Global variation in stroke burden and mortality: estimates from monitoring, surveillance, and modelling

Background Recent improvements in the monitoring and modelling of stroke have led to more reliable estimates of stroke mortality and burden worldwide. However, little is known about the global distribution of stroke and its relations to the prevalence of cardiovascular disease risk factors and sociodemographic and economic characteristics. Methods National estimates of stroke mortality and burden (measured in disability-adjusted life years [DALYs]) were calculated from monitoring vital statistics, a systematic review of studies that report disease surveillance, and modelling as part of the WHO Global Burden of Disease programme. Similar methods were used to generate standardised measures of the national prevalence of cardiovascular risk factors. Risk factors other than diabetes and disease burden estimates were age-adjusted and sex-adjusted to the WHO standard population. Findings There was a ten-fold difference in rates of stroke mortality and DALY loss between the most-affected and the least-affected countries. Rates of stroke mortality and DALY loss were highest in eastern Europe, north Asia, central Africa, and the south Pacific. National per capita income was the strongest predictor of mortality and DALY loss rates (p<0.0001) even after adjustment for cardiovascular risk factors (p<0.0001). Prevalences of cardiovascular risk factors measured at a national level were generally poor predictors of national stroke mortality rates and burden, although raised mean systolic blood pressure (p=0.028) and low body-mass index (p=0.017) predicted stroke mortality, and greater prevalence of smoking predicted both stroke mortality (p=0.041) and DALY-loss rates (p=0.034). Interpretation Rates of stroke mortality and burden vary greatly among countries, but low-income countries are the most affected. Current measures of the prevalence of cardiovascular risk factors at the population level poorly predict overall stroke mortality and burden and do not explain the greater burden in low-income countries.