Melatonin for Pediatric Insomnia in Neurodevelopmental Disorders: A Dose-Response Meta-Analysis

Abstract

Introduction: Insomnia affects 50–80% of children with neurodevelopmental disorders (NDDs), including autism spectrum disorder (ASD) and ADHD. Melatonin is widely used off-label, but optimal dosing and long-term efficacy are poorly characterized.

Methods: Systematic review and one-stage dose-response meta-analysis of RCTs of exogenous melatonin (any formulation) for sleep onset latency (SOL) in children (2–18 years) with NDDs. Databases: PubMed, Embase, CENTRAL. Primary outcome: minutes to sleep onset (actigraphy or diary). Restricted cubic splines modeled dose-response (0.5–10 mg). Secondary: total sleep time (TST), adverse events.

Results: Twelve RCTs (n=873 children) were included. Melatonin reduced SOL by a mean of 39 minutes (95% CI 31–47) versus placebo. Dose-response analysis showed a logarithmic curve: 1 mg reduced SOL by 28 minutes, 3 mg by 38 minutes, 6 mg by 42 minutes, and 10 mg by 44 minutes (plateau). TST increased by 32 minutes (22–42). Adverse events: morning drowsiness (18% melatonin vs. 9% placebo, number needed to harm=11), no serious events. Extended-release formulations did not differ from immediate-release for SOL but improved sleep maintenance (fewer night wakings, mean difference –0.6 per night).

Discussion: Melatonin produces clinically meaningful SOL reduction in pediatric NDDs with dose-response plateau at 6 mg. Higher doses (>6 mg) offer no added benefit and increase drowsiness risk. Extended-release benefits sleep maintenance. Limitations: short trial duration (median 4 weeks), no long-term safety data.

Conclusions: Starting dose of 1–3 mg melatonin is effective for sleep onset insomnia in children with NDDs, titrating to maximum 6 mg if needed. Extended-release formulations preferred when night wakings are prominent.