Understanding Discrepancies and Predictors of Self- versus Proxy-Rated Quality of Life in Chinese Community-Dwelling Older Adults with Mild Dementia: A Cross-Sectional Study.

BACKGROUND

Accurate evaluation of quality of life (QoL) is essential for optimal dementia management, yet notable discrepancies exist between patient self-rated and proxy-rated QoL. This study aimed to analyse the discrepancies and agreement between self-rated and proxy-rated QoL and identify their independent predictors among community-dwelling persons with mild dementia (PwMD).

METHODS

This cross-sectional study included 129 PwMD and their primary caregivers. Assessments included sociodemographic information, the QoL-Alzheimer's Disease scale, Mini-Mental State Examination, Geriatric Depression Scale, Activities of Daily Living, Functional Activities Questionnaire and Neuropsychiatric Inventory Questionnaire. Differences and agreement were analyzed using Wilcoxon tests, Spearman correlations, and intraclass correlation coefficients (ICC). Independent predictors were identified through multiple linear regression.

RESULTS

Patients rated their QoL significantly higher than their caregivers did, with poor-to-fair agreement at the total, dimensional and item levels (ICC = 0.11-0.50). Higher depressive symptoms (β = -0.354, p = 0.001), lower severity of elation (β = 0.214, p = 0.013), and greater disinhibition (β = -0.174, p = 0.041) independently predicted poorer self-rated QoL. By contrast, greater neuropsychiatric symptom severity (β = -0.283, p < 0.001), poorer Activities of Daily Living function (β = -0.230, p = 0.006), and polypharmacy (β = -0.208, p = 0.009) predicted lower proxy-rated QoL.

CONCLUSION

Self- and proxy-rated QoL reflect distinct evaluative perspectives in PwMD. Self-reports are influenced by emotional and psychological states, whereas proxy ratings are shaped by observable symptoms, functional dependency and treatment burden. These findings suggest the need to integrate both assessment perspectives in clinical practice.