The Efficacy Of Psychotherapy For Social Anxiety Disorder: Systematic Review & Meta-Analysis

Social anxiety disorder (SAD) is commonly treated with various psychotherapy approaches, including cognitive-behavioral therapy (CBT), exposure therapy, and third-wave therapies such as mindfulness-based interventions.

CBT focuses on changing negative thought patterns and behaviors, while exposure therapy gradually introduces anxiety-provoking situations. Third-wave therapies incorporate mindfulness and acceptance strategies.

Reviewing these treatments is crucial because SAD significantly impacts quality of life and social functioning for many individuals.

By synthesizing current research on psychotherapy effectiveness, clinicians can make informed decisions about treatment options, researchers can identify areas for further study, and policymakers can allocate resources effectively to improve access to evidence-based interventions for SAD.

de Ponti, N., Matbouriahi, M., Franco, P., Harrer, M., Miguel, C., Papola, D., Sicimoğlu, A., Cuijpers, P., & Karyotaki, E. (2024). The efficacy of psychotherapy for social anxiety disorder, a systematic review and meta-analysis. Journal of Anxiety Disorders, 102881. https://doi.org/10.1016/j.janxdis.2024.102881

Key Points

• Psychotherapy is an effective treatment for social anxiety disorder (SAD), with a large overall effect size of g = 0.88 (95% CI: 0.76 to 1.0) compared to control conditions.
• The number needed to treat (NNT) was 3.8, indicating that approximately 4 individuals would need to be treated with psychotherapy to observe a positive outcome in one person.
• There were no significant differences in effectiveness between different types of psychotherapy for SAD, including cognitive behavioral therapy (CBT), exposure therapy, and third-wave therapies.
• Treatment delivery format, recruitment strategy, target group, mean age of participants, and number of sessions were significantly associated with treatment outcomes in a multivariable meta-regression.
• Group therapy formats and guided self-help showed larger effects compared to individual therapy and unguided self-help formats.
• The majority of included studies (62.1%) had a high risk of bias, and there was evidence of significant publication bias, suggesting results should be interpreted cautiously.
• Heterogeneity across studies was high (I2 = 74%), indicating substantial variability in treatment effects.
• The research provides an up-to-date, comprehensive synthesis of psychotherapy outcomes for SAD, but is limited by the predominance of waitlist control groups and studies conducted in high-income countries.
• The findings support the use of various psychotherapy options and delivery formats for SAD treatment, with implications for increasing treatment accessibility and scalability.

Rationale

This meta-analysis aimed to provide an updated and comprehensive examination of the efficacy of psychotherapy for social anxiety disorder (SAD) compared to control conditions.

SAD is one of the most common mental disorders, with a lifetime prevalence ranging from 4% to 15.4% worldwide (Koyuncu et al., 2019; Stein et al., 2017).

It is characterized by persistent fear and anxiety in social situations and can significantly impair functioning and quality of life (American Psychiatric Association, 2013; Barrera & Norton, 2009).

Previous meta-analyses have demonstrated positive outcomes for psychotherapies in treating SAD (Acarturk et al., 2009; Mayo-Wilson et al., 2014; Powers et al., 2008).

However, these studies are now nearly a decade old. Given the exponential growth in research on psychotherapy for SAD in recent years, there was a clear need for an updated synthesis of the evidence.

Additionally, technological advancements have led to the development of new treatment delivery formats, such as internet-based and self-guided interventions (Andersson et al., 2019; Clark et al., 2023). These formats have the potential to address barriers to treatment access and uptake.

The current meta-analysis aimed to incorporate these newer studies and delivery formats to provide a more current understanding of psychotherapy efficacy for SAD.

The rationale for this study was further strengthened by the need to examine potential moderators of treatment effects, such as type of therapy, delivery format, and patient characteristics.

Understanding these factors can inform clinical decision-making and help tailor treatments to individual needs.

By conducting this comprehensive meta-analysis, the authors sought to provide the most up-to-date and robust estimates of psychotherapy efficacy for SAD, synthesize findings across various treatment types and formats, and identify factors that may influence treatment outcomes.

Method

This systematic review and meta-analysis adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The protocol was pre-registered on the Open Science Framework.

Search strategy and terms:

The authors searched four bibliographic databases (PubMed, PsycINFO, Embase, and Cochrane CENTRAL) from their inception to January 1st, 2024.

