A psychometric assessment of the factor structure and reliability of the Center for Epidemiological Studies Depression (CES-D) questionnaire in four African nations

BACKGROUND: The Center for Epidemiological Studies Depression (CES-D) is commonly used to identify people at risk for clinical depression in HIV care and research settings in Africa, though it has not been validated in these settings. We examined the psychometric properties of the CES-D in people living with (PLWH) and without HIV (PLWoH) in four African countries. METHODS: The African Cohort Study (AFRICOS) administers the CES-D biannually to participants aged 15 years or older in Kenya, Nigeria, Tanzania, and Uganda. The CES-D is completed in English, Kiswahili, or Luganda. Exploratory (EFA) and confirmatory factor analysis (CFA) were each performed in separate subsamples. Longitudinal measurement invariance and invariance by nation, language, sex, and HIV status were examined. CES-D score distributions, internal consistency, and test-retest reliability were reported. RESULTS: There were 3320 participants: 1663 (50%) exploratory sample, 1397 (42%) males, and 2732 (82%) PLWH. EFA supported a 1- or 2-factor solution. CFA confirmed 1-factor solution with a method factor for reverse-scored items. Configural, metric, and either scalar or partial scalar invariance were established for all variables examined. Collapsing across samples, the median CES-D score was 3; internal consistency was α = 0.89; and 6- and 12-month test-retest reliabilities were ρ = 0.37 and 0.33, respectively. CONCLUSIONS: These results have positive implications for using the CES-D in African nations and PLWH. That the four reverse-scored items are not a second factor, but covary due to methodological issues, has implications that transcend specific samples. Further research is needed to establish validity in African nations and PLWH. CLINICAL TRIAL NUMBER: Not applicable.