Predicting depression among rural and urban disabled elderly in China using a random forest classifier

Area of Science:

• Gerontology
• Public Health
• Mental Health Research

Background:

• Increasing global aging populations are accompanied by a rise in elderly individuals with physical disabilities.
• Disabled elderly, particularly those who lose independence, exhibit higher rates of depression symptoms.
• Early recognition and management of depression can potentially mitigate the disability progression in this demographic.

Purpose of the Study:

• To investigate the predictive influence of demographic, health, and social factors on depression among rural and urban disabled elderly.
• To enhance early recognition of depression symptoms in disabled older adults.
• To differentiate predictive factors for depression between rural and urban disabled elderly populations.

Main Methods:

• Utilized data from 1460 disabled adults aged 60+ from the China Family Panel Studies (CFPS).
• Employed the random forest classifier to predict depression based on demographic characteristics, health status, health behavior, family, and social relationships.
• Assessed depression using The Center for Epidemiologic Studies Depression Scale (CES-D); model trained on 70% and tested on 30% of the data.

Main Results:

• Depression rates were 57.67% in rural and 44.59% in urban disabled elderly.
• Random forest models achieved predictive performance with AUCs of 0.71 (rural) and 0.78 (urban).
• Common predictors included self-rated health, perceived health changes, recent illness, life satisfaction, trust, BMI, and future outlook; distinct predictors were found for rural and urban groups.

Conclusions:

• Random forest modeling shows promise for the early detection of depression in disabled elderly populations.
• Identifying unique and shared predictors can inform targeted interventions for depression in rural versus urban disabled older adults.
• Addressing factors like self-rated health, social trust, and life satisfaction is crucial for mitigating depression in this vulnerable group.