab
a ab
ab abab ab
ISR
AU
C
(pm
ol/kg)
0 5 30 0 5 30 0 5 30
ab
b
a
ab
ab
b
abab ab
Figure 4.3 Mean (±SEM) effects of 50 g glucose plus 0, 5, or 30 g fat or protein on glucose,
insulin, C-peptide, the insulin secretion rate (ISR) expressed as the area under the curve (AUC),
and hepatic insulin extraction (HIE) in healthy nondiabetic subjects with different levels of
fasting serum insulin (FSI): low (FSI < 40 pmol/L, white bars), medium (40 ≤ FSI < 70 pmol/L,
66
grey bars), and high (FSI ≥ 70 pmol/L, black bars). Means not sharing a common letter are
significantly different [repeated measures ANOVA, PROC MIXED procedure (SAS Institute,
Cary, NC)]. Age and BMI were included in the model as covariates (Tukey‘s post hoc test, P <
0.05). n = 25.
67
4.5 Discussion and conclusions
We found that the ability of fat and protein to lower glucose responses was not influenced
by the degree of insulin sensitivity of the subjects. Protein significantly decreased the glucose
response regardless of whether the subjects had a low, medium or high FSI. We also found that
protein increased the insulin response but had no effect on the C-peptide response or insulin
secretion rate, which may be explained by either a decreased insulin clearance or increased rate
of C-peptide clearance.
Previous studies suggest that the ability of fat (38) and protein (33) to lower glucose
response was attenuated in subjects with insulin resistance. In these studies, the correlation
between measures of insulin sensitivity and the ability of fat or protein to lower the glucose
response was weak because of the wide scatter of data around the line of best fit. Nevertheless, a
few possible reasons may contribute to the discrepancy between our study and previous studies.
First, different sources of fat and protein were used. Moghaddam et al (38) used corn oil and soy
protein and Brand-Miller et al (33) used lean beef steak, whereas we used canola oil and whey-
protein. It is known that the fat source (272, 273) and protein source (274-277) influence
postprandial glucose responses differently. In particular, whey-protein is considered a ―fast‖
protein, which is digested and absorbed rapidly (278-282). Second, Moghaddam et al (38) used
capillary blood for the measurement of glucose, whereas we used venous blood. The glucose
responses measured in venous plasma were shown to be lower and more variable than those in
capillary blood (283). Finally, the variations may be due to the type of subjects studied. Brand-
Miller et al (33) studied lean healthy male white subjects with very small differences in insulin
sensitivity (assessed by euglycemic-hyperinsulinemic clamp), whereas we studied multiethnic
subjects with a much wider range of insulin sensitivity, as evidenced by their significantly
different oral glucose insulin sensitivity and whole-body insulin sensitivity (Table 4.2).
68
We found that protein decreased the glucose response, whereas fat had no such effect.
This was unexpected because, usually, when fat is added to carbohydrate, it results in a lower
glucose response (32, 147, 246). This may be due to the difference between solid and liquid
meals — if fat delays gastric emptying, it may have more of an effect with solid than with liquid
meals. Another reason is that the increase in insulin after fat intake may be too small to result in
a decreased glucose response (Figure 4.2). It is also possible that fat and protein may
differentially modulate the release of cholecystokinin (CCK) and peptide YY (PYY), the 2 gut
hormones that delay gastric emptying (284, 285). It is known that the rate of gastric emptying is
a major determinant of postprandial glycemia; even modest changes may have a substantial
effect on the magnitude of postprandial increases in glucose and insulin (286, 287). Although
CCK and PYY were not measured in this study, a previous study showed that protein is more
effective than fat at delaying gastric emptying (37), and, in normal-weight and obese humans,
protein intake induces a greater release of PYY than does fat (288).
Gram per gram, the insulin-raising effect of protein was 2 times that of fat. This may be
because hepatic insulin extraction was not affected by fat, whereas it was significantly decreased
with protein intake (Figure 4.2). It is also possible that fat and protein may differentially
modulate incretin hormone release and inactivation. We found that both 30g fat and protein
significantly increased the GLP-1 response; however, we do not know whether GIP increased in
a similar fashion. We did not measure GIP, but it is known that whey protein is a potent
stimulator of GIP release (289), and it was previously shown that the early (30 min) GIP
response was higher after protein ingestion than after fat ingestion (290). In terms of incretin
inactivation, a rodent model showed that whey protein inhibits proximal small intestine
dipeptidyl peptidase-4 activity (the enzyme deactivates GLP-1), thus prolonging the action of
69
GLP-1, whereas fat has no such effect (37). It is not known whether such effect also exists in the
small intestine of humans.
We found that adding whey protein to 50g oral glucose had no effect on C-peptide or the
insulin secretion rate, whereas a previous study showed that feeding protein (beef) alone
increased the C-peptide response (291). This discrepancy is unlikely because of a slower gastric
emptying time and lower glycemic response to oral glucose because the glucose response curves
were identical over the first 15 minutes, regardless of whether whey protein was added to oral
glucose or not (Figure 4.1). Nevertheless, why did whey protein dose-dependently increase the
insulin concentration but have no effect on C-peptide or insulin secretion rate? The elevation of
peripheral insulin concentrations after meal ingestion is due to both the stimulation of pancreatic
β-cell insulin secretion and a reduction in hepatic insulin clearance (201, 202); therefore, one
possible mechanism is that the increased insulin after whey protein may be due to reduced liver
insulin clearance. Indeed, our data show that, although whey protein had no effect on the insulin
secretion rate, it dose-dependently reduced hepatic insulin extraction (Figure 4.2). A previous
study described a close relation between whey protein induced insulin response and a
concomitant increase in postprandial amino acids (277), and leucine, isoleucine, valine, lysine
and threonine are the major amino acids responsible for increased insulinemia (289). Beside
amino acids, milk proteins are a highly abundant source of bioactive peptides (292). We
postulate that certain branched-chain amino acids (i.e. leucine, isoleucine, valine, lysine and
threonine) and/or whey-derived bioactive peptides (Appendix 7) may act as competitive
inhibitors and that they compete with insulin by binding to insulin receptors on hepatocytes; thus
less insulin is cleared by the liver, and peripheral insulin concentrations increase. The other
possible mechanism for the unaltered C-peptide and insulin secretion rate is that whey protein
70
may affect the C-peptide metabolic clearance rate. C-peptide is susceptible to renal uptake, with
>85% the amount primarily metabolized by the kidney and only a small proportion of C-peptide
being excreted intact in urine (293). It is known that protein intake increases urea nitrogen
production, and a strong positive correlation was observed between the changes in urea nitrogen
excretion and the changes in creatinine clearance in humans (the measure of glomerular filtration
rate), and also between protein intake and creatinine clearance (294). Therefore, it is possible
that protein may increase glomerular filtration rate via increased urea nitrogen excretion, as the
result of which more C-peptide may end up cleared by the kidney, causing no change in plasma
C-peptide to be observed.
In conclusion, we found that the glucose-lowering effect of fat and protein was not
blunted by insulin resistance. Fat has no effect on liver insulin clearance, whereas protein
significantly decreased it. Protein significantly increased peripheral insulin concentration
without any effect on C-peptide and insulin secretion rate, which may be explained by either
decreased insulin clearance or increased rate of C-peptide clearance.
71
CHAPTER 5
ARE THE GLYCEMIC AND INSULINEMIC INDEX VALUES OF CARBOHYDRATE
FOODS SIMILAR IN HEALTHY CONTROL, HYPERINSULINEMIC AND TYPE 2
DIABETIC PATIENTS?
EFFECTS OF METABOLIC STATUS ON GI AND INSULINEMIC INDEX
A partial content of this chapter is published in European Journal of Clinical Nutrition (370).
72
5.1 Abstract
Background: A criticism of Glycemic Index (GI) is that it does not indicate the insulin response
of foods (Insulinemic Index). However, it is unknown if the GI and Insulinemic Index of foods
are equivalent in all subjects, a necessary criterion for clinical utility.
Objective: To compare GI and Insulinemic Index in non-diabetic subjects with fasting-serum-
insulin (FSI) <40pmol/L (Healthy Control) or with FSI≥ 40pmol/L (Hyperinsulinemic) and
subjects with type 2 diabetes (T2DM), and to see whether GI and Insulinemic Index were related
to the serum-glucose concentrations, insulin sensitivity, β-cell function, and hepatic insulin
extraction of the subjects.
Design: Serum-glucose, -insulin and -C-peptide responses after 50g available-carbohydrate
portions of glucose (tested 3 times by each subject), sucrose, instant mashed-potato, white-bread,
polished-rice and pearled-barley were measured in Healthy Control (n=9), Hyperinsulinemic
(n=12) and T2DM (n=10) subjects using standard GI methodology.
Results: Food GI values did not differ significantly among the 3 subject-groups, whereas
Insulinemic Index values were higher in T2DM (100±7) than Healthy Control (78±5) and
Hyperinsulinemic subjects (70±5) (mean ±SEM, p=0.05). Insulinemic Index was inversely
associated with insulin sensitivity (r=-0.66, p<0.0001) and positively related to fasting- and
postprandial-glucose (both r=0.68, p<0.0001) and hepatic insulin extraction (r=0.62, p=0.0002).
In contrast, GI was not related to any of the biomarkers (p>0.05).
Conclusion: The GI is a valid property of foods because its value is similar in Healthy Control,
Hyperinsulinemic and T2DM subjects, and is independent of subjects‘ metabolic status.
However, Insulinemic Index may depend upon the glycemic control, insulin sensitivity and
hepatic insulin extraction of the subjects.
73
5.2 Introduction
The blood glucose raising potential of carbohydrate in foods is numerically classified as
Glycemic Index (GI), which is the incremental area under the glycemic response curve (AUC)
elicited by a 50g available-carbohydrate portion of a food expressed as a percentage of that after
50g glucose in the same subject (15). Many foods have been tested for their GI values in healthy
or diabetic subjects (17). Since the GI values of foods are similar in healthy and diabetic
subjects (19-21), GI values tested in healthy subjects can be applied in the nutritional
management of diabetes (295). However, because GI has never been tested in non-diabetic
subjects with insulin resistance/hyperinsulinemia, it is not known whether GI is valid in this
population.
Because hyperinsulinemia may play a role in the pathogenesis of insulin resistance and
associated chronic diseases (22-27), a major concern about the GI concept is that it does not
consider concurrent insulin response. Thus, some investigators have begun to report values for
Insulinemic Index (28-30). Insulinemic Index is calculated in a similar way to GI to measure the
extent to which a food raises plasma insulin (31). However, for Insulinemic Index to be a valid
property of a food, its value should be similar in different subjects regardless of their degree of
insulin sensitivity, β-cell function or glucose tolerance status. There is evidence that relative
insulin responses differ in lean and obese subjects with normal glucose tolerance (NGT), subjects
with impaired glucose tolerance (IGT) and subjects with type 2 diabetes (296). In addition,
relative insulin responses were inversely related to subjects‘ fasting insulin, suggesting that
Insulinemic Index may dependent on subject‘s insulin sensitivity (297). However, these studies
used mixed meals rather than individual foods and did not use standard GI methodology; thus,
74
their results cannot be used to draw conclusions about the validity of the Insulinemic Index of
carbohydrate foods.
Therefore, we investigated whether the GI and Insulinemic Index of a variety of
carbohydrate foods are similar in healthy control, hyperinsulinemic and type 2 diabetic subjects,
and whether metabolic status such as insulin sensitivity, β-cell function, severity of glycemia
(fasting- and postprandial-glucose), hepatic insulin extraction and plasma GLP-1 response of the
subjects were correlated with the GI and Insulinemic Index obtained.
5.3 Subjects and Methods
5.3.1 Subjects and study design
We recruited male and non-pregnant, non-lactating female subjects aged 18-70 yr (BMI <
35kg/m²) with and without T2DM. Subjects were excluded for any of the following: history of
gastrointestinal disease or gastroparesis, evidence of liver disease (aspartate transaminase,
alanine transaminase, and gamma glutamyl transpeptidase > 2 times upper limit of normal) or
kidney disease (creatinine > 1.2 times upper limit of normal), use of α-glucosidase or lipase
inhibitors or insulin, or any acute medical or surgical event requiring hospitalization within 6
months. Subjects without diabetes had fasting glucose <7.0mmol/L and were divided
prospectively into those with normal fasting serum insulin (FSI) (Healthy Control, n=9, FSI < 40
pmol/L), or high-FSI (Hyperinsulinemic, n=12, FSI ≥ 40 pmol/L). Currently, there are no
criteria by which an individual could be classified as insulin sensitive or resistant, the rationale
for using FSI is the fact that FSI is strongly correlated with insulin resistance measured by
euglycemic- hyperinsulinemic clamp (260), and the 40 pmol/L cut-off point was chosen because
this represents approximately the 67th
percentile for non-diabetic subjects in our laboratory (244).
75
In previous studies, non-diabetic subjects with FSI > 40 pmol/L had significantly greater waist
circumference (WC), BMI, homeostasis model assessment (HOMA) insulin resistance index,
total and LDL cholesterol and triglycerides, and lower HDL cholesterol than those with FSI <
40pmol/L (38).
Subjects with T2DM (n=10) had fasting glucose > 7.0mmol/L or HbA1C > 6.0% and
were treated by diet alone or any combination of sulfonylurea, metformin, thiazolidinedione
and/or meglitinide. After screening 25 subjects with T2DM, 10 were recruited who had HbA1C
of 7.3 ± 0.3% (mean±SEM) and duration of diabetes of 7.0±1.0 y (mean±SEM). Subjects were
excluded for any of the following: history of gastrointestinal disease or gastroparesis, evidence of
liver disease (aspartate transaminase, alanine transaminase, and gamma glutamyl transpeptidase
> 2 times upper limit of normal) or kidney disease (creatinine > 1.2 times upper limit of normal),
use of α-glucosidase or lipase inhibitors or insulin, or any acute medical or surgical event
requiring hospitalization within 6 months. Eight patients were on metformin alone, 1 on
metformin and pioglitazone and 1 on metformin and sulfonylurea. The patients took their usual
medication on study days after the fasting blood sample and prior to starting the test meal. None
of the patients had a history of micro- or macro-vascular complications.
There were originally 13 Hyperinsulinemic subjects, one subject missed 2 test meals;
thus her data were excluded from the statistical analysis. In a T2DM patient, the AUC of insulin
to one of the oral glucose tests was exceedingly low; therefore, the data from that oral glucose
test were not used in calculating the reference AUCs for GI, Insulinemic Index and C-peptide
index.
The protocol was designed to conform to standard GI testing methodology for subjects
with and without diabetes. Subjects were instructed to maintain their usual daily routine and
76
food intake patterns between study days and refrain from exercise on the mornings of the test.
After an overnight fast (10-14hr), they came to the Risk Factor Modification Center at St.
Michael‘s hospital on 8 separate mornings between 7:30-9:30 am. Healthy controls and
Hyperinsulinemic subjects had venous blood samples drawn just before and at 15, 30, 45, 60, 90
and 120 min after starting to eat. T2DM subjects had venous blood samples drawn fasting and at
30, 60, 90, 120 and 180 min after starting to eat. The samples were used to measure plasma
glucose, insulin, C-peptide and GLP-1. The protocol was reviewed and approved by Research
Ethics Boards at the University of Toronto and St. Michael‘s Hospital. All subjects gave written
informed consent.
5.3.2 Test foods
Subjects were fed test meals consisting of 50g available carbohydrate (defined as total
carbohydrate minus dietary fibre) as 50g anhydrous glucose (Grain Processing Enterprising
LTD, Scarborough, Ontario, Canada), 50g sucrose (Redpath brand, Toronto, Ontario, Canada),
71.8 g instant mashed potato (General Mills, Mississauga, Ontario, Canada), 107g white bread
(Weston Bakeries LTD, Toronto, Ontario, Canada), 62.5 g polished rice (Dainty brand, Windsor,
Ontario, Canada) and 80.6 g pearled barley (Quik kook brand, JVF Canada Inc, Toronto,
Ontario, Canada). The nutrient composition of the test foods was shown in Table 5.1. Sugars
(glucose + sucrose) were dissolved in 250ml water. Instant potato was weighed dry, and boiling
water added according to package directions. Two to four portions of rice and barley were
weighed dry, boiled in salted water according to package directions on the morning of the test.
The resulting cooked amount was weighed and divided into portions. The test meals were served
77
with a glass of water. Each subject tested glucose 3 times (first, fourth and last tests) and the
other 5 test meals were once each in randomized order.
5.3.3 Blood analysis
Venous blood samples for glucose, insulin and C-peptide were collected in BD
VacutainerTM
SSTTM
tubes (BD, Franklin Lakes, NJ, USA). Serum glucose was measured by a
glucose oxidase method (SYNCHRON LX Systems, Beckman Coulter, Brea, California, USA),
with inter-assay coefficient of variation (CV) of 1.9%. Insulin was measured using one-step
immunoenzymatic (―sandwich‖) assay (Beckman Access Ultrasensitive Insulin Assay, Beckman
Coulter, Brea, California, USA), with inter-assay CV of 2.5 to 4.3 %. Insulin has no cross-
reactivity with proinsulin. C-peptide was measured using double antibody competitive
radioimmunoassay (Siemens Medical Solutions Diagnostics, Los Angeles, CA), with inter-assay
precision of 10% or less.
Venous blood samples for GLP-1 were collected in BD VacutainerTM
EDTATM
tubes
(BD, Franklin Lakes, NJ, USA). The dipeptidyl peptidase-4 inhibitor (Linco Research, St.
Charles, Missouri) was added immediately (less than 30 seconds) after collection. Plasma GLP-
1 was measured by capturing the active GLP-1 from the sample by a monoclonal antibody
(specific binding to the N-terminal region) using GLP-1 (active) ELISA Kit (Linco Research, St.
Charles, Missouri, USA), with inter-assay CV ranges from 1-13%. All the samples were stored
at -70°C before analysis.
78
5.3.4 Calculations and statistical analysis
The sample size was determined based on previous studies (296, 298), with a power of
80% and p<0.05 (with GI and Insulinemic Index as the dependent measures). Incremental areas
under the curve (AUC), ignoring area below baseline, for glucose, insulin, C-peptide and GLP-1
were calculated using the trapezoid rule (31). Hepatic insulin extraction (HIEauc) was determined
by the AUC of C-peptide divided by that of insulin (268). Insulin secretion rate (ISR) was
calculated from deconvolution of the plasma C-peptide concentration using ISEC software
package developed by Hovorka et al (269). Glycemic Index (GI) was calculated as the AUC of
the test food expressed as a percentage of the mean AUC of 3 tests of oral glucose in the same
subject (15); the mean of the resulting values was the GI of the food. Insulinemic Index was
calculated in a similar fashion as GI. C-peptide is co-secreted with insulin in equimolar
amounts, but is not subjected to hepatic insulin clearance, which varies considerably; thus, C-
peptide is regarded as a much better estimate of insulin secretion than levels of insulin itself
(239). Therefore, C-peptide index was also calculated using the same method for GI and
Insulinemic Index calculation.
The mean glucose and insulin values of the 3 oral glucose tests were used to calculate
insulin sensitivity index, which was calculated using two validated OGTT-derived indices of
insulin sensitivity: 1) Matsuda‘s insulin sensitivity index, which incorporates both hepatic and
muscle components of insulin resistance, correlates well with euglycemic-hyperinsulinemic
clamp, and was calculated as 10,000/(fasting glucose × fasting insulin × mean glucose × mean
insulin )0.5
(173); 2) Mari‘s glucose-insulin model-based oral glucose insulin sensitivity index
constructed on established principles of glucose kinetics and insulin action (174). Oral glucose
79
insulin sensitivity has been validated against the euglycemic-hyperinsulinemic clamp in healthy,
obese and type 2 diabetic subjects (174).
The β-cell compensation for insulin resistance was estimated using the insulin
secretion/insulin resistance (disposition) index derived from OGTT(270), which was shown to be
the best predictor of future development of type 2 diabetes in subjects with NGT compared with
other predictive models such as San Antonio Diabetes Prediction Model (271) (including age,
sex, ethnicity, BMI, blood pressure, fasting plasma glucose, triglycerides, and HDL) and 2-hour
plasma glucose concentration. To calculate insulin secretion/insulin resistance (disposition)
index, first, insulin secretion was expressed as the changes in AUC of insulin secretion rate
(∆ISRAUC) relative to changes in AUC of plasma glucose response (∆ISRAUC/∆GAUC ), then
∆ISRAUC/∆GAUC is divided by the severity of insulin resistance (∆ISRAUC/∆GAUC÷ IR), as
measured by the inverse of Matsuda‘s insulin sensitivity index.
Data were expressed as mean ± SEM for normally distributed variables or median
(interquartile range) for non-normally distributed variables. Normality was assessed using the
Shapiro and Wilk statistic and the normality plots (PROC UNIVARIATE procedure of SAS).
Skewed variables were log-transformed prior to analysis. The values of GI, Insulinemic Index
and C-peptide index were subjected to repeated measures ANOVA (PROC MIXED) to test for
the main effects of food and subject group (Healthy Control, Hyperinsulinemic and T2DM) and
the food×group interaction. Age, BMI and WC were included in the model as covariates to
control for the differences in these variables between subjects. Tukey‘s post hoc test was
performed to compare individual means if the main effects or interactions were statistically
significant.
80
The correlations between GI, Insulinemic Index and the metabolic indices [insulin
sensitivity, β-cell function, hepatic insulin extraction, severity of glycemia (fasting glucose,
mean postprandial glucose and glucose AUC), and GLP-1 response] were determined by simple
linear regressions. Step-wise multiple regression analysis was used to examine the extent to
which the different variables accounted for the variability of GI or Insulinemic Index. When GI
is the dependent variable, the independent variables are age, BMI, oral glucose insulin
sensitivity, insulin secretion/insulin sensitivity (disposition index), hepatic insulin extraction, and
the AUC of glucose and GLP-1. When Insulinemic Index is the dependent variable, the
independent variables are GI, age, BMI, oral glucose insulin sensitivity, insulin secretion/insulin
sensitivity (disposition index), hepatic insulin extraction, and the AUC of glucose and GLP-1.
Collinearity was determined by including variance inflation factor in the model, with variance
inflation factor of 5 or 10 and above indicates a multicollinearity problem (299). No collinearity
was apparent for the variables included in the regression analysis. All analyses were done using
SAS 9.2, (SAS Institute Inc, Cary). Differences were considered significant if 2-tailed p<0.05.
5.4 Results
T2DM subjects were significantly older and had higher BMI and WC than both Healthy
Control and Hyperinsulinemic subjects (Table 5.2). Fasting glucose and postprandial glucose
responses (glucose AUC) were not significantly different between Healthy Control and
Hyperinsulinemic subjects but were significantly higher in T2DM (Table 5.2). The intra-
individual coefficient of variations (CV) of blood glucose AUC for repeated tests of oral glucose
for Healthy Control, Hyperinsulinemic and T2DM subjects are 24±5, 25±3 and 17±3
(mean±SEM) respectively. Fasting insulin was significantly higher in Hyperinsulinemic and
T2DM than Healthy Control subjects. Fasting C-peptide increased in a step-wise fashion from
81
Healthy Control to Hyperinsulinemic to T2DM (Table 5.2). Hepatic insulin extraction was
significantly higher in T2DM than Hyperinsulinemic subjects. Matsuda‘s insulin sensitivity
index was significantly lower in Hyperinsulinemic and T2DM than Healthy Control subjects,
while oral glucose insulin sensitivity and disposition index decreased step-wisely from Healthy
Control to Hyperinsulinemic to T2DM (Table 5.2). Systolic and diastolic blood pressure, total
cholesterol, triglyceride, total:HDL cholesterol were similar in Healthy Control and
Hyperinsulinemic subjects, but were significantly higher in T2DM. Fasting and postprandial
GLP-1, HDL, LDL and CRP were not significantly different among the 3 subject groups.
Figure 5.1 shows the 2-3hr postprandial glucose, insulin, C-peptide and GLP-1 responses
elicited by different foods in Healthy Control, Hyperinsulinemic and T2DM subjects. For all
foods, the general trend was that serum glucose was elevated in T2DM, whereas it was similar in
Healthy Control and Hyperinsulinemic; serum insulin was much higher in Hyperinsulinemic
than Healthy Control and T2DM subjects; C-peptide was elevated in both Hyperinsulinemic and
T2DM subjects; GLP-1 was lower in T2DM but similar in Healthy Control and
Hyperinsulinemic subjects.
Table 5.3 shows that the mean AUC for glucose and insulin differed significantly across
subject-groups for each individual food as well as for the mean of all carbohydrate foods. C-
peptide AUC was significantly higher in Hyperinsulinemic than Healthy Controls for oral
glucose, rice and the mean of all foods (Table 5.3). GLP-1 AUC was not significantly different
among subject groups for any food individually nor for the mean of all carbohydrate foods
(Table 5.3). The AUCs of ISR were significantly higher in Hyperinsulinemic subjects than
Healthy Control and T2DM subjects except for white bread and barely (Table 5.3). Hepatic
82
insulin extraction in response to oral glucose was significantly higher in T2DM than
Hyperinsulinemic subjects (Table 5.3).
There were significant food effects on GI, Insulinemic Index and C-peptide index
(p<0.0001) (Table 5.4), demonstrating that different carbohydrate foods produced different GI,
Insulinemic Index or C-peptide index. For GI values, there was neither significant subject-group
effects (p=0.20) nor significant food × subject-group interactions (p=0.26) (Table 5.4); thus, the
GI values were not significantly different among Healthy Control, Hyperinsulinemic and T2DM
for any food individually, nor for the mean of all carbohydrate foods (Table 5.5). For
Insulinemic Index, there was a tendency of significant subject-group effects (p=0.05) and food ×
subject-group interaction effects (p=0.09); thus the Insulinemic Index of white bread was
significantly higher in T2DM than Healthy Control and Hyperinsulinemic subjects (p=0.01) and
the mean Insulinemic Index of all carbohydrate foods was much higher in T2DM than Healthy
Control and Hyperinsulinemic subjects (p=0.05) (Table 5.5). For C-peptide index, there was a
significant food × subject-group interactions (p=0.03), but no significant subject-group effects
(p=0.40) (Table 5.4). The values of C-peptide index were not significantly different among
Healthy Control, Hyperinsulinemic and T2DM for any food individually, nor for the mean of all
carbohydrate foods (Table 5.5).
