A phenome-wide association study reveals novel phenotypic determinants of migraine risk in the UK Biobank

To identify genetic and phenotypic determinants of migraine susceptibility and progression using a large-scale, hypothesis-free approach.

Migraine is a common neurological disorder with a substantial individual and societal burden. While risk factors are known, hypothesis-driven approaches may overlook contributors.

We analyzed data from 502,364 UK Biobank participants (aged 37-73 years; recruited 2006-2010) to assess associations between migraine and 2824 phenotypes, including observational associations and associations with migraine polygenic risk scores. Time-dependent associations were evaluated, and causality was assessed. Follow-up extended from baseline through to November 30, 2024, with a median of ~14.7 years.

Our analyses identified 253 phenotypes significantly associated with migraine risk, 191 phenotypes associated with migraine PRS, and 87 overlapping between the two approaches. Cox regression analysis confirmed 84 robust time-dependent associations and highlighted novel risk factors, such as oral ulcers (adjusted hazard ratio [aHR] = 1.14, 95% confidence interval [CI]: 1.02-1.28, q = 2.54 × 10^-2), lower limb obesity (aHR = 1.01, 95% CI: 1.00-1.02, q = 3.52 × 10^-4), and hormone replacement therapy (aHR = 1.39, 95% CI: 1.29-1.50, q = 1.95 × 10^-17). MR analysis also revealed causal associations with emotional fluctuations (inverse-variance weighting odds ratio [IVW OR] 4.10, 95% CI = 1.76-9.53, q = 1.00 × 10^-2), neuroticism (IVW OR = 1.12, 95% CI = 1.07-1.08, q = 1.24 × 10^-5), general health ratings (IVW OR = 1.81, 95% CI = 1.43-2.29, q = 3.30 × 10^-2), and fatigue (IVW OR = 1.94, 95% CI = 1.35-2.79, q = 5.73 × 10^-36). However, the association with paracetamol use (IVW OR = 9.65 × 10⁴, 95% CI = 1.67 × 10⁴-5.57 × 10⁵, q = 5.73 × 10^-36) had uncertain directionality on Steiger testing and should be interpreted cautiously. In total, 87 genetic and phenotypic determinants of migraine were identified, offering new therapeutic targets involving inflammatory and insulin-like growth factor 1 (IGF-1) -related pathways.

This comprehensive analysis provides new insights into the genetic and phenotypic factors that influence migraine. The findings suggest new targets for therapeutic intervention and provide a basis for the development of precision prevention and clinical management strategies.