Background/Objective: Genome-wide association studies (GWASs) demonstrate a complex genetic landscape for migraine risk. Migraine polygenic risk scores (PRSs) developed from GWAS data may have utility for predicting disease course. We analyzed the strength of association between an integrative migraine PRS and age at onset and chronification. Methods: In this retrospective clinical/genetic case–control study, PGS004799 was calculated for adults with European ancestry from two real-world community cohorts. In the DodoNA cohort, 1653 treated, deeply phenotyped migraine cases, diagnosed using International Classification of Headache Disorders 3rd edition criteria, were followed for a mean (range) of 2.3 (1–10) years and compared to 3460 controls (without migraine diagnosis). In the GHI cohort, 2443 cases were identified using the first migraine ICD code as a proxy for migraine onset and compared to 8576 controls (without migraine ICD codes). PRS associations with age at onset (DodoNA) or first migraine ICD code (GHI) and chronification (DodoNA) were evaluated. Results: In both cohorts, PRS was higher in cases (DodoNA mean (range) cases: 0.82 (0.07–1.76), controls: 0.78 (0.04–1.56); t (5111) = −6.1, p = 1.4 × 10⁻⁹, GHI: cases: 0.79 (0.003–1.68), controls: 0.75 (−0.06–1.53); t (11,017) = −7.69, p = 1.6 × 10⁻¹⁴), and a higher PRS was associated with earlier onset in females (HR [95% CI] DodoNA: 2.1 [1.6–2.6, p < 0.001; GHI: 1.8 [1.4–2.1], p < 0.001) and in males (DodoNA: 2.5 [1.3–4.7], p = 0.005; GHI: 1.6 [1.1–2.6], p = 0.027). PRS was not different in cases with or without chronification (t (1651) = −1.67, p = 0.094) and was not associated with earlier chronification (1.2 [0.8–1.6], p = 0.424). Conclusions: Higher genetic risk was associated with earlier onset and increased risk of migraine well into adulthood, but not with chronification. This suggests that the PRS quantifies genetic susceptibility that is distinct from factors influencing disease course.

Background/Objectives: Migraine is a common neurological disorder with highly variable characteristics. While genome-wide association studies have identified genetic risk factors that implicate underlying pathways, the influence of genetic susceptibility on disease characteristics or treatment response is incompletely understood. We examined the relationships between a previously developed standardized integrative migraine polygenic genetic risk score (PRS) and migraine characteristics in a real-world, treated patient cohort. Methods: This retrospective cohort study used covariate-adjusted regression to comprehensively evaluate associations between the PRS and clinical characteristics in 1653 treated migraine cases with European ancestry at baseline and, in 800 cases, after one year. Cases were deeply phenotyped by neurologists during extensive interviews, using structured clinical documentation tools to record ~200 discrete data elements. Results: In treated patients, higher standardized PRS showed associations with two common migraine symptoms: photophobia (odds ratio [confidence interval]: 1.33 [1.13–1.56], p = 0.001) and stabbing pain (1.21 [1.08–1.36], p = 0.001]; both retained significance at Q = 0.05. Associations with phonophobia, nausea, emesis, and unilateral headache had similar effect sizes but did not survive correction for multiple tests. In this population, the PRS was not associated with other symptoms of migraine attacks, objective measures of migraine disability, frequency, severity, average duration, time-to-peak intensity of migraine attacks, chronification, emergency department visits, triptan responsiveness, or changes at follow-up. Conclusions: In treated patients, genetic risk was associated with common migraine symptoms but not with the severity of migraine characteristics or treatment outcomes. This suggests that in treated patients, other genetic and non-genetic factors influence migraine symptom severity and disease course more strongly than genetic susceptibility.