Suicidal Behavior After FDA Warnings

On June 18, the British Medical Journal published an article by Christine Lu et al, titled Changes in antidepressant use by young people and suicidal behavior after FDA warnings and media coverage: quasi-experimental study

Here’s the conclusion paragraph from the abstract:

“Safety warnings about antidepressants and widespread media coverage decreased antidepressant use, and there were simultaneous increases in suicide attempts among young people. It is essential to monitor and reduce possible unintended consequences of FDA warnings and media reporting.”

Note the slightly rebuking tone directed against the FDA and the media.


The researchers interrogated the claims databases (2000-2010) of eleven medical insurance groups who collectively cover about 10 million people in 12 geographically scattered US locations.  All groups were members of the Mental Health Research Network (MHRN).

The general concept behind this sort of research network is that computerized insurance claims data represent an enormous repository of potentially very useful information, which researchers can readily tap for answers to questions that are difficult to resolve with smaller-scale data.  MHRN’s website, under the tab “Funding“, states:

“Initial funding for the MHRN is through a 3-year cooperative agreement with the National Institute of Mental Health (U19 MH092201 “Mental Health Research Network: A Population-Based Approach to Transform Research”) and through a supplement from NIMH to the existing Cancer Research Network funded by the National Cancer Institute.

During the initial funding cycle, the MHRN developed core infrastructure for collaborative research and conducted four developmental research projects to test and leverage that infrastructure in specific clinical areas.”

The Lu et al study is one of these four projects.

MHRN’s “virtual data warehouse” contains information on inpatient and outpatient treatment, and outpatient pharmacy data.  It also contains, for deceased members, the date and cause of death.

The present study included all adolescents (10-17), young adults (18-24), and adults (20-64) in the eleven insurance groups.


From the pharmacy data, the authors calculated the quarterly percentages of individuals who were dispensed an antidepressant.

The frequency of suicide attempts was measured by the quarterly incidence of “poisoning by psychotropic agents,” (ICD-9 code 969), that resulted in inpatient or ER treatment.

The frequency of completed suicides per 100,000 members was also calculated quarterly.  Although the other data elements are presented for each quarter up to the end of 2010, the completed suicide data stops at the end of 2008.  The authors explain:  “There is generally a lag time of 12-24 months for both reporting of deaths and availability of data; therefore we analyzed data on deaths up to and including 2008.”


The study’s time frame was from the first quarter of 2000 to the last quarter of 2010.  The researchers divided this time frame into three segments:

1.  The pre-warning phase (2000, q1 – 2003, q3)
2.  The phase-in period (2003, q4 – 2004, q4)
3.  The post-warning period (2005, q1 – 2010, q4)

The FDA black box warnings were mandated in October 2004.

The authors justify the use of a phase-in period as follows:

“To deal with the possibility of an anticipatory response to the warnings, we considered the last quarter of 2003 to the last quarter of 2004 as a ‘phase-in’ period that spanned the entire period of FDA advisories, the boxed warning, and intense media coverage, and excluded these five data points from the regression models.”

Regression, in this context, basically means finding a line-of-best-fit through a series of points which represent data on a graph.


The study included approximately the following numbers of individuals:

Adolescents (10-17)                1.1 million
Young adults (18-29)              1.4 million
Adults (30-64)                            5    million

The results for the three age groups are set out in graph form and are designated Figures 1, 2, and 3.

Let’s start with the adolescent graphs (Fig 1)

Fig. 1 Rates of antidepressant use, psychotropic drug poisonings, and completed suicides per quarter before and after the warnings among adolescents enrolled in 11 health plans in nationwide Mental Health Research Network

Fig 1 Lu et al



Adolescents (10-17)












The graph is divided into three parts.  The upper part shows antidepressant use, quarter by quarter, for the period of the study (2000, q1 – 2010, q4).  Each dot represents the percentage of the 1.1 million adolescents that were taking antidepressants in that particular quarter.  So in the first quarter of 2000, a little over 1.5%  were taking the drugs, and so on.

The middle part of the figure shows the percentage of adolescents who were treated for psychotropic drug poisoning in each quarter.

And the lower part shows the number of completed suicides (per 100,000) in the group, plotted quarterly, and graphed as a continuous line (the dark green line) rather than as dots.  The lighter lines represent the 95% confidence interval.  In other words, the chances are 95% that the true suicide rate for the population lies between these lines.

The slightly darker vertical band that goes up through the middle of all three graphs is the phase-in period (2003, q 4 – 2004, q4).  The solid lines in the upper and middle sections are the regression lines for the pre and post warnings data.  The post lines are curves because the authors report that this fitted the data better.  The orange dotted lines in the post area are simply the projection of the pre-warning regression lines; i.e. a projection of what purportedly would have happened if the warnings had not been issued.  Completed suicide data for the last two years of the study was not available.

. . . . . . . . . . . . . . . .

The data for the young adults (Fig 2) is presented similarly.

Fig 2 Rates of antidepressant use, psychotropic drug poisonings, and completed suicides per quarter before and after the warnings among young adults enrolled in 11 health plans in nationwide Mental Health Research Network

Fig 2 Lu et al


Young Adults (18-29)












The data for older adults is shown in Fig. 3


Fig 3 Rates of antidepressant use, psychotropic drug poisonings, and completed suicides per quarter before and after the warnings among adults enrolled in 11 health plans in nationwide Mental Health Research Network 

Fig 3 Lu et al


Older Adults (30-64)













The FDA warnings were not aimed at adults.  The authors state that the adults were included in the study  “…as a ‘control’ group.”

From all of this data, the authors draw the following results/conclusions:

1.  The FDA safety warnings, plus the media coverage, “decreased antidepressant use.”

2.  There were simultaneous increases in suicide attempts by young people.

3.  It is essential to monitor “unintended consequences” of black box warnings and media reporting.

4.  Completed suicides did not change for any group.

The main thrust of the article is expressed in the final paragraph:

“Undertreated mood disorders can have severe negative consequences. Thus, it is disturbing that after the health advisories, warnings, and media reports about the relation between antidepressant use and suicidality in young people, we found substantial reductions in antidepressant treatment and simultaneous, small but meaningful increases in suicide attempts. It is essential to monitor and reduce possible unintended effects of FDA warnings and media reporting.”

