SEARCH

SEARCH BY CITATION

Keywords:

  • glycemic index;
  • glucose tolerance;
  • insulin sensitivity;
  • insulin secretion;
  • C-reactive protein

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Objective: To compare the effects of two calorie-restricted diets that differ in glycemic load (GL) on glucose tolerance and inflammation.

Research Methods and Procedures: Thirty-four healthy overweight adults, ages 24 to 42 years, were randomized to 30% provided calorie-restricted diets with high (HG) or low (LG) glycemic load for 6 months. Outcomes were changes in glucose-insulin dynamics and C-reactive protein (CRP) levels.

Results: Compared with baseline, levels of fasting insulin, homeostasis model assessment of insulin resistance, post-load insulin at 30 minutes, and incremental area-under-the-curve-insulin during the oral glucose tolerance test were significantly lower in both groups at 6 months (p range, 0.01 to 0.05), but after adjustment for baseline values and weight change, there were no differences between the two groups with regard to changes over time in any parameter. The mean percentage change in insulin sensitivity by a frequently sampled intravenous glucose tolerance test was +26% in the HG group and +24% in the LG group (p = 0.83); first-phase acute insulin release was −20% in the HG group and −21% in the LG group (p = 0.77). More participants on the LG diet (14 of 16 subjects) had a decline in serum CRP, compared with those on the HG diet (7 of 16 subjects) (p < 0.05).

Discussion: In healthy overweight adults provided with food for 6 months, the dietary GL did not seem to influence chronic adaptations in glucose-insulin dynamics above that associated with weight loss. This finding highlights the importance of absolute weight loss over the dietary macronutrient composition used to achieve weight loss. The finding of greater declines in CRP concentration after consumption of a low-GL diet warrants further investigation.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The incidence and prevalence of type 2 diabetes (t2DM)1 are increasing at an alarming rate both in the United States and worldwide (1). Insulin resistance, impaired pancreatic β cell function, and low-grade systemic inflammation are important contributors to the development of glucose intolerance and t2DM (2, 3). Weight loss through energy restriction and increased energy expenditure is recognized to be successful in reducing t2DM risk through improvements in insulin sensitivity, insulin secretion, and systemic inflammation (4, 5, 6). Specific dietary factors or patterns may also be important, but this area remains controversial (7).

One dietary factor that may influence the development of t2DM is the amount and type of carbohydrate as defined by the dietary glycemic load (GL) [GL = glycemic index (GI) × carbohydrate amount. The GI is defined as the area under the glycemic response curve during a 2-hour period after consumption of 50 g of carbohydrate from a test food, and values are expressed relative to the effect of either white bread or glucose] (8). Several observational studies have suggested that diets high in GI (9, 10) or GL (11, 12) increase t2DM risk, but others have not found such an association (10, 13, 14). Weight gain, impaired insulin secretion, increased insulin resistance, and systemic inflammation have all been implicated as potential mechanisms mediating the association between the dietary GL and t2DM (15, 16). Diets low in GL are cautiously recommended by the American Diabetes Association for prevention of t2DM (17). However, questions remain whether integrating the GL concept in energy-restricted nutritional interventions will decrease risk for glucose intolerance and t2DM above and beyond any benefits seen by weight loss (17). The controversy persists, at least in part, because of lack of data from long-term intervention studies with controlled diets that have focused on GL as a nutritional intervention modality for prevention of glucose intolerance.

As part of a larger investigation of the effect of calorie restriction in healthy overweight adults, we used data from the first 6 months of a longer trial, when all food was provided, to test the hypothesis that two calorie-restricted diets that differ in GL have differential effects on major risk factors for development of t2DM, in particular glucose-insulin dynamics and plasma C-reactive protein (CRP) concentration.

Research Methods and Procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

This randomized controlled trial was conducted at the Human Nutrition Research Center on Aging at Tufts University with approval from the Tufts-New England Medical Center Human Investigation Review Committee, and written informed consent by all participants (Clinicaltrials.gov identification # NCT00099099). The study was one of three independent trials that make up the first phase of the Comprehensive Assessment of Long-term Effects of Reducing Intake of Energy study.

Study Participants

Healthy adults ages 24 to 42 years with a BMI of 25 to 29.9 kg/m2 and a fasting plasma glucose level of <100 mg/dL (5.6 mM) were recruited from the greater Boston metropolitan area. Participants underwent a three-step screening process, during which they were excluded if they met the following criteria: greater than a 15-pound weight change during the previous year; known serious medical condition (such as diabetes, cancer, heart disease, endocrine disorder, acquired immunodeficiency syndrome, depression, eating disorder); anemia; hypertension; abnormal electrocardiogram; liver, kidney, or thyroid dysfunction; strong family history of heart disease, cancer, or diabetes; pregnancy, lactation, or plans to become pregnant in the following year; or heavy participation in sports activities (>12 h/wk). We also excluded participants with known nutritional and lifestyle issues that could prevent participation in and completion of the study and those unwilling or unable to complete an accurate food record. Data presented here were acquired during the first 6 months of the trial, when all food was provided to the participants.

