PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Diabetes. Author manuscript; available in PMC 2008 September 3.
Published in final edited form as:
Published online 2007 March 15. doi:  10.2337/db07-0009
PMCID: PMC2528279
NIHMSID: NIHMS58685

Body Size and Shape Changes and the Risk of Diabetes in the Diabetes Prevention Program

Abstract

The researchers conducted this study to test the hypothesis that risk of type 2 diabetes is less following reductions in body size and central adiposity. The Diabetes Prevention Program (DPP) recruited and randomized individuals with impaired glucose tolerance to treatment with placebo, metformin, or lifestyle modification. Height, weight, waist circumference, and subcutaneous and visceral fat at L2-L3 and L4-L5 by computed tomography were measured at baseline and at 1 year. Cox proportional hazards models assessed by sex the effect of change in these variables over the 1st year of intervention upon development of diabetes over subsequent follow-up in a subset of 758 participants. Lifestyle reduced visceral fat at L2-L3 (men −24.3%, women −18.2%) and at L4-L5 (men −22.4%, women −17.8%), subcutaneous fat at L2-L3 (men −15.7%, women −11.4%) and at L4-L5 (men −16.7%, women −11.9%), weight (men −8.2%, women −7.8%), BMI (men −8.2%, women −7.8%), and waist circumference (men −7.5%, women −6.1%). Metformin reduced weight (−2.9%) and BMI (−2.9%) in men and subcutaneous fat (−3.6% at L2-L3 and −4.7% at L4-L5), weight (−3.3%), BMI (−3.3%), and waist circumference (−2.8%) in women. Decreased diabetes risk by lifestyle intervention was associated with reductions of body weight, BMI, and central body fat distribution after adjustment for age and self-reported ethnicity. Reduced diabetes risk with lifestyle intervention may have been through effects upon both overall body fat and central body fat but with metformin appeared to be independent of body fat.

Type 2 diabetes has rapidly become a global health problem (1). Moreover, diabetes is a major cause of morbidity and mortality from coronary artery, cerebrovascular, retinal, neurological, and renal complications (2,3). One solution may be to prevent or delay diabetes in at-risk populations (46). The Diabetes Prevention Program (DPP), a multicenter randomized clinical trial, showed that this could be accomplished with lifestyle intervention or metformin among individuals with impaired glucose tolerance. At the end of 3.2 years, diabetes incidence rates were 10.8, 7.7, and 5.0 per 100 person-years in the placebo, metformin, and lifestyle intervention groups, respectively (this average duration of follow-up is 4 months longer than that reported in the primary outcome report [6]). Treatment effects were consistent by sex and race/ethnicity.

Individuals who have type 2 diabetes are usually over-weight or obese. Moreover, a central pattern of body fat distribution, particularly an increased amount of intra-abdominal or visceral fat, is an independent risk factor for type 2 diabetes (717). The effect of lifestyle modification and metformin on body size and central obesity was assessed in the DPP by both anthropometric measurements and computed tomography (CT).

In this article, we tested the hypothesis that reductions in body size and central adiposity over the 1st year are associated with a decreased risk in developing type 2 diabetes over subsequent follow-up. This was done in a subset of 758 participants who had measurements of body fat and body fat distribution at both baseline and 1 year and who remained nondiabetic at 1 year.

RESEARCH DESIGN AND METHODS

Subject eligibility and recruitment

Because the design and methods of the DPP (18), the baseline characteristics of the cohort (19), and the primary outcomes (6) have been previously described in detail, they are briefly summarized here. Participants were recruited across 27 clinical centers located throughout the U.S. Key eligibility requirements were age ≥25 years, BMI ≥24 kg/m2 (≥22 kg/m2 for Asian Americans), a fasting plasma glucose level of 5.3–7.0 mmol/l (95–125 mg/dl) (≤7.0 mmol/l [≤125 mg/dl] in the American Indian centers), and a 2-h plasma glucose level of 7.8–11.1 mmol/l (140–199 mg/dl) following 75 g glucose given orally. Individuals were excluded if they had diabetes, any condition likely to limit life span or increase risk from the interventions, any condition that would likely affect the ability to conduct the trial as designed, or if they were taking any medication or had any medical condition likely to confound the assessment of diabetes status.

