Social anxiety disorder (SAD) is highly prevalent (Kessler et al., 1994, Kessler et al., 2005a, Kessler et al., 2005b, Stein et al., 1994, Magee et al., 1996, Grant et al., 2005 and Stinson et al., 2007), associated with significant social and occupational impairment (Schneier et al., 1992, Davidson et al., 1993, Wittchen and Beloch, 1996, Keller, 2003 and Kessler, 2003), and often co-occurs with other psychiatric disorders (Rapee et al., 1988 and Reich et al., 1994). SAD generally has an onset during adolescence or early adulthood (Rapee et al., 1988, Reich et al., 1994 and Chartier et al., 1998) and is typically considered to be chronic (Reich et al., 1994, Yonkers et al., 2003 and Grant et al., 2005).
Data from clinical samples have suggested that SAD seldom remits (Reich et al., 1994, Massion et al., 2002 and Yonkers et al., 2003). Initial short-term treatment studies suggested that early age of onset (Yonkers et al., 2003) and comorbid alcohol use disorders (Versiani et al., 1988) were associated with persistence of SAD, whereas gender (Cameron et al., 1986) and duration of illness (Mersch et al., 1991) did not seem to affect its chronicity. More recently, prospective data (Reich et al., 1994, Massion et al., 2002 and Yonkers et al., 2003) from the Harvard/Brown Anxiety Research Program (HARP), which followed treatment-seeking patients with anxiety disorders for up to eight years, indicated that gender, age of onset, and comorbid anxiety and mood disorders did not predict persistence of SAD (Reich et al., 1994 and Massion et al., 2002), but avoidant personality disorder (APD) did (Massion et al., 2002). In primary care settings, comorbid panic disorder and lower psychosocial functioning have also been found to predict the persistence of SAD (Beard et al., 2010).
Several studies using community samples have also examined persistence of SAD. Data from the Duke site of the Epidemiological Catchment Area study suggested that earlier age of onset and presence of psychiatric comorbidity were associated with the persistence of SAD (Davidson et al., 1993), whereas results from the National Comorbidity Survey indicated that endorsing more than one social fear predicted persistence of SAD (Kessler et al., 1998). Conflicting results have been reported from longitudinal studies. For instance, the Dresden Predictor Study found, in an 18-month follow-up of 91 young women with SAD, that having a lifetime history of depression, a higher number of comorbid psychiatric disorders, higher levels of stress, poorer mental status, and being unemployed at baseline predicted persistence of SAD (Vriends et al., 2007). However, data from the Early Developmental Stages of Psychopathology Study on 183 adolescents and young adults with SAD followed for up to 4 years indicated that most sociodemographic factors, clinical characteristics, comorbid psychiatric disorders, familial factors, and stressful life events did not predict the persistence of SAD (Müller, 2002).
Although these previous studies have advanced our knowledge of the course of SAD, they have been constrained by having modest sample size (Davidson et al., 1993 and Vriends et al., 2007), use of cross-sectional data (Kessler et al., 1998 and DeWit et al., 1999), or reliance on clinical samples (Beard et al., 2010), limiting the generalizability of their findings. Large prospective studies in the community are essential for the identification of predictors of persistence of SAD in the general population.
To expand current knowledge on this topic, we used data from a large and nationally representative community sample of United States (US) adults, the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) (n = 43,093) and its 3-year prospective follow-up (n = 34,653). In the NESARC, the 12-month prevalence of SAD at baseline (i.e., Wave 1) was 2.76% ( Grant et al., 2005). The large size and representiveness of the NESARC sample and the longitudinal design of the study provide an unusual opportunity to simultaneously examine a broad range of variables that have been hypothesized to predict the course of SAD. Specifically, we sought to: 1) investigate differences in sociodemographic characteristics among individuals with persistent and remitted SAD; 2) compare rates of psychiatric comorbidity among individuals with persistent and remitted SAD; 3) examine clinical characteristics of SAD among individuals with persistent and remitted SAD; and, 4) examine childhood and adulthood risk factors among individuals with remitted and persistent SAD.