Study Protocol

Nutritional Intervention

After a 7-week baseline period, when usual energy requirements for weight stability were assessed as total energy expenditure measured using the doubly labeled water method (18), participants were randomized for 24 weeks to either a high-GL diet (HG) (60% carbohydrate, 20% protein, 20% fat, 1 kcal/g energy density, fiber 15 g/1000 kcal, mean estimated daily GI of 86, and mean estimated daily GL of 116 g/1000 kcal) or a low-GL diet (LG) (40% carbohydrate, 30% protein, 30% fat, 1 kcal/g energy density, fiber 15 g/1000 kcal, mean estimated daily GI of 53, and mean estimated daily GL of 45 g/1000 kcal). Both diets were provided at 30% calorie restriction compared with individual baseline weight maintenance energy requirements. There were two additional small groups (n = 6 in each group) that received the two diets at 10% caloric restriction. These groups are not included in the present analysis because their main function was to provide us with experience in recruiting a control group and there were insufficient data for the present analyses. The GI and GL of the diets were determined using international tables of GI and GL (19).

Both diets approximated current dietary recommendations for healthy macronutrient ranges and contained Dietary Reference Intakes of micronutrients and essential fatty acids (20). The HG diet had higher GL, with adequate fiber to meet current dietary recommendations (20), emphasized low-energy-dense foods, including use of whole grains rather than refined carbohydrates, limited liquid calories, and had a higher variety of low-energy-dense foods such as fruits and vegetables and a low variety of high-energy-dense foods. The LG diet was similar in all respects to the HG diet except that the macronutrient balance was changed and the carbohydrate sources were of low GI, based on published GI of different carbohydrate sources (19). The diets were matched initially for dietary variety and palatability. Participants were also provided with a multivitamin supplement and calcium 500 mg/d to ensure that Dietary Reference Intakes for micronutrients were met.

During the 6-month intervention period, all food was provided by the research center and collected for home consumption twice weekly by the participants or their designated representative. Participants were requested to consume only this food; however, they were to report additional foods and drinks if any were eaten. To maximize adherence to the study diet, participants attended regular behavioral group meetings and individual sessions with a dietitian. From participants’ reports of leftover food and extra items, actual daily nutrient intake during the intervention period was calculated (21).

Randomization-Sequence Generation and Allocation Concealment

The study participants were randomized to the two diet groups and block-stratified by sex and BMI (27.5 or >27.5 kg/m2). Participants were not informed of their randomization for the first 12 weeks of the intervention. The dietary intervention study personnel were non-blinded, whereas all outcome and data management study personnel were blinded to the treatment allocation.

Study Outcomes

Body weight (to ±100 grams) was measured weekly using an electronic calibrated scale (Model CN-20; Detecto-Cardinal Scale Manufacturing Co., Webb City, MO). Height (to ±0.1 cm) was measured at baseline using a wall-mounted stadiometer. At the end of the baseline period and in the intervention period at Weeks 12 and 24, participants resided at the Human Nutrition Research Center on Aging for 1 and 2 days, respectively, while continuing to eat their assigned diet. Metabolic measurements were performed after an overnight 12-hour fast, as described below.

Homeostasis Model Assessment of Insulin Resistance (HOMAIR)

Insulin sensitivity in the fasting state was estimated by HOMAIR, an index based on fasting glucose and insulin values, which is calculated as: HOMAIR = [glucose (mM) × insulin (mU/L)]/22.5 (22). This model assumes that normal participants have an insulin resistance of 1. High HOMAIR scores denote low insulin sensitivity (increased insulin resistance), and they correlate well with insulin sensitivity obtained from the euglycemic clamp procedure (23).

Oral Glucose Tolerance Test (OGTT)

After a 12-hour overnight fast, 75 grams of glucose was given orally and a blood draw for glucose and insulin was done before (0 minutes) and at 30, 60, 90, and 120 minutes after the glucose load. The insulin level at 30 minutes after the glucose load (INS30) was used as a measure of pancreatic β cell early-phase insulin release (24). Incremental areas under the curve (AUC) for the glucose (AUC-glucose) and insulin (AUC-insulin) responses were determined as measures of glucose tolerance and insulin sensitivity, respectively (25).