Again, we see the rebuking tone aimed at the FDA and the media.


As mentioned earlier, the study was funded by the National Institute of Mental Health (NIMH), which is a division of the US National Institutes of Health (NIH).  The study was published on June 18.  On June 19, NIH issued a two-page press release via its own website, Medline Plus.

Here are some quotes from the press release, interspersed with my comments and observations.

“Teen suicide attempts rose nearly 22 percent after the U.S. Food and Drug Administration (FDA) warned about dangers of antidepressants, a new study finds.”

This statement is false.  What the study found was that within the HMO Research Network’s insurance plans in ten states, psychotropic drug poisonings (including, incidentally, according to the press release, poisonings by marijuana, amphetamines, and Ecstasy), that required medical treatment, rose by the stated amount.  This is only one method of attempting suicide, and in many cases might not be attempted suicide at all.  The study provided no information on the incidence of confirmed suicide attempts measured as such.  Incidentally, ICD-9 code 969, which Lu et al used as a proxy measurement for suicide attempts, also includes “poisoning by caffeine”!

The statement is also misleading, in that the changes in the rates of psychotropic poisonings were measured not against the previous rate of poisoning as is implied in the above quote, but against the projected rate, based on the 2000 – 2003 regression line.  This is critical because regression lines can be extraordinarily poor predictors, as any stock market analyst can attest.

Regression lines are also easy to manipulate by the simple expedient of delineating the time boundaries to show a desired trend.  In the present study, for instance, the exclusion of the phase-in data from the pre-warning regression analysis definitely had the effect of lowering the poisoning trend projection in the young adult group (Fig 2), and probably had this effect in the adolescent group (Fig 1).  Excluding the phase-in data from the regression analysis was an arbitrary decision.  The researchers might just as readily have used a simple pre-post cutoff based on the quarter in which the warnings were mandated.

It is also noteworthy that the percentage changes in antidepressant prescriptions and psychotropic poisonings were assessed solely on the 4th quarter of 2006.  The authors’ statement on this matter is interesting:

“In addition, we also provided absolute and relative differences (with 95% confidence intervals)…in the second year after the warnings (that is, in the last quarter of 2006), which were estimated by comparing the overall changes in outcome attributable to the warnings with counterfactual estimates of what would have happened without the warnings.”

Note the unwarranted causal implication (“attributable”), and the assumption that the authors knew what would have happened had the warnings not been issued.  “Counterfactual” according to my Webster’s means “contrary to fact”.  I’m not sure what the authors had in mind in describing their projections this way.

There is also, I think, a suggestion in the above quote that the result applies to the general population.  This, however, would only be the case if in fact the group of individuals studied had been selected at random from the US population.  This is emphatically not the case.  The study group comprised all the individuals between ages 10 and 65 who were enrolled in the eleven MHRN insurance groups in the 12 states..  This group, for instance, is certainly not representative of uninsured people, nor of people in other locations.

. . . . . . . . . . . . . . . .

“Following the warnings, antidepressant prescriptions for young people fell by more than a fifth. At the same time, suicide attempts rose, possibly because depression was being undertreated, according to background information in the study.”

This statement is false.  What the study found was that antidepressant prescriptions for adolescents (10-17) declined from a peak of about 1.9% in 2003, q3 to a low of about 1.3% in 2008, q2, and then climbed to about 1.7% in 2010, q 3 (Fig 1).  But the figures for young adults (18-24), who are presumably included in the term “young people” as used above, remained remarkably constant at about 4% from 2004, q1 to 2010, q4.  The only “fall” for this group is a notional fall from the trend line. (Fig. 2)

Again in this quote, we see the implied causal link to suicide attempts, and even a suggested explanation:  “possibly because depression was being undertreated.”  The study provides no information on general nationwide trends in the frequency of suicide attempts, and even if suicide attempts were trending upwards during the period in question, there are many other possible explanations.  Also in this quote we see the invalid substitution of the term “suicide attempts” for psychotropic drug poisonings.

. . . . . . . . . . . . . . . .

“‘We found a substantial reduction in use of antidepressants in youth, and also in adults — who were not targeted by the warning,’ said lead author Christine Lu, an instructor in population medicine at the Harvard Pilgrim Health Care Institute in Boston.”

Here again, this statement is false.  The antidepressant use data for young (Fig 2) and older (Fig 3) adults remained almost perfectly flat in the post-warning period.  As noted earlier, the only falls were from the projected trend lines.

. . . . . . . . . . . . . . . .

“Lu attributes the drop in prescriptions to the FDA’s warning and resulting media coverage. ‘To a certain extent, the FDA’s black box warning was legitimate, but the media emphasis was really on suicide without noting the potential risk of undertreatment of depression. Because of that, there has been an overreaction, and that overreaction has sent alarming messages to parents and young people,’ she said.”

Note the skilful tightrope walking.  Dr. Lu and the NIH want to bash the media for publicizing the link between antidepressants and suicidal activity, but at the same time not overly antagonize the FDA.  The media ignored the “potential risk of undertreatment”; the media overreacted; the media caused alarm.  Big bad journalists need to get into line!

. . . . . . . . . . . . . . . .

Then we have a truly magnificent piece of psychiatric spin:

“Although the initial studies showing an increased risk of suicide in teens taking antidepressants prompted the black box warnings, researchers never proved that the medications were the cause of the increased risk of suicide, only that there was a link.

Likewise, though the current research finds a strong association between the uptick in suicides and the drop in antidepressant use, Lu and her colleagues weren’t able to definitively show that a decrease in antidepressant prescriptions was directly responsible for the recent increase in suicide attempts.”

The deliberations and the data on which the FDA black box warnings were based were comprehensive, and showed a clear link between antidepressant use in young people and suicidal activity.  And the evidence was sufficient to convince the FDA to take action.

The findings in the Lu et al study do not come even close to this standard.  But by counterposing them in this way, the NIH is trying to convey the impression that the findings of the present study are comparable in quality to the earlier work.

Note also the blatant falsehood:  “…the current research finds a strong association between the uptick in suicides and the drop in antidepressant use…”  [Emphasis added] In fact, the Lu et al study found no increase in completed suicides.  They stated clearly in their abstract:  “Completed suicides did not change for any group.”

