Antidepressant use in children, adolescents may double aggression, suicidality risk

View on the News

Serious harm may be underreported in clinical trials

The analysis by Tarang Sharma and her colleagues suggests that clinical study reports probably underestimate the possible negative effects of antidepressants, Dr. Joanna Moncrieff wrote in an accompanying editorial.

“To reevaluate antidepressants properly, we need access to original data from trials, but we also need research that clarifies the whole range of antidepressant-induced … alterations” after acute treatment, use, and withdrawal, she wrote.

Dr. Moncrieff, a senior lecturer in the division of psychiatry at the University College London, is a cochair of the Critical Psychiatry Network. Her editorial also was published in BMJ online (2016 Jan 27. doi: 10.1136/bmj.i217.




Children and adolescents who take certain SSRIs and SNRIs appear to be at double the risk of aggression and suicidality, however, no such associations are found in adults, according to results of a systematic review and meta-analysis published online Jan. 27.

Those are some of the results of the review and meta-analysis involving more than 18,500 patients conducted by Tarang Sharma of the Nordic Cochrane Centre in Copenhagen, and her colleagues of five antidepressants. The researchers sought to assess the associations between selective serotonin reuptake inhibitors and four outcomes: mortality, suicidality, aggression, and akathisia. Their report was published online in BMJ (2016 Jan 27. doi: 10.1136/bmj.i65).


Ms. Sharma and her colleagues analyzed 68 clinical study reports involving 70 trials investigating duloxetine, fluoxetine, paroxetine, sertraline, and venlafaxine. Of the included trials, 11 involved children and adolescents, and 63% (n = 44) excluded patients at risk for suicide.

Overall, the researchers found no significant differences in akathisia (odds ratio, 2.04; 95% confidence interval, 0.93-4.48), suicidality (OR, 1.21; 95% CI, 0.81-1.74), or mortality (OR, 1.28; 95% CI, 0.40-4.06). However, they did find an association with aggressive behavior (OR, 1.93; 95% CI, 1.26-2.95).

When they separated the adult data from those of younger patients, however, they found that children and adolescents had double the risk of aggression (OR 2.79, 95% CI: 1.62-4.81) and suicidality (OR, 2.39; 95% CI, 1.31-4.33). The risk of akathisia was lower (OR, 2.19; 95% CI, 0.48-9.65). Data for adults did not demonstrate a significant increase in aggression (OR, 1.09; 95% CI, 0.55-2.14), suicidality (OR, 0.81; 95% 0.51-1.28), or akathisia (OR, 2.00; 95% CI, 0.79-5.04).

The authors noted several limitations in the trials, such as insufficient lead-in periods. In addition, there were inconsistencies between the trials in how the events were coded. Further, 46% of the trials included in the analysis did not note which coding dictionary was used. The authors point out that the limitations of the trials in terms of clinical study reports “may have led to serious underestimation of the harms” and that there is still uncertainty of the true risks.

The Laura and John Arnold Foundation funded the study. The authors report no disclosures.

Next Article: