Introduction:
Observational studies have suggested an association between vitamin D deficiency and the risk of neonatal jaundice; however, it remains unclear whether this relationship is causal. We conducted a Mendelian randomization (MR) study to evaluate whether vitamin D levels influence the risk of neonatal jaundice.

Material and methods:
Single nucleotide polymorphisms (SNPs) highly associated with vitamin D levels were selected as instrumental variables from publicly available genome-wide association studies (GWAS). MR analysis was conducted using five different models, including MR-Egger regression and inverse-variance weighting (IVW). Sensitivity analysis included MR-PRESSO (MR-pleiotropy residual sum and outlier) testing, Cochran’s Q heterogeneity testing, the MR-Egger method, leave-one-out analysis, and Bayesian colocalization analysis to determine whether there were shared causal loci between vitamin D levels and neonatal jaundice.

Results:
A total of 160 SNPs with genome-wide significance for vitamin D levels were identified, explaining 1.4% of the genetic variance in vitamin D levels. The MR-PRESSO test did not detect any outlier values, and heterogeneity testing did not identify significant heterogeneity. However, pleiotropy testing revealed significant horizontal pleiotropy, prompting the use of the MR-Egger regression model for MR analysis. The results indicated a significant negative causal association between vitamin D levels and the risk of neonatal jaundice (OR = 0.04, 95% CI: 0.004–0.43, p = 0.0026). Sensitivity analysis and colocalization analysis further confirmed the accuracy and robustness of the results.

Conclusions:
Genetically reduced vitamin D levels are causally associated with an increased risk of neonatal jaundice.