Share |





The rush to black label (or blackball) SSRIs.

Sally Satel, MD
September 30, 2004

Last June, Eliot Spitzer kicked off a long hot summer for SSRI[1] drugs with his lawsuit against makers of the antidepressant Paxil. The New York Attorney General accused GlaxoSmithKline of "concealing" evidence that Paxil led to an increased risk of suicide in children. As autumn sets in, the controversy isn't showing any signs of cooling off anytime soon, especially since the FDA is itself now in the hot seat.[2]

As the pressure mounts, there is a real risk that the FDA will try to appease its critics rather than making a principled stand on the best research available. And that may wind up hurting the very children everyone is trying to protect.

Right now, a major point of debate is whether antidepressants should require a "black box" warning against suicidal ideation and behavior.[3] Politicians who believe that the FDA went out of its way to downplay risks, and anguished parents who blame their children's suicide on the drugs, are pressing for black box warnings on all SSRIs. Such a warning (appearing on package inserts in a bolded box) is the most stringent caution the FDA can issue short of declaring a medication contraindicated for a certain purpose or banning it outright.

Proponents of the black box warning believe that it will put the brakes on excessive prescribing and effectively force physicians to refer patients to child psychiatrists, who are better equipped to diagnose and treat depression. On the other hand, some worry that a black box will exert a chilling effect on prescribing and that children who need the medication will be denied it by skittish doctors and parents.

But there is a more basic reason for rejecting the black box: the evidence does not warrant it. Consider the findings published in a 131 page report called Review and Evaluation of the Clinical Data[4], completed on August 16, 2004:

  • Within the 15 clinical trials of pediatric depression they examined, youngsters on SSRIs were no more likely to experience an "emergence of suicidality" than those on placebo, nor was there any discrepancy between the two groups with respect to "worsening of suicidality."(p. 38-40)
  • What the report did find was that youngsters on SSRIs for other conditions (such as anxiety, or obsessive-compulsive disorder) as well as for depression, were 1.78 times as likely as those on placebo to exhibit "definitive suicidal behavior/ideation." It was this finding that received all the media attention. But not mentioned in the coverage was how this elevated risk emerged only when two discrete categories, "suicidal behavior" and "suicidal thinking," were combined to make a third, composite category called "definitive suicidal behavior/ideation." (p. 38)
  • None of the individual trials (for depression or other conditions) showed a statistically significant indication of suicide thinking or behavior compared to placebo (p.45)
  • Most of the suicidal thinking/behavior that did occur, occurred in trials with the highest proportion of patients with suicidal thinking/behavior before the trails. (p.46)

After the September 13 Advisory Committee hearing, media accounts universally trumpeted the observation that 2-3% of children taking SSRIs will have suicidal thoughts or behaviors because of the medications. On September 14 The New York Times reported "if 100 [youth] are given antidepressants, two or three will become suicidal who otherwise would not have been had they been given placebos."

This is not quite what the data say. What the FDA reported was that 5-6% of children on medications were noted to have suicidal thoughts or activity compared to 2-3% of children on placebo. This raises the intriguing question of why placebo treatments would produce suicidal reactions in the first place - unless such phenomena simply reflected the natural worsening of depression. Thus, the difference between drug and placebo effects is 2-3%. This is of interest, to be sure - after all, the expected result would be a lower risk with drug than placebo - but it is also puzzling. Indeed, those figures still do not answer the question of whether children on the drugs enjoyed some improvement.

A key piece of information might have helped - that is, what were the baseline values? In other words, what percentage of the children displayed suicidal thoughts or behavior at baseline, before they entered the clinical trial? It is possible, after all, that suicidality decreased from a higher pre-trial baseline, suggesting that interventions did indeed have some impact.

This was in fact the outcome of a trial published in the August 18, 2004 issue of the Journal of the American Medical Association. In brief, the study called the Treatment for Adolescent Depression Study (TADS) led by John March of Duke University Medical Center, measured suicidal ideation at baseline (pre-trial) - unfortunately, it did not obtain a baseline reading on recent self-harm or suicide attempts. Twenty nine percent of the entire sample expressed ideas of suicide before the trial began and at the end of the 12-week trial all subjects experienced a decline. "The impact of treatment with fluoxetine [Prozac] on reduction of suicidal ideation was identical to that of placebo," write the authors, "suggesting that fluoxetine on average does not increase suicidal ideation." As far as suicide-related events, however, 9 subjects on Prozac (8%) displayed "suicide-related events" compared to 4 (4%) with placebo.

