Association Between Genetically Predicted 25-Hydroxyvitamin D and Type 2 Diabetes Risk in South Asians: An...

Vitamin D deficiency and Type 2 diabetes (T2D) are highly prevalent in South Asian populations. Observational studies report inverse associations between serum 25-hydroxyvitamin D [25(OH)D] and T2D risk, but these findings may reflect confounding or reverse causation. We used Mendelian randomization (MR) to assess the association between genetically predicted 25(OH)D and T2D risk in South Asians.

We conducted a two-sample MR analysis using publicly available OpenGWAS summary statistics. One variant (rs3755967) met the conventional genome-wide significance threshold (p < 5 × 10^-8), while two additional variants (rs2459139 and rs72744878) met an exploratory threshold (p < 1 × 10^-6). The three variants served as instruments for serum 25(OH)D in a South Asian GWAS (ukb-e-30890_CSA; n = 7380; mean F-statistic = 29.6; total R² = 1.06%). The primary outcome was BMI-unadjusted T2D (ebi-a-GCST90093109; n = 50,533; 16,677 cases), with a BMI-adjusted GWAS used for sensitivity analysis. The primary estimator was inverse-variance weighted (IVW), supported by weighted-median, MR-Egger, simple-mode, and weighted-mode analyses. Heterogeneity and directional pleiotropy were assessed using Cochran's Q and the MR-Egger intercept, respectively. Partial participant overlap could not be excluded because UK Biobank contributed to both datasets.

The primary IVW analysis showed no statistically significant association between genetically predicted 25(OH)D and T2D risk (β = 0.058, SE = 0.120, p = 0.630; OR = 1.06, 95% CI: 0.83-1.36). Weighted-median and mode-based estimates were also non-significant. The MR-Egger estimate was highly imprecise (OR = 32.0, 95% CI: 0.11-9103) because only three instruments were available. No significant heterogeneity or directional pleiotropy was detected, although these tests had limited power. The BMI-adjusted sensitivity analysis produced similar findings.

We found no statistically significant evidence of a linear association between genetically predicted serum 25(OH)D and T2D risk. However, the small exposure GWAS, limited instrument set, exploratory instrument threshold, and possible sample overlap mean that small-to-moderate causal effects cannot be excluded. Larger South Asian GWAS are needed for more definitive inference.