The epidemiology of social phobia: findings from the Duke Epidemiological Catchment Area Study

Abstract

SynopsisSocial phobia was studied in a North Carolina community, using DSM-III criteria. Two kinds of comparison were made: social phobia v. non-social phobia, and comorbid social phobia v. non-comorbid social phobia. Six-month and lifetime prevalence rates were 2·7 and 3·8% respectively. Social phobia had an early onset, lasted a long time and rarely recovered. Predictors of good outcome recovery in a logistic regression analysis were onset of phobia after age 11, absence of psychiatric comorbidity and greater education. The disorder was often missed in medical consultation. Increased rates of psychiatric comorbidity existed, especially for other anxiety disorders and for schizophrenia/schizophreniform disorder. There was increased risk of neurological disorder. Social phobia was also associated with an increased rate of suicide attempts, antisocial behaviour and impaired school performance during adolescence, impaired medical health, increased health-seeking behaviour, poor employment performance, reduced social interaction and impaired social support. Comorbidity accounted for some, but not all observed differences.