To investigate the relation between obesity, fat distribution, and weight gain through adulthood and the risk of non-insulin-dependent diabetes melli-tus (NIDDM).

We analyzed data from a cohort of 51,529 U.S. male health professionals, 40-75 years of age in 1986, who completed biennial questionnaires sent out in 1986, 1988, 1990, and 1992. During 5 years of follow-up (1987-1992), 272 cases of NIDDM were diagnosed among men without a history of diabetes, heart disease, and cancer in 1986 and who provided complete health information. Relative risks (RRs) associated with different anthropometrie measures were calculated controlling for age, and multivariate RRs were calculated controlling for smoking, family history of diabetes, and age.

We found a strong positive association between overall obesity as measured by body mass index (BMI) and risk of diabetes. Men with a BMI of ≥35 kg/m² had a multivariate RR of 42.1 (95% confidence interval [CI] 22.0-80.6) compared with men with a BMI <23.0 kg/m². BMI at age 21 and absolute weight gain throughout adulthood were also significant independent risk factors for diabetes. Fat distribution, measured by waist-to-hip ratio (WHR), was a good predictor of diabetes only among the top 5%, while waist circumference was positively associated with the risk of diabetes among the top 20% of the cohort.

These data suggest that waist circumference may be a better indicator than WHR of the relationship between abdominal adiposity and risk of diabetes. Although early obesity, absolute weight gain throughout adulthood, and waist circumference were good predictors of diabetes, attained BMI was the dominant risk factor for NIDDM; even men of average relative weight had significantly elevated RRs.