CBT for anxiety

In short

• Fear and anxiety are not the same thing. Fear is a response to a threat that is here now; anxiety is a response to a threat you expect in the future. Both are normal. Anxiety becomes a disorder when it is out of proportion, lasts for months and gets in the way of life.
• One panic attack is not panic disorder. About 13 people in 100 have at least one panic attack in their lifetime; fewer than 2 in 100 develop panic disorder.
• What keeps anxiety going is the protection against it. In all the main CBT models, fear is maintained not by the threat itself but by avoidance and "safety behaviours", which stop you from finding out that the catastrophe doesn't happen.
• Panic is about how sensations are read. In David Clark's model (1986), an attack is driven by the belief that ordinary body sensations mean disaster: "my heart is pounding, so it's a heart attack".
• "Just relax and breathe" is not a treatment for panic. In a large component analysis, muscle relaxation was linked to worse outcomes, breathing retraining had no proven benefit, and the best results came from face-to-face therapy that combined rethinking the fear with deliberately bringing on the feared sensations.
• For social anxiety, individual CBT is the first choice in the UK's NICE guidelines. In a trial by Clark's group, 84% of people no longer met the diagnosis after cognitive therapy.
• Generalised anxiety has several models: intolerance of uncertainty (Dugas), beliefs about worry itself (Wells), and worry as avoidance (Borkovec). All lead to effective treatments, but the evidence hasn't settled which model is best.
• Health anxiety is kept going by checking and reassurance-seeking. In the CHAMP trial, after 5–10 sessions of CBT, a normal level of health anxiety at one year was about twice as common as with usual care: 14% against 7%. New symptoms are assessed by a doctor first.
• CBT works, but it isn't magic. Against a waitlist the effect is large; against placebo it is moderate. Most key trials were run by the people who created the treatments, and that is worth keeping in mind.

Timeline

1. 1985

   Beck, Emery and Greenberg: a book on the cognitive approach to anxiety disorders and phobias

   Cognitive line

2. 1986

   David Clark (Oxford): the cognitive model of panic

   Cognitive line

3. 1988

   Barlow and Craske: Mastery of Your Anxiety and Panic, the basis of Panic Control Treatment

   Behavioural line

4. 1989

   Lars-Göran Öst: one-session treatment for specific phobias

   Behavioural line

5. 1990

   Borkovec and Hu: worry dampens the heart's response to a frightening image

   Evidence

6. 1991

   Paul Salkovskis: the concept of safety behaviours

   Cognitive line

7. 1994

   Clark and colleagues' trial: cognitive therapy vs relaxation vs imipramine for panic

   Evidence

8. 1995

   Clark and Wells: model of social phobia; Wells: metacognitive model of generalised anxiety

   Cognitive line

9. 1998

   Dugas and colleagues: the intolerance of uncertainty model

   Cognitive line

10. 2000

    Barlow and colleagues' trial: CBT, imipramine and their combination for panic (312 people)

    Evidence

11. 2006

    Clark and colleagues' trial: cognitive therapy for social phobia, 84% vs 42%

    Evidence

12. 2008

    Wells's book on metacognitive therapy

    Third wave

13. 2011

    NICE CG113: guideline on generalised anxiety and panic

    Evidence

14. 2013

    NICE CG159: NICE's first guideline on social anxiety

    Evidence

15. 2014

    Network meta-analysis of 101 trials in social anxiety

    Evidence

16. 2017

    Barlow's Unified Protocol matches disorder-specific protocols

    Evidence

17. 2018

    Meta-analysis of placebo-controlled CBT trials; component analysis of panic treatment

    Evidence

How to read these notes: evidence labels

Each important claim carries a label in brackets, followed by its source after a semicolon. Proven by research means randomised trials (participants are split into groups by chance and the results compared) or meta-analyses (studies that pool many trials). Clinical consensus means clinical guidelines (such as those from the UK's NICE), disease classifications and professional bodies. The author's own account means a model or claim by the person who created the method, not independently checked. Interpretation means a reading of the evidence that can be challenged. Documented fact means dates, publications and names. Not verified means we could not open the primary source. Where sources disagree, both versions are given.

Fear and anxiety: what's the difference?

Disease classifications draw the line in time: fear is about "now", anxiety is about "what if, later".

Imagine you're crossing the road and a car shoots round the corner. You jump back before you've had time to think. That's fear. Now imagine a difficult conversation with your boss is a week away, and you're already sleeping badly, rehearsing what they'll say, and feeling your shoulders tighten. That's anxiety.

The World Health Organization's classification, ICD-11, describes fear as a reaction to a perceived imminent threat in the present, and anxiety as more future-oriented, a reaction to a perceived anticipated threat clinical consensus; ICD-11, WHO, v2026-01, via Find-A-Code. The American manual DSM-5 adds that fear is more often linked to surges of fight-or-flight arousal and escape, and anxiety to muscle tension, vigilance and avoidance clinical consensus; DSM-5, via Eppel & Swinson, 2024. The current edition is DSM-5-TR (2022); we could not check this wording against it word for word not verified.

Both reactions are normal and useful: fear gets you out of the car's path, anxiety makes you prepare for the exam. The trouble starts when anxiety is out of proportion to the situation, lasts for months, and starts deciding where you go and who you talk to. That is when clinicians speak of an anxiety disorder.

The main anxiety disorders

ICD-11 has a separate group called "Anxiety or fear-related disorders". The disorders in it differ in what the person is afraid of clinical consensus; ICD-11, WHO, v2026-01, via Find-A-Code; Domschke & Zwanzger, 2025.

In ICD-11, as in DSM-5, panic disorder and agoraphobia are diagnosed separately and can occur together documented fact; Domschke & Zwanzger, 2025. For generalised anxiety, ICD-11 dropped the strict six-month rule and the fixed symptom count: it requires marked symptoms "for at least several months, for more days than not" documented fact; Domschke & Zwanzger, 2025. OCD and PTSD sit in other ICD-11 groups; they are covered in Episode 9.

How common is it?

The two WHO estimates refer to different years: 301 million for 2019 and 470 million for 2023; the second is the current one documented fact; WHO, 2023 and 2026. Both point the same way: anxiety disorders are among the most common mental disorders, and most people who have one get no help.

The common logic of CBT: protection that keeps fear alive

Each anxiety disorder has its own model, but the mechanism they share is the same.

Cognitive behavioural therapy (CBT) for anxiety is not one method but a family of protocols. A protocol is a written-out therapy plan with clear steps. Each disorder has its own cognitive model: what exactly the person sees as dangerous, and what they do to protect themselves. Therapy then targets whatever keeps the anxiety going documented fact; Barlow et al., 2017.

The shared idea sounds counter-intuitive: anxiety is maintained not by the threat itself but by what people do to protect themselves from it.

