Social determinants and population attributable fractions for risk factors of cognitive frailty among community-dwelling older adults in Shanghai, China (Preprint)

BACKGROUND Cognitive frailty (CF), defined as the co-occurrence of physical frailty and cognitive impairment without dementia, is a promising target for dementia prevention. However, the relative contributions of structural and intermediary social determinants and the burden attributable to modifiable factors remain unclear in Chinese communities. OBJECTIVE This study aimed to identify social determinants of CF, quantify the contributions of different determinant layers, and prioritize modifiable factors for community-based prevention. METHODS This multicenter cross-sectional study analyzed baseline data from 13,170 community-dwelling adults aged ≥60 years recruited from 14 primary medical institutions in Shanghai between May 2024 and November 2025. CF was operationalized as the co-occurrence of physical pre-frailty or frailty, assessed using a modified Fried phenotype, and cognitive impairment, assessed using the Mini-Mental State Examination, in individuals without dementia. Guided by the World Health Organization Commission on Social Determinants of Health framework, candidate factors were classified into structural and intermediary blocks. Hierarchical multivariable logistic regression examined independent associations, while dominance analysis to quantify each block’s relative contribution. Model discrimination was assessed using the area under the receiver operating characteristic curve (AUC). Population attributable fractions (PAFs) for potentially modifiable factors were estimated from exposure prevalence and adjusted prevalence ratios derived using modified Poisson regression, with communality weighting to account for shared variance among factors. RESULTS Overall, 1451 participants (11.0%) met the criteria for CF. Older age, female sex, and lower educational attainment were independently associated with higher odds of CF. After full adjustment, more frequent social, leisure, and housework activities, greater dietary diversity, and higher daily water intake were associated with lower odds of CF. Conversely, smoking, tea drinking, non-light dietary taste, poor sleep quality, dyslipidemia, receipt of drug or surgical treatment, and anxiety symptoms were associated with higher odds. Intermediary and structural determinants accounted for 53.1% and 46.9% of the explained variance, respectively. Adding intermediary determinants improved model discrimination from an AUC of 0.699 to 0.765. Collectively, 30.9% (95% CI 17.3%-42.1%) of CF cases were potentially attributable to the 10 modifiable factors. After communality weighting, infrequent social activity remained the leading contributor (10.2%), followed by infrequent leisure activity (4.2%), poor sleep quality (3.4%), and no housework (3.0%). CONCLUSIONS CF is common and socially patterned among older adults in Shanghai. Structural determinants can guide equity-oriented screening toward older, female, and less-educated populations, whereas intermediary determinants identify actionable targets for multidomain community interventions, particularly social participation, leisure activity, sleep, housework, hydration, and diet. Given the cross-sectional design, the PAF estimates represent theoretical preventable burdens rather than confirmed causal effects and require longitudinal and interventional validation. CLINICALTRIAL The study protocol was registered with the Chinese Clinical Trial Registry (registration number: ChiCTR2400085431).