. . . . . . . . . . . . . . . .

“Coverage of the warning may have had unintended consequences, Lu said. Doctors may have been less willing to prescribe antidepressants and parents may have been fearful of letting their children take them, she said.

The lesson, Lu said, is that the media and the FDA should strive for the right balance so potential overreactions don’t occur.”

Again, big bad media needs to get in line and print only what psychiatrists tell them to print (“the right balance”).

. . . . . . . . . . . . . . . .

“Undertreating depression is worse than the slight increase in suicidal thoughts antidepressants may cause, Lu said. ‘It’s also a reminder for doctors to weigh the risk of a drug with the risk of not treating or undertreating the condition,’ she said.”

This is extremely misleading.  Note how the concerns on which the boxed warnings were based are dismissed:  “…the slight increase in suicidal thoughts antidepressants may cause.”

The pediatric meta-analysis data on which the FDA deliberated in 2003-2004 are written up in detail (131 pages) here.  Tarak Hammad et al published the formal journal report in 2006.  It is clear that great pains were taken to include in the primary outcome measure only adverse events that had clear suicidal potential.  The primary outcome, which was labeled “suicidal behavior or ideation”, contained three elements:

  • Suicide attempts, and/or
  • Preparatory actions towards imminent suicidal behavior, and/or
  • Suicidal ideation

The overall risk ratio (drug vs. placebo) for the three elements combined was found to be 1.95 (95% CI, 1.28 – 2.98).  The risk ratio for suicidal behavior (first two items) was 1.90 (95% VCI, 1.00 – 3.63).  In other words, the individuals in the studies who took the drug were about twice as likely to have made a suicide attempt, and/or made preparations for imminent suicide, and/or had been actively thinking about suicide than those who took the placebo.  This is emphatically not something to be dismissed as “a slight increase in suicidal thoughts…”

Hammad et al acknowledged the limitations of their study, but went on to state:

“Despite the limitations, the observed signal of risk for suicidality represents a consistent finding across trials, with many showing RRs [risk ratios] of 2 or more. Moreover, the finding of no completed suicides among the approximately 4600 patients in the 24 trials evaluated does not provide much reassurance regarding a small increase in the risk of suicide because this sample is not large enough to detect such an effect.”

They also provide the following interpretation of their findings:

“…when considering 100 treated patients, we might expect 1 to 3 patients to have an increase in suicidality beyond the risk that occurs with depression itself owing to short-term treatment with an antidepressant.” [Emphasis added]

In addition, since the issuing of the black box warnings in 2004, there have been two studies confirming the link between suicidal activity and antidepressants.

Olfson et al 2006 found:

“…in children and adolescents (aged 6-18 years), antidepressant drug treatment was significantly associated with suicide attempts (OR [odds ration], 1.52; 95% CI, 1.12-2.07 [263 cases and 1241 controls]) and suicide deaths (OR, 15.62; 95% CI, 1.65-infinity [8 cases and 39 controls]).”

And Olfson et al 2008 found:

“Among children, antidepressant treatment was associated with a significant increase in suicide attempts (odds ratio [OR] = 2.08, 95% confidence interval [CI] = 1.06 to 4.10; cases, N = 51; controls, N = 239; p = .03).”


The Lu et al article, and the authors’ conclusions, received wide coverage in the general media. Most of the mainstream media simply regurgitated the gist of the study plus the press release with additional quotes from two of the authors, Christine Lu, PhD, and Steven Soumerai, ScD, both from Harvard’s Department of Population Medicine.

Both the Washington Post and The Boston Globe delivered the authors’ message pretty much without reservation.  Here’s a quote form the Washington Post:

“…Wednesday’s study wrote that while the government’s actions in 2003 and 2004 were legitimate and thorough, ‘FDA advisories and boxed warnings can be crude and inadequate ways to communicate new and sometimes frightening scientific information to the public.’ Likewise, researchers argue that while media attention can create much-needed awareness…sometimes ‘the information may be oversimplified and distorted when communicated in the media.'”

And from The Boston Globe:

“The study’s authors say that patients and doctors, frightened by news coverage that exaggerated the risk of antidepressants, shunned treatment that might have prevented the suicide attempts.”

In fact, there is no information in the report as to whether the victims of the poisonings had taken antidepressants or not.

And amazingly,

“Steven B. Soumerai, a coauthor and a Harvard professor of population medicine, said black box warnings typically have little effect on physician behavior, unless they are accompanied by news reports.”

In his attempt to castigate the press for publicizing the black box warnings, has Dr. Soumerai inadvertently shot psychiatry in the foot?  Do psychiatrists actually pay more attention to news reports than to FDA warnings?

Bloomberg also ran the standard story:

“A widely publicized warning by U.S. regulators a decade ago about risks for teenagers taking antidepressants led to plummeting prescriptions and increased suicide attempts, Harvard University researchers said.”

Note the clearly implied causality (“led to”), the exaggeration (“plummeting”) and the mischaracterization of psychotropic poisonings as “suicide attempts.”

“‘After the widely publicized warnings we saw a substantial reduction in antidepressant use in all age groups,’ said Lu, an instructor in population medicine at Harvard Pilgrim Health Care Institute, in a telephone interview. ‘Warnings, especially widely publicized warnings, may have unintended consequences.'”

“Lu said the focus by the media on the risk of suicide, even though the review of data found no increase in completed suicides, frightened patients and parents.”

Again, the big bad media frightening patients and parents.

But Bloomberg also obtained some rebuttal quotes from Marc Stone, MD, a senior medical reviewer at the FDA:

“‘It’s a stretch to say that the people that are committing suicide or the increase in suicide attempts has to do with the prescription of antidepressants,’ Stone said in a telephone interview. ‘There’s absolutely nothing in the study to say that these are the people who would have been prescribed the antidepressants if it weren’t for the warnings.'”

“There are other issues that could be influencing the drop in antidepressant prescription and rise in poisonings, Stone said. The data presented in the study shows a steady increase before the warnings in antidepressant prescription rates for adolescents while drug poisonings remained relatively steady. The rise in poisonings after the warnings when prescriptions declined slightly in that group doesn’t show a link between the two events, he said.

