Critical Psychiatry Textbook, Chapter 10: Anxiety Disorders


Editor’s Note: Over the next several months, Mad in America is publishing a serialized version of Peter Gøtzsche’s book, Critical Psychiatry Textbook. In this blog, he critiques the way the textbooks discussed treatment for anxiety, panic attacks, and OCD. Each Monday, a new section of the book is published, and all chapters are archived here.

Although the textbooks advised psychotherapy for anxiety disorders, their focus was on drugs.

One book noted that cognitive behavioural therapy is the best documented intervention for anxiety disorders in children, possibly supplemented with an SSRI, and it recommended this also for obsessive-compulsive disorder (OCD).19:143,19:167 Drugs were only indicated if there was a lack of effect of psychosocial support and cognitive behavioural therapy and where the anxiety was severely invalidating.19:157

This is the standard script for psychiatry. Even when drugs are generally not recommended, they must always be used if the condition is severe enough.

Closeup photo of pills on a sheet that reads "Anxiety Disorder"Another book noted that psychotherapy is the preferred treatment of agoraphobia,18:136 while a third book was more positive towards drugs. It noted that agoraphobia can be treated with cognitive behavioural therapy, depression pills, or both in combination.16:349 For social phobia and generalised anxiety,16:351,16:357 cognitive behavioural therapy was considered first choice, but it was mentioned that SSRIs also have well documented effects.

There is no doubt that shyness—now called social phobia, as it is better for industry’s marketing7:214—should be treated with psychotherapy. A 24-week trial that randomised 375 patients with social phobia to sertraline or to gradual exposure to the feared symptoms found a similar effect of exposure and sertraline, but during an additional six-month follow-up, the exposure group continued to improve, which the sertraline group did not.545 This was expected. People on drugs don’t learn anything about how to cope with their anxiety. It is like alleviating the tension with alcohol. In contrast to drugs, psychotherapy usually has enduring effects on psychiatric disorders.180,497-501

A Cochrane review of 41 trials in children and adolescents with anxiety showed very large effects from cognitive behavioural therapy.546 The outcome was assessed blindly in 32 of the 41 trials. The odds ratio for remission, compared with waiting-list controls, was 7.85 (5.31 to 11.60), and the reduction in anxiety symptoms had an effect size of -0.98 (-1.21 to -0.74). Other psychological therapies were similarly effective.

A Cochrane review of anxiety and depressive disorders did not find a difference between the results obtained by paraprofessionals and professionals (psychiatrists or psychotherapists), effect size 0.09 (-0.23 to 0.40).547 These results agree with those from numerous other studies.14,547 Patients can also help themselves. A Cochrane review of self-help where printed materials, audio or video recordings, computers, or the internet were used to teach adult patients behavioural or cognitive behavioural therapy for anxiety found a clear effect compared with no intervention, effect size 0.67 (0.55 to 0.80).548

For OCD, a book recommended psychoeducation, self-help material, and cognitive behavioural therapy, and SSRIs in severe cases.16:360 The evidence for psychotherapy is strong. A Cochrane review of trials in adults found that psychotherapy resulted in far fewer symptoms than if the patients had received treatment as usual, effect size -1.24 (-1.61 to -0.87).549 The effect of SSRIs was substantially smaller, effect size -0.46 (-0.55 to -0.37) (calculated by me).550 There were few direct comparisons, but a review found that psychotherapy was better than depression pills, effect size -0.36 (-0.72 to 0.00) (calculated by me, three trials with 118 patients).551

The book that recommended SSRIs in severe cases also stated that psychosis pills could possibly be used as augmentation.16:360 There were no references to original research, 16:369 only to a national guideline from 2007,552 which was very brief: “There are no studies of monotherapy with antipsychotics that have shown an effect on OCD. Various open and a few double-blind placebo-controlled studies of the effect of combination therapy with a serotonergic antidepressant and an antipsychotic (p. 162) have been carried out. On this background, it is concluded that there is some evidence that risperidone and quetiapine may have an effect in augmentation treatment of OCD.”

The guideline referred to page 162 in a NICE report, which was not illuminating.553 It describes a few small trials and there was no systematic review of these trials.

It is difficult to understand the thinking behind the weak conclusion that risperidone and quetiapine “may” have an effect based on “some evidence.” Severe OCD can ruin the lives not only for the patients but also for their relatives, but it is not a deadly disease. In contrast, psychosis pills are some of the deadliest drugs ever invented (see page 46), apart from cancer chemotherapy, and they should be avoided, also for patients with OCD.

As SSRIs double the risk of suicide and have many other important harms, these pills should also be avoided. The book noted that SSRIs and SNRIs may increase anxiety, and that it takes longer than for depression before they work, but that there is continued improvement after several months.16:368 To say that it takes longer than for depression before they work means they don’t work, but the psychiatric mindset doesn’t allow such admissions.

On the same page, this book offered horrible advice also about benzodiazepines.16:368 It mentioned that a study had found an effect after years of treatment, especially with alprazolam and clonazepam, but that generally only a few weeks of treatment is recommended while treatment with a depression pill is started. Alprazolam is a very harmful drug. After a few weeks, many people have become dependent on it, and the rebound effect when it is stopped is so pronounced that the patients become worse than they were when they started therapy.5:295

This book also claimed that pregabalin, an antiepileptic, works and is approved for anxiety disorders. It is bad medicine to use antiepileptics for anxiety given their many serious harms, including a doubling of the suicide risk.390,439

The literature list did not provide support to the harmful recommendations launched by a psychiatrist as sole author.

