Social anxiety disorder (SAD), also called social phobia, is characterized by marked and persistent fear of embarrassment or humiliation in situations involving performance or interaction with or scrutiny by others (see Table 1 for DSM-IV criteria). The situation(s) are either avoided or endured with marked distress. The affected individual will often experience marked anticipatory anxiety before a feared interactional or performance situation and may experience a panic attack during the exposure. This anticipatory anxiety, avoidance, or distress in the social situation has a negative effect on the individual's social, academic, or occupational function and interpersonal relationships and/or causes marked distress.
DSM-IV Diagnostic Criteria for Social Anxiety Disorder
The true extent of the impairment and dysfunction associated with SAD may have been unrecognized in earlier years because the diagnostic term "social phobia" tended to conflate individuals with both subtypes of the disorder: those with nongeneralized or performance anxiety and those with the more impairing generalized subtype, in which the severity of symptoms and extent of impairment in function and quality of life are amplified (1, 2).
GENERALIZED SOCIAL ANXIETY DISORDER
Generalized social anxiety disorder (GSAD) accounts for two-thirds of individuals with SAD (2) and is characterized by fear and avoidance of numerous interactional as well as performance social situations. Compared with the nongeneralized subtype, GSAD is more pervasive and is associated with greater distress and dysfunction in affected individuals, increased alcohol and drug abuse, depression, suicide attempts, poor marital functioning, vocational impairment, financial dependence, increased utilization of health care resources, and decreased educational attainment (3—8).
NONGENERALIZED SOCIAL ANXIETY DISORDER
Nongeneralized SAD or "performance anxiety" refers to marked anticipatory anxiety, distress, and avoidance associated with public speaking or other performance-type situations. Although it is less pervasive and generally considered less disabling than the generalized subtype (4), nongeneralized SAD may result in significant impairment and underachievement at school and work as well (9). Most individuals with GSAD also experience performance anxiety.
A number of both national and international epidemiologic studies suggest that SAD is a common psychiatric disorder (10). Recently, the National Comorbidity Survey (NCS) Replication documented that SAD has a lifetime prevalence of 12.1%, making it the fourth most common psychiatric condition in the United States behind major depressive disorder, alcohol abuse, and specific phobias (11). A more stringent reanalysis of data from the Epidemiological Catchment Area Study and the NCS included the requirement that the disorder be clinically significant, defined as requiring treatment or causing impairment. Even with this more rigorous assessment, the disorder remained relatively common with a reported 12-month prevalence of 3.7% (12).
SAD frequently presents comorbidly with other psychiatric disorders including other anxiety disorders, major depressive disorder, bipolar disorder, and alcohol and substance use. The NCS documented that 81% of individuals with SAD report at least one other lifetime DSM-III-R psychiatric diagnosis (5). Not surprisingly given its overall greater severity, GSAD was more commonly associated with comorbid psychiatric illnesses than was the nongeneralized subtype (1, 2).
The onset of SAD frequently precedes and may in fact be a risk factor for the development of other comorbid disorders such as major depressive disorder (5, 6, 13, 14). In the NCS, secondary major depressive disorder was present in 37% of those with SAD (15); the lifetime rate of comorbid major depressive disorder was reported to be near 60% in clinical samples (16). Similarly, 22% of patients from a large study of bipolar disorder had SAD (17).
Comorbid anxiety disorders including panic disorder, posttraumatic stress disorder, generalized anxiety disorder, and obsessive-compulsive disorder are also relatively common among individuals with SAD. For instance, in the NCS, a lifetime history of posttraumatic stress disorder was present in 16% of individuals with SAD, panic disorder was present in 11%, and generalized anxiety disorder was present in 13% (5). Individuals with SAD also have an increased risk for alcohol abuse and dependence and other substance abuse disorders. Socially anxious individuals may use alcohol in an attempt to decrease anticipatory anxiety and reduce avoidance of feared social and/or performance situations. In the NCS, the lifetime prevalence rate of alcohol dependence was 24% among those with social phobia (5); thus, individuals with SAD have a two- to threefold increased risk of developing alcohol abuse or dependence compared with the general population. As with other comorbidities, SAD onset typically precedes that of alcohol abuse; in a prospective study of individuals with social phobia or subclinical social fears, the risk of the developing alcohol abuse or dependence was more than twice that of the general population without such fears (18). The rate of alcohol abuse among those with social phobia in the clinical setting approaches 40% (19).
The aim of treatment of SAD is to ultimately eliminate the patient's anticipatory and phobic anxiety around social interaction and performance situations, eradicating attendant avoidance behavior, and improving overall quality of life and function.