The GI and Insulinemic Index were well correlated in Healthy Control (r=0.61,
p<0.0001), Hyperinsulinemic subjects (r=0.62, p<0.0001) and T2DM patients (r=0.29, p=0.04)
respectively (Figure 5.2); similarly, the GI and C-peptide index were also correlated in Healthy
Control (r=0.36, p=0.02), Hyperinsulinemic subjects (r=0.48, p=0.0001) and T2DM patients
(r=0.46, p=0.0009) respectively (Figure 5.2). The regression lines of GI vs. Insulinemic Index
83
/C-peptide index were nearly identical for Healthy Control, Hyperinsulinemic and T2DM
patients (Figure 5.2).
The GI was not related to any of the anthropometric (age and BMI) or metabolic indices
(oral glucose insulin sensitivity, hepatic insulin extraction, disposition index and GLP-1
response) for each food individually (p>0.05), nor for the mean of all carbohydrate foods (p >
0.05) (Figure 5.3). Multiple regression analysis using GI as dependent variable and age, BMI,
oral glucose insulin sensitivity, disposition index, hepatic insulin extraction and AUC of glucose
and GLP-1as independent variables found that none of the variables met the 0.05 significance
level for entry into the model.
Insulinemic Index was inversely associated with oral glucose insulin sensitivity (r = -
0.66, p<0.0001) and positively related to hepatic insulin extraction (r = 0.62, p=0.0002). GI was
not related to either oral glucose insulin sensitivity (r = -0.07, p=0.73) or hepatic insulin
extraction (r=-0.09, p=0.65) (Figure 5.3). C-peptide index was not significantly related to oral
glucose insulin sensitivity (r=-0.17, p=0.37) and hepatic insulin extraction (r=0.35, p=0.057).
The GI, Insulinemic Index, and C-peptide index were not related to either GLP-1 response or
disposition index (p>0.05). Multiple regression analysis using Insulinemic Index as dependent
variable and age, BMI, GI, oral glucose insulin sensitivity, disposition index, hepatic insulin
extraction and the AUC of glucose and GLP-1 as independent variables showed that oral glucose
insulin sensitivity and hepatic insulin extraction together predicted Insulinemic Index
(Insulinemic Index =106-0.09×oral glucose insulin sensitivity +2.75×hepatic insulin extraction),
and explained approximately 51% of the variation in Insulinemic Index (r2=0.51, p<0.0001). In
particular, oral glucose insulin sensitivity alone explained 43% of the variation in Insulinemic
Index (r2=0.43, p<0.001).
84
GI was not related to any of the markers of the severity of glycemia (p>0.05) (Figure
5.4) whereas Insulinemic Index was positively associated with fasting glucose (r=0.68,
p<0.0001), mean postprandial glucose (r=0.68, p<0.0001) and glucose AUC (r=0.46, p=0.009).
Figure 5.5 shows the comparison of the overall means of GI, Insulinemic Index and C-
peptide index for each carbohydrate food for all 31 subjects. Barley, rice, sucrose and the mean
of all carbohydrate foods had similar GI, Insulinemic Index and C-peptide index; however,
mashed potato had disproportionately higher Insulinemic Index and C-peptide index than GI and
white bread had higher Insulinemic Index than GI (Figure 5.5).
85
Table 5.1 Nutrition composition of the test foods
Weight
(g)
Total CHO
(g)
Fat
(g)
Protein
(g)
Dietary fiber
(g)
Available CHO
(g)
Oral glucose 50.0 50.0 0 0 0.0 50.0
Sucrose 50.0 50.0 0 0 0.0 50.0
Mashed potato 71.8 56.3 0.9 6.2 6.3 50.0
White bread 107 51.4 2.9 10 1.4 50.0
Rice 62.5 50.0 0 5.6 0.0 50.0
Barley 80.6 62.5 0.9 9.0 12.5 50.0
Source: the information on carbohydrate, fat, protein and dietary fiber are from the nutrient
composition table on the package of the foods. Available CHO = total CHO - dietary fiber.
86
Table 5.2 Anthropometric and metabolic characteristics of the study groups1
Control (n=9) Hyper[I] (n=12) T2DM (n=10) P2
Age (yr) 24(23-29)a,3
26(23-37)a 57(53-58)
b <0.0001
BMI (kg/m2) 22(21-22)
a 25(22-30)
b 32(29-34)
c <0.0001
Waist circumference (cm) 72(70-76)a 88(77-95)
b 102(94-115)
c <0.0001
Fasting glucose (mmol/L) 4.6 (4.5-4.7)a 4.9 (4.7-5.2)
a 9.8 (7.2-11.3)
b <0.0001
Glucose AUC (mmol×min/L)4 198(138-291)
a 247(186-278)
a 831(699-978)
b 0.0002
Fasting insulin (pmol/L) 29 (24-31)a 51 (47-57)
b 51 (42-63)
b 0.0002
Fasting C-peptide (pmol/L) 231 (213-241)a 412 (345-445)
b 784 (711-878)
c <0.0001
Fasting GLP-1 (pmol/L) 6.1±2.2 6.5±2.2 4.4±0.6 0.70
2hr-postprandial GLP-1 (pmol ×min/L4) 211(94-340) 183(113-294) 281(133-420) 0.64
Hepatic insulin extraction4 3.8(3.3-6.4)
ab 3.4(2.7-3.8)
a 6.8(5.7-10.4)
b 0.002
Matsuda‘s insulin sensitivity index4 28 (27-32)
a 14 (12-17)
b 12 (9-15)
b <0.0001
Oral glucose insulin sensitivity4 497 (478-521)
a 459 (424-485)
b 280 (259-362)
c 0.001
Disposition index4 63±5
a,5 38±4
b 29±4
c 0.0008
Systolic BP (mm Hg) 101±2.4a 109±2.2
a 130±4.1
b <0.0001
Diastolic BP (mm Hg) 61±2.2a 67±2.2
a 79±2.9
b 0.0001
Total cholesterol (mmol/L) 4.1±0.3a 4.4±0.3
ab 5.5±0.5
b 0.02
Triglycerides (mmol/L) 0.7 (0.5-0.9)a 0.9 (0.8-1.1)
a 1.5 (1.2-1.9)
b 0.02
HDL-C (mmol/L) 1.4±0.1 1.2±0.1 1.2±0.1 0.41
LDL-C (mmol/L) 2.4±0.3 2.7±0.2 3.1±0.3 0.27
Total:HDL cholesterol 2.8 (2.7-3.8)a 3.6 (3.2-3.9)
ab 3.9 (3.3-6.6)
b 0.02
C-reactive protein (mg/L) 0.7 (0.2-1.2) 1.5 (0.5-3.8) 3.7 (1.1-5.4) 0.10
87
1 Healthy, non-diabetic subjects with fasting serum insulin (FSI) <40pmol/L; Hyperinsulinemic,
non-diabetic subjects with FSI≥40pmol/L; T2DM, subjects with type 2 diabetes; BP, blood
pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein.
2 P represents overall significant differences across groups. P-values were derived from one-way
ANOVA except those for 2-hour postprandial GLP-1 response, hepatic insulin extraction,
Matsuda‘s insulin sensitivity index, oral glucose insulin sensitivity and disposition index.
3 Median; interquartile range in parentheses (all such values).
Values in the same row with
different superscript letters differ significantly, P<0.05 (Tukey‘s post hoc test).
4 Calculated by using the mean of 3 oral-glucose-tolerance test results. PROC MIXED repeated
measures ANOVA, adjusted for age, BMI and WC. The units for Matsuda‘s insulin sensitivity
index, oral glucose insulin sensitivity and disposition index are (mmol ×pmol)-1
×L, (ml×min-1
×m-2
), and (mmol)-2
×kg-1
×min-1 ×L respectively. Hepatic insulin extraction doesn‘t have units.
5 Mean±SEM (all such values).
88
Table 5.3 The glucose, insulin, C-peptide, GLP-1 and insulin secretion rate (ISR) expressed as
area under the curve (AUC) and hepatic insulin extraction (HIE) for each individual food and the
mean of all carbohydrate foods in the study groups1
Subjects Glucose Sucrose Potato Bread Rice Barley Mean2
Glucose AUC (mmol ×min/L)
Control FSI<40 198(138-291)a,3
123(84-165)a 256(139-265)
a 125(66-149)
a 134(82-151)
a 140(69-175)
a 139(92-200)
a
Hyper[I] FSI≥40 247(186-278)a 163(118-213)
a 163(105-356)
a 173(126-258)
b 212(119-232)
b 111(58-161)
a 174(120-239)
b
Type 2 Diabetes 831(699-978)b 473(402-747)
b 708(562-1145)
b 651(436-818)
c 609(403-851)
c 468(255-564)
b 609(437-826)
c
P5 0.0002 0.002 0.0004 0.0004 <0.0001 0.01 <0.0001
Insulin AUC (nmol×min/L)
Control FSI<40 16(12-18)a 8(8-14)
a 20(15-24)
a 11(10-17)
a 9(7-11)
a 6(4-9)
a 11(8-17)
a
Hyper[I] FSI≥40 33(29-49)b 23(19-37)
b 39(23-60)
b 25(16-46)
b 19(14-22)
b 12(8-20)
b 24(17-36)
b
Type 2 Diabetes 14(10-25)a 12(7-17)
a 17(14-29)
ab 18(11-25)
ab 11(8-19)
ab 9(6-11)
ab 14(8-19)
a
P <0.0001 <0.0001 0.002 0.03 0.002 0.04 <0.0001
C-peptide AUC (nmol×min/L)
Control FSI<40 64±4a 42±7
4 87±10 56±8 28±7
a 30±4 51(29-74)
a
Hyper[I] FSI≥40 117±9b 82±8 108±12 92±12 89±12
b 50±6 89(59-122)
b
Type 2 Diabetes 111±15ab
93±14 139±20 122±15 85±15ab
62±9 102(67-138)ab
P 0.0009 0.049 0.37 0.11 0.01 0.19 0.0003
GLP-1 AUC (pmol ×min/L)
Control FSI<40 211(94-340) 66(46-217) 169(90-329) 66(49-184) 59(26-87) 52(34-71) 86(47-217)
Hyper[I] FSI≥40 183(113-294) 94(40-151) 159(99-281) 80(8-213) 33(18-184) 23(1-89) 104(29-198)
Type 2 Diabetes 281(133-420) 62(20-227) 216(90-270) 169(34-259) 102(70-186) 72(21-111) 121(52-265)
P 0.64 0.85 0.99 0.91 0.60 0.95 0.91
89
Table 5.3 (continued)
Subjects Glucose Sucrose Potato Bread Rice Barley Mean2
ISR AUC (pmol/kg)
Control FSI<40 404±32a 281±45
a 546±63
a 395±55 233±46
a 225±32 361±21
a
Hyper[I] FSI≥40 641±44b 504±60
b 692±61
b 540±85 565±75
b 364±53 576±27
b
Type 2 Diabetes 559±29a 498±50
ab 568±62
ab 555±43 456±52
ab 419±47 522±18
b
P 0.001 0.02 0.04 0.24 0.003 0.14 <0.0001
Hepatic insulin extraction
Control FSI<40 3.8(3.3-6.4)ab
5.4(2.0-7.0) 4.4(3.8-5.5) 5.1±0.8 3.0(1.7-4.5) 5.2(2.6-7.6) 4.9(2.8-6.6)
Hyper[I] FSI≥40 3.4(2.7-3.8)a 3.2(2.3-4.2) 3.1(2.3-3.7) 3.3±0.5 4.0(2.8-6.6) 3.9(2.7-6.1) 3.3(2.5-4.4)
Type 2 Diabetes 6.8(5.7-10.4)b 7.3(5.3-8.4) 6.4(4.7-8.4) 6.9±0.9 6.9(5.7-7.7) 6.4(5.0-9.7) 6.8(5.4-9)
P 0.002 0.26 0.06 0.07 0.24 0.60 0.05
1 Healthy control subjects with fasting serum insulin (FSI) <40pmol/L (n=9); Hyperinsulinemic,
non-diabetic subjects with FSI≥40pmol/L (n=12); T2DM, subjects with type 2 diabetes (n=10).
2 Refers to the mean of all carbohydrate foods.
3 Median; interquartile range in parentheses (all such values). Values in the same column with
different superscript letters are significantly different, P < 0.05 (Tukey‘s post hoc test).
4 Mean±SEM (all such values).
5 P
refers to overall significant differences across subject-groups (repeated measures ANOVA,
PROC MIXED, Tukey‘s post hoc test, adjusted for age, BMI and WC).
90
Table 5.4 Main and interaction effects of food, subject group on GI, Insulinemic Index and
C-peptide index
GI2 Insulinemic Index
2
C-peptide index2
p-value
Food <0.0001 <0.0001 <0.0001
Subjects group1 0.20 0.05 0.40
Food×subjects group 0.26 0.09 0.03
1Healthy Control vs. Hyper [I] vs. T2DM subjects. The p-values are derived from repeated
measures of ANOVA (PROC MIXED). Age, BMI and WC are included in the model as
covariates.
2GI was normally distributed. Insulinemic Index and C-peptide index were not normally
distributed and were log transformed prior analysis.
91
Table 5.5 The GI, Insulinemic Index and C-peptide index of different carbohydrate foods in the
study groups1
Subjects Glucose Sucrose Potato Bread Rice Barley Mean2
Glycemic Index (%)
Control 100±0 68±9 101±7 63±11 61±7 58±6 70±4
Hyper[I] 100±0 69±6 81±10 70±6 72±7 44±6 67±3
T2DM 100±0 68±5 98±5 71±6 70±10 47±5 70±4
P3 0.95 0.20 0.13 0.10 0.55 0.20
Insulinemic Index (%)
Control 100±0 70±8 128±10 86±9ab
60±6 46±5 78±5
Hyper[I] 100±0 74±8 114±10 74±7a 56±7 35±4 70±5
T2DM 100±0 81±10 137±12 138±13b 73±7 70±16 100±7
P 0.85 0.28 0.01 0.81 0.31 0.05
C-peptide index (%)
Control 100±0 65±11 142±18 85±10 45±11 47±7 77±7
Hyper[I] 100±0 74±9 96±12 77±8 77±10 44±6 74±4
T2DM 100±0 83±9 123±14 113±11 72±8 57±6 90±6
P 0.44 0.07 0.14 0.24 0.70 0.40
1 Healthy control subjects with fasting serum insulin (FSI) <40pmol/L (n=9); Hyperinsulinemic,
non-diabetic subjects with FSI≥40pmol/L (n=12); T2DM, subjects with type 2 diabetes (n=10).
Values (Mean±SEM) in the same column with different superscript letters are significantly
different, P < 0.05 (Tukey‘s post hoc test).
2 Refers to the mean of all carbohydrate foods.
92
3 p-values refer to overall significant differences across subject-groups, and were derived from
repeated measures of ANOVA (PROC MIXED, Tukey‘s post hoc test). Age, BMI and WC
were included in the model as covariates.
93
0
200
400
600
Insulin
(pm
ol/L
)
barley
4
8
12
16
Control FSI < 40 pmol/L
Hyper[I] FSI 40pomol/L
T2DM
Glu
cose
(mm
ol/L
)
rice white bread Sucrose mashed potato
0
500
1000
1500
2000
C-P
ep
tid
e(p
mo
l/L
)
0 30 60 90 1201501800
5
10
GL
P-1
(pm
ol/L
)
0 30 60 90 120150180 0 30 60 90 120150180 0 30 60 90 120150180 0 30 60 90 120150180
Time (min)
Figure 5.1 The 2-3hr postprandial glucose, insulin, C-peptide and GLP-1 responses
(Mean±SEM) to different carbohydrate foods in Healthy Control (FSI < 40pmol/L, open circle,
dotted line), Hyperinsulinemic (FSI ≥ 40pmol/L, grey circle, solid line) and T2DM subjects
(black circle, solid line).
94
0
100
200
300
Hyper[I] FSI 40pomol/L
Control FSI < 40 pmol/L
T2DMIn
sulinem
ic Index (
II)
0 50 100 1500
50
100
150
200
Glycemic Index (GI)
C-p
eptide In
dex
Figure 5.2 The linear correlation between Glycemic Index (GI) and Insulinemic Index (top) and
C-peptide index (bottom) respectively in Healthy Control (FSI< 40pmol/L, open circle and solid
line), Hyperinsulinemic (FSI ≥ 40pmol/L, grey circle and dotted line) and T2DM (black circle
and dashed line). The lines are regression line.
95
50
100
150
Hyper[I] FSI 40pomol/L
Control FSI < 40 pmol/L
T2DM
r=-0.07p=0.73
Gly
cem
ic Index (
GI)
0
50
100
150
200 r = -0.66p<0.0001
Insulin
em
ic Index (
II)
200 300 400 500 6000
50
100
150
r = -0.17 p=0.37
OGIS
(ml/min/m2)
C-p
eptide Index
r=-0.09p=0.65
r = 0.62 p=0.0002
0 5 10 15
r = 0.35p=0.057
HIE
Figure 5.3 The linear correlation between GI (top) or Insulinemic Index (middle) or C-peptide
index (bottom) and oral glucose insulin sensitivity (OGIS) (left) and hepatic insulin extraction
(HIE) (right) respectively in all subject-groups (Healthy Control, FSI < 40pmol/L, white circle;
Hyperinsulinemic, FSI ≥ 40pmol/L, grey circle; and T2DM, black circle). r: Pearson‘s
correlation coefficient. The values of GI, Insulinemic Index and C-peptide index are the mean of
five carbohydrate foods for each subject (n=31). Oral glucose insulin sensitivity and hepatic
96
insulin extraction are calculated from the mean of 3 oral-glucose-tests. The lines are regression
line.
97
0
50
100
150
Hyper[I] FSI 40pomol/L
Control FSI < 40 pmol/L
T2DM
r=-0.10p=0.58
Gly
cem
ic Index (
GI)
r=0.0003p=0.99
r=0.19p=0.31
0
50
100
150
r=0.68p<0.0001
Fasting Glucose(mmol/L)
Insulin
em
ic Index (
II)
r=0.68p<0.0001
Mean Postprandial Glucose(mmol/L)
r=0.46p=0.009
Glucose AUC
(pmol min/L)
Figure 5.4 The linear correlation between GI (top) or Insulinemic Index (bottom) and fasting
glucose (left) and mean postprandial glucose (middle) and glucose AUC (right) respectively in
all subject-groups (Healthy Control, FSI < 40pmol/L, white circle; Hyperinsulinemic, FSI ≥
40pmol/L, grey circle; and T2DM, black circle). r: Pearson‘s correlation coefficient. The values
of GI and Insulinemic Index are the mean of five carbohydrate foods for each subject (n=31).
Fasting glucose, mean postprandial glucose and glucose AUC are calculated from the mean of 3
oral-glucose-tests. The lines are regression line.
98
barley mashed potato rice sucrose white bread all CHO foods
0
50
100
150
GI II C-peptide index
p=0.02
p=0.002
p=0.01
Figure 5.5 The comparison of overall mean±SEM of GI (white bar), Insulinemic Index (grey
bar) and C-peptide index (black bar) for each food and all carbohydrate foods. Data are from all
subject groups (Healthy Control, Hyperinsulinemic and T2DM). The p-values were derived
from repeated measures of ANOVA (PROC MIXED, Tukey‘s post hoc test). Age, BMI and WC
were included in the model as covariates.
99
5.5 Discussion and conclusions
The results of this study confirm that the GI values of carbohydrate foods are similar in
all subjects regardless of the severity of glycemia or degree of insulin resistance. This shows
that GI is a property of foods and affirms its clinical utility in a broad population. However, the
Insulinemic Index values of carbohydrate foods were inversely associated with insulin sensitivity
but positively related to the severity of glycemia and hepatic insulin extraction of the subjects,
suggesting that Insulinemic Index is not solely a property of foods but also depends on the
metabolic status of the subjects in whom it is measured.
We found that between-subject variation of GI values was not explained by any
demographic, anthropometric or metabolic variable measured. Previous studies have shown that
GI values are similar in healthy controls vs. T2DM (241), individuals with type 1 diabetes
(T1DM) vs. individuals with T2DM (242), adults vs. children with T1DM (21), T2DM subjects
on oral agents vs. insulin (300), and T2DM subjects in good vs. poor metabolic control (301).
We showed here that the GI values of foods in hyperinsulinemic subjects were similar to those in
healthy control and T2DM subjects. While not unexpected, this is important because of evidence
that GI may be particularly useful for obese and/or insulin resistant subjects to assist with weight
management (302) and/or the prevention of T2DM (104) and stroke (303). Thus, it is valid to
utilize the GI values of foods tested in healthy control subjects in the dietary management of
hyperinsulinemic/ insulin resistant subjects.
The GI values of carbohydrate foods depend on differences in their relative rates of
digestion and absorption (that is, the rate of glucose appearance from the gut) (31) which,
presumably, do not differ in healthy, hyperinsulinemic and diabetic subjects. We included
sucrose as a test meal because its glycemic response depends, at least in part, on the hepatic
100
metabolism of fructose which, in turn, may depend on insulin sensitivity. We previously showed
that the GI of fruit leather, over 50% of the available-carbohydrate of which consisted of
fructose, was inversely related to fasting insulin and to waist circumference (304). However, the
present results are not consistent with this, in that the mean GI of sucrose in Hyperinsulinemic
subjects was, if anything, slightly higher than in the healthy control.
One of the major criticisms of GI is that it does not take into account the concurrent
insulin response (34); thus, some researchers have advocated using Insulinemic Index in the
treatment of diabetes (30, 131); however, for Insulinemic Index to have clinical utility, it must be
applicable to a broader population and be similar in all subject groups regardless of their degree
of insulin sensitivity and glucose tolerance status. We found that the mean Insulinemic Index
values of all 5 carbohydrate foods were higher in T2DM than healthy control and
hyperinsulinemic subjects (p=0.05). Furthermore, Insulinemic Index values were inversely
related to oral glucose insulin sensitivity and positively related to hepatic insulin extraction and
the severity of glycemia. In addition, oral glucose insulin sensitivity alone explained 43% of the
variation in Insulinemic Index. These results suggest that the Insulinemic Index values of
carbohydrate foods vary depending on the metabolic status, thus limiting its clinical utility.
There is a physiological basis for the observed higher mean Insulinemic Index of the 5
carbohydrate foods for T2DM (100±7) than healthy control (78±5) and hyperinsulinemic
subjects (70±5) (p=0.05). Glucose is not the only stimulus for insulin secretion, gastrointestinal
hormones, mainly gastric inhibitory polypeptide (GIP) and glucagon-like peptide 1 (GLP-1) are
known to potentiate the stimulatory effect of glucose and mediate postprandial insulin secretion
(305-308). In addition, the activity of the entero-insulin axis and hepatic insulin extraction all
play a role in modulating postprandial insulin concentration. We found that oral glucose insulin
101
sensitivity and β-cell function decreased in a step-wise fashion from healthy control to
hyperinsulinemic to T2DM and hepatic insulin extraction was much higher in T2DM than
healthy and hyperinsulinemic subjects. These metabolic aberrations, especially reduced β-cell
function (less insulin secretion) and increased hepatic insulin extraction may reduce plasma
insulin response to oral glucose (the reference meal) in T2DM patients; thus other components of
foods (the test meals) on insulin secretion become more prominent, which resulted in increased
Insulinemic Index in T2DM subjects.
Though the mean Insulinemic Index of the 5 carbohydrate foods was higher in T2DM
than healthy and hyperinsulinemic subjects, for each individual food, only in white bread a
significant difference in Insulinemic Index was observed. How can the fact that Insulinemic
Index was correlated with metabolic status (oral glucose insulin sensitivity and the severity of
glycemia) be reconciled with the lack of significant difference in Insulinemic Index among the
subject groups for each individual food (except white bread)? This may be because Insulinemic
Index, oral glucose insulin sensitivity and markers of the severity of glycemia are continuous
variables, whereas the subject groups (healthy control, hyperinsulinemic, and T2DM) are
categorical variables. It has been demonstrated that continuous variables (regression analysis)
has greater statistical power than categorical variables (ANOVA analysis) due to increased
precision, a simpler and more informative interpretation of the results, and greater parsimony
(309); therefore, it is possible that categorical variables analysis (ANOVA) run the risk of
missing significant effects.
We found that the GI and Insulinemic Index/C-peptide index are highly correlated in
healthy control, hyperinsulinemic subjects and type 2 diabetic patients respectively, and the
regression lines are nearly identical (Figure 5.2); however, the Insulinemic Index values of white
102
bread and mashed potato were much higher than GI (Figure 5.5). We are not the only group to
show the dissociation between insulin and glucose responses in certain foods. Dissociation of
the glycemic and insulinemic response has been found in whole and skimmed milk, and the
protein fraction in milk was suggested to be responsible for milk‘s insulinotropic effect (126).
Also it is known that coingestion of fat and protein can evoke additional or synergistic insulin
secretion (249, 310, 311); however, our finding of the higher Insulinemic Index than GI for white
bread and mashed potato but similar GI and Insulinemic Index for rice and barley is hard to
explain because white bread and mashed potato had fat and protein content similar to those of
rice and barley (Table 5.1). The mostly likely explanation may be that both mashed potato and
white bread are highly processed and refined carbohydrate foods. There is evidence that bakery
products (rich in fat and refined carbohydrate) elicited insulin responses that were
disproportionately higher than their glycemic responses (30).
It is concluded that the GI of carbohydrate foods is not significantly different among
healthy control, hyperinsulinemic and T2DM patients, and the GI is not influenced by subject‘s
metabolic status. This finding supports the clinical utility of GI in the prevention and
management of diabetes. On the contrary, Insulinemic Index values were inversely related to
insulin sensitivity and positively associated with the severity of glycemia and hepatic insulin
extraction of the subjects, suggesting that Insulinemic Index is not a property of food but is
subject-dependent. The results suggest that Insulinemic Index may have limited clinical utility.
103
CHAPTER 6
THE RELATIONSHIP BETWEEN PLASMA GLP-1 RESPONSE AND INDIRECT
MEASURES OF INSULIN SENSITIVITY, β-CELL FUNCTION AND HEPATIC
INSULIN EXTRACTION
104
6.1 Abstract
Background: The relationships among insulin resistance, β-cell function and hepatic insulin
extraction are well characterized; however, the roles of glucagon-like peptide-1 (GLP-1) in each
of these underlying disorders remain unclear.
Objective: To determine whether endogenous postprandial GLP-1 response to oral glucose is
related to the indirect measures of insulin sensitivity, β-cell function, and hepatic insulin
extraction.