Frequently Sampled Intravenous Glucose Tolerance Test (FSIVGTT)

In a subgroup of participants (those who consented to the procedure and in whom it was technically feasible, n = 15 in the HG group and n = 11 in the LG group), glucose and insulin dynamics were assessed with an FSIVGTT at 0 and 6 months (26). After a 12-hour overnight fast, an intravenous cannula was inserted into each arm. One was used for infusion of glucose and insulin, and the other was used for withdrawal of blood. Dextrose (300 mg/kg total body weight) was injected intravenously over 2 minutes at time 0, and insulin (0.03 U/kg total body weight; Novolin R; NovoNordisk, Princeton, NJ) was injected rapidly at time 20 minutes. Blood for insulin and glucose was sampled at −15, −10, −5, 1, 2, 3, 4, 5, 6, 7, 8, 10, 12, 14, 16, 19, 22, 23, 24, 25, 27, 30, 40, 50, 60, 70, 80, 90, 100, 120, 140, 160, and 180 minutes. The data were analyzed through minimal model analysis using MinMod software (MinMod Millenium, version 5.18; MinMod, Inc., Pasadena, CA) to estimate insulin sensitivity index (Si). The incremental first-phase acute insulin response to glucose (AIRg) was determined by calculating the AUC-insulin above the baseline in the first 10 minutes after glucose infusion. The disposition index (DI, the product of Si and AIRg), was calculated as a measure of pancreatic β cell function that captures both insulin sensitivity and the ability of the β cell to secrete insulin in response to the prevailing insulin sensitivity (27). A low DI value indicates decreased β cell function.

Assays

Glucose was measured by the hexokinase method in a Cobas Mira Analyzer (Roche Diagnostics, Indianapolis, IN) with intra- and inter-assay coefficients of variation (CV) of 1.4 and 2.1%, respectively. Insulin was measured by a radioimmunoassay commercial kit (Linco Research, Inc., St. Charles, MO) with intra- and inter-assay CVs of 3.1 and 3.8%, respectively. High-sensitivity CRP, as a marker of systemic inflammation (28), was measured using a commercially available solid-phase chemiluminescent immunometric assay (Diagnostic Products Corp., Los Angeles, CA) with intra- and inter-assay CVs of 4.0 and 6.5%, respectively.

Statistical Analysis

Data on 32 (of 34) participants who completed the 6-month intervention period were analyzed. To examine differences in baseline characteristics between groups, we used Student's t test for differences in means for continuous data and the χ2 test for differences in proportions for categorical variables. To compare within-group mean changes in outcomes (baseline, 3 months, and 6 months after the intervention), we used Student's paired t test. To compare between-group differences over time for all outcomes, we used general linear models adjusting for baseline values and changes in weight (PROC GLM procedure in SAS software; SAS Institute, Inc., Cary, NC). Statistical significance was set at p < 0.05. Data are presented as means ± standard error of the mean (SEM), unless otherwise noted. Statistical analysis was performed using SAS, version 8.2. The study was independently monitored for overall compliance and data accuracy by an external clinical trial monitor from the Duke Clinical Research Institute, Durham, NC. Clinical trial safety and efficacy were monitored by a data safety monitoring board.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Baseline Characteristics and Dietary Intervention

At baseline, the mean age of the 32 participants who entered the analysis was 34.6 years, and their mean BMI was 27.5 kg/m2 (Table 1). The mean target energy intake for the entire cohort was 1966 kcal/d, and the mean reported daily energy intake at 6 months did not differ between the two groups (2017 kcal on the HG diet vs. 1972 kcal on the LG diet, p = 0.70).

Table 1.  Baseline characteristics of study participants who completed the 6-month dietary intervention
CharacteristicHG (n = 16)LG (n = 16)p
  • HG, high glycemic; LG, low glycemic; HOMAIR, homeostasis model assessment of insulin resistance; CRP, C-reactive protein; OGTT, oral glucose tolerance test; AUC, area under the curve; FSIVGTT, frequently sampled intravenous glucose tolerance test; Si, insulin sensitivity index; AIRg, acute insulin response to glucose; DI, disposition index. Data are presented as means ± standard error of the mean, unless otherwise indicated. p Values are for Student's t test for differences in means or the χ2 test for differences in proportions for categorical variables in comparisons between the two groups.

  • *

    For FSIVGTT measurements, data included 16 participants in the HG group and 12 participants in the LG group.

Sex [no. (%) women]13 (76)12 (75)0.92
Age (years)34.3 ± 1.235.0 ± 1.50.72
Race [no. (%) white]14 (82)14 (88)0.62
Weight (kg)79.3 ± 3.178.8 ± 2.30.92
BMI (kg/m2)27.6 ± 0.427.6 ± 0.30.99
Energy requirement by doubly labeled water method (kcal/d)2822 ± 1222758 ± 990.69
Prescribed energy intake (kcal/d)1956 ± 881931 ± 700.83
Fasting measurements and indices   
 Plasma glucose (mg/dL)83.8 ± 1.783.8 ± 1.60.75
 Fasting insulin (mU/L)11.1 ± 1.012.2 ± 1.20.22
 HOMAIR2.3 ± 0.22.5 ± 0.30.18
 CRP (mg/L)2.2 ± 0.63.1 ± 0.70.65
OGTT measurements and indices   
 2-hour glucose (mg/dL)107.1 ± 8.0118.2 ± 6.40.24
 2-hour insulin (mU/L)49.2 ± 8.654.6 ± 9.50.56
 AUC-glucose585 ± 28609 ± 250.40
 AUC-insulin338 ± 35335 ± 270.57
FSIVGTT measurements*   
Si (10−4/min/mU/mL)4.2 ± 0.54.4 ± 0.50.79
AIRg (mU/L/min)448 ± 80401 ± 330.64
DI1571 ± 2321730 ± 2260.65

Metabolic Outcomes Related to Risk of t2DM

Body Weight

Participants maintained a stable body weight during the baseline period while on their usual diet. At 3 months and 6 months, both groups achieved statistically significant (p < 0.001) weight loss compared with their baseline weight. Adjusted for baseline weight, weight loss was equivalent in the two groups (7.2 kg in the HG group vs. 7.7 kg in the LG group at 6 months, p = 0.69).

