The indirect effects of cognitive function in the association between stroke and quality of life among older adults in a rural community: A cross-sectional study

Purpose We examined whether cognitive function statistically accounted for associations between stroke and quality of life (QOL). Methods A cross-sectional survey included adults aged ≥ 60 years in Segamat, Malaysia. Cognition was assessed using the Identification and Intervention for Dementia in Elderly Africans instrument, and QOL was assessed using the WHOQOL-BREF. Stroke survivors were matched 1:3 with non-stroke participants by propensity scores based on age, sex, and education. Stroke status was the exposure, cognition the proposed statistical mediator, and four QOL domains, overall QOL, and overall health the outcomes. Models were adjusted for sociodemographic and vascular factors, accounted for matching and multiple testing, and p < 0.05 was considered statistically significant. Results Among 596 participants, 149 reported stroke; the mean age was 70.36 years (SD 6.81); 41.6% female; 58.4% male. An indirect association between stroke and physical QOL via cognition was observed (β = −0.67, 95% CI − 1.30 to − 0.18; BH-adjusted p = 0.018). A residual correlation of ρ = 0.16 between cognition and physical-QOL models would attenuate this association to the null. The indirect association did not differ by sex (index β = −0.49, 95% bootstrap CI − 1.48 to 0.18; p = 0.164). Item-level physical-QOL analysis identified an indirect association with greater pain interference through lower cognition (β = −0.051, 95% CI − 0.113 to − 0.009; BH-adjusted p = 0.028). Conclusion Reduced cognition partly accounted for associations of stroke with reduced physical QOL and greater pain interference. Findings were sensitive to modest unmeasured confounding but require confirmation from longitudinal studies.