Review of: Strawn JR et al, J Am Acad Child Adolesc Psychiatry 2018;57(4):235–244.e2

Antidepressants are part of the first-line treatment for severe childhood anxiety disorders when removal of stressors and psychotherapy are not enough, but are all antidepressants created equal in this situation?

A recent meta-analysis shows that antidepressants have a moderate effect size of 0.56 for treating anxiety disorders in children (see CCPR, Jan/Feb 2018), but do we have the data to further break that down? Another meta-analysis was recently performed that can further guide us in tailoring our medication choices for pediatric anxiety disorders.

In this meta-analysis, the authors pooled data from 9 randomized placebo-controlled trials that compared either an SSRI or an SNRI to placebo for the treatment of social, generalized, and/or separation anxiety disorders. Total sample size was 1,805 children ages 5–17 years, with 53% male. All studies were done in outpatient clinics and had a mix of federal and industry funding sources. The follow-up periods varied from 8 to 16 weeks, with a median of 10 weeks. Four SSRIs (fluoxetine, fluvoxamine, paroxetine, and sertraline) and three SNRIs (atomoxetine, venlafaxine, and duloxetine) were used in the studies. The primary outcomes were the time it took to see improvement, how treatment response differed between SSRIs and SNRIs, and differences in low-dose vs high-dose SSRIs. Rating scales, most commonly the Pediatric Anxiety Rating Scale (PARS), were administered every 2 weeks.

Overall, children improved quickly compared to placebo, with a statistically significant difference in the rating scales by week 2 (p = 0.005) and a clinically significant difference seen by week 6 (p = 0.001). SSRIs outperformed SNRIs over the entire treatment course, with a statistically significant difference emerging by week 2 (p = 0.021), but both classes of medications resulted in significant improvement compared to placebo by week 2. For the high-dose vs low-dose SSRI comparison, high-dose was considered > 1.5 fluoxetine equivalents (> 49.5 mg) per day. High-dose SSRI treatment resulted in earlier improvement (week 2), while low-dose resulted in later improvement (week 6). However, over time, there was no significant difference (p = 0.638), but the variance was greater for the low-dose group (p < 0.001).

This meta-analysis found that, overall, SSRIs resulted in greater improvement in childhood anxiety disorders than SNRIs, and that high-dose SSRIs led to earlier improvement. The authors postulate that the differences may be due to an underdeveloped noradrenergic system in children compared to the serotonergic system, or due to anxiety disorders themselves being caused by more dysfunction in the serotonergic system.

CCPR’S Take
When making medication decisions, the more information we have, the better. This study confirms that both SSRIs and SNRIs are effective in treating pediatric anxiety disorders. And, all other things being equal, SSRIs may give better results. Unless you have a reason to avoid SSRIs, using them as the first-line medication choice makes sense. High-dose SSRIs may give faster results but may come at a cost of increased side effects. Always be on the lookout for activation (which is generally more common with SSRIs than SNRIs) and other side effects.

Editor’s note: Generally speaking, a moderate effect size tells you that, if you pick randomly from the treated group vs the control group, you have a better than 50% chance that the person responded.