Factorial invariance of the Centre for Epidemiological Studies Depression Scale (CES-D)

This foundational research concerns the measurement properties of the Centre for Epidemiological Studies Depression Scale (CES-D) 10-item short form (CES-D-10). Understanding measurement properties of the CESD-10 is crucial for any future work using CLSA CES-D-10 data. Existing data suggest the CES-D-10 measures two distinct factors related to depressive symptoms: depressed affect and lack of positive affect. Research with the longer form of the CES-D, from which the CES-D-10 has been derived, suggest the four factor structure seen in the longer version varies depending on demographic variables, which has implications for research with the CES-D. Fundamental to understanding all future research using the CLSA CES-D is the factor structure of the CES-D-10 for all participants. The purpose of the present methodological study is to explore the psychometric properties of the CES-D-10; particularly whether the two factor structure of the CES-D-10 is invariant to language, age, gender, ethnic background, or cognitive status.

We found that the Center for Epidemiologic Studies Depression Scale (CES-D) is invariant to sex, age, education, ethnic background, and more recently cognitive impairment. The two-factor structure of depressed mood and lack of positive affect was consistent with theory in this area and with some prior published works that also found a two-factor structure. However, the two-factor structure was not in line with other work suggesting a three- or single-factor for a short-form of this depression scale. The final analysis will include cognitive impairment indicator(s). The final analysis is important because self-report of depression even in the presence of cognitive impairment appears to be similar to how depressive symptoms are reported for cognitively intact people.