Social anxiety disorder, also known as social phobia, is characterized by an excessive or unreasonable fear of scrutiny in social situations (American Psychiatric Association, 2000). The lifetime prevalence of social anxiety disorder is 5%, and the mean age of onset is 15 years (Grant et al., 2005). Two types of social anxiety disorder have been identified. Generalized social anxiety disorder indicates that the affected individual fears multiple social situations (i.e., eating, writing, speaking in public), whereas individuals who have only one social fear (typically public speaking fears) are considered to have the non-generalized type (Mannuzza et al., 1995). The generalized type is considered to be more severe and more debilitating (Kessler, Stein, & Berglund, 1998; Wittchen & Beloch, 1996; Wittchen & Fehm, 2001) and is the type most likely to seek treatment (Kessler et al., 1998).
One of the challenges in treating generalized social anxiety disorder is that it is often complicated by an additional Axis I disorder, most commonly other anxiety disorders, affective disorders, and/or substance use disorders (Burns & Teesson, 2002; Chartier, Walker, & Stein, 2003; Kessler et al., 1998; Magee, Eaton, Wittchen, McGonagle, & Kessler, 1996; Merikangas & Angst, 1995; Rosenbaum, 1995). These patients tend to have more severe social anxiety and greater functional impairment (Lecrubier, 1998). Unfortunately, little is known about how to treat social anxiety disorder in the context of a current comorbid psychiatric condition, as clinical trials typically exclude individuals with comorbid disorders (Herbert, 1995). In fact, a recent consensus paper on investigating treatments for social anxiety disorder specifically recommends excluding comorbidity because it may interfere with the measurement of efficacy (Montgomery et al., 2004). Consequently, there is a lack of evidence-based research to guide the treatment of social anxiety in individuals with comorbid conditions. A recent review of pharmacotherapy for social anxiety disorder and a recent review of treatment recommendations for persons with a co-occurring affective or anxiety and substance use disorders specifically have called for more research on comorbid samples (Stein, Ipser, & Van Balkom, 2006; Watkins, Hunter, Burnam, Pincus, & Nicholson, 2005), including alcohol use disorders (i.e., alcohol abuse or alcohol dependence), which occurs in about 20% of individuals seeking treatment for social anxiety disorder (Burns and Teesson, 2002 and Morris et al., 2005; Van Ameringen, Mancini, Styan, & Donison, 1991).
Given that one in five individuals who seek treatment for social anxiety disorder has an alcohol use disorder, an important question is whether a first-line treatment for social anxiety disorder, such as an SSRI, is a safe and effective treatment for social anxiety in this patient population. To our knowledge, the only study addressing this question was a pilot project conducted by our group (Randall et al., 2001). Those results, based on a small sample size and 8-week treatment period, were encouraging, but a larger scale study was warranted to confirm the pilot study results.
The present study was conducted to follow-up and extend our pilot study to include a larger sample size and a longer treatment period. We chose paroxetine as the treatment for social anxiety disorder, as its efficacy has been demonstrated with well-controlled clinical trials (Baldwin, Bobes, Stein, Scharwaechter, & Faure, 1999; Liebowitz, Gelenberg, & Munjack, 2005; Stein et al., 1998), and it was used in our pilot study. The randomized clinical trial was 16 weeks and double blind. It included individuals seeking treatment for social anxiety disorder who also met DSM-IV criteria for an alcohol use disorder. All individuals reported deliberate drinking to cope with social stress. Data on social anxiety severity and alcohol use were collected weekly using standardized and validated instruments. Research and clinical assessments of social anxiety were conducted separately to insure independent ratings of outcome variables. Data analysis used state-of-the-art statistical methods. The effects of paroxetine treatment on drinking outcomes (e.g., quantity/frequency of drinking, drinking to cope with social anxiety) is presented in more detail in a stand-alone publication. The current paper focuses only on social anxiety outcomes. The a priori hypothesis tested was that the paroxetine-treated group would demonstrate improvements in social anxiety compared to the placebo-treated group.