• In panic, someone sits down and holds on to the wall "so I don't collapse".
• In social anxiety, someone stays quiet and hides their hands "so nobody sees them shaking".
• In generalised anxiety, someone keeps worrying "so I'm ready for the worst".

Safety behaviours

In 1991 the British psychologist Paul Salkovskis described these actions as safety-seeking behaviours: things a person does because they believe they prevent a catastrophe. Because the person believes these actions are what saves them, "spontaneous disconfirmation" of the threat becomes especially unlikely, which is why safety behaviours may be a key factor in maintaining anxiety disorders the author's own account; Salkovskis, 1991.

What it looks like in real life. Someone is afraid of fainting in the supermarket and always holds on to the trolley. They don't faint. The brain's conclusion: "Good thing I held on". They never get to find out whether they would have fainted without the trolley, so the fear stays exactly where it was.

Alongside safety behaviours there is avoidance: simply not going where it feels frightening. Avoidance brings instant relief, but the bill comes later: life gradually gets smaller.

How this was tested

In an experiment by Salkovskis, Clark and colleagues (1999), 18 patients with panic and agoraphobia spent 15 minutes in a frightening situation. One group kept their usual safety behaviours; the other deliberately dropped them. On a repeat test, those who had dropped them showed significantly less anxiety and less belief in catastrophe. The authors themselves note the small sample, and that the effect of dropping safety behaviours can't be fully separated from the effect of the explanation participants were given proven by research, one small study; Salkovskis et al., 1999.

The table is a teaching summary of the models described below (interpretation; based on Clark, 1986; Clark & Wells, 1995; Wells, 1995; Salkovskis, 1991).

How therapy is usually structured

Broadly, the path looks like this: assessment (interview and questionnaires) → ruling out medical causes of the symptoms, especially in panic → a personal map of the problem based on the disorder's model, drawn together with the client using their most recent episode → explaining how anxiety works → rethinking fears and running behavioural experiments → dropping safety behaviours → preparing for setbacks (interpretation; the usual structure of the protocols, based on NICE, 2011, 2013). How the personal map is built is covered in Episode 4, and how thoughts are reconsidered in Episode 5.

Because safety behaviours matter so much, modern protocols usually present exposure (deliberately facing the fear) as a behavioural experiment: the person tests a specific prediction, such as "if I let go of the trolley, I'll collapse", and does it without their usual protections the author's own account; Clark, 1986; Salkovskis, 1991. Behavioural experiments and exposure as techniques are covered in Episode 6.

Panic: how the body frightens itself

A panic attack is a surge of fear; panic disorder is the fear of the next surge.

Panic attack and panic disorder

A panic attack is a sudden surge of intense fear or discomfort that peaks within minutes. DSM requires at least four symptoms from a list including a pounding heart, sweating, trembling, shortness of breath, chest pain, nausea, dizziness, a sense of unreality, and fear of going mad or dying clinical consensus; DSM-5, via Cackovic et al., 2023. On average an attack lasts about 10 minutes proven by research, review; Tunnell et al., 2024.

Panic disorder is when attacks keep coming unexpectedly, followed by at least a month of persistent worry about further attacks or their consequences, or changes in behaviour to avoid them clinical consensus; DSM-5, via Cackovic et al., 2023. ICD-11 (code 6B01) is similar: recurrent unexpected attacks not tied to particular triggers, plus persistent worry or avoidance that impairs functioning clinical consensus; ICD-11, WHO, v2026-01, via Find-A-Code.

In these surveys the median age of onset of panic disorder was 32, and 80.4% of people with panic disorder also had another mental disorder at some point in their life proven by research; de Jonge et al., 2016, 2018. Up to 12% of emergency department patients meet criteria for panic disorder, and 20–35% of those whose complaint is chest pain proven by research, review; Tunnell et al., 2024.

Clark's model: the vicious circle

In 1986 the Oxford psychologist David Clark published "A cognitive approach to panic". His idea: panic attacks come from the catastrophic misinterpretation of certain body sensations, mostly those that are part of normal anxiety, such as a pounding heart, breathlessness and dizziness. The main example in the paper is perceiving palpitations as a sign of an impending heart attack the author's own account; Clark, 1986.

The circle goes like this the author's own account; Clark, 1986; the sequence of steps follows the Open University's OpenLearn course:

1. Trigger: external (a stuffy train, a queue) or internal (a racing heart after coffee, the thought "what if it happens again").
2. Perceived threat.
3. Apprehension: anxiety rises.
4. Body sensations: heart pounding, short of breath, dizzy, everything feels unreal.
5. Catastrophic interpretation: "it's a heart attack", "I'll suffocate", "I'll collapse", "I'm going mad", "I'll lose control".
6. The interpretation increases the sense of threat, the sensations get stronger, and the circle closes within seconds.

Why the circle doesn't break on its own. First, safety behaviours: sitting down, holding the rail, carrying water and a pill, only going out with someone. Every time an attack ends without catastrophe, the person credits the safety behaviour the author's own account; Salkovskis, 1991. Second, attention: someone afraid for their heart keeps checking it and notices beats others simply wouldn't. Third, avoiding situations and activities that bring on the sensations (exercise, heat, coffee), which leads to agoraphobia (interpretation; based on the model as described by OpenLearn and Psychology Tools).

The model also explains attacks "out of the blue": a harmless sensation, for example after running for a bus, gets linked to panic. And night-time attacks: a sensation during sleep starts the circle, and the attack wakes the person up the author's own account; as presented by OpenLearn. Critics point to night-time attacks with no conscious frightening thought and to biological models of panic; we did not check these lines of criticism against primary sources not verified.

The model has also been tested indirectly: cognitive measures at the end of treatment predicted longer-term outcome in the 1994 and 1999 trials, which is consistent with the model proven by research; Clark et al., 1994; Clark et al., 1999.

How panic is treated

The most effective part of panic treatment sounds strange: the feared sensations are brought on deliberately.

Cognitive therapy for panic (Oxford)

In outline, the protocol developed by Clark and Salkovskis includes these steps (interpretation; general description of the protocol, the manual itself was not consulted):

1. Map a specific episode with the client using the model: sensations → thought → how strongly believed → what they did.
2. Re-examine the interpretation: what supports "heart attack" and what doesn't, and what else could explain it ("my heart is racing from anxiety and coffee").
3. Behavioural experiments: bring on the sensations in a harmless way (fast breathing, running on the spot) to see where they come from.
4. Drop safety behaviours as an experiment: "what happens if I don't sit down?"
5. Go into avoided situations as a test of a prediction.
6. Prepare for possible setbacks.

A standard course of CBT for panic usually takes 12–15 one-hour sessions (interpretation; review by Rayburn & Otto, 2003). In a 1999 trial (43 people), a brief version, about 6.5 hours of therapist time plus self-study modules, did as well as the full version: both beat the waitlist, and results held at 12 months proven by research; Clark et al., 1999.

