A nonlinear, causal link between serum 25-hydroxyvitamin D and prolactin in type 2 diabetes: Evidence from clinical and Mendelian randomization analyses

Abstract

Background

Type 2 diabetes mellitus is increasingly prevalent and often accompanied by vitamin D insufficiency. Prolactin, once considered solely lactogenic, has emerged as a metabolic regulator, yet its relationship with vitamin D in T2DM remains unclear.

Methods

We conducted a cross-sectional study of 221 male T2DM patients aged 25–75 years recruited between January 2022 and December 2024. Serum 25-hydroxyvitamin D (25(OH)D) and prolactin were measured, defining vitamin D sufficiency as ≥ 20 ng/mL. Associations were evaluated using group comparisons, Spearman correlation and multivariable regression adjusting for confounders. Restricted cubic splines assessed nonlinearity and thresholds. Causality was examined via Mendelian randomization employing 227 25(OH)D-associated variants.

Results

Vitamin D deficiency affected 59.7 % of participants. Median prolactin levels were higher in vitamin D-sufficient than in deficient patients. Serum 25(OH)D correlated positively with prolactin and remained significant after adjustment. Spline analysis suggested a nonlinear relationship with an inflection of 18.48 ng/mL: below this threshold prolactin decreased as 25(OH)D increased, whereas above it prolactin rose steeply. Mendelian randomization provided evidence for a causal association: the inverse variance weighted estimate was non-significant, but MR-Egger and weighted median analyses indicated a significant link without pleiotropy or heterogeneity.

Conclusion

These findings demonstrate a nonlinear, threshold-dependent association between vitamin D status and prolactin in male T2DM. Levels of 25(OH)D at 18.48 ng/mL were associated with suppressed prolactin, while sufficient concentrations enhanced prolactin, suggesting that maintaining adequate vitamin D may modulate prolactin and benefit metabolic outcomes. Further research is warranted to validate these observations.