REVIEW OF: Solmi M et al, World Psychiatry 2020;19(2):214–232

Adverse effects are important considerations when choosing psychotropic medications, especially in children and adolescents. This study pooled available research to compare the safety profiles of various psychotropics used in the pediatric population. The authors screened thousands of studies of psychotropics for 78 possible adverse effects. The combined sample of these studies included roughly 337,000 children and adolescents: 120,000 on antidepressants, 66,000 on antipsychotics, 148,000 on ADHD meds, and 1,600 on mood stabilizers.

Among antidepressants, escitalopram and fluoxetine fared best, while venlafaxine was the worst owing to anorexia, abdominal pain, hypertension, and suicidality. Sertraline also performed poorly due to gastrointestinal issues, insomnia, and weight gain. Among antipsychotics, lurasidone was found to be the least problematic, with asenapine in second place. Olanzapine was the most problematic with a host of side effects including sedation, metabolic syndrome, and extrapyramidal side effects, with aripiprazole in second-last place.

Turning to ADHD medications, methylphenidate stood out to be the safest, although it came out only slightly ahead of lisdexamfetamine, while atomoxetine was the worst of the lot due to gastrointestinal issues and weight loss. Surprisingly, guanfacine also had significant issues such as reports of abdominal pain, sedation, and (notably) QT prolongation. Finally, among the mood stabilizers, lithium was the most well tolerated, while sodium valproate trailed the pack with significant adverse effects like weight gain, sedation, and cytopenia.

A quick tally of the safest available agents favors escitalopram and fluoxetine for depression, lurasidone for schizophrenia, methylphenidate for ADHD, and lithium for bipolar disorder.

CCPR’s Take
This study yielded some surprises and is worth a close read. While venlafaxine’s higher side effect profile was expected, we had not imagined sertraline to do so much worse than escitalopram and fluoxetine. Furthermore, we had, perhaps incorrectly, considered guanfacine and atomoxetine mild “starter” medications, even placebo-like, but here the stimulants stood out for their advantages in both safety and efficacy. Lastly, lithium and lurasidone deserve more consideration, despite having a narrow therapeutic range and insurance issues respectively.