Melatonin at the right time (not 'right before bed')

Double-blind placebo-controlled RCT, 105 medication-free children with ADHD + chronic sleep-onset insomnia. 3-6 mg melatonin for 4 weeks advanced dim-light melatonin onset (DLMO) by ~44 min vs a 13-min delay on placebo, advanced sleep onset and added ~20 min total sleep time. Direct demonstration that melatonin acts as a circadian phase-shifter in ADHD.

3-arm RCT, 51 adults with ADHD + DSPS. Low-dose 0.5 mg melatonin (timed) advanced DLMO by ~1.5 h (2 h with added bright light) and reduced ADHD symptoms ~14%. Important caveat: despite the phase advance, melatonin did not by itself advance actual sleep timing or improve subjective sleep without accompanying behavioral sleep scheduling.

Meta-analysis of 5 RCTs in ASD: melatonin significantly improved sleep duration and sleep-onset latency vs placebo, with documented abnormal endogenous melatonin circadian rhythms in ASD supporting a chronobiological mechanism.

More recent systematic review/meta-analysis confirming melatonin improves total sleep time, sleep-onset latency and sleep efficiency vs placebo in autistic individuals; well tolerated.

Supports the timing principle: melatonin for DSWPD is most effective at the minimal effective dose given ~3-5 h before DLMO (i.e. hours before bedtime), acting as a chronobiotic rather than a hypnotic; larger/late doses are not superior.