Social anxiety disorder in adults: Treatment overview

INTRODUCTION — Social anxiety disorder (SAD), also known as social phobia, is a condition marked by extreme fear of situations that involve possible scrutiny by others. The individual avoids these situations due to fear of humiliation or embarrassment or endures them with intense anxiety.

SAD is a prevalent condition, estimated to affect between 4 and 10 percent of the adult United States population over a 12-month period. SAD typically begins in childhood or adolescence and, untreated, can be associated with the subsequent development of major depression, substance use disorder, and other mental health problems. The disorder can be associated with extensive functional impairment and reduced quality of life [1].

This topic presents our treatment overview for SAD and SAD, performance only. The epidemiology, pathogenesis, clinical manifestations, assessment, and diagnosis of SAD are discussed separately, as is discussion of treatment with psychotherapy. (See "Social anxiety disorder in adults: Epidemiology, clinical features, assessment, and diagnosis" and "Social anxiety disorder in adults: Psychotherapy".)

PRIOR TO TREATMENT

Decision to treat — Patient preference is the primary factor in the determining the need for treatment. After the diagnosis of social anxiety disorder (SAD) is established, we review the impact that the disorder has had on the individual’s quality of life. We review the individual’s future goals and how these may be affected by the disorder. We use these discussions to help the patient decide on whether to pursue treatment.

Not all patients with SAD require immediate treatment. In particular, young adults may benefit from education about their disorder and may need time to reflect on the extent to which social anxiety and avoidance have negatively impacted their lives. It is not unusual for a newly diagnosed patient with SAD to be surprised that these symptoms can diminish or resolve with treatment as opposed to being an inalterable aspect of their self. The need for treatment of SAD is rarely an emergency. Most patients have had SAD for many years, and the decision to seek a diagnosis may have taken a long time.

Selection of treatment modality

Social anxiety disorder, generalized — Our choice of treatment for SAD, generalized, is based on informed patient preference.

Treatment of SAD with either cognitive-behavioral therapy (CBT) or pharmacotherapy has been found to reduce symptoms of social anxiety, as compared with placebo [2-5]. While some studies suggest improved response when using a combination of medication and psychotherapy (ie, CBT), other studies have not shown substantive differences compared with either modality alone. Given the uncertainties, we favor initiating therapy with either psychotherapy or medication while reserving combination therapy for those that prefer combination treatment and for selected individuals with a suboptimal response.

●Comparing CBT and pharmacotherapy – Evidence from randomized trials comparing CBT with pharmacotherapy for SAD has not demonstrated clear superiority of one treatment modality over the other [2-5].

As an example, in a randomized trial, 128 patients with SAD were assigned to treatment in one of the following: the monoamine oxidase inhibitor (MAOI) phenelzine, group CBT tailored for SAD, the combination of phenelzine and group CBT, or placebo pill [2]. After 12 weeks, response rates (as measured by the Clinical Global Improvement scale) between the CBT group and the phenelzine group were similar and greater than response rates compared with the placebo group (47 versus 54 versus 33 percent, respectively).

While the onset of symptom response may be faster with pharmacotherapy, speed of onset is rarely an important attribute in the treatment of SAD, as most patients have had the disorder for many years and a difference of a few weeks in response time is largely irrelevant. CBT appears to result in a more durable response [6].

●Combining CBT and pharmacotherapy – Clinical trials have yielded mixed results comparing the combination of CBT and medication to either modality individually [2-4]. As examples:

•In a randomized trial, 295 subjects with SAD were assigned to one of five treatment groups: fluoxetine, group CBT, combined group CBT and placebo, combined group CBT and fluoxetine, or placebo [3]. A higher proportion of patients responded to fluoxetine (51 percent), group CBT (52 percent), combined group CBT and placebo (51 percent) and combined group CBT and fluoxetine (54 percent) than to placebo (32 percent). However, the group receiving combined medication and CBT did not differ significantly in response compared with groups receiving either monotherapy.

•In a randomized trial of 128 patients with SAD (described above), combined group CBT-phenelzine resulted in higher response rates compared with either CBT or phenelzine as monotherapy (72 versus 47 versus 54 percent, respectively) [2].

