Critical Psychiatry Textbook, Chapter 8: Depression and Mania (Affective Disorders) (Part Six)


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 describes the way the pharmaceutical industry, and the drug regulators, spun the results of clinical trials to hide suicide attempts and deaths on depression pills. Each Monday, a new section of the book is published, and all chapters are archived here.

Concealing suicide and homicide: fraud, organised crime, and FDA’s complicity

Fluoxetine (Prozac or Fontex) was the first SSRI that came into widespread use. The story behind it is a grim one which students of psychiatry should know about, also because this drug’s approval paved the way for a host of similar drugs. It illustrates that marketing trumps science totally in influencing if doctors use drugs, and if so, which drugs they use.2,6:202

Illustration depicting a pencil eraser erasing a person from a row of businessmen in suitsFluoxetine is such a terrible drug that senior management in Eli Lilly wanted to shelve it after having considered marketing it for eating disorders.2 But Lilly was in serious financial trouble, and if fluoxetine failed, Lilly could “go down the tubes.”197,343,344

The FDA noted serious flaws in Lilly’s trials.2 Patients who didn’t do well after two weeks had their code broken, and if they were on placebo, they were switched to fluoxeine.345 In this way, six weeks of fluoxetine was compared to two weeks on placebo. It also turned out that 25% of the patients had taken an additional drug, and when the FDA in 1985 removed patients on benzodiazepines and other drugs from Lilly’s trials, there was no significant effect of fluoxetine.

The FDA went to extremes to make it look like fluoxetine worked.345 Perhaps the fact that Lilly is an American company played a role. Fluoxetine was approved when Bush senior was president and he had been a member of the board of directors of Lilly. Vice President Dan Quayle was from Indiana where Lilly’s headquarters are, and he had former Lilly personnel on his own staff and sat on an FDA oversight committee.21

The German drug regulator considered fluoxetine “totally unsuitable for the treatment of depression” and furthermore noted that, according to the patients’ self-ratings, there was little or no response, in contrast to doctors’ ratings.2,5,346 This is also the case for other depression pills, and also for children. When the patients evaluate the effect themselves, it is non-existent (effect size 0.05 or 0.06).347,348,349 Only the psychiatrists think they work (effect sizes 0.25 to 0.29) but they are not the ones to be treated.

When Lilly showed some of its data to Swedish psychiatrists, they laughed and didn’t think Lilly was serious.350 But it was crucial to get fluoxetine approved in Sweden, as it would then be easier to get it approved by the FDA. Lilly’s Swedish director, John Virapen, invited doctors to the Caribbean for a week, with plenty of relaxation, including “diving, surfing, sailing, pretty girls and hot nights.”350 He came to Copenhagen to visit me to tell me more about this than he published in his book,350 and official documents confirm his story.7:59

By planting indirect questions to the secretaries of prominent psychiatrists, Virapen identified the independent expert, psychiatrist Anders Forsman, who was going to examine the clinical documentation for the Swedish drug agency. Forsman was one of those who had laughed about the idea of ever getting fluoxetine approved, but already at their second meeting, he suggested $20,000 as a reasonable sum for a speedy approval, which, moreover, shouldn’t become known to the taxman but was to be handled by Lilly’s office in Genève. He furthermore demanded a good deal of research money. The money was split and the second half was to be paid when the drug was approved. This is how the mob operates when it orders a murder.

Forsman even suggested falsifying the registration application, e.g. suicide attempts were called “miscellaneous effects,” and he placed his own personal letter of recommendation.

As the criteria for the depression diagnosis were much more stringent and relevant than today, there weren’t many depressed people at the time, and fluoxetine was therefore marketed as a mood lifter, like street pushers sell cocaine.

The approval in Germany also followed “unorthodox lobbying methods exercised on independent members of the regulatory authorities,” as Virapen called it.350 After having been so enormously helpful to Lilly, Virapen was fired. This is also like in the mob. When a lower-ranked person has been asked to murder a well-known political figure, it is safest to kill the assassin afterwards. The official explanation was that Lilly had certain ethical principles. When journalists ask me what I think of the ethical principles of the drug industry, I say I have no answer as I cannot describe what doesn’t exist. The only industry principle is money, and the worse the crime, the more money will be earned.6

Forsman’s name became known in the press, but he just went on and came to work for the court, as a psychiatric assessor for Sweden. Virapen tried to prosecute him, but that wasn’t possible because he wasn’t an employee of the health authority. After this affair, the Swedish anti-corruption law was amended.