Panic Control Treatment (Barlow and Craske)

The American psychologists David Barlow and Michelle Craske developed Panic Control Treatment, based on their manual Mastery of Your Anxiety and Panic documented fact; Barlow & Craske, 1988. It includes explaining how anxiety and panic work, a panic diary, breathing skills, rethinking fears (especially overestimating the probability of disaster), interoceptive exposure, facing avoided situations, and relapse prevention the author's own account; Craske & Barlow, via OUPblog, 2007.

Interoceptive exposure: spinning in a chair with your therapist

Interoceptive exposure means deliberately bringing on feared body sensations. "Interoception" is the sense of signals from inside the body: heartbeat, breathing, balance. The point is for the person to have a new experience: "my heart is pounding and my head is spinning, and nothing terrible happens".

Craske and Barlow described a standard set of exercises documented fact; Craske & Barlow, 1993, as reproduced in a Macmillan textbook table:

The exercises most similar to a panic attack are rapid deep breathing (hyperventilation), spinning and straw breathing proven by research; Schmidt & Trakowski, 2004; Antony et al., 2006. In Germany's Panic Net project, with 301 patients, repeating interoceptive exercises reliably reduced both the intensity of the sensations and anxiety, and the more repetitions, the greater the reduction proven by research; Westphal et al., 2015.

How it is done in therapy (interpretation; based on Schmidt & Trakowski, 2004):

1. The therapist runs all the exercises in the session; the client rates the intensity of the sensations, their anxiety, and how similar it feels to panic.
2. The exercises are ranked into a ladder, from mildest to most panic-like.
3. Each exercise is done at full strength, then the client stays with the sensations without distraction or safety behaviours. Doing it "half-heartedly" is itself a safety behaviour.
4. Before the exercise the client writes down a prediction ("I'll lose control"); afterwards, what actually happened.
5. Then the skill is carried into daily life: coffee, stairs, sauna, exercise.

What works in panic treatment, and what doesn't

In 2018 Pompoli and colleagues broke CBT for panic down into its components: 72 trials, 4,064 participants proven by research; Pompoli et al., 2018. Results as odds ratios for remission (above 1 means better, below 1 means worse):

The most effective combination was face-to-face delivery, psychoeducation, support, cognitive restructuring and interoceptive exposure. Compared with the least effective combination, its odds ratio for remission was 7.69; compared with a standard CBT package, 2.33. The authors conclude that effective CBT packages for panic should include face-to-face delivery and interoceptive exposure and leave out muscle relaxation and virtual reality. They list the limitations themselves: small studies, wide intervals, no long-term data, and possible researcher allegiance proven by research; Pompoli et al., 2018.

CBT, medication, and what lasts

In Clark and colleagues' 1994 trial (64 people), cognitive therapy was compared with applied relaxation, the antidepressant imipramine, and a waitlist. At 3 months, cognitive therapy beat both alternatives on most measures. At 6 months it was level with imipramine, and both beat relaxation. Between months 6 and 15, some patients on imipramine relapsed, and at 15 months cognitive therapy was again better than both proven by research; Clark et al., 1994.

In Barlow and colleagues' large US trial (2000; 312 people, 4 centres, double-blind and placebo-controlled) there were five groups. Over 3 months, 48.7% responded to CBT, 45.8% to imipramine, 60.3% to CBT plus imipramine, and 21.7% to placebo; CBT and imipramine both beat placebo. In the acute phase the combination was not significantly different from the other groups; in the maintenance phase it beat each treatment alone but not CBT plus placebo. Six months after treatment ended, improvement held in 41.0% of the CBT-plus-placebo group, 31.9% after CBT alone, 26.3% after the combination and 19.7% after imipramine alone proven by research; Barlow et al., 2000. After treatment stopped, the groups that had CBT without an active drug did best.

NICE recommends CBT for panic totalling 7–14 hours, completed within 4 months, and advises against benzodiazepines because of poorer long-term outcomes clinical consensus; NICE, 2011. The American Psychological Association's Society of Clinical Psychology rates CBT for panic as having "strong research support" clinical consensus; Society of Clinical Psychology.

Social anxiety: "everyone is looking at me"

In social anxiety, people see themselves not through others' eyes but from inside their own sensations.

Social anxiety disorder (ICD-11 code 6B04; formerly called social phobia) is marked, excessive fear in social situations. The person fears they will act, or show anxiety symptoms, in a way that will be judged negatively. The situations are avoided or endured with intense fear, this lasts several months, and it significantly affects life clinical consensus; ICD-11, WHO, v2026-01, via Find-A-Code.

In the World Mental Health Surveys (28 surveys, 142,405 people), 4.0% of people have social anxiety disorder at some point in their life and 2.4% in the past year; it is more common in high-income countries and usually starts early proven by research; Stein et al., 2017.

The Clark and Wells model (1995)

In 1995 David Clark and Adrian Wells described a cognitive model of social phobia documented fact; Clark & Wells, 1995. Its core: in social situations, the person turns their attention inwards and builds an impression of how they look from their own sensations and images, not from what others actually see the author's own account; Clark & Wells, 1995, as summarised by Brown, 2007, and Psychology Tools. We could not open the 1995 chapter itself; the components below follow Brown's thesis (2007), Psychology Tools and the treatment targets listed in the NICE guideline.

The model's components the author's own account; based on Brown, 2007; Psychology Tools; NICE, 2013:

1. Negative beliefs of three kinds: excessively high standards ("I must sound brilliant"), conditional beliefs ("if I stumble over my words, they'll think I'm weak") and unconditional beliefs about the self ("I'm boring"). They switch on when the person enters the situation, and the situation is judged as dangerous.
2. Self-focused attention and an observer-perspective self-image. The impression of oneself is built from the inside: "I feel anxious, so I must look anxious". In a study by Hackmann, Surawy and Clark (1998), 77% of people with social phobia and 10% of controls described negative images of themselves as seen from outside proven by research, as summarised by Brown, 2007.
3. Safety behaviours: rehearsing sentences, avoiding eye contact, holding a glass with both hands, saying little or, the opposite, talking non-stop. They stop the fear being tested, create an odd impression in their own right (someone who stays quiet and avoids eye contact may come across as cold), and increase self-focus.
4. Physical symptoms such as blushing, trembling and a "blank mind", which feel obvious to everyone.
5. Anticipatory processing: before the event, running through scenarios of failure and recalling past ones.
6. Post-event processing, the "post-mortem": afterwards, going over in detail what went wrong, based on the distorted self-image, so that the event becomes worse in memory than it was.

Not every experiment supports the model: in some studies self-focus did not affect heart rate, and false feedback about heart rate did not strengthen the observer perspective (interpretation; based on the review in Brown, 2007).

