Social Stratification of Psycho-Physical Health Outcomes Among Older Adults in China: Evidence from China Family Panel Studies

Objective: This study aimed to examine whether socioeconomic circumstances (education attainment and pension level) are associated with physical impairment and depressive symptoms among elderly Chinese population. 

Methods: Data were obtained from the 2014 national survey of the China Family Panel Studies (CFPS). The sample included 8757 older adults (4472 women versus 4285 men) aged 60 years or older across China. Participants who reported needing help with at least one out of seven activities of daily living (ADLs) were flagged as having functional limitation. A score of 12 on the 6-item cross-culturally adapted versions of Center for Epidemiological Studies Depression Scale (CES-D) was used as the cutoff for depressed versus not depressed. Education was categorized into five levels: illiterate/semi-literate, primary school, junior high school, senior high school/secondary school/technical school/vocational senior school, and 3-year college or above. Monthly retirement income in Chinese yuan (RMB) was operationalized into four categories: none, low (1 to 500), medium (501 to 2000), and high (higher than 2000). Respondents’ demographic characteristics (age and marital status) and health status (chronic conditions and medical consultation) were covariates. Four age categories were created (60 to 69, 70 to 79, 80 to 89, and 90 years or older). Binary logistic regressions were conducted for men and women separately and the results were compared.

Results: Educational achievement was a robust indicator of functional limitations for both genders in this representative sample of older Chinese, even after adjustments were made for age, marital status, chronic conditions, and utilization of health care services. The odds of having functional limitations for participants who were in the lowest educational group rose more than five times for women (OR=5.288, 95% CI=1.221 22.894, p = 0.026) and more than threefold for men (OR=3.788, 95% CI=1.581 9.077, p = 0.003) compared to participants in the highest educational group. Among older women, pension income was negatively associated with functional limitation in the unadjusted model, but the inclusion of education fully attenuated the relationship.

Regarding mental health, education level was associated with depression in men in the adjusted models, but not women. Pension income was negatively associated with depression for both genders, in both the crude and adjusted models. In the adjusted logistic regression analyses, the odds of being depressed among older adults with no pension income were 83.5% higher for women (OR=1.835, 95% CI=1.173 2.871, p = 0.008) and 67.9% higher for men (OR=1.679,95% CI=1.143 2.465, p = 0.008) than their respective peers in the highest pension bracket.

Conclusions: This nationally representative study revealed that the prevalence and odds of both functional limitation and depression were inversely related to social class across the full spectrum of the socioeconomic gradient. The association was stronger for the educational gradient with physical impairment, whereas the association with depression was stronger for the income gradient. Important underlying gender differences were observed for two relationships: the association of old age income with functional limitations for older women and the association of education with mental health for older men.