Abstract
BACKGROUND: Although social anxiety disorder (SAD) is classified in the fourth edition of The Diagnostic and Statistical Manual (DSM-IV) into generalized and non-generalized subtypes, community surveys in Western countries find no evidence of disjunctions in the dose-response relationship between number of social fears and outcomes to support this distinction. We aimed to determine whether this holds across a broader set of developed and developing countries, and whether subtyping according to number of performance versus interactional fears would be more useful. METHODS: The World Health Organization's World Mental Health Survey Initiative undertook population epidemiological surveys in 11 developing and 9 developed countries, using the Composite International Diagnostic Interview to assess DSM-IV disorders. Fourteen performance and interactional fears were assessed. Associations between number of social fears in SAD and numerous outcomes (age-of-onset, persistence, severity, comorbidity, treatment) were examined. Additional analyses examined associations with number of performance fears versus number of interactional fears. RESULTS: Lifetime social fears are quite common in both developed (15.9%) and developing (14.3%) countries, but lifetime SAD is much more common in the former (6.1%) than latter (2.1%) countries. Among those with SAD, persistence, severity, comorbidity, and treatment have dose-response relationships with number of social fears, with no clear nonlinearity in relationships that would support a distinction between generalized and non-generalized SAD. The distinction between performance fears and interactional fears is generally not important in predicting these same outcomes. CONCLUSION: No evidence is found to support subtyping SAD on the basis of either number of social fears or number of performance fears versus number of interactional fears.