Causal relationships between irritable bowel syndrome and brain functional networks: A Mendelian randomization study.

BACKGROUND AND

OBJECTIVE

The association between irritable bowel syndrome (IBS) and brain function remains unclear. We conducted Mendelian randomization (MR) analysis to investigate the possible causal relationships between them.

METHODS

We performed bidirectional MR analysis using genome-wide association study data from IBS and 191 brain resting-state functional magnetic resonance imaging (rsfMRI) traits. Odds ratio (OR) with 95% confidence intervals was used to estimate the causal effects. Enrichment analyses were further conducted on genes corresponding to involved single nucleotide polymorphisms.

RESULTS

In forward analysis, IBS potentially increased the connectivity between postcentral/ precentral gyrus and cerebellum (OR = 1.15, P = 0.014), and decreased the connectivity between occipital/precuneus and superior frontal lobe (OR = 0.87, P = 0.037), as well as the connectivity between cerebellum and subcortical region (OR = 0.86, P = 0.017). In reverse analysis, increased neural activity in parietal/postcentral/precuneus region was potentially negatively associated with IBS (OR = 0.93, P = 0.036); the connectivity between cerebellum and subcortical region (OR = 1.16, P = 0.042) was potentially positively associated with IBS; the connectivity between cerebellum and precentral/frontal/superior motor area (OR = 0.85, P = 0.032), calcarine/lingual and paracentral gyrus (OR = 0.88, P = 0.031), postcentral/precentral and subcortical region (OR = 0.89, P = 0.019) were potentially negatively associated with IBS. The abnormally connected brain networks mainly included subcortical-cerebellum, motor, default mode and attention networks. Enrichment analyses demonstrated that the genes were mainly enriched in cell junctions and long-term potentiation pathways.

CONCLUSIONS

Our findings provided suggestive evidence for several causal relationship between IBS and rsfMRI phenotypes, in which, cerebellum, subcortical regions, postcentral and precentral gyrus played important roles. These putative causal relationships warrant further validation in independent, ideally larger-scale datasets.