In late 2003, the F.D.A. asked the pharmaceutical companies for all relevant information on every so-called adverse event in each of these pediatric antidepressant trials.

This is one reason that in antidepressant trials the placebo effect is so large and unpredictable.

Their findings suggest that patients in current antidepressant trials represent only a minority of patients treated in routine clinical practice for depression.

Dr. Posner said the Glaxo findings should be treated with caution, because the antidepressant trials done to date were not designed to evaluate suicide risk.

The mean number of past antidepressant trials was seven, and more than 55 percent of participants failed to respond to electroconvulsive therapy.

But depressed people who enroll in antidepressant clinical trials are a very select group who are not representative of depressed patients in general.

In one of the new analyses, researchers at the University of Ottawa re-evaluated data from 345 antidepressant trials for depression and other conditions, involving 36,455 men and women.

However, the more antidepressants an individual had already tried, the less likely they were to benefit from a new antidepressant trial.

After reviewing 24 antidepressant trials involving over 4,400 children and teens, the FDA concluded that young people using antidepressants are more likely to have suicidal thoughts and behavior.

Moreover, Fournier and his colleagues write, in some antidepressant trials prospective participants are given a placebo for several days or more before the trial officially begins.