The role of dietary fiber in the etiology of noninsulin-dependent diabetes mellitus

Thesis (Ph. D.)--University of Washington, 1987Diet histories of cases of NIDDM and controls were compared to determine whether low dietary fiber intake was positively associated with the occurrence of NIDDM. The study group included prevalent, medically diagnosed diabetics between the ages of 20 and 74 years, residing in Alamosa and Conejos counties in Colorado and controls selected by geographically-based multi-stage sampling. Fiber intake prior to NIDDM diagnosis, or a comparable time in the past for controls, was obtained by food frequency questionnaire (FFQ). Current fiber intake was ascertained by FFQ and 24-hour diet recall. Three analyses were performed: (1) cross-sectional analysis of current fiber intake and plasma insulin concentrations among the controls; (2) comparison of prior fiber intake in previously known diabetics and nondiabetic controls; and (3) comparison of current fiber intake in previously undiagnosed diabetics and normal controls. Analyses were adjusted for age, sex, ethnicity, and body mass index. Other factors studied included disease duration, waist to hip ratio, parental history of diabetes, calories, carbohydrate, and season of interview.Among controls, dietary fiber was inversely associated with fasting plasma insulin concentration adjusted for calories and carbohydrate. However, in a comparison of previously known diabetics and nondiabetic controls there was a tendency toward higher reported intakes of fiber prior to diagnosis among the diabetics. A decrease in fiber of 10 g/day was associated with a decrease in risk of NIDDM of 0.80 (95% CI = 0.65,0.98). When the diabetic group was limited to cases of less than five years duration, the association was no longer present. When previously undiagnosed diabetics were compared to normal controls, the odds ratio relating a 10 g/day decrease in fiber intake to NIDDM was 1.61 (95% CI = 0.94, 2.77) after adjusting for calories. The odds ratio relating fiber to NIDDM was reduced to 1.12 (95% CI = 0.62,2.00) when carbohydrate was included in the model. This study has limitations including retrospective recall of diet, fiber measurement, and relatively small numbers of previously undiagnosed diabetics. Nonetheless, these findings do not support the hypothesis that increasing dietary fiber intake could reduce the future occurrence of NIDDM