Design: Thirty-eight healthy, non-diabetic subjects classified by their fasting serum insulin (FSI)
[13 with low-FSI (FSI < 40pmol/L); 16 with medium-FSI (40 ≤ FSI < 70pmol/L); and 9 with
high-FSI (FSI ≥ 70 pmol/L)] and 10 type 2 diabetic patients (T2DM) consumed 50g oral glucose
on 3 separate occasions for the measurement of fasting and postprandial (2-hours) glucose,
insulin, C-peptide and GLP-1. The mean responses from the 3 oral glucose tests were used to
calculate indices of insulin sensitivity, β-cell function, and hepatic insulin extraction using
validated methods.
Results: Plasma GLP-1 concentration was inversely related to BMI (r=-0.33, p=0.03) and waist
circumference (r=-0.28, p=0.05), but was not associated with the indirect measures of insulin
sensitivity, β-cell function and hepatic insulin extraction (p>0.05).
Conclusion: The endogenous plasma GLP-1 response to oral glucose was not related to the
indirect measures of insulin sensitivity, β-cell function and hepatic insulin extraction.
105
6.2 Introduction
Postprandial glucose homeostasis after a meal is mediated not only by the direct
stimulation of insulin but also through the secretion of incretin hormones, namely glucose-
dependent insulinotropic peptide (GIP) and GLP-1(305-308). GLP-1 has received much more
attention than GIP because, unlike GIP, the insulinotropic effect of GLP-1 is well preserved in
type 2 diabetic patients (312, 313). Moreover, in recent years, because of the strong antidiabetic
effects of GLP-1, various GLP-1-based pharmacological agents have been developed and are
now becoming established as effective therapies for type 2 diabetes (314).
GLP-1 is released from the L-cells of the small intestine in response to nutrient ingestion,
particularly glucose and monounsaturated fatty acids (315). The GLP-1 secretion pattern in type
2 diabetic patients is not clear. Some studies have shown that type 2 diabetic patients have
reduced plasma GLP-1 response after a mixed meal (251, 316) and decreased insulinotropic
potency of GLP-1(317) whereas other studies did not detect any diminished GLP-1 response in
diabetic subjects (318, 319).
Patients with type 2 diabetes exhibit a cluster of pathophysiological abnormalities, the
major ones are insulin resistance, islet β-cell dysfunction (320), and reduced hepatic insulin
extraction (224). Moreover, these parameters are related to each other. For example, insulin
resistance is accompanied by compensatory hyperinsulinemia, which is attributed to increased
insulin secretion and reduced hepatic insulin extraction (206). The state of insulin resistance is
also associated with impaired β-cell function (320). For example, the abnormalities in β-cell
function are present in high-risk individuals such as the first-degree relatives of patients with
type 2 diabetes (321, 322), women with a history of gestational diabetes (323-325) or polycystic
ovarian syndrome (321, 326), and older subjects (327, 328).
106
Though the relationships among insulin resistance, β-cell function and hepatic insulin
extraction are well characterized, the role of GLP-1 in each of these underlying disorders is
unclear. First, it is unclear whether insulin sensitivity per se is of regulatory importance for
GLP-1 secretion, since both reduced (329, 330) and unaltered (331-333) GLP-1 responses have
been reported in non-diabetic insulin resistant subjects. Second, in cultured β-cells and rodent
models of diabetes, GLP-1 stimulates pancreatic islet β-cell differentiation (334), proliferation
(335, 336), and inhibits apoptosis of β-cells (337, 338); however, whether GLP-1 has similar
effects in humans is not known (339). It is not known whether GLP-1 per se exerts direct effects
on β-cells in humans or the improvements of β-cell function with GLP-1-based pharmacological
agents are secondary effects due to amelioration of hyperglycemia (340). Finally, it is unclear
whether GLP-1 has independent effect on liver insulin clearance. Reduced liver insulin
clearance was found in mice treated with GLP-1(341) and exenatide (GLP-1 receptor agonist)
(342), whereas in humans, both endogenously secreted and exogenously administered GLP-1
had no independent effect on liver insulin clearance (208, 343). These controversies prevent a
better understanding of the role of GLP-1 in the three systems (insulin sensitivity, β-cell function
and hepatic insulin extraction) important for glucose homeostasis.
Therefore, we utilized the oral-glucose data from two clinical studies (Chapter 4 & 5) to
determine whether there is any association between the postprandial endogenous GLP-1
response elicited by oral glucose and indirect measure of insulin sensitivity, β-cell function and
hepatic insulin extraction. It is hypothesized that the endogenous plasma GLP-1 response to oral
glucose is associated with insulin sensitivity, β-cell function and hepatic insulin extraction.
107
6.3 Materials & Methods
6.3.1 Subjects
This analysis is a part of two clinical studies comparing the effects of carbohydrate, fat
and protein on postprandial metabolic responses in healthy control, hyperinsulinemic and type 2
diabetic patients. In total, there were 48 subjects (38 healthy, non-diabetic subjects, 10 type 2
diabetic patients) who completed 3 separate oral-glucose-tests. We divided the 38 healthy
subjects into groups with different levels of fasting serum insulin (FSI): low (n=13, FSI < 40
pmol/L), medium (n=16, 40 ≤ FSI < 70 pmol/L) and high (n=9, FSI ≥ 70 pmol/L). The rationale
for this classification was based on the fact that FSI is strongly correlated with insulin resistance
measured by euglycemic-hyperinsulinemic clamp (260), and the 40 pmol/L cut-off point was
chosen because this represents approximately the 67th
percentile for non-diabetic subjects in our
laboratory (244). In our preliminary study, in which FSI > 40 pmol/L was used to define
hyperinsulinemia (38), subjects with FSI > 40 pmol/L had significantly greater waist
circumference (WC), BMI, HOMA insulin resistance index, total and LDL cholesterol and
triglycerides, and lower HDL cholesterol than those with FSI < 40pmol/L (38). We divided
subjects with FSI > 40pmol/L into medium (FSI < 70 pmol/L) and high (FSI ≥ 70 pmol/L)
groups because people with newly diagnosed diabetes have mean FSI > 70 pmol/L (261, 262).
To recruit type 2 diabetic subjects, twenty-five patients were screened by measuring their
height, weight, WC, hip circumference, blood pressure, fasting serum glucose and insulin,
HbA1C, C-reactive protein, liver enzymes, and creatinine. The inclusion criteria were: 18-70 y,
BMI < 35kg/m², fasting glucose > 7.0mmol/L or HbA1C over upper limit or normal, treated by
diet alone or with any combination of sulfonylurea, metformin, thiazolidinedione and/or
meglitinide. Exclusion criteria: gastrointestinal disease, liver disease (aspartate transaminase,
108
alanine transaminase, and gamma glutamyl transpeptidase > 2 times upper limit of normal
respectively) or kidney disease (creatinine > 1.2 times upper limit of normal), gastroparesis,
treated with acarbose, orlistat or insulin and recent acute medical or surgical event. Ten type 2
diabetic patients were recruited, with HbA1C (%) of 7.3± 0.3% (mean ±SEM). On average, the
patients had had diabetes for 7.0±3.3 y (mean±SD). Eight patients were on metformin, 1 patient
was on metformin + pioglitazone, and 1 patient was on metformin + sulfonylureas. The patients
took their usual medication during the study period. None of the patients had a history of micro-
or macro-vascular complications.
The research protocol was reviewed and approved by Research Ethics Boards at the
University of Toronto and St. Michael‘s Hospital. All subjects gave written informed consent.
6.3.2 Study Design
The subjects were instructed to maintain their usual daily routine and dietary pattern
between study days and refrain from exercise on the morning of the test. After an overnight fast
(10-14hr), they came to the Risk Factor Modification Center at St. Michael‘s hospital on three
separate mornings between 7:30-9:30am. For healthy, non-diabetic subjects, venous blood
samples were drawn just before and at 15, 30, 45, 60, 90 and 120 min after starting to ingest oral
glucose (50g anhydrous glucose dissolved in 250ml bottled water). For type 2 diabetic patients,
venous blood samples were drawn at 0, 30, 60, 90, 120 and 180 min. The oral glucose test was
done 3 times over 3 month in each subject because OGTT-derived indices of β-cell function and
insulin sensitivity exhibit high within-subject variability (344). Use of the mean of the 3 tests
reduces within-subject variability and increases the power of the study (263).
109
Blood samples were collected within 2-hours period for the healthy, non-diabetic subjects
and 3-hours for type 2 diabetic patients. In order to match the time period, only data from the 0-
120 min were used for statistical analysis.
6.3.3 Blood analysis
Venous blood samples for the measures of glucose, insulin and C-peptide were collected
in BD VacutainerTM
SSTTM
tubes (BD, Franklin Lakes, NJ, USA). Serum glucose was measured
by a glucose oxidase method (SYNCHRON LX Systems, Beckman Coulter, Brea, California,
USA), with inter-assay coefficient of variation (CV) of 1.9%. Insulin was measured using one-
step immunoenzymatic (―sandwich‖) assay (Beckman Access Ultrasensitive Insulin Assay,
Beckman Coulter, Brea, California, USA), with inter-assay CV of 2.5 to 4.3%. Insulin has no
cross-reactivity with proinsulin. C-peptide was measured using double antibody competitive
radioimmunoassay (Siemens Medical Solutions Diagnostics, Los Angeles, CA), with inter-assay
precision of 10% or less.
Venous blood samples for GLP-1 were collected in BD VacutainerTM
EDTATM
tubes
(BD, Franklin Lakes, NJ, USA). The dipeptidyl peptidase-4 inhibitor (Linco Research, St.
Charles, Missouri) was added immediately (less than 30 seconds) after collection. Plasma GLP-
1 was measured by capturing the active GLP-1 from the sample by a monoclonal antibody
(specific binding to the N-terminal region) using GLP-1 (active) ELISA Kit (Linco Research, St.
Charles, Missouri, USA), with inter-assay CV ranges from 1-13%. All the samples were stored
at -70°C before analysis.
110
6.3.4 Calculations
Pre-hepatic insulin secretion rate (ISR) for each subject during the oral glucose tests was
calculated from deconvolution of the plasma C-peptide concentration using ISEC software
package developed by Hovorka et al (269). Incremental area under the curve (AUC) for plasma
glucose, insulin, C-peptide, insulin secretion rate, and GLP-1 was calculated by applying the
trapezoid rule, with determination of the incremental area above baseline (31). Hepatic insulin
extraction (HIEauc) was determined by the AUC of C-peptide divided by that of insulin (268).
Insulin sensitivity was calculated using two validated OGTT-derived indices of insulin
sensitivity: 1) Matsuda‘s insulin sensitivity index, which incorporates both hepatic and muscle
components of insulin resistance, correlates well with euglycemic-hyperinsulinemic clamp, and
was calculated as 10,000/(fasting glucose × fasting insulin × mean glucose × mean insulin
)0.5
(173); 2) Mari‘s glucose-insulin model-based oral glucose insulin sensitivity index
constructed on established principles of glucose kinetics and insulin action (174). Oral glucose
insulin sensitivity has been validated against the euglycemic-hyperinsulinemic clamp in healthy,
obese and type 2 diabetic subjects (174).
β-cell function was calculated using three validated methods: 1) the insulin
secretion/insulin resistance (disposition) index (270), which was shown to be the best predictor
of future development of type 2 diabetes in subjects with NGT compared with other predictive
models such as San Antonio Diabetes Prediction Model (271) (including age, sex, ethnicity,
BMI, blood pressure, fasting plasma glucose, triglycerides, and HDL) and 2-hour plasma glucose
concentration. To calculate insulin secretion/insulin resistance (disposition) index, first, insulin
secretion was expressed as the changes in AUC of insulin secretion rate (∆ISRAUC) relative to
changes in AUC of plasma glucose response (∆ISRAUC/∆GAUC ), then ∆ISRAUC/∆GAUC is divided
111
by the severity of insulin resistance (∆ISRAUC/∆GAUC÷ IR), as measured by the inverse of
Matsuda‘s insulin sensitivity index. 2) Disposition index calculated as total AUCins/gluc ×
Matsuda‘s insulin sensitivity index. This method is considered valid for oral glucose based
measure of β-cell function because a hyperbolic relationship was demonstrated between OGTT-
derived AUCins/gluc and Matsuda‘s insulin sensitivity index across a range of glucose tolerance
(345). 3) Disposition index calculated as insulinogenic index (ΔI0–30/ΔG0–30) × 1/fasting insulin.
The insulinogenic index (ΔI0–30/ΔG0–30) also demonstrated a hyperbolic relation with 1/fasting
insulin, and the product of the two variables is predictive of development of diabetes for over 10
years (346).
6.3.5 Statistical Analysis
Data were expressed as mean ± SEM for normally distributed continuous variables or
median (interquartile range) for non-normally distributed continuous variables. The normality of
the variables was assessed using the Shapiro and Wilk statistic and the normality plots (PROC
UNIVARIATE procedure of SAS). The skewed variables were transformed into their natural
logarithms before being subjected to statistical analysis.
Indices of insulin sensitivity, β-cell function, hepatic insulin extraction and integrated
responses of biomarkers were analyzed with an ANCOVA model with the subject group as fixed
effect, and age, BMI and WC as the covariates. Tukey‘s post hoc correction was performed to
compare the variables among the 4 subject groups. Multiple regression analysis was carried out
with the plasma GLP-1 concentration as the dependent variable and age, BMI, oral glucose
insulin sensitivity, disposition index (∆I0-30/∆G0-30) × 1/FSI), hepatic insulin extraction, and
AUCs of glucose and insulin as independent variables. Collinearity was determined by including
112
variance inflation factor in the model, with variance inflation factor of 5 or 10 and above
indicates a multicollinearity problem (299). No collinearity was apparent for the variables
included in the regression analysis. All analyses were done using SAS 9.2, (SAS Institute Inc,
Cary, NC, USA). Differences were considered significant if 2-tailed p < 0.05. The graphs were
plotted using GraphPad Prim, version 5 (GraphPad Software, Inc, California, USA).
6.4 Results
Type 2 diabetic patients and the subjects with high-FSI were older, and had significantly
higher BMI, WC, and waist:hip ratio than their counterparts in the low- and medium-FSI groups
(Table 6.1). Among the non-diabetic subjects (low-, medium- and high-FSI groups), fasting
insulin and C-peptide increased progressively as the FSI of the subjects increased; however, their
fasting and 2-hour glucose were not significantly different (Table 6.1). Type 2 diabetic patients‘
fasting insulin and C-peptide were comparable to either the medium or high-FSI group; however,
as expected, they had significantly elevated fasting and 2-hour postprandial glucose
concentrations (Table 6.1). Basal insulin secretion rate (ISRbasal) significantly increased as the
fasting insulin of the subject increased, and type 2 diabetic patients had ISRbasal comparable to
those of the medium- and high-FSI groups. After adjusting for fasting glucose, ISRbasal was the
highest in the high-FSI group. Fasting GLP-1 was not significantly different among the 4 groups
(Table 6.1). Systolic and diastolic blood pressure and triglycerides were not significantly
different among the low-, medium- and high-FSI groups; however, these parameters were
significantly higher in type 2 diabetic patients (Table 6.1). Total cholesterol, HDL and LDL
cholesterol, total:HDL cholesterol ratio and C-reactive protein were not significantly different
among the 4 groups (Table 6.1).
113
The glucose AUC was significantly higher in the high-FSI group and type 2 diabetic
patients compared to that of the low- and medium-FSI groups. As expected, the AUC of insulin
and C-peptide increased as subjects‘ fasting insulin increased, and type 2 diabetic patients had
the lowest insulin and C-peptide concentrations (Table 6.2). The postprandial insulin secretion
rate (ISR AUC) was significantly higher in the medium and high-FSI groups than that of the
low-FSI group and type 2 diabetic patients (Table 6.2). In addition, it was positively related to
basal insulin secretion rate (adjusted for fasting glucose) (r=0.70, p<0.0001) (Figure 6.1).
The postprandial GLP-1 response (GLP-1 AUC) was not significantly different among
the subject-groups (Table 6.2, Figure 6.2). Gender difference in the GLP-1 response after oral
glucose ingestion was observed: plasma GLP-1 concentration was more than 2 times greater in
female than that of the male subjects (294±52 pmol × min/L vs. 131±23 pmol × min/L;
p<0.0001).
Oral glucose insulin sensitivity decreased in a step-wise fashion across the groups, with
type 2 diabetic patients the least insulin sensitive (Table 6.3). All three calculations of β-cell
function decreased across the groups, with type 2 diabetic patients exhibiting the most
compromised β-cell function (Table 6.3). Even among the healthy, non-diabetic subjects, insulin
secretion/insulin resistance (disposition) indices (∆ISRAUC/∆GAUC ÷ IR) significantly decreased
as the subjects‘ fasting insulin increased, highlighting the reduced ability of pancreatic β-cells to
compensate for insulin resistance even in subjects with normal glucose tolerance.
Type 2 diabetic patients had significantly higher hepatic insulin extraction compared to
the low- and medium-FSI subjects (Table 6.3), which partially explains their very low
postprandial insulin concentration (Table 6.2). Hepatic insulin extraction was inversely related
114
to postprandial insulin concentration (r = -0.76, p<0.0001) but positively associated with age
(r=0.62, p<0.0001) (Figure 6.3).
Plasma GLP-1 response was inversely related to BMI (r=-0.33, p=0.03) and WC (r=-
0.28, p=0.05) respectively (Figure 6.4). However, no correlation was observed between the
GLP-1 response and oral glucose insulin sensitivity and all three calculations of β-cell function
(Figure 6.5). There was still no significant correlation even in the non-diabetic subjects alone
(Figure 6.6) or in the T2DM subjects alone (Figure 6.7). Plasma GLP-1 response was not
associated with hepatic insulin extraction either (r=-0.02, p=0.88).
Multivariate regression analysis show that age, BMI and disposition index (∆I0-30/∆G0-30)
× 1/FSI) explained 16% of the variation in GLP-1 response (GLP-1response = 218 +5.2×age -
10.5×BMI +35.2×disposition index, r2=0.16, p=0.06) after oral glucose ingestion.
115
Table 6.1 Anthropometric and metabolic characteristics of the study groups1
Low (n=13) Medium (n=16) High (n=9) T2DM (n=10) P2
Age, y 24(23-35)a,3
24(22-33)a 40(34-41)
b 57(53-58)
c <0.0001
Gender (M:F), n:n 5:8 9:7 4:5 5:5
BMI, kg/m2 22(21-23)
a 24(22-30)
a 32(29-32)
b 32(29-34)
b <0.0001
WC, cm 73(71-77)a 82(77-92)
a 99(97-108)
b 102(94-115)
b <0.0001
Waist:Hip Ratio 0.77±0.02a,4
0.80±0.02a 0.90±0.03
b 0.93±0.02
b <0.0001
FSI, pmol/L 27(25-32)a 47(43-55)
b 90(75-107)
c 62(47-73)
b <0.0001
FC-peptide, pmol/L 310±33a 511±41
b 897±67
c 786±33
c <0.0001
FG, mmol/L 4.5(4.4-4.7)a 4.7(4.7-5.1)
a 4.9(4.9-5.4)
a 9.8(7.2-11.3)
b <0.0001
2h glucose, mmol/L 6.1(5.6-6.5)a 6.4(5.7-6.9)
a 7.6(7.1-7.8)
a 13.1(10.2-15.1)
b <0.0001
ISRbasal, pmol/kg/min 1.13±0.12a 1.89±0.16
bc 2.62±0.17
d 2.36± 0.10
cd <0.0001
ISRbasal/FG 0.21(0.18-0.29)a 0.37(0.3-0.5)
bc 0.50(0.43-0.63)
b 0.26(0.19-0.33)
ac <0.0001
Fasting GLP-1, pmol/L 2.7(2.2-4.8) 3.2(2.7-5.3) 2.1(1.9-5.1) 3.1(2.4-4.8) 0.67
SBP, mm Hg 107±4a 110±3
a 115±2
ab 130±4
b 0.0004
DBP, mm Hg 68±4ab
66±2a 74±3
ab 79±3
b 0.016
TC, mmol/L 4.7±0.3 4.4±0.2 4.8±0.4 5.5±0.5 0.09
TG, mmol/L 1.00(0.67-1.48)ab
0.88(0.66-1.23)a 1.03(0.74-1.67)
ab 1.49(1.20-1.86)
b 0.04
HDL-C, mmol/L 1.19±0.08 1.20±0.07 1.13±0.06 1.20±0.10 0.93
LDL-C, mmol/L 2.95±0.22 2.74±0.19 3.07±0.32 3.10 ±0.33 0.70
TC: HDL 4.5(2.8-4.8) 3.4(3.2-3.8) 4.1(3.1-4.8) 3.9(3.3-6.6) 0.36
CRP (mg/L) 0.9(0.4-2.3) 1.0(0.3-2.4) 2.3(0.3-3.5) 3.7(1.1-5.4) 0.30
116
1 n = 48 except where otherwise noted. low (n=13, FSI < 40 pmol/L), medium (n=16, 40 ≤ FSI <
70 pmol/L), high (n=9, FSI ≥ 70 pmol/L), and type 2 diabetic patients (T2DM) (n=10). WC,
waist circumference; FSI, fasting serum insulin; FC-peptide, fasting C-peptide; FG, fasting
glucose; ISR, insulin secretion rate; SBP, systolic blood pressure; DBP, diastolic blood pressure;
TC, total cholesterol; TG, triglycerides; HDL, high-density-lipoprotein; LDL, low-density-
lipoprotein; CRP, C-Reactive Protein. Values in the same row with different superscript letters
are significantly different, P < 0.05 (Tukey‘s post hoc test). The unit for ISRbasal/FG is
pmol/mmol/kg/min.
2 P represents overall significant differences across groups by one-factor ANOVA.
3 Median; interquartile range in parentheses (all such values).
4 Mean ± SEM (all such values).
117
Table 6.2 The AUCs of glucose, insulin, C-peptide, insulin secretion rate (ISR) and GLP-1 in
the study groups1
Low
(n=13)
Medium
(n=16)
High
(n=9)
T2DM
(n=10)
P2
Glucose AUC 225(140-272)a,3
209(168-276)a 359(260-387)
b 678(597-749)
c <0.0001
Insulin AUC 37±6ab,4
48±4a
56±8a 12±2
b 0.007
C-peptide AUC 77±8a 116±8
b 146±13
b 71±9
a <0.0001
ISR AUC 351±41a 485±40
b 579±48
b 376±24
a 0.0002
GLP-1 AUC 195(82-445) 156(104-365) 163(56-184) 164(80-281) 0.60
1 n = 48 except where otherwise noted. low (n=13, FSI < 40 pmol/L), medium (n=16, 40 ≤ FSI <
70 pmol/L), high (n=9, FSI ≥ 70 pmol/L), and type 2 diabetic patients (T2DM) (n=10). AUC:
area under the curve. The units for AUCs of glucose, insulin, C-peptide, ISR and GLP-1 are
mmol ×min/L, nmol×min/L, nmol×min/L, pmol/kg, and pmol ×min/L respectively. Values in
the same row with different superscript letters are significantly different, P < 0.05 (Tukey‘s post
hoc test).
2 P represents overall significant differences across groups by repeated measures ANOVA
(PROC-MIXED, adjusted for age, BMI and WC). Data were calculated from the mean of 3 oral
glucose tests.
3 Median; interquartile range in parentheses (all such values).
4 Mean±SEM (all such values).
118
Table 6.3 Insulin sensitivity, β-cell function and hepatic insulin extraction in the study groups1
Low
(n=13)
Medium
(n=16)
High
(n=9)
T2DM
(n=10)
P2
Insulin sensitivity
OGISMari (ml min-1
m-2
) 505(485-509)a, 3
465(433-498)a 375(367-378)
b 294(250-369)
c <0.0001
β-cell function
∆ISRAUC/∆GAUC ÷ IR 54.6±4.3a,4
38.3±4.4b 16.8±2.1
c 13.6±1.7
c <0.0001
(∆I0-30/∆G0-30)× 1/FSI ([mmol]-1
) 2.6±0.3a 2.9±0.3
a 1.8±0.3
ab 0.4±0.1
b 0.006
TAUCins/glu×ISIMatsuda ([mmol]-2
) 818±53a 807±61
a 536±62
b 141±28
c <0.0001
Hepatic insulin extraction 2.8(1.6-3.3)a 3.2(1.7-3.5)
a 2.8(2.5-3.6)
ab 5.7(4.7-8.3)
b 0.04
1 n = 48 except where otherwise noted. Low (n=13, FSI < 40 pmol/L), medium (n=16, 40 ≤ FSI
< 70 pmol/L), high (n=9, FSI ≥ 70 pmol/L), and type 2 diabetic patients (T2DM) (n=10). OGIS,
oral glucose insulin sensitivity; IR, insulin resistance; FSI, fasting serum insulin; TAUC, total
AUC; ISI, insulin sensitivity index. The unit for ∆ISRAUC/∆GAUC ÷ IR5
is [mmol]-2
kg-1
min-1
L.
Values in the same row with different superscript letters are significantly different, P < 0.05
(Tukey‘s post hoc test).
2 P represents overall significant differences across groups by repeated measures ANOVA
(PROC-MIXED, adjusted for age, BMI and WC). Data were calculated from the mean of 3 oral
glucose tests.
3Median; interquartile range in parentheses (all such values).
4 Mean±SEM (all such values).
119
0.2 0.4 0.6 0.80
200
400
600
800
1000
r=0.70p<0.0001
ISRbasal
(pmol /kg/ min)
ISR
AU
C(p
mo
l/kg
)
Figure 6.1 The correlation between basal insulin secretion rate (ISRbasal, adjusted for fasting
glucose) and postprandial insulin secretion rate (ISRauc). r, Pearson‘s correlation coefficient.
The line is regression line, n=48.
120
0 30 60 90 120
4
8
12
16High FSI
Low FSI
Medium FSI
T2DM
Time (min)
GL
P-1
Re
sp
on
se
(pm
ol/L
)
Figure 6.2 Mean (±SEM) postprandial plasma GLP-1 response to oral glucose in type 2
diabetic patients (solid circle) and subjects with different levels of fasting serum insulin (FSI):
low (open triangle, FSI < 40 pmol/L), medium (open circle, 40 ≤ FSI < 70 pmol/L) and high
(open square, FSI ≥ 70 pmol/L). Data were calculated from the mean of 3 oral glucose tests.
121
0 20 40 60 80 100
4
8
12
r=-0.76 p<0.0001
InsulinAUC 0-120min
(nmol/L x min)
C-p
eptide/Insulin
Ratio
(AU
C0-1
20m
in)
10 20 30 40 50 60 70
4
8
12
r=0.62p<0.0001
age (yrs)
C-p
eptid
e/In
sulin
Ra
tio
(AU
C0-1
20m
in)
Figure 6.3 The correlations between hepatic insulin extraction (ratio of C-peptide AUC and
insulin AUC) and postprandial insulin response (top) and age (bottom) respectively. r: Pearson‘s
correlation coefficient. The lines are regression line, n=48.
