Inferring the genetic effects of serum homocysteine and vitamin B levels on autism spectral disorder through Mendelian randomization.

purposeThe previous studies have suggested that serum homocysteine (Hcy) and vitamin B levels are potentially related to autism spectrum disorder (ASD). However, the causality between their concentrations and ASD risk remains unclear. To elucidate this genetic association, we used a Mendelian randomization (MR) design.

methodsFor this MR analysis, 47 single-nucleotide polymorphisms (SNPs)-13 related to Hcy, 13 to folate, 14 to vitamin B6, and 7 to vitamin B12-were obtained from a large-scale Genome-Wide Association Studies (GWAS) database and employed as instrumental variables (IVs). Our study used three approaches to calculate the MR estimates, including inverse-variance weighted (IVW) method, MR-Egger method, and weighted median (WM) method. Among these, the IVW method served as our primary MR method. False discovery rate (FDR) was implemented to correct for multiple comparisons. We also performed a series of sensitivity analyses, including Cochran's Q test, MR-Egger's intercept, MR-PRESSO, leave-one-out analysis, and the funnel plot.

resultsUnivariable Mendelian randomization (UVMR) analysis revealed a statistical association between serum vitamin B12 levels and ASD risk (OR = 1.68, 95% CI 1.12-2.52, P = 0.01) using the IVW method. However, neither the WM method (OR = 1.57, 95% CI 0.93-2.66, P = 0.09) nor the MR-Egger method (OR = 2.33, 95% CI 0.48-11.19, P = 0.34) was significantly association with higher levels of serum vitamin B12 and ASD risk. Additionally, we found no evidence of causal relationships between serum levels of vitamin B6, folate, Hcy, and ASD risk. After correcting for the FDR, the causality between serum vitamin B12 levels and ASD risk remained significant (q value = 0.0270). Multivariate Mendelian randomization (MVMR) analysis indicated an independent association between elevated serum vitamin B12 levels and the risk of ASD (OR = 1.74, 95% CI 1.03-2.95, P = 0.03) using the IVW method, but this finding was inconsistent when using the WM method (OR = 1.73, 95% CI 0.89-3.36, P = 0.11) and MR-Egger method (OR = 1.60, 95% CI 0.95-2.71, P = 0.08). Furthermore, no causal associations were observed for serum levels of vitamin B6 and folate in MVMR analysis. Sensitivity analyses confirmed that these results were reliable.

conclusionOur study indicated that elevated serum vitamin B12 levels might increase the risk of ASD. The potential implications of our results for ASD risk warrant validation in randomized clinical trials.