‘They’re describing a very strange phenomenon in society that’s supposedly being held back by antidepressant use,’ Stone said. ‘It doesn’t stand up to what we know about the mental health situation in the United States.'”

I could find nothing in the New York Times, the Wall Street Journal, or the LA Times about the study or the press release.

Medscape, a web resource for physicians and other health professionals, toed the party line:

“The safety warnings were covered widely in the media and led to a decrease in antidepressant use by young people, but at the same time, there was an increase in suicide attempts, Christine Y. Lu, PhD, of the Department of Population Medicine, Harvard Medical School in Boston, Massachusetts, and colleagues found.”

Again, note the inaccurate characterization of drug poisonings as “suicide attempts”.

“The researchers say that it is possible that the warnings and extensive media attention led to ‘unexpected and unintended population level reductions in treatment for depression and subsequent increases in suicide attempts among young people.'”

And the focus on the “extensive media attention”.

The APA drew attention to the Lu et al study in a Psychiatric News Alert, but the tone of the piece was appropriately and, I must say, surprisingly, skeptical.  The only quote was from Mark Olfson:

“Mark Olfson, M.D., M.P.H., a professor of psychiatry at Columbia University Medical Center and an expert in mood disorders, told Psychiatric News that ‘the new findings shed little light on the complex associations between anxiety and depressive disorders, antidepressant treatment, and the risk of self-harm and suicide. The measure of suicide attempts used in this study, psychotropic poisonings (ICD-9 code 969), is only loosely related to suicide attempts. Most suicide attempts in young people do not involve poisoning by psychotropic drugs and most intoxications do not represent suicide attempts.’

Because of the recent substantial increase in unintentional poisonings from stimulants, Olfson stated that the increase in psychotropic overdose could be a result of an underlying substance use disorder rather than suicide. ‘This trend [of psychotropic overuse by youth], which may be driven by complex societal factors, deserves study and clinical attention,’ Olfson concluded.”

. . . . . . . . . . . . . . . .

Finally, the most reasonable interpretation of the Lu et al data is that following the black box warnings, the long-standing increase in the use of antidepressant drugs was halted, but suicide rates remained about the same (N = 7.5 million).


This is not the first attempt to discredit/undermine the black box warnings for these drugs.  In September 2007, Robert Gibbons, PhD et al published Early evidence on the effects of regulators’ suicidality warnings on SSRI prescriptions and suicide in children and adolescents in the American Journal of Psychiatry.  The paper was based on a study of US and Dutch SSRI prescription rates to children and adolescents from 2003 to 2005, and was funded by NIMH grant.  The authors reported the following results:

“SSRI prescriptions for youths decreased by approximately 22% in both the United States and the Netherlands after the warnings were issued. In the Netherlands, the youth suicide rate increased by 49% between 2003 and 2005 and shows a significant inverse association with SSRI prescriptions. In the United States, youth suicide rates increased by 14% between 2003 and 2004, which is the largest year-to-year change in suicide rates in this population since the Centers for Disease Control and Prevention began systematically collecting suicide data in 1979.”

And drew the following conclusions:

“In both the United States and the Netherlands, SSRI prescriptions for children and adolescents decreased after U.S. and European regulatory agencies issued warnings about a possible suicide risk with antidepressant use in pediatric patients, and these decreases were associated with increases in suicide rates in children and adolescents.”

The report contains a graph showing a sharp increase in suicide rates for children and adolescents (5 – 19) between 2003 and 2004.

Gibbons Fig 2

The data for this graph was extracted from the CDC WONDER Compressed Mortality Database (up to 2002) and from the CDC WISQARS Injury Mortality Report Database (2003 – 2004).  On the face of it, this graph does seem to show a dramatic increase in completed suicides for this age group between 2003 and 2004.  But when we look at the bigger picture, this trend is not so obvious.

On July 1, I extracted completed suicide rate and population data from the CDC website for the period 1979 to 2010.  The Gibbons et al graph is for ages 5 – 19.  I was unable to obtain that graph from the CDC site, but I was able to collate the raw data and draw the graph using Excel software.  Here’s the graph:

Suicides per 100,000, 1979-2010

As can be readily seen, the increase in the suicide rates for this age group from 2003 to 2004 does not appear as marked or as noteworthy when viewed against the wider background.  Drawing conclusions from short-term trends is fraught with potential for error.  Dr. Gibbons et al were obviously aware of these issues, and provided a lengthy discussion/argument in support of the relevance of the associated trends.  Nevertheless, here’s their closing paragraph:

“In December 2006, the FDA’s Psychopharmacologic Drugs Advisory Committee recommended that the black box warning be extended to cover young adults, and in May 2007, the FDA asked drug manufacturers to revise their labels accordingly. If the intent of the pediatric black box warning was to save lives, the warning failed, and in fact it may have had the opposite effect; more children and adolescents have committed suicide since it was introduced. If as a result of extending the black box warning to adults there is a 20% decrease in SSRI prescriptions in the general population, we predict that it will result in 3,040 more suicides (a 10% increase) in 1 year (17). If the FDA’s goal is to ensure that children and adults treated with antidepressants receive adequate follow-up care to better detect and treat emergent suicidal thoughts, the current black box warning is not a useful approach; what should be considered instead is better education and training of physicians.”

The Washington Post ran an article on the Gibbons et al study, Youth Suicides Increased As Antidepressant Use Fell, September 6, 2007.  The article included the following quotes:

“Thomas Insel, director of the National Institute of Mental Health, said, ‘We may have inadvertently created a problem by putting a ‘black box’ warning on medications that were useful.’ He added, ‘If the drugs were doing more harm than good, then the reduction in prescription rates should mean the risk of suicide should go way down, and it hasn’t gone down at all — it has gone up.'”

“The trend lines do not prove that suicides rose because of the drop in prescriptions, but Gibbons, Insel and other experts said the international evidence leaves few other plausible explanations.” 

Given the behavior of the trend line since 2004, these quotes suggest, at the very least, an over-reaction from the authors and from the NIMH.  And the fact that the NIMH funded both studies suggests that they may have an agenda:  to promote antidepressant drugs as safe, and to downplay any indications to the contrary.