Culture shapes what is feared. ICD-11 recognises a Japanese form of social anxiety, taijin kyofusho: the fear is not of being judged but of causing discomfort or offence to others through one's appearance or behaviour documented fact; Domschke & Zwanzger, 2025.

How social anxiety is treated

The NICE guideline (CG159) lists what individual CBT based on the Clark and Wells model should include clinical consensus; NICE, 2013:

1. Education about social anxiety.
2. Experiments showing the harm of self-focused attention and safety behaviours.
3. Video feedback to correct a distorted negative self-image.
4. Systematic training in externally focused attention.
5. Behavioural experiments in session, with linked homework.
6. Work on traumatic social memories.
7. Examining and changing core beliefs.
8. Working on anticipatory and post-event processing.
9. Relapse prevention.

The dose is up to 14 sessions of 90 minutes over about 4 months clinical consensus; NICE, 2013.

The safety-behaviour experiment. As it's usually done in therapy: the client has a conversation twice, first focusing on themselves and using all their usual safety behaviours, then without them and with attention on the other person, and compares their anxiety and how they think they came across (interpretation; general description of the protocol, the manual was not consulted).

Video feedback. A conversation is recorded. Before watching, the client describes in detail how they think they looked ("red as a tomato, voice shaking, endless pauses"), pictures those signs, and then watches the recording as if seeing a stranger. In an experiment by Harvey, Clark and colleagues (2000), this preparation strengthened the effect: participants who prepared rated their performance more positively than those who didn't proven by research, experiment not in patients; Harvey et al., 2000.

The numbers. In Clark and colleagues' 2006 trial (62 people), cognitive therapy was compared with an established treatment, exposure plus applied relaxation. After treatment, 84% no longer met the diagnosis after cognitive therapy, 42% after exposure with relaxation, and none on the waitlist; the differences held at one year proven by research; Clark et al., 2006.

In Mayo-Wilson and colleagues' network meta-analysis (2014; 101 trials, 13,164 participants), individual CBT had a large effect against a waitlist (SMD −1.19) and beat psychological placebo (−0.56); SSRI and SNRI antidepressants beat pill placebo (−0.44). Only individual CBT and SSRIs/SNRIs outperformed their respective placebos. The authors conclude that individual CBT is the best first choice, with medication for those who prefer it proven by research; Mayo-Wilson et al., 2014.

NICE names individual CBT (based on the Clark and Wells or Heimberg models) as the first-line treatment, advises against routinely offering group CBT instead, and advises against, for example, benzodiazepines for social anxiety, as well as botulinum toxin for sweating and surgery for blushing clinical consensus; NICE, 2013.

Generalised anxiety: why you can't stop worrying

In generalised anxiety, worry feels useful, yet it can't be switched off.

Generalised anxiety disorder (GAD) (ICD-11 code 6B00) means marked anxiety symptoms for at least several months, on more days than not, with general apprehension or excessive worry about many everyday things: family, health, money, study, work clinical consensus; ICD-11, WHO, v2026-01, via Find-A-Code.

In the World Mental Health Surveys (26 countries, 147,261 adults), 3.7% of people have GAD at some point in their life and 1.8% in the past year; in high-income countries the lifetime figure is 5.0%. Severe impairment at work, in study or in relationships was reported by 50.6% of people with GAD, and 49.2% had sought treatment proven by research; Ruscio et al., 2017.

CBT has no single model here. A review by Behar and colleagues (2009) compares five models of GAD and finds a common denominator: avoidance of internal experiences, that is, thoughts, beliefs and emotions (interpretation; Behar et al., 2009). Here are the three best known.

Dugas: intolerance of uncertainty

The Canadian psychologist Michel Dugas and colleagues (1998) proposed a model with four elements: intolerance of uncertainty, beliefs about worry, poor problem orientation and cognitive avoidance the author's own account; Dugas et al., 1998. In plain words: not knowing feels unbearable even when a bad outcome is unlikely; worry feels like a way to prepare ("if I think through everything that could go wrong, I'll be ready"); the person doubts their ability to solve problems; and they escape frightening images by switching to verbal "going over and over" (interpretation; common description of the model).

In the first test (24 patients with GAD and 20 people without it), all four variables helped tell the groups apart, intolerance of uncertainty was the strongest, and 82% of participants were correctly classified proven by research, small sample; Dugas et al., 1998. Group CBT based on this model (14 sessions, 52 adults) beat a waitlist, and improvement continued over 2 years of follow-up proven by research; Dugas et al., 2003.

Wells: worrying about worry

The British psychologist Adrian Wells (1995) proposed a metacognitive model. "Metacognitions" are thoughts about thinking. For Wells, the problem lies less in what people worry about than in their beliefs about worrying. On the one hand, worry is seen as useful ("it helps me prepare"); on the other, as dangerous and uncontrollable ("I can't stop it", "it'll drive me mad"). This produces meta-worry, worry about worry, and a sense that one's thinking is out of control the author's own account; Wells, 1995, 1999.

Metacognitive therapy (MCT; Wells, 2008) works on these beliefs rather than on the content of each "what if". In the protocol used in a 2018 trial, it did not use cognitive restructuring of worry content, exposure or relaxation; its methods were challenging beliefs about the uncontrollability and danger of worry, behavioural experiments, and postponing further "going over it" documented fact; Nordahl et al., 2018. MCT as part of the "third wave" is covered in Episode 13.

Borkovec: worry as avoidance

The American psychologist Thomas Borkovec proposed the avoidance theory of worry. In an experiment by Borkovec and Hu (1990), people afraid of public speaking imagined a frightening scene after relaxing, thinking neutral thoughts or worrying. After worrying, their heart rate rose least in response to the frightening image, even though they felt more afraid. The authors concluded that worry may suppress emotional processing and so preserve the fear structure proven by research; Borkovec & Hu, 1990.

In plain words: verbal worrying briefly dampens the body's reaction, gives short-term relief, and is reinforced for that reason. And because almost nothing people worry about actually happens, the brain draws a false conclusion: "I worried, and it was fine". This second part is the common interpretation of the theory; we did not find it stated in so many words in the sources we could open (interpretation). The 2004 review chapter by Borkovec and colleagues was not consulted not verified.

How GAD is treated

Bringing the different protocols together, therapy commonly includes (interpretation; summary based on Dugas et al., 2003; Nordahl et al., 2018; NICE, 2011):

• A worry diary: what about, for how long, what you did.
• Separating a solvable problem ("the bill isn't paid") from a hypothetical "what if" ("what if I get ill"). Solvable problems get solved step by step; with hypothetical ones, the work is learning to tolerate not knowing.
• Postponing worry to a set time, as an experiment against the belief "I can't control it".
• Re-examining beliefs about worry: does it actually help, is it actually dangerous?
• Cutting down on checking and asking "are you sure everything will be fine?" Each reassurance soothes briefly and works as a safety behaviour.

