Psychological Vulnerability Triples Stroke Risk in Middle-Aged and Older Adults
A large Chinese cohort study finds that higher psychological frailty scores strongly and linearly predict incident stroke, independent of traditional risk factors.
Summary
Using data from 15,284 adults aged 45+ in the CHARLS cohort, researchers found that psychological vulnerability — measured by the Psychological Frailty Index (PFI) — was strongly associated with new-onset stroke over a two-year follow-up. Each interquartile range increase in PFI raised stroke risk by 87%, and those in the highest PFI quartile had more than three times the stroke risk of those in the lowest quartile. The dose-response relationship was linear. The association was strongest in older adults and those without a partner, suggesting that age and social support moderate the psychological frailty–stroke link. These findings point to mental health screening as a potentially valuable component of stroke prevention.
Detailed Summary
Stroke remains one of the leading causes of death and disability worldwide, and while traditional cardiovascular risk factors are well-established, psychosocial contributors are increasingly recognized. This study investigated whether psychological vulnerability — quantified using the Psychological Frailty Index (PFI) — independently predicts incident stroke in a large, nationally representative Chinese cohort.
Researchers used Wave 4 (2018) data from the China Health and Retirement Longitudinal Study (CHARLS) as baseline, following 15,284 adults aged 45 and older through Wave 5 (2020). The PFI is a composite score derived from four equally-weighted subdimensions: depressive symptoms (8-item simplified CES-D), subjective cognitive complaints, coping ability, and emotional instability. Each subscale was Z-score normalized and summed to yield the PFI, which was then standardized by its interquartile range. Incident stroke was self-reported at the two-year follow-up. Cox proportional hazards regression, restricted cubic spline (RCS) analysis, and subgroup analyses with Bonferroni correction were employed.
During follow-up, 323 of 15,284 participants (2.11%) experienced a first stroke. In fully adjusted Cox models, each one-IQR increase in PFI was associated with an 87% higher stroke risk (HR = 1.87, 95% CI 1.54–2.27; P < 0.001). Compared to the lowest PFI quartile (Q1), individuals in the highest quartile (Q4) had more than three times the stroke risk (HR = 3.12, 95% CI 1.99–4.91; P < 0.001). RCS analysis confirmed a significant, strictly linear dose-response relationship between PFI and stroke risk, with no evidence of nonlinearity (P for nonlinearity > 0.05). Covariates adjusted for included age, sex, residence, marital status, education, smoking, alcohol use, hypertension, diabetes, dyslipidemia, and heart disease.
Subgroup analyses revealed that the PFI–stroke association was significantly modified by age and marital status (P for interaction < 0.05), with stronger associations observed among older individuals and those who were unmarried. The findings suggest that social isolation and aging amplify the cerebrovascular consequences of psychological vulnerability. Proposed biological mechanisms include chronic low-grade inflammation (elevated IL-6 and CRP), neuroimmune disruption of the neurovascular unit, and behavioral pathways such as sedentary lifestyle, poor diet, smoking, and suboptimal management of hypertension and diabetes.
This study represents one of the first large-scale investigations linking a composite psychological frailty measure to incident stroke in a low- or middle-income country context. Key limitations include reliance on self-reported stroke diagnosis without imaging confirmation, a short two-year follow-up, and the observational design precluding causal inference. Nevertheless, the results strongly support integrating psychological vulnerability assessment into comprehensive stroke prevention frameworks, especially for elderly and socially vulnerable populations.
Key Findings
- Each IQR increase in PFI was associated with 87% higher stroke risk (HR = 1.87, 95% CI 1.54–2.27).
- Highest PFI quartile (Q4) carried more than triple the stroke risk versus lowest quartile (HR = 3.12).
- The PFI–stroke dose-response relationship was strictly linear with no evidence of a threshold effect.
- Association was significantly stronger in older adults and unmarried individuals (P for interaction < 0.05).
- Findings held after adjusting for hypertension, diabetes, dyslipidemia, heart disease, and lifestyle factors.
Methodology
Prospective cohort study using CHARLS Wave 4 (2018) as baseline and Wave 5 (2020) for outcome ascertainment in 15,284 adults aged ≥45 years. Cox proportional hazards regression and restricted cubic spline analysis assessed PFI–stroke associations; Bonferroni-corrected subgroup analyses evaluated effect modification.
Study Limitations
Stroke was ascertained by self-report without clinical or imaging confirmation, introducing potential information bias. The two-year follow-up is relatively short, and the observational design prevents establishing causality between psychological frailty and stroke.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.