Fluoxetine: A Problematic Trial in Ages 4-6

Portrait by Mona Nelson

The first randomized test of an antidepressant under age six, but the study has problems

STUDY: Shakibaei F et al, Advanced Biomedical Research 2026

STUDY TYPE: Randomized, double-blind, placebo-controlled trial

FUNDING: Isfahan University of Medical Sciences (government/academic grant)

Background

I’m reposting this review after discovering more problems in the trial, thanks to psychiatrist Martin Plöderl and colleagues for their thoughtful critique.

Fluoxetine stands as the antidepressant with the most robust evidence in children, but is only FDA-approved down to age eight. This trial asked how children as young as four respond to the SSRI.

The Study

• 40 children ages 4 to 6 with major depression, diagnosed by DSM-5 criteria and confirmed with the Depression scale (PFC-S).
• Fluoxetine (0.7 mg/kg daily, up to 15 mg) versus placebo for 60 days.

Results

Depression scores dropped sharply with fluoxetine, from about 4.9 to 3.1, while the placebo group barely moved, from 5.4 to 5.1. The difference showed up early, by 30 days, and held through 60 days. Parents reported no side effects on the antidepressant checklist (ASEC) in either group.

The Problems

1. Selective reporting of side effects. Two children on fluoxetine dropped out of the trial due to agitation, enough to leave the study entirely, but they were not counted in the final tally.
2. Moving the goal posts. The primary outcome in the protocol was the Revised Child Anxiety and Depression Scales–Parent Version, but shifted to the Preschool Feelings Checklist-Scale’ (PFC-S). Likewise, entry criteria changed from the protocol to the final report.
3. Additional discrepancies (detected with RegCheck, which compares registered to published study data: Changes in assessment periods (4 weeks and 3 months vs. days 30 and 60) and in requirement for a psychiatrist-confirmed diagnosis.
4. Data error and poor reporting. The paper states that 55 patients were assessed for the trial, and then 90 were randomized, and this kind of typo suggests poor reporting of data. Also, only p-values are reported, without test statistics (t, χ2, F), or confidence intervals.
5. Unusual responses. The placebo response was unusually low for a child study (6.5%), resulting in an unusually large effect size (1.5-1.7) for the medication and raising questions in the context of the other problems.
6. Other limitations. Small, single-center pilot study with just two months of follow-up.

Practice Implications

1. No trial has tested antidepressants in children of preschool age, and this study has too many flaws to change that. It needs a second editorial review and consideration for Retraction Watch.
2. Psychotherapy is first-line in this age group. Parent-Child Interaction Therapy has the strongest empirical support for depression in children under age eight.
3. Context (eg, sleep, diet, relationships, activity, and trauma) and family history (bipolar? schizophrenia?) are important in all assessments, particularly in the young.