The models have been compared in trials. In Nordahl and colleagues' Norwegian trial (2018; 81 people), metacognitive therapy was compared with Borkovec's CBT. Recovery (on the PSWQ worry questionnaire) was 65% vs 38% after treatment and 57% vs 31% at 2 years proven by research; Nordahl et al., 2018. The supervisors were the protocols' own authors, Wells and Borkovec, both of whom receive royalties from related publications; the sample was small and from a single clinic documented fact; Nordahl et al., 2018. In a Dutch trial, metacognitive therapy also outperformed intolerance-of-uncertainty therapy proven by research; van der Heiden et al., 2012, via NICE Evidence Update, 2012. Earlier, Borkovec and Costello (1993) found applied relaxation and CBT roughly equal for GAD, and both better than non-directive therapy proven by research; Borkovec & Costello, 1993.

The APA's Society of Clinical Psychology rates CBT and applied relaxation for GAD as having strong research support, and calls metacognitive therapy "promising" for now clinical consensus; Society of Clinical Psychology. NICE recommends stepped care: first information and monitoring, then CBT-based self-help, then, for marked impairment, CBT or applied relaxation (usually 12–15 weekly one-hour sessions) or medication, as the person prefers; for complex GAD or a high risk of self-harm, specialist care clinical consensus; NICE, 2011.

Specific phobias: the most "behavioural" case

For phobias the main remedy is exposure, and sometimes one long session is enough.

A specific phobia is marked fear of a particular object or situation, such as heights, animals, flying, blood or injections. The fear is out of proportion to the real danger, and the person avoids it clinical consensus; ICD-11 code 6B03, WHO, v2026-01, via Find-A-Code.

It is one of the most common anxiety disorders. In the World Mental Health Surveys (25 surveys in 22 countries, 124,902 people), lifetime prevalence of specific phobia was 7.4% and 12-month prevalence 5.5%; it is more common in women (9.8% vs 4.9% lifetime). The median age of onset was 8. Only 23.1% of people with a phobia in the past year had received any treatment proven by research; Wardenaar et al., 2017.

The main method is exposure: gradually and according to a plan, facing what you fear. In a meta-analysis by Wolitzky-Taylor and colleagues (2008; 33 randomised studies), exposure had a large effect compared with no treatment and outperformed both placebo and other active psychotherapies; real-life exposure beat imaginal and virtual exposure right after treatment, but not at follow-up proven by research; Wolitzky-Taylor et al., 2008.

In 1989 the Swedish psychologist Lars-Göran Öst described one-session treatment: up to three continuous hours of exposure, therapist modelling, testing feared predictions, and encouragement. In his first series of 20 patients, 90% were "much improved or completely recovered" the author's own account; Öst, 1989, via Davis, Ollendick & Öst, 2012. This was a case series with no control group. A 2008 review calls the method promising but asks for independent replication and comparisons with other active treatments proven by research, review; Zlomke & Davis, 2008.

How exposure is explained today. Michelle Craske and colleagues (2014) suggest building exposure around an inhibitory learning model rather than aiming only for "getting used to" the fear. Their practical advice includes designing exercises so that the expected catastrophe is clearly disconfirmed, removing safety signals, and varying the contexts and conditions of practice documented fact; Craske et al., 2014. The model is usually explained this way: exposure doesn't erase the old fear but creates new learning, "this is safe", that competes with it (interpretation).

Health anxiety: when checking doesn't reassure

Why searching your symptoms online, or one more test "for peace of mind", often makes the anxiety worse.

Health anxiety (once called hypochondriasis) is a persistent fear that you have, or are about to develop, a serious illness, even though tests and doctors don't confirm it. The person isn't pretending: the sensations are real and the fear is distressing.

In 1986 Paul Salkovskis and Hilary Warwick proposed looking at this through a CBT lens. The core of their model is behaviour that works like avoidance: seeking reassurance ("tell me it isn't cancer") and repeatedly checking the body. According to the model, this is what keeps the preoccupation with health going. In two single-case reports, medical reassurance brought only short-lived relief, while a treatment that cut down reassurance-seeking worked quickly (the authors' own account; Salkovskis & Warwick, 1986). These are two cases with no control group, so they are an idea, not proof.

The cycle is usually described like this (interpretation; based on the Salkovskis and Warwick model): the person notices a sensation, say a tingling in the chest → reads it as a sign of illness ("it's my heart") → feels anxious → checks: takes their pulse, reads about symptoms, asks family, requests another test → feels better for a while → but attention to the body grows, they notice more sensations and start doubting ("what if the test was wrong?") → the circle repeats. The logic is the same as with OCD rituals (Episode 9).

In therapy, client and therapist draw this circle together, test an alternative explanation ("the problem is not an illness but the fear of illness"), use behavioural experiments to see how focusing on the body intensifies sensations, and gradually reduce checking and reassurance-seeking (interpretation; general description of the approach).

The largest trial is CHAMP. In six general hospitals in England, 444 adults aged 16–75 with marked health anxiety were randomly allocated either to 5–10 sessions of CBT (about 6 on average) or to usual care. The therapy was delivered by nurses and other staff with no previous CBT expertise, after training and under supervision. At one year, 13.9% of those who had CBT had a normal level of health anxiety, against 7.3% with usual care — roughly twice as many; the improvement held at 2 years proven by research; Tyrer et al., 2014. An honest caveat: even after CBT, most participants did not reach a normal level. It is a clear improvement, not a cure for everyone (interpretation).

One protocol for all anxiety?

If anxiety disorders share a mechanism, can one method treat them all?

David Barlow thinks so. His idea is that emotional disorders share a vulnerability: a tendency to react strongly to negative emotions and to avoid them. The Unified Protocol (UP) is a transdiagnostic approach, meaning one protocol aimed at mechanisms shared by different disorders the author's own account; Barlow & Gallagher, Society of Clinical Psychology.

Its modules: the adaptive nature of emotions; motivation; present-moment emotion awareness; cognitive flexibility; identifying and preventing emotional avoidance; tolerating physical sensations; and emotion-focused exposure to situations and sensations documented fact; Barlow et al., 2017.

In a 2017 trial (223 patients with panic disorder, GAD, OCD or social anxiety; up to 16 sessions), the Unified Protocol was equivalent to disorder-specific protocols on the severity of the main diagnosis, and more people completed treatment: 87.5% vs 69.2% proven by research; Barlow et al., 2017. Critics note that the sample was well educated and mostly white, more than half were taking medication, and the trial was run by Barlow's own centre (interpretation; Newby & Norton, 2018, The Mental Elf).

How well does CBT for anxiety really work?

The answer depends on what you compare it with.

