Anorexia nervosa and bulimia nervosa: a Mendelian randomization study of gut microbiota

Abstract

Background: Anorexia nervosa (AN) and bulimia nervosa (BN) poses a significant challenge to global public health. Despite extensive research, conclusive evidence regarding the association between gut microbes and the risk of AN and BN remains elusive. Mendelian randomization (MR) methods offer a promising avenue for elucidating potential causal relationships.

Materials and methods: Genome-wide association studies (GWAS) datasets of AN and BN were retrieved from the OpenGWAS database for analysis. Independent single nucleotide polymorphisms closely associated with 196 gut bacterial taxa from the MiBioGen consortium were identified as instrumental variables. MR analysis was conducted utilizing R software, with outlier exclusion performed using the MR-PRESSO method. Causal effect estimation was undertaken employing four methods, including Inverse variance weighted. Sensitivity analysis, heterogeneity analysis, horizontal multivariate analysis, and assessment of causal directionality were carried out to assess the robustness of the findings.

Results: A total of 196 bacterial taxa spanning six taxonomic levels were subjected to analysis. Nine taxa demonstrating potential causal relationships with AN were identified. Among these, five taxa, including Peptostreptococcaceae, were implicated as exerting a causal effect on AN risk, while four taxa, including Gammaproteobacteria, were associated with a reduced risk of AN. Similarly, nine taxa exhibiting potential causal relationships with BN were identified. Of these, six taxa, including Clostridiales, were identified as risk factors for increased BN risk, while three taxa, including Oxalobacteraceae, were deemed protective factors. Lachnospiraceae emerged as a common influence on both AN and BN, albeit with opposing effects. No evidence of heterogeneity or horizontal pleiotropy was detected for significant estimates.

Conclusion: Through MR analysis, we revealed the potential causal role of 18 intestinal bacterial taxa in AN and BN, including Lachnospiraceae. It provides new insights into the mechanistic basis and intervention targets of gut microbiota-mediated AN and BN.

Keywords: Mendelian randomization analysis; anorexia nervosa; bulimia nervosa; eating disorders; gut microbiota.

Figure 1

Flow chart of the study. GWAS, genome wide association study; SNPs, single nucleotide polymorphisms; MR, Mendelian randomization; LD, linkage disequilibrium; IVW, inverse variance weighted.

Figure 2

Scatter plots of MR analysis of potential causal effect of 9 gut microbiota on AN. (A) Cyanobacteria, (B) Gammaproteobacteria, (C) Mollicutes RF9, (D) Peptostreptococcaceae, (E) Coprococcus3, (F) Escherichia Shigella, (G) Eubacterium brachy group, (H) Lachnospiraceae NC2004 group, (I) Lachnospiraceae UCG010. IVW, inverse variance weighted; AN, anorexia nervosa; MR, Mendelian randomization; SNP, single nucleotide polymorphism.

Figure 3

Scatter plots of MR analysis of potential causal effect of 9 gut microbiota on BN. (A) Clostridiales, (B) Rhodospirillales, (C) Oxalobacteraceae, (D) Bilophila, (E) Coprobacter, (F) Holdemania, (G) Lachnospiraceae UCG008, (H) Ruminococcaceae UCG009, (I) Slackia. IVW, inverse variance weighted; BN, bulimia nervosa; MR, Mendelian randomization; SNP, single nucleotide polymorphism.