If the NIMH genuinely wants to explore the sources/causes of suicide, they need to recognize three facts:

1.  Suicide is a complex, psychosocial phenomenon, and its roots/causes vary enormously from individual to individual, and from community to community.  Psychiatry routinely asserts that the root, or major cause, of suicide is depression, but this assertion is meaningless because suicidal activity/rumination is one of psychiatry’s defining features of depression.  A genuine understanding of human actions involves so much more than assigning labels.

2.  The facts surrounding any completed suicide are always, by the nature of the matter, at least partially hidden.

3.  Any attempt to understand suicide must involve a detailed, open-minded exploration of individual cases.  Variables that appear associated with suicide trend lines provide, at best, suggestions that might inform these explorations, but can also be very misleading.


At the present time, there is a great deal of anecdotal information, on the Internet and elsewhere, to the effect that individuals who were not previously suicidal, became so, shortly after starting SSRI’s.  The psychiatrist Joseph Glenmullen drew attention to this graphically and convincingly in Prozac Backlash (2000).  Here’s a quote:

“…the key elements in these stories appeared to be the ‘dramatic change’ observed in these people after starting Prozac, how ‘out of character’ their behavior was on the drug, and the often extraordinary degree of violence not only toward themselves but toward others.”

Elsewhere in the book, Dr. Glenmullen makes it clear that his comments apply, not only to Prozac, but to all drugs in the SSRI category.  Psychiatrist Peter Breggin has made similar points.  AntiDepAware has accumulated an enormous amount of anecdotal information on this matter.

Psychiatry tends to dismiss these kinds of concerns as “anecdotal”.  In this context, “anecdotal” usually means:  based on casual or incidental information rather than on systematic observations and evaluations.  In particular, the term suggests a subjective, rather than an objective approach.

But in fact, the raw data for virtually all psychiatric research is anecdotal in this sense.  If a person reports suicidal ruminations after taking an SSRI, this is anecdotal.  But if a study participant reports that he’s feeling better after taking an SSRI, this is also anecdotal.  And if 20 people make essentially similar statements, that’s 20 anecdotal statements.  And if they make these statements by answering 17 questions on the Hamilton Rating Scale for Depression, then we have 340 anecdotal statements.  The fact that the HDRS-17 is widely used, and yields numerical data which can be collated into rows and columns and analyzed with sophisticated computer programs doesn’t alter the fact that the raw data is anecdotal.

No one on this side of the issue is suggesting that a large proportion of people taking antidepressant drugs kill themselves.  If that were the case, then the drugs would have been removed from the market long ago.  What is being contended, however, is that these drugs are inducing strong suicidal urges in a relatively small proportion of individuals who had not previously had thoughts of this kind, and that some of these people are succumbing to these urges, and are taking their own lives.  That is indeed anecdotal information.  But it’s also very important, and it warrants investigation even if it involves only a fraction of one percent of the people taking the drugs.  To dismiss these widespread and credible contentions on the basis of the dogmatic insistence that the drugs are wholesome, or that the individuals were probably suicidal to begin with, is simply unconscionable.  This is particularly the case in that firstly, the proof of the wholesomeness of these products is also, ultimately, based on anecdotal information, and secondly, there is no way to reliably identify the individuals concerned prior to prescribing the drugs.

Prescribing these psychoactive drugs is like playing Russian roulette with someone else’s life.  The notion that this information should be suppressed because it might scare people from taking the products is cruel, callous, and irresponsible.  Psychiatry’s persistent and self-serving efforts to suppress this information is a national, and indeed, worldwide, scandal.

The Lu et al study, and the NIH press release, reminds me of a statement made by Patrick B. Kwanashie, Assistant Attorney General in Connecticut, after the Adam Lanza murders/suicide:

“Even if you can conclusively establish that Adam Lanza’s murderous actions were caused by antidepressants, you can’t logically from that conclude that others would commit the same actions as a result of taking antidepressants.  So it’s simply not legitimate, and not only is it not the use to which they are proposing to put the information not legitimate, it is harmful, because you can cause a lot of people to stop taking their medications, stop cooperating with their treating physicians just because of the heinousness of what Adam Lanza did.”


The Lu et al study has been subjected to a great deal of criticism.  These criticisms have been cogent and valid, but there is also a bigger picture.  The Lu et al study was never meant to be about science.  It was about spin and PR.  Its purpose was to attack and embarrass the media in order to keep them in line.  The fundamental message in the study, and in the NIH press release, is:  If you reporters print bad things about antidepressants, this will lead to reduced usage of these products, which in turn will lead to more suicides, and you will have blood on your hands!  Psychiatry is intellectually and morally bankrupt.  It has no valid response to its critics and is increasingly resorting to this kind of spin and PR.

Organized psychiatry realizes that it has lost the present debate on its logical/scientific merits.  They realize that their only hope for survival as a profession hinges on their ability to control the media.  In the pursuit of this objective, there truly are no depths to which they will not sink.

Over the past two decades, I have become convinced of two things:

1.  SSRI’s, and possibly other psychiatric drugs, are inducing strong suicidal and violent urges in some people.
2.  This information is being systematically suppressed by psychiatry and by the NIMH.  To the extent that they acknowledge it at all, they pretend that this suppression is in the public interest.

If there was ever a time when we need courageous and well-informed reporting, it is now.  Let us hope that the press has sufficient integrity and courage to resist psychiatry’s tawdry blackmail, to follow the evidence, and to print the truth.   In this context, it is very encouraging that neither the New York Times, nor the LA Times, nor the Wall Street Journal took the Lu et al bait.

It is time to end the charade.  The link between psychiatric drugs and suicide/violence needs urgent study by independent, adequately funded investigators who are given access to all the information.

  • Francesca Allan

    Thank you for the thoughtful analysis, Phil. I can only hope that prescribers and potential patients take the time to gather real information beyond just the brief and misleading synopsis of Lu’s research that the media are providing. It troubles me that correlation only suggests cause when it suits the agenda of the researcher.

  • Congrats. Exposing flaws in statistical analysis can be tedious and frustrating. How does stuff like this meet peer review standards? (rhetorical question)

    Ironically the article and the promotional activities of the authors to get media coverage of it have the potential of causing suicide. They chide the media and then irresponsibly attempt to create some hysteria by use of the media.

