The tripartite model of anxiety and depression: A new look at theory and measurement (original) (raw)

The relationship between anxiety and depression remains unclear, despite ongoing revision of diagnostic concepts over the last sixty years. Various proposals have been suggested which attempt to explain this relationship. Some have suggested that anxiety and depression are distinct disorders, with different etiologies, symptoms, natural courses, and response to treatment (the pluralist position). Others have suggested that anxiety and depression are a single disorder, or at least different manifestations of the same underlying pathology (the unitary position). Clark & Watson (1991) proposed an alternative model in which anxiety and depression symptoms are divided into three clusters-a specific set of symptoms for each, as well as a shared set of non-specific symptoms seen in both disorders. They suggested that such a tripartite model of anxiety and depression symptoms provided the best description of the manifestation of anxiety and depression symptoms. Clark & Watson (1991) further suggested that by focusing on the specific symptoms of anxiety and depression, the discriminant validity of psychometric assessments of the two could be improved. The current research reviews the literature in support of unitary, pluralist and tripartite models of anxiety and depression symptoms, and concludes that a tripartite model is the most consistently supported by the literature, including studies of symptoms, course, treatment, etiology, family patterns and genetics in both. It examines in detail the research into the tripartite structure of anxiety and depression symptoms, and concludes that previous research also supports a three-factor structure of anxiety and depression symptoms in existing self-report measures. The Mood and Anxiety Symptom Questionnaire (MASQ) is reviewed as the only existing clinical symptom measure based on the tripartite model. Although previous authors have reported broad support for the MASQ, the present research uniquely used a sample of anxious and depressed patients and confirmatory factor analytic methods, and failed to support the MASQ as a valid tripartite measure. Specifically, there was no support at the item or subscale level for any of the two-factor or three-factor models tested. Furthermore, the MASQ subscales did not show the expected differences when compared between anxious and depression patients, and the instrument was found to be inadequate in discriminating between those diagnosed with anxiety versus those diagnosed with depression. In response to the psychometric inadequacy of the MASQ, the Tripartite Clinical Symptom Inventory (TCSI) was proposed as a measure based on the tripartite model which would be clinically useful, as well as showing strong reliability and validity. It was also proposed that the TCSI would further validate several aspects of the tripartite theory of anxiety and depression symptoms. The original 45-item set was reduced over a series of exploratory factor analyses, designed to yield a core subset of symptoms that were reliably and stably associated with