They used a combination of index terms and free text related to psychotherapy and anxiety disorders, filtered for randomized controlled trials (RCTs). The specific search strings are provided in the supplementary material.

Inclusion and exclusion criteria:

Studies were included if they met the following criteria:

1. Randomized controlled trial
2. Examined a psychological intervention
3. Compared to a control group (e.g., waitlist, care-as-usual, or other inactive control)
4. Participants were adults (18 years and older)
5. Participants had a diagnosis of current SAD according to an operationalized diagnostic manual (e.g., DSM, ICD)

Studies were excluded if they were not in a language comprehensively understood by one of the authors (English, Dutch, German, Spanish, Italian, Greek, Persian, and Turkish).

Statistical measures:

The authors calculated the standardized mean difference (Hedges’ g) at post-test for each comparison between a psychological intervention and a control group.

They used a random-effects model for all analyses and estimated between-study heterogeneity using restricted maximum likelihood. The I2 statistic was calculated to assess heterogeneity.

Multiple sensitivity analyses were conducted, including different pooling methods, outlier and influence analyses, and publication bias assessments.

Subgroup analyses and meta-regressions were performed to examine potential moderators of treatment effects.

Results

Study characteristics:

The meta-analysis included 66 randomized controlled trials with 98 comparisons between psychotherapy and control groups.

A total of 5,560 participants were included, with 3,573 in intervention groups and 1,987 in control groups. The mean age of participants was 33.3 years, and 56.7% were women.

Main effect of psychotherapy:

The overall effect size for psychotherapy compared to control conditions was large (g = 0.88, 95% CI: 0.76 to 1.0), with a number needed to treat (NNT) of 3.8. Heterogeneity was high (I2 = 74%, 95% CI: 69 to 79).

Sensitivity analyses:

Most sensitivity analyses yielded similar results, with effect sizes ranging from g = 0.76 to 1.00. The outlier analysis, which removed 19 outliers, reduced heterogeneity to 0% (95% CI: 0 to 27) and slightly decreased the effect size (g = 0.76, 95% CI: 0.71 to 0.81).

Publication bias:

There was evidence of significant publication bias (Egger’s test, p < 0.001). After adjusting for publication bias, the estimated effect size was reduced, ranging from g = 0.30 to 0.86, depending on the method used.

Subgroup analyses and meta-regressions:

No significant differences were found between different types of psychotherapy (p = 0.709).

In the multivariable meta-regression, several factors were significantly associated with treatment effects:

1. Recruitment strategy: Clinical recruitment was associated with larger effects compared to community recruitment (β = -0.675, p = 0.003).
2. Treatment delivery format: Individual (β = -0.373, p = 0.035) and unguided self-help formats (β = -0.539, p = 0.043) were associated with smaller effects compared to group formats.
3. Target group: Other target groups (mainly students) showed larger effects compared to adults (β = 0.619, p = 0.019).
4. Number of sessions: A higher number of sessions was associated with larger effects (β = 0.168, p = 0.023).

Risk of bias:

Most studies (62.1%) had a high risk of bias, with only 6.1% demonstrating a low risk of bias. The main sources of bias were related to the randomization process, missing outcome data, and selective reporting.

Insight

This meta-analysis provides a comprehensive and up-to-date synthesis of the efficacy of psychotherapy for social anxiety disorder (SAD).

The large overall effect size (g = 0.88) and relatively low number needed to treat (NNT = 3.8) suggest that psychotherapy is an effective treatment option for adults with SAD.

This finding is consistent with previous meta-analyses but provides a more current and robust estimate based on a larger pool of studies.

One of the most informative aspects of this study is the lack of significant differences in effectiveness between various types of psychotherapy, including cognitive behavioral therapy (CBT), exposure therapy, and third-wave therapies.

This suggests that clinicians and patients have multiple effective options to choose from when selecting a treatment approach for SAD.

The flexibility in treatment options may help increase treatment uptake and adherence, as individuals can select an approach that aligns with their preferences and values.

The finding that group therapy formats showed larger effects than individual therapy is particularly interesting and somewhat contradicts previous research (Mayo-Wilson et al., 2014).

This could have important implications for treatment delivery, as group formats may be more cost-effective and allow clinicians to treat more patients simultaneously.

However, the authors rightly caution that this finding should be interpreted carefully, as there may be selection bias in terms of which patients are willing to participate in group treatments.

The effectiveness of guided self-help formats is another key insight from this study. While unguided self-help showed smaller effects, guided self-help performed comparably to traditional face-to-face treatments.

