Longitudinal trajectories of stroke risk among Chinese adults: a four-year community screening cohort study

Stroke remains a leading cause of death and disability worldwide, with a growing burden in China. Traditional stroke risk assessments are often static, relying on single-time point measurements. This study aimed to identify heterogeneous longitudinal stroke risk trajectories and examine factors influencing these trajectories in a Chinese community cohort.

We used data from the National Stroke Screening and Intervention Program, which followed 1,376 participants across three waves (2018, 2020, 2022). Group-based trajectory modeling (GBTM) was applied to identify stroke risk trajectories based on cumulative risk scoring. We evaluated associations between baseline characteristics and trajectory group membership using multinomial logistic regression.

Three distinct stroke risk trajectories were identified: (1) Stable low-risk (41.42%); (2) Increasing-risk (41.21%); (3) Stable high-risk (17.37%). Significant baseline differences in demographic, lifestyle, and health-related factors were observed across the three groups, including drinking habits, taste preference, dietary pattern, fruit intake frequency, waist circumference and homocysteine levels. Multinomial logistic regression revealed that older age, frequent heavy drinking, vegetarian diet preference, larger waist circumference, and higher homocysteine levels were significantly associated with higher odds of progressing to the increasing-risk and stable high-risk trajectories.

This study identified heterogeneous stroke risk trajectories among Chinese adults over time. The identification of individuals on an increasing-risk trajectory may provide an opportunity for earlier intervention. These findings suggest the potential utility of trajectory-based approaches for improving stroke risk stratification and identifying individuals who may benefit from earlier and more targeted prevention.