Compared with a waitlist (people simply waiting for treatment), the effect is large: SMD −1.19 for social anxiety proven by research; Mayo-Wilson et al., 2014. Compared with placebo (for example, conversations without active techniques, or a sugar pill), the effect is moderate.

Carpenter and colleagues' meta-analysis (2018) pooled 41 placebo-controlled trials (2,835 patients) of CBT for anxiety and related disorders proven by research; Carpenter et al., 2018:

How to read this: g is a standardised difference between groups; 0 means no difference, around 0.2 is small, around 0.5 moderate, and 0.8 or more large (interpretation; Cohen's conventional benchmarks). Publication bias is when successful studies are more likely to be published than unsuccessful ones, inflating the average effect. In the non-PTSD studies, participants were on average 82% white proven by research; Carpenter et al., 2018.

Other caveats:

• Evidence quality for panic was rated low or very low by a Cochrane review proven by research; Pompoli et al., 2016.
• Allegiance. Key trials (Clark 1994, 1999, 2006; Dugas 2003; Barlow 2017) were run by the protocols' own developers. The component analysis of panic treatment lists the risk of researcher-allegiance bias among its limitations, which is why independent replication matters (interpretation; Pompoli et al., 2018; Newby & Norton, 2018).
• People over 55. Against an active control, the effect of CBT for anxiety was small and not significant proven by research; Gould et al., 2012, via NICE Evidence Update, 2012.
• Culture. We found no studies of CBT for anxiety in Russian-speaking or Azerbaijani samples for these notes.

This doesn't mean CBT doesn't work. It means it helps many people, but not everyone, and not by magic. The evidence for and criticism of CBT are covered in Episode 14.

Myths about anxiety: a quick check

Try this: translating the language of anxiety

A one-minute exercise. It's a first step in self-observation, not treatment.

The point is to notice that between the sensation and the fear there is a thought. It follows the same logic as Clark's model, but it doesn't replace therapy. Interoceptive exercises are not for doing on your own: they are done with a therapist, after a medical check.

When to see a professional

• If attacks keep happening, you've started avoiding places or activities, or anxiety is getting in the way of work, study or relationships: first a doctor, to rule out medical causes, then a psychotherapist who works with CBT.
• A psychiatrist if there is marked depression, thoughts of death or self-harm, dependence on alcohol or sedatives, or a decision about medication to make. If you are having thoughts of self-harm, see If you need help now.
• A psychologist does not rule out physical illness or prescribe medication; that is a doctor's job.

Check yourself

1. How does fear differ from anxiety? Fear is a response to a threat now; anxiety, to a threat expected in the future.
2. What, according to Clark, turns ordinary sensations into a panic attack? Catastrophic misinterpretation: reading the sensations as a sign of immediate disaster.
3. Why do safety behaviours keep anxiety going? The person credits their "rescue" to the safety behaviour and never learns that the catastrophe wouldn't have happened anyway.
4. According to the component analysis, which two elements should CBT for panic include, and which one not? Include face-to-face delivery and interoceptive exposure; muscle relaxation is linked to worse outcomes.
5. How do the Dugas, Wells and Borkovec models of GAD differ? Intolerance of uncertainty; beliefs about worry itself; worry as avoidance of emotion and physical arousal.
6. Why doesn't one more test "for peace of mind" help with health anxiety? Reassurance eases the anxiety only briefly, while it increases attention to the body and doubt, so the checking circle repeats.

What to remember

• Anxiety is normal; it becomes a disorder when it is out of proportion, lasts and gets in the way of life.
• In all CBT models, anxiety is maintained by avoidance and safety behaviours, which stop the fear from being tested.
• Panic is a circle: sensation → catastrophic thought → more sensations. It is broken by rethinking the fear and deliberately bringing on sensations with a therapist, not by relaxing.
• In social anxiety, people see themselves from inside their own sensations; individual CBT is the first choice.
• In GAD, therapy works on how the person relates to uncertainty and to worry itself, rather than arguing with each "what if".
• For phobias, the main method is planned exposure.
• In health anxiety, CBT reduces checking and reassurance-seeking; new symptoms are assessed by a doctor first.
• Against placebo, the effect of CBT is moderate, and that's an honest benchmark. Medical causes of symptoms and decisions about medication are for a doctor.

