Distress‐Related Internalizing Symptoms Interact With Externalizing Symptoms to Predict Alcohol Problems in an Inpatient Adolescent Sample

Background and Objectives Previous research has yielded equivocal findings regarding whether internalizing symptoms are risk factors for adolescent hazardous alcohol use (AU), specifically in the presence of externalizing symptoms. This may be due to the type of internalizing symptoms examined (ie, distress vs fear), and the use of primarily normative rather than clinical samples. Thus, we tested internalizing and externalizing symptom interactions as they relate to adolescent hazardous AU in a high‐risk, clinical sample of adolescents. Methods Adolescents (N = 101; 66% female; Mage = 15) were recruited from an inpatient psychiatric unit and assessed for current symptoms. Hypotheses were tested by running a series of moderation models regressing hazardous AU on a set of internalizing symptoms (ie, depression, generalized anxiety disorder [GAD], social anxiety, separation anxiety, school avoidance, and panic), each interacting with externalizing symptoms. Results We observed significant interactions between several internalizing symptoms (depression, GAD, and panic symptoms) and externalizing symptoms predicting hazardous AU. These internalizing symptoms were unrelated to AU at low and average levels of externalizing symptoms, but were positively related to AU at high levels of externalizing symptoms. Discussion and Conclusions In this clinical sample, distress‐related internalizing symptomatology was associated with risk for, and not protection against, hazardous AU, but only when accompanied by high externalizing symptoms. Scientific Significance Findings suggest that whether or not internalizing symptoms operate synergistically with externalizing symptoms to predict hazardous AU depends on the subdomain of internalizing symptoms (distress vs fear) and perhaps the type of sample (ie, clinical vs community). (Am J Addict 2019:00:00–00)