    Clustering following media reports regarding suicide is known to occur.

    Here are couple recent Lancet articles on suicide.

  • Anonymous

    ” SSRI’s, and possibly other psychiatric drugs, are inducing strong suicidal and violent urges in some people.”

    How do you know it is the SSRIs “inducing” these thoughts? You need to look at what else is happening concurrently to them ingesting these drugs. In my opinion the most significant thing occurring concurrently is the initial period of time taking these drugs may very well be the first time in their lives they’ve been led to believe they have no free will and are just a meat machine brain / a sick brain, at the mercy of their brain, not in control, the passive victim of an active “disease”. Concurrent to the first few days/weeks of taking the drugs is the initiation into a powerlessness ideology, that is to say, people tend to focus on the pharmaceutical chemicals introduced into the person’s body, and not on the ideology introduced into their minds during the same period. I don’t dispute that drugs can inhibit/dis-inhibit people, and I do believe a deadly mix of dis-inhibition combined with psychiatric brainwashing that teaches the person they have no free will, no personal power to master their life problems, is at play in many of these cases. I just don’t like to see these parents that blame drugs for their kid’s suicide, as if that’s the whole story. Who took little Johnny or Janie to a biological determinist quack to be brainwashed into believing he was a passive brain disease sufferer whose only salvation lay in the technical wizardly of the neuro-poisoners? Parents need to own that they had their kid’s mind filled with dis-empowering spurious lies. But it’s easier to jump up and down and sue Astra Zeneca. And why is it always suicide that the drugs are alleged to “cause”? Why not regicide or insecticide? Why don’t they “cause” people write symphonies or get an MBA? Or hire a clown for kid’s parties? Suicide is frowned upon and there are endless evasions of responsibility, ascribing a human action to the magical powers of drugs is magical thinking drugs are only ever part of the story.

    I also think that the psychiatric term “completed suicide” is ridiculous. You can have a partially completed building, or film, but you cannot have a partially completed killing. One is either killed or not killed, anything less than being killed, is an injury.

  • Francesca Allan

    A philosopher would say we can never really PROVE a causal relationship, just demonstrate a correlation. In the case of violence following SSRIs, that correlation is strong enough to be statistically meaningful.

    I prefer the term “completed” suicide over “committed” suicide as the latter implies a crime. And I believe an attempted suicide is more serious than a self injury and ought to be labelled as such.

  • Phil_Hickey


    Tragically, it’s the headlines that people retain. Psychiatry knows this, and is exploiting this like shabby politicians.

    Best wishes.

  • Anonymous

    Stop breaking the truce please.

  • Phil_Hickey


    Yes. I was struck by the fact that the BMJ published it. The flaws were pretty obvious.

    Thanks for the links.

  • Phil_Hickey


    Thanks for coming in. These are all good points. The initiation into the drugged community is indeed a disempowering experience, and I’m sure that you’re correct in stressing the potential consequences.

    But I think there is also a very specific toxic effect with SSRI’s in some people. Here are two quotes from Joseph Glenmullen’s book Prozac Backlash (2000):

    “Joanne said the agitation and suicidal thoughts had begun at the same time, at the end of her first week on Prozac. She described the same kind of depersonalization that I had heard attributed to the drug before: feeling she was outside herself, observing her own behavior in a detached, ‘surreal’ way. Joanne had not been suicidal before starting Prozac. She was convinced it had triggered the obsession.” (p 143)

    “In her second week on Prozac, Sunita’s clinical picture changed dramatically. Trembling because she felt so ‘panicky,’ Sunita told me she felt like she was being ‘brutalized’ on campus.

    ‘What do you mean?’ I asked, alarmed.

    ‘My body parts are strewn all over the campus.’

    ‘Your body parts?’

    ‘My arms, my legs, my torso…when I walk around campus I see them strewn through the trees.’


    ‘Yes. They’re hanging in trees, all bloodied, I’ve been dismembered.’

    Sunita went on to tell me the professor in one of her most difficult premed courses was now forcing her to ‘fellate’ him.

    ‘He is?’

    ‘Yes. Most days when I go to class.’

    ‘When does this happen?’

    In class. He’s been making me perform oral sex on him in front of the whole class!’

    ‘In front of the class?’

    ‘The whole class, one hundred fifty students,’ Sunita insisted, desperately. ‘He keeps staring at me, singling me out, talking to me instead of the others, looking at me sexually until I come down and perform oral sex on him in front of everyone, and then he stops bothering me. It’s so humiliating.’

    Obviously this could not possibly be happening in front of a huge class of students. Sunita was psychotic.

    Still worse, she was also preoccupied with killing herself. Disjointedly, she said, ‘I’m dead anyway. People are killing me. My body parts are strewn all over campus. I can’t stop thinking I should just kill myself and get it over with.’

    Sunita had never been psychotic or suicidal before.” (p 145-146)

    I’ve heard, and read of, a great many similar stories, and it’s hard not to believe that there’s some kind of specific neurotoxic effect, probably from pathologically high levels of serotonin. Often the individuals don’t even appear to be particularly depressed.

    Suicides are described as “complete” simply to distinguish them from suicide attempts. The correct word to indicate this distinction would be “successful,” but it is avoided, I think, largely out of deference to the feelings of family members, friends, etc…

    Best wishes.

  • Cledwyn B’stard

    “The suicide rate, so I am told by an intelligent mortician, is going up. It is good news to his profession, which has been badly used of late by the progress of medical science, and scarcely less so by the rise of cut-throat, go-getting competition within its own ranks. It is also good news to those romantic optimists who like to believe that the human race is capable of rational acts. What could be more logical than suicide? What could be more preposterous than keeping alive?”

    H.L. Mencken

    Maybe these drugs do make some people more rational…..

  • Francesca Allan

    Anonymous, you are posting on a public forum. Anybody can reply to any post. Deal with it.

  • cledwyn barstard

    In venenum veritas.

  • Always the cynic. Quit being such an INFJ anomalously sensitive, don’t you know your putting yourself at huge risk for misdiagnosis?

  • Cledwyn B’stard

    What’s INFJ?