Sources

1. Antony, M. M., Ledley, D. R., Liss, A., & Swinson, R. P. (2006). Responses to symptom induction exercises in panic disorder. Behaviour Research and Therapy, 44(1), 85–98. https://experts.mcmaster.ca/scholarly-works/675531
2. Barlow, D. H., & Craske, M. G. (1988). Mastery of Your Anxiety and Panic. Graywind.
3. Barlow, D. H., & Gallagher, M. W. The promise of transdiagnostic treatments for anxiety disorders. Society of Clinical Psychology (APA Division 12). https://div12.org/?p=7723
4. Barlow, D. H., Gorman, J. M., Shear, M. K., & Woods, S. W. (2000). Cognitive-behavioral therapy, imipramine, or their combination for panic disorder: A randomized controlled trial. JAMA, 283(19), 2529–2536. https://pubmed.ncbi.nlm.nih.gov/10815116/
5. Barlow, D. H., Farchione, T. J., Bullis, J. R., et al. (2017). The Unified Protocol for transdiagnostic treatment of emotional disorders compared with diagnosis-specific protocols for anxiety disorders: A randomized clinical trial. JAMA Psychiatry, 74(9), 875–884. https://open.bu.edu/handle/2144/33638
6. Beck, A. T., Emery, G., & Greenberg, R. L. (1985). Anxiety Disorders and Phobias: A Cognitive Perspective. Basic Books.
7. Behar, E., DiMarco, I. D., Hekler, E. B., Mohlman, J., & Staples, A. M. (2009). Current theoretical models of generalized anxiety disorder: Conceptual review and treatment implications. Journal of Anxiety Disorders. https://read.qxmd.com/read/19700258
8. Borkovec, T. D., & Costello, E. (1993). Efficacy of applied relaxation and cognitive-behavioral therapy in the treatment of generalized anxiety disorder. Journal of Consulting and Clinical Psychology, 61(4), 611–619. https://pure.psu.edu/en/publications/efficacy-of-applied-relaxation-and-cognitive-behavioral-therapy-i/
9. Borkovec, T. D., & Hu, S. (1990). The effect of worry on cardiovascular response to phobic imagery. Behaviour Research and Therapy, 28(1), 69–73. https://pure.psu.edu/en/publications/the-effect-of-worry-on-cardiovascular-response-to-phobic-imagery/
10. Brown, M. A. (2007). Cognitive biases in social anxiety (PhD thesis). University of Southampton. https://eprints.soton.ac.uk/466366/
11. Cackovic, C., Nazir, S., & Marwaha, R. (2023). Panic disorder. StatPearls. https://www.ncbi.nlm.nih.gov/books/NBK430973/
12. Carpenter, J. K., Andrews, L. A., Witcraft, S. M., Powers, M. B., Smits, J. A. J., & Hofmann, S. G. (2018). Cognitive behavioral therapy for anxiety and related disorders: A meta-analysis of randomized placebo-controlled trials. Depression and Anxiety, 35(6), 502–514. https://pmc.ncbi.nlm.nih.gov/articles/PMC5992015
13. Clark, D. M. (1986). A cognitive approach to panic. Behaviour Research and Therapy, 24(4), 461–470. DOI 10.1016/0005-7967(86)90011-2. https://ora.ox.ac.uk/objects/uuid:96416d34-2558-4cef-9622-401c7b11490a
14. Clark, D. M., Salkovskis, P. M., Hackmann, A., Middleton, H., Anastasiades, P., & Gelder, M. (1994). A comparison of cognitive therapy, applied relaxation and imipramine in the treatment of panic disorder. British Journal of Psychiatry, 164, 759–769.
15. Clark, D. M., Salkovskis, P. M., Hackmann, A., Wells, A., Ludgate, J., & Gelder, M. (1999). Brief cognitive therapy for panic disorder: A randomized controlled trial. Journal of Consulting and Clinical Psychology, 67(4), 583–589. https://research.manchester.ac.uk/en/publications/brief-cognitive-therapy-for-panic-disorder-a-randomized-controlle/
16. Clark, D. M., & Wells, A. (1995). A cognitive model of social phobia. In R. G. Heimberg, M. R. Liebowitz, D. A. Hope, & F. R. Schneier (Eds.), Social Phobia: Diagnosis, Assessment, and Treatment. Guilford.
17. Clark, D. M., Ehlers, A., Hackmann, A., McManus, F., Fennell, M., Grey, N., Waddington, L., & Wild, J. (2006). Cognitive therapy versus exposure and applied relaxation in social phobia: A randomized controlled trial. Journal of Consulting and Clinical Psychology, 74(3), 568–578. https://ora.ox.ac.uk/objects/uuid:76bb2579-2c73-4188-aea9-d23ea66506b6
18. Craske, M. G., & Barlow, D. H. (2007). Therapist guide: Mastery of Your Anxiety and Panic (excerpt). OUPblog. https://blog.oup.com/2007/07/panic/
19. Craske, M. G., Treanor, M., Conway, C. C., Zbozinek, T., & Vervliet, B. (2014). Maximizing exposure therapy: An inhibitory learning approach. Behaviour Research and Therapy, 58, 10–23. DOI 10.1016/j.brat.2014.04.006
20. Davis, T. E., Ollendick, T. H., & Öst, L.-G. (Eds.) (2012). Intensive One-Session Treatment of Specific Phobias. Springer. https://link.springer.com/book/10.1007/978-1-4614-3253-1
21. de Jonge, P., Roest, A. M., Lim, C. C. W., et al. (2016). Cross-national epidemiology of panic disorder and panic attacks in the World Mental Health Surveys. Depression and Anxiety, 33(12), 1155–1177. https://diposit.ub.edu/dspace/handle/2445/128129
22. Domschke, K., & Zwanzger, P. (2025). Taxonomy of anxiety disorders: A comparison of ICD-10 and ICD-11. Der Nervenarzt. DOI 10.1007/s00115-025-01842-6
23. Dugas, M. J., Gagnon, F., Ladouceur, R., & Freeston, M. H. (1998). Generalized anxiety disorder: A preliminary test of a conceptual model. Behaviour Research and Therapy, 36(2), 215–226.
24. Dugas, M. J., Ladouceur, R., Léger, E., & Freeston, M. H. (2003). Group cognitive-behavioral therapy for generalized anxiety disorder: Treatment outcome and long-term follow-up. Journal of Consulting and Clinical Psychology, 71(4).
25. Eppel, A., & Swinson, R. P. (2024). Anxiety and related disorders in medical settings. McMaster Textbook of Internal Medicine. https://empendium.com/mcmtextbook/chapter/B31.II.21.5.
26. Golovacheva, V. A., & Parfenov, V. A. (2017). Anxiety disorders under the mask of vegetative-vascular dystonia [in Russian]. Meditsinskiy Sovet, (17), 26–30. DOI 10.21518/2079-701X-2017-17-26-30
27. Gupta, A. 5 tips for using interoceptive exposure to face your fears. Anxiety and Depression Association of America. https://adaa.org/learn-from-us/from-the-experts/blog-posts/consumer/5-tips-using-interoceptive-exposure-face-your
28. Harvey, A. G., Clark, D. M., Ehlers, A., & Rapee, R. M. (2000). Social anxiety and self-impression: Cognitive preparation enhances the beneficial effects of video feedback following a stressful social task. Behaviour Research and Therapy, 38(12), 1183–1192.
29. Javaid, S. F., Hashim, I. J., Hashim, M. J., Stip, E., Samad, M. A., & Al Ahbabi, A. (2023). Epidemiology of anxiety disorders: Global burden and sociodemographic associations. Middle East Current Psychiatry, 30, 44.
30. Johns Hopkins Bloomberg School of Public Health (2014). Talk therapy — not medication — best for social anxiety disorder, large study finds. https://publichealth.jhu.edu/2014/talk-therapy-not-medication-best-for-social-anxiety-disorder-large-study-finds
31. 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. Lancet Psychiatry, 1(5), 368–376. https://pubmed.ncbi.nlm.nih.gov/26361000/
32. Newby, J., & Norton, P. (2018). Transdiagnostic and disorder-specific anxiety disorder treatments have equivalent treatment effects. The Mental Elf. https://nationalelfservice.net/mental-health/anxiety/transdiagnostic-and-disorder-specific-anxiety-disorder-treatments-have-equivalent-treatment-effects
