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 continues to detail the way the pharmaceutical industry and the drug regulators hide the increase in suicide attempts and deaths on depression pills. Each Monday, a new section of the book is published, and all chapters are archived here.
In 2014, ten years after the FDA had issued a black-box warning on depression pills because the rate of suicidal thinking or behaviour was twice as high among young patients on a depression pill as on placebo,303,337 a psychiatrist argued in New England Journal of Medicine that the FDA should consider removing the warning entirely.337
His arguments were untenable. He found it disturbing that the warning had decreased the use of depression pills also in adults, “for whom there is solid evidence of a positive effect of anti-depressant medication on suicide risk.” As we shall see, the truth is the opposite.
He opined that “the risk posed by untreated depression—in terms of morbidity and mortality—has always been far greater than the very small risk associated with antidepressant treatment. We need to better educate physicians, to help them understand that although they cannot ignore that small risk, they can safely manage it by carefully monitoring their patients, particularly children and adolescents, during pharmacotherapy.”
It is typical for the journal, which is so beholden to drug companies that it is nicknamed the New England Journal of Medicalisation, to publish such nonsense. The harms are far greater than the benefits, which are invisible, and the suicide risk cannot be safely managed. Many children and young people have committed suicide by violent means, e.g. hanging, while their parents or peers had no idea that they were endangered.2,7:79
But this is how psychiatrists and drug regulators think. In 2007, the FDA humbly “proposed” to the drug makers that they update their black box warning:7,371
“All patients being treated with antidepressants for any indication should be monitored appropriately and observed closely for clinical worsening, suicidality, and unusual changes in behavior, especially during the initial few months of a course of drug therapy, or at times of dose changes, either increases or decreases. The following symptoms, anxiety, agitation, panic attacks, insomnia, irritability, hostility, aggressiveness, impulsivity, akathisia (psychomotor restlessness), hypomania, and mania, have been reported in adult and pediatric patients being treated with antidepressants.”
The FDA also noted that, “Families and caregivers of patients should be advised to look for the emergence of such symptoms on a day-to-day basis, since changes may be abrupt.”
The FDA finally admitted—after 20 years of foot dragging—that SSRIs can cause madness at all ages and that the drugs are very dangerous; otherwise daily monitoring wouldn’t be needed. But since this is a fake fix, the FDA, instead of “proposing” label changes, should have taken the drugs off the market.
The FDA also admitted, at least indirectly, that depression pills increase the suicide risk in adults, too.
Three years earlier, in 2004, the FDA had issued a warning that depression pills can cause a cluster of activating or stimulating symptoms such as agitation, panic attacks, insomnia and aggressiveness.353 Such effects were expected, as fluoxetine is similar to cocaine in its effects on serotonin. However, when the EMA in 2000 continued to deny that the use of SSRIs leads to dependence, it nonetheless stated that SSRIs “have been shown to reduce intake of addictive substances like cocaine and ethanol. The interpretation of this aspect is difficult.”372 It is only difficult for those who are so blind that they will not see.
It has been difficult to demonstrate the danger of depression pills because many suicidal events are missing in the trials.2,6,7 This has been demonstrated by the FDA itself. When the FDA in 2006 published its meta-analysis of 100,000 patients who had received depression pills or placebo in randomised trials, after having asked the companies how many suicides they had, the suicide rate on the pills was 1 per 10,000 patients.7,303
Five years earlier, Thomas Laughren, who chaired the large FDA meta-analysis, had published his own meta-analysis of the drugs, based on data in FDA’s possession, and this time the suicide rate on pills was 10 per 10,000 patients, or 10 times as many.373 Laughren interpreted his findings in a dishonest way: “There is obviously no suggestion of an excess suicide risk in placebo-treated patients.” Surely not, but there were four times as many suicides—not just suicidal thoughts—on depression pills than on placebo, which was statistically significant (P = 0.03, my calculation).373 Laughren left the FDA and established Laughren Psychopharm Consulting to help pharmaceutical companies “meet the high standards of FDA and other regulatory agencies.”7:74 He certainly knows how to speak and behave like a drug company.
What is abundantly clear—and which has been demonstrated by many researchers—is that the companies have deliberately concealed many cases of suicide and suicide attempts in their trials and in their reports to the drug regulators.
It is difficult to comprehend discrepancies of this magnitude, but it can be explained. When the FDA asked the companies to adjudicate possibly suicide-related adverse events, the agency didn’t verify if they were correct or if some had been left out. Why would the companies, which had cheated shamelessly earlier about suicidal events caused by their drugs, not continue cheating when they knew that FDA didn’t check what they reported? If they didn’t cheat this time, it would be too obvious how much they had cheated earlier.