This finding supports the potential of internet-based and other technology-assisted interventions to increase treatment accessibility, particularly in areas with limited access to mental health professionals.

The significant associations found in the meta-regression between treatment outcomes and factors such as recruitment strategy, target group, and number of sessions provide valuable information for tailoring treatments and designing future studies.

For example, the positive association between the number of sessions and treatment effect suggests that longer treatments may be more beneficial, although this needs to be balanced with practical considerations and patient preferences.

However, the high risk of bias in many included studies, and evidence of publication bias highlight the need for more high-quality research in this area.

Future studies should focus on improving methodological rigor, including better randomization procedures, handling of missing data, and transparent reporting of outcomes.

To extend this research, future studies could:

1. Conduct more direct comparisons between different types of psychotherapy and delivery formats.
2. Investigate long-term outcomes of psychotherapy for SAD.
3. Examine the efficacy of psychotherapy for SAD in diverse populations and low- and middle-income countries.
4. Explore the mechanisms of change in effective psychotherapies for SAD.
5. Investigate combinations of psychotherapy with other interventions, such as pharmacotherapy or neurostimulation techniques.

Strengths

The study had many methodological strengths, including:

1. Comprehensive search strategy: The authors searched multiple databases and included a wide range of psychotherapy types and delivery formats.
2. Large sample size: With 66 RCTs and 5,560 participants, this is the largest meta-analysis on psychotherapy for SAD to date.
3. Rigorous statistical analysis: The authors conducted multiple sensitivity analyses, subgroup analyses, and meta-regressions to examine the robustness of their findings and explore potential moderators.
4. Adherence to guidelines: The study followed PRISMA guidelines and pre-registered its protocol, enhancing transparency and reproducibility.
5. Assessment of risk of bias: The authors used the Cochrane risk of bias tool 2 to evaluate the quality of included studies.
6. Examination of publication bias: Multiple methods were used to assess and adjust for potential publication bias.
7. Up-to-date evidence: The search included studies up to January 2024, providing a current synthesis of the literature.
8. Consideration of various treatment formats: The inclusion of different delivery formats (e.g., individual, group, guided self-help) allows for comparisons between these approaches.

Limitations

The study had several limitations that should be considered when interpreting the results:

1. High risk of bias: Most included studies (62.1%) had a high risk of bias, which may have inflated effect sizes.
2. Publication bias: There was significant evidence of publication bias, which could lead to an overestimation of treatment effects.
3. Heterogeneity: The high heterogeneity across studies (I2 = 74%) indicates substantial variability in treatment effects, which was not fully explained by the examined moderators.
4. Predominance of waitlist control groups: Most studies used waitlist control groups, which may inflate effect sizes compared to active control conditions.
5. Short-term outcomes: The meta-analysis focused on post-treatment outcomes and could not assess long-term effects of psychotherapy for SAD.
6. Limited generalizability: Most included studies were conducted in high-income countries, limiting the generalizability of findings to low- and middle-income settings.
7. Language restrictions: The exclusion of studies not in languages understood by the authors may have led to the omission of relevant research.
8. Lack of individual patient data: The use of study-level data limits the ability to examine patient-level moderators of treatment effects.

These limitations suggest that the results should be interpreted cautiously and highlight areas for improvement in future research on psychotherapy for SAD.

Implications

The findings of this meta-analysis have several important implications for clinical practice, research, and policy:

1. Treatment options: The comparable effectiveness of various psychotherapy types suggests that clinicians and patients have multiple evidence-based options for treating SAD. This flexibility may help increase treatment uptake and adherence.
2. Delivery formats: The effectiveness of group therapy and guided self-help formats has implications for increasing treatment scalability and accessibility. These formats may be particularly useful in settings with limited mental health resources.
3. Personalized treatment: The identification of factors associated with treatment outcomes (e.g., number of sessions, recruitment strategy) can inform treatment planning and help tailor interventions to individual needs.
4. Research priorities: The limitations identified in this study highlight the need for more high-quality RCTs, particularly those examining long-term outcomes and conducted in diverse settings.
5. Policy and funding: The effectiveness of psychotherapy for SAD supports continued investment in psychological treatments and efforts to increase their availability.
6. Training: The findings underscore the importance of training mental health professionals in various evidence-based psychotherapies for SAD.
7. Public health: Given the high prevalence and significant impact of SAD, the demonstrated effectiveness of psychotherapy supports its inclusion in public health initiatives aimed at reducing the burden of mental disorders.
8. Technology integration: The effectiveness of guided self-help formats supports the continued development and implementation of technology-assisted interventions for SAD.
9. Cost-effectiveness: The potential for group and guided self-help formats to be as effective as individual therapy has implications for improving the cost-effectiveness of SAD treatment.
10. Patient choice: The range of effective treatment options highlighted in this study supports a patient-centered approach, allowing individuals to choose treatments that align with their preferences and circumstances.