Plasma Glucose, Serum Insulin, and Insulin Resistance by HOMAIR

The mean ± SEM fasting plasma glucose of the participants was 84.0 ± 0.2 mg/dL (4.7 ± 0.01 mM). At baseline, HOMAIR correlated with BMI, as anticipated (r = 0.48, p = 0.01). Changes in fasting glucose, insulin, and HOMAIR during the course of the trial are shown in Figure 1. Compared with baseline, within-group small declines in fasting glucose at 3 and 6 months were not statistically significant (Figure 1A); however, within groups, fasting insulin and HOMAIR were lower at 6 months compared with baseline levels (Figure 1B and 1C). After adjusting for baseline values and changes in weight, there were no statistically significant differences between the two groups at 3 or 6 months in fasting glucose, insulin, or HOMAIR.

image

Figure 1. Mean ± SEM plasma glucose (A), serum insulin (B), and HOMAIR (C) during a 6-month feeding study of an HG diet (n = 16) vs. an LG diet (n = 16) in overweight adults with normal fasting glucose at baseline. * p < 0.05 for within-group change from baseline. p = not significant for changes between groups (adjusted for baseline values and change in weight). To convert to SI units, glucose (mM = 0.0555 × mg/dL), and insulin (pM = 7.175 × mU/L).

Download figure to PowerPoint

OGTT Measures

The glucose and insulin response to a 75-gram OGTT before and after the 6-month dietary intervention is shown in Figure 2. Within groups, there were no statistically significant differences in post-load timed glucose concentrations compared with baseline (Figure 2A, 2C and 2E); however, INS30 after the glucose load (Figure 2B and 2D) and AUC-insulin (Figure 2F) were significantly lower at 6 months in both groups compared with baseline (p < 0.01). After adjustment for baseline values and change in weight, no statistically significant differences were observed between the two groups in post-load glucose or insulin values at any individual time-points, or in the AUC for either glucose or insulin.

image

Figure 2. Mean ± SEM plasma glucose (A and C), serum insulin (B and D), and AUC (E and F) after a 75-gram OGTT before and after a 6-month feeding study of an HG diet (A and B, n = 16) vs. an LG diet (C and D, n = 16) in overweight adults with normal fasting glucose at baseline. * p < 0.01 for within-group change from baseline. p = not significant for changes between groups (adjusted for baseline values and change in weight).

Download figure to PowerPoint

FSIVGTT Measures

In those who underwent an FSIVGTT, Si correlated with BMI at baseline, as anticipated (r = 0.33, p < 0.05). The mean percentage change at 6 months in Si was +26% in the HG group and +24% in the LG group. The mean percentage change in AIRg was −20% in the HG group and −21% in the LG group. After adjusting for baseline values and change in weight, there were no statistical differences in Si, AIRg, or DI between the two groups (data not shown). Figure 3 shows the relationship between Si and AIRg at baseline and at 6 months. The changes in these two parameters were of very similar degree and direction in both groups, and the direction of change was along the prediction curve after weight loss for individuals with normal glucose tolerance.

image

Figure 3. Relationship between Si and AIRg at baseline and after 6 months. The DI curve for euglycemic persons (constructed from the baseline data of the entire cohort) is shown as a dashed line.

Download figure to PowerPoint

CRP

At 6 months, mean plasma CRP concentration decreased from baseline by 35% (p < 0.01) in the LG group, whereas it remained essentially unchanged in the HG group (Figure 4). The difference in the mean CRP change at 6 months between the two groups did not achieve statistical significance after adjusting for baseline values and change in weight (−1.44 ± 0.44 mg/L in the LG group vs. 0.41 ± 0.91 mg/L in the HG group, p = 0.13). Owing to the expected large individual variability in CRP levels, we also analyzed the data by examining whether participants experienced an increase or a decline in CRP values from baseline to follow-up. In this analysis, more subjects (14 of 16 participants) on the LG diet experienced a decline in CRP compared with those on the HG diet (7 of 16 participants) (p <0.05 for χ2).

image

Figure 4. Mean ± SEM change in CRP after a 6-month feeding study of an HG diet (n = 16) vs. an LG diet (n = 16) in overweight adults with normal fasting glucose at baseline. * p < 0.01 for within-group change from baseline. p = not significant for changes between groups (adjusted for baseline values and change in weight). More participants (14 of 16) on the LG diet experienced a decline in CRP than those on the HG diet (7 of 16) (p <0.05 for χ2).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

In overweight individuals with normal fasting plasma glucose, we found that energy-restricted provided diets with low or high GL for 6 months had equivalent weight-loss adjusted effects on chronic adaptations in glucose-insulin dynamics. Our very well-controlled study highlights the importance of absolute weight loss, rather than dietary macronutrient composition, as the primary determinant of improvements in glucose-insulin dynamics during weight loss treatment programs.

