| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
Department of Physiology, Adelaide University (L.K.H.), Adelaide, South Australia, AUSTRALIA
Health Sciences & Nutrition (M.N., P.M.C.), Adelaide, South Australia, AUSTRALIA
Dr. Peter Clifton, CSIRO, Health Science and Nutrition, PO Box 10041 BC, Adelaide SA, 5000, AUSTRALIA. E-mail: peter.clifton{at}hsn.csiro.au
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Methods: Twenty-three female and twenty-two male overweight subjects participated in 12 weeks of energy restriction (average BMI 33.2 kg/m2, age 56.7 years, glycated hemoglobin (GHb) 6.7%). After a four-week run-in on a high saturated fat (SFA) diet (1540 kcal/day, 17% SFA), the free-living subjects were randomly assigned to either a high- (75 GI units) or low- (43 GI units) GI diet (1440 kcal/day, 60% carbohydrate, 5% SFA) for eight weeks. Weight, serum lipids, plasma glucose and glycated hemoglobin were measured every four weeks. An oral glucose tolerance test (OGTT) was also performed at baseline, weeks 4 and 12. From the baseline OGTT results subjects were divided into three groups of low, median and high glucose tolerance.
Results: At baseline, BMI, age and glycated hemoglobin concentrations were not different between subjects allocated to the high- or low-GI diets. After four weeks, weight loss was 3.6 ± 0.3 kg. Fasting glucose (-5.6%), glycated hemoglobin (-2.8%), area under the glucose curve (-13.0%) and triglyceride (-13.8%) concentrations were reduced (p < 0.02). Between weeks 4 and 12 reductions were observed in weight (-4.9%), fasting glucose (-4.6%), area under glucose curve (-10.1%), glycated hemoglobin (-7.2%), triglyceride (-7.5%) and LDL-C (-13.2%) concentrations. Weight loss was not different between low and high-GI diets. However, glycated hemoglobin was reduced twofold more in subjects consuming a low-GI diet as compared to subjects consuming a high-GI diet, but this was not statistically significant. LDL concentrations were also reduced more in subjects with low glucose tolerance on the low-GI diet (p = 0.02).
Conclusion: Weight loss produces substantial improvements in glycemic control and lipoprotein metabolism. Lowering the glycemic index of high carbohydrate, low fat diets increases the fall in LDL cholesterol in subjects with type 2 diabetes with low glucose tolerance, but has little effect on glycemic control.
Key words: type 2 diabetes, weight loss, lipids, glycemic index, diet, glucose tolerance
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Weight loss in obese patients promotes substantial improvements in insulin sensitivity, blood pressure, triglycerides and LDL [8]. Generally, weight loss diets are low in fat and high in carbohydrate. However, it is unclear whether reducing the glycemic index of high carbohydrate diets will confer a similar benefit during energy restriction as that observed in energy balance [3]. Therefore, in this study we investigated if GI impacts on glucose and lipid metabolism during weight loss in overweight subjects with variable glucose tolerance. Based on previous trials in weight maintenance we hypothesised that low GI energy restricted diets will further enhance the improvement in glucose tolerance and plasma lipids as compared to high-GI energy restricted diets.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
11.1 mmol/L at baseline. Twelve women and 13 men (n = 25) were classified with low glucose tolerance (LGT). Ten subjects were classified as having median glucose tolerance (MGT, >7.8 mmol/L at two hours) while ten subjects had high glucose tolerance (HGT, <7.8 mmol/L at two hours).
