WHY ARE ANXIETY AND DEPRESSIVE SYMPTOMS COMORBID IN YOUTH? A MULTI-WAVE, LONGITUDINAL EXAMINATION OF COMPETING ETIOLOGICAL MODELS By

Comorbidity, defined as the manifestation of multiple disorders within an individual, has become the rule, rather than the exception. In youth, the co-occurrence of depression and anxiety is not only common, but leads to a more severe course of mental illness and poorer treatment response. Thus, researchers have aimed to understand the development of this frequent, and deleterious, relation. A collection of research points to two developmental explanations for why depression and anxiety may relate: A causal model, in which anxiety predicts depression, and a correlated liabilities model, where shared vulnerability factors predict both anxiety and depression. While promising trends have been identified for both models, a consistent pattern for the development of comorbid symptoms has yet to emerge. The present study sought to clarify past research by introducing a diathesis-anxiety approach to understanding comorbidity in youth. Specifically, we predicted that specific cognitive vulnerabilities would interact with anxiety symptoms to predict prospective depressive symptoms. For this study, 678 3 rd (n=208), 6 th (n=245), and 9 th (n=225) grade girls (n=380) and boys (n=298) completed iii self-report measures at baseline assessing cognitive vulnerabilities (rumination, negative inferential styles, and self-criticism), stressors, depression, and anxiety. Every 3 months over the next 18 months, youth completed follow-up measures of depression, anxiety, and stressors. Findings supported a diathesis-anxiety approach for self-criticism (t(2494) = 3.36, p < .001) and rumination (t(2505) = 2.40, p = .05). On the other hand, partial support was found for a correlated liabilities model, as girls with a negative inferential style were more likely to experience both depressive (t(2518) = 2.66, p = 0.008) and anxiety (t(1436) = 2.08, p = 0.03) symptoms following a negative event. Finally, no support was found for a causal model, as anxiety symptoms did not uniquely predict depressive symptoms (p > .05). These results clarify past results concerning comorbidity by suggesting that multiple pathways exist for comorbid depressive and anxiety states, and by introducing diathesis-anxiety models as an important and novel explanation for comorbid emotional distress in youth. Clinical implications and future developmental psychopathology research on this important topic are discussed. iv ACKNOWLEDGEMENTS