this stage, reducing our ability to model stroke risk in the age strata where the majority of events occur. Furthermore, the duration of follow-up in UKB is relatively limited and, because of the limited number of assessments, we could not model the cumulative effect of BP and smoking over time; however, we accounted for potential regression dilution bias in SBP measurements via the use of diagnosed hypertension, which showed stronger associations with stroke. Family history of stroke in UKB may be less comprehensive than in stroke-specific studies, limiting its predictive power, and overall the UKB study population is healthier than the general UK population42, which could have led to underestimation of some of the effects of risk factors. Further independent validation of the stroke metaGRS in other cohorts would be necessary before considering its clinical use; however, this is challenging given that the majority of available stroke GWAS studies have been included in MEGASTROKE. Nonetheless, recent successful validation of our previous CAD metaGRS21 in French-Canadian cohorts43 suggests that scores developed on the UK Biobank are generalisable to other cohorts of European ancestry. Our modelling assumes that risk factors, such as SBP and BMI, can be varied independently of each other.
