An item response analysis of the DSM-IV criteria for major depression: findings from the Australian National Survey of Mental Health and Wellbeing

Abstract

Background: This study examines the psychometric properties and presence of gender bias in the major depression criteria using data from the Australian general population.

Methods: Data came from a subsample of respondents from the 1997 National Survey of Mental Health and Wellbeing (NSMHWB; n=2061). A two-parameter logistic model was employed to yield severity and discrimination parameters, and the IRT log-likelihood-ratio test for differential item functioning (IRTLRDIF) procedure was utilized to evaluate gender bias.

Results: DIF analyses indicated that the psychomotor difficulties criterion was endorsed at lower levels of severity by males than females. In general, the criteria were arrayed along a continuum of depression severity. Discrimination was greatest for concentration difficulties/indecision and lowest for death/suicidal thoughts and worthlessness/guilt. Worthlessness/guilt, psychomotor difficulties, and death/suicidal thoughts tapped the severe end of the depression continuum, whereas concentration difficulties/indecision and sleep disturbance tapped the mild range.

Limitations: The inclusion of stem questions precluded examination of two core symptoms of depression (depressed mood and loss of interest).

Conclusions: Collectively, the criteria performed well in defining a latent continuum of major depression. Few gender differences were observed, with the exception of the psychomotor difficulties criterion. Quantitative and qualitative analyses collectively hold promise of providing a scientifically rigorous basis for empirically-based modifications to the psychiatric classification system.