Social anxiety disorder, performance only — For SAD, performance only, our preference is treatment with CBT, particularly if the individual will encounter performance situations on a recurring basis (see 'Patients who prefer psychotherapy' below). However, for a patient requesting help with a single or rarely recurring performance (ie, giving a toast at a friend’s wedding) pharmacotherapy is a suitable alternative. Further discussion of the treatment of SAD, performance only, is found below. (See 'Social anxiety disorder, performance only' below and "Social anxiety disorder in adults: Epidemiology, clinical features, assessment, and diagnosis", section on 'Performance anxiety'.)

PATIENTS WHO PREFER PSYCHOTHERAPY — For patients with either social anxiety disorder (SAD), generalized or SAD, performance only, who have chosen treatment with psychotherapy, we suggest first-line treatment with either individual or group cognitive-behavioral therapy (CBT) customized for SAD.

CBT as preferred initial modality — In CBT tailored for SAD, the therapist works with the patient to identify and challenge maladaptive cognitions associated with social situations and to reduce avoidance or safety behaviors when confronting feared situations [7,8]. (See "Social anxiety disorder in adults: Psychotherapy".)

Administration, components and comparative efficacy of various forms of psychotherapy including CBT, CBT tailored for SAD, attention retraining [9-11], interpersonal psychotherapy [12], and psychodynamic therapy in the treatment of SAD are discussed elsewhere. (See "Social anxiety disorder in adults: Psychotherapy".)

Placebo-controlled studies suggest that CBT may be the most efficacious form of psychotherapy for SAD [8,13-16].

●A meta-analysis of five randomized trials totaling 318 patients found traditional CBT to be efficacious for SAD compared with placebo control (odds ratio 4.21, 95% CI 2.07-8.98) [13]. Medium to large positive effects (Hedges g 0.84, 95% CI 0.72; 0.97) on social anxiety symptoms have also been seen for group CBT tailored for SAD compared with waitlist [14].

●A network meta-analysis including 101 clinical trials suggested that CBT was superior to no treatment, pill placebo, psychological control conditions, and psychotherapies including psychodynamic psychotherapy, interpersonal psychotherapy, mindfulness, and supportive psychotherapy [15].

However, those clinical trials directly comparing CBT with other forms of psychotherapy did not consistently favor either treatment [17-19]. As an example, in a trial 495 subjects with SAD were randomly assigned to treatment with manual-guided CBT, manual-guided psychodynamic psychotherapy or waitlist condition [18]. Both CBT and psychodynamic therapy were superior to wait list on measures of response rate and remission rate. While individuals receiving CBT demonstrated superior response compared with psychodynamic therapy on some measures of social anxiety at postassessment, differences were small in magnitude and not found during follow-up at 24 months [17].

For good response to CBT — We consider a response to treatment to entail substantial reduction in anticipatory fear and phobic avoidance, along with reduced anxiety when in social situations.

We encourage patients with a response to CBT to continue practicing the cognitive and behavioral skills learned in session on a maintenance basis. We recommend the inclusion of maintenance sessions (eg, two to four visits per year) as part of a relapse-prevention plan.

For partial or no response to CBT

Initial adjustments — We reassess patients with partial or no response to a full trial of CBT for remaining cognitive errors or avoidance strategies contributing to maintenance of social anxiety; we address these in additional weekly CBT sessions over 8 to 12 weeks. Additionally, we reassess the patient for comorbid disorders such as major depression, complicated grief, interpersonal trauma, or substance use disorder. We address these as needed. (See "Social anxiety disorder in adults: Psychotherapy", section on 'Strategies for nonresponse'.)

If initial adjustments do not lead to desired response to treatment, we typically augment the psychotherapy with a selective serotonin reuptake inhibitor (SSRI) or a serotonin-norepinephrine reuptake inhibitor (SNRI). However, we prioritize patient preference. For example, if an individual wishes to continue treatment with psychotherapy only, we follow their preference. Choice of medication and subsequent pharmacologic management of individuals with SAD is discussed below. (See 'Patients who prefer psychotherapy' above and 'Selecting a medication' below and 'Subsequent pharmacologic management' below.)

MEDICATION MANAGEMENT — For patients with social anxiety disorder (SAD), generalized type, who prefer medication management (with or without psychotherapy) or in those who have had a partial or no response to psychotherapy, our pharmacologic management is discussed below.

Selecting a medication — Our preference for pharmacologic management of SAD is first-line treatment with a selective serotonin reuptake inhibitor (SSRI) or a serotonin-norepinephrine reuptake inhibitor (SNRI). As there is no evidence supporting the superiority of one SSRI or SNRI over another, our choice is based on patient preference, clinician familiarity with medications, past history, side effects, and family history (table 1 and table 2). (See 'Efficacy of SSRIs and SNRIs' below.)

