Weight history, glucose intolerance, and insulin levels in middle-aged Swedish men

The association between weight history and glucose intolerance was examined in a cross-sectional study consisting of 3,128 Swedish men aged 35-56 years, 52 percent of whom had a family background of diabetes mellitus. Oral glucose tolerance testing detected 55 cases of type 2 (non-insulin-dependent) diabetes and 172 cases of impaired glucose tolerance. Among men with no family history of diabetes, the estimated odds ratios for impaired glucose tolerance associated with short (<5 years) and long (greater than or equal to 10 years) durations of obesity (body mass index (weight (kg)/height(2) (m(2)) greater than or equal to 25.0) were 1.3 (95% confidence interval (Cl) 0.2-7.7) and 11.8 (95% CI 3.3-41.9), respectively. Among men with a family history of diabetes, the odds ratios were 2.0 (95% CI 0.8-4.7) and 4.0 (95% CI 1.8-9.1), respectively. Corresponding estimates of the odds of type 2 diabetes, adjusted for family history of diabetes, were 1.9 (95% CI 0.5-7.1) and 7.3 (95% CI 2.2-23.7), respectively. The odds of high (greater than or equal to 30.0 mU/liter) fasting insulin levels in subjects with impaired glucose tolerance were 6.9 (95% CI 0.6-74.2) and 21.0 (95% CI 2.1-206.4) for short and long durations of obesity, respectively. Corresponding estimated odds of low 2-hour insulin response (less than or equal to 71.9 mU/liter) were 0.7 (95% CI 0.2-2.9) and 3.3 (95% CI 1.2-8.9). Homeostasis model assessment of insulin resistance yielded an odds ratio of 6.7 (95% CI 0.6-73.4) for a short duration of obesity and 20.0 (95% CI 2.0-200.6) for a long duration. Examination of beta-cell function with homeostasis model assessment resulted in odds ratios of 0.2 (95% CI 0.0-1.6) and 2.0 (95% CI 0.7-5.4) for short and long durations of obesity, respectively. These data indicate that obesity decreases glucose tolerance by way of progressively increased insulin resistance and, in the case of prolonged duration, by decreased insulin secretion as well.