Evaluating the seven-item Center for Epidemiologic Studies depression scale short-form: a longitudinal U.S. community study.

PURPOSE: The current study aims to examine the potential use of the seven-item Center for Epidemiologic Studies depression scale (CES-D) short form (CES-D-SF).
METHODS: Data were examined from the National Longitudinal Survey of Youth 1979. Participants responded to the 20-item CES-D (n = 8,858) in 1992, and to the 7-item CES-D-SF in 1994 (n = 8,500) and from 1998 to 2010 if aged 40 (n = 7,972) or 50 (n = 1,574) or over. Variables examined in 1979 were race, SES, and sex and in 1981 cognitive functioning. The CES-D-SF was examined for internal and test-retest reliability, unidimensionality with confirmatory factor analysis, and a cutoff score with receiver operator curve characteristics. Survival analysis was used to examine time period of first CES-D-SF suspected major depression episode, multinomial regression to examine the chronicity of CES-D-SF suspected major depression, and the course of depression with a Generalized Estimating Equation model.
RESULTS: Compared to the CES-D, the CES-D-SF had higher internal consistency, and better unidimensionality based on confirmatory factor analysis. A CES-D-SF cutoff score ≥8 had acceptable specificity (0.97, 95% CI 0.96, 0.97) and modest sensitivity (0.69, 95% CI 0.67, 0.71) with the standard CES-D cutoff score of 16. Female sex and lower cognitive functioning were significantly (p < 0.05) associated with more CES-D-SF suspected depression that was more chronic based on a multinomial regression model, and occurred at a younger age based on a Cox regression model.
CONCLUSIONS: The seven-item CES-D-SF has acceptable psychometric properties, is associated with exposures documented to be associated with an increased likelihood of depression, and may be used to screen for suspected major depressive disorder in US community studies.