Intervention groups

Participants were randomized to placebo, metformin (850 mg twice daily), or an intensive lifestyle modification program. At baseline and annually thereafter, all participants received written information and an individual session with their case manager to address the importance of a healthy lifestyle for the prevention of type 2 diabetes.

The intensive lifestyle intervention goals were to 1) achieve and maintain a weight reduction of at least 7% of initial body weight through healthy eating and physical activity and 2) achieve and maintain a level of physical activity of at least 150 min/week (equivalent to ~700 kcal/week) through moderate-intensity activity (such as walking or bicycling). Key elements included training in diet, exercise, and behavior modification skills; frequent contact with the interventionist (at least monthly); and cultural sensitivity. The weight loss goal was attempted initially through a reduction in dietary fat intake to <25% of calories, and if this was not successful, caloric restriction was added. Details are available at http://www.bsc.gwu.edu/dpp/index.htmlvdoc (20). Because CT was not initiated at the very start of the DPP, the subset of participants who are in this report was followed for an average 2.5 years, compared with 3.2 years for all DPP participants.

Anthropometry

Each measurement was recorded twice during the baseline screening visit and averaged. A third measurement was taken if the variability was greater than a predefined value. Staff members performing these measurements were certified annually. Height was measured on a stadiometer to the nearest 0.5 cm, body weight was measured on a calibrated balance scale to the nearest 0.1 kg, and BMI (kg/m2) was computed from body weight and height. Waist was defined as the midpoint between the highest point of the iliac crest and the lowest point of the costal margin at the midaxillary line, and waist circumference was measured with a cloth tape. Hip circumference was measured at the level of the greater femoral trochanters. The waist-to-hip ratio (WHR) was the waist circumference divided by the hip circumference.

Measurements of adipose tissue by CT

Baseline measurement of adipose tissue by CT was done in 1,106 participants from 18 of the 27 sites. Of these, 159 had a poor-quality scan at baseline, 113 had a poor-quality scan at follow-up, and 59 did not return for a second scan, resulting in 777 participants who had CT scans at both baseline and follow-up of acceptable quality. Among the participants who had acceptable CT scans at baseline, those who did not have follow-up scans were larger in terms of hip circumference, BMI, and waist circumference (P < 0.001) than those who did receive follow-up scans. The instruments used included GE High Speed Advantage (five centers), Picker PQ 5000 (five centers), Siemens and Siemens Somatom Plus (two centers), GE 9800 (three centers), and GE Highlite (two centers). The clinic variability of measurement for both subcutaneous and visceral adipose tissue at L2-L3 and L4-L5 for men and women was considerably smaller than the residual variance, indicating that different clinic CT methods did not have an important influence upon the scan measurements.

An antero-posterior scout and two 10-mm thick axial images at the L2-L3 and the L4-L5 disc spaces were submitted to a central reading facility at the University of Colorado Health Sciences Center by tape or optical disc. Each subject was used as his/her own control to create a bimodal histogram depicting the distribution of Hounsfield units in the image, resulting in separate peaks for muscle and fat. The area under the fat peak (number of fat pixels) multiplied by the area of one pixel equals the fat area for that CT slice. This peak also defines the range of fat density for a patient. Four such measurements were obtained (two at L2-L3 and two at L4-L5). Averages and SDs were reported for each level. A line drawn manually through the body wall divided the subcutaneous and visceral compartments, and subcutaneous fat area was calculated by subtracting visceral fat area from total fat area. Those analyzing the CT scans were blinded as to treatment assignment.