122
60 80 100 120
200
400
600
800
r = -0.28p = 0.05
Waist Circumference (cm)
GLP
-1 R
esponse
(pm
ol
min
/L)
15 20 25 30 35 40
r = -0.33p=0.03
BMI
(kg/m2)
Figure 6.4 The correlations between GLP-1 response after ingestion of 50g oral glucose and
waist circumference (WC) and BMI respectively, in type 2 diabetic patients (solid circle, n=10)
and subjects with different levels of fasting serum insulin (FSI): low (open triangle, n=13, FSI <
40 pmol/L), medium (open circle, n=16, 40 ≤ FSI < 70 pmol/L) and high (open square, n=9, FSI
≥ 70 pmol/L). Dashed line indicates the upper and lower 95% CI. r, Pearson‘s correlation
coefficient. The lines are regression line, n=48.
123
40
80
120r=0.17p=0.26
IS
R(A
UC
)/
G(A
UC
)÷ IR
0 200 400 600 800
500
1000
1500
r=0.15p=0.30
GLP-1 AUC(pmol x min/L)
Tota
l A
UC
ins/g
lu ×
ISI
0 200 400 600 8000
2
4
6
r=0.17 p=0.26
GLP-1 AUC(pmol x min/L)
(I0
-30/
G0-3
0)×
1/F
SI)
200
400
600
r = 0.16p = 0.29
OG
IS
(ml/m
in/m
2)
Figure 6.5 The correlations between GLP-1 response after ingestion of 50g oral glucose and
oral glucose insulin sensitivity (OGIS) and markers of β-cell function in type 2 diabetic patients
(solid circle, n=10) and subjects with different levels of fasting serum insulin (FSI): low (open
triangle, n=13, FSI < 40 pmol/L), medium (open circle, n=16, 40 ≤ FSI < 70 pmol/L) and high
(open square, n=9, FSI ≥ 70 pmol/L). Dashed line indicates the upper and lower 95% CI. r,
Pearson‘s correlation coefficient. The lines are regression lines. n=48.
124
40
80
120r=0.19p=0.26
IS
R(A
UC
)/
G(A
UC
)÷ IR
0 200 400 600 800
500
1000
1500
r=0.20p=0.22
GLP-1 AUC(pmol x min/L)
Tota
l A
UC
ins/g
lu ×
ISI
0 200 400 600 8000
2
4
6
r=0.19 p=0.26
GLP-1 AUC(pmol x min/L)
(I0
-30/
G0-3
0)×
1/F
SI)
250
350
450
550
r = 0.31p = 0.06
OG
IS
(ml/m
in/m
2)
Figure 6.6 The correlations between GLP-1 response after ingestion of 50g oral glucose and
oral glucose insulin sensitivity (OGIS) and markers of β-cell function in non-diabetic subjects
with different levels of fasting serum insulin (FSI): low (open triangle, n=13, FSI < 40 pmol/L),
medium (open circle, n=16, 40 ≤ FSI < 70 pmol/L) and high (open square, n=9, FSI ≥ 70
pmol/L). Dashed line indicates the upper and lower 95% CI. r, Pearson‘s correlation coefficient.
The lines are regression lines. n=38.
125
0
5
10
15
20
25
r=-0.24p=0.51
ISR
(AU
C)/
G(A
UC
)÷ IR
0 100 200 300 4000
100
200
300
400r=-0.16p=0.65
GLP-1 AUC(pmol x min/L)
To
tal A
UC
ins/g
lu ×
ISI
0 100 200 300 400
0.5
1.0
1.5
r=0.04p=0.91
GLP-1 AUC(pmol x min/L)
(I0
-30/
G0
-30
)× 1
/FS
I)
100
300
500
r=-0.35p=0.31
OG
IS
(ml/m
in/m
2)
Figure 6.7 The correlations between GLP-1 response after ingestion of 50g oral glucose and
oral glucose insulin sensitivity (OGIS) and markers of β-cell function in type 2 diabetic patients.
Dashed line indicates the upper and lower 95% CI. r, Pearson‘s correlation coefficient. The lines
are regression lines. n=10.
126
6.5 Discussion and Conclusion
In this analysis combining 3 separate oral-glucose-test data from two clinical studies
(Chapter 4 & 5), we found that plasma intact GLP-1 concentration after oral glucose ingestion
was not significantly different among different subject groups. Moreover, no correlation was
found between intact GLP-1 response and insulin sensitivity, β-cell function and hepatic insulin
extraction, the classic triad of type 2 diabetes.
Our finding that the GLP-1 response after oral glucose in T2DM subjects was not
significantly different from that of the healthy, non-diabetic subjects was in agreement with some
studies. For example, Theodorakis et al found that plasma GLP-1 concentration actually
increased between 20-80 min in newly diagnosed patients with type 2 diabetes (319) and
Vollmer et al found no decrease in GLP-1 levels in type 2 diabetic patients after oral glucose or
mixed meal ingestion (318). However, Toft-Nielsen et al reported approximately 30% lower
postprandial GLP-1 levels in type 2 diabetic patients compared with normal oral glucose tolerant
subjects (347), and Vilsboll et al found that both total and intact GLP-1 after a mixed meal were
markedly reduced in type 2 diabetic patients (251). The discrepancies in findings may be
attributable to the different levels of metabolic aberrations (i.e. degree of hyperglycemia,
duration of diabetes) amongst the patients in the different studies. In both Theodorakis and
Vollmer‘s studies, the patients were either newly diagnosed diabetics or in relatively good
glycemic control, with HbA1C of 6.8±0.9% (318) and 7.0±0.2% (mean±SEM) (319)
respectively. In contrast, the patients studied by Vilsboll et al and Toft-Nielsen et al exhibited
much poorer glycemic control, with HbA1C of 8.4±1.7% (mean±SEM) (347) and 9.2% (range
7.0-12.5) (251). In our study, the patients had had a relatively good glycemic control, with
127
HbA1C of 7.3± 0.3% (mean ±SEM), which may explain the lack of significant difference in
GLP-1 levels between the diabetic patients and non-diabetic subjects.
We also found that the AUCs of postprandial intact GLP-1 response was not
significantly different among the low-, medium-, and high-FSI groups and was not related to the
insulin sensitivity of the subjects. Previous studies have shown that the postprandial GLP-1
secretion after both meal (331) and oral glucose (332) was not significantly different between
first degree relatives of type 2 diabetes and the controls. And in women with a history of
gestational diabetes, the GLP-1 response after oral glucose was not different from that of the
controls (333). Our data confirms that the degree of insulin sensitivity per se does not determine
GLP-1 response, and the reduction in GLP-1 does not precede diabetes. This further suggests
that the reduction in GLP-1 response as observed in long-term diabetic patients with poorly
controlled glycemia (251, 347) may not be the cause but the consequence of the diabetic state.
The postprandial intact plasma GLP-1 levels were not associated with any of the indirect
measures of β-cell function (Figure 6.5). This is not expected because there is a biological basis
for the positive effect of GLP-1 on β-cell function. In cell culture studies or rodent models of
diabetes, GLP-1 acts on β-cells acutely to enhance insulin biosynthesis and stimulate insulin
gene transcription (348) and chronically, stimulates β-cell proliferation (335, 336), and inhibits
β-cell apoptosis, thus promoting expansion of β-cell mass (337, 338). In addition, there is some
indirect clinical evidence that GLP-1 may have β-cell trophic effect in vivo in humans. For
example, patients with clinically significant hypoglycemia after gastric bypass surgery exhibited
exaggerated GLP-1 and insulin secretory responses to a mixed meal (349). Since symptomatic
hypoglycemia in post-gastric bypass patients only develops after 1-5 y, it was suggested that
GLP-1 may have β-cell trophic effects and may increase β-cell mass over time after surgery
128
(349). The other clinical evidence is demonstrated in the recent LEAD-3 study, patients with
early type 2 diabetes were treated with liraglutide (analogue of human GLP-1, with 97%
homology to the endogenous protein) (350) or glimepiride for 52 weeks (351). Ligraglutide
treated patients have greater reduction in HbA1C than glimepiride. The most important finding
is the sustained reduction in HbA1C at 52 weeks, indicating improved β-cell function (351).
The lack of association between GLP-1 response and the indirect measures of β-cell
function maybe due to the following reasons: first, we only measured the biologically active
intact GLP-1-(7-36 amide), which is only a small part of intestinal L-cell secretion, and does not
represent GLP-1 secretion (352). In addition, it does not reflect GLP-1 action since GLP-1
appears to act via sensory afferents before its degradation (352). A non-specific assay should be
used to measure total GLP-1 concentrations [active GLP-1-(7-36) amide + inactive GLP-1-(9-
36) amide], which reflects the secretary rate of the L cells (347). Second, β-cell function was
characterized by indirect calculations related to insulin secretion/insulin sensitivity in the basal
state and after stimulation with oral glucose. Under these circumstances, changes in glucagon,
glucose, free fatty acids, GLP-1, and GIP may determine insulin (and proinsulin) secretion (318).
To better characterize β-cell behavior, tests assessing the influence of single factors (i.e. glucose
bolus injection or hyperglycemic clamp) should be used. Third, the data is pooled from two
different studies, which are not designed to address the research question (are there any
association between GLP-1 response and the indirect measures of β-cell function?). This
obviously is the weakness of this analysis.
We did not find any association between hepatic insulin extraction and GLP-1 response
(r=-0.02, p=0.88), which is consistent with previous studies in humans (208, 343), suggesting
GLP-1 has no independent effect on liver insulin clearance. In support of this, there is evidence
129
that hepatocytes lack GLP-1 receptor expression (353); hence, it is unlikely that GLP-1 can exert
any direct influence on liver insulin clearance. However, a significant inverse association was
found between hepatic insulin extraction and insulin concentrations after oral glucose ingestion
(r = -0.76, p<0.0001). This result is consistent with our previously published data (354) and
Meier et al‘s study (208). These data suggest that insulin concentration itself may determine its
own rate of clearance from the liver.
Previous literature on hepatic insulin extraction in diabetic patients is inconsistent as both
normal (217, 218), decreased (219, 220), or even increased (221) hepatic insulin clearance had
been reported. We found that hepatic insulin extraction was significantly increased in type 2
diabetic patients compared to the non-diabetic subjects (Table 6.3). Three possible factors may
explain such discrepancies: first, the variations in liver fat content (222, 223) may contribute to
this. It is known that increased liver fat content is associated with impaired hepatic insulin
clearance (224). Second, the medications the patients took may affect hepatic insulin extraction.
For example, all our patients were on metformin. Metformin enhances hepatic insulin extraction
in non-diabetic subjects at high-risk of developing diabetes (234). Hence, it is possible that
metformin may also enhance hepatic insulin extraction in diabetic subjects. Third, the age of the
patients may have played a role. It has been reported that hepatic insulin clearance was
increased in elderly men and women (210-212). Indeed, a positive correlation was observed
between hepatic insulin extraction and age (r=0.62, p<0.0001), suggesting that the older the
subjects, the higher the liver insulin clearance.
We found that plasma GLP-1 concentration after oral glucose ingestion was 2 times
greater in female than male subjects. This finding is consistent with previous results (318, 347).
In women, delayed gastrointestinal (GI) motility (355), decreased gall bladder contraction (356),
130
and increased colonic transit and gastric emptying time (357) have all been reported. Therefore,
slower movement of foods through the GI tract may provide more contact time for the food
material with intestinal mucosa, which might lead to increased GLP-1 secretion.
We found that intact GLP-1 was inversely associated with the measures of adiposity
(BMI and WC); the higher the BMI and WC of the subjects, the lower the GLP-1 response
(Figure 6.4). This result is consistent with previous findings. Naslund et al showed that obese
subjects had an attenuated GLP-1 release in response to meals (358). Toft-Nielsen et al
demonstrated that BMI was negatively related to GLP-1 response (347), and Ranganath et al
found a pronounced attenuation of plasma GLP-1 secretion to oral carbohydrate in obese
subjects compared to lean subjects (359). It was suggested that increased proximal absorption
could hypothetically explain the decreased GLP-1 secretion with increasing obesity (360). The
rationale being that obese subjects may have increased proximal absorption rate (361), thus less
food will reach the distal intestine, where the GLP-1-producing L cells are most abundant. On
the other hand, the presence of nutrients in the proximal small intestine can stimulate GLP-1
release independently of the presence of nutrients within the ileum and colon (362).
In conclusion, the endogenous intact GLP-1 response to oral glucose was not related to
indirect measures of insulin sensitivity, β-cell function and hepatic insulin extraction.
131
CHAPTER 7
GENERAL DISCUSSION
132
7.1 General discussion
Controlling postprandial glucose excursion is essential in the prevention and management
of diabetes. Many substances in foods can affect postprandial glycemia; the major ones are
macronutrient (carbohydrate, fat and protein). Numerous studies have compared carbohydrate,
fat and protein metabolism in healthy and diabetic subjects; however, there are meager data on
the impact of insulin resistance or hyperinsulinemia on macronutrient metabolism. Our studies
are the first to address this issue by systematically comparing the postprandial metabolic
responses elicited by carbohydrate, fat and protein in healthy control, hyperinsulinemic and type
2 diabetic subjects. The results generated from these studies have important implications for
both diabetic patients and people at high risk for diabetes, namely those with insulin
resistance/hyperinsulinemia.
We found that the ability of fat and protein to suppress glucose response was not
influenced by the degree of insulin sensitivity of the subjects, and the mechanisms through which
fat and protein modulate glucose responses differ, specifically, fat had no effect on liver insulin
clearance, whereas protein significantly decreased it. For the GI values of various carbohydrate
foods, we found that they were similar in healthy control, hyperinsulinemic and type 2 diabetic
subjects, suggesting that carbohydrates induce similar increases in relative glycemic response
independent of the subjects‘ metabolic status. Currently, people with diabetes are the primary
users of GI. Our finding suggests that GI is also a useful tool in the dietary management of
obese and/or non-diabetic, insulin resistant subjects. For the Insulinemic Index of carbohydrate
foods, we found that the mean Insulinemic Index of the 5 carbohydrate foods was much higher
for T2DM (100±7) than healthy controls (78±5) and hyperinsulinemic subjects (70±5)
(mean±SEM, p=0.05). Moreover, Insulinemic Index was inversely associated with the insulin
133
sensitivity and positively related to the severity of glycemia and hepatic insulin extraction of the
subjects, suggesting that Insulinemic Index is not a property of foods but depends on the
metabolic status of the subjects in whom it is measured; therefore, unlike GI, Insulinemic Index
may has limited clinical utility.
We hypothesized that adding fat and protein to carbohydrate would reduce glucose
responses less in non-diabetic hyperinsulinemic subjects than healthy control subjects, which is
likely due to reduced GLP-1 secretion and insulin sensitivity. However, contrary to this
hypothesis, the glucose-lowering effect of fat and protein was not dependent on insulin
sensitivity of the subjects, and the hypoglycemic effect of protein is equally potent in all subject
groups (low-, medium-, and high-FSI). This could be that GLP-1 response to fat or protein was
similar in subjects with low, medium or high-FSI (Figure 4.1, Appendix 4 & 6), thus the GLP-1
mediated insulin secretion after fat or protein may be similar as well in all subject groups.
Indeed, fat and protein did not appear to affect insulin secretion differentially in the 3 subject
groups as there were no significant FSI × fat and FSI × protein interaction effects on both C-
peptide (true marker of insulin secretion) and insulin secretion rate (ISR), and C-peptide
response and ISR were similar in all 3 subject groups (Figure 4.3).
Besides GLP-1, other gut hormones such as cholecystokinin (CCK) and PYY may also
play a role in modulating postprandial glycemia. CCK and PYY were not measured in this
study, but it is known that fat and protein are the most important stimulators of CCK secretion
(363), and dietary protein is the most potent stimulant of PYY release (288). In subjects with
insulin resistance, both CCK and PYY responses to meal may be altered. It was found that
healthy female first-degree relatives of people with type 2 diabetes had significantly lower
fasting serum PYY levels than controls, but their PYY response to a high fat meal was not
134
significantly different from the controls (364). For CCK, the plasma CCK response after glucose
drink was higher in hyperinsulinemic men compared to the control (365).
The regulation of glucose homeostasis after a meal involves not only β-cell function,
hepatic insulin extraction and hormones of the entero-insulin axis, but also the newly discovered
pathway of the upper-intestinal-lipids-activated gut-brain-liver neural axis, through which liver
insulin sensitivity was enhanced, and hepatic glucose production was reduced (39). However,
the gut-brain-liver neural pathway may be impaired in insulin resistant state. Indeed, in high-fat-
diet-induced insulin resistant rats, intraduodenal lipid infusion failed to suppress hepatic glucose
production, suggesting that insulin resistance may dampen the ability of the gut-brain-liver
neuronal network to sense and respond to lipid, thus resulting in impaired glucose regulation
(39). Based on these findings, one might logically assume, therefore, that insulin resistant
humans may also acquire such defects in hypothalamic nutrient sensing; however, such
hypothesis is extremely difficult to prove in humans. This is because the hypothalamus is
located deep in the midbrain, and even techniques such as computed tomography scans and
magnetic resonance imaging do not have sufficient spatial resolution to detect changes in
specific nuclei (366). On the other hand, our finding that the glucose-lowering effect of fat is not
influenced by insulin sensitivity of the subjects indirectly suggests that unlike those observed in
animal models, the lipid-activated gut-brain-liver neural circuit may not be impaired in insulin
resistant humans.
Then, what is the implication that the ability of fat and protein to suppress the glycemic
response to oral glucose is not altered in the presence of insulin resistance/hyperinsulinemia?
This suggests that for the dietary management of insulin resistance/hyperinsulinemia, there may
not be a need for a specifically designed diet since the glucose-lowering effects of fat and protein
135
are not influenced by the degree of insulin sensitivity of the subjects. On the other hand,
functionality in terms of lowering glucose alone should not be the sole determining factor of
macronutrient composition of a diet for insulin resistant humans. The effects on lipids,
inflammatory markers (i.e. CRP, TNFα, interleukin-6, etc), food intake, and body weight
regulation also need to be taken into consideration.
It is commonly held that the ability of protein to increase insulin concentrations is
primarily due to amino acid mediated insulin secretion (127, 156). We have identified a novel,
alternative pathway through which protein mediates peripheral insulin concentration. We found
that whey protein increased insulin but had no effect on C-peptide or insulin secretion rate,
whereas hepatic insulin extraction was decreased. We proposed that whey-induced
hyperinsulinemia was probably due to a decrease in hepatic insulin extraction (thus conserves
insulin) rather than increased secretion. Then, how does whey protein decrease hepatic insulin
extraction? It is known that both amino acids and the biologically active (bioactive) peptides
released after protein digestion regulate physiological functions (367) and whey proteins are
precursors of many bioactive peptides encrypted in their amino acid sequences (292). Thus, it is
hypothesized that the effects on hepatic insulin extraction may be executed by certain branched-
chain amino acids (BCAA) (i.e. leucine, isoleucine, valine, lysine and threonine), or whey-
derived bioactive peptides (Appendix 7), or synergistic actions among them, which exert their
function by acting as competitive inhibitors for insulin receptors on hepatocytes; thus less insulin
is cleared by liver.
Proteins are highly heterogeneous in physical-chemical properties and biological
functions. For example, varying sources of protein differ in the rate of digestion and absorption
(i.e. whey is a fast protein, whereas casein is a slow protein). Moreover, different sources of
136
protein differ in the composition of amino acids (153) and bioactive peptides (368). All these
differences in physical-chemical properties likely will contribute to their different physiological
effects. Therefore, our finding of whey-induced reduction in liver insulin clearance does not
necessary imply that other proteins such as casein or soy protein also has such an effect.
Previous study has shown that in healthy adults, the GI and Insulinemic Index of
carbohydrate foods are well correlated (r=0.69, p<0.001) (29). We found that GI and
Insulinemic Index of carbohydrate foods are well correlated not only in healthy subjects, but also
in subjects with hyperinsulinemia and type 2 diabetes, suggesting that the ability of GI to reliably
predict insulin demand after carbohydrate foods is independent of subject‘s metabolic status.
This finding expands our knowledge of insulin response to carbohydrate foods in subjects with
varying metabolic states.
7.2 Weaknesses
There are a number of limitations to the studies. Due to the expense and technical
difficulties associated with performing the euglycemic-hyperinsulinemic clamp, we chose
FSI>40pmol/L (based on pilot data) instead as indicative of insulin resistance/hyperinsulinemia.
Using fasting insulin as the surrogate measure of insulin sensitivity and arbitrarily defining
FSI>40pmol/L as the cut-off point posed the risk of misclassifying subjects. In the future, if
similar studies were to be carried out, oral glucose insulin sensitivity would be a better candidate
as an indirect measure of insulin sensitivity because it has been found to be the most accurate
surrogate of M values derived from the clamp method (369). Nevertheless, this shortcoming
does not negate the conclusion that the hypoglycemic effect of fat and protein was not attenuated
by insulin resistance. This is because the regression analysis showed that the changes in relative
137
glycemic response per gram of fat or protein were not related to fasting insulin. Through
regression analysis, we treated the subjects as continuous variables instead of as categorical
variables (subjects group), thus avoiding the bias that may be brought about by the potential risk
of misclassification of subjects.
In the first study (Chapter 4), the subjects with different levels of fasting insulin were not
matched for age and adiposity (BMI & WC). In the second study (Chapter 5), the healthy
control, hyperinsulinemic and type 2 diabetic subjects were also not matched for age, BMI and
WC. This issue was addressed by adjusting the covariates (age, BMI and WC) using analysis of
covariance (ANCOVA).
In the second study, the blood sampling schedule for the non-diabetic subjects (healthy
controls and hyperinsulinemic subjects) was 2 hours; however, for the type 2 diabetic subjects, it
was 3 hours. In future, if a similar study were to be carried out, the blood sampling schedules
should be the same for all the subject groups. Blood sampling time should also be extended to 3
hours as it takes longer for insulin and C-peptide to return to baseline.
7.3 Strengths
Our studies systematically investigated whether metabolic responses to macronutrient are
different among healthy control, hyperinsulinemic and diabetic subjects, which clearly is a
strength as previous studies only focused on the comparison between the healthy and diabetic
subjects. In addition, we examined the effects of macronutrient, both singly and in combination,
on postprandial glucose and insulin responses. Contrary to the results from animal studies, we
found that in humans, the ability of fat to suppress postprandial glycemia after oral glucose
ingestion is not influenced by insulin sensitivity of the subjects. Another novel finding is that
138
protein decreases hepatic insulin extraction whereas fat has no such effect, demonstrating that fat
and protein modulate postprandial metabolic responses through different mechanisms.
We are also the first to evaluate both the GI and Insulinemic Index of different
carbohydrate foods in healthy control, hyperinsulinemic and type 2 diabetic subjects. These
findings affirm the clinical utility of GI in the dietary prevention and management of type 2
diabetes. Furthermore, it suggests that GI is a useful tool for people with insulin
resistance/hyperinsulinemia in guiding their choices of carbohydrate foods. However, unlike GI,
insulinemic index is not a property of food because it varies depending on insulin sensitivity, the
severity of glycemia and the hepatic insulin extraction of the subjects. Thus, it may have limited
therapeutic utility. This finding is important as the nutrition community needs this type of
information to determine whether food insulin index is a valid and useful tool in clinical settings
and in epidemiological research.
In order to evaluate pancreatic β-cell insulin secretion following ingestion of different
macronutrient, we measured not only insulin but also C-peptide (the true secretion of insulin)
and gut hormone GLP-1. This is an important strength as these biomarkers assist in better
understanding of the complex factors involved in altering insulin secretion after food digestion
and absorption.
139
CHAPTER 8
FUTURE DIRECTIONS
140
Based on the results of current studies, a few potential promising areas for future studies
are identified: first, from a mechanistic point of view, it is necessary to find out how does whey
protein, specifically whey-derived amino acids/bioactive peptides influence liver insulin
clearance; second, it will be interesting to find out whether fat and protein differentially affect
hypothalamic nutrient sensing; third, since Insulinemic Index was inversely associated with oral
glucose insulin sensitivity, and oral glucose insulin sensitivity alone explained 43% of the
variation in Insulinemic Index (r2=0.43, p<0.001), it is necessary to compare Insulinemic Index
in different subject groups classified by oral glucose insulin sensitivity. Both mechanism-finding
studies (in animal models) and clinical studies are needed to address these issues:
In animal models:
1. Test the effects of whey-derived amino acids and bioactive peptides on hepatic insulin
extraction. Hepatic insulin extraction can be measured directly using hepatic vein
catheterization technique, an invasive method with catheters placed in artery and hepatic
veins.
2. The much more potent glucose-lowering effect of protein than fat suggests that other
factors besides insulin and gut hormones might play a role. Therefore, in animal models,
we can test whether whey protein regulates glucose homeostasis through hypothalamic
amino acid sensing (i.e. effect on hepatic glucose production) and compare the effects to
those of fat. Further, it is important to explore whether the hypothalamic protein-sensing
mechanism is impaired in insulin resistance and diabetes.
141
In humans:
1. Insulinemic Index is inversely associated with oral glucose insulin sensitivity; therefore,
it is important to find out whether Insulinemic Index values were similar in subjects with
varying degrees of insulin sensitivity classified by oral glucose insulin sensitivity.
2. In this study, only 5 carbohydrate foods were tested, thus, it is impossible to do a
correlation analysis to see if Insulinemic Index of healthy subjects were correlated with
those of the hyperinsulinemic or diabetic subjects. Future studies can include more
different sources of carbohydrate foods (i.e. 20 or more carbohydrate foods) and to see if
there were any correlation between Insulinemic Index derived from the same
carbohydrate foods in healthy control vs. hyperinsulinemic, healthy control vs. T2DM, or
hyperinsulinemic vs.T2DM. If there were no correlation, this will further support our
finding that Insulinemic Index may not have relevance in dietary intervention in
hyperinsulinemic and diabetic subjects.
142
CHAPTER 9
CONCLUSIONS
143
Conclusions to the specific hypotheses are:
1. Adding fat and protein to oral glucose reduces postprandial glycemic responses to a
similar extent in subjects with low-, medium- or high-FSI, suggesting that the
hypoglycemic effect of fat and protein is not attenuated by insulin resistance.
2. The effects of macronutrient on postprandial responses are not affected by habitual
intakes of fat and dietary fibre. One possible explanation could be that the sample size is
most likely not adequate to address habitual diet.