Another issue is that the collection of adverse events was limited to within one day of stopping randomised treatment, although stopping an SSRI increases the risk of suicide for several weeks. As I have documented in detail, the huge FDA meta-analysis303 grossly underestimates the risk of suicide.6,7 In trials on some drugs included in FDA’s analysis, there were more suicides than in the whole FDA analysis of all the drugs. For example, a memo from Lilly Germany listed nine suicides in 6,993 patients on fluoxetine in the trials.374 This is a suicide rate 14 times bigger than the five suicides in total in FDA’s analysis of 52,960 patients on SSRI drugs.303
Many suicides disappeared and the data I found were remarkably consistent. There were likely 15 times more suicides on depression pills than reported by the FDA in its large metaanalysis.7:70 This is an error of 1,400%. The fraud has been so massive that it is difficult to comprehend and it has killed many patients all over the world. I consider it a crime against humanity.
Even missing by far most of the suicides and other suicidal events, FDA found that paroxetine increased suicide attempts significantly in adults with psychiatric disorders, odds ratio 2.76 (1.16 to 6.60).303 GSK limited its analysis to adults with depression, but it also found that paroxetine increases suicide attempts, odds ratio 6.7 (1.1 to 149.4).375 GSK USA sent a “Dear Doctor” letter that pointed out that the risk of suicidal behaviour was increased also above age 24.376
Does anyone think that paroxetine is an exception and that all other depression pills do not increase the suicide risk in adults? Apparently, many psychiatrists think so, but this is irrational.
In their submissions to drug agencies, several companies obscured the suicide risk by using patient-years as the denominator instead of the number of randomised patients. This introduced considerable bias because several of the trials had a follow-up phase where all patients could receive the active drug. As those who continue with the drug are those who tolerate it, patient-years are added “for free” to the drug group in terms of suicidality.7:78
In 2016, my research group found that, compared to placebo, depression pills double the occurrence of FDA-defined precursor events for suicide and violence in healthy adult volunteers.377 In 2017, we demonstrated with similar methods, based on unpublished clinical study reports submitted to drug regulators, that duloxetine increased the risk of suicide and violence by 4-5 times in middle-aged women with stress urinary incontinence, and that twice as many women experienced a core or potential psychotic event than those who got placebo.378 Later, the FDA announced that, in the open label extension phase of the randomised trials in urinary incontinence, the suicide attempt rate was 2.6 times higher on duloxetine than for other women of similar age.379
Leading psychiatrists did not like our results and criticised our use of precursor events, but this is a red herring. Precursor events are used throughout medicine, e.g. prognostic factors for heart disease. As smoking and inactivity increase the risk of heart attacks, we recommend people to stop smoking and to start exercising.
Suicide attempts and suicides are not only concealed during the trial. Most often, they are also omitted when they occur just after the randomised phase is over.8:52 When Pfizer in 2009 did a meta-analysis of its trials of sertraline used in adults, they reported a halving of the suicidal events (risk ratio 0.52).380 But when they included events occurring up to 30 days after the trial phase ended, there was an increase in suicidality events of about 50% (risk ratio 1.47).
A 2005 meta-analysis conducted by independent researchers using UK drug regulator data found a doubling in suicide or self-harm when subsequent events were included.381 These researchers noted that the companies had underreported the suicide risk in their trials, and they also found that nonfatal self-harm and suicidality were seriously underreported compared to reported suicides.
Another 2005 meta-analysis was also carried out by independent researchers, but this time of the published trials.382 It found twice as many suicide attempts on drug than on placebo, odds ratio (which is about the same as risk ratio when events are rare) 2.28 (1.14 to 4.55). The investigators reported that many suicide attempts must have been missing. Some of the trial investigators told them that there were suicide attempts they had not reported, while others didn’t even look for them. Further, events occurring shortly after active treatment was stopped were not counted. These researchers found that, for every 1,000 patients treated for one year, there were 5.6 additional suicide attempts on active drug compared to placebo (all ages). Thus, by treating 179 patients for a year with an SSRI, one additional patient will attempt suicide.
The reason why it is so important to include suicidal events that occur after the randomised phase is over is that it reflects what happens in real life rather than in a tightly controlled trial where the investigators motivate the patients to take every single dose of the trial drug. In real life, patients miss doses because they forget to take the pills to work, school or a weekend stay, or they introduce a drug holiday because the pills have prevented them from having sex.383
It differs from trial to trial what happens when it is over. Sometimes, the patients are offered active treatment, sometimes only the treated patients continue with active treatment, and sometimes there is no treatment.
In 2019, two researchers reanalysed the FDA data and included harms occurring during follow-up.384 Leading psychiatrists disliked the results and criticised the researchers, who then published additional analyses.385 Like other researchers, they found that suicide events had been manipulated, e.g. they removed two suicides that had erroneously been assigned to the placebo group in the paroxetine data.385 They reported twice as many suicides in the active groups than in the placebo groups, odds ratio 2.48 (1.13 to 5.44).
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.