Statistical analysis

Of 777 participants who had CT at baseline and 1 year, 19 diagnosed with diabetes at 6 months or 1 year were excluded from the analyses reported here, leaving a final sample size of 758. Descriptive statistics of baseline and year 1 body composition variables were computed by sex, race, and treatment groups. Differences between treatment groups in baseline and year 1 body fat measurements were tested for each sex using general linear models adjusted for self-reported race/ethnicity and age. The Holm’s procedure (21) was used to control type 1 error rate between treatment comparisons. A P value of ≤0.05 was considered statistically significant. Cox regression models were used to assess effect of change in body fat variables (visceral fat area at L2-L3 and L4-L5, subcutaneous fat area at L2-L3 and L4-L5, weight, waist, BMI, and WHR) over the 1st year (year 1 measurement minus baseline measurement) on the subsequent risk of developing diabetes. Hazard ratios are reported per 1 SD of the change in body fat measurements. Models were run separately for each treatment group adjusted for age and self-reported ethnicity, and a test of heterogeneity was used to see if the effect of change in a body fat variable differed across treatment groups. Madalla’s likelihood ratio R2LR (22) was computed to determine the approximate variation explained by addition of a body fat measurement to a Cox model with age and race/ethnicity. The SAS analysis system version 8.2 was used for all analyses (SAS Institute, Cary, NC).

RESULTS

None of the baseline measurements of body fat differed significantly by treatment group (Table 1). On average, participants were obese, with a mean BMI of 32.1 ± 5.2 kg/m2 in men and 33.0 ± 5.5 kg/m2 in women. The waist circumference was large. Waist girth and visceral fat area were larger in men than women, whereas the subcutaneous fat area at both L2-L3 and L4-L5 was larger in women than men. There was more visceral fat, and sex differences appeared to be greater at the L2-L3 than the L4-L5 level. All body fat measurements differed significantly by sex at a P < 0.01, except BMI (P = 0.03).

TABLE 1
Baseline characteristics by treatment group in men and women

Every anthropometric measurement decreased from baseline in both sexes in both the metformin and lifestyle intervention arms (Table 2 and Fig. 1). Reductions in visceral fat observed with lifestyle intervention in both men (−24.3% at L2-L3 and −22.4% at L4-L5; Fig. 1) and women (−18.2% at L2-L3 and −17.8 at L4-L5; Fig. 1) were especially dramatic. Much smaller reductions in subcutaneous fat, waist, BMI, and weight were found after 1 year of lifestyle intervention.

FIG. 1
Percent change (mean [95% confidence limits]) in body fat measurement year minus baseline by treatment group in men and women. ●, placebo; ■, metformin; ▲, lifestyle intervention.
TABLE 2
Change in body fat measurement (year 1 minus baseline) by treatment group in men and women

Cox regression models predicting diabetes were run separately for each treatment group (Table 3). In men, reductions in all of the fat variables were significant predictors of decreased diabetes risk in the lifestyle intervention arm, whereas in women, weight, BMI, and waist reduction were significant predictors. Visceral fat reduction was nearly significant as a predictor in women. In the placebo group, only subcutaneous fat reduction predicted decreased diabetes risk and only in men. In both men and women, none of the fat variables predicted diabetes in the metformin group.

TABLE 3
Cox regression models estimating hazard ratios for change in body fat measurement (year 1 minus baseline) in predicting diabetes by treatment group and sex*

DISCUSSION

A central pattern of body fat distribution has been associated with both insulin resistance and risk for diabetes independent of BMI (817). A central pattern of body fat distribution may be identified by waist or WHR or by skinfold thickness on the trunk compared with the extremities. In the DPP, weight, BMI, waist girth, hip girth, and WHR each significantly predicted the development of diabetes in the placebo and lifestyle intervention groups in both sexes (23). Skinfolds were less predictive. In the metformin group, none of these measurements predicted the future risk of diabetes, and the reduction in diabetes incidence was much less than with lifestyle modification. We have now shown that reduction of diabetes risk with lifestyle modification was closely associated with reduction of body size and central adiposity. Reduction of diabetes risk with metformin, however, appeared independent of changes in body size or central adiposity.

Increased subcutaneous abdominal fat, increased visceral fat, or a combination of the two may underlie central adiposity. Techniques such as CT or magnetic resonance imaging are necessary to distinguish between subcutaneous and visceral abdominal fat (24). Greater visceral adiposity, and not subcutaneous fat, was shown by CT to precede the development of diabetes in a prospective study among Japanese Americans (15,17). But there appear to be racial differences in the relationship between the amount of visceral fat and diabetes as well as in the amount of visceral fat at similar levels of overall obesity. African-American men, despite higher risk for type 2 diabetes, were shown to have less visceral fat than white men (25). Pima Indians also have less visceral fat than whites at similar levels of obesity (26). Although others have argued that subcutaneous fat has as important a role in insulin resistance as does visceral fat (27), most studies have shown the visceral adipose depot to be more important. Although comprising only a small portion of total body fat stores, visceral fat shows rapid turnover, and, during weight loss attributable to diet and physical activity, there is a proportionately greater reduction of visceral fat than of total body fat stores (2730).

