Marshall J A, Weiss N S, Hamman R F
Department of Preventive Medicine and Biometrics, University of Colorado Health Sciences Center, Denver 80262.
Ann Epidemiol. 1993 Jan;3(1):18-26. doi: 10.1016/1047-2797(93)90005-o.
To investigate the hypothesis that a low intake of dietary fiber could increase the risk of developing non-insulin-dependent diabetes mellitus (NIDDM), we ascertained prior dietary intake of 242 persons with known diabetes and 460 persons without a prior diagnosis of diabetes among 20- to 74-year-old residents of two counties in southern Colorado from 1984 to 1986. When persons with diabetes were compared to nondiabetic controls, a higher reported fiber intake prior to diagnosis was found among persons with diabetes. A decrease in fiber of 10 g/d was associated with a decrease in risk of NIDDM of 0.75 (95% confidence interval: 0.59 to 0.96), rather than an increase as hypothesized. However, when the diabetic group was limited to those with diabetes for less than 5 years, this association was no longer present. Two further analyses were carried out on 1317 persons without a prior diagnosis of diabetes seen between 1984 and 1988. Among these persons, current fiber intake was inversely associated with fasting plasma insulin concentration. However, fiber explained less than 1% of the variation in fasting insulin levels. When persons with previously undiagnosed NIDDM were compared to normal controls, the odds ratio relating a decrease in fiber consumption of 10 g/d to NIDDM was 1.21 (95% confidence interval: 0.70 to 2.10) adjusting for calorie and carbohydrate intake. All analyses were adjusted for age, sex, ethnicity, and body mass index. The inconsistent findings reported here do not support the hypothesis that increasing dietary fiber intake could reduce the future occurrence of NIDDM.