Now, the interesting question is why subjects on placebo engaged in self-harm and suicide attempts less often than those on Prozac? No one can say with confidence, but a plausible explanation is that the activating side effects of SSRIs may offset its benefits.

As early as 1960 psychiatrists were aware that antidepressants (then it was mainly tricyclic antidepressants, such as Elavil, in use) could precipitate suicide in some patients. The neurotransmitter changes they induced often caused the patient's energy to return before his hopelessness faded. Patients no longer immobilized by their misery now had enough energy to end it. Another small fraction of patients could flip into mania or a manic psychosis, or develop a severe form of agitation (akathisia) that might also trigger a suicide attempt. This is why standard teaching for beginning psychiatry residents emphasizes that they watch patients very closely when starting an antidepressant.

In the mid-eighties the SSRI drugs came on the scene. Eventually they too were observed to spark suicidal activity and akathisia in a small percentage of patients.

Is it possible, then, that SSRIs have precipitated some actual suicides? Yes. (Though keep in mind, that during the clinical trials under review by the FDA no child killed himself). This is why doctors must watch patients very carefully after starting antidepressants - something pediatricians and primary care doctors may not do as often as they should.

But the larger question the FDA must answer for the public is whether these medications have prevented more suicidal activity than they have caused. Almost surely they have. For example, epidemiologic trends raise the possibility that antidepressants have helped prevent youth suicide. If lots of young people were committing suicide as a side effect of medication, it seems reasonable that we'd be seeing a spike in the adolescent suicide rate. Since the early 1990's, the incidence of suicidal ideation among adolescents is down over 40%, and actual suicides are down over 25%, according to the Centers for Disease Control.

Along these lines, a study in the Archives of General Psychiatry last year looked at the numbers of antidepressant prescriptions written for youth according to zip codes across the country and compared them with census data on youth suicide. The greater the increase in the prescriptions in particular geographic areas, the sharper the reduction in the teen suicide rate in those locations over the last decade.

The findings are far from definitive: the study did not control for potential risk factors such as substance use and so cannot be said to constitute definitive proof. To confirm the findings we need studies designed specifically to assess aggravation of suicidal thoughts and behavior in kids who are already suicidal - and we don't have this yet. Nor do we have studies that look at the long-term effects of treatment on the risk of suicide.

We also need more studies with children who are severely depressed - suicidal or not. Most of the studies the FDA examined were conducted with young subjects who were mildly or moderately afflicted. This is why the drugs rated as only slightly more effective than placebo in the trials. The sickest - and most medication-responsive - children are often excluded from participating. This leaves too many subjects enrolled in clinical trials with relatively mild to moderate depressions that would get better anyway, without medication.

In summary, stripped of the aggravating glare of the media, the data about suicide risk and SSRIs are real but not alarming. Warnings (not surrounded by a black box) are already on the SSRI package inserts. Right now, the important lesson is that childhood depression must be diagnosed expertly and, if treated with medications, done so by experienced child psychiatrists (or specially trained primary care docs) and there are just too few of both.

This is the message the FDA should be defending, while calling for further research into the underlying issues surrounding treatment of pediatric depression. After all, the FDA's mission is not to approve risk-free drugs, but drugs whose benefits exceed the harms caused by the diseases they are meant to treat - and suicide, despite its enormous downturn over the past decade, is still the third leading cause of death among adolescents.

The SSRI-suicide controversy is still unfolding, but the FDA shouldn't allow sensationalism in the media to unduly influence its treatment of SSRIs.


  1. Selective serotonin re-uptake inhibitors.
  2. Ten days after an FDA advisory committee reviewed the agency's data on SSRIs and pediatric depression, FDA procedures were the subject of scathing criticism from some members of the oversight subcommittee of the House Energy and Commerce Committee. "There is something terribly rotten at the FDA," said Rep. Peter Deutsch (D-Fla.) at the hearing, "No agency charged with protecting public health should have behaved with such indifference." (Baltimore Sun, Sept. 24)
  3. The recommendation for a black box came out of the proceedings of an FDA Advisory Council meeting on September 13-14. The council, which comprises many outside experts, was divided on the warnings (15 in favor, 8 against). The FDA will decide whether to adopt the recommendation quickly, perhaps as early as September 30.
  4. briefing/2004-4065b1-10-TAB08-Hammads-Review.pdf

Sally Satel, M.D. is a Resident Scholar at the American Enterprise Institute.

home   spotlight   commentary   research   events   news   about   contact   links   archives
Copyright Manhattan Institute for Policy Research
52 Vanderbilt Avenue
New York, NY 10017
(212) 599-7000