Stroke is the leading cause of death among residents in China and a major global public health issue, characterized by high incidence, high mortality, high disability, and high recurrence rates. Homocysteine (Hcy) is a potential risk factor for stroke; however, the association between the two remains controversial. Therefore, exploring Hcy as an early warning biomarker is of great significance for the prevention and control of stroke.

To investigate the association between Hcy levels and the risk of incident stroke among residents aged 40 years and older in Pudong New Area, and to provide evidence and guidance for community-based chronic disease health management as well as stroke prevention and control.

Using multistage stratified cluster sampling, permanent residents aged 40 years and above in Pudong New Area were recruited in October 2016 and October 2019, respectively, and were enrolled in the Pudong New Area Chronic Disease and Risk Factor Surveillance Cohort Study. A total of 13 395 participants were finally included (6 266 enrolled in 2016 and 7 129 newly enrolled in 2019). The date of the first enrollment was used as the start of follow-up, and the follow-up ended on December 31, 2023. The outcome events were newly diagnosed stroke cases or stroke-related deaths. Univariate and multivariate Cox regression analyses were performed to explore the association between serum Hcy and the risk of incident stroke, and the restricted cubic spline (RCS) regression model was used to examine the potential non-linear dose-response relationship between them. Subgroup analyses were conducted to further evaluate the association between Hcy and stroke risk across different subgroups.

During the follow-up period, 923 incident stroke cases occurred among the 13 395 participants, with a cumulative incidence of 7.14% in men and 6.73% in women. Multivariable Cox regression analysis showed that history of hypertension (HR=1.211, 95%CI=1.047-1.400), history of diabetes (HR=1.282, 95%CI=1.103-1.490), dyslipidemia (HR=1.178, 95%CI=1.016-1.366), overweight/obesity (HR=1.243, 95%CI=1.071-1.443), elevated Hcy levels (HR=1.009, 95%CI=1.001-1.016), and age (50-<60 years: HR=2.670, 95%CI=1.717-4.152; 60-<70 years: HR=4.508, 95%CI=2.947-6.897; 70-79 years: HR=7.478, 95%CI=4.834-11.569) were risk factors for incident stroke, whereas education level of junior high school (HR=0.769, 95%CI=0.634-0.931), senior high school (HR=0.675, 95%CI=0.538-0.847), and college diploma or above (HR=0.663, 95%CI=0.486-0.906) were protective factors (all P<0.05). RCS regression analysis revealed a non-linear association between Hcy and the risk of incident stroke (P_{overall association} =0.011, P_{non-linearity}=0.027). When Hcy was greater than 15.218 μmol/L, the risk of stroke increased with rising Hcy levels. Subgroup analysis showed a significant interaction effect between different CRP subgroups (P_interaction =0.027); in the low CRP group (Q1, 0-0.3 mg/L), the association between Hcy and stroke was linear (P_{overall association}<0.001, P_{non-linearity}=0.251).

This study confirms the non-linear association between Hcy levels and the risk of incident stroke, and that this association is modified by CRP levels. These findings suggest that community-based chronic disease health management should prioritize comprehensive multi-factorial intervention, with special attention to the management of elderly individuals, those with chronic disease histories, and those with obesity. Enhanced screening for inflammatory factors and close monitoring of key indicators, including blood pressure, blood glucose, blood lipids, CRP, and Hcy, are recommended. Hcy can serve as a predictor of stroke risk, but it should be combined with inflammatory markers such as CRP for comprehensive assessment.