Committee members voted unanimously in Lilly’s favor, and Prozac remained on the market. No warning that the drug might induce violent or suicidal urges was added to its label or to the labels of similar selective serotonin reuptake inhibitors (SSRIs) that the FDA would soon approve. Though the risk continued to concern some researchers and fuel lawsuits against drug manufacturers, the issue of SSRI-induced suicide and violence faded from the public eye. Until this summer.

In May 2003, new data presented to U.S. and British regulators showed that among 1,100 children enrolled in clinical trials of the antidepressant Paxil, those taking the drug were three times as likely to develop suicidal thoughts as children taking placebos. In early June, British regulators warned doctors not to prescribe the drug to children. Nine days later, the FDA announced it would conduct a detailed review of pediatric trials of Paxil, a review soon broadened to include other antidepressants. In August, Wyeth Pharmaceuticals warned doctors that twice as many children taking its antidepressant Effexor developed hostile behavior or suicidal thinking as did children taking a placebo.

Then, last month, British regulators went further and officially warned that they were urging doctors to stop prescribing to children a group of six antidepressants, including Paxil, Zoloft, and Effexor, because they caused an increase in suicidal thoughts and actions. “These products should not be prescribed as new therapy for patients under 18 years of age with depressive illness,” wrote Gordon Duff, chairman of the Committee on Safety of Medicines, in a letter to British physicians.

Prozac, the only SSRI approved for use in depressed children, was not included in the new warning. The British review didn’t find a significant increase in the risk of suicide-related events among children taking that drug.

These warnings, and the FDA’s decision, were a surprising turnaround that raised troubling questions: How did a concern that was dismissed so thoroughly more than a decade ago suddenly reemerge? How did drugs commonly prescribed by pediatricians and child psychiatrists, and widely viewed as nearly risk free, come to be seen as potentially dangerous?

Since Prozac came on the market in 1987, followed a few years later by Zoloft and Paxil, all SSRIs have benefited from a carefully cultivated reputation as revolutionary new drugs, vast improvements over previous antidepressants. “The SSRIs have sold themselves very heavily as safe and clean drugs compared to the old antidepressants,” says David Healy, a psychopharmacologist from the University of Wales College of Medicine and a leading critic of SSRI overuse. “Well, it’s not clear that they’re safer, and it’s not clear that they cause fewer side effects.”

One big advantage SSRIs do have over older drugs is that distressed patients who attempt to kill themselves by overdosing are unlikely to succeed; they can usually tolerate the pills. But that advantage may be offset by a side effect that SSRIs have been known to cause for at least 15 years, which may lie at the center of the current controversy. In some people, SSRIs induce a sensation called akathisia, a restless agitation that ranges from mere jitteriness to feeling you’re “jumping out of your skin.”

Researchers have been aware that SSRIs could trigger akathisia at least since 1990, when Harvard investigators reported on a group of six adult patients taking Prozac for depression who developed “intense violent suicidal preoccupation” after taking Prozac for two to seven weeks. Their fixation with violence and death abated when they stopped taking the drug. Similar symptoms were noted the next year in a paper describing six children aged 10 to 17 who developed “intense self-injurious ideation or behavior” on Prozac. One 14-year-old girl, who’d been depressed but never suicidal, began cutting herself and chanting that she wanted to die after three weeks on the drug.

These papers, along with thousands of spontaneous reports submitted to the FDA about Prozac-induced suicidal or violent acts—more reports than for any drug in FDA history—helped set the stage for the 1991 hearings. But since these cases occurred outside clinical trials and without a control group of depressed patients not taking the drug, they were discounted in favor of the evidence supplied by Lilly.

With the issue of suicide and violence largely put to rest, the full marketing might of the pharmaceutical industry was turned loose to promote the new antidepressants. The use of antidepressants and other psychotropic medications by children and adolescents exploded, tripling from 1987 to 1996, according to a recent study in the Archives of Pediatric and Adolescent Medicine, with most of that increase occurring after 1991. By 1996, the study found, six percent of American children and teenagers were taking psychotropic medications, one-third of which were antidepressants.

For its review, the FDA is going over data from all pediatric trials of antidepressants, reanalyzing the way reports of suicidal ideation among children were categorized. Critics contend that the agency has a long history of protecting the drug industry and are concerned that this reanalysis may minimize the risk, to the benefit of drug companies. They point to company memos, uncovered through legal actions, which reveal a sense of confidence within the companies that the FDA was on their side. For example, a memo from an Eli Lilly executive described one FDA official as “our defender.” Another from an executive of SmithKline Beecham, maker of Paxil, discussed the suicide issue and quoted an FDA official as saying the agency “does not see this as a real issue, but rather as a public relations problem.”

The FDA’s Thomas Laughren rejects the notion that his agency is protecting drug makers. “The goal here is to get to the truth,” he says.

As the agency reviews the data, experts will continue to debate the core question: how to reconcile reports that antidepressants trigger suicidal behavior with studies suggesting that antidepressants reduce suicide rates. Can antidepressants lower the suicide risk in some people while raising it in others?

David Healy says the drugs simply have different effects on different people. “My hunch is that, just as with adults, there’s a group of children who are suited to the pills and do very well on them, and an equally large group of kids who aren’t.” Among those who don’t do well, he says, are some who get much worse.

The emerging field of pharmacogenomics, which studies how people’s individual genetic makeup can affect their response to drugs, may help identify in advance those people who are likely to respond poorly to antidepressants. It’s now known that about seven percent of Caucasians have a variation of a gene (CYP450-2D6) that leaves them unable to efficiently metabolize a wide range of drugs, including SSRIs. Because their bodies can’t efficiently break down a drug like Prozac, some experts believe it may accumulate in their bodies and cause toxic reactions. Tests now available can identify people who are poor metabolizers, and some experts believe such testing may prove to be a valuable tool.

In any case, most experts agree that patients, especially children, should be monitored closely for side effects from the day they start taking the drugs. “Doctors have been educated to think that the SSRIs take four, five, six weeks to work,” Healy says. “But they can cause problems long before that.” Some people experience agitation just a few days after their first pill.

Therapists need to learn to recognize the signs of adverse side effects in children, or adults, taking SSRIs. “They may be more anxious or have unusual thoughts,” Healy says. “They may think about harming others or themselves. One of the things to ask would be simply, ‘Since you’ve been on these pills, have you had any strange dreams or nightmares?’ Move on from there to ask, ‘Have you had any strange thoughts during the day?’ The other thing to look out for is the opposite effect: kids who become absolutely fearless; they just don’t feel anxious at all.”

In February, an FDA advisory committee will revisit the issue of suicide among children taking antidepressants in a hearing certain to be contentious and emotional. Parents of teenagers will describe the suicides of their children. Experts will duel over data. And critics of current practices will, once again, urge the agency to require warnings about the risk of suicide on drug labels. A great deal will be at stake: the health and well-being of thousands of children, along with millions of dollars in drug sales. The question now is whether 12 more years of research and science will provide some useful guidance, and add some light to all of the heat.

top of page