**Background:** Stroke is the leading cause of life expectancy loss in China and a major global cause of death and disability. While China has established stroke prevention centers and green channel programs, the prognosis remains poor, with most survivors experiencing sequelae. Given China's rapid economic growth and lifestyle changes, this study investigated whether high-risk factors and high-risk populations for stroke have shifted over time, aiming to inform updated prevention strategies.
**Methods:** Data were derived from the China Health and Nutrition Survey (CHNS), an ongoing open cohort covering 12 Chinese provinces. A total of 42,419 middle-aged and elderly residents aged ≥45 years were included across survey waves from 1989 to 2015. Participants with missing data on key variables were excluded (N=92,790 excluded for age <45; additional exclusions for missing stroke diagnosis, marital status, education, occupation, income, smoking, alcohol, activity, sleep, hypertension, diabetes, atrial fibrillation, height/weight). Stroke was defined by self-reported physician diagnosis of stroke or transient ischemic attack. Risk factors assessed included sociodemographic factors (gender, age, residence, marital status, education, income, occupation), lifestyle factors (smoking, alcohol, activity participation, sleep duration), historical health variables (hypertension, diabetes, atrial fibrillation, BMI), and regional information (province). Chi-square tests, Shapley value decomposition models, and CHAID decision tree models were used for analysis.
**Key Results:** Overall stroke prevalence was 2.44% (1,035/42,419). Stroke prevalence decreased from 1989 to 2015. Hypertension was the leading risk factor across all time points, but its contribution rate declined from 50.91% in 2009 to 44.51% in 2011 and 36.39% in 2015. Age was consistently the second-ranked risk factor, with contribution declining from 19.78% to 11.01%. The contribution of diabetes increased from 2.77% (rank 7) in 2009 to 9.68% (rank 4) in 2015. Sleep duration contribution rose from 2.89% (rank 6) to 9.55% (rank 5). Province contribution increased from 1.71% (rank 11) to 10.29% (rank 3). The decision tree analysis identified specific high-risk subgroups: in 2009, the highest-risk group was hypertensive patients aged ≥75 years (stroke prevalence 18.0%). By 2015, the first high-risk group shifted to residents without spouse living in Beijing and Liaoning (stroke prevalence 9.5%). The second high-risk group in 2015 was male hypertensive patients living in urban areas (9.1%), and the third was female patients with both hypertension and diabetes (7.5%). The hypertension treatment rate increased from 20.18% in 1991 to 84.26% in 2015.
**Clinical Implications:** The declining contribution of hypertension to stroke risk, alongside rising treatment rates, suggests that improved hypertension management has been effective. However, the emergence of regional factors (Beijing, Liaoning, Shanghai forming a 'Stroke Belt'), lifestyle factors (sleep duration >10 hours associated with 22.4% stroke prevalence in hypertensive patients), and marital status (no spouse) as key risk determinants indicates that current screening and prevention strategies need updating. The authors recommend integrating stroke screening with routine health examinations, conducting targeted health education for high-risk populations, advocating healthy lifestyles, standardizing antihypertensive drug use, and building community-level health files linked to China's family doctor contract system. Limitations include reliance on self-reported stroke diagnosis, lack of distinction between ischemic and hemorrhagic stroke, and potential cohort effects from participant turnover across survey waves.