Diet Composition
The dietary interventions were designed to be similar in energy, and restricted subjects to approximately 1500 kcal/day. For the first four weeks all subjects consumed a diet that was relatively high in SFA (32% fat, 17% saturated fat, 50% CHO, 20% protein) before either high or low-GI diets were consumed. The GI diets were high in carbohydrate (60%) and low in fat (15%) and contained similar amounts of protein (20%). The GI of each diet was calculated by constructing menus of foods differing in their published GI [9]. The GI score for the low- and high-GI diets based on previous experiments was 43 and 75 units respectively. To maximise compliance, subjects were provided with ‘key’ foods, including breakfast cereals, breads, biscuits and 4 low fat frozen protein foods per week. This provided approximately 60% of total energy intake. Key foods for the high-GI diet included a high GI cereal and fruits, wholemeal bread, potato flakes and plain sweet biscuits, while the low-GI diet included a low GI cereal and fruits, wholegrain bread, pasta and wheatmeal biscuits (Table 2). Subjects attended dietary consultations every two weeks with the research dietitian to ensure compliance and were given detailed instruction on how to manipulate the diets in keeping with appropriate glycemic index and saturated fat contents. Subjects were required to complete three-day detailed records of their dietary intake every two weeks. The food records were analysed for daily energy and macronutrient intakes using "Diet 4 Nutritional Calculation" software (Xyris Software, Highgate Hill, QLD, Australia) based on nutritional data obtained from Australian food tables and food manufacturers [10].
|
Statistical Analysis
Data are expressed as means ± SEM. Data from two consecutive visits at weeks 0, 4, 8 and 12 were averaged for statistical analysis. Data was analysed at weeks 0 and 4 for all subjects together and then at weeks 4 and 12 to test the differential effects of GI with repeated-measures GLM and age, gender, BMI, saturated fat intake and diabetes status as co-factors using SPSS 10.0 for Windows statistical software (Chicago, USA). A one-way ANOVA was used to test baseline differences and differences in energy and nutrient composition. Post hoc analysis was performed using Bonferonni where necessary. The analysis was also conducted with week 0 variables as cofactors, but this did not appreciably alter the significance of the diet by time interaction. Correlations were performed using Pearson correlation coefficients. Chi square analyses were used to determine the distribution of gender and diabetes status by high- and low-GI diet. Significance was set at p 
0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
1440 kcal/day), and protein intake was significantly higher than expected (p < 0.001). Energy intake was not different between high- and low-GI groups. However, carbohydrate intake was 2.8% higher (p < 0.001), and saturated fat intake was 1% higher (p = 0.02) in subjects consuming a high-GI diet as compared to subjects consuming a low-GI diet. No other macronutrients were statistically different between high and low GI groups.
|
High SFA Phase (n = 45)
Average weight loss was 3.6 ± 0.3 kg after the first four weeks of energy restriction when all subjects were consuming a diet that was relatively high in saturated fat. Fasting glucose was reduced 6% (p = 0.008), area under the curve (AUC) for glucose was reduced 13% (p = 0.001), glycated hemoglobin was reduced 3% (p = 0.02), and urinary glucose excretion was reduced 74% (p = 0.05). Triglyceride concentrations were also significantly reduced by 14% (p = 0.007), and total cholesterol was reduced 4% (p = 0.004). However, LDL and HDL concentrations were unchanged. We also observed that systolic but not diastolic blood pressure was significantly reduced (p = 0.017). Strong correlations were observed between weight loss and the change in (
) AUC (r = 0.381, p = 0.01), GHb (0.474, p = 0.001), triglyceride (r = 0.306, p = 0.04), total cholesterol (r = 0.547, p < 0.001) and LDL (r = 0.374, p = 0.01).
