Causal effects of gut microbiota on autism spectrum disorder: A two-sample mendelian randomization study

Abstract

There is increasing evidence that alterations in gut microbiota (GM) composition are associated with autism spectrum disorder (ASD), but no reliable causal relationship has been established. Therefore, a 2-sample Mendelian randomization (MR) study was conducted to reveal a potential causal relationship between GM and ASD. Instrumental variables for 211 GM taxa were obtained from genome-wide association studies (GWAS) and Mendelian randomization studies to estimate their impact on ASD risk in the iPSYCH-PGC GWAS dataset (18,382 ASD cases and 27,969 controls). Inverse variance weighted (IVW) is the primary method for causality analysis, and several sensitivity analyses validate MR results. Among 211 GM taxa, IVW results confirmed that Tenericutes (P value = .0369), Mollicutes (P value = .0369), Negativicutes (P value = .0374), Bifidobacteriales (P value = .0389), Selenomonadales (P value = .0374), Bifidobacteriaceae (P value = .0389), Family XIII (P value = .0149), Prevotella7 (P value = .0215), Ruminococcaceae NK4A214 group (P value = .0205) were potential protective factors for ASD. Eisenbergiella (P value = .0159) was a possible risk factor for ASD. No evidence of heterogeneous, pleiotropic, or outlier single-nucleotide polymorphism was detected. Additionally, further sensitivity analysis verified the robustness of the above results. We confirm a potential causal relationship between certain gut microbes and ASD, providing new insights into how gut microbes mediate ASD. The association between them needs to be further explored and will provide new ideas for the prevention and treatment of ASD.

Figure 1.

Overview of MR analyses process and major assumptions, and Gut-brain axis. (A) Overview of the analysis process of the causal relationship between the gut microbiome and ASD through MR analyzes. (B) Three core assumptions of Mendelian randomization analysis. (C) Gut microorganisms relay messages to the brain via various direct and indirect mechanisms). ASD = Autism spectrum disorder, MR = Mendelian randomization.

Figure 2.

Associations of genetically predicted GM with risk of ASD (Associations of genetically predicted GM with risk of ASD). CI = confidence interval, OR = odds ratio.

Figure 3.

Summary of scatter plots potential associations between gut microbiota and ASD risk. Figure A-I shows the positive associations between gut microbiota and ASD risk, while Figure J shows the opposite. Each dot in the graph represents an SNP locus. The vertical axis of the graph is the effect of the instrumental variable on the outcome, the horizontal axis is the effect of the instrumental variable on the exposure, and the ratio of the 2 effects is the effect of exposure on the outcome, that is, the slope of the regression line corresponds to the causal effect of exposure on the outcome in the graph. Horizontal and vertical crosses show a 95% confidence interval for each association. The estimates for the MR Analysis were slightly different, but the overall downward-sloping trend suggests that exposure (this gut microbiota) may have a negative causal effect on the outcome (ASD). ASD = Autism spectrum disorder, SNP = single-nucleotide polymorphisms, MR = mendelian randomization.