Dysfunctional adiposity and the risk of prediabetes and type 2 diabetes in obese adults

Abstract

Context: The risk of type 2 diabetes mellitus is heterogeneous among obese individuals. Factors that discriminate prediabetes or diabetes risk within this population have not been well characterized. A dysfunctional adiposity phenotype, characterized by excess visceral fat and insulin resistance, may contribute to diabetes development in those with obesity.

Objective: To investigate associations between adiposity phenotypes and risk for incident prediabetes and diabetes in a multiethnic, population-based cohort of obese adults.

Design, setting, and participants: Among 732 obese participants (body mass index ≥30) aged 30 to 65 years without diabetes or cardiovascular disease enrolled between 2000 and 2002 in the Dallas Heart Study, we measured body composition by dual energy x-ray absorptiometry and magnetic resonance imaging (MRI); circulating adipokines and biomarkers of insulin resistance, dyslipidemia, and inflammation; and subclinical atherosclerosis and cardiac structure and function by computed tomography and MRI.

Main outcome measures: Incidence of diabetes through a median 7.0 years (interquartile range, 6.6-7.6) of follow-up. In a subgroup of 512 participants with normal fasting glucose values at baseline, incidence of the composite of prediabetes or diabetes was determined.

Results: Of the 732 participants (mean age, 43 years; 65% women; 71% nonwhite), 84 (11.5%) developed diabetes. In multivariable analysis, higher baseline visceral fat mass (odds ratio [OR] per 1 SD [1.4 kg], 2.4; 95% CI, 1.6-3.7), fructosamine level (OR per 1 SD [1.1 μmol/L], 2.0; 95% CI, 1.4-2.7), fasting glucose level (OR per 1 SD [1.1 μmol/L], 1.9; 95% CI, 1.4-2.6), family history of diabetes (OR, 2.3; 95% CI, 1.3-4.3), systolic blood pressure (OR per 10 mm Hg, 1.3; 95% CI, 1.1-1.5), and weight gain over follow-up (OR per 1 kg, 1.06; 95% CI, 1.02-1.10) were independently associated with diabetes, with no associations observed for body mass index, total body fat, or abdominal subcutaneous fat. Among the 512 participants with normal baseline glucose values, the composite outcome of prediabetes or diabetes occurred in 39.1% and was independently associated with baseline measurements of visceral fat mass; levels of fasting glucose, insulin, and fructosamine; older age; nonwhite race; family history of diabetes; and weight gain over follow-up (P < .05 for each) but not with measurements of general adiposity.

Conclusion: Excess visceral fat and insulin resistance, but not general adiposity, were independently associated with incident prediabetes and type 2 diabetes mellitus in obese adults.

Figure 1

Participant Selection and Follow-up CVD indicates cardiovascular disease; DHS, Dallas Heart Study.

Figure 2

Incidence of Type 2 Diabetes Among Obese Individuals Stratified by Sex-Specific Median Values for Visceral Fat Mass and by HOMA-IR and Fructosamine Levels The median cut points for visceral fat mass were 3.2 kg for men and 2.2 kg for women. A, For homeostasis model assessment of insulin resistance (HOMA-IR), the median cut point was 4 units for both men and women. B, For fructosamine, the median cut points were 204 μmol/L for men and 196 μmol/L for women.