High postprandial glucose excursions are predictors of development of glucose intolerance and t2DM (4), and carbohydrates are the component of the diet that influences the glycemic response the most. Therefore, decreasing daily postprandial glucose excursions and hyperinsulinemia with a low GL diet may decrease t2DM risk, presumably by means of changes in insulin sensitivity and/or pancreatic β cell function (8). In our study, we examined glucose-insulin dynamics by several methods: while participants were fasting, after a standard oral glucose challenge, and after an intravenous glucose challenge, in an effort to discern chronic influences of the dietary GL on various physiological measures that define glucose tolerance. Our finding that glycemia remained essentially unchanged in both groups during the weight loss intervention can be explained on the basis that participants had normal glucose tolerance at baseline, and, therefore, one would not expect glycemia to improve any further even after significant weight loss. Our finding of no change in glucose is in accord with previous studies of diets that differ in GL in subjects with normal baseline glucose tolerance (29, 30, 31, 32, 33, 34), including studies with very low-carbohydrate diets (35, 36). The declines in fasting and post-challenge timed and AUC-insulin levels after weight loss in both groups in the setting of maintaining euglycemia suggest an improvement in insulin sensitivity, which was confirmed by other measures of insulin sensitivity such as HOMAIR and Si. However, there were no weight loss-adjusted differences in insulin sensitivity measures between the two groups, which is in accord with most other trials (duration over 1 month) with glucose-tolerant subjects (31, 36, 37, 38, 39, 40), including studies with very-low-carbohydrate diets (35). In the two studies that showed improvements in HOMAIR with low-GL diets, glucose and insulin values were not provided and changes in HOMAIR were not adjusted for weight changes, to assess the weight-independent effect of the dietary GL (41, 42).

A limitation of using the OGTT to assess the effect of diets that vary in GL is that although the prevailing glycemia and insulinemia may be decreased with low-GI diets, this effect may not be evident during a standard OGTT. Furthermore, priming with a high-carbohydrate diet, such as a high-GL diet, improves the pancreatic β cell response, whereas chronic exposure to a low-GL diet may attenuate the response to a standard glucose load. Nevertheless, the OGTT does remain an accepted method for assessing glucose tolerance, and it also provides useful information regarding the ability of the pancreatic β cell to respond to a standard glucose load. In the subgroup of participants who underwent an FSIVGTT, Si increased and AIRg declined equivalently in both groups, resulting in a DI value that remained unchanged from baseline. Although there is a range of values for Si and AIRg, DI tends to remain constant in the euglycemic population. Individuals with low values of Si, AIRg, and DI are at increased risk for progressing to glucose intolerance and t2DM (43). In our study, the reciprocal changes in Si and AIRg resulting in an unchanged DI were expected, given that both groups remained euglycemic during the study and lost equivalent weight (27). A previous study in individuals with impaired glucose tolerance found no improvements in Si or AIRg among three diets that varied in GI or GL but found improvements in DI only in the group that lowered the dietary GI (40). In that study, the high-GI group achieved greater weight loss, which was not adjusted for in the analysis.

There are several possible explanations as to why risk factors for glucose tolerance did not differ between groups in our study. First, the effects of high- and low-GL/GI diets may be stronger in patients at very high risk for t2DM (obese, sedentary, genetically susceptible) (15). Although our participants were overweight, they had normal fasting glucose at baseline and only mild insulin resistance. Therefore, the potential beneficial effects of a low-GL diet may not be evident in this population, because they are less susceptible to changes in glucose-insulin dynamics. Second, because weight is the most important risk factor for t2DM, the significant and equivalent weight loss achieved in both groups in response to the two prescribed caloric-restricted diets may have masked any differences between the two dietary patterns that would be attributed to the individual macronutrient composition. Next, the short-term effects of low-GI or low-GL diets, seen in various studies, may not be seen in longer-term studies such as ours, because there is evidence that some adaptation to diets that vary in GI or GL takes place over time (44). Finally, the acute effects of a low-GI or low-GL diet on lowering postprandial glycemia may be beneficial per se and need to be distinguished from its chronic effects, measured here (32).