High- vs. Low-GI Phase
The GI diets were implemented at week 4. At week 12, weight loss was not different between high- and low-GI groups (Fig. 1). Fasting plasma glucose was reduced a further 4% (high-GI, p = 0.01) and 5% (low-GI, p = 0.001) with no significant difference between diets (Table 4). GHb was reduced from 6.35% to 6.06% on the high-GI diet, a reduction of 4.6% (p = 0.03) (Fig. 2). GHb was also reduced on the low-GI diet from 6.65% to 6.04%, a reduction of 9.1% (p = 0.002); this twofold difference between diets was not, however, statistically significant, and, adjusting for baseline GHb or two-hour plasma glucose levels did not change this. Almost all of the difference between the two diets was accounted for by two individuals on the low-GI diet who had absolute changes in GHb of 3.7% and 2.1%. The ratio between the change in GHb in the first four weeks on the high saturated fat phase and the change in the last eight weeks on the high-carbohydrate diets was not different between the low- and high-GI diets (p = 0.7). Area under the curve for glucose was reduced by a further 12% (high-GI) and 8% (low-GI) by week 12 (p < 0.001), with no significant effect of diet. Urinary glucose excretion was not significantly changed between weeks 4 and 12 in either group. No significant differences were observed in glycemic control between high- and low-GI groups when only subjects with low glucose tolerance (and or median glucose tolerance) were analysed.
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Before admission to the study, all subjects reported that their physicians had diagnosed them with type 2 diabetes in the previous ten years. In accordance with our previous experience, we found that only 25 subjects were classified as type 2 diabetic according to WHO criteria. These subjects were classified with low glucose tolerance. A further ten subjects had impaired glucose tolerance which we classified as median glucose tolerance, but surprisingly ten subjects had high glucose tolerance. Presumably they had lost weight since the original diagnosis although the men with high glucose tolerance were 6 kg heavier than the men with low glucose tolerance. As the benefits of a low glycemic index diet have mainly been shown in diabetic populations, we analysed our results by degree of glucose tolerance, but this made very little difference to the results. We also observed that subjects with low glucose tolerance lost significantly less weight as compared to subjects with median or high glucose tolerance. However, the reason for this was unclear, as energy intake, age and initial BMI were not different between the two groups. One possibility is that urinary glucose loss was substantial at baseline in the low glucose tolerance group and glucose excretion was reduced with weight loss. However, no correlation was found between total weight loss and initial urinary glucose excretion.
A large difference in glycemic index was achieved between groups. We provided 60% of daily energy intakes and supplied subjects with daily diet check-lists to promote compliance to the subjects’ allocated glycemic index. Furthermore, subjects completed three-day weighed food records on Sunday, Monday and Tuesday every two weeks, which were then individually reviewed by a dietitian providing a good estimate of actual nutrient and energy consumption by the subjects. We elected to use international tables of glycemic index rather than test all of the foods used in this study, as this is the way glycemic index is used in clinical practice and is equivalent to the use of food composition tables for assessment of dietary intake. No statistical difference was found between high and low GI diets in improving glycemic control during weight loss. However, GHb levels were reduced twofold more in subjects consuming a low GI diet, indicating that better glycemic control may be achieved on the low GI diet although in this study most of the difference was due to two very responsive subjects. This is in agreement with other GI studies during weight maintenance [5,6]. The lack of statistical significance may have been because not all subjects had type 2 diabetes or IGT as defined by WHO classifications, and GHb levels were normal in subjects with high glucose tolerance and did not change from weeks 4 to 12. However, based on the low glucose tolerant subjects alone, a study of 250 subjects would be required to achieve statistical significance in the difference observed, indicating that this study was insufficiently powered to detect this small effect. Furthermore, the differential effects of GI may be masked during energy restriction because of its powerful insulin sensitizing effect [21]. A weight maintenance period following weight loss may have allowed for statistical significance between groups. The effects of altering diet GI on glycemic control may also be modulated by the total carbohydrate content of the diet [22]. Wolever et al. [22] observed that second meal effects after a low GI breakfast were abolished when total carbohydrate intake was reduced. In the current study, the proportion of calories eaten as carbohydrate was similar to that of other energy balance studies that have shown a significant effect of GI on glycemic control [6,7]. However, total carbohydrate intake was approximately 30% lower than that in energy balance studies. Potentially, the reduced intake of carbohydrate consumed during energy restriction may reduce the insulin sensitising ability of the low GI diet.
