Anxiety Disorders

Reviewed population: Adults with GAD at step 3, moderate-to-severe panic disorder, or social anxiety disorder; use the disorder-specific CBT protocol and delivery recommended for each diagnosis.

NICE CG113 offers high-intensity CBT or applied relaxation when adults with GAD have marked impairment or insufficient response to lower-intensity care; it does not prefer psychological treatment over medication regardless of patient preference. For moderate-to-severe panic disorder, it says to consider referral for CBT, delivered by suitably trained and supervised practitioners using established protocols. For adult social anxiety, NICE CG159 offers individual CBT specifically based on the Clark and Wells or Heimberg model; group CBT is not routinely preferred. These are three diagnosis-specific recommendations with different formats and treatment steps. They do not endorse every CBT adaptation for all anxiety presentations, or establish that a generic CBT course is interchangeable with the named protocols.

Reviewed 2026-09-06

• NICE CG113 (2011), GAD and panic: recommendation 1.2.17
• NICE CG113 (2011), GAD and panic: recommendation 1.2.18
• NICE CG113 (2011), GAD and panic: recommendation 1.2.19
• NICE CG113 (2011), GAD and panic: recommendation 1.3.12
• NICE CG113 (2011), GAD and panic: recommendation 1.3.14
• NICE CG159 (2013), social anxiety: recommendation 1.3.2
• NICE CG159 (2013), social anxiety: recommendation 1.3.3
• NICE CG113 1.3.13 (2011): CBT within the psychological-intervention pathway

Reviewed population: Adults with social anxiety disorder who decline cognitive behavioral and pharmacological interventions; a short-term psychodynamic protocol specifically developed for social anxiety.

NICE CG159 says to consider short-term psychodynamic psychotherapy specifically developed for social anxiety when an adult declines cognitive behavioral and pharmacological interventions. The guideline explicitly notes more limited clinical effectiveness and lower cost effectiveness than CBT, self-help and medication. Its delivery recommendation describes a focused social-anxiety protocol addressing relationship themes, shame and engagement with feared social situations. This is a conditional alternative in a particular treatment pathway, not a first-choice recommendation for all anxiety disorders. It does not establish the same status for GAD, panic disorder, every brief psychodynamic variant or intensive short-term dynamic psychotherapy (ISTDP). The record reflects this 2013 guideline position, rather than an updated whole-literature efficacy comparison.

Reviewed 2026-09-06

• NICE CG159 (2013), social anxiety: recommendation 1.3.7
• NICE CG159 (2013), social anxiety: recommendation 1.3.16

Reviewed population: Adults with primary generalized anxiety disorder receiving Roemer and Orsillo's acceptance-based behavior therapy (ABBT).

A 2013 trial randomized 81 adults to 16 sessions of ABBT or applied relaxation. Both groups improved through six-month follow-up, without significant between-treatment differences. This is evidence for the named GAD protocol, not formal equivalence or superiority over applied relaxation, and not evidence for every anxiety diagnosis. The predominantly White research-clinic sample limits generalization.

Reviewed 2026-09-08

• Hayes-Skelton et al. (2013): ABBT versus applied relaxation for GAD

Reviewed population: Adults with principal or co-principal generalized social anxiety disorder receiving individual ACT; this assessment does not cover every anxiety disorder.

Craske et al. randomized 100 participants to ACT, CBT or waiting list; the modified intention-to-treat analysis included 87 after pretreatment exclusions. Both treatments outperformed waiting list, with no detected ACT–CBT differences on the reported outcomes. This establishes a randomized comparison in social anxiety, not formal proof of equivalence or ACT superiority. The restricted analysis and sample limit generalization.

Reviewed 2026-09-08

• Craske et al. (2014): ACT and CBT trial for social phobia
• Craske et al. (2014): social-phobia trial author manuscript

Reviewed population: Adults with co-occurring clinically elevated insomnia and anxiety symptoms, selected using ISI and GAD-7 thresholds, receiving a supported four-lesson internet CBT-I program.

Mason et al. (2023) randomized 120 adults to internet CBT-I or internet CBT for anxiety. Both groups reported reduced anxiety; GAD-7 improvement did not differ significantly between groups through three-month follow-up. CBT-I improved insomnia more at post-treatment, but that sleep advantage did not persist significantly at follow-up. The active-comparator study does not establish equivalence, anxiety-diagnosis remission, or effects in anxiety without insomnia. Entry used symptom thresholds, and diagnostic interviews were not repeated after treatment. This scoped randomized-study label does not borrow the separate insomnia guideline recommendation.

Reviewed 2026-09-08

• Mason et al. (2023), internet CBT-I versus anxiety CBT for co-occurring insomnia and anxiety

Reviewed population: Adults with primary generalized anxiety disorder, with or without co-occurring major depression, receiving Mennin and Fresco's Emotion Regulation Therapy.