Other medications with efficacy in SAD are supported by fewer trials, have more problematic side effect profiles (monoamine oxidase inhibitors [MAOIs] [2,20] and benzodiazepines [21,22]), or show less robust treatment effects (gabapentin [23] and pregabalin [24-26]) when compared with placebo [15,20,27-30]. These agents are generally reserved for treatment of those with suboptimal response to SSRIs or SNRIs. (See 'Other agents with limited support' below.)

Our preference for pharmacologic management of SAD, performance only, is described below. (See 'Social anxiety disorder, performance only' below.)

Initiating treatment — To avoid side effects, when starting treatment with an antidepressant (eg, SSRI, SNRI), we are vigilant to begin at the lower end of the starting dose range and titrate gradually.

●SSRIs – SSRIs are the best studied and the most commonly prescribed of the medication treatments for SAD [31]. Higher doses of SSRIs typically result in better outcomes, so the dose is usually pushed to the maximum tolerated by the individual (unless an excellent response is attained at a lower dose, at which point the dose would be held there to ascertain stability of the response).

•We typically start paroxetine at 10 mg/day orally and increase to a therapeutic dose of 20 mg/day after a few days. We maintain at this dose for six weeks while monitoring symptomatically. If the patient does not have an adequate response, we increase by 10 mg every two to three weeks, as tolerated, until desired effect is seen or maximum dose of 60 mg per day is reached.

•We typically start sertraline at 25 mg/day orally and after one week increase to 50 mg per day. We maintain at this dose for four to six weeks. If the patient does not have an adequate response, we increase by 50 mg increments every two to three weeks, as tolerated, until desired effect is seen or a maximum of 250 mg/day is reached.

Common side effects of SSRIs include restlessness, agitation, headache, diarrhea, nausea, and insomnia. SSRIs also cause sexual dysfunction in as many as 50 percent of patients. Starting and maintenance doses of SSRIs and their side effects are presented on the tables (table 1 and table 2).

●SNRIs – When prescribing extended-release venlafaxine, we begin at 37.5 mg/day taken orally. We continue for three to seven days then increase to 75 mg (low end of the therapeutic range) and monitor for four to six weeks. If further increases are needed, we increase by 37.5 mg on a weekly basis, as tolerated, until response is seen or a maximum dose of 225 mg daily is reached.

As venlafaxine can increase blood pressure, particularly in higher doses, we routinely check blood pressure at each visit in individuals with a history of hypertension [32]. Although usually small, the increases can be significant in some patients.

Other common side effects of venlafaxine are nausea, dizziness, insomnia, sedation, and constipation. Starting and usual maintenance doses of SNRIs and their side effects are found on the tables (table 1 and table 2). (See "Serotonin-norepinephrine reuptake inhibitors: Pharmacology, administration, and side effects".)

Length of trial, monitoring, homework — We consider 10 to 12 weeks of treatment including six weeks at the highest tolerated dose within the therapeutic range to be a full therapeutic trial. The clinical effects of SSRI and SNRI treatment for SAD typically require four to six weeks to have a significant impact; maximal benefit can require as long as 16 weeks [6].

We use the Liebowitz Social Anxiety Scale (LSAS) to monitor treatment response at each visit. The scale has good psychometric properties and can be used for clinician assessment or patient self-report [33,34]. (See "Social anxiety disorder in adults: Epidemiology, clinical features, assessment, and diagnosis", section on 'Monitoring'.)

Whenever pharmacotherapy is prescribed for SAD, we recommend that the patient enter the previously feared social situations to “test out” the effects of the medication and to learn that these situations are safe and can be tolerated. Although there are no research data to support this claim, there is some reason to believe that those patients who do the most “practicing” and “expanding of their social horizons” while on medication are the ones who have the best longer-term outcomes. One of the first indications of response is the patient’s report of feeling less “self-conscious” in typical social situations.

Good response to treatment — For patients who have an adequate response to initial trial of SSRI or SNRI, we continue treatment for at least one year. In individuals who have maintained improvements over that period of time we gradually (over two to three months) taper off of medications. However, our treatment plan is individualized to address other factors such as difficulty reaching stabilization, level of psychosocial impairment, history of prior exacerbations after medication discontinuation, or the presence of active comorbid disorders. For example, in an individual whose symptoms caused significant psychosocial disruption and in whom stabilization may have been difficult, we would likely continue medication for greater than one year.

