Abstract
BACKGROUND: Cognitive impairment is widely reported in migraineurs. A Mendelian randomisation (MR) approach, similar to a randomised-controlled trial, employs single-nucleotide polymorphisms (SNPs) to investigate causal relationships. METHODS: This study comprised one- and two-sample MR analyses of the Taiwan Biobank. Three strategies were used to obtain causal estimates: (1) a polygenic risk score (PRS) method-several SNPs associated with migraines were constructed as a single instrument variable; (2) a meta-analysis of genome-wide association study (GWAS) statistics for traits of migraines and cognitive impairment in the framework of a one-sample MR; and (3) a two-sample MR analysis with a meta-analysis of GWAS statistics in two distinct datasets (IEU GWAS database and the Taiwan Biobank). RESULTS: In strategy 1, the PRS constructed by 18 selected SNPs exhibited a causal association with cognitive impairment (β = -2.31, 95% confidence interval [CI]: -4.56 to -0.06). In strategy 2, a one-sample MR showed migraines were causally associated with cognitive impairment (inverse-variance weighted [IVW] estimator β = 2.90; 95% CI: 0.90-4.89). In strategy 3, a two-sample MR validated migraines to be causally associated with cognitive impairment (IVW estimator β = 2.43; 95% CI: 1.08-3.78). CONCLUSIONS: Migraine, a polygenic disorder, is causally associated with cognitive impairment.