Dietary carbohydrate intake, presence of obesity and the incident risk of type 2 diabetes in Japanese men

Abstract

Aims/introduction: The present cohort study assessed the risk among Japanese men for developing type 2 diabetes, based on the percentage of energy intake from carbohydrates and degree of obesity.

Participants and methods: The participants were 2,006 male factory employees, and the macronutrient intake of each patient was measured using a self-administered diet history questionnaire. The incidence of diabetes was determined in annual blood examinations over a 10-year period.

Results: During the study, 232 participants developed diabetes. The crude incidence rates (/1,000 person-years) for different levels of carbohydrate intake as a percentage of calories consumed (<50.0, 50.0-57.4, 57.5-65.0, >65.0% of energy intake) were 16.5, 14.4, 12.7 and 17.6. Overall, carbohydrate intake was not associated with the risk of diabetes. However, there was significant interaction between carbohydrate intake and degree of obesity on the incidence of diabetes (P for interaction = 0.024). Higher carbohydrate intake was associated with elevated risk for diabetes among participants with a body mass index ≥25.0 kg/m(2) (P for trend = 0.034). For obese participants, the multivariate-adjusted hazard ratio for those with carbohydrate intakes >65% energy was 2.01 (95% confidence interval 1.08-3.71), which was significantly higher than that of participants with carbohydrate intakes 50.0-57.4% energy.

Conclusions: Higher carbohydrate intake was associated with higher risk of diabetes in obese participants, but not in non-obese participants. Obese participants with carbohydrate intakes >65% energy should reduce their intakes to levels within the desirable carbohydrate energy proportion for Japanese (50-65% energy) to prevent development of type 2 diabetes.

Keywords: Cohort study; Dietary carbohydrates; Incidence.

Figure 1

Adjusted hazard ratio for type 2 diabetes based on dietary carbohydrate intake by the presence of obesity in Japanese men. Hazard ratios were adjusted for age, family history of diabetes, smoking, alcohol drinking, habitual exercise, total energy intake and total fiber intake. CI, confidence interval.