The mediating effect of depression on new-onset stroke in diabetic population: Evidence from the China health and retirement longitudinal study.

Background: Diabetes has a high incidence in China, which may cause stroke and depression. However, the relationship between diabetes and the incidence of new-onset stroke and depression has not been fully studied.

Methods: The data from the China Longitudinal Study on Health and Retirement (CHARLS) from 2013 to 2018 were used. A total of 8530 respondents aged ≥45 years old were included in the follow-up study. Logistic regression model, Cox regression, and Mediation analyses were used to explore the association between diabetes, depression, and new-onset stroke.

Results: The depression score of patients with diabetes history was higher (HR,95%CI = 1.02, 1.01-1.04) and were more likely to experience new-onset stroke events (HR, 95%CI = 1.046, 1.02-1.07). With a history of hypertension (HR,95%CI = 1.747, 1.381-2.208), older (HR,95%CI = 1.033, 1.020-1.046) with high BMI (HR,95%CI = 1.056, 1.027-1.086) have a high risk of new-onset stroke. In the combined subgroup analysis, the incidence of new-onset stroke in the subgroup with diabetes depression was higher than in others. The mediating effect of depression on new-onset stroke events in diabetic patients is more pronounced in the medium to long term (>3 years) after adjusting covariates.

Limitations: We defined new-onset stroke by patient self-report, there might be some memory bias. In addition, new-onset stroke was not classified in the CHARLS questionnaire, which would hinder us to evaluate the mediating effect of depression on different types of new-onset stroke.

Conclusion: Our results showed that depression has a partial mediating effect between diabetes and new-onset stroke in the middle-aged and elderly population in China.