Summary: Canadian Medical Association Journal 1996; 154: 207
An enrollment questionnaire in 1976 inquired about height, weight, medical history and health-related behaviours; biennial follow-up questionnaires to 1990 ascertained interim weight changes and diagnoses of major illnesses. Additional information was obtained from women who reported a diagnosis of diabetes. NIDDM was confirmed in 2204 participants. Incidence rates of NIDDM during the 14 years of follow-up were calculated for categories of body mass index (BMI) at baseline and for categories of weight change between age 18 and the year 1976. Women who reported major illnesses were excluded from subsequent follow-up. The relative risk of NIDDM for women in a given BMI category was calculated as the incidence rate of NIDDM among women in that category divided by the incidence rate among women with a BMI below 22.0.
An attained increase in BMI was associated with increased risk for NIDDM. Women with an average BMI (24.0 to 24.9) had an age- adjusted relative risk of 5.0 (95% confidence interval [CI] 3.6 to 6.6) compared with women with a BMI below 22.0; a BMI of 31.0 or more was associated with relative risks of 40.0 or greater. This association was not limited to any age strata; increased BMI was a strong risk factor even for women over 65.
Early adiposity was also found to increase risk. Women who had had a BMI above 22.0 at age 18 had a higher risk of NIDDM than women whose BMI was less than 22.0 at that age. The age-adjusted relative risk for women whose BMI was 33.0 to 34.9 at age 18 was 6.6 (95% CI 4.7 to 9.4).
Weight change between age 18 and the year 1976 was strongly associated with an increased risk for diabetes. Women with modest gains of 5.0 to 7.9 kg had a risk of 1.9 (95% CI 1.5 to 2.3) relative to those who had gained less than 5.0 kg, and women who gained 8.0 to 10.9 kg had a risk of 2.7 (95% CI 2.1 to 3.3). Subjects who had lost 5 to 10.9 kg between age 18 and the year 1976 had a significantly reduced risk for NIDDM (relative risk 0.54 [95% CI 0.4 to 0.8]). For each category of BMI at age 18 even modest weight gains were associated with an increase in risk.
These findings suggest that the risk of NIDDM can be minimized by achieving a lean build in early adulthood and avoiding weight gain throughout life. The researchers concluded that the USDA's recommendations may be misleading.