Diabetes, glycemic profile and risk of vitiligo: A Mendelian randomization study.

Abstract

Previous observational studies have shown that vitiligo usually co-manifests with a variety of dysglycemic diseases, such as Type 1 diabetes mellitus (T1DM) and Type 2 diabetes mellitus (T2DM). Mendelian randomization (MR) analysis was performed to further evaluate the causal association between fasting plasma glucose, glycosylated hemoglobin (HbA1c), T1DM, T2DM and vitiligo. We used aggregated genome-wide association data from the Integrative Epidemiology Unit (IEU) online database of European adults vitiligo; HbA1c data were from IEU. Fasting blood glucose data were obtained from the European Bioinformatics Institute (EBI). T1DM and T2DM data were from FinnGen. We used bidirectional two-sample and multivariate MR analyses to test whether dysglycemic measures (fasting blood glucose, HbA1c), diabetes-related measures (T1DM, T2DM) are causatively associated with vitiligo. Inverse variance weighting (IVW) method was used as the main test method, MR-Egger, Weighted mode and Weighted median were used as supplementary methods. We found no statistically significant evidence to support a causal association between dysglycemic traits and vitiligo, but in the correlation analysis of diabetic traits, our data supported a positive causal association between T1DM and vitiligo (p=0.018). In the follow-up multivariate MR analysis, our results still supported this conclusion (p=0.016), and suggested that HbA1c was not a mediator of T1DM affecting the pathogenesis of vitiligo. No reverse causality was found in any of the reverse MR Analyses of dysglycemic traits and diabetic traits. Our findings support that T1DM is a risk factor for the development of vitiligo, and this conclusion may explain why the co-presentation of T1DM and vitiligo is often seen in observational studies. Clinical use of measures related to T1DM may be a new idea for the prevention or treatment of vitiligo.