Blood pressure control transitions and subsequent reported stroke among Chinese adults with hypertension in a landmark cohort study
Year: 2026
Type: preprint
Abstract: Background Blood pressure measured at successive visits may identify different subsequent stroke risks, but changes in measured control should be distinguished from treatment effects. We examined blood pressure control transitions from 2011 to 2013 in the China Health and Retirement Longitudinal Study (CHARLS). Methods This landmark analysis included adults aged at least 45 years with baseline hypertension, valid blood pressure measurements at both visits, and no reported stroke before the landmark. Control was defined as systolic pressure below 140 mmHg and diastolic pressure below 90 mmHg. Subsequent first reported physician-diagnosed stroke through the 2020 survey was analysed with an interval-censored Weibull model and community-clustered uncertainty. The primary comparison was uncontrolled-to-controlled versus uncontrolled-to-uncontrolled pressure among initially uncontrolled participants. Covariates included baseline demographic and clinical characteristics and continuous blood pressure. Results The complete-case stroke sample comprised 2902 participants with 256 events. Among 2126 initially uncontrolled participants, 190 events occurred. In the primary Weibull model, transition to control was associated with a lower modelled stroke hazard than uncontrolled pressure at both visits (hazard ratio [HR], 0.68; 95% confidence interval [CI], 0.49–0.95). Multiple imputation gave an HR of 0.70 (95% CI, 0.51–0.97). A semiparametric interval-censored model gave an HR of 0.69 (community-bootstrap 95% interval, 0.46–0.96). Several exposure-definition analyses could not exclude no association. A lognormal model had a lower Akaike information criterion than the Weibull model and yielded an uncertain time ratio of 1.22 (95% CI, 0.97–1.53). Conclusions In the primary Weibull model, transition to measured control was associated with a lower modelled hazard of subsequently reported stroke. The evidence depended on distributional assumptions and exposure definitions. Death-related underascertainment, measurement error, and selection limit interpretation; these estimates do not establish a treatment effect. Clinical trial registration Clinical trial number: not applicable. This study was an observational secondary analysis of existing CHARLS cohort data and did not involve the assignment of any healthcare intervention to participants.
Source: Research Square
Language: English
Subfield: Cardiology and Cardiovascular Medicine
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W7225046659