Lilly turned their awful drug, which they didn’t even like themselves, into a blockbuster, which contributed to making the company one of the world’s ten biggest.

Lilly promoted fluoxetine illegally for several non-approved ailments, e.g. shyness, eating disorders, and low self-esteem, and concealed that the drug causes suicide and violence.2,122,351

In 1990, only two years after fluoxetine came on the market, Martin Teicher et al. described six patients who had become suicidal and reacted in bizarre ways with intense, violent suicidal preoccupation while receiving the drug, which was something completely new to them.352 Teicher’s observations were very convincing. Later, however, internal Lilly documents that came to light during a litigation case353 revealed that the FDA worked with Lilly on the suicide issue. The psychiatrists Lilly had corrupted came in handy while Lilly’s own scientist left out information at the subsequent 1991 FDA hearings that demonstrated that fluoxetine increases the risk of suicide.122 Earlier, Lilly had submitted data to the German drug agency showing that suicide attempts almost doubled on fluoxetine compared to placebo.

The chair of the FDA committee, psychiatrist Daniel Casey, brutally interrupted Teicher so that he couldn’t present his findings and reasons. He was only allowed to present a few slides while Lilly staff presented many. A few years later, Teicher’s wife was offered a job at Lilly as their top scientist in oncology without having applied, which she accepted.

in 2004, the BMJ received a series of internal Lilly documents and studies on fluoxetine from an anonymous source, which had been available ten years earlier in a litigation case.353 They revealed that Lilly had known since 1978—ten years before fluoxetine came on the market—that fluoxetine can produce in some people a strange, agitated state of mind that can trigger in them an unstoppable urge to commit suicide or murder.344 In 1985, two years before fluoxetine was approved, the FDA’s safety reviewer noted under the headline “Catastrophic and serious events” that some psychotic episodes had not been reported by Lilly but were detected by the FDA by examining case reports on microfiche. The reviewer noted that fluoxetine’s profile of adverse effects resembled that of a stimulant drug, which might be the reason why Lilly marketed fluoxetine as a mood lifter.

Already in 1985, an in-house analysis of placebo-controlled trials found 12 suicide attempts on fluoxetine versus one on placebo, but after the code was broken, Lilly’s hired consultants threw out six of the attempts on fluoxetine.111:258

Lilly was keen to root out the word “suicide” altogether from its database of adverse events experienced by patients and suggested that, when doctors reported a suicide attempt on fluoxetine, Lilly staff should code it as “overdose.”

Lilly’s fraud was second to none. It is hardly possible to kill yourself by overdosing fluoxetine, and the suicides occur on normal doses. Furthermore, Lilly excluded 76 of 97 cases of suicidality on fluoxetine in a post-marketing surveillance study it submitted to the FDA.354,355 Lilly instructed its staff to code “suicidal ideation” as “depression,”197 which is the usual script for drug companies, drug regulators, and psychiatrists: Blame the disease, not the drug.7:208

Lilly also kept completed suicides from public view. In 2004, the body of a 19-year-old college student was found hanging by a scarf from a shower rod in an Indianapolis laboratory run by Lilly.354 She had entered a clinical study as a healthy volunteer in order to help pay her college tuition after having undergone thorough medical testing to screen out depression or suicidal tendencies. She had taken duloxetine, another Lilly drug. When a BMJ journalist, Jeanne Lenzer, filed Freedom of Information Act requests for all safety data related to duloxetine she received a database that included 41 deaths and 13 suicides. Missing from the database was any record of the college student and at least four other volunteers known to have committed suicide while taking duloxetine for depression.354

One of the leaked documents noted that in clinical trials, 38% of fluoxetine-treated patients reported new activation compared to only 19% of placebo-treated patients. Activation may lead to agitation or akathisia, and Lilly recommended early on that, in their trials of fluoxetine, such patients should also take benzodiazepines,2 which reduce the symptoms. We therefore don’t know what the true extent of akathisia is. Other companies adopted the same strategy, and minor tranquillisers were permitted in 84% of placebo-controlled trials of depression pills.356

