Causal association between vitamin D and gestational diabetes mellitus: a two-sample Mendelian randomization study.

Abstract

Background: Previous articles on the relationship between vitamin D and gestational diabetes mellitus (GDM) were inconsistent. Their relationship has been observed primarily through observational studies, and the causality of this association has not been established.

Methods: A two-sample Mendelian randomization (MR) research was conducted to test the causal association between vitamin D and GDM, utilizing publically available statistics from genome-wide association studies (GWAS). This study obtained genetic variants from GWAS including vitamin D (N = 373,045,10,783,672 Single Nucleotide Polymorphisms SNPs), and GDM (5687 cases and 117,892 controls). The major technique was the inverse variance weighted approach (IVW), although there were other approaches as well, such as MR-Egger regression, weighted median, weighted mode, and simple mode. Additionally, we conducted sensitivity analyses to detect any potential diversity and horizontal pleiotropy.

Results: The study suggested that there was no causal link between vitamin D and GDM (all methods p > 0.05). For heterogeneity, MR egger Q value was 113.7, p < 0.05; IVW Q value was 114.7, p < 0.05. Therefore, random- effects IVW approach was applied. Regarding pleiotropy, the MR Egger regression intercept was 0.0046, which was close to zero with a p value of 0.452, suggesting the absence of pleiotropy.

Conclusions: We observed no assosiation between genetically predicted vitamin D and the risk of GDM, implying that insufficient vitamin D may do not confer an increased susceptibility to GDM.