3. The GI values of carbohydrate foods do not differ in healthy control, hyperinsulinemic
and type 2 diabetic subjects, which support the clinical utility of GI as a property of the
available carbohydrate in foods. The GI values of sucrose are also not significantly
different among the subject groups.
4. Unlike GI, Insulinemic Index values are dependent on the metabolic status of the
subjects, suggesting that Insulinemic Index may have limited clinical utility.
5. The endogenous GLP-1 response to oral glucose is not related to indirect measures of
insulin sensitivity, β-cell function and hepatic insulin extraction.
The overall conclusion of the thesis:
The ability of fat and protein to suppress postprandial glycemia to oral glucose is not
influenced by insulin sensitivity of the subjects. Moreover, fat and protein modulate glucose
response through different mechanisms. Protein reduces liver insulin clearance whereas fat
exhibits no such effect. The insulin-raising effect of protein may be due to reduced liver insulin
clearance (thus conserves insulin) rather than increased secretion; however, the exact
mechanisms of protein on liver insulin clearance warrant further investigation in animal models.
144
Carbohydrates induce similar increase in relative glycemic response (GI) in healthy
control, hyperinsulinemic and type 2 diabetic subjects. This finding supports the concept that GI
is the property of the available carbohydrate in foods rather than the metabolic characteristics of
the individual consuming the foods. On the contrary, relative insulin responses (Insulinemic
Index) are dissimilar in the three subject groups. Moreover, the associations between
Insulinemic Index and the subjects‘ metabolic status (insulin sensitivity, severity of glycemia and
hepatic insulin extraction) suggest that Insulinemic Index is not a property of foods but is
subject-dependent; therefore, unlike GI, Insulinemic Index may have limited clinical utility.
No association was found between endogenous GLP-1 concentration and insulin
sensitivity, β-cell function and hepatic insulin extraction. This may be because biologically
active GLP-1 was measured instead of total GLP-1concentration that represents GLP-1 secretion.
Moreover, insulin sensitivity, β-cell function and hepatic insulin extraction were characterized
using indirect calculations instead of direct measurements (i.e. euglycemic hyperinsulinemic
clamp for insulin sensitivity, hyperglycemic clamp for β-cell function, and hepatic vein
catheterization for hepatic insulin extraction).
145
CHAPTER 10
REFERENCES
146
References
1. Diagnosis and classification of diabetes mellitus. Diabetes Care 2011;34:S62-S69.
2. Coutinho M, Gerstein HC, Wang Y, Yusuf S. The relationship between glucose and incident
cardiovascular events: A metaregression analysis of published data from 20 studies of 95,783
individuals followed for 12.4 years. Diabetes Care 1999;22:233-40.
3. Saad MF, Knowler WC, Pettitt DJ, Nelson RG, Mott DM, Bennett PH. The natural history of
impaired glucose tolerance in the Pima Indians. New Engl J Med 1988;319:1500-6.
4. Balkau B, Shipley M, Jarrett RJ, et al. High blood glucose concentration is a risk factor for
mortality in middle-aged nondiabetic men: 20-year follow-up in the Whitehall Study, the Paris
Prospective Study, and the Helsinki Policemen Study. Diabetes Care 1998;21:360-7.
5. Gannon MC, Ercan N, Westphal SA, Nuttall FQ. Effect of added fat on plasma glucose and
insulin response to ingested potato in individuals with NIDDM. Diabetes Care 1993;16:874-80.
6. Van Loon LJC, Kruijshoop M, Menheere PPCA, Wagenmakers AJM, Saris WHM, Keizer
HA. Amino acid ingestion strongly enhances insulin secretion in patients with long-term type 2
diabetes. Diabetes Care 2003;26:625-30.
7. Rushakoff RA, Goldfine ID, Beccaria LJ, Mathur A, Brand RJ, Liddle RA. Reduced
postprandial cholecystokinin (CCK) secretion in patients with noninsulin-dependent diabetes
mellitus: Evidence for a role for CCK in regulating postprandial hyperglycemia. J Clin
Endocrinol Metab 1993;76:489-93.
8. English PJ. Evidence for PYY 3-36 resistance in Type 2 diabetes. Diabetic Med 2004;21:36-
108.
9. Ferrannini E. Is insulin resistance the cause of the metabolic syndrome? Ann Med 2006;38:42-
51.
10. Lillioja S. Insulin resistance and insulin secretory dysfunction as precursors of non- insulin-
dependent diabetes mellitus: Prospective studies of Pima Indians. New Engl J Med,
1993;329:1988-92.
11. DeFronzo RA. Pathogenesis of type 2 diabetes: Metabolic and molecular implications for
identifying diabetes genes. Diabetes reviews 1997;5:177-269.
12. Bergman RN. The evolution of β-cell dysfunction and insulin resistance in type 2 diabetes.
Eur J Clin Invest 2002;32:35-45.
13. Petersen K, Dufour S, Savage D, et al. The role of skeletal muscle insulin resistance in the
pathogenesis of the metabolic syndrome. Proc Natl Acad Sci U S A 2007;104:12587-94.
147
14. Reaven G. The insulin resistance syndrome: definition and dietary approaches to treatment.
Annu Rev Nutr 2005;25:391-406.
15. Jenkins DJA, Wolever TMS, Taylor RH. Glycemic index of foods: A physiological basis for
carbohydrate exchange. Am J Clin Nutr 1981;34:362-6.
16. Wolever TMS. Effect of blood sampling schedule and method of calculating the area under
the curve on validity and precision of glycaemic index values. Br J Nutr 2004;91:295-300.
17. Foster-Powell K, Holt SHA, Brand-Miller JC. International table of gylcemic index and
glycemic load values: 2002. Am J Clin Nutr 2002;76:5-56.
18. Wolever TMS, Jenkins DJA, Vuksan V, Josse RG, Wong GS, Jenkins AL. Glycemic index
of foods in individual subjects. Diabetes Care 1990;13:126-32.
19. Jenkins DJA, Wolever TMS, Jenkins AL. The glycaemic index of foods tested in diabetic
patients: A new basis for carbohydrate exchange favouring the use of legumes. Diabetologia
1983;24:257-64.
20. Wolever TMS, Jenkins DJA, Josse RG, Wong GS, Lee R. The glycemic index: Similarity of
values derived in insulin-dependent and non-insulin-dependent diabetic patients. J Am Coll Nutr
1987;6:295-305.
21. Wolever TMS, Jenkins DJA, Collier GR, et al. The glycaemic index: Effect of age in insulin
dependent diabetes mellitus. Diabetes Res 1988;7:71-4.
22. DeFronzo RA, Bonadonna RC, Ferrannini E. Pathogenesis of NIDDM: A balanced
overview. Diabetes Care 1992;15:318-68.
23. DeFronzo RA, Ferrannini E. Insulin resistance: A multifaceted syndrome responsible for
NIDDM, obesity, hypertension, dyslipidemia, and atherosclerotic cardiovascular disease.
Diabetes Care 1991;14:173-94.
24. Modan M, Halkin H, Almog S. Hyperinsulinemia. A link between hypertension obesity and
glucose intolerance. J Clin Invest 1985;75:809-17.
25. Despr s J-, Lamarche B, Mauri ge P, et al. Hyperinsulinemia as an independent risk factor
for ischemic heart disease. New Engl J Med 1996;334:952-7.
26. Nilsson P, Nilsson J-, Hedblad B, Eriksson K-, Berglund G. Hyperinsulinaemia as long-term
predictor of death and ischaemic heart disease in nondiabetic men: The Malmö Preventive
Project. J Intern Med 2003;253:136-45.
27. Takahashi F, Hasebe N, Kawashima E, et al. Hyperinsulinemia is an independent predictor
for complex atherosclerotic lesion of thoracic aorta in non-diabetic patients. Atherosclerosis
2006;187:336-42.
148
28. Lee BM, Wolever TMS. Effect of glucose, sucrose and fructose on plasma glucose and
insulin responses in normal humans: Comparison with white bread. Eur J Clin Nutr 1998;52:924-
8.
29. Miller JB, Pang E, Broomhead L. The glycaemic index of foods containing sugars:
Comparison of foods with naturally-occurring v. Added sugars. Br J Nutr 1995;73:613-23.
30. Holt SHA, Brand Miller JC, Petocz P. An insulin index of foods: The insulin demand
generated by 1000-kJ portions of common foods. Am J Clin Nutr 1997;66:1264-76.
31. Wolever TMS, Jenkins DJA, Jenkins AL, Josse RG. The glycemic index: Methodology and
clinical implications. Am J Clin Nutr 1991;54:846-54.
32. Collier G, O'Dea K. The effect of coingestion of fat on the glucose, insulin, and gastric
inhibitory polypeptide responses to carbohydrate and protein. Am J Clin Nutr 1983;37:941-4.
33. Brand-Miller JC, Colagiuri S, Gan ST. Insulin sensitivity predicts glycemia after a protein
load. Metabolism: Clinical and Experimental 2000;49:1-5.
34. Xavier Pi-Sunyer F. Glycemic index and disease. Am J Clin Nutr 2002;76:290S-298S.
35. Feinle C, O'Donovan D, Doran S, et al. Effects of fat digestion on appetite, APD motility,
and gut hormones in response to duodenal fat infusion in humans. Am J Physiol Gastrointest
Liver Physiol 2003;284: G798-G807.
36. Collier GR, Greenberg GR, Wolever TMS, Jenkins DJA. The acute effect of fat on insulin
secretion. J Clin Endocrinol Metab 1988;66:323-6.
37. Gunnarsson PT, Winzell MS, Deacon CF, et al. Glucose-induced incretin hormone release
and inactivation are differently modulated by oral fat and protein in mice. Endocrinology
2006;147:3173-80.
38. Moghaddam E, Vogt JA, Wolever TMS. The effects of fat and protein on glycemic responses
in nondiabetic humans vary with waist circumference, fasting plasma insulin, and dietary fiber
intake. J Nutr 2006;136:2506-11.
39. Wang PYT, Caspi L, Lam CKL, et al. Upper intestinal lipids trigger a gut-brain-liver axis to
regulate glucose production. Nature 2008;452:1012-6.
40. Brand-Miller JC. Postprandial glycemia, glycemic index, and the prevention of type 2
diabetes. Am J Clin Nutr 2004;80:243-3.
41. Dietary reference intakes for energy, carbohydrate, fiber, fat, fatty acids, cholesterol, protein,
and amino acids. Washington, D.C.: National Academies Press, 2005.
149
42. Wolever TMS. Carbohydrate and the regulation of blood glucose and metabolism. Nutr Rev
2003;61:S40-S48.
43. Sheard NF. Dietary carbohydrate (amount and type) in the prevention and management of
diabetes: A statement by the American Diabetes Association. Diabetes Care 2004;27:2266-71.
44. Lee BM. Effect of glucose, sucrose and fructose on plasma glucose and insulin responses in
normal humans: Comparison with white bread. Eur J Clin Nutr 1998;52:924-8.
45. Pribylova J. Glucose and galactose infusions in newborns of diabetic and healthy mothers.
Biol Neonate 1979;36:193-7.
46. Henry RR. Current issues in fructose metabolism. Annu Rev Nutr 1991;11:21-39.
47. Mayes PA. Intermediary metabolism of fructose. Am J Clin Nutr 1993;58:754S-765S.
48. Crapo PA. Effects of oral fructose in normal, diabetic, and impaired glucose tolerance
subjects. Diabetes Care 1980;3:575-81.
49. Crapo PA. Comparison of the metabolic responses to fructose and sucrose sweetened foods.
Am J Clin Nutr 1982;36:256-61.
50. Teff KL. Dietary fructose reduces circulating insulin and leptin, attenuates postprandial
suppression of ghrelin, and increases triglycerides in women. J Clin Endocrinol Metab
2004;89:2963-72.
51. Chong MF-. Mechanisms for the acute effect of fructose on postprandial lipemia. Am J Clin
Nutr 2007;85:1511-20.
52. Hallfrisch J. Blood lipid distribution of hyperinsulinemic men consuming three levels of
fructose. Am J Clin Nutr 1983;37:740-8.
53. Reiser S. Blood lipids, lipoproteins, apoproteins, and uric acid in men fed diets containing
fructose or high-amylose cornstarch. Am J Clin Nutr 1989;49:832-9.
54. Wolever TMS, Bolognesi C. Source and amount of carbohydrate affect postprandial glucose
and insulin in normal subjects. J Nutr 1996;126:2798-806.
55. Wolever TMS. Glycaemic index of 102 complex carbohydrate foods in patients with
diabetes. Nutr Res 1994;14:651-69.
56. Englyst KN. Rapidly available glucose in foods: An in vitro measurement that reflects the
glycemic response. Am J Clin Nutr 1999;69:448-54.
57. Jenkins DJA. Relationship between rate of digestion of foods and post-prandial glycaemia.
Diabetologia 1982;22:450-5.
150
58. Wolever TMS. Ileal loss of available carbohydrate in man: Comparison of a breath hydrogen
method with direct measurement using a human ileostomy model. Am J Gastroenterol
1986;81:115-22.
59. Jenkins DJ. Slow release dietary carbohydrate improves second meal tolerance. Am J Clin
Nutr 1982;35:1339-46.
60. Jenkins DJA. Metabolic effects of reducing rate of glucose ingestion by single bolus versus
continuous sipping. Diabetes 1990;39:775-81.
61. Wursch P. Cell structure and starch nature as key determinants of the digestion rate of starch
in legume. Am J Clin Nutr 1986;43:25-9.
62. Haber GB. Depletion and disruption of dietary fibre. Effects on satiety, plasma-glucose, and
serum-insulin. Lancet 1977;2:679-82.
63. O'Dea K. Physical factors influencing postprandial glucose and insulin responses to starch.
Am J Clin Nutr 1980;33:760-5.
64. Jenkins DJA. Effect of processing on digestibility and the blood glucose response: A study of
lentils. Am J Clin Nutr 1982;36:1093-1101.
65. Krezowski PA. The effect of protein ingestion on the metabolic response to oral glucose in
normal individuals. Am J Clin Nutr 1986;44:847-56.
66. Thorne MJ. Factors affecting starch digestibility and the glycemic response with special
reference to legumes. Am J Clin Nutr 1983;38:481-8.
67. Hansen WE. Effect of dietary fiber on pancreatic lipase activity in vitro. Pancreas
1987;2:195-8.
68. Yoon JH. The effect of phytic acid on in vitro rate of starch digestibility and blood glucose
response. Am J Clin Nutr 1983;38:835-42.
69. Englyst HN. Englyst HN. Classification and measurement of nutritionally important starch
fractions. Eur J Clin Nutr 1992;46:S33-S50.
70. Englyst HN. Measurement of rapidly available glucose (RAG) in plant foods: A potential in
vitro predictor of the glycaemic response. Br J Nutr 1996;75:327-37.
71. Topping DL. Short-chain fatty acids and human colonic function: Roles of resistant starch
and nonstarch polysaccharides. Physiol Rev 2001;81:1031-64.
72. Cummings JH. Short chain fatty acids in the human colon. Gut 1981;22:763-79.
151
73. Pouteau E. Whole-body, peripheral and intestinal endogenous acetate turnover in dogs using
stable isotopes. J Nutr 1998;128:111-5.
74. Pouteau E. Production rates and metabolism of short-chain fatty acids in the colon and whole
body using stable isotopes. Proc Nutr Soc 2003;62:87-93.
75. Scheppach W. Effect of gut-derived acetate on oral glucose tolerance in man. Clin Sci
1988;75:355-61.
76. Crouse JR. Role of acetate in the reduction of plasma free fatty acids produced by ethanol in
man. J Lipid Res 1968;9:509-12.
77. Wolever TMS. Effect of rectal infusion of short chain fatty acids in human subjects. Am J
Gastroenterol 1989;84:1027-33.
78. Wolever TMS. Interaction between colonic acetate and propionate in humans. Am J Clin
Nutr 1991;53:681-7.
79. Jenkins DJA. Glycemic index: Overview of implications in health and disease. Am J Clin
Nutr 2002;76:266S-273S.
80. Salmeron J. Dietary fiber, glycemic load, and risk of non-insulin-dependent diabetes mellitus
in women. JAMA 1997;277:472-7.
81. Vega-López S, Mayol-Kreiser SN. Use of the glycemic index for weight loss and glycemic
control: A review of recent evidence. Current Diabetes Reports 2009;9:379-88.
82. Reaven GM. Diet and Syndrome X. Curr Atheroscler Rep 2000;2:503-7.
83. Coulston AM. Much ado about (almost) nothing. Diabetes Care 1997;20:241-3.
84. Franz MJ. The glycemic index: Not the most effective nutrition therapy intervention.
Diabetes Care 2003;26:2466-8.
85. Flint A. The use of glycaemic index tables to predict glycaemic index of composite breakfast
meals. Br J Nutr 2004;91:979-89.
86. Wolever TMS. Prediction of glucose and insulin responses of normal subjects after
consuming mixed meals varying in energy, protein, fat, carbohydrate and glycemic index. J Nutr
1996;126:2807-12.
87. Wolever TMS. Food glycemic index, as given in Glycemic Index tables, is a significant
determinant of glycemic responses elicited by composite breakfast meals. Am J Clin Nutr
2006;83:1306-12.
152
88. Coulston AM. Effect of source of dietary carbohydrate on plasma glucose and insulin
responses to mixed meals in subjects with NIDDM. Diabetes Care 1987;10:395-400.
89. Alfenas RCG. Influence of glycemic index/load on glycemic response, appetite, and food
intake in healthy humans. Diabetes Care 2005;28:2123-9.
90. Venn B,J., Green T,J. Glycemic index and glycemic load: measurement issues and their
effect on diet-disease relationships. Eur J Clin Nutr 2007;61: S122-S131.
91. Mann J, Cummings J,H., Englyst H,N., et al. FAO/WHO Scientific Update on carbohydrates
in human nutrition: conclusions. Eur J Clin Nutr 2007;61: S132-S137.
92. Standards of medical care in diabetes-2011. Diabetes Care 2011;34:S11-S61.
93. Wolever TMS. Beneficial effect of low-glycemic index diet in overweight NIDDM subjects.
Diabetes Care 1992;15:562-4.
94. Brand JC. Low-glycemic index foods improve long-term glycemic control in NIDDM.
Diabetes Care 1991;14:95-101.
95. Fontvieille AM. The use of low glycaemic index foods improves metabolic control of
diabetic patients over five weeks. Diabetic Med 1992;9:444-50.
96. Frost G. Dietary advice based on the glycaemic index improves dietary profile and metabolic
control in type 2 diabetic patients. Diabetic Med 1994;11:397-401.
97. Jarvi AE. Improved glycemic control and lipid profile and normalized fibrinolytic activity on
a low-glycemic index diet in type 2 diabetic patients. Diabetes Care 1999;22:10-18.
98. Giacco R. Long-term dietary treatment with increased amounts of fiber-rich low-glycemic
index natural foods improves blood glucose control and reduces the number of hypoglycemic
events in type 1 diabetic patients. Diabetes Care 2000;23:1461-6.
99. Gilbertson HR. The effect of flexible low glycemic index dietary advice versus measured
carbohydrate exchange diets on glycemic control in children with type 1 diabetes. Diabetes Care
2001;24:1137-43.
100. Lafrance L. Effects of different glycaemic index foods and dietary fibre intake on glycaemic
control in type I diabetic patients on intensive insulin therapy. Diabetic Med 1998;15:972-8.
101. Luscombe ND. Diets high and low in glycemic index versus high monounsaturated fat
diets: Effects on glucose and lipid metabolism in NIDDM. Eur J Clin Nutr 1999;53:473-8.
102. Heilbronn LK. The effect of high- and low-glycemic index energy restricted diets on plasma
lipid and glucose profiles in type 2 diabetic subjects with varying glycemic control. J Am Coll
Nutr 2002;21:120-7.
153
103. Brand-Miller J. Low-glycemic index diets in the management of diabetes: A meta-analysis
of randomized controlled trials. Diabetes Care 2003;26:2261-7.
104. Barclay AW. Glycemic index, glycemic load, and chronic disease risk - A metaanalysis of
observational studies. Am J Clin Nutr 2008;87:627-37.
105. Thomas D. Low glycaemic index, or low glycaemic load, diets for diabetes mellitus. The
Cochrane database of systematic reviews 2009; issue1. Art.No.CD006296.
106. Thomas DE. The use of low-glycaemic index diets in diabetes control. Br J Nutr
2010;104:797-802.
107. Holman RR. A randomized double-blind trial of acarbose in type 2 diabetes shows
improved glycemic control over 3 years (U.K. Prospective Diabetes Study 44). Diabetes Care
1999;22:960-4.
108. Holman RR. United Kingdom prospective diabetes study (UKPDS) 13: Relative efficacy of
randomly allocated diet, sulphonylurea, insulin, or metformin in patients with newly diagnosed
non-insulin dependent diabetes followed for three years. BMJ 1995;310:83-8.
109. Ludwig DS. The glycemic index: Physiological mechanisms relating to obesity, diabetes,
and cardiovascular disease. JAMA 2002;287:2414-23.
110. Jenkins DJA. Effect of a low-glycemic index or a high-cereal fiber diet on type 2 diabetes:
A randomized trial. JAMA 2008;300:2742-53.
111. Reaven GM. Effects of low-fat, high-carbohydrate diets on risk factors for ischemic heart
disease in postmenopausal women. Am J Clin Nutr 1997;65:1027-33.
112. Frost G. Glycaemic index as a determinant of serum HDL-cholesterol concentration. Lancet
1999;353:1045-8.
113. Liu S. Dietary glycemic load assessed by food-frequency questionnaire in relation to plasma
high-density-lipoprotein cholesterol and fasting plasma triacylglycerols in postmenopausal
women. Am J Clin Nutr 2001;73:560-6.
114. Mente A. A systematic review of the evidence supporting a causal link between dietary
factors and coronary heart disease. Arch Intern Med 2009;169:659-69.
115. McKeown-Eyssen G. Epidemiology of colorectal cancer revisited: Are serum triglycerides
and/or plasma glucose associated with risk? Cancer epidemiology biomarkers prevention
1994;3:687-95.
116. Giovannucci E. Insulin, insulin-like growth factors and colon cancer: A review of the
evidence. J Nutr 2001;131:3109S-3120S.
154
117. Key TJ. Carbohydrates and cancer: An overview of the epidemiological evidence. Eur J
Clin Nutr 2007;61:S112-S121.
118. Mulholland HG. Glycemic index, glycemic load, and risk of digestive tract neoplasms: A
systematic review and meta-analysis. Am J Clin Nutr 2009;89:568-76.
119. Larsson SC. Glycemic load, glycemic index and breast cancer risk in a prospective cohort
of Swedish women. International journal of cancer 2009;125:153-7.
120. Wen W. Dietary carbohydrates, fiber, and breast cancer risk in Chinese women. Am J Clin
Nutr 2009;89:283-9.
121. Mulholland HG. Dietary glycaemic index, glycaemic load and breast cancer risk: A
systematic review and meta-analysis. The British Journal of Cancer 2008;99:1170-5.
122. Gnagnarella P. Glycemic index, glycemic load, and cancer risk: A meta-analysis. Am J Clin
Nutr 2008;87:1793-1801.
123. Van MME. Methodological challenges in the application of the glycemic index in
epidemiological studies using data from the european prospective investigation into cancer and
nutrition. J Nutr 2009;139:568-75.
124. Coulston AM. Utility of studies measuring glucose and insulin responses to various
carbohydrate-containing foods. Am J Clin Nutr 1984;39:163-5.
125. Miller JB. The glycaemic index of foods containing sugars: Comparison of foods with
naturally-occurring v. Added sugars. Br J Nutr 1995;73:613-23.
126. Hoyt G, Hickey MS, Cordain L. Dissociation of the glycaemic and insulinaemic responses
to whole and skimmed milk. Br J Nutr 2005;93:175-7.
127. Gannon MC. The insulin and glucose responses to meals of glucose plus various proteins in
type II diabetic subjects. Metabolism, clinical and experimental 1988;37:1081-8.
128. Collier G. The effect of coingestion of fat on the glucose, insulin, and gastric inhibitory
polypeptide responses to carbohydrate and protein. Am J Clin Nutr 1983;37:941-4.
129. Gentilcore D. Effects of fat on gastric emptying of and the glycemic, insulin, and incretin
responses to a carbohydrate meal in type 2 diabetes. J Clin Endocrinol Metab 2006;91:2062-7.
130. Collier GR. The acute effect of fat on insulin secretion. J Clin Endocrinol Metab
1988;66:323-6.
131. Bao J, De Jong V, Atkinson F, Petocz P, Brand-Miller JC. Food insulin index: Physiologic
basis for predicting insulin demand evoked by composite meals. Am J Clin Nutr 2009;90:986-
92.
155
132. Collier G. Effect of co-ingestion of fat on the metabolic responses to slowly and rapidly
absorbed carbohydrates. Diabetologia 1984;26:50-54.
133. Rasmussen O. Differential effects of saturated and monounsaturated fat on blood glucose
and insulin responses in subjects with non-insulin-dependent diabetes mellitus. Am J Clin Nutr
1996;63:249-53.
134. Normand S. Influence of dietary fat on postprandial glucose metabolism (exogenous and
endogenous) using intrinsically 13C-enriched durum wheat. Br J Nutr 2001;86:3-11.
135. Welch I. Duodenal and ileal lipid suppresses postprandial blood glucose and insulin
responses in man: Possible implications for the dietary management of diabetes mellitus. Clin
Sci 1987;72:209-216.
136. Cecil JE. Comparison of the effects of a high-fat and high-carbohydrate soup delivered
orally and intragastrically on gastric emptying, appetite, and eating behaviour. Physiology
behavior 1999;67:299-306.
137. Herrmann C, Goke R, Richter G, Fehmann H-, Arnold R, Goke B. Glucagon-like peptide-1
and glucose-dependent insulin releasing polypeptide plasma levels in response to nutrients.
Digestion 1995;56:117-26.
138. Owen B. Effect of fat on glycaemic responses in normal subjects: A dose-response study.
Nutr Res 2003;23:1341-7.
139. Gatti E. Differential effect of unsaturated oils and butter on blood glucose and insulin
response to carbohydrate in normal volunteers. Eur J Clin Nutr 1992;46:161-6.
140. Beysen C. Interaction between specific fatty acids, GLP-1 and insulin secretion in humans.
Diabetologia 2002;45:1533-41.
141. Thomsen C. Differential effects of saturated and monounsaturated fatty acids on
postprandial lipemia and incretin responses in healthy subjects. Am J Clin Nutr 1999;69:1135-
43.