33. NICE (2011, updated to 2022). Generalised anxiety disorder and panic disorder in adults: management (CG113). https://www.nice.org.uk/guidance/cg113
34. NICE (2012). Evidence Update 22: Generalised anxiety disorder and panic disorder. https://www.nice.org.uk/guidance/cg113/evidence
35. NICE (2013). Social anxiety disorder: recognition, assessment and treatment (CG159). https://www.nice.org.uk/guidance/cg159
36. Nordahl, H. M., Borkovec, T. D., Hagen, R., et al. (2018). Metacognitive therapy versus cognitive–behavioural therapy in adults with generalised anxiety disorder. BJPsych Open, 4, 393–400. https://doaj.org/article/f7bdb3ec667f4028b12908bd56a16fcb
37. Öst, L.-G. (1989). One-session treatment for specific phobias. Behaviour Research and Therapy, 27(1), 1–7.
38. Pompoli, A., Furukawa, T. A., Imai, H., Tajika, A., Efthimiou, O., & Salanti, G. (2016). Psychological therapies for panic disorder with or without agoraphobia in adults: A network meta-analysis. Cochrane Database of Systematic Reviews, CD011004. DOI 10.1002/14651858.CD011004.pub2
39. Pompoli, A., Furukawa, T. A., Efthimiou, O., Imai, H., Tajika, A., & Salanti, G. (2018). Dismantling cognitive-behaviour therapy for panic disorder: A systematic review and component network meta-analysis. Psychological Medicine, 48, 1945–1953. https://boris.unibe.ch/110766/
40. Psychology Tools. Cognitive behavioral model of panic (Clark, 1986); Cognitive behavioral model of social phobia (Clark & Wells, 1995). https://www.psychologytools.com
41. Rayburn, N. R., & Otto, M. W. (2003). Cognitive-behavioral therapy for panic disorder: A review of treatment elements, strategies, and outcomes. CNS Spectrums, 8(5), 356–362.
42. Ruscio, A. M., Hallion, L. S., Lim, C. C. W., et al. (2017). Cross-sectional comparison of the epidemiology of DSM-5 generalized anxiety disorder across the globe. JAMA Psychiatry. https://hcp.hms.harvard.edu/publication/cross-sectional-comparison-epidemiology-dsm-5-generalized-anxiety-disorder-across-globe
43. Salkovskis, P. M. (1991). The importance of behaviour in the maintenance of anxiety and panic: A cognitive account. Behavioural and Cognitive Psychotherapy, 19(1), 6–19.
44. Salkovskis, P. M., Clark, D. M., Hackmann, A., Wells, A., & Gelder, M. G. (1999). An experimental investigation of the role of safety-seeking behaviours in the maintenance of panic disorder with agoraphobia. Behaviour Research and Therapy, 37, 559–574.
45. Schmidt, N. B., & Trakowski, J. (2004). Interoceptive assessment and exposure in panic disorder: A descriptive study. Cognitive and Behavioral Practice, 11, 81–92.
46. Society of Clinical Psychology (APA Division 12). Research-supported psychological treatments: Panic disorder; Generalized anxiety disorder; Social anxiety disorder. https://div12.org
47. Stein, D. J., Lim, C. C. W., Roest, A. M., et al. (2017). The cross-national epidemiology of social anxiety disorder: Data from the World Mental Health Survey Initiative. BMC Medicine, 15. https://doaj.org/article/30cfb5847c0d42be9b821122fa64ac8a
48. The Open University, OpenLearn. Panic attacks: what they are and what to do about them. https://www.open.edu/openlearn/health-sports-psychology/panic-attacks-what-they-are-and-what-do-about-them/content-section-2.1
49. Tunnell, N. C., Corner, S. E., Roque, A. D., Kroll, J. L., Ritz, T., & Meuret, A. E. (2024). Biobehavioral approach to distinguishing panic symptoms from medical illness. Frontiers in Psychiatry, 15. DOI 10.3389/fpsyt.2024.1296569
50. Wardenaar, K. J., Lim, C. C. W., Al-Hamzawi, A. O., et al. (2017). The cross-national epidemiology of specific phobia in the World Mental Health Surveys. Psychological Medicine. https://hcp.hms.harvard.edu/publication/cross-national-epidemiology-specific-phobia-world-mental-health-surveys
51. Wegner, D. M., Schneider, D. J., Carter, S. R., III, & White, T. L. (1987). Paradoxical effects of thought suppression. Journal of Personality and Social Psychology, 53(1), 5–13.
52. Wells, A. (1995). Meta-cognition and worry: A cognitive model of generalized anxiety disorder. Behavioural and Cognitive Psychotherapy, 23(3), 301–320.
53. Wells, A. (1999). A metacognitive model and therapy for generalized anxiety disorder. Clinical Psychology and Psychotherapy, 6(2), 86–95.
54. Wells, A. (2008). Metacognitive Therapy for Anxiety and Depression. Guilford.
55. Westphal, D., Gerlach, A. L., Lang, T., Wittchen, H.-U., et al. (2015). Die Effekte interozeptiver Expositionsübungen in der Kognitiven Verhaltenstherapie von Panikstörung mit Agoraphobie. Verhaltenstherapie, 25(4), 268–276.
56. Wolitzky-Taylor, K. B., Horowitz, J. D., Powers, M. B., & Telch, M. J. (2008). Psychological approaches in the treatment of specific phobias: A meta-analysis. Clinical Psychology Review, 28, 1021–1037.
57. World Health Organization. Anxiety disorders — Fact sheet (2023 and September 2026 versions). https://www.who.int/news-room/fact-sheets/detail/anxiety-disorders
58. Zlomke, K., & Davis, T. E. (2008). One-session treatment of specific phobias: A detailed description and review of treatment efficacy. Behavior Therapy, 39(3), 207–223. DOI 10.1016/j.beth.2007.07.003
59. World Health Organization (2026). ICD-11 for Mortality and Morbidity Statistics, v2026-01: block 6B0 "Anxiety or fear-related disorders", 6B00–6B04. Official browser: https://icd.who.int (could not be opened during checking). Wording checked against the verbatim Find-A-Code mirror: https://www.findacode.com/icd-11/block-1336943699.html
60. James, A. C., Reardon, T., Soler, A., James, G., & Creswell, C. (2020). Cognitive behavioural therapy for anxiety disorders in children and adolescents. Cochrane Database of Systematic Reviews, CD013162. DOI 10.1002/14651858.CD013162.pub2. https://www.cochrane.org/evidence/CD013162/cognitive-behavioural-therapy-anxiety-children-and-young-people
61. NICE (2013). Social anxiety disorder: recognition, assessment and treatment (CG159). Information for the public: Treatment for children and young people. https://www.nice.org.uk/guidance/cg159/ifp/chapter/Treatment-for-children-and-young-people
62. Salkovskis, P. M., & Warwick, H. M. C. (1986). Morbid preoccupations, health anxiety and reassurance: A cognitive-behavioural approach to hypochondriasis. Behaviour Research and Therapy, 24(5), 597–602. DOI 10.1016/0005-7967(86)90041-0. https://www.psy.ox.ac.uk/publications/1028094
63. Tyrer, P., Cooper, S., Salkovskis, P., et al. (2014). Clinical and cost-effectiveness of cognitive behaviour therapy for health anxiety in medical patients: A multicentre randomised controlled trial. The Lancet, 383(9913), 219–225 (online 18 Oct 2013). DOI 10.1016/S0140-6736(13)61905-4. Summary: The Mental Elf, https://www.nationalelfservice.net/mental-health/anxiety/the-champ-rct-finds-that-cbt-is-more-effective-than-standard-care-for-reducing-health-anxiety/

If you need help now

If your life is in danger, call your local emergency number: 112 in most of Europe, 911 in the US and Canada, 999 in the UK, 000 in Australia. For support lines by country, see Help now.

Image credits

This page uses no third-party images.