High circulating CRP levels have been shown to be a risk factor of t2DM independent of conventional risk factors (6). Therefore, a decline in CRP is desirable when considering specific dietary interventions. Our finding of greater declines in CRP concentration in participants randomized to a low-GL diet is consistent with observational studies (16) and some, but not all, previous clinical trials (37, 42, 45, 46). In a 3-month feeding study, a decline in CRP with a low-GL diet was seen despite equivalent weight loss (42). In a 6-month caloric-restricted ad libitum study of extremely low-carbohydrate vs. low-fat diets, CRP declined modestly in both groups, but the decline was greater in the low-carbohydrate diet only among participants with a high baseline CRP (46). In contrast, a study of high-protein vs. high-carbohydrate diet and a study of sucrose vs. artificial sweeteners found equivalent changes in CRP (37, 45). Although our finding of greater CRP declines with lower-GL diets may have important health implications, it warrants further investigation because CRP measurements have substantial variability and the population size in this investigation was relatively small.

Our study has several strengths, including long duration, high retention rate, provided foods, use of the doubly labeled water method to estimate total energy requirement for calculations of individual energy-restricted prescriptions during the study, and measurements of glucose tolerance and glucose-insulin dynamics in several ways, both at basal state (fasting) and during dynamic testing with oral and intravenous glucose challenge. Despite these strengths, potential limitations also warrant consideration in relation to interpreting our results. Our study provided hypocaloric diets, and, therefore, the effect of the dietary GL independent of caloric restriction cannot be assessed. Although the two diets differ in carbohydrate quality and quantity, they also differ in protein and fat content. The latter two macronutrients also contribute to glucose and insulin response, so the findings should be attributed to macronutrient balance rather than carbohydrate content per se. Finally, our results may not be widely generalizable in individuals with glucose intolerance at baseline or others who are at higher risk of diabetes because of advanced insulin resistance or defective β cell function at baseline (47).

In conclusion, healthy overweight individuals with normal fasting glucose responded to calorie-restricted diets of varying GL with equivalent changes in glucose-insulin dynamics, which highlights the importance of absolute weight loss over the macronutrient composition of the diet used to achieve weight loss. The finding of greater declines in CRP concentration after the low-GL diet warrants further investigation. Long-term studies aimed toward individuals at high risk for t2DM are needed before embracing the GL as a viable option for effective prevention of t2DM.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

This research was supported by NIH Grants K23-DK61506 (to A.G.P.), U01-AG20480 (to S.B.R.), and P30DK046200 to Boston Obesity Nutrition Research Center and the U.S. Department of Agriculture cooperative agreement no. 58-1950-4-401 (to S.B.R. and A.S.G.). We thank the study participants and staff of the Metabolic Research Unit at the Jean Mayer USDA Human Nutrition Research Center on Aging at Tufts University and Gayle Petty and the staff of the Nutrition Evaluation Laboratory.

Footnotes
  • 1

    Nonstandard abbreviations: t2DM, type 2 diabetes mellitus; GL, glycemic load; GI, glycemic index; CRP, C-reactive protein; HG, high-GL diet; LG, low-GL diet; HOMAIR, homeostasis model assessment of insulin resistance; OGTT, oral glucose tolerance test; INS30, insulin level at 30 minutes after glucose load; AUC, area under the curve; FSIVGTT, frequently sampled intravenous glucose tolerance test; Si, insulin sensitivity index; AIRg, acute insulin response to glucose; DI, disposition index; CV, coefficient of variation; SEM, standard error of the mean.