Partial or no response to treatment

Initial adjustments — We reassess patients who do not respond to a full trial of SSRI or SNRI medication for adherence to treatment including reviewing side effects that may be limiting adherence. Additionally, we reassess for the presence of comorbid disorders such as major depression, complicated grief, substance use, or interpersonal trauma. We address these as needed.

Augmentation versus changing medication — Our preference is augmentation of a partial response to a medication and change in agent for nonresponse to a medication.

●Partial response – For a partial response to the first or second SSRI/SNRI, our next step is augmentation of the SSRI. Our choice of augmenting agent is based on whether there is a history of a substance use disorder. (See 'No substance use disorder' below and 'Substance use disorder' below.)

●No response – For all individuals with no response to initial SSRI or SNRI, we try a full trial of a second SSRI or SNRI (see 'Length of trial, monitoring, homework' above). We use the same factors in choosing the second antidepressant as we do for the first. (See 'Selecting a medication' above.)

For no response to two SSRIs/SNRIs, we change to another agent. Our preference for the next agent depends on whether there is a history of a substance use disorder. (See 'No substance use disorder' below and 'Substance use disorder' below.)

A randomized clinical trial of patients with no or partial response after 10 weeks of sertraline treatment compared continued use of sertraline at the same dose with switching to venlafaxine [35]. Both groups improved somewhat over time, but no difference in response was seen between the two groups. While some patients may accrue additional benefit by remaining on their initial SSRI or SNRI beyond 10 weeks of treatment, it is difficult to justify leaving a patient with minimal or no response on the same medication for an extended duration.

Efficacy of SSRIs and SNRIs — Evidence supporting the use of SSRIs or the SNRI, venlafaxine, for the treatment of SAD in adults includes:

●In a meta-analysis of trials comparing SSRIs with placebo (seven trials, n = 896), treatment with SSRIs resulted in greater symptom reduction than placebo on the LSAS. The effect size was moderate [20]. Additionally, treatment with SSRIs led to greater reduction in symptoms such as depression, or associated disability than placebo.

●In a trial investigating the efficacy of treatment of SAD with paroxetine, 183 patients with SAD were randomly assigned to receive 11 weeks of paroxetine (20 to 50 mg) versus placebo [36]. At treatment end, subjects in the paroxetine group were more likely to be “much improved” or “very much improved” as compared with patients treated with placebo (55 versus 24 percent). Additionally, the percent improvement on the LSAS was greater in the paroxetine group as compared with placebo group (39 versus 17 percent).

●In a randomized trial, 440 patients with SAD were assigned to treatment with venlafaxine (75 to 225 mg daily extended release), paroxetine (20 to 50 mg daily), or placebo [27]. After 12 weeks, response rates were similar in the venlafaxine and paroxetine groups, both of which were superior to placebo (59 and 63 versus 36 percent response, respectively). Medication discontinuation rates were similar with venlafaxine and paroxetine, but adverse events leading to a reduction in dosage were more common with venlafaxine (16 versus 8 percent).

SUBSEQUENT PHARMACOLOGIC MANAGEMENT — For individuals who have had a poor or partial response to the above measures (ie, poor response to two selective trials of serotonin reuptake inhibitor [SSRI] and serotonin-norepinephrine reuptake inhibitor [SNRI], or partial response to either trial), our preference for subsequent treatment is based on whether there is a history of a substance use disorder. We typically avoid benzodiazepines in individuals with substance use disorders.

No substance use disorder — We prefer to use benzodiazepines as the next agent in those without an active or history of a substance use disorder. As above, we augment the SSRI/SNRI with a benzodiazepine for partial response, we discontinue the SSRI/SNRI and switch to a benzodiazepine for those with no response to SSRI/SNRI treatment. Choice, administration, and titration of benzodiazepines are discussed below. (See 'Benzodiazepine treatment' below.)

If symptoms persist, our next choice of agent depends on the level of response to the benzodiazepine.

●For no response to treatment with the benzodiazepine our preference is discontinuation of the benzodiazepine and treatment with a monoamine oxidase inhibitor (MAOI) as monotherapy. If MAOI is ineffective, unable to be taken (eg, due to inability to follow dietary restrictions), or if side effects limit treatment, our next steps are sequential trials of buspirone, gabapentin, or pregabalin. (See 'Monoamine oxidase inhibitors' below and 'Other agents with limited support' below.)

