UC San Diego

The link between anxiety disorders and alcohol abuse and dependence has been established in adults, but the relationship between anxiety and adolescent alcohol involvement is less clear. In this dissertation, we compared explanatory models of substance use including the self-medication hypothesis, the social learning model, and the rebound hypothesis to characterize the association between anxiety and alcohol use in groups with various levels of experience with drinking. Additionally, we included depression symptoms in our studies to assess whether the relationships we found between anxiety and alcohol were unique or similar to those found with other forms of negative affect. In chapter 1, we investigated the relationships between several types of social anxiety symptoms and the onset of alcohol use in middle school students. We found that social anxiety significantly predicted initiation of drinking at both extremely high and low levels, and sex differences were observed. Chapter 2 follows with an analysis of the impact of social anxiety, depression, and alcohol expectancies on youths' drinking. We found that the self-medication hypothesis characterized the relationship between depression symptoms and drinking behavior, while the relationship between social anxiety and alcohol use was explained by social learning theory. Chapters 3 and 4 examined alcohol and drug use relapse characteristics in an adolescent substance use disordered treatment sample. In chapter 3 we found that compared to youth with only a substance use disorder, comorbid youth relapsed more often and more rapidly after treatment. In chapter 4, anxiety symptoms were more likely reported to improve than worsen immediately after relapse and in the two weeks following the relapse episode, indicating support for the self-medication hypothesis, while depression symptoms worsened in youth who relapsed with stimulants, which is consistent with a rebound effect. Chapter 5 is a replication of the study in chapter 4 with adults post-treatment for substance use disorders, with and without comorbid psychopathology. Unlike in the adolescent sample, almost no adults reported improvements in any psychiatric symptoms following relapse, however depression symptoms were rated as worsening more than anxiety symptoms