The 2018 trial randomized 53 patients, 43% with comorbid major depression, to immediate ERT or modified attention control. ERT produced greater improvement on GAD and related outcomes. The control involved periodic contact while awaiting ERT, rather than an established active psychotherapy. This small trial supports a scoped randomized-study label; it does not establish superiority over CBT or treatment of every anxiety disorder.

Reviewed 2026-09-08

• Mennin et al. (2018): Emotion Regulation Therapy trial for GAD
• Mennin et al. (2018): ERT trial author manuscript

Reviewed population: Adults with a structured-interview-confirmed principal diagnosis of generalized anxiety disorder receiving individual Greenberg-derived Emotion-Focused Therapy in Irish public primary-care counseling services.

Timulak et al. (2022) randomized 58 clients to EFT or CBT, offered in 16–20 sessions by therapists trained in both approaches. GAD-7 symptoms improved over time, but the primary group-by-time interaction was not significant. The study followed participants through its six-month follow-up assessment and was designed to inform a future definitive noninferiority trial; it does not establish equivalence or noninferiority. Small sample, moderate treatment adherence and missing follow-up data limit inference. The 2023 correction clarifies six-month EFT research attrition. This concerns individual EFT for GAD, not couples EFT or tapping.

Reviewed 2026-09-08

• Timulak et al. (2022), EFT versus CBT for generalized anxiety: feasibility trial
• Correction (2023) to Timulak et al. (2022): six-month research attrition

Reviewed population: Adults with GAD or social anxiety disorder using the specific supported internet-delivered CBT programs studied; iCBT here is a delivery format, not inference-based CBT.

A GAD trial randomized 150 people to clinician-assisted or technician-assisted online CBT or delayed treatment; 145 were analyzed, and both supported programs improved outcomes versus delay. A separate 126-person social-anxiety trial found therapist-supported internet CBT noninferior to group CBT within its prespecified margin through six months. These findings concern particular supported programs and diagnoses, not every app, unguided use, or equivalence to all face-to-face CBT. GAD follow-up attrition and the social-anxiety trial's permissive noninferiority margin limit interpretation.

Reviewed 2026-09-08

• Robinson et al. (2010): supported internet CBT for GAD
• Hedman et al. (2011): internet CBT versus group CBT for social anxiety

Reviewed population: Adults aged 18–75 with primary GAD, social anxiety, panic disorder or agoraphobia; the TAME sample predominantly had GAD or social anxiety.

TAME randomized 276 adults to MBSR or escitalopram. At week 8, MBSR met the trial's prespecified noninferiority margin on clinician-rated overall severity in the 208-person primary per-protocol analysis, supported by imputed intention-to-treat analyses. Treatment time differed and participants were unblinded. Few participants had primary panic disorder or agoraphobia, so the pooled finding does not establish equivalence for each diagnosis, other drugs, or all mindfulness programs.

Reviewed 2026-09-08

• Hoge et al. (2022/2023): MBSR versus escitalopram, TAME trial

Reviewed population: Adults with long-standing generalized anxiety disorder treated with Wells-model metacognitive therapy at a Norwegian university outpatient clinic.

Nordahl et al. randomized 81 patients to metacognitive therapy, CBT or waiting list. Both therapies improved outcomes; metacognitive therapy showed greater improvement on the primary worry measure than CBT, with differences maintained at two years. Outcomes were predominantly self-reported and the planned one-year follow-up was abandoned after collection problems. This comparison concerns GAD and the studied protocols, not superiority for all anxiety disorders.

Reviewed 2026-09-08

• Nordahl et al. (2018): metacognitive therapy versus CBT for GAD

Reviewed population: Adults aged 18–65 with both social anxiety disorder and avoidant personality disorder as principal diagnoses, receiving group schema therapy in Dutch specialist outpatient services.

Baljé et al. (2024) randomized 154 patients to group schema therapy or group CBT, with 30 weekly group sessions. The primary longitudinal analyses found no significant treatment-by-time interactions through one-year follow-up. Separate intention-to-treat endpoint effect-size intervals favored CBT for social anxiety at one year and avoidant-personality severity at three months. More patients completed schema therapy, but attrition was substantial in both arms. This superiority trial evaluates this comorbid population; nonsignificant longitudinal differences do not establish equivalence or noninferiority. Findings do not establish effects for other anxiety disorders or schema-therapy formats.

Reviewed 2026-09-08

• Baljé et al. (2024), group schema therapy versus CBT for social anxiety with avoidant personality disorder

Reviewed population: Adult outpatients with principal GAD, panic disorder with or without agoraphobia, social anxiety disorder or OCD in a mixed-diagnosis trial.

Barlow et al. randomized 223 patients to the Unified Protocol, diagnosis-specific protocols or waiting list. Both active treatments outperformed waiting list; pooled principal-diagnosis severity changes met prespecified equivalence criteria at treatment end and six months. Equivalence was assessed across diagnoses, not independently established for every anxiety disorder. The specialist center developed the Unified Protocol and most comparators, limiting generalization to routine settings.

Reviewed 2026-09-08

• Barlow et al. (2017): Unified Protocol versus diagnosis-specific protocols