142. Dworatzek PDN. Postprandial lipemia in subjects with the threonine 54 variant of the fatty
acid-binding protein 2 gene is dependent on the type of fat ingested. Am J Clin Nutr
2004;79:1110-7.
143. Karamanlis A. Effects of protein on glycemic and incretin responses and gastric emptying
after oral glucose in healthy subjects. Am J Clin Nutr 2007;86:1364-8.
144. Nuttall FQ. Effect of protein ingestion on the glucose and insulin response to a standardized
oral glucose load. Diabetes Care 1984;7:465-70.
156
145. Spiller GA. Effect of protein dose on serum glucose and insulin response to sugars. Am J
Clin Nutr 1987;46:474-80.
146. Westphal SA. Metabolic response to glucose ingested with various amounts of protein. Am
J Clin Nutr 1990;52:267-72.
147. Wolever TMS., ed. The glycemic index: a physiological classification of dietary
carbohydrate. Wallingford(UK): CABI Publishing, 2006.
148. Boirie Y. Slow and fast dietary proteins differently modulate postprandial protein accretion.
Proc Natl Acad Sci U S A 1997;94:14930-5.
149. Dangin M. The digestion rate of protein is an independent regulating factor of postprandial
protein retention. American journal of physiology: endocrinology and metabolism [serial online]
2001;280:E340-E348.
150. Tessari P. Slow versus fast proteins in the stimulation of beta-cell response and the
activation of the entero-insular axis in type 2 diabetes. Diabetes Metab Res 2007;23:378-85.
151. Nuttall FQ. Metabolic response to egg white and cottage cheese protein in normal subjects.
Metabolism, clinical and experimental 1990;39:749-55.
152. Lang V. Varying the protein source in mixed meal modifies glucose, insulin and glucagon
kinetics in healthy men, has weak effects on subjective satiety and fails to affect food intake. Eur
J Clin Nutr 1999;53:959-65.
153. Van LJC. Plasma insulin responses after ingestion of different amino acid or protein
mixtures with carbohydrate. Am J Clin Nutr 2000;72:96-105.
154. Frid AH. Effect of whey on blood glucose and insulin responses to composite breakfast and
lunch meals in type 2 diabetic subjects. Am J Clin Nutr 2005;82:69-75.
155. Mahe S. Gastrojejunal kinetics and the digestion of [15N]β-lactoglobulin and casein in
humans: The influence of the nature and quantity of the protein. Am J Clin Nutr 1996;63:546-52.
156. Nilsson M, Stenberg M, Frid AH, Holst JJ, Bj rck IME. Glycemia and insulinemia in
healthy subjects after lactose-equivalent meals of milk and other food proteins: The role of
plasma amino acids and incretins. Am J Clin Nutr 2004;80:1246-53.
157. Groop LC, Bonadonna RC, DelPrato S, et al. Glucose and free fatty acid metabolism in
non-insulin-dependent diabetes mellitus. Evidence for multiple sites of insulin resistance. J Clin
Invest 1989;84:205-13.
158. Ferrannini E, Simonson DC, Katz LD, et al. The disposal of an oral glucose load in patients
with non-insulin-dependent diabetes. Metabolism: Clinical and Experimental 1988;37:79-85.
157
159. Chen Y-I. Resistance to insulin suppression of plasma free fatty acid concentrations and
insulin stimulation of glucose uptake in noninsulin-dependent diabetes mellitus. J Clin
Endocrinol Metab 1987;64:17-21.
160. Bergman RN, Finegood DT, Ader M. Assessment of insulin sensitivity in vivo. Endocr Rev
1985;6:45-86.
161. Del Prato S, Leonetti F, Simonson DC, Sheehan P, Matsuda M, DeFronzo RA. Effect of
sustained physiologic hyperinsulinaemia and hyperglycaemia on insulin secretion and insulin
sensitivity in man. Diabetologia 1994;37:1025-35.
162. Kahn, S E Prigeon, R L McCulloch, D K Boyko, E J Bergman, R N Schwartz, M W
Neifing, J L Ward, W K Beard, J C Palmer,J P. Quantification of the relationship between
insulin sensitivity and beta-cell function in human subjects. Evidence for a hyperbolic function.
Diabetes 1993;42:1663-72.
163. Diamond MP, Thornton K, Connolly-Diamond M, Sherwin RS, DeFronzo RA. Reciprocal
variations in insulin-stimulated glucose uptake and pancreatic insulin secretion in women with
normal glucose tolerance. J Soc Gynecol Investig 1995;2:708-15.
164. Ferrannini E, Balkau B, Coppack SW, et al. Insulin resistance, insulin response, and obesity
as indicators of metabolic risk. J Clin Endocrinol Metab 2007;92:2885-92.
165. Kim SH, Reaven GM. Insulin resistance and hyperinsulinemia. Diabetes Care
2008;31:1433-8.
166. Shen SW, Reaven GM, Farquhar JW. Comparison of impedance to insulin-mediated
glucose uptake in normal subjects and in subjects with latent diabetes. J Clin Invest
1970;49:2151-60.
167. DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying
insulin secretion and resistance. Am J Physiol 1979;237:E214-E223.
168. Bergman RN, Finegood DT, Ader M. Assessment of insulin sensitivity in vivo. Endocr Rev
1985;6:45-86.
169. Matthews DR, Hosker JP, Rudenski AS. Homeostasis model assessment: Insulin resistance
and β-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia
1985;28:412-9.
170. Katz A, Nambi SS, Mather K, et al. Quantitative insulin sensitivity check index: A simple,
accurate method for assessing insulin sensitivity in humans. J Clin Endocrinol Metab
2000;85:2402-10.
158
171. Hays NP, Starling RD, Sullivan DH, Fluckey JD, Coker RH, Evans WJ. Comparison of
insulin sensitivity assessment indices with euglycemic-hyperinsulinemic clamp data after a
dietary and exercise intervention in older adults. Metab Clin Exp 2006;55:525-32.
172. Abdul-Ghani MA, Matsuda M, Balas B, DeFronzo RA. Muscle and liver insulin resistance
indexes derived from the oral glucose tolerance test. Diabetes Care 2007;30:89-94.
173. Matsuda M, DeFronzo RA. Insulin sensitivity indices obtained from oral glucose tolerance
testing: Comparison with the euglycemic insulin clamp. Diabetes Care 1999;22:1462-70.
174. Mari A, Pacini G, Murphy E, Ludvik B, Nolan JJ. A model-based method for assessing
insulin sensitivity from the oral glucose tolerance test. Diabetes Care 2001;24:539-48.
175. Laakso M. How good a marker is insulin level for insulin resistance? Am J Epidemi
1993;137:959-65.
176. Clausen JO, Borch-Johnsen K, Ibsen H, et al. Insulin sensitivity index, acute insulin
response, and glucose effectiveness in a population-based sample of 380 young healthy
Caucasians: Analysis of the impact of gender, body fat, physical fitness, and life-style factors. J
Clin Invest 1996;98:1195-209.
177. Samaras K. Insulin levels in insulin resistance: Phantom of the metabolic opera? Med J
Aust 2006;185:159-61.
178. Li C, Ford ES, McGuire LC, Mokdad AH, Little RR, Reaven GM. Trends in
hyperinsulinemia among nondiabetic adults in the U.S. Diabetes Care 2006;29:2396-402.
179. Li C, Ford ES, Zhao G, Mokdad AH. Prevalence of pre-diabetes and its association with
clustering of cardiometabolic risk factors and hyperinsulinemia among U.S. adolescents:
National health and nutrition examination survey 2005-2006. Diabetes Care 2009;32:342-7.
180. Reaven GM. Role of insulin resistance in human disease. Diabetes 1988;37:1595-607.
181. DeFronzo RA. Pathogenesis of type 2 diabetes mellitus. Med Clin North Am 2004;88:787-
835.
182. Moller LF, Jespersen J. Fasting serum insulin levels and coronary heart disease in a Danish
cohort: 17-year follow-up. J Cardiovasc Risk 1995;2:235-40.
183. Pyorala K, Savolainen E, Kaukola S, Haapakoski J. Plasma insulin as coronary heart
disease risk factor: Relationship to other risk factors and predictive value during 9 1/2 -year
follow-up of the Helsinki Policemen Study population. Acta Med Scand 1985;217:38-52.
184. Ferrannini E, Natali A, Capaldo B, Lehtovirta M, Jacob S, Yki-Järvinen H. Insulin
resistance, hyperinsulinemia, and blood pressure: Role of age and obesity. Hypertension
1997;30:1144-9.
159
185. Ingelsson E, Sundström J, Ärnlöv J, Zethelius B, Lind L. Insulin resistance and risk of
congestive heart failure. J Am Med Assoc 2005;294:334-41.
186. Dunaif A. Insulin resistance and the polycystic ovary syndrome: Mechanism and
implications for pathogenesis. Endocr Rev 1997;18:774-800.
187. Sanyal AJ, Campbell-Sargent C, Mirshahi F, et al. Nonalcoholic steatohepatitis: Association
of insulin resistance and mitochondrial abnormalities. Gastroenterology 2001;120:1183-92.
188. Luchsinger JA. Hyperinsulinemia and risk of Alzheimer disease. Neurology 2004;63:1187-
92.
189. Young SE, Mainous III AG, Carnemolla M. Hyperinsulinemia and cognitive decline in a
middle-aged cohort. Diabetes Care 2006;29:2688-93.
190. Hsing AW, Gao Y-, Chua Jr. S, Deng J, Stanczyk FZ. Insulin resistance and prostate cancer
risk. J Natl Cancer Inst 2003;95:67-71.
191. Komninou D, Ayonote A, Richie Jr. JP, Rigas B. Insulin resistance and its contribution to
colon carcinogenesis. Exp Biol Med 2003;228:396-405.
192. Bruning PF, Bonfrer JMG, Van Noord PAH, Hart AAM, De Jong-Bakker M, Nooijen WJ.
Insulin resistance and breast-cancer risk. International Journal of Cancer 1992;52:511-6.
193. Reaven GM. The kidney: an unwilling accomplice in syndrome X. American Journal of
Kidney Diseases 1997;30:928-31.
194. Krieger DR, Landsberg L. Mechanisms in obesity-related hypertension: Role of insulin and
catecholamines. American Journal of Hypertension 1988;1:84-90.
195. Plymate SR. Inhibition of sex hormone-binding globulin production in the human hepatoma
(Hep G2) cell line by insulin and prolactin. J Clin Endocrinol Metab 1988;67:460-4.
196. Chen WY. Exogenous and endogenous hormones and breast cancer. Best Practice and
Research: Clinical Endocrinology and Metabolism 2008;22:573-85.
197. Rodin A. Hyperandrogenism in polycystic ovary syndrome: Evidence of dysregulation of
11β-hydroxysteroid dehydrogenase. New Engl J Med 1994;330:460-5.
198. Gasparini L. Potential roles of insulin and IGF-1 in Alzheimer's disease. Trends Neurosci
2003;26:404-6.
199. Farris W, Mansourian S, Chang Y, et al. Insulin-degrading enzyme regulates the levels of
insulin, amyloid β-protein, and the β-amyloid precursor protein intracellular domain in vivo.
Proc Natl Acad Sci U S A 2003;100:4162-7.
160
200. Frolich L. Brain insulin and insulin receptors in aging and sporadic Alzheimer's disease. J
Neural Transm 1998;105:423-38.
201. Nauck MA, Homberger E, Siegel EG. Incretin effects of increasing glucose loads in man
calculated from venous insulin and C-peptide responses. J Clin Endocrinol Metab 1986;63:492-
8.
202. Shapiro ET, Tillil H, Miller MA, et al. Insulin secretion and clearance. Comparison after
oral and intravenous glucose. Diabetes 1987;36:1365-71.
203. Duckworth WC. Insulin metabolism and degradation. Endocr Rev 1981;2:210-33.
204. Bratusch PR. Hepatic disposal of biosynthetic human insulin and porcine C-peptide in
humans. Metabolism, clinical and experimental 1984;33:151-7.
205. Sacca L. Direct assessment of splanchnic uptake and metabolic effects of human and
porcine insulin. J Clin Endocrinol Metab 1984;59:191-6.
206. Polonsky KS, Given BD, Hirsch L, et al. Quantitative study of insulin secretion and
clearance in normal and obese subjects. J Clin Invest 1988;81:435-41.
207. Meier JJ. The reduction in hepatic insulin clearance after oral glucose is not mediated by
Gastric inhibitory polypeptide (GIP). Regul Pept 2003;113:95-100.
208. Meier JJ, Holst JJ, Schmidt WE, Nauck MA. Reduction of hepatic insulin clearance after
oral glucose ingestion is not mediated by glucagon-like peptide 1 or gastric inhibitory
polypeptide in humans. American Journal of Physiology - Endocrinology and Metabolism
2007;293:E849-E856.
209. Nauck MA, Homberger E, Siegel EG. Incretin effects of increasing glucose loads in man
calculated from venous insulin and C-peptide responses. Journal of Clinical Endocrinology and
Metabolism 1986;63:492-8.
210. Basu R. Effects of age and sex on postprandial glucose metabolism differences in glucose
turnover, insulin secretion, insulin action, and hepatic insulin extraction. Diabetes 2006;55:2001-
14.
211. Basu R. Mechanisms of the age-associated deterioration in glucose tolerance: Contribution
of alterations in insulin secretion, action, and clearance. Diabetes 2003;52:1738-48.
212. Ahren B. Age-related reduction in glucose elimination is accompanied by reduced glucose
effectiveness and increased hepatic insulin extraction in man. J Clin Endocrinol Metab
1998;83:3350-6.
213. Kautzky-Willer A. Elevated hepatic insulin extraction in essential hypertension.
Hypertension 1993;21:646-53.
161
214. Meistas MT. Hyperinsulinemia of obesity is due to decreased clearance of insulin. Am J
physiol: endocrinology and metabolism 1983;245:E155-E159.
215. Erdmann J, Mayr M, Oppel U, Sypchenko O, Wagenpfeil S, Schusdziarra V. Weight-
dependent differential contribution of insulin secretion and clearance to hyperinsulinemia of
obesity. Regulatory Peptides 2009;152:1-7.
216. Letiexhe MR. Insulin secretion, clearance, and action on glucose metabolism in cirrhotic
patients. J Clin Endocrinol Metab 1993;77:1263-68.
217. Polonsky KS. Alterations in immunoreactive proinsulin and insulin clearance induced by
weight loss in NIDDM. Diabetes 1994;43:871-7.
218. Gastaldelli A. Relationship Between Hepatic/Visceral Fat and Hepatic Insulin Resistance in
Nondiabetic and Type 2 Diabetic Subjects. Gastroenterology 2007;133:496-506.
219. Frost DP. The kinetics of insulin metabolism in diabetes mellitus. Postgrad Med J
1973;49:949-54.
220. Stimmler L. Insulin disappearance after intravenous injection and its effect on blood
glucose in diabetic and non-diabetic children and adults. Clin Sci 1972;42:337-44.
221. Orskov H. Plasma disappearance rate of injected human insulin in juvenile diabetic,
maturity-onset diabetic and nondiabetic subjects. Diabetes 1969;18:653-9.
222. Strang BD. Relationship of Triglyceride Accumulation to Insulin Clearance and Hormonal
Responsiveness in Bovine Hepatocytes. J Dairy Sci 1998;81:740-7.
223. Goto T. The influence of fatty liver on insulin clearance and insulin resistance in non-
diabetic Japanese subjects. Int J Obes 1995;19:841-5.
224. Kotronen A, Juurinen L, Tiikkainen M, Vehkavaara S, Yki-Järvinen H. Increased Liver Fat,
Impaired Insulin Clearance, and Hepatic and Adipose Tissue Insulin Resistance in Type 2
Diabetes. Gastroenterology 2008;135:122-30.
225. Peiris AN. Splanchnic insulin metabolism in obesity. Influence of body fat distribution. J
Clin Invest 1986;78:1648-57.
226. Svedberg J. Free-fatty acid inhibition of insulin binding, degradation, and action in isolated
rat hepatocytes. Diabetes 1990;39:570-4.
227. Hennes MMI. Receptor and postreceptor effects of free fatty acids (FFA) on hepatocyte
insulin dynamics. Int J Obes 1990;14:831-41.
228. Hennes MM. Mechanism of free fatty acid effects on hepatocyte insulin receptor binding
and processing. Obes Res 1993;1:18-28.
162
229. Wiesenthal SR. Free fatty acids impair hepatic insulin extraction in vivo. Diabetes
1999;48:766-74.
230. Kotronen A. Effect of liver fat on insulin clearance. Am J Physiol: endocrinology and
metabolism 2007;293:E1709-E1715.
231. Escobar O. Hepatic insulin clearance increases after weight loss in obese children and
adolescents. Am J Med Sci 1999;317:282-6.
232. Adrogue HJ. Glyburide increases the secretion, tissue uptake, and action of insulin in
conscious normal dogs. Metabolism, clinical and experimental 1996;45:579-86.
233. Chen H. Alteration of rat hepatic insulin metabolism by glyburide and glipizide. J
Pharmacol Exp Ther 1993;264:1293-8.
234. Schuster D. Impact of metformin on glucose metabolism in nondiabetic, obese African
Americans: A placebo-controlled, 24-month randomized study. Diabetes Care 2004;27:2768-9.
235. Osei K. Thiazolidinediones increase hepatic insulin extraction in African Americans with
impaired glucose tolerance and type 2 diabetes mellitus. A pilot study of rosiglitazone.
Metabolism, clinical and experimental 2007;56:24-29.
236. Gravholt CH. Epidemiological, endocrine and metabolic features in Turner syndrome.
European journal of endocrinology 2004;151:657-87.
237. Monti LD. Glucose turnover and insulin clearance after growth hormone treatment in girls
with turner's syndrome. Metabolism, clinical and experimental 1997;46:1482-8.
238. Krakower GR. Female sex hormones, perinatal, and peripubertal androgenization on
hepatocyte insulin dynamics in rats. Am J Physiol 1993;264:E342-E347.
239. Polonsky KS, Rubenstein AH. C-peptide as a measure of the secretion and hepatic
extraction of insulin. Pitfalls and limitations. Diabetes 1984;33:486-94.
240. Toffolo G. A minimal model of insulin secretion and kinetics to assess hepatic insulin
extraction. Am J Physiol: endocrinology and metabolism 2006;290:E169-E176.
241. Jenkins DJA, Wolever TMS, Jenkins AL. The glycaemic index of foods tested in diabetic
patients: A new basis for carbohydrate exchange favouring the use of legumes. Diabetologia
1983;24:257-64.
242. Wolever TMS, Jenkins DJA, Josse RG, Wong GS, Lee R. The glycemic index: Similarity
of values derived in insulin-dependent and non-insulin-dependent diabetic patients. J Am Coll
Nutr 1987;6:295-305.
163
243. Wolever TMS, Chiasson J-, Hunt JA, Palmason C, Ross SA, Ryan EA. Similarity of
relative glycaemic but not relative insulinaemic responses in normal, IGT and diabetic subjects.
Nutr Res 1998;18:1667-76.
244. Wolever TMS, Campbell JE, Geleva D, Anderson GH. High-fiber cereal reduces
postprandial insulin responses in hyperinsulinemic but not normoinsulinemic subjects. Diabetes
Care 2004;27:1281-5.
245. Simpson RW, McDonald J, Wahlqvist ML. Macronutrients have different metabolic effects
in nondiabetics and diabetics. Am J Clin Nutr 1985;42:449-53.
246. Gannon MC, Nuttall FQ, Westphal SA, Seaquist ER. The effect of fat and carbohydrate on
plasma glucose, insulin, C-peptide, and triglycerides in normal male subjects. J Am Coll Nutr
1993;12:36-41.
247. Gannon MC, Ercan N, Westphal SA, Nuttall FQ. Effect of added fat on plasma glucose and
insulin response to ingested potato in individuals with NIDDM. Diabetes Care 1993;16:874-80.
248. Nuttall FQ, Gannon MC, Wald JL, Ahmed M. Plasma glucose and insulin profiles in
normal subjects ingesting diets of varying carbohydrate, fat, and protein content. J Am Coll Nutr
1985;4:437-50.
249. Nuttall FQ, Mooradian AD, Gannon MC. Effect of protein ingestion on the glucose and
insulin response to a standardized oral glucose load. Diabetes Care 1984;7:465-70.
250. Vilsbøll T, Knop FK, Krarup T, et al. The Pathophysiology of Diabetes Involves a
Defective Amplification of the Late-Phase Insulin Response to Glucose by Glucose-Dependent
Insulinotropic Polypeptide - Regardless of Etiology and Phenotype. J Clin Endocrinol Metab
2003;88:4897-903.
251. Vilsbøll T, Krarup T, Deacon CF, Madsbad S, Holst JJ. Reduced postprandial
concentrations of intact biologically active glucagon-like peptide 1 in type 2 diabetic patients.
Diabetes 2001;50:609-13.
252. Morgan LM, Hampton SM, Tredger JA, Cramb R, Marks V. Modifications of gastric
inhibitory polypeptide (GIP) secretion in man by a high-fat diet. Br J Nutr 1988;59:373-80.
253. Cunningham KM, Daly J, Horowitz M, Read NW. Gastrointestinal adaptation to diets of
differing fat composition in human volunteers. Gut 1991;32:483-6.
254. Morgan LM, Tredger JAT, Hampton SM, French AP, Peake JCF, Marks V. The effect of
dietary modification and hyperglycaemia on gastric emptying and gastric inhibitory polypeptide
(GIP) secretion. Br J Nutr 1988;60:29-37.
164
255. Reimer RA, Mcburney MI. Dietary fiber modulates intestinal proglucagon messenger
ribonucleic acid and postprandial secretion of glucagon-like peptide-1 and insulin in rats.
Endocrinology 1996;137:3948-56.
256. Massimino SP, McBurney MI, Field CJ, et al. Fermentable dietary fiber increases GLP-1
secretion and improves glucose homeostasis despite increased intestinal glucose transport
capacity in healthy dogs. J Nutr 1998;128:1786-93.
257. Frost GS, Brynes AE, Dhillo WS, Bloom SR, McBurney MI. The effects of fiber
enrichment of pasta and fat content on gastric emptying, GLP-1, glucose, and insulin responses
to a meal. Eur J Clin Nutr 2003;57:293-8.
258. Freeland KR. Adaptation of colonic fermentation and glucagon-like peptide-1 secretion
with increased wheat fibre intake for 1 year in hyperinsulinaemic human subjects. Br J Nutr
2010;103:82-90.
259. Akhavan T. Effect of premeal consumption of whey protein and its hydrolysate on food
intake and postmeal glycemia and insulin responses in young adults. Am J Clin Nutr
2010;91:966-75.
260. Laakso M. How good a marker is insulin level for insulin resistance? Am J Epidemiol
1993;137:959-65.
261. Jensen CC, Cnop M, Hull RL, Fujimoto WY, Kahn SE. β-cell function is a major
contributor to oral glucose tolerance in high-risk relatives of four ethnic groups in the U.S.
Diabetes 2002;51:2170-8.
262. Turner R. United Kingdom prospective diabetes study 24: A 6-year, randomized, controlled
trial comparing sulfonylurea, insulin, and metformin therapy in patients with newly diagnosed
type 2 diabetes that could not be controlled with diet therapy. Ann Intern Med 1998;128:165-75.
263. Bruce K. Armstrong, Emily White and Rodolfo Saracci. Chapter 5. Reducing measurement
error and its effects. In: Principles of exposure measurement in epidemiology. New York:
Oxford University Press, 1994:115-118.
264. Marr JW. Within- and between-person variation in dietary surveys: number of days needed
to classify individuals. Human nutrition Applied nutrition 1986;40:347-64.
265. Burstein R. Acute reversal of the enhanced insulin action in trained athletes. Association
with insulin receptor changes. Diabetes 1985;34:756-60.
266. Oshida Y. Effects of training and training cessation on insulin action. Int J Sports Med
1991;12:484-6.
267. Baecke JAH, Burema J, Frijters JER. A short questionnaire for the measurement of habitual
physical activity in epidemiological studies. Am J Clin Nutr 1982;36:936-42.
165
268. Polonsky KS, Rubenstein AH. C-peptide as a measure of the secretion and hepatic
extraction of insulin. Pitfalls and limitations. Diabetes 1984;33:486-94.
269. Hovorka R, Soons PA, Young MA. ISEC: A program to calculate insulin secretion. Comput
Methods Programs Biomed 1996;50:253-64.
270. Abdul-Ghani MA, Williams K, DeFronzo RA, Stern M. What is the best predictor of future
type 2 diabetes? Diabetes Care 2007;30:1544-8.
271. Stern MP, Williams K, Haffner SM. Identification of persons at high risk for type 2 diabetes
mellitus: Do we need the oral glucose tolerance test? Ann Intern Med 2002;136:575-81.
272. Dworatzek PDN, Hegele RA, Wolever TMS. Postprandial lipemia in subjects with the
threonine 54 variant of the fatty acid-binding protein 2 gene is dependent on the type of fat
ingested. Am J Clin Nutr 2004;79:1110-7.
273. Gatti E, Noe D, Pazzucconi F, et al. Differential effect of unsaturated oils and butter on
blood glucose and insulin response to carbohydrate in normal volunteers. Euro J Clin Nutr
1992;46:161-6.
274. Nuttall FQ, Gannon MC. Plasma glucose and insulin response to macronutrients in
nondiabetic and NIDDM subjects. Diabetes Care 1991;14:824-38.
275. Holt SHA, Brand Miller JC, Petocz P. An insulin index of foods: The insulin demand
generated by 1000-kJ portions of common foods. Am J Clin Nutr 1997;66:1264-76.
276. von Post-Skageg rd M, Vessby B, Karlstr m B. Glucose and insulin responses in healthy
women after intake of composite meals containing cod-, milk-, and soy protein. Eur J Clin Nutr
2006;60:949-54.
277. Nilsson M, Stenberg M, Frid AH, Holst JJ, Björck IME. Glycemia and insulinemia in
healthy subjects after lactose-equivalent meals of milk and other food proteins: The role of
plasma amino acids and incretins. Am J Clin Nutr 2004;80:1246-53.
278. Boirie Y, Dangin M, Gachon P, Vasson M-, Maubois J-, Beaufrère B. Slow and fast dietary
proteins differently modulate postprandial protein accretion. Proceedings of the National
Academy of Sciences of the USA 1997;94:14930-5.
279. Dangin M, Boirie Y, Garcia-Rodenas C, et al. The digestion rate of protein is an
independent regulating factor of postprandial protein retention. American Journal of Physiology -
Endocrinology and Metabolism 2001;280: E340-E348.
