Melatonin Is Not a Sleeping Pill

The investigation

The claim

“Melatonin is a natural sleep aid. Take five or ten milligrams before bed.”

The evidence

Melatonin is a timing signal, not a sedative. It does not knock you out; it tells your body what time it thinks it is. Reviews of its use recommend physiologic doses of 0.1 to 0.3 mg taken at bedtime, precisely to avoid the effects of overshooting.

A double-blind, placebo-controlled trial in adults over 50 compared 0.1, 0.3 and 3.0 mg against placebo with sleep measured by polysomnography. The physiologic 0.3 mg dose restored sleep efficiency and returned night-time melatonin to a normal range. The 3.0 mg dose also improved sleep — and induced hypothermia, and left plasma melatonin still elevated into the daylight hours.

That last part is the part that matters. An oversized night-time dose does not stop working at dawn. It keeps telling your body it is night, into a morning you are trying to be awake for.

How much sleep does it actually buy? A meta-analysis of 19 randomised placebo-controlled trials in 1,683 people found melatonin reduced the time taken to fall asleep by a weighted mean of about 7 minutes, and increased total sleep time by about 8 minutes. Statistically real. Modest.

Then the bottles. Researchers assayed 31 commercial melatonin supplements and found actual melatonin content ranging from 83% below to 478% above the labelled amount. Lot-to-lot variation within a single product ran as high as 465%. More than 71% missed their own label by more than 10%. And serotonin — a controlled substance used to treat neurological disorders — was found in eight of them.

Now the part almost nobody selling it will tell you, and the reason this episode exists in a show about autoimmune disease at all. Melatonin is not a neutral sleep molecule. It is immunostimulatory. It acts on helper T cells and on T and B cell precursors, and it is antiapoptotic — it keeps immune cells alive that would otherwise have died. In a body whose problem is an immune system attacking its owner, those are not obviously the properties you want to add at bedtime, every night, indefinitely.

The animal work is the cleanest version of the worry. In rats given Freund's complete adjuvant — the standard laboratory model of rheumatoid arthritis — removing the pineal gland, and with it the animal's own melatonin, produced a significantly LESS pronounced inflammatory response. Giving melatonin back at a low dose restored the arthritis to normal. Giving it at a pharmacological dose made it worse. The dose-response runs in the direction that should make anyone pause before taking ten milligrams a night for a year.

In people with rheumatoid arthritis, melatonin is not only circulating — it is in the joint. Researchers found it present in the synovial fluid of RA patients, and found that synovial macrophages, the cells doing the damage, carry melatonin binding sites. That is the missing piece a mechanism needs: not just a hormone that is around, but a receptor on the cell that is causing the disease.

So it was tested. Seventy-five people with rheumatoid arthritis were randomised to melatonin 10 mg at night or an identical placebo, on top of their existing treatment, for six months. The inflammatory markers moved, and they moved the wrong way: erythrocyte sedimentation rate rose in the melatonin group against placebo, neopterin rose, and kynurenine rose. In the placebo group ESR and neopterin were significantly trending DOWN over the same six months; in the melatonin group they were not. The authors' own word for the pattern was proinflammatory.

And here is where an honest reading has to slow down, because that trial is quoted as if it ended there. It did not. Interleukin-1 beta, interleukin-6 and TNF-alpha — the three cytokines that actually drive rheumatoid arthritis, and the three that every modern RA drug was built to block — did not change at all. Neither did the patients. Across six months there was no significant effect on any clinical assessment of symptoms. What that trial found was biochemistry moving while the disease stood still.

Then the foundation moves. Much of the disease-promoting argument rests on a 2002 finding that night-time melatonin is HIGHER in people with RA than in healthy controls — too much of a proinflammatory hormone, in the people with the inflammatory disease. In 2023 a study of eighty-four RA patients against controls measured the opposite: plasma melatonin markedly LOWER in RA, 117 picograms per millilitre against 330, and lowest of all in the patients with the highest disease activity. Both cannot be true. The newer, larger measurement points the other way, and any argument built on the older one has to be rebuilt or dropped.

And the direction is not even the same across autoimmune diseases. In multiple sclerosis the relationship is inverted and the evidence is considerably stronger. A 2015 paper in Cell showed that melatonin levels correlate NEGATIVELY with MS activity — relapses cluster in the seasons when nights are short and melatonin is low — and went on to show the mechanism: melatonin induces a transcription factor called Nfil3, which blocks the differentiation of pathogenic Th17 cells, while boosting protective Tr1 cells through interleukin-10. In the laboratory model of MS, melatonin made the disease better. Th17 cells drive rheumatoid arthritis too. The same molecule appears to suppress them in one disease and is suspected of promoting another.