Further ahead, another author wrote more soberly about benzodiazepines, contradicting the first author:16:585 The information on the anxiolytic effect is conflicting and there is a lack of long-term studies. Furthermore, there is development of tolerance (the effect vanishes over time), and cessation causes physical abstinence symptoms, including anxiety, restlessness, irritability, difficulty sleeping, tremor, photo- and phonophobia, flu-like symptoms and rebound phenomena.

Even though the abstinence symptoms are very much the same for SSRIs and SNRIs as for benzodiazepines (see Chapter 8, Part Three),554they were not called abstinence symptoms in any of the textbooks.

The psychiatrist author wrote that when stopping, 10-20% of the starting dose should be removed every other week, but in the last part of withdrawal it may be necessary to reduce with even smaller doses and extend the intervals.16:586 Another book had similar advice, a dose reduction of 10-20% with 1-2 weeks intervals and possibly even slower in the final phase.18:71

It is of utmost importance that the dose reductions are much smaller by the end of a taper-ing.281 But none of the books explained that the binding curves for psychiatric drugs are hyperbolic (as I will explain in Chapter 15), and that the tapering therefore needs to be exponential. This is unfortunate, as very few doctors know about correct withdrawal and cause terrible harms by withdrawing the drugs much too quickly.

A third book recommended either cognitive behavioural therapy or SSRIs for social phobia, for 6-12 months, if there is effect.18:136 It noted that benzodiazepines should not be used long-term due to dependence, and because abstinence symptoms can be difficult to distinguish from the primary anxiety symptoms. It is true that rebound anxiety is a very common abstinence symptom. But why was the same not said about SSRIs and SNRIs in any of the textbooks? The same problems occur,554 but the authorities also failed badly, as they ignored the dependence problems with depression pills for two decades.304

In the same book, other authors contradicted this, as they said that benzodiazepines are used long term for anxiety such as panic attacks when cognitive behavioural therapy or depression pills have not had sufficient effect.18:240 This is horrible advice.

The two remaining books escalated the confusion. Two psychologists claimed that SSRIs and cognitive behavioural therapy should often be combined to get the best result in OCD and that most studies had shown remission in 60% of the patients,20:485 a meaningless statement, as there is no control group. There were 47 references but none of them were about the effect of SSRIs.

The fifth book contradicted this, noting that, according to the National Board of Health,555 the effect is not increased by adding depression pills to psychotherapy,17:420 which, also considering the harms, was the reason the Board does not recommend pills.

The authors noted that, on SSRIs, 60-70% will experience a 50% reduction in panic symptoms and 50-60% will have an effect on social phobia, but they added that 60% will experience an effect of placebo on panic attacks.17:404 What is the reader supposed to conclude based on this?

The authors also claimed that, according to a meta-analysis, about half of the patients with OCD will come in remission but none of their 13 references were to a meta-analysis.17:420 There was a reference to an article about escitalopram, but it was irrelevant and there was no mention of it in the text.556

From then on, it became worse in this book.17:423 The authors spoke about extensive evidence for the effect of SSRIs and that we should try another one or increase the dose beyond the maximum (in rare cases) if the effect is insufficient. We could also add a small dose of a psychosis pill, which is effective according to clinical experience. But they added that the National Board of Health says that no clinically relevant effect has been shown and that there is risk of harms and that, in some cases, psychosis pills can cause or worsen OCD.

This is confusing and contradictory, and the authors felt that clinical experience is more important that advice from the Board of Heath. Furthermore, it seems that the 2007 national guideline for anxiety disorders552 is in conflict with the one specifically for OCD.555 The guideline for OCD was updated in 2019. During these 12 years, the apparent effect of psychosis pills in 2007 disappeared:

“As there is insufficient evidence from the paediatric literature on augmentation therapy with antipsychotics, the question is solely addressed in adults … Use only after careful consideration an atypical antipsychotic as augmentation therapy for adults with severe OCD who have had no effect of treatment with cognitive behavioural therapy and antidepressants (SSRIs), as no clinically relevant effect has been demonstrated and as there is a risk of side effects.”

This textbook opined that the primary drug treatment for anxiety is depression pills.17:664 Benzodiazepines should not be used for more than four weeks but can be used for longer, e.g. if the patient has panic attacks. We are even told that pregabalin can be used because the harms are relatively mild. Antiepileptics have many harms, one of which is to double the risk of suicide.390,439

This horrible and harmful advice suggests that patients with anxiety disorders should avoid seeing a psychiatrist. It is too dangerous.


To see the list of all references cited, click here.



Mad in America hosts blogs by a diverse group of writers. These posts are designed to serve as a public forum for a discussion—broadly speaking—of psychiatry and its treatments. The opinions expressed are the writers’ own.


Mad in America has made some changes to the commenting process. You no longer need to login or create an account on our site to comment. The only information needed is your name, email and comment text. Comments made with an account prior to this change will remain visible on the site.


  1. “This horrible and harmful advice suggests that patients with anxiety disorders should avoid seeing a psychiatrist. It is too dangerous.” Lol, but I agree, today it’s “too dangerous” for anyone to see a psychiatrist, for any reason. Thank you for your truth speaking, Dr. Peter.

    Report comment

  2. I am almost 7 years post psychiatry and still recovering from almost three decades of lies and carnage inflicted by the Canadian mental health system.
    It was your research that gave me the strength I needed to stand behind my decision to run for my life, from the mental health system. I was in a rapid decline, with very little cognitive range left. At one time I was an A-student and I held on to the importance of good research and I didn’t let go.
    I laugh out loud now. You were a huge help in this exodus and journey back to myself. I have life again.
    Many thanks- Sherri

    Report comment