  • The costs of publication of this article were defrayed, in part, by the payment of page charges. This article must, therefore, be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  • 1
    Mokdad, A. H., Ford, E. S., Bowman, B. A., et al (2003) Prevalence of obesity, diabetes, and obesity-related health risk factors, 2001. JAMA 289: 7679.
  • 2
    Stumvoll, M., Goldstein, B. J., van Haeften, TW (2005) Type 2 diabetes: principles of pathogenesis and therapy. Lancet 365: 13331346.
  • 3
    Pittas, A. G., Joseph, N. A., Greenberg, AS (2004) Adipocytokines and insulin resistance. J Clin Endocrinol Metab. 89: 447452.
  • 4
    Knowler, W. C., Barrett-Connor, E., Fowler, S. E., et al (2002) Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med. 346: 393403.
  • 5
    Uusitupa, M., Lindi, V., Louheranta, A., Salopuro, T., Lindstrom, J., Tuomilehto, J. (2003) Long- term improvement in insulin sensitivity by changing lifestyles of people with impaired glucose tolerance: 4-year results from the Finnish Diabetes Prevention Study. Diabetes 52: 25322538.
  • 6
    Hu, F. B., Meigs, J. B., Li, T. Y., Rifai, N., Manson, JE (2004) Inflammatory markers and risk of developing type 2 diabetes in women. Diabetes 53: 693700.
  • 7
    Hu, F. B., Manson, J. E., Stampfer, M. J., et al (2001) Diet, lifestyle, and the risk of type 2 diabetes mellitus in women. N Engl J Med. 345: 790797.
  • 8
    Ludwig, DS (2002) The glycemic index: physiological mechanisms relating to obesity, diabetes, and cardiovascular disease. JAMA 287: 24142423.
  • 9
    Hodge, A. M., English, D. R., O'Dea, K., Giles, GG (2004) Glycemic index and dietary fiber and the risk of type 2 diabetes. Diabetes Care 27: 27012706.
  • 10
    Schulze, M. B., Liu, S., Rimm, E. B., Manson, J. E., Willett, W. C., Hu, FB (2004) Glycemic index, glycemic load, and dietary fiber intake and incidence of type 2 diabetes in younger and middle-aged women. Am J Clin Nutr. 80: 348356.
  • 11
    Salmeron, J., Ascherio, A., Rimm, E. B., et al (1997) Dietary fiber, glycemic load, and risk of NIDDM in men. Diabetes Care 20: 545550.
  • 12
    Salmeron, J., Manson, J. E., Stampfer, M. J., Colditz, G. A., Wing, A. L., Willett, WC (1997) Dietary fiber, glycemic load, and risk of non-insulin-dependent diabetes mellitus in women. JAMA 277: 472477.
  • 13
    Meyer, K. A., Kushi, L. H., Jacobs, D. R., Jr, Slavin, J., Sellers, T. A., Folsom, A. R. (2000) Carbohydrates, dietary fiber, and incident type 2 diabetes in older women. Am J Clin Nutr. 71: 921930.
  • 14
    Stevens, J., Ahn, K., Juhaeri, Houston, D., Steffan, L.D., Couper, (2002) Dietary fiber intake and glycemic index and incidence of diabetes in African-American and white adults: the ARIC study. Diabetes Care 25: 17151721.
  • 15
    Roberts, S. B., Pittas, AG (2003) The role of glycemic index in type 2 diabetes. Nutr Clin Care 6: 7378.
  • 16
    Liu, S., Manson, J. E., Buring, J. E., Stampfer, M. J., Willett, W. C., Ridker, PM (2002) Relation between a diet with a high glycemic load and plasma concentrations of high-sensitivity C-reactive protein in middle-aged women. Am J Clin Nutr. 75: 492498.
  • 17
    Sheard, N. F., Clark, N. G., Brand-Miller, J. C., et al (2004) Dietary carbohydrate (amount and type) in the prevention and management of diabetes: a statement by the American Diabetes Association. Diabetes Care 27: 22662271.
  • 18
    Schoeller, DA (1999) Recent advances from application of doubly labeled water to measurement of human energy expenditure. J Nutr. 129: 17651768.
  • 19
    Foster-Powell, K., Holt, S. H., Brand-Miller, JC (2002) International table of glycemic index and glycemic load values: 2002. Am J Clin Nutr. 76: 556.
  • 20
    Standing Committee on the Scientific Evaluation of Dietary Reference Intakes Food and Nutrition Board, Institute of Medicine (2002) Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids (Macronutrients) National Academy Press Washington, DC.
  • 21
    Schakel, S. F., Sievert, Y. A., Buzzard, IM (1988) Sources of data for developing and maintaining a nutrient database. J Am Diet Assoc. 88: 12681271.
  • 22
    Matthews, D. R., Hosker, J. P., Rudenski, A. S., Naylor, B. A., Treacher, D. F., Turner, RC (1985) Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 28: 412419.
  • 23
    Bonora, E., Targher, G., Alberiche, M., et al (2000) Homeostasis model assessment closely mirrors the glucose clamp technique in the assessment of insulin sensitivity: studies in subjects with various degrees of glucose tolerance and insulin sensitivity. Diabetes Care 23: 5763.
  • 24
    Hanson, R. L., Pratley, R. E., Bogardus, C., et al (2000) Evaluation of simple indices of insulin sensitivity and insulin secretion for use in epidemiologic studies. Am J Epidemiol. 151: 190198.
  • 25
    Yeni-Komshian, H., Carantoni, M., Abbasi, F., Reaven, GM (2000) Relationship between several surrogate estimates of insulin resistance and quantification of insulin-mediated glucose disposal in 490 healthy nondiabetic volunteers. Diabetes Care 23: 171175.
  • 26