Previous weight stable studies have shown that glycemic control is improved by reducing diet glycemic index [3]. However, the majority of these studies reduced diet GI by including legumes and removing bread [5,6]. Legumes may be responsible for the observed improvements in glycemic control and arguably may not be accepted as a daily diet choice by the general population. The low GI diet in the present study was achieved by altering only the types of bread consumed as well as substituting pasta for instant potato on the low GI diet. We also allowed subjects on the low GI diet to consume only fruits that have been reported to have very low GI and restricted very low GI fruit consumption in the high GI group. This made the diets similar without the possible confounding effects of legumes. Weight maintenance studies that have not added legumes to low GI diets have had different outcomes. Luscombe et al. [23] found no difference in the glucose response between low and high GI diets, but Jarvi et al. [24] found fructosamine and incremental blood glucose levels were lower after the low GI phase, which was achieved by modification of starch. Tsihlias et al. [26] found no difference between high and low GI cereals in a large six-month study. The negative results from this study were attributed to the use of diabetic subjects who had poor metabolic control and a small 10 unit change in GI, although one of the most successful studies only achieved a 13 unit change in GI [5]. In the present study only minimal differences were found in glycemic control between diets, indicating the low GI diet does not substantially enhance the benefit of energy restriction and weight loss.
The reductions in triglyceride and LDL were not different between groups overall, indicating the reductions were a result of weight loss and reduced saturated fat intake. However, in subjects with low glucose tolerance the fall in LDL cholesterol did appear to be dependent on GI (-0.61 mmol/L (low-GI) vs. -0.14 mmol/L (high-GI)). We may have underestimated the ability of the high-GI diet to reduce LDL, as saturated fat intake was marginally higher in this group. However, a 1% difference in saturated fat intake is unlikely to account for the large difference in LDL observed between the 2 groups, although adjustment for saturated fat intake rendered the difference in LDL statistically insignificant. Based on the formula of Mensink and Katan [26], LDL differences between high and low-GI groups should have been 0.03–0.04 mmol/L. Low GI diets have previously been shown to reduce LDL as compared to high GI diets in weight maintenance studies in subjects with hypertriglyceridemia and subjects with poorly controlled type 2 diabetes [24,27].
As the response to dietary change depends on a subject’s normal diet, we created a standard baseline diet before testing the effects of glycemic index. This diet approximated the average Australian diet and enabled us to test the effects of energy restriction, independently of fat restriction. During the initial high saturated fat phase, reductions in triglyceride and total cholesterol concentrations and improvements in glucose tolerance were observed. LDL concentrations were unchanged despite a 4% weight reduction. In the following eight weeks of low saturated fat consumption, LDL was reduced by 13%. Other energy restriction studies in this laboratory have produced similar results [20,28]. In ten subjects with type 2 diabetes no change was observed in LDL on a high SFA (17%), energy restricted diet at week 4 [20]. However, an 8% decrease was observed in 18 overweight non-diabetic subjects consuming a high SFA (17%) weight loss diet at week 4, but this was less than the 19% decrease in LDL observed in non-diabetic subjects consuming low saturated fat diets [28]. Although LDL changes are best assessed in an energy stable state at the end of weight loss, the fact that LDL, which normally falls with energy restriction, did not change after four weeks on a high saturated fat weight loss diet suggests that weight loss diets based on high levels of saturated fat may not produce the most optimal results. HDL concentrations were unchanged over the entire period of energy restriction. This conflicts with other studies, which have found that HDL levels drop after the implementation of energy restriction [20,21,28]. This suggests that increased fat consumption at least initially may prevent or attenuate the characteristic fall in HDL in response to energy restriction. This may be important in overweight subjects with type 2 diabetes with typically low HDL in whom monounsaturated fat could be substituted for saturated fat.