Lilly’s widespread criminal activities and corruption of doctors worked. In 1997, Prozac was the fifth most prescribed drug in the United States.357 It also became the most complained-about drug.1:287

In relation to lawsuits, Healy found early drafts of Prozac’s package insert that stated that psychosis might be precipitated in susceptible patients by depression pills.357 The warning about psychosis wasn’t included in the final package insert for the United States, and is not even included today,33 whereas the German drug agency required it. By 1999, the FDA had received reports of over 2000 Prozac-associated suicides and a quarter of the reports specifically referred to agitation and akathisia.2:171 As always, the FDA protected the drug and not the patients, as it said it would not have allowed a company to put a warning about akathisia or suicide on the label; it would have considered it mislabelling.357

Other companies also indulged in fraud and organised crime.6:208 SmithKline Beecham, later merged into GSK, started marketing paroxetine (Paxil or Seroxat) in 1992 and falsely claimed for the next 10 years that it wasn’t habit forming358 even though the licence application showed that paroxetine leads to withdrawal reactions in 30% of the patients.359 The UK drug regulator also denied there was a problem whereas the BBC reported in 2001 that WHO had found Paxil to have the hardest withdrawal problems of any depression pill. Until 2003, the UK drug regulator propagated the falsehood that SSRIs are not addictive, but the same year, WHO published a report that noted that three SSRIs (fluoxetine, paroxetine and sertraline) were among the top 30 highest-ranking drugs for which drug dependence had ever been reported.307

The UK drug regulator also misrepresented the data when it described withdrawal reactions as generally being rare and mild. Independent researchers showed that the reactions had been classified as moderate in 60% of the cases and as severe in 20% by the same UK regulator that told the public that they were mild.360

In 2003, GSK quietly and in small print revised its previous estimate of the risk of withdrawal reactions in the prescribing instructions from 0.2% to 25%,307 a 100-times increase.

From 2002 onwards, the BBC presented four documentaries about SSRIs in its Panorama series, the first one called Secrets of Seroxat. The GSK spokesperson, doctor Alastair Benbow, lied in front of a running camera. He denied that paroxetine could cause suicidality or self-harm, while he sent data to the drug regulator one month later that showed exactly this, and which immediately led to a ban on using the drug in children.

The drug regulator claimed that this information was completely new to GSK, which, however, had known about it for ten years. In addition, the head of the drug agency echoed the drug companies’ false assertion that it was the disease, not the drug, that caused the suicidal events.

US senator Charles Grassley asked GSK for how long the company had known that paroxetine carried a suicide risk.361 GSK lied when it wrote back that they “detected no signal of any possible association between Paxil and suicidality in adult patients until late February 2006.” Government investigators found that the company had the data in 1998 and Healy found evidence in internal company documents that 25% of healthy volunteers experienced agitation and other symptoms of akathisia while taking paroxetine.357

Healy performed a study of sertraline in 20 healthy volunteers, and to his big surprise two of them became suicidal.2:179 One was on her way out the door to kill herself in front of a train or a car when a phone call saved her. Both volunteers remained disturbed several months later and seriously questioned the stability of their personalities.

Pfizer’s own studies in healthy volunteers showed similar deleterious effects, but they hid most of the data in company files.

Drug regulators also hid the lethal harms. When FDA reviewers and independent researchers had found that the drug companies had concealed cases of suicidal thoughts and acts by labelling them “emotional lability,” the FDA bosses suppressed this information.2,362 When the FDA’s own safety officer, Andrew Mosholder, concluded that SSRIs increase the suicide risk among teenagers, the FDA prevented him from presenting his findings at an advisory meeting and suppressed his report. When the report was leaked, the FDA’s reaction was to do a criminal investigation into the leak.355,363

There were other types of fraud. In data submitted by GSK to the FDA in the late 1980s and early 1990s, the company had added suicide attempts from the washout period before the patients were randomised to the results for the placebo group, but not to those for the paroxetine group. At least three companies, GSK, Lilly, and Pfizer, added cases of suicide and suicide attempts in patients to the placebo arm of their trials, although they didn’t occur while the patients were randomised to placebo.2,141,353,364,365