    Bergman, R. N., Prager, R., Volund, A., Olefsky, JM (1987) Equivalence of the insulin sensitivity index in man derived by the minimal model method and the euglycemic glucose clamp. J Clin Invest. 79: 790800.
  • 27
    Kahn, S. E., Prigeon, R. L., McCulloch, D. K., et al (1993) Quantification of the relationship between insulin sensitivity and beta-cell function in human subjects: evidence for a hyperbolic function. Diabetes 42: 16631672.
  • 28
    Ridker, P. M., Hennekens, C. H., Buring, J. E., Rifai, N. (2000) C-reactive protein and other markers of inflammation in the prediction of cardiovascular disease in women. N Engl J Med. 342: 836843.
  • 29
    Slabber, M., Barnard, H. C., Kuyl, J. M., Dannhauser, A., Schall, R. (1994) Effects of a low-insulin-response, energy-restricted diet on weight loss and plasma insulin concentrations in hyperinsulinemic obese females. Am J Clin Nutr. 60: 4853.
  • 30
    Meckling, K. A., O'Sullivan, C., Saari, D. (2004) Comparison of a low-fat diet to a low-carbohydrate diet on weight loss, body composition, and risk factors for diabetes and cardiovascular disease in free-living, overweight men and women. J Clin Endocrinol Metab. 89: 27172723.
  • 31
    Bouche, C., Rizkalla, S. W., Luo, J., et al (2002) Five-week, low-glycemic index diet decreases total fat mass and improves plasma lipid profile in moderately overweight nondiabetic men. Diabetes Care 25: 822828.
  • 32
    Brynes, A. E., Edwards, Mark C., Ghatei, M. A., et al (2003) A randomised four-intervention crossover study investigating the effect of carbohydrates on daytime profiles of insulin, glucose, non-esterified fatty acids and triacylglycerols in middle-aged men. Br J Nutr. 89: 207218.
  • 33
    Frost, G., Leeds, A., Trew, G., Margara, R., Dornhorst, A. (1998) Insulin sensitivity in women at risk of coronary heart disease and the effect of a low glycemic diet. Metabolism 47: 12451251.
  • 34
    Heilbronn, L. K., Noakes, M., Clifton, PM (2001) Energy restriction and weight loss on very-low-fat diets reduce C-reactive protein concentrations in obese, healthy women. Arterioscler Thromb Vasc Biol. 21: 968970.
  • 35
    Stern, L., Iqbal, N., Seshadri, P., et al (2004) The effects of low-carbohydrate versus conventional weight loss diets in severely obese adults: one-year follow-up of a randomized trial. Ann Intern Med. 140: 778785.
  • 36
    Meckling, K. A., Gauthier, M., Grubb, R., Sanford, J. (2002) Effects of a hypocaloric, low- carbohydrate diet on weight loss, blood lipids, blood pressure, glucose tolerance, and body composition in free-living overweight women. Can J Physiol Pharmacol. 80: 10951105.
  • 37
    Noakes, M., Keogh, J. B., Foster, P. R., Clifton, PM (2005) Effect of an energy-restricted, high-protein, low-fat diet relative to a conventional high-carbohydrate, low-fat diet on weight loss, body composition, nutritional status, and markers of cardiovascular health in obese women. Am J Clin Nutr. 81: 12981306.
  • 38
    Sloth, B., Krog-Mikkelsen, I., Flint, A., et al (2004) No difference in body weight decrease between a low-glycemic-index and a high-glycemic-index diet but reduced LDL cholesterol after 10-wk ad libitum intake of the low-glycemic-index diet. Am J Clin Nutr. 80: 337347.
  • 39
    Ebbeling, C. B., Leidig, M. M., Sinclair, K. B., Seger-Shippee, L. G., Feldman, H. A., Ludwig, DS (2005) Effects of an ad libitum low-glycemic load diet on cardiovascular disease risk factors in obese young adults. Am J Clin Nutr. 81: 976982.
  • 40
    Wolever, T. M., Mehling, C. (2002) High-carbohydrate-low-glycaemic index dietary advice improves glucose disposition index in subjects with impaired glucose tolerance. Br J Nutr. 87: 477487.
  • 41
    Ebbeling, C. B., Leidig, M. M., Sinclair, K. B., Hangen, J. P., Ludwig, DS (2003) A reduced-glycemic load diet in the treatment of adolescent obesity. Arch Pediatr Adolesc Med. 157: 773779.
  • 42
    Pereira, M. A., Swain, J., Goldfine, A. B., Rifai, N., Ludwig, DS (2004) Effects of a low-glycemic load diet on resting energy expenditure and heart disease risk factors during weight loss. JAMA 292: 24822490.
  • 43
    Weyer, C., Bogardus, C., Mott, D. M., Pratley, RE (1999) The natural history of insulin secretory dysfunction and insulin resistance in the pathogenesis of type 2 diabetes mellitus. J Clin Invest. 104: 787794.
  • 44
    Kiens, B., Richter, EA (1996) Types of carbohydrate in an ordinary diet affect insulin action and muscle substrates in humans. Am J Clin Nutr. 63: 4753.
  • 45
    Sorensen, L. B., Raben, A., Stender, S., Astrup, A. (2005) Effect of sucrose on inflammatory markers in overweight humans. Am J Clin Nutr. 82: 421427.
  • 46
    Seshadri, P., Iqbal, N., Stern, L., et al (2004) A randomized study comparing the effects of a low-carbohydrate diet and a conventional diet on lipoprotein subfractions and C-reactive protein levels in patients with severe obesity. Am J Med. 117: 398405.
  • 47
    Pittas, A. G., Das, S. K., Hajduk, C. L., et al (2005) A low-glycemic load diet facilitates greater weight loss in overweight adults with high insulin secretion but not in overweight adults with low insulin secretion in the CALERIE Trial. Diabetes Care 28: 29392941.