A multi-site study of alcohol subtypes: classification and overlap of unidimensional and multi-dimensional typologies

Aims The current study examined the usefulness of four prevailing alcohol typologies, assessed in terms of replicability of the multi-dimensional schemas, percentage of the sample classified by each typology, distribution of subtypes by gender and treatment site and construct validity of Type 1/Type 2. In addition, overlap of classification systems was examined to determine whether the four typologies may be narrowed to a smaller set of meaningful, non-redundant subtype schemas. Design Baseline data from five treatment outcome studies were used to facilitate subtyping according to four alcohol typologies: antisocial (ASP) versus non-ASP, early versus late onset, Type 1/Type 2 and Type A/Type B. Setting The studies were conducted at several in-patient and out-patient treatment sites. Participants The sample included 342 participants (23% female) who met DSM-III-R criteria for alcohol abuse or dependence. Measurements Subjects were assessed on substance use severity, family history, psychopathology, personality and psychosocial functioning. Findings Type 1/Type 2 and Type A/B typologies were replicated. Type 1/Type 2 had poor construct validity due to symptom cluster overlap between the two subgroups. Only 14% of the subjects met criteria for ASP The two strongest associations including that between early versus late onset and Type 1/Type 2 and that between ASP versus non-ASP and Type A/Type 13, were expected and reflect overlaps in conceptual definitions. In comparison, the relationship between early versus late onset and ASP versus non-ASP and Type A/B was weak. Conclusions Clinical utility of the ASP/non-ASP typology is limited in non-Veterans Administration (VA) samples, due to low prevalence. The Type 1/Type 2 subtyping schema was redundant with the age of onset schema, and was the least internally valid of all four typologies. In general, the Type A/B schema was most promising of prevailing typologics studied. It was relatively inclusive, and the A, B groups were distinct from one another. However, dichotomous typologies may not be complex enough to be clinically useful descriptors of alcoholic samples. Aside from ASP and Type B, there appears to be heterogeneity within groups typically considered homogeneous, such as 'early versus late onset' alcoholics.