Abstract text (incl. figure legends and references)

Question: Migraine has been linked to inflammatory bowel disease (IBD) and coeliac disease. This paper assesses whether the link may be explained by a shared genetic basis or could be causal.

Methods: Linkage disequilibrium score regression and two-sample bidirectional Mendelian randomisation (MR) analyses were performed using summary statistics from genome-wide association studies of migraine (59,674 cases; 316,078 controls), IBD (25,042 cases; 34,915 controls) and coeliac disease (11,812 or 4533 cases; 11,837 or 10,750 controls). Migraine with and without aura (MA and MO) were analysed separately, as were the two IBD subtypes Crohn"s disease and ulcerative colitis. Positive control analyses and conventional MR sensitivity analyses were performed.

Results: Migraine was not genetically correlated with IBD or coeliac disease. No evidence was observed for IBD or coeliac disease causing migraine or vice versa when all migraineurs were analysed jointly (p > 0.05 for all). There was some indication of causality between coeliac disease and MA (odds ratio 1.04, 95% confidence interval 1.00–1.08, p = 0.045) and between coeliac disease and MO (0.95, 0.92–0.99, p = 0.006), as well as between MO and ulcerative colitis (1.15, 1.02–1.29, p = 0.025). The results were, however, not significant after multiple testing correction.

Conclusions: We found no evidence of a shared genetic basis or of a causal association between migraine and either IBD or coeliac disease, although we obtained some indication of causality with migraine subtypes.

Figure 1. Effects of genetic liability to gastrointestinal conditions on migraine. Forest plot of two-sample MR effect estimates for IBD and coeliac disease on migraine based on the inverse-variance weighted method.

Figure 2. Effects of genetic liability to migraine on gastrointestinal conditions. Forest plot of two-sample MR effect estimates for migraine on IBD and coeliac disease based on the inverse-variance weighted method.