Healy wrote in 2002364 that, based on data he had obtained from the FDA, three of five suicide attempts on placebo in a sertraline trial366 had occurred during washout rather than on placebo and that two suicides and three of six attempts on placebo in a paroxetine trial366 had also occurred in the washout period. Healy’s observations weren’t denied by Pfizer and GSK,367,368 but GSK provided another glaring example that their lies are not of this world:368

The “drug” v. “true placebo” analysis Dr Healy describes is not only scientifically invalid, but also misleading. Major depressive disorder is a potentially very serious illness associated with substantial morbidity, mortality, suicidal ideation, suicide attempts and completed suicide. Unwarranted conclusions about the use and risk of antidepressants, including paroxetine, do a disservice to patients and physicians.

The systematic fraud can be important for the companies in court cases. In 2001, when a man on paroxetine had murdered his wife, daughter, and granddaughter and committed suicide, GSK said in its defence that its trials didn’t show an increased risk of suicide on paroxetine.369 This seemed to be incorrect. In 2004, a researcher published a meta-analysis based on the full reports of GSK’s trials that were made available on the Internet as a result of litigation. He found that paroxetine increased significantly suicidal tendencies, odds ratio 2.77 (1.03 to 7.41).370

The clinical study reports we analysed also included trials in adults.326 We could not address the harms fully because some of them appeared only in patient listings in appendices, which we had for only 32 of our 70 included trials. Furthermore, we didn’t have case report forms. But we found many alarming events, which you will never see in medical journals and here are some:

Four deaths were misreported by the company, in all cases favouring the active drug.

A patient receiving venlafaxine attempted suicide by strangulation without forewarning and died five days later in hospital. Although the suicide attempt occurred on day 21 out of the 56 days of randomised treatment, the death was called a post-study event as it occurred in hospital and the drug had been discontinued because of the suicide attempt.

Although patient narratives or individual patient listings showed they were suicide attempts, 27 of 62 such attempts were coded as emotional lability or worsening depression, which is what you see in the publications, not the suicide attempts.

A suicide attempt (intentional overdose with paracetamol in a patient on fluoxetine) was described in the adverse events tables as “elevated liver enzymes,” which you can get if you drink alcohol.

For Eli Lilly’s drugs, fluoxetine and duloxetine, we compared our findings with the summary trial reports on the company’s website. Lilly’s reporting was seriously misleading.8,326 In most cases, adverse events were only shown if they occurred in, for example, at least 5% of the patients. In this way, the companies may avoid reporting many serious harms. Only 2 of 20 suicide attempts (17 on drug, 3 on placebo) were documented. None of 14 suicidal ideation events (11 vs 3) were mentioned, and only 3 akathisia events (15 vs 2) were mentioned.

In three sertraline trials where we had access to both the verbatim and the coded preferred terms, akathisia was coded as “hyperkinesia,” and miscoding seemed to have been prevalent also in paroxetine trials since we didn’t find a single case of akathisia.

As explained earlier, akathisia increases the risk of suicide, violence, and homicide. We could only identify akathisia if we had access to the verbatim terms, but we nonetheless found that, like aggression, akathisia was seen twice as often on the pills than on placebo.326

It is of particular relevance for the many school shootings that the following events for 11 patients on a depression pill were listed under aggression in patient narratives for serious adverse events: homicidal threat, homicidal ideation, assault, sexual molestation, a threat to take a gun to school, damage to property, punching household items, aggressive assault, verbally abusive and aggressive threats, and belligerence.

Many of the killers were on depression pills. The authorities routinely hide such information in order not to raise concerns about the safety of the pills, and it therefore took quite a while before we learned that the Germanwings pilot that took a whole plane with him when he committed suicide in the Alps, and that the Belgian bus driver who killed many children by driving his bus into a wall, also in the Alps, were both on a depression pill.


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. Mental illness is complex and varies from person to person. Medication while under a doctors care can be apart of a range of coping tools. Take personal responsibility and radical acceptance of your life . Life is a continual process. I wish you all well.

    Report comment

    • “Take personal responsibility” is in general a good rule when it is life’s misfortunes (like getting cancer or being born with no legs) that are the problem. You can’t blame anyone for that.

