Given that there is limited research examining the specific role social isolation plays in stroke risk, we aimed to estimate the controlled direct effect of social isolation on stroke risk in China. A nationally representative sample (n = 12,662) of persons aged 45 years or more at baseline (2011), with corresponding follow-ups in 2013 and 2015, was taken from the China Health and Retirement Longitudinal Study. Stroke was assessed through a self- or proxy-reported physician's diagnosis. Social isolation was measured by incorporating marital status, frequency of contact with friends, family, and children, and participation in social activities. A marginal structural model with stabilized inverse probability weights was applied to examine the controlled direct effect of social isolation on stroke risk. Overall, 245 persons had a stroke within the 4-year timeline under study. The total effect model indicated that persons experiencing social isolation had a 64% increased risk of stroke (odds ratio (OR) = 1.64, 95% confidence interval (CI): 1.26, 2.13). Results from the marginal structural model also indicated that socially isolated persons had an increased risk of stroke (OR = 2.39, 95% CI: 1.49, 3.82) after adjustment for depression. Interventions to reduce social isolation may be particularly beneficial in preventing the occurrence of stroke among middle-aged and older adults in China.
Keywords: China; depression; marginal structural models; social isolation; stroke.
© The Author(s) 2019. Published by Oxford University Press on behalf of the Johns Hopkins Bloomberg School of Public Health. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.