We found that in both men and women, lifestyle intervention dramatically reduced visceral fat, whereas metformin had no effect. There were, however, some sex differences in the response of other body fat variables. Both lifestyle intervention and metformin reduced weight and BMI in both men and women. In women, both lifestyle intervention and metformin also reduced waist circumference and subcutaneous fat, whereas in men lifestyle intervention but not metformin reduced both of these. However, although lifestyle changes and metformin both had an effect, the effect of lifestyle changes was far greater in both men and women. In men, changes in subcutaneous fat predicted time to develop diabetes in the placebo group, whereas in the lifestyle modification group decreases in all the fat variables predicted diabetes independent of age and self-reported race/ethnicity. It is not surprising that changes in subcutaneous fat predicted diabetes in the placebo group, because this was the measure that changed the most, whereas both visceral fat and total body fat remained relatively constant. For this group, measures of body fatness at baseline were strong predictors of diabetes.

These results suggest that lifestyle intervention decreases diabetes risk in men through reductions in overall, as well as central, body fat. In women, no fat measurement predicted diabetes in the placebo group and only changes in weight, BMI, and waist circumference were predictive in the lifestyle intervention group. Change in visceral fat, which is highly correlated with change in waist circumference, was nearly significant. In contrast, whereas diabetes risk reduction appears to be mediated through changes in body fat and body fat distribution in both sexes, the reduction of diabetes risk by metformin appears to be through a mechanism independent of such changes.

The Da Qing Study (4) and the Finnish Diabetes Prevention Study (5), which also demonstrated prevention of type 2 diabetes through diet and exercise interventions, examined a limited number of adipose variables that did not include CT. Both studies showed a significant reduction in body weight, whereas in the Finnish Study there was also significant reduction in waist circumference.

In the DPP, improvements at 1 year in insulin sensitivity and preservation of insulin secretion were associated with lower hazard rates for diabetes (31). The lifestyle intervention group had the greatest improvement, there was no significant change in the placebo group, and metformin was intermediate.

In a small lifestyle modification study among Japanese Americans with impaired glucose tolerance, dietary restriction of saturated fat plus participation in physical activity for 24 months resulted in significant weight loss, reduced percent body fat, decreased visceral and subcutaneous abdominal fat by CT, and increased insulin sensitivity (32,33). Improvement of insulin sensitivity was the result of changes in visceral rather than subcutaneous fat when examined with multiple regression analyses.

A strength of our study is the large number of participants in whom we were able to obtain baseline and follow-up measurements of body fat and body fat distribution. A limitation is that we assessed body fat change over only the 1st year of intervention, whereas diabetes risk was assessed over the subsequent mean 1.5 years of follow-up. The relationship between body fat change and change in diabetes risk may have been greatly improved if we had assessed the former over 2.5 years. Another limitation is that because not all DPP participants were in the CT substudy, there were some differences in results between this study and other DPP publications. For example, weight did not predict diabetes among the metformin participants in this substudy. Such differences may be due to the smaller number of participants in the substudy but are more likely due to differences in baseline characteristics such as lower body weight, BMI, and hip and waist circumferences in the substudy participants than in the entire DPP cohort (P < 0.001). Although we focused specifically upon body fat and its distributional change, differential lean body mass change by treatment arm and gender is possible but was not examined. Because muscle is a major glucose clearance site, change in muscle mass could influence diabetes risk. Although the physical activity program was predominantly aerobic, strength training activities were not prohibited.

In conclusion, lifestyle intervention was associated with dramatic reductions in visceral fat as well as smaller, but still significant, decreases in subcutaneous fat, body weight, BMI, and waist circumference in both men and women. Moreover, reduction in the risk of developing diabetes was associated with decreases in both body fat and central adiposity only in the lifestyle intervention group, suggesting that these changes in body fat were important in determining the benefit seen with the DPP lifestyle intervention.