  • Anonymous

    It refers to this

    Mankind’s addiction to putting labels on everything and everyone. But I respect T.A’s right to believe in it.

  • Nobody get’s our jokes.

  • Francesca Allan

    “Personality predicts behavior and behavior predicts personality.”

    This is true as far as it goes but it doesn’t shed much light on the matter. All you’re saying is that X type people behave in X type ways and that when you observe X type ways it’s likely you’re looking at an X type person.

  • James Woods
    Los Angeles, California

    Via Telepathic Facsimile and/or Kevin Bacon

    Re: Mohammad Atta

    Dear Brother Jim,

    Ah Brother Jim the INFJ/INFP . . . I can sense you’re grokking me already. LMAO. But are you fully grokking me or has cognitive dissonance got you consciously blinded? No, I think your eyes have been wide open for a long time.

    Your 9/11 story sucks and I’m the perfect person to dissect it. At the critical moment on the video you slipped too much. LOL. And I like this account on Wikipedia: “On the flight he noticed four men near him acting suspiciously. He said that they never drank anything, ordered food service or talked to anybody. They were just whispering to each other.”

    You are one of the ten smartest people alive today What would it mean to the world if you copped to the paranormal? I’m sure the FBI was so kind as to remind you of Shirley MacLean. I call that a threat. I’m sure they made an appeal to your patriotism and spent some time convincing you that it would be a threat to our ordered society and national security if the existence of the paranormal became common knowledge. I call that bullshit. This is the same agency which considered John Lennon a threat. What’s more shameful the surveillance of John and Yokko or the ignorance demonstrated in conducting it? I say the latter.

    If Stanley Krippner wants the world to understand the paranormal, then
    rest assured that would be a good thing. We both understand that psychic premonitions are not generally reliable enough to act upon. The world will understand that too.

    You have in your grasp the opportunity to transform humanity by giving
    the world something it desperately needs—knowledge of true spirituality.

    Your kindred spirit,

    Gerard Scott

  • September 12, 2001.

    Hickey could you please delete the other photos.

  • Phil_Hickey


    I’m not a big fan of the concept of “personality” or “personality type” for the following reasons.

    1. Like mental health “diagnoses” it has, in my view, limited explanatory value.

    Q. Why is he always so assertive and pushy?

    A. Because he has an assertive personality.

    Q. How do you know that he has an assertive personality?

    A. Because he is so assertive and pushy.

    2. It tends to downplay the role of context in eliciting behavior.

    3. I think “habit” is a better concept. A person might, for instance, have acquired the habit of speaking quietly; another might have developed the habit of speaking forcefully and assertively. So the above conversation would become:

    Q. Why is he always so assertive and pushy?

    A. I’ve no idea. I don’t know him very well.


    A. I’ve known him all his life. His parents strongly encouraged assertive behavior, and apparently it became habitual. But I’ve also seen him be extremely quiet and unassuming in certain situations.

    The best book I’ve come across on personality is Walter Mischel’s Personality and Assessment (1968). Here’s his final paragraph:

    “Global traits and states are excessively crude, gross units to encompass adequately the extraordinary complexity and subtlety of the discriminations that people constantly make. Traditional trait-state conceptions of man have depicted him as victimized by his infantile history, as possessed by unchanging rigid trait attributes, and as driven inexorably by unconscious irrational forces. This conceptualization of man, besides being philosophically unappetizing, is contradicted by massive experimental data. The traditional trait-state conceptualizations of personality, while often paying lip service to man’s complexity and to the uniqueness of each person, in fact lead to a grossly over-simplified view that misses both the richness and the uniqueness of individual lives. A more adequate conceptualization must take full account of man’s extraordinary adaptiveness and capacities for discrimination, awareness, and self-regulation; it must also recognize that men can and do reconceptualize themselves and change, and that an understanding of how humans can constructively modify their behavior in systematic ways is the core to a truly dynamic personality psychology.”

  • Jorge_Videla

    It tends to downplay the role of context

    hear! hear! BUT…

    it isn’t subtle, but it seems beyond the grasp of almost everyone that the phenotype (and the psychical phenotype, a fortiori) is not the arithmetic sum of genome and environment and is not fixed at age 18. P = G + E is a very very crude approximation.

    unfortunately hereditists and “enduring trait”-ists have this model as an “unknown known”, so that to any criticism they respond, “humbug! you’re just a blank slater!”

    in the end one cannot always say, “had i his experience…” BUT nor can one say, “he was born that way.”

    each genome responds differently to each environment.

    heritability, h^2, at a particular place and time, a particular society, is almost meaningless. and, obviously, is based on the P = G + E model.

    even theoretical physicist and BGI cognitive genomics leader steve hsu seems incapable of understanding this.

    lead to a grossly over-simplified view

    and due to psychology and psychiatry’s cultural presence and conservative hereditist ideology my experience is that laymen think of people this way too. and it does enormous harm.

  • Jorge_Videla

    as the “elvis of cultural criticism” has said so many times, human affairs are shot through with ideology even in this “post-ideological” age. “they don’t know it, but they do it.”

    it’s even gotten worse in a post-scarcity economy where 80% of the formally employed don’t really do anything. if you aren’t really doing anything you have to sell what you’re doing constantly and you have to sell it to the salesmen too. so a jive talk develops.

    there’s nothing ideological about farming or building a house. is there?

  • Jorge_Videla

    i know one schizophrenic. his father is an oral surgeon and his parents still married, so he had a good home as far as i know. he was treated for depression before he developing delusions and having hallucinations.

    my uncle on my father’s side died from an unknown organic brain disease…in the us. his doctors had no diagnosis. he lost his ability to speak, then his ability to walk, then he died. he had been treated for depression for a few years beforehand. so it could be his depression was really this organic problem or

    might the drugs have killed him?

  • cledwyn bastardo

    Personally, I think the idea that these drugs induce suicidal impulses further befuddles the issue. Granted, there is more truth in it than is to be found in psychiatric explanations of violence, be it against the self or others, but nevertheless I don’t think these drugs literally induce suicidal impulses. I would surmise that anterior to such events there are often to be found a concatenation of circumstances and events so numerous, anything other than a partial understanding is impossible.

    There are so many imponderables, imponderables that people in their hatred of uncertainty refuse to acknowledge.

