Do antidepressants work? They do. New evidence comes from a study of children and adolescents, patients in age groups where the efficacy of antidepressants has been hard to demonstrate. It’s just the sort of research that critics of drug trials ought to find convincing.
Antidepressants seem to do less good in the young. For decades, there were few studies showing that the medications helped at all in patients under age 18. Scientists had even put forth theories – I review some in Against Depression – explaining why antidepressants ought not to have the same effects in children as in adults.
Then in 1997 and again in 2002 and 2004, Prozac was shown in a series of trials to have substantial benefits. The grounds for debate moved to other issues. How do antidepressants affect the developing brain? How often do the medications arouse suicidal thoughts? And when the medicines help, how long should they be taken?
The trials published in 2004—there has been follow-up work as well—were of particular interest. They were conducted by a research collaborative, the focus was psychotherapy as much as pharmacotherapy, placebo response rates were moderate, and the results were bound to be published whatever the outcomes. The findings were that psychotherapy and medication work best, and also—surprisingly—that when it comes to “monotherapies,” Prozac works much better than psychotherapy alone. Placebo came in a distant fourth.
Prozac did seem to stir up suicidal thoughts in some patients who did not have them initially; the addition of psychotherapy muted that effect. Looking only at averages, Prozac lowered the frequency of suicidal thoughts, and the combination of medication and psychotherapy had a very substantial effect in decreasing suicidality.
It’s not clear in my mind whether psychotherapy got a fair shake in that trial. The model used was cognitive-behavioral, an approach that I think has been hyped. But the medication worked as a treatment for depression, with a response rate of over 60% when placebo responses were under 35%.
The current report, by Graham J. Emslie and others at the University of Texas Southwestern Medical Center, has similar credentials. The NIMH funded the research. The trial was large enough, with over 300 subjects between the ages of seven and 18 (mean age 11.5 years) initially evaluated. Whatever the findings, they were very likely to find their way into print. Once again, the drug-placebo difference was substantial, 27%—this, in a population that responds poorly to antidepressants. The researchers went on to ask the length-of-treatment question: if a subject was doing well on Prozac at 12 weeks, was it safe to discontinue the medication?
The answer is that early-onset depression simply is a hard condition to treat. The study resumed at the 12-week point, enrolling patients who were then doing well, and continued for six more months. In that interval, almost 70 per cent of the patients who stopped taking Prozac relapsed, most in the first month or two. But 42 per cent of young patients who stayed on the drug also fell back into depression. A second analysis, looking at a smaller group of patients and using a stricter definition of relapse found that Prozac halved the relapse rate.
The results argue for quite vigorous treatment. Patients whose disease had remitted but who had even one lingering symptom of depression were much more likely to relapse. Among those who had completely remitted (with no residual symptoms), Prozac was more effective, with a 25% subsequent relapse rate; placebo was shockingly ineffective — 67% relapsed.
Patients and their families seemed well disposed to the medication. Of those who declined to enter the follow-on trial (to study the effect of discontinuing Prozac), almost half gave as their reason a fear that the patient would be given a placebo rather than the active drug. On the downside, one patient in the Prozac group attempted suicide. This was not a case of “new-onset suicidal ideation.” The patient had suicide plans before being put on the medication and had a “history of self-injurious behavior”—but this attempt took place after the medicine had apparently worked, that is, after the patient’s depression had initially remitted.
None of this research proves that antidepressants should be used in children or adolescents. We know too little about long-term outcomes. But the research does speak to the question of short-term efficacy.
Overall, the Emslie study confirms what we know about depression, antidepressants, and antidepressant studies. Depression is a stubborn, recurrent illness. Antidepressants have limitations, and the drugs come with risks. But the medications are of help to a substantial percentage of patients. In terms of demonstrating this efficacy, the tougher the population under study and the more painstaking the research, the greater the gap between medication and placebo.
The reception of this study also confirms what I suspect, that when it comes to antidepressants, only controversies—really, only negative or critical studies—have news value. Although the American Journal of Psychiatry featured this story (in an editorial and in its educational material for doctors), and although there is enormous interest in the subject of adolescents and antidepressants, so far as I can tell, there has been no press coverage of the research.