ACKNOWLEDGMENTS

Supported by a cooperative agreement from the National Institute of Diabetes and Digestive and Kidney Diseases.

Glossary

CT
computed tomography
DPP
Diabetes Prevention Program
WHR
waist-to-hip ratio

REFERENCES

1. Wild S, Green A, Sicree R, King H. Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care. 2004;27:1047–1053. [PubMed]
2. Herman WH, Eastman RC. The effects of treatment on the direct costs of diabetes. Diabetes Care. 1998;21 Suppl. 3:C19–C24. [PubMed]
3. Amos AF, McCarty DJ, Zimmet P. The rising global burden of diabetes and its complications: estimates and projections to the year 2010. Diabet Med. 1997;14:81–85. [PubMed]
4. Pan XR, Li GW, Hu YH, Wang JX, Yang WY, An ZX, Hu ZX, Lin J, Xiao JZ, Cao HB. Effects of diet and exercise in preventing NIDDM in people with impaired glucose tolerance: the Da Zing IGT and Diabetes Study. Diabetes Care. 1997;20:537–544. [PubMed]
5. Tuomilehto J, Lindstrom J, Eriksson JG, Valle TT, Hamalainen H, Ilanne-Parikka P, Keinanen-Kiukaanniemi S, Laakso M, Louheranta A, Rastas M. Prevention of type 2 diabetes mellitus by changes in lifestyle among subjects with impaired glucose tolerance. N Engl J Med. 2001;344:1343–1350. [PubMed]
6. Diabetes Prevention Program Research Group. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med. 2002;346:393–403. [PMC free article] [PubMed]
7. The Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care. 1997;20:1183–1197. [PubMed]
8. Vague J. The degree of masculine differentiation of obesities determining predisposition to diabetes, gout and uric calculus disease. Am J Clin Nutr. 1956;4:20–34. [PubMed]
9. Feldman R, Sender AJ, Siegelaub AB. Difference in diabetic and nondiabetic fat distribution patterns by skinfold measurements. Diabetes. 1969;18:478–186. [PubMed]
10. Kissebah AH, Bydelingum N, Murray R, Evans DJ, Hartz AJ, Kalkhoff RK, Adams PW. Relation of body fat distribution to metabolic complications of obesity. J Clin Endocrinol Metab. 1982;54:254–260. [PubMed]
11. Ohlson LO, Larsson B, Svardsudd K, Welin L, Eriksson H, Wilhelmsen L, Bjorntorp P, Tibblin G. The influence of body fat distribution on the incidence of diabetes mellitus: 13.5 years of follow-up of the participants in the study of men born in 1913. Diabetes. 1985;34:1055–1058. [PubMed]
12. Sparrow D, Borkan GA, Gerzof SB, Wisniewiski C, Silbert CK. Relationship of fat distribution to glucose tolerance. Diabetes. 1986;35:411–415. [PubMed]
13. Fujioka S, Matsuzawa Y, Tokunaga K, Tarui S. Contribution of intra-abdominal fat accumulation to the impairment of glucose and lipid metabolism in human obesity. Metabolism. 1987;36:54–59. [PubMed]
14. Despres JP, Nadeaur A, Tremblay A, Ferland M, Moorjani S, Lupien PF, Theriault G, Pinault S, Bouchard C. Role of deep abdominal fat in the association between regional adipose tissue distribution and glucose tolerance in obese women. Diabetes. 1988;38:304–309. [PubMed]
15. Bergstrom RW, Newell-Morris LL, Leonetti DL, Shuman WP, Wahl PW, Fujimoto WY. Association of elevated fasting C-peptide level and increased intra-abdominal fat distribution with development of NIDDM in Japanese-American men. Diabetes. 1990;39:104–111. [PubMed]
16. Lemieux S, Despres JP. Metabolic complications of visceral obesity: contribution to the etiology of type 2 diabetes and implications for prevention and treatment. Diabete Metab. 1994;20:375–393. [PubMed]
17. Boyko EJ, Fujimoto WY, Leonetti DL, Newell-Morris L. Visceral adiposity and risk of type 2 diabetes. Diabetes Care. 2000;23:465–471. [PubMed]
