Is This Really the Best Drug to Treat Depression in Children?

William T. Basco, Jr, MD, MS


December 14, 2016

Treating Depression in Children: Which Drug?

Major depressive disorder is experienced by 2%-3% of early school-aged children and 5%-6% of all adolescents.[1] Data show an increase in the use of antidepressants by children and adolescents in recent decades.

In a recent study, Cipriani and colleagues[1] reviewed existing data to determine whether there was a preferred antidepressant to treat these children. This was a meta-analysis of double-blind, randomized, controlled trials (published in English or other languages) that compared antidepressant drugs with either placebo or other antidepressants. The participants in all studies were children and adolescents aged 9-18 years.

Studies were included in the analysis only if they used the suggested therapeutic dosing range of the antidepressant medications. These included amitriptyline, citalopram, clomipramine, desipramine, duloxetine, escitalopram, fluoxetine, imipramine, mirtazapine, nefazodone, nortriptyline, paroxetine, sertraline, and venlafaxine. The outcomes of interest were the overall change in depressive symptoms, as well as the frequency of discontinuing therapy because of adverse events. The researchers also evaluated suicidal ideation or behavior as potential side effects.

More than 5700 potential citations were identified, from which 34 randomized controlled trials that met criteria were chosen for analysis. The total number of enrollees was 5200, with an average of 159 per trial. The mean age of the study children was 13.6 years, and the sample was 53% female.

The median duration of acute treatment was 8 weeks, and more than 80% of the trials focused on children with at least moderate or severe depressive symptoms. Two thirds of the trials were supported by industry. Only four trials (12%) were deemed by the researchers to have a low risk for bias. Most of the drugs were evaluated in only one or two trials, with the exceptions being paroxetine (five trials) and fluoxetine (10 trials).

Only three drugs demonstrated greater efficacy than placebo: fluoxetine, escitalopram, and sertraline. The tolerability of duloxetine, imipramine, sertraline, and venlafaxine was lower than that of placebo.

When the combined outcome of efficacy and tolerability was examined, only fluoxetine consistently performed better than placebo. Venlafaxine was the only drug associated with a significantly greater risk for suicidal behavior or ideation compared with placebo, as well as in comparison with five other drugs. In the studies of fluoxetine, 76.6% of the children experienced a positive response as determined by accepted measures.

The study authors concluded that only fluoxetine was significantly more effective than placebo, and this effect size was considered to be in the medium range. They also noted that the overall quality of the evidence was low or very low, on the basis of objective scoring criteria.


This study is valuable to the pediatric community on several levels. First, it should give us pause that the drug with the most pediatric clinical trial testing had only 10 studies that met the review criteria. Two of the drugs had only a single placebo-controlled trial, and seven had only two placebo-controlled trials. The conclusion that the general state of knowledge about the effectiveness and tolerability of these agents in adolescents is very limited will not surprise anyone.

There is a very good take-home point—that fluoxetine is the drug with the most pediatric data behind it, and the only drug that has consistently shown efficacy and tolerability. Therefore, for providers with limited training or experience with these medications, it would seem that fluoxetine should be the primary drug to consider prescribing.

However, the apparent support for fluoxetine in these studies doesn't mean that the other drugs are not effective. I can't help wondering whether the apparent effectiveness of fluoxetine compared with other drugs is a result of having 10 trials, which provides more stable estimates of its effectiveness. It may well be that other drugs are effective, but with only two or three trials averaging 159 children with which to judge the effectiveness of these drugs, we simply may not have enough data. I chose not to summarize some of the drug/drug comparison trials that were evaluated, because many of them were only single trials.

The article supplies lots of interesting reading and extensive appendices for anyone who wants to learn more, but the take-home message is simple.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.