●For partial response to the benzodiazepine, our preference is augmentation. We do not use MAOI as augmentation in individuals on SSRI/SNRI due to potential for serotonin syndrome. In these cases, our preference is sequential augmentation trials of buspirone, gabapentin, or pregabalin. (See 'Monoamine oxidase inhibitors' below and 'Other agents with limited support' below.)

As minimal evidence supports the use of buspirone, gabapentin, or pregabalin as agents for monotherapy or augmentation treatment of social anxiety disorder (SAD), our choice is made based on patient preference and physician level of comfort and familiarity with the various agents.

Some individuals may wish to discontinue medication and switch to cognitive-behavioral therapy (CBT) treatment only. Additionally, CBT can be added to any pharmacologic regimen at any time. The choice of treatment is typically made based on patient preference [31]. (See 'Patients who prefer psychotherapy' above.)

Substance use disorder — We typically avoid the use of benzodiazepines in individuals with a history of a substance use disorder. If no response to SSRI treatment is seen, we prefer changing from the SSRI to an MAOI as monotherapy. In this case, a washout period is needed prior to starting the MAOI. We do not add the MAOI to the SSRI due to the potential for serotonin syndrome. Choice of agent, initial dose, titration, and dietary and pharmacologic precautions necessary when using MAOI are discussed below.(See 'Monoamine oxidase inhibitors' below.)

If treatment with an MAOI leads to no response, we try sequential trials of buspirone, pregabalin, or gabapentin. Our choice is made based on patient preference, physician level of comfort and familiarity with the various agents.

There is little evidence available to guide augmentation of an SSRI/SNRI for patients with a history of a substance use disorder and a partial response to the serotonergic drug [37]. The MAOI phenelzine cannot be used, because of the risk of serotonin syndrome with concurrent use of phenelzine and an SSRI/SNRI. For individuals with a partial response to SSRI/SNRI treatment, our choice is subsequent augmentation trials of buspirone, pregabalin, or gabapentin. As minimal evidence supports that choice of these agents in the treatment of SAD, our choice is made based on patient preference and physician level of comfort and familiarity with the various agents.

If these strategies are not effective, patients with a history of substance use disorder could be judiciously prescribed an adjunctive benzodiazepine provided that the substance use disorder was not active or recent (eg, in the past 12 months) and the patient’s use of the medication was carefully monitored. (See 'No substance use disorder' above.)

Some individuals may wish to discontinue medication and switch to CBT treatment only. Additionally, CBT can be added to any pharmacologic regimen at any time. The choice of treatment is typically made based on patient preference [31]. (See 'Patients who prefer psychotherapy' above.)

Benzodiazepine treatment — To minimize side effects such as sedation, we begin treatment with the benzodiazepine at a low dose and titrate slowly until satisfactory response is seen. Our preference is to a use an agent with slower onset and longer-half life, such as clonazepam. When using clonazepam, as either monotherapy or as an augmenting agent, we begin at 0.25 to 0.50 mg twice daily and we increase by 0.5 mg weekly to the maximum tolerated dose within the therapeutic range up to a maximum dose of 2 mg twice daily.

The potential for misuse and physiologic dependence limit the use of these agents. We are vigilant to discuss these issues with the patient and their family prior to beginning a benzodiazepine.

Efficacy — Small randomized trials have found the high-potency benzodiazepines alprazolam [21] and clonazepam [22], to be efficacious as monotherapy in the reduction of SAD symptoms.

Randomized trials have found evidence in support of augmentation of an SSRI with clonazepam [35,38]. For example, in a trial, 181 patients with SAD who had no more than a partial response to 10 weeks of sertraline were randomly assigned to 12 weeks of continuation on sertraline augmented by either clonazepam or placebo [35]. Treatment with sertraline plus placebo was associated with a greater drop in symptom severity and disability as measured by the Liebowitz Social Anxiety Scale (LSAS). Supplemental analysis showed that the overall response rate (as defined by LSAS score ≤50) was greater for those treated with sertraline plus clonazepam as compared with sertraline plus placebo (56 versus 36 percent).

Monoamine oxidase inhibitors — The MAOIs have the longest track record of use for SAD, having been essentially the only pharmacotherapy available (though never approved by the US Food and Drug Administration [FDA]) for SAD prior to the advent of the SSRIs. However, MAOI-related side effects, and dietary restrictions limit their use.

