Scott A. Baldwin and Dr. Diane L. Spangler, Psychology
Depression and anxiety are often comorbid and several theories have been proposed to account for their co-occurrence. The tripartite model accounts for the comorbidity between anxiety and depression in terms of the underlying symptoms that are associated with the two disorders. According to the tripartite model, low positive affect (PA) is unique to depression, while autonomic arousal (AA) is unique to anxiety. Negative affect (NA) is common to both disorders; therefore, the relationship between the disorders in due to a common set of symptoms—NA.
In contrast to the tripartite model, Beck proposed the cognitive-specificity model of anxiety and depression. In this model, depression and anxiety share some common cognitive components resulting in their comorbidity. However, depression and anxiety also have unique cognitive components. Depression is associated with thoughts of loss, while anxiety is associated with fearful thoughts of the future.
Both the tripartite and cognitive-specificity models have been researched individually and have received much support. But both models have been shown to have weaknesses. To remedy the weaknesses, some researchers have suggested that the combination of the tripartite and cognitive models of anxiety and depression would improve prediction and differentiation of these emotional states. Indeed, one study, which combined the affective and cognitive models, found that the combination did improve prediction of anxiety and depression. In that study, however, AA was not included and the data were cross-sectional, rather than longitudinal. The purpose of the current study was to test whether a combination of the tripartite and cognitive models of anxiety and depression would predict the two disorders over time better than either model alone. In this study AA was controlled for and the data were longitudinal, both of which make the study statistically stronger.
Three hundred and fifty-three undergraduates received measures of anxiety, depression, cognition, PA, NA, and AA at two time points. Longitudinal regression models were used to analyze the data. In the models, anxious cognitions, depressive cognitions, NA, PA, and AA were the predictor variables, while anxiety or depression was the criterion variable. Depressive cognitions, NA, PA, and AA all predicted depression while anxious cognitions did not. Additionally, AA and NA predicted anxiety while anxious cognitions, depressive cognitions, and PA did not. Overall, for depression, the data supported the cognitive-specificity model, but not the tripartite model, because AA significantly predicted depression. In contrast, for anxiety, the data supported the tripartite model, but not the cognitive-specificity model because while depressive cognitions did not predict anxiety, neither did anxious cognitions. The results suggest that a combination of the tripartite and cognitive models does not improve prediction of anxiety and depression over either model alone.