In conclusion this study suggests that altering the glycemic index of high carbohydrate energy restricted diets does not impact on the improvement in glycemic control that is observed normally during energy restriction. However, low glycemic index, high carbohydrate diets may be slightly better in improving glycemic control and lipoprotein metabolism in subjects who have low glucose tolerance, but this clearly needs confirmation.
|
ACKNOWLEDGMENTS |
|---|
Received January 16, 2001. Accepted November 30, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Pal, S. Lim, and G. Egger The Effect of a Low Glycaemic Index Breakfast on Blood Glucose, Insulin, Lipid Profiles, Blood Pressure, Body Weight, Body Composition and Satiety in Obese and Overweight Individuals: A Pilot Study J. Am. Coll. Nutr., June 1, 2008; 27(3): 387 - 393. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Livesey, R. Taylor, T. Hulshof, and J. Howlett Glycemic response and health a systematic review and meta-analysis: the database, study characteristics, and macronutrient intakes Am. J. Clinical Nutrition, January 1, 2008; 87(1): 223S - 236S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Murakami, S. Sasaki, Y. Takahashi, H. Okubo, Y. Hosoi, H. Horiguchi, E. Oguma, and F. Kayama Dietary glycemic index and load in relation to metabolic risk factors in Japanese female farmers with traditional dietary habits Am. J. Clinical Nutrition, May 1, 2006; 83(5): 1161 - 1169. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. W Krieger, H. S Sitren, M. J Daniels, and B. Langkamp-Henken Effects of variation in protein and carbohydrate intake on body mass and composition during energy restriction: a meta-regression 1 Am. J. Clinical Nutrition, February 1, 2006; 83(2): 260 - 274. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. F. Sheard, N. G. Clark, J. C. Brand-Miller, M. J. Franz, F. X. Pi-Sunyer, E. Mayer-Davis, K. Kulkarni, and P. Geil Dietary Carbohydrate (Amount and Type) in the Prevention and Management of Diabetes: A statement by the American Diabetes Association Diabetes Care, September 1, 2004; 27(9): 2266 - 2271. [Full Text] [PDF]
|
|
|
|
|
|
S. W. Rizkalla, L. Taghrid, M. Laromiguiere, D. Huet, J. Boillot, A. Rigoir, F. Elgrably, and G. Slama Improved Plasma Glucose Control, Whole-Body Glucose Utilization, and Lipid Profile on a Low-Glycemic Index Diet in Type 2 Diabetic Men: A randomized controlled trial Diabetes Care, August 1, 2004; 27(8): 1866 - 1872. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Franz Meta-Analysis of Low-Glycemic Index Diets in the Management of Diabetes: Response to Brand-Miller et al. and Mann Diabetes Care, December 1, 2003; 26(12): 3364 - 3365. [Full Text] [PDF]
|
|
|
|
|
|
J. Brand-Miller, S. Hayne, P. Petocz, and S. Colagiuri Low-Glycemic Index Diets in the Management of Diabetes: A meta-analysis of randomized controlled trials Diabetes Care, August 1, 2003; 26(8): 2261 - 2267. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Franz The Glycemic Index: Not the most effective nutrition therapy intervention Diabetes Care, August 1, 2003; 26(8): 2466 - 2468. [Full Text] [PDF]
|
|
|
|
|
|
A. Jimenez-Cruz, M. Bacardi-Gascon, W. H. Turnbull, P. Rosales-Garay, and I. Severino-Lugo A Flexible, Low-Glycemic Index Mexican-Style Diet in Overweight and Obese Subjects With Type 2 Diabetes Improves Metabolic Parameters During a 6-Week Treatment Period Diabetes Care, July 1, 2003; 26(7): 1967 - 1970. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. M. Bravata, L. Sanders, J. Huang, H. M. Krumholz, I. Olkin, C. D. Gardner, and D. M. Bravata Efficacy and Safety of Low-Carbohydrate Diets: A Systematic Review JAMA, April 9, 2003; 289(14): 1837 - 1850. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Franz and J. P. Bantle Response to Irwin Diabetes Care, July 1, 2002; 25(7): 1262 - 1263. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
