Does melatonin work for you?
The only real answer is your own nights. Mark fourteen nights on or off, pick one thing to watch before you start, and see if the difference beats your normal swing.
5.5 minutes to fall asleep faster on melatonin (median 19.5 against 25), with a normal swing of about 7.7. 1 night dropped.
The difference is smaller than your nights already vary. It may be doing a little, or nothing you can see. Try it at the right hour before you decide.
Pick the outcome before you start, and don't change it halfway. The verdict rule is ours: a difference bigger than the standard deviation of your off nights counts as real for you.
What your verdict means
Keep it. Your melatonin nights beat your own night-to-night swing. Stay on the smallest dose that did it, at the same hour, and run the test again in a few months to see if you still need it.
Inside your normal swing. The difference is smaller than your nights already vary on their own, which is what most people should expect, since the average effect is about 7 minutes. Check the hour first. If it was already right, it's fine to stop.
Try the right hour first. You took it at lights out, or your on nights came out worse. Move the dose 1 to 3 hours before bed and test again, since earlier doses did better in a 2024 analysis of 26 trials.
A few more nights. Fewer than 4 usable nights of each, once the drinks and late nights are dropped. Keep going until you have them.
The example, worked
| Nights | Minutes to fall asleep | Median |
|---|---|---|
| Off | 25, 18, 32, 22, 40, 20, 28 | 25 |
| On | 20, 15, 24, 19, 22, 17, with one drinks night dropped | 19.5 |
That's 5.5 minutes faster, inside a normal swing of about 8, and about the size the trials found when they measured it: across 19 trials, people fell asleep about 7 minutes faster, 5.5 when it was measured rather than self-reported. Verdict: Inside your normal swing.
How the test works
Drop the nights with drinks or a late night, since a moderate evening drink cut people's own melatonin by about a fifth. Compare the median of your on nights with the median of your off nights. Your normal swing is the standard deviation of your off nights, and a difference bigger than that counts as real for you. That rule is ours.
Take it at the right hour while you test. A 2024 analysis of 26 trials found earlier doses did better, so a test run at lights out mostly measures the wrong hour.
3 sources
- Ferracioli-Oda E, Qawasmi A, Bloch MH, Meta-analysis: melatonin for the treatment of primary sleep disorders, PLoS One, 2013pubmed.ncbi.nlm.nih.gov
- Cruz-Sanabria F and others, Optimizing the time and dose of melatonin as a sleep-promoting drug: a systematic review of randomized controlled trials and dose-response meta-analysis, Journal of Pineal Research, 2024pubmed.ncbi.nlm.nih.gov
- Rupp TL, Acebo C, Carskadon MA, Evening alcohol suppresses salivary melatonin in young adults, Chronobiology International, 2007pubmed.ncbi.nlm.nih.gov
