Dietary manganese and type 2 diabetes mellitus: two prospective cohort studies in China

Abstract

Aims/hypothesis: The association between dietary Mn and type 2 diabetes is unclear. We aimed to elucidate whether dietary Mn is associated with type 2 diabetes, to investigate whether this association is independent of dietary total antioxidant capacity (TAC) and to explore the underlying mechanisms in their association. Methods: Two prospective cohorts of 3350 and 7133 Chinese adults (20–74 years old) were enrolled including, respectively, 244 and 578 individuals newly diagnosed with type 2 diabetes, with mean values of 4.2 and 5.3 years of follow-up. Cox’s proportional-hazards regression and linear regression were performed to investigate the association between dietary Mn and type 2 diabetes (diagnosed by OGTT) or HbAlc and to analyse the joint association between dietary Mn and TAC. Restricted cubic spline (RCS) regression was applied to the non-linear association between dietary Mn and incidence of type 2 diabetes. Mediation analysis was applied to explore potential mediators in their association in a subgroup of 500 participants. Results: Dietary Mn intakes were 4.58 ± 1.04 and 4.61 ± 1.08 (mean ± SD) mg/day in the two cohorts. Dietary Mn was inversely associated with type 2 diabetes incidence and HbAlc concentration in both cohorts (ptrend < 0.01 and <0.01 for type 2 diabetes, and ptrend < 0.01 and =0.02 for HbAlc, respectively, in each cohort) independent of TAC, adjusted for age, sex, BMI, tobacco use, alcohol consumption, physical activity, diabetes inheritance, total energy, carbohydrate, total fatty acids, fibre, calcium, Mg, hypertension, hyperlipidaemia, and impaired glucose tolerance or FBG (all at baseline). Their inverse association was stronger in the presence of diets with high, compared with low, TAC. In RCS, intakes of >6.01 and 6.10–6.97 mg/day were associated with a significantly lower type 2 diabetes incidence in the two respective cohorts. Mediation analysis showed that high plasma Mn and low oxidative stress (increased Mn superoxide dismutase and decreased 8-hydroxydeoxyguanosine) contributed to the association between dietary Mn and both type 2 diabetes and HbAlc. Conclusions/interpretation: Dietary Mn was inversely associated with type 2 diabetes independently of TAC. In addition, this association was stronger in a high- rather than low-TAC diet. Plasma Mn and oxidative stress were mediators in the association between dietary Mn and type 2 diabetes. Future studies on absolute Mn intake should be conducted to study the potential non-linearity and optimal levels of dietary Mn and type 2 diabetes.