Measuring depressive symptoms in illness populations: Psychometric properties of the Center for Epidemiologic Studies Depression (CES-D) Scale

Abstract

Psychometric properties of the Center for Epidemiologic Studies Depression (CES-D) scale were examined among five groups that vaned in physical health and illness. Participants included 175 healthy undergraduates, 176 individuals attending family physicians, 107 progressive renal disease, 135 end-stage renal disease, and 120 cancer patients. Individual item and total CES-D scores were relatively symmetrically distributed and varied across the entire range of potentially obtainable scores. Reliability analyses yielded internal consistency (alpha) coefficients ranging from 0.63 to 0.93 across the groups. Test-retest reliability (3-month lag) was 0.61. The CES-D's factorial composition was highly similar to that observed among community volunteers. Varimax-rotated principal-components analyses extracted four factors, corresponding to “depressive affect”, “positive affect”, “somatic and retarded activity”, and “interpersonal”. Moreover, this pattern did not change despite differences in physical health-illness across the groups. although a significant Groups X Items (repeated measures) interaction indicated that “positive affect” items were lower among progressive renal failure patients than among the other groups. The psychometric characteristics of the CES-D scale, thus, appear to be relatively constant across the five health-illness populations.