Association between sugar and starch intakes and type 2 diabetes risk in middle-aged adults in a prospective cohort study

Objectives We aimed to investigate the association between sugar or starch intake and the risk of type 2 diabetes (T2D) in middle-aged Japanese adults. Subjects/methods Participants comprised 27,797 men and 36,880 women aged 45–75 years with no history of diabetes and critical illness before the second survey in the Japan Public Health Center-based Prospective Study. We calculated sugar (total sugar, total fructose, and sugar subtypes) and starch intakes (% energy/d) using a validated 147-item food frequency questionnaire, to estimate the average dietary intake over the previous year. T2D onset was defined by validated self-reports. ORs adjusted for potential confounders were estimated using multiple logistic regression with categorical and cubic spline models. Results During the 5-year follow-up, 690 men and 500 women were identified with T2D. In women, the quartiles of total sugar or total fructose intakes were not significantly associated with T2D risk; however, the spline curves showed an increased risk at extremely high intake levels (ORs [95% CI]: 1.88 [1.07–3.31] at 30% energy/d for total sugar and 1.87 [1.10–3.16] at 14% energy/d for total fructose). Starch intake was positively associated with T2D risk among women in the categorical and spline models (ORs [95% CI]: 1.55 [1.13–2.12] at 50% energy/d). In men, sugar and starch intakes were not associated with T2D risk. Conclusions In this large-scale population-based cohort study, starch intake was associated with an increased T2D risk in Japanese women. An increased risk with extremely high intake of total sugar or total fructose among women cannot be disregarded.