The new analysis, based in part on previously unpublished data, also concludes that treatment with antidepressants decreases the risk of suicide among adults of all ages.

"These results have to instill some additional confidence that prescribing these medications is not necessarily going to lead to suicidal thoughts or behavior," said Robert Gibbons, a professor at the University of Chicago and lead author of the study, published in Archives of General Psychiatry.

The findings -- based on data for kids and adults using fluoxetine (Prozac) and for adults on venlafaxine (Effexor) -- are not enough to change everyone's view of the risks of antidepressants, especially to kids.

"The authors in this study examined the risk of suicidal thinking or behavior associated with one drug, fluoxetine," said Jeff Bridge, a researcher at Nationwide Children's Hospital in Columbus, Ohio. "My view is that the weight of evidence shows a small but significant increased risk of suicidal ideation/suicidal behavior in pediatric patients treated with antidepressants."

Bridge's position is in line with the FDA's current stance on suicide risk for children taking antidepressants.

In 2004, the agency asked manufacturers of antidepressants to include what's called a "black box" warning on its packaging for the medications, alerting physicians, patients and parents to an increased risk of suicide among children taking the drugs. (See FDA announcement at 1.usa.gov/yjXP1G).

Three years later, the FDA expanded that warning to include young people up to age 25.

Gibbons has long been opposed to the labeling. As an advisory board member to the FDA, he voted against adding the warning to antidepressant packaging.

"I didn't think the data were very convincing, and I was concerned physicians would stop prescribing antidepressants," he told Reuters Health.

The FDA had looked for any reports of suicidal thinking or action among 4,400 children who were in clinical trials comparing an antidepressant drug to a fake drug called a placebo.

They found that suicidal thoughts or attempts were twice as common among the kids taking an antidepressant, although none of the children had committed suicide.

To get a better handle on the risk of suicide over the course of treatment, Gibbons' team gathered data from experiments that compared the antidepressants to placebo and that had measured suicide risk from the get-go.

Some results came from a study of adolescents by the National Institute of Mental Health and the rest came from two drug makers, Eli Lilly, which markets Prozac, and Wyeth, whose parent company markets Effexor.

The authors have served as consultants or have received research money from drug makers in the past, but Gibbons said neither company had access to this study -- which was funded by the federal government -- before it was published.

The analysis found that among the 708 children in the reviewed studies, the risk of having suicidal thoughts or attempts after eight weeks was no different between the kids who took Prozac and the kids who took the placebo.

Although fluoxetine is the only antidepressant drug approved for use in children, doctors can prescribe other drugs "off-label" to treat depression in kids.

"I think it's premature to extrapolate these findings to other antidepressants," Bridge told Reuters Health in an email.

The studies in the FDA review that found an increased risk of suicide looked at fluoxetine and eight other drugs.

In Gibbons' review, treatment with fluoxetine or venlafaxine resulted in a 90 percent decreased risk of suicidal thoughts or behaviors after eight weeks among adults and the elderly, compared to a 79 percent decrease after eight weeks of taking a placebo.

Gibbons said the drop in suicide risk seen in adults was tightly linked to the improvement in depression symptoms.

"What that means -- and it's not a surprising result - is if you don't treat depressive severity, you continue to have a high rate of suicide," he said.

At the beginning of the studies, children had higher rates of suicidal thoughts and behaviors overall than adults; about 20 percent of kids and three to five percent of adults started out thinking about or attempting suicide.

Gibbons said it was interesting that the kids' suicide risk didn't fall like the adults' did, even though the results indicated the antidepressants did work to relieve the kids' depression symptoms too.

"Suicide and depression are very strongly linked in adults and the elderly, and apparently not so strongly linked in children," he said.

Bridge said he'd like to see future studies examine whether antidepressants are tied to behavioral problems, such as hostility and agitation, and if an increased risk of those behaviors is related to more suicidal thoughts and behaviors.

As for the black box warning, Gibbons said he would support the FDA in gathering more data to better evaluate whether the warning is warranted, but he would not say whether it should be removed.

The FDA did not respond to requests for comment.

Keri McGrath-Happe, a communications manager at Eli Lilly and Company, wrote in a statement to Reuters Health that "at this time, we do not plan to discuss this matter further with the FDA."

She added that "some depressed individuals, on treatment and off, have worsening suicidal ideas and acts. It is prudent for clinicians and patients to remain vigilant for this possibility."

SOURCE: bit.ly/xmhAJq Archives of General Psychiatry, online February 6, 2012.

• A Permanent Economic Emergency