Factorial Validation of the Short Form Center for Epidemiologic Studies Depression (CES-D) Scale Among Community-Dwelling Older Adults in China

Background and Purpose. The prevalence rate of depressive symptoms among elders increases with the population aging. With the rapidly growing older population in China, there is an increasing need for a valid and reliable instrument for screening depression among Chinese older adults population. The objective of this study is to explore the factorial structure of a short form 10-item Center for Epidemiologic Studies Depression Scale (CES-D 10) among community-dwelling older adults from a national representative survey and examines the psychometric properties of CES-D 10, including reliability, factorial structure among male and female group, criterion validity, and prevalence rates of depressive symptoms in China.

Methods. The secondary data used in this study were drawn from the baseline (2011-2012) of the China Health and Retirement Longitudinal Survey (CHARLS) of older adults aged 60 and above (N = 7,669) in 447 communities. A 10% of the subsample (n = 763) was randomly selected from the total sample to avoid type I error. Confirmatory factor analysis with LISREL 8.80 was used to test the factorial structure of CES-D 10 through three competing models (one, two, and three-factor model) and determine the best fit model in the Chinese sample. Factorial structure by gender, reliability test, and criterion validity were also examined in this study.

Results. Confirmatory factor analysis supported the two-factor structure of CES-D 10 in Chinese older adult sample through model comparisons; the model comparison indicated that the two-factor model showed the best model fit (χ²₍₃₄₎ = 76.46, p < .001, GFI = .98, CFI = .99, RMSEA = .044, SRMR = .030, AIC = 118.46). Two factors (depressive affect and positive affect) were validated in this study, and the scale had satisfactory internal reliability (Cronbach’s α = .819). In addition, the two-factor structure was supported in both males and females; factor loading and model fit were similar among males and females. The CES-D 10 also demonstrated good criterion validity with life satisfaction (r = −.362, p < .01), self-rate health (r = −.426, p < .01), and pain (r = .509, p < .01), which corresponds with findings in previous studies. Lastly, the prevalence of depressive symptoms differed between males (22.19%) and females (27.90%), which is similar to what has been documented in previous research.

Discussion. The CES-D 10 demonstrated satisfactory reliability and validity among community-dwelling older adults in China. A two-factor structure corresponds with findings established in Western societies. It also showed the same factorial structure among females and males. Results from this study provide evidence for future cross-culture comparisons and gender comparison in China.