Background. Patterns of co-occurrence among the common mental disorders may provide information about underlying dimensions of psychopathology. The aim of the current study was to determine which of four models best fits the pattern of co-occurrence between 10 common DSM-IV and 11 common ICD-10 mental disorders.

Method. Data were from the Australian National Survey of Mental Health and Well-Being (NSMHWB), a large-scale community epidemiological survey of mental disorders. Participants consisted of a random population-based sample of 10641 community volunteers, representing a response rate of 78%. DSM-IV and ICD-10 mental disorder diagnoses were obtained using the Composite International Diagnostic Interview (CIDI), version 2.0. Confirmatory factor analysis (CFA) was used to assess the relative fit of competing models.

Results. A hierarchical three-factor variation of a two-factor model demonstrated the best fit to the correlations among the mental disorders. This model included a distress factor with high loadings on major depression, dysthymia, generalized anxiety disorder (GAD), post-traumatic stress disorder (PTSD) and neurasthenia (ICD-10 only); a fear factor with high loadings on social phobia, panic disorder, agoraphobia and obsessive–compulsive disorder (OCD); and an externalizing factor with high loadings on alcohol and drug dependence. The distress and fear factors were best conceptualized as subfactors of a higher order internalizing factor.

Conclusions. A greater focus on underlying dimensions of distress, fear and externalization is warranted.