      But saying “take personal responsibility” as though pointing out the damage that psychiatry/psychiatrists (even the mentality of other patients) have done to people is like asking the victim of a crime to take responsibility for what the criminal has done to them. There is no “personal responsibility” on the part of those people here. Seldom do you see psychiatrists and psychiatric supporters talk about the damaging aspects of psychiatry.

      People’s problems are varied like you said. Correspondingly, the solutions psychiatrists dish out to different people also work/help or don’t work/damage. And it is not “some incompetent professionals which exist in all fields”. It is pretty much the standard practices of the field I’m talking about, no matter the competence of the professional involved.

      Report comment

  2. It is not a minor issue if scientific studies are used as a marketing channel for products because there is worldwide trust in the scientific method. If there is an issue with that method, then any argument based on the idea of a trusted source becomes meaningless.

    Using a trusted source as proof is common among those with a scientific worldview to distinguish fact from fiction. It is even employed on platforms like Wikipedia, and in debates, it takes the form of “The Scientific Worldview Proof,” which is a fallacy presented as follows:

    “The scientific method always leads to the correct answer based on evidence. ⇔ The majority of psychiatrists believe that psychiatric medicines help patients. Therefore, it implies that the majority of psychiatrists likely speak the truth, while those opposing it are likely ignorant.”

    Individuals with a scientific worldview can provide examples from any scientific field to support the notion that some make irrational claims, implying that only the majority can be trusted. This indirectly suggests that individuals like Gøtzsche or Moncrieff are akin to those making unreasonable statements.

    Those using the scientific worldview proof can dismiss all competing evidence as meaningless without even having to read or argue against it. They may argue that while the majority is not always correct, it is highly likely, thus reinforcing their position. This proof is based on the idea of a trusted source derived from the trusted method.

    This kind of proof is challenging to counter because it is often an approximation and a rule of thumb that leads to the correct answer. Some employ it regularly to combat climate denialists and the promotion of dangerous natural treatments, such as silver water or nut bars marketed to individuals with nut allergies.

    It often leads to ad hominem arguments because individuals who hold a scientific worldview and are proud of it tend to mock and ridicule anyone claiming something contrary to their method of proof, suggesting that the opposing party does not understand science as they do. This behavior aims to protect the scientific community.

    To definitively counter the scientific worldview proof, one must first disprove the proposition that the scientific method always leads to the correct answer. This would require conducting trials and finding a way to identify cases where the majority in certain scientific fields cannot be reasonably trusted.

    However, this is not an easy task as it necessitates deriving a common, testable, and proven theory from specific cases, such as the marketing of fluoxetine and other psychiatric drugs, and refining it through trials.

    Does fraud always occur when the majority of research is conducted by those producing a product, or are there other contributing factors? What are those variables?

    If a washing machine manufacturer conducts research on their product, why is it considered more reliable than research conducted by Eli Lilly? Is it because marketing of washing machines is easier to recognize and falsify, or is it because washing machines are developed for the needs of the user, while psychiatric drugs are developed for the needs of authorities? If that is the case, why don’t all other cases where doctors make decisions for the benefit of the user exhibit a similar tendency towards fraud? What variables can be observed to confidently distinguish these cases?

    These would be genuinely fascinating research subjects since the researcher would need to test numerous scenarios and identify suitable trials to develop a consistent theory that could serve as a simple yet accurate tool. There is already a considerable amount of research on biases in psychology that can provide a foundation.

    However, it is not easy to convince decision-makers to recognize the necessity for this kind of research to accurately identify cases with a high degree of confidence. It would require a group of individuals responsible for budget and research goals to acknowledge the global problem concerning the reliability of certain scientific fields and deem it significant enough to warrant investment.

    Report comment

    • There is also a big difference between those who use scientific EVIDENCE as proof of their position and those who use scientific CONSENSUS (what ‘scientists’ tend to agree on) as proof of their positions. As has become very clear, many of the so-called “scientists” in the psychiatric field not only don’t rely on science for their conclusions, they actively attack any real science that contradicts their desired narrative. Corrupt “scientists” who benefit from their position of power should not be accorded any kind of special status. Only the science itself is worthy of discussion if we’re really interested in facts and knowledge!

      Report comment