MAOIs cause inhibition of the MAO enzyme. Irreversible MAOIs, such as phenelzine, are incompatible with certain foods and medications (eg, cheese, aged meats, alcohol, over-the-counter cold preparations). A potentially fatal hypertensive reaction can occur if users of MAOIs consume food containing tyramine; thus, use of irreversible MAOIs must be accompanied by a low tyramine diet. Their use is also contraindicated with other antidepressants, and over the counter or prescribed medications containing sympathomimetic stimulants or dextromethorphan.

A washout period is needed after stopping the SSRI/SNRI and before starting the MAOI to avoid precipitation of a hypertensive reaction. We typically wait one to two weeks for most SSRIs/SNRIs and five to six weeks following cessation of fluoxetine. (See "Monoamine oxidase inhibitors (MAOIs): Pharmacology, administration, safety, and side effects".)

We begin phenelzine at 15 mg once or twice daily for two weeks then increase by 15 mg weekly as tolerated until desired effect is seen or a total daily dose of 60 to 90 mg/day is reached. It can take four to six weeks for a response to occur, and longer for maximal response to be attained.

Several randomized trials have found that phenelzine is an efficacious treatment for SAD, with approximately one-half of patients responding with a clinically significant reduction in symptoms [20]. A randomized trial that included a phenelzine arm found a response rate of 54 percent for phenelzine compared with 33 percent for placebo [2].

MAOIs are generally reserved for SAD refractory to other treatments. Some practitioners feel that MAOIs are more effective than SSRIs or SNRIs, though no comparative efficacy trials have been conducted to substantiate this belief. (See "Monoamine oxidase inhibitors (MAOIs): Pharmacology, administration, safety, and side effects".)

Other adverse effects common at therapeutic doses include postural hypotension, insomnia, paradoxical daytime sedation, sexual dysfunction, and weight gain. In many instances, these side effects can be dose limiting.

OTHER AGENTS WITH LIMITED SUPPORT — Clinical trials have found modest reductions in social anxiety disorder (SAD) symptoms with pregabalin or gabapentin monotherapy however, augmentation with these drugs has not been tested.

Other medication have yielded mixed or negative results in clinical trials or have not been sufficiently tested. Our choice of medication is based on clinical experience and patient preference.

●Buspirone – A small uncontrolled trial found that 7 of 10 patients partially responsive to selective serotonin reuptake inhibitors (SSRIs) appeared to benefit from augmentation with buspirone [37,39].

●Gabapentin – Gabapentin (an anticonvulsant) has demonstrated modest effectiveness (response rates <45 percent) for SAD. In a randomized trial, 69 patients with SAD were assigned to treatment with either gabapentin (flexible dose between 900 and 3600 mg daily in three divided doses) or placebo for 14 weeks [23]. Patients receiving gabapentin experienced reduced SAD symptoms compared with patients receiving placebo. Adverse events were consistent with the known side effect profile of gabapentin (eg, dizziness, dry mouth, somnolence).

Based upon case reports and human studies, the US Food and Drug Administration (FDA) issued a safety alert in 2019, warning that gabapentin may be associated with respiratory depression when administered to patients receiving central nervous system depressants or patients with underlying respiratory impairment [40]. In prescribing gabapentin to patients with these risk factors, we start the medication at a low dose (eg, 100 mg three times daily) and monitor for symptoms of respiratory depression while increasing slowly over several weeks.

●Pregabalin – Pregabalin, a compound related to gabapentin, has demonstrated modest effectiveness (response rates <45 percent) in the treatment of generalized SAD [25,26]. Side effects include sedation and dizziness. Tolerance, withdrawal, and dependence are possible, but pregabalin is generally better tolerated than benzodiazepines. Doses of pregabalin for SAD typically range from 150 to 600 mg per day.

The FDA issued a safety alert, warning that gabapentinoids may be associated with respiratory depression [40].

●Mirtazapine – Mirtazapine has shown mixed results in patients with SAD. In a randomized trial of 66 women with SAD, mirtazapine (30 mg a day) was found to be efficacious for SAD compared with placebo [41]. In a more recent randomized controlled trial of 60 males and females with generalized SAD randomized to either 30 to 45 mg/day of the medication or placebo for 12 weeks, mirtazapine was no more effective than placebo (13 versus 13 percent response rate) [42].

●Beta-blockers – Beta-blockers are not an effective treatment for generalized SAD [43-45].

