Evaluating dimensional models of psychopathology in outpatients diagnosed with emotional disorders: A cautionary tale

Mental disorders cluster together systematically. Through factor analysis of disorder comorbidity, investigators are establishing the latent dimensions that underlie the development of related syndromes. However, these dimensions have not been validated across diverse patient samples, in which comorbidity patterns vary widely.

We assessed 4,928 outpatients seeking treatment for emotional disorders with a semistructured diagnostic interview. This was the largest patient sample as yet for an evaluation of the latent structure of mental disorders. We compared several competing dimensional models of common mental disorders via confirmatory factor analysis.

The hypothesized confirmatory factor model-anchored by internalizing and externalizing spectra-fit the diagnostic data poorly. Neither a one-factor model, reflecting a unitary liability to all mental disorders, nor a three-factor model, wherein the internalizing dimension bifurcated into distress and fear subfactors, fit appreciably better.

These data provide novel evidence that the internalizing and externalizing spectra are not structurally sound in all clinical contexts. We speculate about the causes of model misfit and advise additional research into the generalizability-with respect to sample, input data, developmental stage, and more-of dimensional models of mental disorder.