280. Dangin M, Boirie Y, Guillet C, Beaufrère B. Influence of the protein digestion rate on
protein turnover in young and elderly subjects. J Nutr 2002;132: 3228S-3233S.
166
281. Hall WL, Millward DJ, Long SJ, Morgan LM. Casein and whey exert different effects on
plasma amino acid profiles, gastrointestinal hormone secretion and appetite. Br J Nutr
2003;89:239-48.
282. Morifuji M, Sakai K, Sugiura K. Dietary whey protein modulates liver glycogen level and
glycoregulatory enzyme activities in exercise-trained rats. Experimental Biology and Medicine
2005;230:23-30.
283. Wolever TMS, Bolognesi C. Source and amount of carbohydrate affect postprandial
glucose and insulin in normal subjects. J Nutr 1996;126:2798-806.
284. Pironi L, Stanghellini V, Miglioli M, et al. Fat-induced ileal brake in humans: A dose-
dependent phenomenon correlated to the plasma levels of peptide YY. Gastroenterology
1993;105:733-9.
285. Moran TH, Kinzig KP. Gastrointestinal satiety signals. II. Cholecystokinin. American
Journal of Physiology - Gastrointestinal and Liver Physiology 2004;286: G183-G188.
286. Horowitz M, Edelbroek MAL, Wishart JM, Straathof JW. Relationship between oral
glucose tolerance and gastric emptying in normal healthy subjects. Diabetologia 1993;36:857-62.
287. Rayner CK, Samsom M, Jones KL, Horowitz M. Relationships of upper gastrointestinal
motor and sensory function with glycemic control. Diabetes Care 2001;24:371-81.
288. Batterham RL, Heffron H, Kapoor S, et al. Critical role for peptide YY in protein-mediated
satiation and body-weight regulation. Cell Metabolism 2006;4:223-33.
289. Nilsson M, Holst JJ, Björck IME. Metabolic effects of amino acid mixtures and whey
protein in healthy subjects: Studies using glucose-equivalent drinks. Am J Clin Nutr
2007;85:996-1004.
290. Carr RD, Larsen MO, Winzell MS, et al. Incretin and islet hormonal responses to fat and
protein ingestion in healthy men. American Journal of Physiology - Endocrinology and
Metabolism 2008;295:E779-E784.
291. Gannon MC, Nuttall JA, Damberg G, Gupta V, Nuttall FQ. Effect of protein ingestion on
the glucose appearance rate in people with type 2 diabetes. J Clin Endocrinol Metab
2001;86:1040-7.
292. Dziuba M. Milk proteins as precursors of bioactive peptides. Acta Scientarum Polonorum -
Technologia Alimentaria 2009;8:71-90.
293. Zavaroni I, Deferrari G, Lugari R. Renal metabolism of C-peptide in man. J Clin Endocrinol
Metab 1987;65:494-8.
167
294. Lew SQ, Bosch JP. Effect of diet on creatinine clearance and excretion in young and elderly
healthy subjects and in patients with renal disease. Journal of the American Society of
Nephrology 1991;2:856-65.
295. Atkinson FS. International tables of glycemic index and glycemic load values: 2008.
Diabetes Care 2008;31:2281-3.
296. Wolever TMS, Chiasson J-, Csima A, et al. Variation of postprandial plasma glucose,
palatability, and symptoms associated with a standardized mixed test meal versus 75 g oral
glucose. Diabetes Care 1998;21:336-40.
297. Wolever TMS, Campbell JE, Geleva D, Anderson GH. High-fiber cereal reduces
postprandial insulin responses in hyperinsulinemic but not normoinsulinemic subjects. Diabetes
Care 2004;27:1281-5.
298. Wolever TMS. Determination of the glycaemic index of foods: Interlaboratory study. Eur J
Clin Nutr 2003;57:475-82.
299. O'Brien RM. A caution regarding rules of thumb for variance inflation factors. Quality
quantity 2007;41:673-90.
300. Jenkins DJA, Wolever TMS, Jenkins AL. Low glycemic response to traditionally processed
wheat and rye products: Bulgur and pumpernickel bread. Am J Clin Nutr 1986;43:516-20.
301. Wolever TMS, Jenkins DJA, Kalmusky J. Comparison of regular and parboiled rices:
Explanation of discrepancies between reported glycemic responses to rice. Nutr Res 1986;6:349-
57.
302. Ebbeling CB. Effects of a low-glycemic load vs low-fat diet in obese young adults: A
randomized trial. JAMA 2007;297:2092-102.
303. Oh K. Carbohydrate intake, glycemic index, glycemic load, and dietary fiber in relation to
risk of stroke in women. Am J Epidemiol 2005;161:161-9.
304. Wolever TMS. The glycaemic index values of foods containing fructose are affected by
metabolic differences between subjects. Eur J Clin Nutr 2009;63:1106-14.
305. Creutzfeldt W. The incretin concept today. Diabetologia 1979;16:75-85.
306. Creutzfeldt W, Ebert R. New developments in the incretin concept. Diabetologia
1985;28:565-73.
307. Meier JJ, Nauck MA. Glucagon-like peptide 1(GLP-1) in biology and pathology. Diabetes
Metab Res 2005;21:91-117.
168
308. Drucker DJ, Nauck MA. The incretin system: glucagon-like peptide-1 receptor agonists and
dipeptidyl peptidase-4 inhibitors in type 2 diabetes. Lancet 2006;368:1696-705.
309. Lazic SE. Why we should use simpler models if the data allow this: Relevance for ANOVA
designs in experimental biology. BMC physiology 2008;8:16.
310. Krezowski PA, Nuttall FQ, Gannon MC, Bartosh NH. The effect of protein ingestion on the
metabolic response to oral glucose in normal individuals. Am J Clin Nutr 1986;44:847-56.
311. Gannon MC, Nuttall FQ, Westphal SA, Seaquist ER. The effect of fat and carbohydrate on
plasma glucose, insulin, C-peptide, and triglycerides in normal male subjects. J Am Coll Nutr
1993;12:36-41.
312. Nauck MA, Heimesaat MM, Orskov C, Holst JJ, Ebert R, Creutzfeldt W. Preserved incretin
activity of glucagon-like peptide 1 [7-36 amide] but not of synthetic human gastric inhibitory
polypeptide in patients with type- 2 diabetes mellitus. J Clin Invest 1993;91:301-7.
313. Meier JJ, Hücking K, Holst JJ, Deacon CF, Schmiegel WH, Nauck MA. Reduced
Insulinotropic Effect of Gastric Inhibitory Polypeptide in First-Degree Relatives of Patients with
Type 2 Diabetes. Diabetes 2001;50:2497-504.
314. Wick A, Newlin K. Incretin-based therapies: Therapeutic rationale and pharmacological
promise for type 2 diabetes. J Am Acad Nurse Pract 2009;21:623-30.
315. Lim GE, Brubaker PL. Glucagon-like peptide 1 secretion by the L-cell: The view from
within. Diabetes 2006;55:S70-S77.
316. Muscelli E, Mari A, Casolaro A, et al. Separate impact of obesity and glucose tolerance on
the incretin effect in normal subjects and type 2 diabetic patients. Diabetes 2008;57:1340-8.
317. Kjems LL, Holst JJ, Vølund A, Madsbad S. The influence of GLP-1 on glucose-stimulated
insulin secretion: Effects on β-cell sensitivity in type 2 and nondiabetic subjects. Diabetes
2003;52:380-6.
318. Vollmer K, Hoist JJ, Bailer B, et al. Predictors of incretin concentrations in subjects with
normal, impaired, and diabetic glucose tolerance. Diabetes 2008;57:678-87.
319. Theodorakis MJ, Carlson O, Michopoulos S, et al. Human duodenal enteroendocrine cells:
Source of both incretin peptides, GLP-1 and GIP. American Journal of Physiology -
Endocrinology and Metabolism 2006;290: E550-E559.
320. Kahn SE. The relative contributions of insulin resistance and beta-cell dysfunction to the
pathophysiology of Type 2 diabetes. Diabetologia 2003;46:3-19.
169
321. Ehrmann DA, Sturis J, Byrne MM, Karrison T, Rosenfield RL, Polonsky KS. Insulin
secretory defects in polycystic ovary syndrome. Relationship to insulin sensitivity and family
history of non-insulin-dependent diabetes mellitus. J Clin Invest 1995;96:520-7.
322. Elbein SC, Wegner K, Kahn SE. Reduced β-cell compensation to the insulin resistance
associated with obesity in members of caucasian familial type 2 diabetic kindreds. Diabetes Care
2000;23:221-7.
323. Ryan EA, Imes S, Liu D, et al. Defects in insulin secretion and action in women with a
history of gestational diabetes. Diabetes 1995;44:506-12.
324. Ward WK, Johnston CLW, Beard JC. Insulin resistance and impaired insulin secretion in
subjects with histories of gestational diabetes mellitus. Diabetes 1985;34:861-9.
325. Buchanan TA, Xiang AH, Kjos SL, Trigo E, Lee WP, Peters RK. Antepartum predictors of
the development of type 2 diabetes in Latino women 11-26 months after pregnancies
complicated by gestational diabetes. Diabetes 1999;48:2430-6.
326. Dunaif A, Finegood DT. β-Cell dysfunction independent of obesity and glucose intolerance
in the polycystic ovary syndrome. Journal of Clinical Endocrinology and Metabolism
1996;81:942-7.
327. Chen M, Bergman RN, Pacini G, Porte Jr. D. Pathogenesis of age-related glucose
intolerance in man: Insulin resistance and decreased β-cell function. Journal of Clinical
Endocrinology and Metabolism 1985;60:13-20.
328. Kahn SE, Larson VG, Beard JC, et al. Effect of exercise on insulin action, glucose
tolerance, and insulin secretion in aging. American Journal of Physiology - Endocrinology and
Metabolism 1990;258: E937-E943.
329. Rask E, Olsson T, S derberg S, et al. Impaired incretin response after a mixed meal is
associated with insulin resistance in nondiabetic men. Diabetes Care 2001;24:1640-5.
330. Forbes S, Moonan M, Robinson S, et al. Impaired circulating glucagon-like peptide-1
response to oral glucose in women with previous gestational diabetes. Clinical Endocrinology
2005;62:51-5.
331. Nyholm B, Walker M, Gravholt CH, et al. Twenty-four-hour insulin secretion rates,
circulating concentrations of fuel substrates and gut incretin hormones in healthy offspring of
Type II (non-insulin-dependent) diabetic parents: Evidence of several aberrations. Diabetologia
1999;42:1314-23.
332. Nauck MA, El-Ouaghlidi A, Gabrys B, et al. Secretion of incretin hormones (GIP and GLP-
1) and incretin effect after oral glucose in first-degree relatives of patients with type 2 diabetes.
Regulatory Peptides 2004;122:209-17.
170
333. Meier JJ, Gallwitz B, Askenas M, et al. Secretion of incretin hormones and the
insulinotropic effect of gastric inhibitory polypeptide in women with a history of gestational
diabetes. Diabetologia 2005;48:1872-81.
334. Egan JM, Bulotta A, Hui H, Perfetti R. GLP-1 receptor agonists are growth and
differentiation factors for pancreatic islet beta cells. Diabetes/Metabolism Research and Reviews
2003;19:115-23.
335. Stoffers DA, Kieffer TJ, Hussain MA, et al. Insulinotropic glucagon-like peptide 1 agonists
stimulate expression of homeodomain protein IDX-1 and increase islet size in mouse pancreas.
Diabetes 2000;49:741-8.
336. Xu G, Stoffers DA, Habener JF, Bonner-Weir S. Exendin-4 stimulates both β-cell
replication and neogenesis, resulting in increased β-cell mass and improved glucose tolerance in
diabetic rats. Diabetes 1999;48:2270-6.
337. Buteau J, El-Assaad W, Rhodes CJ, Mosenberg L, Joly E, Prentki M. Glucagon-like
peptide-1 prevents beta cell glucolipotoxicity. Diabetologia 2004;47:806-15.
338. Farilla L, Bulotta A, Hirshberg B, et al. Glucagon-Like Peptide 1 Inhibits Cell Apoptosis
and Improves Glucose Responsiveness of Freshly Isolated Human Islets. Endocrinology
2003;144:5149-58.
339. Kahn SE. Mechanisms linking obesity to insulin resistance and type 2 diabetes. Nature
2006;444:840-6.
340. Wajchenberg BL. β-cell failure in diabetes and preservation by clinical treatment.
Endocrine Reviews 2007;28:187-218.
341. Ahr n B, Thomaseth K, Pacini G. Reduced insulin clearance contributes to the increased
insulin levels after administration of glucagon-like peptide 1 in mice. Diabetologia
2005;48:2140-6.
342. Li L, Yang G, Li Q, et al. Exenatide prevents fat-induced insulin resistance and raises
adiponectin expression and plasma levels. Diabetes, Obesity and Metabolism 2008;10:921-30.
343. Brandt A, Katschinski M, Arnold R, Polonsky KS, G ke B, Byrne MM. GLP-1-induced
alterations in the glucose-stimulated insulin secretory dose-response curve. Am J Physiol-
Endocrinology and Metabolism 2001;281:E242-E247.
344. Utzschneider KM, Prigeon RL, Tong J, et al. Within-subject variability of measures of beta
cell function derived from a 2 h OGTT: Implications for research studies. Diabetologia
2007;50:2516-25.
171
345. Retnakaran R, Shen S, Hanley AJ, Vuksan V, Hamilton JK, Zinman B. Hyperbolic
relationship between insulin secretion and sensitivity on oral glucose tolerance test. Obesity
2008;16:1901-7.
346. Utzschneider KM, Prigeon RL, Faulenbach MV, et al. Oral disposition index predicts the
development of future diabetes above and beyond fasting and 2-h glucose levels. Diabetes Care
2009;32:335-41.
347. Toft-Nielsen M-, Damholt MB, Madsbad S, et al. Determinants of the impaired secretion of
glucagon-like peptide-1 in type 2 diabetic patients. J Clin Endocrinol Metab 2001;86:3717-23.
348. Fehmann H-, Habener JF. Insulinotropic hormone glucagon-like peptide-I(7-37) stimulation
of proinsulin gene expression and proinsulin biosynthesis in insulinoma βTC-1 cells.
Endocrinology 1992;130:159-66.
349. Goldfine AB, Mun EC, Devine E, et al. Patients with neuroglycopenia after gastric bypass
surgery have exaggerated incretin and insulin secretory responses to a mixed meal. J Clin
Endocrinol Metab 2007;92:4678-85.
350. Knudsen LB, Nielsen PF, Huusfeldt PO, et al. Potent derivatives of glucagon-like peptide-1
with pharmacokinetic properties suitable for once daily administration. J Med Chem
2000;43:1664-9.
351. Garber A, Henry R, Ratner R, et al. Liraglutide versus glimepiride monotherapy for type 2
diabetes (LEAD-3 Mono): a randomised, 52-week, phase III, double-blind, parallel-treatment
trial. Lancet 2009;373:473-81.
352. Holst JJ, Deacon CF. Glucagon-like peptide-1 mediates the therapeutic actions of DPP-IV
inhibitors. Diabetologia 2005;48:612-5.
353. Bullock BP, Heller RS, Habener JF. Tissue distribution of messenger ribonucleic acid
encoding the rat glucagon-like peptide-1 receptor. Endocrinology 1996;137:2968-78.
354. Lan-Pidhainy X, Wolever TM. The hypoglycemic effect of fat and protein is not attenuated
by insulin resistance. Am J Clin Nutr 2010;91:98-105.
355. Fried GM. Comparison of cholecystokinin release and gallbladder emptying in men and in
women at estrogen and progesterone phases of the menstrual cycle. Surgery 1984;95:284-9.
356. Sacchetti G. Influence of age and sex on gallbladder emptying induced by a fatty meal in
normal subjects. AJR, American journal of roentgenology 1973;119:40-45.
357. Phillips SF. Variability of gastrointestinal transit in healthy women and men. Gut
1996;39:299-305.
172
358. Naslund E, Hellstrom PM. Glucagon-like peptide-1 in the pathogenesis of obesity. Drug
News and Perspectives 1998;11:92-7.
359. Ranganath LR, Beety JM, Morgan LM, Wright JW, Howland R, Marks V. Attenuated GLP-
1 secretion in obesity: Cause or consequence? GUT 1996;38:916-9.
360. Vilsbøll T, Krarup T, Sonne J, et al. Incretin secretion in relation to meal size and body
weight in healthy subjects and people with type 1 and type 2 diabetes mellitus. J Clin Endocrinol
Metab 2003;88:2706-13.
361. Wisen O, Johansson C. Gastrointestinal function in obesity: Motility, secretion, and
absorption following a liquid test meal. Metabolism: Clinical and Experimental 1992;41:390-5.
362. Roberge JN, Brubaker PL. Secretion of proglucagon-derived peptides in response to
intestinal luminal nutrients. Endocrinology 1991;128:3169-74.
363. Liddle RA. Cholecystokinin: Its role in health and disease. Curr Opin Endocrinol Diabetes
2003;10:50-4.
364. Boey D. Low serum PYY is linked to insulin resistance in first-degree relatives of subjects
with type 2 diabetes. Neuropeptides 2006;40:317-24.
365. Samra RA. Enhanced food intake regulatory responses after a glucose drink in
hyperinsulinemic men. Int J Obes 2007;31:1222-31.
366. Sandoval DA. Targeting the CNS to treat type 2 diabetes. Nature Reviews.Drug Discovery
2009;8:386-98.
367. Anderson GH, Aziz A. Multifunctional roles of dietary proteins in the regulation of
metabolism and food intake: Application to feeding infants. J Pediatr 2006;149: S74-S79.
368. Saito T. Antihypertensive peptides derived from bovine casein and whey proteins. Adv Exp
Med Biol 2008;606:295-317.
369. Mari A. Comparative evaluation of simple insulin sensitivity methods based on the oral
glucose tolerance test. Diabetologia 2005;48:748-51.
370. Lan-Pidhainy X, Wolever TM. Are the glycemic and insulinemic index values of
carbohydrate foods similar in healthy control, hyperinsulinemic and type 2 diabetic patients
EJCN 2011; 65:727-734.
173
CHAPTER 11
APPENDICES
174
Appendix 1
20 40 60 80 1001201401605
10
15
20
25
r= -0.70 p<0.0001
Fasting Insulin(pmol/L)
C-p
eptide / Insulin
Ratio
(Basal)
20 40 60 80 1000
2
4
6
8
10
r= -0.80 p<0.0001
InsulinAUC 0-120min
(nmol/L x min)
C-p
eptide/insulin
Ratio
(AU
C0-1
20m
in)
20 40 60 80 1001201401600
2
4
6
8
10
r=-0.72p<0.0001
Fasting Insulin(pmol/L)
C-p
eptide/Insulin
Ratio
(AU
C0-1
20m
in)
Appendix 1. The correlations between fasting insulin and basal hepatic insulin extraction
(HIEbasal) (top), fasting insulin and postprandial hepatic insulin extraction (HIEauc) (middle), and
postprandial insulin response (insulinauc) and postprandial hepatic insulin extraction (HIEauc)
(bottom). n=25, the lines are regression lines.
175
Appendix 2
0 200 400 600 800
2
4
6
8
10
r=-0.61 p=0.001
Hepatic insulin resistance
C-p
eptide/Insulin
Ratio
(AU
C0-1
20m
in)
Appendix 2. The correlation between hepatic insulin extraction (C-peptideauc/Insulinauc) and
hepatic insulin resistance. n=25, the line is a regression line.
176
Appendix 3 Physical activity indices of the subjects by FSI1
Low
(FSI<40pmol/L)
Medium
(40 ≤ FSI <70pmol/L)
High
(FSI ≥ 70 pmol/L)
P2
Work 2.6±0.2
2.1±0.1
2.5±0.2
0.07
Sport 2.8±0.2
2.2±0.2
2±0.1
0.44
Leisure-time 3.3±0.2
3±0.2
2.5±0.1
0.83
Mean physical activity index3 2.9±0.1
2.4±0.1
2.3±0.1
0.37
1 n = 25 except where otherwise noted. Values are expressed as mean±SEM.
2P represents overall significant differences across groups by one way ANOVA.
3 Mean physical activity index comprising the work, sport, and leisure-time physical activity
indices.
177
Appendix 4
Low FSI
2
4
6
8
10
og fat
5g fat
30g fat
Glu
cose
(mm
ol/L)
Medium FSI High FSI
0
400
800
1200
Insu
lin
(pm
ol/L)
0
1000
2000
3000
C-p
eptide
(pm
ol/L)
0 30 60 90 1200 30 60 90 120
8
12
16
GL
P-1
(pm
ol/L)
0 30 60 90 120
Time (min)
Appendix 4. Mean (±SEM) 2-h postprandial plasma glucose, insulin, C-peptide, and glucagon-
like peptide 1 (GLP-1) concentrations after 50 g oral glucose plus 0, 5, and 30 g fat in non-
diabetic humans with different concentrations of fasting serum insulin (FSI): low (FSI < 40
pmol/L), medium (40 ≤ FSI < 70 pmol/L), and high (FSI ≥ 70 pmol/L). n = 25. The error bars
are not shown if they overlap or are smaller than the symbol.
178
Appendix 5
0
2
4
6r = -0.06p=0.77
R
IR/g
fat (%
/g)
0 40 80 120 160
0
2
4
6
8
r = 0.04p=0.86
R
IR/g
pro
tein
(%
/g)
Fasting insulin (pmol/L)
0 40 80 120 160
-4
-3
-2
-1
0
r = -0.08p=0.70
RG
R/g
pro
tein
(%
/g)
-1
0
1
2r = -0.05p=0.82
R
GR
/g fat (%
/g)
Appendix 5. The correlations between relative glucose responses (RGR) per g of fat or protein
and fasting insulin (left) and between relative insulin responses (RIR) per g of fat or protein and
fasting insulin (right). r: pearson‘s correlation coefficient. The lines are regression lines.
179
Appendix 6
0 5 30 0 5 30 0 5 300
2
4
6a
b b bab ab abab ab
Lo
g(G
LP
AU
C)
(pm
ol x m
in/L
)
0 5 30 0 5 30 0 5 30
a
bb
a
bb bab ab
g of fat added g of protein added
Appendix 6 Mean (±SEM) effects of 50 g glucose plus 0, 5, or 30 g fat or protein on 2hr GLP-
1 response in healthy nondiabetic subjects with different levels of fasting serum insulin (FSI):
low (FSI < 40 pmol/L, white bars), medium (40 ≤ FSI < 70 pmol/L, grey bars), and high (FSI ≥
70 pmol/L, black bars). Means not sharing a common letter are significantly different [repeated
measures ANOVA, PROC MIXED procedure (SAS Institute, Cary, NC)]. Age and BMI were
included in the model as covariates (Tukey‘s post hoc test, P < 0.05). n = 25.
180
Appendix 7
Nutrient composition of 30g whey protein
whey powder (37g)
Energy (kcal) 153
Protein (g) 30
Fat (g) 2.3
Total CHO (g) 2.6
Proteose peptones (g) 1.2
α-lactalbumin (g) 3.3
β-lactoglobulin (g) 14.3
Bovine serum albumin (g) 0.4
Immunoglobulins (g) 1.2
Glycomacropeptide (g) 4.6
Lactoferrin (g) 0.1
Source: Product bulletin of Whey protein concentrate 392; Fonterra Inc, Camp Hill, PA.
181
Appendix 8 Screening form for fat and protein study
XIAPP (CIHR PP Study)
Postprandial responses elicited by fat and protein in normal and hyperinsulinaemic subjects
Contact Person: Xiaomiao Lan-Pidhainy [email protected]
Contact Phone: 7438
Physician: Dr. Thomas Wolever, 61 Queen, 6th
floor, rm 6-143
Study duration: March-Nov 07
Please enter:
1. Subject ID: ______ Subject Initial: _______ DOB: _______/_______/ _____
(day) (month) (year)
2. Enter Doctor Code. 14147 Dr. Wolever
3. Enter Location Code: XIAPP
4. Register the following tests:
□ Phlebotomy
□ Glucose
□ Insulin
□ Total cholesterol
□ HDL cholesterol
□ Triglycerides
□ Creatinine
□ AST
□ CRP
182
Appendix 9
3-day food records
Initial: ________ Subject ID#: _______ Day of Record: __________
Day of Week (circle): Mon Tue Wed Thu Fri Sat Sun Date: __________
Meal location Food Portion Condiment Portion Beverage Portion
Breakfast
Time:
Snack
Time:
Lunch
Time:
Snack
Time:
Dinner
Time:
Snack
Time:
Portion
control
guide
¼ cup = golf ball ½ cup = tennis or racquet ball 1 cup = small fist
1 oz. = one handful or matchbox 4 oz. fish filet = eyeglass case
3 oz. portion of cooked meat = a deck of playing cards or cassette tape
1 teaspoon = quarter or tip of your thumb 3 teaspoons = 1 tablespoon 8 fl. oz. = 1 cup
1. Overall, do you feel the food choices and amounts you ate today were typical of your usual
diet? Yes____ No____ Somewhat_____
2. For the most part, when did you record food items eaten? Immediately after eating____ A
while after eating_____At the end of the day________
Other___________________________
183
Appendix 10 Consent form for the 1st study: Postprandial responses elicited by fat and protein
in normal and hyperinsulinaemic subjects
Risk Factor Modification Center
St. Michael‘s Hospital
61 Queen‘s St. East
(416) 867-7438
Consent to Participate in a Research Study
Study Title: Postprandial responses elicited by fat and protein in normal and
hyperinsulinaemic subjects
Before agreeing to take part in this research study, it is important that you read the information in
this research consent form. It includes details we think you need to know in order to decide if
you wish to take part in the study. If you have any questions, ask a study doctor or study staff.
You should not sign this form until you are sure you understand the information. All research is
voluntary. You may also wish to discuss the study with your family doctor, a family member or
close friend. If you decide to take part in the study, it is important that you are completely
truthful about your health history and any medications you are taking. This will help prevent
unnecessary harm to you.
Investigator(s)
INVESTIGATORS: Thomas MS Wolever, PhD, DM,
Endocrinologist, St. Michael‘s Hospital
Professor, Study Advisor, Department of Nutritional Sciences
University of Toronto
Toronto, ON M5S 3E2
Tel: 416-978-5556 e-mail: [email protected]
Xiaomiao Lan-Pidhainy, Study Coordinator;
MSc, Ph.D student
Department of Nutritional Sciences
University of Toronto
Toronto, ON M5S 3E2
Tel: 416-867-7438 e-mail: [email protected]
184
Doctor of Philosophy Candidate Research Project
This study is to be conducted as part of the study coordinator, Ms Lan-Pidhainy‘s Ph. D degree
requirements. Ms Lan-Pidhainy has completed her Master‘s of Science (MSc) degree at the
University of Toronto. Dr. Wolever is Ms Lan-Pidhainy‘s advisor and Principle Investigator on
this study. This means he will be supervising the conduct of the study.