●Tricyclic antidepressants – The TCAs have not been well studied in clinical trials of patients with SAD. A small open-label trial of imipramine suggested that it was not effective. It is possible that TCAs with more potent serotonergic reuptake blockade (eg, clomipramine) might be useful, but this is unknown.

●Second-generation antipsychotics – Very small randomized controlled trials of olanzapine and quetiapine suggested that these atypical (or “second-generation”) antipsychotics might be effective for SAD [46,47]. Further study is needed as well as consideration of metabolic side effects of these agents [48].

●Atomoxetine – In one small trial, atomoxetine was not shown to be effective for SAD [49]. Some evidence suggests that atomoxetine may be helpful in adults with comorbid ADHD and SAD [50].

●Ketamine – More work is needed to determine the efficacy and safety of ketamine infusions for SAD. In a small trial including 18 adults with SAD, a single dose of intravenous ketamine (0.5 mg/kg over 40 minutes) was associated with a higher rate of treatment response (defined as >35 percent reduction on the Liebowitz Social Anxiety Scale (LSAS) as compared with placebo in the first two weeks following infusion (33.3 versus 0 percent) [51]. The response rate to both treatments is low.

MEDICATION RESISTANCE — For patients who do not respond to any one or all of these medications, cognitive-behavioral therapy (CBT) remains an option, either as monotherapy or in conjunction with the medication providing some benefit. (See "Social anxiety disorder in adults: Psychotherapy", section on 'Cognitive-behavioral therapy'.)

SOCIAL ANXIETY DISORDER, PERFORMANCE ONLY — For most patients with social anxiety disorder (SAD), performance only, particularly those who will encounter performance symptoms on a recurring basis, we suggest first-line treatment with cognitive-behavioral therapy (CBT) for SAD.

For a patient requesting help with a single or rarely recurring performance (eg, giving the toast at a friend’s wedding), pharmacotherapy may be a suitable alternative. When medication is used, either a beta-adrenergic blocker or a benzodiazepine can be given 30 to 60 minutes before the performance. We avoid using benzodiazepines if the person is planning on drinking alcohol. Patients should try the medication in advance of a potentially precipitating event; a dose adjustment may be needed for effectiveness or side effects.

We would be more inclined to use a beta blocker in patients who have prominent awareness of physiological symptoms such as tachycardia or tremor.

Patients with public-speaking anxiety can benefit from exposure to practicing public speaking in classes or clubs such as Toastmasters, but generally only after the completion of a course of CBT. (See "Social anxiety disorder in adults: Psychotherapy", section on 'Administration'.)

●Beta adrenergic blockers – Several small trials found beta-blockers to reduce performance anxiety, although not in patients diagnosed with SAD [43,52]. Beta-blockers are nonetheless used at times for the management of performance anxiety. Clinical experience suggests that only half (or fewer) people find as-needed beta-blockers useful for performance anxiety.

Our preference is to use propranolol as it is lipophilic and crosses the blood brain barrier. When prescribing propranolol, we tell the patient to take 10 to 20 mg orally 30 to 60 minutes prior to the anxiety-inducing situation. We encourage the patient to try out the medication in advance of a potentially precipitating event to determine how well it is tolerated and to see if it is efficacious. If tolerated but not sufficiently effective, we suggest increasing the dose by 10 or 20 mg for the next time. Some patients may eventually require a dose of 60 mg. Contraindications to the prescription of beta-blockers include: a history of beta-blocker intolerance or allergy, diabetes, and certain cardiac conditions (eg, conduction problems). (See "Major side effects of beta blockers".)

●Benzodiazepines – Benzodiazepines can also be used on an "as needed" basis to treat SAD, performance only [53]. We typically give doses of clonazepam 0.25 to 1 mg or lorazepam 0.5 to 2 mg 30 to 60 minutes before the performance. Sedation can be a side effect of benzodiazepines, particularly at higher doses. For this reason, the patient should be encouraged to try out the medication in advance of a potentially precipitating event to determine how well it is tolerated and to see if it is efficacious.

Tolerance and physical dependence are less of a concern with occasional use. However, the potential for misuse, highest in persons with a history of alcohol or other substance use limit their use in this context. Patients may have used alcohol in the past to cope with similar situations and should be explicitly cautioned not to mix alcohol with benzodiazepines. The efficacy and side effects of benzodiazepines are reviewed above. (See 'No substance use disorder' above.)

●Cannabinoids – As the evidence in support of cannabinoids in the treatment of SAD in adults is sparse, we do not recommend the use of cannabinoids in the treatment of SAD.