    Michael Haneke tried to show this in his “70 fragments of a chronology of chance”, a fragmented narrative of an individual who commits homicide “in a moment of madness”, mostly made up of events of no import looked at in isolation, but in the aggregate form a web of circumstances and events that induce a susceptibility to violence.

    Does this mean that I am denying the disinihibition that often occurs on certain drugs? No. Am I denying that certain drugs induce such inner agitation (as I now live with everyday because of the tardive akathisia, so I’m not being ignorant), that they turn their violent impulses against themselves or, under provocation, against others? No. Just that they don’t literally induce these impulses. Their effects do.

    The effects sometimes are so unpleasant, they do indeed “drive you nuts”, especially something like akathisia, where nothing you do gives you the desired catharsis, nothing gives you relief from or transcendence of your situation (although some drugs can help short term, they are all of limited efficacy, and can make things even worse, if that’s possible), which creates an inner tension too unbearable for words.

    I think much of the violence, as well as the anger, the bitterness, and frustration of man in modern society is rooted in conflict-resolution mechanisms whose cathartic efficacy is poor, mechanisms nowhere near as efficacious as those means purging of ourselves that have either fallen into desuetude (such as dueling. Bring it back I say!), or those found amongst criminal subcultures, such as the Mafia, where, conflict was resolved by murdering your opponent, and thereby purging yourself of the diabolical concoction of emotions, impulses, and thoughts stirred up in our conflicts with others, which undergo intensification with each exchange of hostilities, as any student or observer of human conflict will know.

    Man in modern society must contain his anger, his frustration, his hatred, his atavistic appetites and instincts, and he does so, most of the time, although sometimes he gets pushed beyond the limits of human endurance, or randomly discharges the anger and hatred smouldering beneath the surface. Either which way, the existence of punitive mechanisms, and concepts such as human rights and the social contract, inculcate upon him the awareness that that he must learn to restrain himself, and learn not to give free rein to his vengeful, predatory, cannibalistic impulses, his lust for power, otherwise face the consequences; but now his conflicts rankle him all the more, because he can no longer kill his opponent, or fight him, or humiliate him, or anything that would achieve the catharsis that brings him relief from his torment.

    Perhaps this explains the love of the mob movie. People are getting nostalgic for a world in which you can resolve your conflicts in a much more efficacious way, that is efficacious from the point of view of achieving catharsis (Although the mob movie is actually itself a kind of surrogate cathartic mechanism, because through imaginatively identifying with the characters, you can vicariously experience some form of mild catharsis, but these surrogate mechanisms that allow us to imaginatively simulate the discharge of such pent up aggression and rage don’t ultimately achieve the real thing).

    Indeed, mob movies are excellent arenas for viewing the drama of human conflict, its underlying dynamics, as well as learning about how are avoided what Girard describes as mimetic waves of violence, and peace within communities is maintained (in the mob world, the conflict that threatens to upset the peace and unleash a wave of mimetic violence, is resolved by the killing the necessary person), mimetic waves that inevitably happen when relations of conflictual mimesis (which basically refers to the immense suggestibility of men to the emotions, desires and behaviours of their opponents in relations of conflictual reciprocity) spread throughout a community and engulf it with violence.

    Coming back to the inefficacy of the mechanisms that force man to repress his instincts, they nevertheless often beget the problem they are supposed to remedy, because the tension that builds can be so unbearable, it leaves some men (especially those overexposed to humiliating treatment, who are on the wrong side of conflictual imbalanced power relationships that preclude the possibility of catharsis for such people because of this disparity in power) bursting at the seams with rage and frustration, till it starts eating away at them; so it erupts, sometimes randomly, sometimes in conflictual situations

    This demystifies so much of the so-called “savagery” and anger of mental patients, which is often in complete contrast to psychiatrists, and for a reason that has nothing to do, I would surmise, with any sort of dispositional difference, and everything to do with the much greater reason people on the wrong side of imbalanced power relationships have for being consumed with anger and frustration, and vengeful, hateful impulses; because their situation denies them proper catharsis, catharsis for the frustration and humiliation that builds up as a result of their enslavement, the violence against them, the abuses of power they are on the wrong end of, the hypocrisy and lies they are the victims etc., whereas psychiatrists often have the power to resolve conflicts in their favor because of their power and position (yet the rage of the powerless, the impoverished, the humiliated and the abused is still viewed in dispositional terms, perhaps because it allows their judges to revel all the better in their hatred)

    Continuing with this for a sec, it’s hardly surprising that so many people commit suicide in psychiatric wards when the foregoing is taken into account; there is only so much inner tension you can take before you turn the violent impulses and emotional turmoil, occasioned by the frustration of vengeful desire, against yourself. That’s what I almost did in hospital.

    Concluding this discussion about the the violence in our society, where psychiatric drugs come into all this is that they take away inhibitions, so that people, in their conflicts with others, are less likely to give any thought to the consequences. Nevertheless, it is conflict that is central here, I would surmise, be it with violence against others or against the self. People just aren’t very good at dealing with it, as Peter Watkins show in his psychodrama “Punishment Park”.

  • all too easy

    “These findings fail to support either an overall difference in suicide risk between antidepressant- and placebo-treated depressed subjects in controlled trials or a difference between SSRIs and either other types of antidepressants or placebo.” FDA

    Appears we have one more undisputed genius on our hands. Cleodapuss, watch out honey.

    Offering a medically safe and approved solution to many severely depressed patients undoubtedly makes them feel powerless which obviously increases their longings to leave this life. Poor kids. Just like when a combat soldier has her legs blown off in war, (according to this magnificent boob), she isn’t supposed to seek medical assistance. After all, she might feel powerless!

    But, genius, she might like not bleeding to death. She might enjoy prosthetic devices. Woe is me, Phil! Probably will enjoy all the attention she’s bound to get. Terrible. Just terrible. Our genius knows she’d be much better off bleeding to death on a piping hot desert dune, slowly, agonizingly, fading away into the hellish heat. But surely by doing so, she is departing, “the good old fashioned way.” (misbehaving brat!)

    The warped thinking of these complete idiots always leads to nonsense, needless suffering and tragic results. They need a black box warning on everything they do.