Verified by 
A recent study published in the New England Journal of Medicine found Zoloft and cognitive-behavioral therapy to be highly effective in the treatment of reasonably severe anxiety disorders in children and adolescents. I reported on the findings when they first appeared, hailing them as mostly good news, and promised to come back with further postings about the limitations of the research.
Today: thoughts about the medication arm of the trial.
The Journal is selective when it comes to research studies in psychiatry. I suspect that the editors chose this paper, because it looked at a large sample (489 subjects) and because the results were striking. The children had social anxiety, generalized anxiety, and separation anxiety. On medication, most kids responded, and the responses were substantial. The combination of medication and psychotherapy was particularly effective.
But it pays to read these studies in detail. I suppose that's a prime theme of this blog: The devil is in the data.
Here, the medication doses were high. In adults, the standard Zoloft dose for depression is 50 mg daily. A number of patients in my practice have done well on 25 mg. To treat anxiety, the starting doses tend to be low; at first, antidepressants can increase anxiety. But for unknown reasons, the final dose, to mitigate anxiety, can be substantial. The package insert sets the daily dosage range for obsessive-compulsive disorder, post-traumatic stress disorder, and social anxiety disorder at 50 to 200 mg. Again, in my clinical experience, for the more routine forms of anxiety, often modest doses suffice.
But in the current study of anxiety in children, the average Zoloft dose was between 133 and 146 mg daily. The children started at 25 mg, with the dose raised weekly as needed up to the eighth week of the study. Effectively, the number of pills must have been increased on almost every visit. By comparison, the placebo group got to pills resembling 176 mg, on average—so the active drug was being raised almost as fast as a pill that did nothing. (In this case, obviously, it did not even help through buoying expectations—that's why doctors kept adding pills.)
The average age of children in the study was under 11 years. According to data from the Centers for Disease Control, the average 11-year-old boy carries about half the weight of an average 20-year-old man; with girls and women, the ratio is about 60 percent. This sample had about equal numbers of boys and girls. If we assume that Zoloft distributes throughout the body, the children were taking a Zoloft dose equivalent to over 250 mg daily for an adult. So what the study seems to show is that, in children, if you use medication to pound the hell out of anxiety, it remits.
But it is not at all clear that it would be prudent to keep children on this sort of regimen. We have enough concerns about the hidden effects of chronic medication use on the developing brain. Surely it is worrisome to administer Zoloft to children at levels beyond what we are familiar with in adults.
To test the drug, the scientists designing the trial obviously thought it necessary to raise doses quickly. It just is too expensive to conduct a study where you behave as clinicians do in practice, giving a small dose and waiting for an effect, and then trying a small increase, and so on. The researchers write, "The schedule that we used, which emphasized upward dose adjustment in the absence of response and adverse events, suggests that the average end-point dose of sertraline [ie, Zoloft] in this study is the highest dose consistent with good outcome and tolerability."
Probably, if child psychiatrists can afford to be patient, it will turn out that lower antidepressant doses work fine for anxiety in children. But we don't know as much from this study. It may be that only very high doses work—and whether children can safely be left on those doses is also entirely unknown.
So a fuller account of this research's findings would say that Zoloft, particularly in conjunction with psychotherapy, is remarkably effective for anxiety disorders in children and adolescents—but at drug doses that may not be prudent. Further research will be needed to test the safety of the regimen under study. (Apparently, a six-month follow-up is in progress—which is a start.) And only new, more expensive trials can tell us whether the medication works at more modest doses, equivalent to what the field has been prescribing for adults.
In a future posting, I will explore the limitations of the psychotherapy arm of this same study.
