Going above FDA limits adds little benefit and doubles the dropout rate
STUDY: Farhat LC et al, JAMA Psychiatry 2024;81(2):157–166, PMID: 39025023
STUDY TYPE: Systematic review and meta-analysis
FUNDING: Non-industry: NIHR, Wellcome Trust
Background
Sunlight. Sleep. Water. Even these basic needs have an upper threshold where too much is harmful. This meta-analysis sought to find that threshold for sleep.
The Study
Researchers pooled data from 47 placebo-controlled randomized trials covering 7,714 adults with ADHD treated with methylphenidate or amphetamines. They analyzed dose-response curves and compared outcomes at licensed versus unlicensed (above FDA-limit) doses.
For methylphenidate, unlicensed doses did produce slightly greater ADHD symptom reductions on the ADHD Rating Scale (ADHD-RS) compared to licensed doses (effect size, SMD 0.23) — but the gain was small and likely not clinically meaningful. More importantly, the risk of dropping out due to adverse events more than doubled (odds ratio 2.02). For amphetamines, unlicensed doses showed no additional symptom benefit at all, with the dose-response curve plateauing around 30–35 mg/day.
Practice Implications
- Higher doses are riskier for amphetamines than methylphenidates.
- Though this study doesn’t clarify the harms, psychosis, mania, insomnia, and hypertension are dose-dependent.
- These psychiatric risks are greater in patients with complex comorbidities, like mood and personality disorders.
- Another study of psychosis/mania risk found a greater risk with amphetamine (not methylphenidate), especially above these “high” doses, which are similar to the cut-off of around 30-35 mg Adderall in the above study:
| Formulation | Medium | High |
| Dextroamphetamine | 15 mg | 30 mg |
| Vyvanse | 38 mg | 75 mg |
| Adderall | 18 mg | 36 mg |
—Chris Aiken, MD
Director, Psych Partners
Editor in Chief, Carlat Psychiatry Report