Conflict of Interest
Dr.Wolever is the president and part owner of Glycemic Index Laboratories (GI Labs), a contract
research organization specializing in determining the glycaemic index of foods. However, there
is no foreseeable commercial application of the results. Ms Lan-Pidhainy declares no conflicts
of interest.
Study Sponsor
Canadian Institutes of Health Research
160 Elgin Street, 9th Floor
Address Locator 4809A
Ottawa, ON, K1A 0W9
CANADA
Purpose of the Research
You are invited to participate in a nutrition research study on the after-meal metabolic responses
to fat and protein in normal subjects and subjects with high fasting insulin.
High blood glucose after eating is associated with increased risk of cardiovascular disease,
diabetes and cancer. Many studies have compared the effects of different carbohydrate foods on
blood glucose but less is known about how fat and protein affect glucose response. We recently
found that the glucose-lowering effect of fat was reduced in insulin resistant subjects, and the
ability of protein to lower glucose was increased as fiber intake increased. The purpose of this
study is to find out why this happened by measuring the effects of fat and protein on insulin and
other hormones.
The reason for this study is to test whether the blood glucose lowering effect of fat and protein is
reduced in subjects with high fasting insulin compared to the normal subjects. This is likely due
to their differences in insulin secretion, hepatic insulin extraction, and gut hormone responses.
Procedure
This study will recruit 8 subjects with high fasting insulin (FPI) (FPI > 80 pmol/L) and 16
subjects with normal insulin level (8 with FPI <41 pmol/L and 8 with 41<FPI<80 pmol/L).
Approximately 100 subjects will be screened in order to determine their eligibility. You will be
asked to come to the Risk Factor Modification Center, St. Michael‘s Hospital (61 Queen‘s St.
185
East) on one morning (between 8:00-9:30am) and give a fasting blood sample (10 ml blood,
about 2 tsps). The blood sample will be used to measure insulin and proinsulin levels, and other
routine tests (glucose, cholesterol level) to see if you are eligible. You will be asked to fill out a
questionnaire only once at the beginning of the study; this includes questions about your age,
gender, ethnicity and your parents‘ ethnicities. You will be asked to have your height, weight,
waist, hip circumference and blood pressure measured. You will not be eligible for this study if
you have the following:
BMI is >35kg/m2( a ratio of your weight and height )
Fasting blood sugar >= 7.0mmol/L
Triglycerides (a measure of fat level in the blood) >= 10.0mmol/L
You will also be asked questions regarding your health status. If you answer ‗yes‘ to any of the
following questions you will not be able to participate in this study:
Do you have Diabetes?
Do you have a heart problem such as Congestive Heart Failure (CHF) or a history of
Heart Attack?
Have you had a Stroke or TIA (stroke symptoms that cleared up)?
Do you have any chronic Liver disease such as Hepatitis C?
Do you have Kidney Disease?
Are you HIV positive or diagnosed with AIDS?
Do you have any chronic stomach or bowel conditions such as Inflammatory Bowel
Disease (eg. Crohn‘s Disease or Ulcerative Colitis) or Malabsorption (a condition where
the body can‘t absorb the vitamins and other nutrients eaten)?
Have you had recent [within 3 months] emergency and surgery?
Have you been admitted to Hospital urgently for any other medical reasons [within 3
months]?
Are you pregnant or breast-feeding?
Are you on Diuretics (drugs that promote water loss from the body, often used for High
Blood Pressure and CHF) eg. HCTZ (a thiazide) or Spironolactone.
Are you on oral Corticosteroids (steroid hormones) or Beta-Blockers (a common heart
medication)?
If you are selected for this study, you will be asked to attend 11 clinic visits over a period of 3
months at the Risk Factor Modification Center, St. Michael‘s Hospital (61
Queen‘s St. East). You will be asked to come on mornings (between 8:00-9:30am) after an
overnight fasts (no food or drink for 10-14 hours). Each visit is expected to take about 2 hours of
your time.
At the clinic, you will be fed test meals consisting of 50g anhydrous glucose dissolved in 250ml
water to which will be added 0, 5 and 30g of whey protein concentrate (Fonterra Inc. USA) and
0, 5 and 30g fat (canola oil). Venous blood samples (blood will be taken from a needle inserted
into a vein in your arm) will be collected by the nurse at fasting and at 15, 30, 45, 60, 90 and
120min after starting to eat.
186
For your each visit, you will give 7 blood samples, and each sample will collect 10 ml of blood;
therefore, 70ml blood for each visit (only require one visit in a week). In total, 770 ml blood will
be drawn during approximately 3 month period, which is safe on its own because in blood
donation, donors can donate 450 ml blood every 2 month. You are advised not to donate
blood during and 3 months after the study.
The blood samples will be used to measure the levels of glucose, insulin, and gut hormones.
During the 2 hrs of the test, a short physical activity questionnaire will be administered. It will
only take 5 minutes to fill in the form.
To assess dietary nutrient intake, you will be asked to complete two 3-day food records at the
beginning and end of the study period (appendix A). Each record will take approximately 20
minutes of your time.
You will be asked to repeat test if the results were poor.
Any abnormal measures (i.e. if the blood pressure was 140/90 or above) will be brought to the
attention of Dr. Wolever, a physician at St. Michael's Hospital, who will advise Ms. Lan-
Pidhainy as to whether you should be advised to seek follow-up with your family doctor. Dr.
Wolever will be available to discuss this with you if you desire, and will arrange appropriate
follow-up at St. Michael's Hospital if you does not have a family doctor and desires that course
of action.
Test Meals
You will be given the test meals in a random order. You will be fed the following test meals
during the 11 clinical visits. The control meal will be consumed at the beginning, the middle and
the end of the study.
1. control : 50 g glucose alone ( 0 g fat and protein)
2. 50g glucose + 5g protein
3. 50g glucose + 30g protein
4. 50g glucose + 5g fat
5. 50g glucose + 5g fat + 5g protein
6. 50g glucose + 5g fat + 30 protein
7. 50g glucose + 30g fat
8. 50g glucose + 30g fat + 5g protein
9. 50g glucose + 30g fat + 30g protein
Risks
The risks involved in the study are very low. For each test, 7 venous blood samples (70ml
blood) will be taken. The risks associated with having venous blood drawn include discomfort,
and possible dizziness, fainting or feeling lightheaded, redness, swelling, hematoma (blood
187
accumulating under the skin), infection (a slight risk any time the skin is broken). Since an
experienced registered nurse (RN) will be hired for blood sampling, the risks of these problems
are considered extremely low.
Benefits
You will not receive any direct benefit by participating in this study. You do not have to
participate.
Alternatives to Participation
This is not a treatment study. You do not have to participate in this study.
Protecting Your Health Information
Confidentiality will be respected and no information that discloses your identity will be released
if results of this study are published. Your identity will not be disclosed without your permission
unless required by law.
Your blood samples will be sent for biochemical analysis to the Banting and Best Core Diabetes
Laboratory, located at the Mount Sinai Hospital in Toronto. The labels on your blood samples
will only contain your initial and ID number. The consent forms, data sheets and questionnaires
will be kept securely in locked file cabinets in St.Michael‘s
Hospital. The study co-investigator will require access to your file and identification number
and the principle investigator may require access to your personal health information. The
likelihood of the principle investigator needing access to your identification information is very
low. All data will be stored electronically and will only be identified by initial and ID numbers.
And all data will be saved in password protected files.
At the end of the study, the files linking the ID number with the names will be destroyed by the
study coordinator. The results will be used in scientific meetings and publications, but subject
identity will not be revealed in these presentations.
If you withdraw from the study, your blood samples and any data collected up
until the time you decide to withdraw from the study will be destroyed
immediately.
Study Results
If you were interested in the study results, you may contact either Dr. Wolever or Ms. Lan-
Pidhainy. For the published results, you may find in journals such as American Journal of
Clinical Nutrition, European Journal of Clinical Nutrition, Journal of Nutrition, etc in the near
future.
Payment for Participation
For the screening test, you will be paid $15 for your time.
188
If you are selected for the study, you will be paid $40 for each test completed, and $40 for
completing two 3-day food records (at the beginning and end of the study period), for a total of $
480 if you complete the entire study.
You may be asked to repeat tests if the results are poor. As long as this is not a result of failure
to follow instructions, you will be paid $40 for each test repeated.
Compensation for Injury
If you suffer a physical injury as a direct result of the administration of the study procedures,
medical care may be obtained by you in the same manner as you would ordinarily obtain any
other medical treatment. In no way does this form waive your legal rights nor relieve the
investigator, sponsors or involved institutions from their legal and professional responsibilities.
Participation and Withdrawal
Participation in this research study is voluntary. If you choose not to participate, you and your
family will continue to have access to customary care at St.Michael‘s Hospital. If you decide to
participate in this study you can change your mind without giving a reason, and you may
withdraw from the study at any time without any effect on the care you and your family will
receive at St. Michael‘s Hospital.
New Findings or Information
We may learn new information during the study that you may need to know or that might make
you want to stop participating in the study. If so, you will be notified about any new information
in a timely manner. You may be asked to sign a new consent form discussing these new findings
if you decide to continue in the research study.
Research Ethics Board Contact
If you have any questions regarding your rights as a research participant, you may contact Julie
Spence, Chair, Research Ethics Board at 416-864-6060 ext. 2557 during business hours.
The study protocol and consent form have been reviewed by a committee called the Research
Ethics Board at St. Michael‘s Hospital. The Research Ethics Board is a group of scientists,
medical staff, individuals from other backgrounds (including law
and ethics) as well as members from the community. The committee is established by the
hospital to review studies for their scientific and ethical merit. The Board pays special
attention to the potential harms and benefits involved in participation to the research participant,
as well as the potential benefit to society.
189
This committee is also required to do periodic review of ongoing research studies. As part of this
review, someone may contact you from the Research Ethics Board to discuss your experience in
the research study.
Study Contacts
If you have any questions concerning your participation in this study or if at any time you feel
you have experienced a research-related injury you may contact the study doctor, Dr.Thomas
M.S. Wolever at (416) 978-5556 during business hours (Monday to Friday 9:00-5:00) or e-mail
him at [email protected].
190
Consent
The research study has been explained to me, and my questions have been answered to my
satisfaction. I have been informed of the alternatives to participation in this study. I have the
right not to participate and the right to withdraw without affecting the quality of medical care at
St. Michael‘s Hospital for me and for other members of my family. As well, the potential harms
of participating in this research study have been explained to me. I have been told that I have not
waived my legal rights nor released the investigators, sponsors, or involved institutions from
their legal and professional responsibilities. I know that I may ask now, or in the future, any
questions I have about the study. I have been told that records relating to me and my care will be
kept confidential and that no information will be disclosed without my permission unless
required by law. I have been given sufficient time to read the above information.
I consent to participate. I have been told I will be given a signed copy of this consent form.
Subject Name: _____________________________________________
Subject Signature: ________________________________ Date: ________________
Name and Position of Person Conducting Informed Consent Decision:
Name: _______________________________________ Position: _____________
Signature: _____________________________________ Date: _______________
191
Appendix 11 Screening form for carbohydrate study
Patient ID#:__________
Initial:______
Date of Birth: ____________
Research Study Tel#_ 7438
Location : XIAPP
Send results to: DR.WOLEVER 14147
Contact: XIAOMIAO 7438
XIAPP STUDY
CARBOHYDRATE STUDY
PHLEB
GLUF
LIP2F
CRP
INSU
CR
AST
ALT
GGT
HBAIC
192
Appendix 12 Concent form for the 2nd
study: postprandial response to different carbohydrates in
normal, hyperinsulinemic and type 2 diabetic subjects
Risk Factor Modification Center
St. Michael‘s Hospital
61 Queen‘s St. East
(416) 867-7438
Consent to Participate in a Research Study
Study Title: Postprandial responses to different carbohydrates in normal, hyperinsulinaemic and
type 2 diabetic subjects
Before agreeing to take part in this research study, it is important that you read the information in
this research consent form. It includes details we think you need to know in order to decide if
you wish to take part in the study. If you have any questions, ask a study doctor or study staff.
You should not sign this form until you are sure you understand the information. All research is
voluntary. You may also wish to discuss the study with your family doctor, a family member or
close friend. If you decide to take part in the study, it is important that you are completely
truthful about your health history and any medications you are taking. This will help prevent
unnecessary harm to you.
Investigator(s) PRINCIPAL INVESTIGATOR:
Thomas MS Wolever, PhD, DM,
Staff physician, St. Michael‘s Hospital
Professor, Department of Nutritional
Sciences
University of Toronto
Toronto, ON M5S 3E2
Tel: 416-978-5556
e-mail: [email protected]
CO-INVESTIGATOR
Xiaomiao Lan-Pidhainy,
Study Coordinator;
MSc, Ph.D student
Department of Nutritional Sciences
University of Toronto
Toronto, ON M5S 3E2
Tel: 416-867-7438
e-mail: [email protected]
Doctor of Philosophy Candidate Research Project
This study is to be conducted as part of the study coordinator, Ms Lan-Pidhainy‘s Ph. D degree
requirements. Ms Lan-Pidhainy has completed her Master‘s of Science (MSc) degree at the
University of Toronto. Dr. Wolever is Ms Lan-Pidhainy‘s advisor and Principal Investigator on
this study. This means he will be supervising the conduct of the study.
Conflict of Interest
Dr.Wolever is the president and part owner of Glycemic Index Laboratories (GI Labs), a contract
research organization specializing in determining the glycaemic index of foods. However, there
193
is no foreseeable commercial application of the results. Ms Lan-Pidhainy declares no conflicts
of interest.
Study Sponsor Canadian Institutes of Health Research
160 Elgin Street, 9th Floor
Address Locator 4809A
Ottawa, ON, K1A 0W9
CANADA
Purpose of the Research You are are being asked to consider participating in a nutrition research study on the after-meal
physiological responses to different carbohydrate foods in normal subjects, subjects with high
fasting insulin, and subjects with type 2 diabetes.
The Glycaemic Index (GI) and Insulinaemic Index (II) measure how much the carbohydrate in
foods increases your blood glucose and insulin. For these indexes to have a use in clinics, their
values must be the same in different subjects regardless of whether they have normal, high
fasting insulin or are diabetic. There is evidence that the GI values of foods are similar in normal
and diabetic subjects. However, the GI of foods has not been determined in subjects with high
fasting insulin. Thus we do not know if they are similar in normal vs. subjects with high fasting
insulin vs. type 2 diabetic subjects.
Because of the uncertainty regarding insulin response to carbohydrate foods, insulin index (II)
has also been measured; however, it is not known whether the II values are the same in normal
vs. subjects with high fasting insulin vs. type 2 diabetic subjects. Since insulin response is not
only glucose-dependent, but also connected with the normal functioning of gut hormones and
how insulin is cleared by liver. The II values may be subject-dependent. This means that II may
not be a valid measure of food property. Therefore, to see if II is a valid measure of
carbohydrate in foods, it is important to use ―pure‖ carbohydrates and carbohydrate foods to
measure II values in all conditions.
Procedure
This study will recruit 10 subjects with normal insulin level (FPI <40 pmol/L), 10 subjects with
high fasting insulin (FPI) (FPI > 40 pmol/L) and 10 subjects with type 2 diabetes. You will be
asked to come to the Risk Factor Modification Center, St. Michael‘s Hospital (61 Queen‘s St.
East, the 6th
floor) on one morning (between 8:00-9:30am) and give a fasting blood sample (10
ml blood, about 2 tsps). A fasting blood sample means that you cannot have any food or drinks
10-14 hours before your study visit. The blood sample will be used to measure glucose, insulin,
and lipids level to see if you are eligible. You will be asked to fill out a questionnaire only once
at the beginning of the study; this includes questions about your age, gender, ethnicity and your
parents‘ ethnicities. You will be asked to have your height, weight, waist, hip circumference and
blood pressure measured.
194
For the non-diabetic healthy subjects, you may not be eligible for this study if you have the
following:
BMI is >35kg/m2( a ratio of your weight and height )
Fasting blood sugar >= 7.0mmol/L
Triglycerides (a measure of fat level in the blood) >= 10.0mmol/L
You will also be asked questions regarding your health status. Your answers to these questions
will be used to determine your eligibility to participate in this study:
For subjects with type 2 diabetes, Dr.Wolever will interview you about your health status and
medication. You will be eligible for the study if you are:
18 - 70 y old
BMI<35kg/m2
( a ratio of your weight and height )
Fasting glucose>7.0mmol/L or
HbA1c (a measure of your long-term glucose control) over upper limit or normal or
Treated with any combination of sulfonylures, metformin, thiazolidinedione and/or
meglitinide.
If you are selected for this study, you will be asked to attend 8 clinic visits over a period of 2
months at the Risk Factor Modification Center, St. Michael‘s Hospital (61
Queen‘s St. East). You will be asked to come on mornings (between 8:00-9:30am) after an
overnight fasts (no food or drink for 10-14 hours). For non-diabetic subject, each visit is
expected to take about 2 hours of your time, for diabetic subjects, each visit will take about 3
hours of your time.
At the clinic, you will be fed test meals consisting of 50g available carbohydrate (total
carbohydrate minus dietary fibre) as 50g anhydrous glucose (Fisher Scientific, Mississauga,
ON), 50g sucrose (Lantic Sugar Ltd, Toronto, ON), 50 instant mashed potato, 104g white bread
(Wonder bread, Mississauga, ON), 50 g polished rice (Dainty brand, Peterborough, ON) and 50
pearled barley (Gouda brand, Peterborough, ON). Sugars will be dissolved in 250ml bottled
water. Instant potato will be weighed dry, and boiling water added according to package
directions. Two to four portions of rice and barley will be weighed dry, boiled in salted water
according to package directions on the morning of the test and the resulting cooked amount
weighed and divided into portions. You are asked to consume the food within 10-15 minutes.
For non-diabetic subjects, venous blood samples (blood will be taken from a needle inserted into
a vein in your arm) will be collected by the nurse at fasting and at 15, 30, 45, 60, 90 and 120min
after starting to eat. For type 2 diabetic subjects, blood samples will be collected in the same
manner at fasting and at 30, 60, 90, 120, 150 and 180min after starting to eat.
For your each visit, you will give 7 blood samples, and each sample will collect 10.5 ml of
blood; therefore, 73.5ml (about 5 tablespoons) blood for each visit (only require one visit in a
week). In total, 588 ml (about 1¼ pints) blood will be drawn during approximately 2 month
period, which is safe on its own because in blood donation, donors can donate 450 ml (about 1
195
pint) blood every 2 month. You are advised not to donate blood during the study and for 2
months after the study.
The blood samples will be used to measure the levels of glucose, insulin, C-peptide, free fatty
acids and gut hormone GLP-1.
During the 2 hrs of the test, a short physical activity questionnaire will be administered (at the
beginning and end of the study period). It will only take 5 minutes to fill in the form.
You may be asked to repeat some tests if the results that are obtained are poor.
Any abnormal measures (e.g. if the blood pressure was 140/90 or above) will be brought to the
attention of Dr. Wolever, a physician at St. Michael's Hospital, who will advise as to whether
you should seek follow-up with your family doctor. Dr. Wolever will be available to discuss this
with you if you desire, and will arrange appropriate follow-up at St. Michael's Hospital if you
does not have a family doctor and desire that course of action.
Test Meals
You will be given the test meals in a random order. You will be fed the following test meals
during the 8 clinical visits. The control meal will be consumed at the beginning, the middle and
the end of the study.
10. control : 50 g glucose alone
11. Sucrose
12. Instant mashed potato
13. White bread
14. Polished rice
15. Pearled barley
Potential Harms (Injury, Discomforts Or Inconvenience)
The risks involved in the study are very low. For each test, 7 venous blood samples (70.5ml
blood or 5 tablespoons) will be taken. The risks associated with having venous blood drawn
include discomfort, and possible dizziness, fainting or feeling lightheaded, redness, swelling,
hematoma (blood accumulating under the skin), infection (a slight risk any time the skin is
broken). Since an experienced registered nurse (RN) will be hired for blood sampling, the risks
of these problems are considered extremely low. With the exception of glucose (approximately
10% of subjects may experience nausea); the likelihood of the other test foods (instant mashed
potato, white bread, rice, pearled barley, sucrose) causing nausea is very rare, but you may
experience some minor stomach discomfort and bloating.
Potential Benefits
You will not receive any direct benefit by participating in this study. You do not have to
participate. However, results from this study may further medical or scientific knowledge.
Alternatives to Participation
This is not a treatment study. You do not have to participate in this study.
196
Protecting Your Health Information
The study investigators are committed to respecting your privacy. No other persons will have
access to your personal health information or identifying information without your consent,
unless required by law. Any medical records, documentation, or information related to you will
be coded by study numbers to ensure that persons outside of the study will not be able to identify
you. Your blood samples will be sent for biochemical analysis to Mount Sinai Hospital Core
laboratory and St. Michael‘s Hospital Core Laboratory (both located in Toronto). The labels on
your blood samples will only contain your ID number. No identifying information about you will
be allowed off site. All information that identifies you will be kept confidential and stored and
locked in a secure place that only the study investigators will have access to. In addition,
electronic files will be stored on a secure hospital or institutional network and will be password
protected. It is important to understand that despite these protections being in place, experience
in similar studies indicates that there is the risk of unintentional release of information. The
principal investigator will protect your records and keep all the information in your study file
confidential to the greatest extent possible. The chance that this information will accidentally be
given to someone else is small.
By signing this form, you are authorizing access to your medical records by the study
investigators and St. Michael‘s Hospital Research Ethics Board. Such access will be used only
for purposes of verifying the authenticity of the information collected for the study, without
violating your confidentiality, to the extent permitted by applicable laws and regulations.
National and Provincial Data Protection regulations, including the Personal Information
Protection and Electronic Documents Act (of Canada) or PIPEDA and the Personal Health
Information Protection Act (PHIPA) of Ontario, protect your personal information. They also
give you the right to control the use of your personal information, including personal health
information, and require your written permission for your personal information (including
personal health information) to be collected, used or disclosed for the purposes of this study, as
described in this consent form. You have the right to review and copy your personal
information. However, if you decide to be in this study or choose to withdraw from it, your right
to look at or copy your personal information related to this study will be delayed until after the
research is completed.
At the end of this study, the files linking the ID number with the names will be destroyed by the
study coordinator.
Study Results
The results of this study will be used in scientific meetings and publications, but your identity
will never be revealed in these presentations.
If you were interested in the study results, you may contact either Dr. Wolever or Ms. Lan-
Pidhainy. For the published results, you may find in journals such as American
Journal of Clinical Nutrition, European Journal of Clinical Nutrition, Journal of Nutrition, etc in
the near future.
197
Payment for Participation
For the screening test, you will be paid $15 for your time.
For the non-diabetic subjects, if you are selected for the study, you will be paid $40 ($20/hr, 2
hrs) for each test completed, for a total of $320 if you complete the entire study.
For the diabetic subjects, if you are selected for the study, you will be paid $60 ($20/hr, 3hrs) for
each test completed, for a total of $480 if you complete the entire study.
You may be asked to repeat tests if the results are poor. As long as this is not a result of failure
to follow instructions, you will be paid $40 or $60 for each test repeated.
If you do not complete the study successfully, you will be compensated on a pro-rated basis at
$20/hr for the time you have spent participating in the study.
Compensation for Injury
If you suffer an injury from study procedures or taking the study foods or participation in this
study, medical care will be provided to you in the same manner as you would ordinarily obtain
any other medical treatment. In no way does signing this form waive your legal rights nor release
the study doctor(s), sponsors or involved institutions from their legal and professional
responsibilities.
Participation and Withdrawal
Participation in this research study is voluntary. If you choose not to participate, you and your
family will continue to have access to customary care at St.Michael‘s Hospital. If you decide to
participate in this study you can change your mind without giving a reason, and you may
withdraw from the study at any time without any effect on the care you and your family will
receive at St. Michael‘s Hospital. If you withdraw from the study, your blood samples and any
data collected up until the time you decide to withdraw from the study will be destroyed
immediately (unless you request otherwise).
New Findings or Information
We may learn new information during the study that you may need to know or that might make
you want to stop participating in the study. If so, you will be notified about any new information
in a timely manner. You may be asked to sign a new consent form discussing these new findings
if you decide to continue in the research study.
Research Ethics Board Contact
The study protocol and consent form have been reviewed by a committee called the Research
Ethics Board at St. Michael‘s Hospital. The Research Ethics Board is a group of scientists,
medical staff, individuals from other backgrounds (including law and ethics) as well as members
from the community. The committee is established by the hospital to review studies for their
scientific and ethical merit. The Board pays special attention to the potential harms and benefits
involved in participation to the research participant, as well as the potential benefit to society.
198
This committee is also required to do periodic review of ongoing research studies. As part of this
review, someone may contact you from the Research Ethics Board to discuss your experience in
the research study.
If you have any questions regarding your rights as a research participant, you may contact Julie
Spence, Chair, Research Ethics Board at 416-864-6060 ext. 2557 during business hours.
Study Contacts
If you have any questions concerning your participation in this study or if at any time you feel
you have experienced a research-related injury you may contact the study doctor, Dr.Thomas
M.S. Wolever at (416) 978-5556 during business hours (Monday to Friday 9:00-5:00) or e-mail
him at [email protected].
199
Consent
The research study has been explained to me, and my questions have been answered to my
satisfaction. I have been informed of the alternatives to participation in this study. I have the
right not to participate and the right to withdraw without affecting the quality of medical care at
St. Michael‘s Hospital for me and for other members of my family. As well, the potential harms
of participating in this research study have been explained to me. I have been told that I have not
waived my legal rights nor released the investigators, sponsors, or involved institutions from
their legal and professional responsibilities. I know that I may ask now, or in the future, any
questions I have about the study. I have been told that records relating to me and my care will be
kept confidential and that no information will be disclosed without my permission unless
required by law. I have been given sufficient time to read the above information.
I consent to participate. I have been told I will be given a signed copy of this consent form.
Subject Name: _____________________________________________
Subject Signature: ________________________________ Date: ________________
Name and Position of Person Conducting Informed Consent Decision:
Name: _______________________________________ Position: _____________
Signature: _____________________________________ Date: _______________
