Predictors of Stroke Among U.S. Adults: A Survey-Weighted Analysis of the Behavioral Risk Factor Surveillance System, 2021 to 2023

Background Stroke remains one of the leading causes of mortality, disability, and healthcare burden worldwide. Identifying demographic, socioeconomic, lifestyle, and clinical factors associated with stroke is essential for improving prevention strategies and reducing disease burden. This study aimed to identify independent predictors of stroke among U.S. adults using nationally representative Behavioral Risk Factor Surveillance System (BRFSS) data collected between 2021 and 2023. Methods A cross-sectional analysis was conducted using pooled BRFSS data from 2021 to 2023. Adults with complete information on stroke status and study variables were included in the multivariable analysis. Stroke status was determined from self-reported physician diagnosis. Survey-weighted multivariable logistic regression was performed to estimate adjusted odds ratios (aORs) and 95% confidence intervals (CIs) for demographic, socioeconomic, lifestyle, and clinical predictors while accounting for the complex BRFSS sampling design. Model discrimination was evaluated using receiver operating characteristic (ROC) curve analysis. Results Among 235,571 participants in the pooled dataset, stroke was more common among older adults and individuals with diabetes, poorer self-reported health, lower income, and smoking history. In the adjusted analysis, increasing age (aOR 1.04, 95% CI 1.04 to 1.04), diabetes (aOR 1.55, 95% CI 1.43 to 1.67), current smoking (aOR 1.44, 95% CI 1.31 to 1.58), multiracial ethnicity (aOR 1.44, 95% CI 1.12 to 1.82), Black race (aOR 1.31, 95% CI 1.15 to 1.50), and poorer general health (aOR 1.58, 95% CI 1.53 to 1.64) were independently associated with higher odds of stroke. Conversely, Asian race (aOR 0.65, 95% CI 0.43 to 0.94), Hispanic ethnicity (aOR 0.65, 95% CI 0.54 to 0.77), higher income (aOR 0.92, 95% CI 0.90 to 0.93), and regular physical activity (aOR 0.86, 95% CI 0.80 to 0.92) were associated with lower odds of stroke. The final model demonstrated good discrimination, with an area under the ROC curve of 0.781 (95% CI 0.774 to 0.788). Conclusions Stroke among U.S. adults is independently associated with a combination of demographic, socioeconomic, lifestyle, and clinical factors. Diabetes, smoking, poor general health, and socioeconomic disadvantage remain important potentially modifiable contributors to stroke risk, whereas regular physical activity appears protective. These findings support targeted public health interventions focused on improving cardiometabolic health, promoting smoking cessation and physical activity, and addressing socioeconomic disparities to reduce the burden of stroke in the United States.