These implications suggest that while psychotherapy is an effective treatment for SAD, there is still room for improving access, personalizing treatments, and conducting further research to optimize outcomes.

References

Primary reference

de Ponti, N., Matbouriahi, M., Franco, P., Harrer, M., Miguel, C., Papola, D., Sicimoğlu, A., Cuijpers, P., & Karyotaki, E. (2024). The efficacy of psychotherapy for social anxiety disorder, a systematic review and meta-analysis. Journal of Anxiety Disorders, 102881. https://doi.org/10.1016/j.janxdis.2024.102881

Other references

Acarturk, C., Cuijpers, P., van Straten, A., & de Graaf, R. (2009). Psychological treatment of social anxiety disorder: a meta-analysis. Psychological medicine, 39(2), 241-254.

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.).

Andersson, G., Titov, N., Dear, B. F., Rozental, A., & Carlbring, P. (2019). Internet‐delivered psychological treatments: From innovation to implementation. World Psychiatry, 18(1), 20-28. https://doi.org/10.1002/wps.20610

Barrera, T. L., & Norton, P. J. (2009). Quality of life impairment in generalized anxiety disorder, social phobia, and panic disorder. Journal of anxiety disorders, 23(8), 1086-1090. https://doi.org/10.1016/j.janxdis.2009.07.011

Clark, D. M., Wild, J., Warnock-Parkes, E., Stott, R., Grey, N., Thew, G., & Ehlers, A. (2023). More than doubling the clinical benefit of each hour of therapist time: a randomised controlled trial of internet cognitive therapy for social anxiety disorder. Psychological medicine, 53(11), 5022-5032.

Koyuncu, A., İnce, E., Ertekin, E., & Tükel, R. (2019). Comorbidity in social anxiety disorder: diagnostic and therapeutic challenges. Drugs in context, 8, 212573. https://doi.org/10.7573/dic.212573

Mayo-Wilson, E., Dias, S., Mavranezouli, I., Kew, K., Clark, D. M., Ades, A. E., & Pilling, S. (2014). Psychological and pharmacological interventions for social anxiety disorder in adults: a systematic review and network meta-analysis. The Lancet Psychiatry, 1(5), 368-376. https://doi.org/10.1016/S2215-0366(14)70329-3

Powers, M. B., Sigmarsson, S. R., & Emmelkamp, P. M. (2008). A meta–analytic review of psychological treatments for social anxiety disorder. International Journal of Cognitive Therapy, 1(2), 94-113. https://doi.org/10.1521/ijct.2008.1.2.94

Stein, D. J., Lim, C. C., Roest, A. M., De Jonge, P., Aguilar-Gaxiola, S., Al-Hamzawi, A., … & WHO World Mental Health Survey Collaborators. (2017). The cross-national epidemiology of social anxiety disorder: Data from the World Mental Health Survey Initiative. BMC medicine, 15, 1-21. https://doi.org/10.1186/s12916-017-0889-2

Keep Learning

1. How might the finding that group therapy formats showed larger effects than individual therapy impact the way clinicians approach treatment for social anxiety disorder?
2. Given the effectiveness of guided self-help formats, what are the potential benefits and challenges of implementing more technology-assisted interventions for SAD?
3. How can researchers address the high risk of bias found in many studies on psychotherapy for SAD? What specific improvements in study design and reporting are needed?
4. Considering the limited generalizability to low- and middle-income countries, how might cultural factors influence the effectiveness of psychotherapy for SAD in diverse global settings?
5. What ethical considerations should be taken into account when deciding between different psychotherapy formats (e.g., individual vs. group) for patients with SAD?
6. How might the finding that longer treatments tend to be more effective be balanced with practical considerations such as cost and patient availability?
7. Given the high heterogeneity in treatment effects, what additional moderating factors should future research explore to better understand what works for whom in SAD treatment?
8. How can the insights from this meta-analysis be used to improve treatment guidelines for SAD and inform policy decisions regarding mental health resource allocation?