Assessing the Dimensionality of Lifetime DSM-IV Alcohol Use Disorders and a Quantity-Frequency Alcohol Use Criterion in the Australian Population: a Factor Mixture Modelling Approach

Abstract

With the revision of the DSM-IV underway, two important research issues currently dominate the addiction literature: (a) how can the dimensionality of DSM-IV alcohol use disorders (AUD) diagnostic criteria best be described? and (ii) should a quantity-frequency alcohol use (QF) criterion be added to the existing diagnostic criteria set in the DSM-V? The current study addressed these aims by analysing lifetime data from a recent Australian population survey. Data from adults screened for lifetime DSM-IV AUD in the 2007 National Survey on Mental Health and Wellbeing (NSMHWB) were analysed (n=5409). A series of alternative factor analytic, latent class and factor mixture or 'hybrid' models were used to assess the dimensionality of lifetime DSM-IV AUD diagnostic criteria and a lifetime QF criterion. Examination of the goodness-of-fit indices revealed that a one-factor or a two-factor model, a three-class latent class model or a two-factor zero-class hybrid model, were all acceptable models for the data. A simple structure one-factor model was considered to be the most parsimonious and theoretically meaningful model, given the high correlation between the abuse and dependence factors (0.874) in the two-factor model. The inclusion of the QF criterion did not enhance the fit of the one-factor model. Incorporating both dimensional and categorical conceptions of lifetime AUD did not provide substantial gains over a simple structure unidimensional model of AUD severity. The utility of a QF use criterion in helping to diagnose AUD is questionable. These findings should be of relevance to the DSM-5 substance use disorder workgroup.