IN A BLACK BOX: A new analysis adds to the evidence that antidepressants are effective in young people, prompting some to renew questions about a "black box" warning required on the drugs since 2004. Image: © ISTOCKPHOTO.COM/JORDAN CHESBROUGH
The most comprehensive survey yet finds that the benefits of antidepressants outweigh the risks in children and teens during the first few months of treatment. The finding comes three years after the Food and Drug Administration (FDA) ordered pharmaceutical companies to put black warning labels—the strongest possible—on antidepressants cautioning that the drugs may increase the risk of suicidal behavior in kids.
The FDA ordered the beefed up warnings in response to a hearing by two of its advisory committees that looked at the potential risks of drugs such as Prozac and Zoloft.
The new finding does not dispute that antidepressants may heighten the risk of suicidal thoughts in young people. But researchers found that for every child or teen who experienced such thoughts, up to dozens more were helped by the drugs. And some experts argue that warning kids off of antidepressants may have backfired.
"We wanted to look at benefit and risk," says epidemiologist Jeffrey Bridge of the Columbus Children's Research Institute, lead author of the study, published online today by JAMA The Journal of the American Medical Association. "The FDA should continue to monitor the safety of these medications, [but] at some point we may want to revisit or reconsider the black box warning."
Bridge and colleagues reviewed data from 27 published and unpublished clinical trials in which patients younger than 19 years of age took antidepressants for depression, anxiety or obsessive-compulsive disorder (OCD). All were randomized studies that compared the effects of placebos with one of several medications, including the selective serotonin reuptake inhibitors Prozac, Paxil and Zoloft as well as others, during the first four months of treatment or less.
The team found no reported suicides, but 3 percent of those taking the drugs reported suicidal behavior compared with 2 percent of those taking a placebo. FDA researchers had found rates of 4 and 2 percent, respectively.
Bridge says antidepressants give the best risk–benefit ratio for anxiety. The analysis found that a physician would have to give antidepressants to about four kids to successfully treat one case of anxiety, compared with 143 to trigger one case of suicidal behavior. That ratio was similar for OCD, and lower for depression (one in 10 treated successfully versus one in 112 reporting suicidal behavior).
Researchers compile information from multiple studies in the hope of reaching a more accurate conclusion than any one trial can achieve by itself, but such meta-analyses work best if the pooled studies used comparable procedures and measurements of outcome.
The result bolsters the case that antidepressants are effective in young people, says psychiatrist John Mann of Columbia University, who adds that the black box warning may be harming kids by making them less likely to take the drugs. A recent Centers for Disease Control report found that the adolescent suicide rate increased in 2004 for the first time in a decade, which Mann says goes hand in hand with reduced prescriptions for antidepressants.
"It's not clear that there is a meaningful risk to the use of antidepressants," he says. "It is clear that that there's a meaningful risk to not using antidepressants."
Bridge notes that suicidal behaviors are most pronounced in depression, which accounted for 85 percent of all such behaviors in the study. "There's a risk for suicidal thoughts and behavior associated with depression itself," he says, "and perhaps the language of the black box warning can take that more into account."
The FDA prefers to see efficacy shown in a single trial, not a meta-analysis, Thomas Laughren, director of the agency's division of psychiatry products, says in a written statement. "That said," he adds, "we are not surprised that there is a tendency for the anti-depressants as a group to be effective in this analysis, however, it is quite a modest effect."