18. The Diabetes Prevention Program. Design and methods for a clinical trial in the prevention of type 2 diabetes. Diabetes Care. 1999;22:623–634. [PMC free article] [PubMed]
19. The Diabetes Prevention Program. Baseline characteristics of the randomized cohort. Diabetes Care. 2000;23:1619–1629. [PMC free article] [PubMed]
20. Wing RR, Hamman RF, Bray GA, Delahanty L, Edelstein SL, Hill JO, Horton ES, Hoskin MA, Kriska A, Lachin J, Mayer-Davis EJ, Pi-Sunyer X, Regensteiner JG, Venditti B, Wylie-Rosett J. Diabetes Prevention Program Research Group. achieving weight and activity goals among Diabetes Prevention Program lifestyle participants. Obes Res. 2004;12:1426–1434. [PMC free article] [PubMed]
21. Holm S. A simple sequentially rejective Bonferroni test procedure. Scand J Stat. 1979;6:5–70.
22. Lachin JM. Biostatistical Methods: The Assessment of Relative Risks. New York: John Wiley and Son; 2000.
23. The Diabetes Prevention Program Research Group. Relationship of body size and shape to the development of diabetes in the Diabetes Prevention Program. Obesity. 2006;14:2107–2117. [PMC free article] [PubMed]
24. Snijder MB, van Dam RM, Visser M, Seidell JC. What aspects of body fat are particularly hazardous and how do we measures them? Internat J Epidemiol. 2006;35:83–92. [PubMed]
25. Hoffman DJ, Wang Z, Gallagher D, Heymsfield SB. Comparison of visceral adipose tissue mass in adult African Americans and whites. Obes Res. 2005;13:66–74. [PubMed]
26. Gautier JF, Milner MR, Elam E, Chen K, Ravusin E, Pratley RE. Visceral adipose tissue is not increased in Pima Indians compared with equally obese Caucasians and is not related to insulin action or secretion. Diabetologia. 1999;42:28–34. [PubMed]
27. Abate N, Garg A, Peshock RM, Stray-Gunderson J, Grundy SM. Relationships of generalized and regional adiposity to insulin sensitivity in men. J Clin Invest. 1995;96:88–98. [PMC free article] [PubMed]
28. Ross R, Hudson R. Sensitivity associated with the identification of visceral adipose tissue levels using waist circumference in men and women: effects of weight loss. Internat J Obes Relat Metab Disord. 1996;20:533–538. [PubMed]
29. Colman E, Katyzel LI, Rogus E, Coon P, Muller D, Goldberg AP. Weight loss reduces abdominal fat and improves insulin action in middle-aged and older men with impaired glucose tolerance. Metabolism. 1995;44:1502–1508. [PubMed]
30. Mourier A, Gaugier JF, De Kerviler E, Bigard AX, Villette JM, Garnier JP, Duvallet A, Guezennec CY, Cathelineau G. Mobilization of visceral adipose tissue related to the improvement in insulin sensitivity in response to physical training in NIDDM: effects of branched-chain amino acid supplements. Diabetes Care. 1997;20:385–391. [PubMed]
31. The Diabetes Prevention Program Research Group. Role of insulin secretion and sensitivity in the evolution of type 2 diabetes in the Diabetes Prevention Program: effects of lifestyle intervention and metformin. Diabetes. 2005;54:2404–2414. [PMC free article] [PubMed]
32. Liao D, Asberry PJ, Shofer JB, Callahan H, Matthys C, Boyko EJ, Leonetti D, Kahn SE, Austin M, Newell L, Schwartz RS, Fujimoto WY. Improvement of BMI, body composition, and body fat distribution with lifestyle modification in Japanese Americans with impaired glucose tolerance. Diabetes Care. 2002;25:1504–1510. [PubMed]
33. Carr DB, Utzschneider KM, Boyko EJ, Asberry PJ, Hull RL, Kodama K, Callahan HS, Matthys CC, Leonetti DL, Schwartz RS, Kahn SE, Fujimoto WY. A reduced-fat diet and aerobic exercise in Japanese Americans with impaired glucose tolerance decreases intra-abdominal fat and improves insulin sensitivity but not β-cell function. Diabetes. 2005;54:340–347. [PubMed]