A variety of pharmacologic agents have demonstrated efficacy for the treatment of SAD (Table 2).
Pharmacotherapy Agents, Dosing and Side Effects for Social Anxiety Disorder
Serotonin selective and serotonin-norepinephrine reuptake inhibitors.
The serotonin selective reuptake inhibitors (SSRIs) and serotonin-norepinephrine reuptake inhibitors (SNRIs) have become first-line pharmacotherapy for the treatment of SAD because of their demonstrated efficacy for SAD and for major depressive disorder as well as other anxiety disorders that may often present comorbidly. In addition, they have a more favorable tolerability and safety profile than the monoamine oxidase inhibitors (MAOIs), which were the former "gold standard" agents for SAD and are not associated with the potential for abuse and dependence or ineffectiveness for comorbid major depressive disorder that complicates treatment with benzodiazepines.
The SSRIs paroxetine, sertraline, and fluvoxamine (controlled release) and the SNRI venlafaxine (extended release) have received Food and Drug Administration approval for the SAD indication. Other agents from this class also have demonstrated efficacy in randomized placebo-controlled studies and are likely effective as well, although differences in side effect profiles may be clinically relevant in some cases (20). Administration of SSRIs and SNRIs may be associated with a variety of side effects including sexual dysfunction, weight gain, increased anxiety, sedation, dizziness, headache, and gastrointestinal distress, as well as hypertension with venlafaxine and urinary retention with duloxetine. The onset of therapeutic effects usually takes at least 2—3 weeks, with greater benefits accruing over weeks or months as patients begin to expose themselves to previously feared situations. Roughly one-half to two-thirds of patients respond in acute treatment trials, with about one-half of these experiencing remission. Although the benefits of acute treatment are maintained in long-term follow-up (21), a substantial number of patients remain at least somewhat symptomatic over time (22). Given that most large randomized controlled trials (RCTs) often exclude patients with significant psychiatric or medical comorbidity as well as other complicating factors, it is likely that response and remission rates in clinical practice are even lower, underscoring the need to develop more effective treatment paradigms.
β-Blockers, such as propranolol (10—80 mg/day) and atenolol (50—150 mg/day) are effective for performance anxiety regarding public speaking or other performance situations (23, 24). They are typically administered on an "as needed" basis 1—2 hours before a performance situation, although some patients facing frequent performance challenges take them on a more routine basis. β-Blockers seem to reduce anxiety in performance situations by blunting the symptoms of physiological arousal such as tachycardia and tremor that are often an individual's focus when performing and drive an escalating cycle of fear and further anxiety. However, given their relatively short duration of action and lack of effect on the emotional and cognitive (relative to physiological) symptoms of social anxiety, these agents have not been considered to be first-line agents for GSAD (25). Side effects of β-blockers include lightheadedness, bradycardia, sedation, and nausea. Of interest, pindolol, a β-blocker with serotonin type 1A (5-HT1A) autoreceptor antagonist properties, which may accelerate or augment responses to antidepressants for major depressive disorder (26) and panic disorder (27), was ineffective in a placebo-controlled randomized augmentation trial in social phobics.
Monoamine oxidase inhibitors.
Until supplanted by the better tolerated and safer SSRIs and SNRIs, the monoamine oxidase inhibitors (MAOIs), including phenelzine and tranylcypromine, were the gold standard pharmacotherapy for SAD (25, 28). Early observations of the efficacy of MAOIs in atypical major depressive disorder, a syndrome characterized by marked sensitivity to rejection evocative of the focus of anxiety in those with social phobia (29), led to the use of the these agents in SAD and subsequent demonstration of efficacy in RCTs (25). However, the association of MAOI administration with troubling side effects including orthostatic hypotension, paresthesias, weight gain, and sexual dysfunction, as well as the need for proscribed dietary intake of tyramine-containing foods and sympathomimetic medication because of risk of potentially fatal hypertensive and serotonergic syndromes, has limited the widespread use of these agents and they are now generally reserved for use in refractory cases.