In a small study, 24 subjects with a diagnosis of SAD but without prior treatment were randomly assigned to receive, 600 mg cannabidiol (n = 12) or placebo (n = 12) prior to a public speaking task. Treatment with cannabidiol resulted in reduced anxiety, cognitive impairment, and discomfort during their speech performance as compared with those receiving placebo [54]. However, in another trial, treatment with cannabidiol did not enhance the effects of exposure therapy in subjects with SAD and panic disorder with agoraphobia [55].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Anxiety and anxiety disorders in adults".)

SUMMARY AND RECOMMENDATIONS

●Social anxiety disorder (SAD) – SAD is a disorder manifested by extreme fear of situations that involve possible scrutiny by others leading to embarrassment or humiliation. The individual either avoids such situations or endures them with intense anxiety. SAD may be generalized or limited to performance situations only. (See 'Introduction' above.)

●Selecting treatment – Psychotherapy and medication are each effective for this disorder. Studies do not suggest greater effect of one treatment over the other or combined treatment over either alone. Our choice is based on informed patient preference and/or availability. (See 'Selection of treatment modality' above.)

•Preference for psychotherapy – For patients who prefer psychotherapy, we suggest treatment with individual or group CBT, customized for SAD, rather than other psychotherapies (Grade 2C). (See 'Patients who prefer psychotherapy' above.)

•Preference for medication – For patients who prefer medication, we suggest treatment with a serotonin reuptake inhibitor (selective serotonin reuptake inhibitor [SSRI] or serotonin-norepinephrine reuptake inhibitor [SNRI]) rather than other medications (Grade 2C). (See 'Selecting a medication' above.)

●Inadequate response – For patients with partial or no response to initial management we reassess for the presence of undiagnosed comorbid disorder. After addressing this, if symptoms persist our preference is as follows (see 'Initial adjustments' above):

•Response to CBT – For inadequate response to CBT, we assess for remaining cognitive errors or avoidance strategies and address this with additional CBT. (See 'Patients who prefer psychotherapy' above.)

If this is unsuccessful, we treat with a combination of CBT and an SSRI/SNRI.

•Response to medication – For inadequate response to initial pharmacological management (with or without combined psychotherapy), we assess for and address adherence to medication and treatment limiting side effects. (See 'Initial adjustments' above.)

For individuals with no response to initial SSRI/SNRI agent, we suggest a trial of a second SSRI/SNRI rather than augmentation methods (Grade 2C).

For partial response to SSRI/SNRI we prefer augmentation methods. Our choice depends on whether there is a substance use disorder. (See 'Subsequent pharmacologic management' above.)

For no response to two SSRI/SNRI agents we switch agents. Our next choice depends on whether there is a substance use disorder. (See 'Subsequent pharmacologic management' above.)

●Subsequent treatment

•No substance use disorder – For individuals without a substance use disorder, we suggest a trial of benzodiazepine rather than other medications (Grade 2C). We add the benzodiazepine to the SSRI for those with a partial response; we switch to a benzodiazepine for no response. (See 'No substance use disorder' above.)

If symptoms persist our next choice of agent depends on the level of response to the treatment with the benzodiazepine.

-For no response, we switch to a monoamine oxidase inhibitor (MAOI) as monotherapy. If MAOI is ineffective we treat with sequential trials of buspirone, gabapentin, or pregabalin.

-For partial response, our preference is augmentation methods. Our preference is sequential trials of buspirone, gabapentin, or pregabalin. We do not combine SSRI/SNRI with MAOIs.

•Substance use disorder – For individuals with a substance use disorder, we typically avoid benzodiazepines. (See 'Substance use disorder' above.)

-For no response to SSRI/SNRI, we switch to MAOI monotherapy. If this is ineffective, we prefer sequential trials of buspirone, gabapentin, or pregabalin.

-For partial response to SSRI/SNRI treatment, minimal data guides our treatment. We prefer subsequent augmentation trials of buspirone, gabapentin, or pregabalin. We do not combine MAOI with SSRI/SNRI.

●SAD, performance only – For patients with SAD, performance only, who will encounter performance situations on a reoccurring basis, we suggest first-line treatment with CBT tailored for SAD (Grade 2C). (See 'Social anxiety disorder, performance only' above.)

For a patient needing help with a single or rarely reoccurring performance, pharmacotherapy may be preferred. (See 'Social anxiety disorder, performance only' above.)