Abstract
Conflicting reports exist on the direction of the relationship between social anxiety (SA) and alcohol/cigarette use (AU/CU) and alcohol/nicotine dependence (AD/ND), with both positive and negative associations reported. A prospective, longitudinal sample of Finnish twins (n = 1,906) was used to test potential explanations for these discrepancies. Specifically, this study used peer, parent, and teacher ratings of SA, and a clinical interview screening item for social anxiety disorder (SAD-Sc) to examine associations between SA and AU/CU and AD/ND from early adolescence into young adulthood. Peer-rated SA was negatively associated with AU, CU, and AD from age 14 through age 22, implying a protective effect (β = -0.01 to -.03). Teacher- and parent-rated SA associations were in the same directions but weaker or nonsignificant, indicating that aspects of SA that are recognizable by peers may be most relevant to AU/CU. Self-reported SAD-Sc was also negatively associated with AU, but positively associated with AD symptoms in young adulthood (β = 0.38). Our findings partially support the existence of different associations between SA and AU versus AD, but only in the context of SAD-Sc rather than trait SA. Neither trait SA nor SAD-Sc significantly predicted ND symptoms, although SAD-Sc was associated with both cigarette abstinence and daily smoking. These findings suggest that adolescent SA is modestly associated with lower AU/CU, although there may be some individuals with more severe SA who develop alcohol problems later in life. There was little evidence of a common underlying liability contributing to both SA and alcohol/cigarette use. (PsycINFO Database Record