Benzodiazepines appear to be effective in SAD, with studies in nondepressed individuals treated with clonazepam and alprazolam suggesting efficacy beginning within 2 weeks (3, 30, 31). In addition to rapid onset of effect relative to other classes of effective agents, benzodiazepines have a favorable side effect profile and the flexibility to be used on an as-needed basis for situational anxiety. In addition, they can be used to augment the efficacy of antidepressants for generalized SAD. Data from a double-blind, randomized, placebo-controlled study demonstrated that the addition of clonazepam to paroxetine improved outcome compared with the SSRI alone (32). Adverse effects associated with benzodiazepine administration include sedation, ataxia, and cognitive and psychomotor impairment; further, physiological dependence may develop with regular use. In addition, it is important to recognize that benzodiazepines as monotherapy are generally not effective for the depressive disorders that commonly present comorbidly with SAD and in fact may lead to worsened mood. The abuse liability of benzodiazepines is generally limited to those with a predisposing diathesis or history of alcohol or substance abuse, although given the relatively high rates of concurrent alcohol or substance abuse in SAD, their abuse potential should be taken into account when a treatment plan for comorbidly affected individuals is developed and alternative therapeutic interventions should be used when possible.
Although not subjected to extensive testing for this indication, the tricyclic antidepressants seem to lack efficacy for SAD (33), whereas there is evidence from a small open trial suggesting potential efficacy for bupropion (34) and a small controlled study with mirtazapine that showed positive results (35). The azapirone buspirone, a 5-HT1A partial agonist, has not demonstrated efficacy for SAD as a monotherapy, although it may be useful for augmentation in patients incompletely responsive to SSRI therapy (36). Although small studies suggest the potential efficacy of atypical antipsychotic drugs including olanzapine (37), risperidone (38), aripiprazole (39), and quetiapine (40), they have not been tested in large RCTs in SAD, and given concerns about associated metabolic syndrome, weight gain, and extrapyramidal effects, their use is best reserved for patients remaining symptomatic despite standard interventions.
The anticonvulsants gabapentin, an α2δ calcium channel antagonist, and a related compound pregabalin demonstrated efficacy for social phobia in RCTs (41, 42). An open trial of valproic acid suggested potential efficacy for SAD (43), whereas results in small studies with levetiracetam have been mixed (44, 45) and results of a larger RCT with that agent were negative (Stein MB, Ravindran L, Simon NM, Khan A, Liebowitz M, Brawman-Mintzer O, Lydiard RB, Pollack MH: A Multicenter, Randomized, Double-Blind, Placebo-Controlled, Parallel-Group Study to Assess the Efficacy and Safety of Levetiracetam for Treatment of Generalized Social Anxiety Disorder. Submitted for publication, 2008.).
COGNITIVE BEHAVIORAL THERAPY
Cognitive behavioral therapy (CBT) is typically a time-limited psychosocial intervention, administered either in individual or group settings, that has demonstrated clear efficacy for the treatment of SAD (46, 47). Typical components of CBT include psychoeducation, somatic management techniques such as muscle relaxation, in vivo and imaginal exposure, video feedback, cognitive restructuring, and social skills training (46, 48). CBT has demonstrated efficacy comparable to that of pharmacotherapy, with a slightly slower onset of therapeutic effect but greater persistence of benefit after treatment discontinuation (31, 49—52).
Although combined pharmacotherapy and CBT may be presumed to be more effective than either intervention administered individually, evidence from one large RCT with fluoxetine and CBT did not demonstrate a significant advantage for the combined intervention over each effective monotherapy (53). This finding suggests that it would be reasonable to initiate treatment with either intervention alone and consider adding the alternative intervention for individuals who do not show a satisfactory response to the single intervention, although there are few systematic data addressing this hypothesis.
Recently, another paradigm for combining pharmacological and CBT interventions has emerged, with the aim of using pharmacotherapy to enhance the effects of exposure-based treatment. This approach is based on translational research derived from preclinical work on the neural circuitry underlying fear extinction, demonstrating the importance of the N-methyl-d-aspartate (NMDA) receptor within the amygdala for fear extinction and the effect of NMDA partial agonists administered systemically or in the amygdala to facilitate extinction (54, 55). Subsequent work in humans demonstrated that administration of the antibiotic d-cycloserine (DCS), an NMDA receptor partial agonist, before a CBT session, enhanced its efficacy for acrophobia (56, 57). More recently, DCS demonstrated efficacy for the enhancement of CBT in the treatment of SAD as well (58). If these early findings are confirmed, DCS and other agents active at the NMDA receptor and glutamatergic system may be routinely administered to enhance the effectiveness of exposure-based treatment of SAD and other phobic disorders.
Over the last two to three decades, advances in our understanding of SAD have been spurred by the growing recognition of its prevalence, early onset, chronicity, and morbid impact. Although currently available pharmacotherapies and CBT are clearly effective for the treatment of SAD, many of those treated remain symptomatic or fail to respond at all, and there is a significant unmet need to discover ways to optimize the use of currently available interventions and to develop novel therapies. Translational research derived from growing understanding of the underlying neurobiology of fear-based disorder offers the promise of improving outcomes for the treatment of SAD.