Vitamin D concentration, obesity, and risk of diabetes: a mendelian randomisation study

Background Low plasma 25-hydroxyvitamin D (25[OH]D) concentration and high BMI have been associated with increased risk of diabetes. We tested the hypotheses that genetic variants associated with low concentrations of 25(OH)D are associated with diabetes, and that the effect on diabetes of genetic variants associated with high BMI is partly mediated through reduced plasma 25(OH)D concentration. Methods In this mendelian randomisation study, we genotyped 96 423 white Danes aged 20–100 years from three studies. 5037 of these participants had type 2 diabetes. All individuals were surveyed for diabetes from 1977 to 2011. 31 040 participants had their plasma 25(OH)D concentration measured and 90 169 had their BMI measured. We assessed the effects of genetic variation in DHCR7 (related to endogenous production) and CYP2R1 (related to liver conversion) on plasma 25(OH)D concentration, and the effects of variation in FTO, MCR4, and TMEM18 on BMI. We then assessed the effect of genetic variation in these genes on risk of type 2 diabetes, and the association of measured plasma 25(OH)D concentration and BMI with risk of type 2 diabetes. We did a mediation analysis to assess how much of the effect of BMI genotype on risk of diabetes was mediated through plasma 25(OH)D concentration. Findings The odds ratios for type 2 diabetes for participants who had a 20 nmol/L reduction in plasma 25(OH)D concentration as determined by genetics were 1·51 (95% CI 0·98–2·33) for DHCR7 and 1·02 (0·75–1·37) for CYP2R1. The DHCR7 allele score was significantly associated with increased risk of type 2 diabetes (p for trend=0·04), whereas the allele score for CYP2R1 was not. For participants who had a measured 20 nmol/L reduction in plasma 25(OH)D concentration, the adjusted odds ratio for type 2 diabetes was 1·16 (1·08–1·25). For participants who had a 10 kg/m2 increase in BMI as determined by genetics, the odds ratio for type 2 diabetes was 19·4 (6·4–59·1); this was associated with an 11·1 nmol/L (2·6–19·6) lower plasma 25(OH)D concentration. For a 10 kg/m2 increase in measured BMI, the adjusted odds ratio for type 2 diabetes was 4·33 (3·70–5·07); this was associated with a 9·1 nmol/L (8·4–9·7) lower plasma 25(OH)D concentration. Mediation analysis showed that 3% (1–5) of the effect of BMI on risk of type 2 diabetes was mediated through lowered plasma 25(OH)D concentrations. Interpretation Genetic variants associated with low plasma 25(OH)D concentrations are associated with type 2 diabetes and low plasma 25(OH)D concentrations might be a modest mediator between obesity and increased risk of diabetes. Genetic variants associated with endogenous production of 25(OH)D might partially explain this increased risk; however, as findings for DHCR7 were not statistically significant, our results require independent confirmation. Funding Danish Heart Foundation, Copenhagen University Hospital.