So what should someone with an autoimmune condition actually do? The honest answer has three parts and none of them is the confident one. First: the caution is real, it is mechanistically grounded, and it is specific — it applies to the pharmacological doses people actually buy, five and ten milligrams, not to the 0.3 mg physiologic dose this episode has been arguing for all along. Second: the one randomised trial in humans that looked for harm found laboratory markers moving and patients not moving, over six months. Third: nobody has run the trial that would settle it, because melatonin cannot be patented — which is the same silence this show keeps finding, and it is a fact about the funding rather than about the molecule. If you have an autoimmune disease, this is a conversation to have with the person managing it, and the thing to bring to that conversation is the dose.

The verdict

Right molecule, wrong dose, unreliable bottleThe problem is not that melatonin does nothing. The problem is that the dose sold to you is roughly ten to thirty times the one that was studied, that the benefit is measured in single-digit minutes rather than hours, and that the number printed on the bottle has a documented habit of being wrong in both directions. If you use it, use it for circadian problems — shift work, jet lag, a delayed sleep phase — at a dose near 0.3 mg, taken at a consistent time. If you are using ten milligrams to sedate yourself, you are not treating insomnia. You are overriding a clock.

Change our mind

A dose-ranging, double-blind trial in adults with insomnia that tested 0.3 mg against the 5 and 10 mg doses the shelf actually sells, with polysomnography and next-morning alertness as endpoints. And, separately, a randomised trial of PHYSIOLOGIC-dose melatonin in autoimmune disease measuring clinical activity rather than laboratory markers — the existing 10 mg trial found biochemistry moving while patients did not, and that question has been left open for nearly twenty years.

1. Zhdanova IV, Wurtman RJ, Regan MM, Taylor JA, Shi JP, Leclair OU. Melatonin treatment for age-related insomnia.J Clin Endocrinol Metab · 2001 · 86(10):4727–4730

   doi.org/10.1210/jcem.86.10.7901 · PMID 11600532
2. Ferracioli-Oda E, Qawasmi A, Bloch MH. Meta-analysis: melatonin for the treatment of primary sleep disorders.PLoS One · 2013 · 8(5):e63773

   doi.org/10.1371/journal.pone.0063773 · PMID 23691095
3. Erland LAE, Saxena PK. Melatonin Natural Health Products and Supplements: Presence of Serotonin and Significant Variability of Melatonin Content.J Clin Sleep Med · 2017 · 13(2):275–281

   doi.org/10.5664/jcsm.6462 · PMID 27855744
4. Zhdanova IV. Melatonin as a hypnotic: pro.Sleep Med Rev · 2005 · 9(1):51–65

   doi.org/10.1016/j.smrv.2004.04.003 · PMID 15649738
5. Lerner AB. Melatonin — without the hype.Adv Exp Med Biol · 1999 · 460:1–3

   PMID 10810494
6. Maestroni GJM, Cardinali DP, Esquifino AI, Pandi-Perumal SR. Does melatonin play a disease-promoting role in rheumatoid arthritis?J Neuroimmunol · 2005 · 158(1–2):106–111

   doi.org/10.1016/j.jneuroim.2004.08.015 · PMID 15589043
7. Maestroni GJM, Sulli A, Pizzorni C, Villaggio B, Cutolo M. Melatonin in rheumatoid arthritis: synovial macrophages show melatonin receptors.Ann N Y Acad Sci · 2002 · 966:271–275

   doi.org/10.1111/j.1749-6632.2002.tb04226.x · PMID 12114283
8. Forrest CM, Mackay GM, Stoy N, Stone TW, Darlington LG. Inflammatory status and kynurenine metabolism in rheumatoid arthritis treated with melatonin.Br J Clin Pharmacol · 2007 · 64(4):517–526 · randomised, placebo-controlled, n=75

   doi.org/10.1111/j.1365-2125.2007.02911.x · PMID 17506781
9. Wróbel A, Szklarczyk J, Barańska I, Majda A, Jaworek J. Association between levels of serotonin, melatonin, cortisol and the clinical condition of patients with rheumatoid arthritis.Rheumatol Int · 2023 · 43(5):859–866 · the measurement that contradicts the 2002 finding

   doi.org/10.1007/s00296-023-05296-4 · PMID 36912941
10. Farez MF, Mascanfroni ID, Méndez-Huergo SP, et al. Melatonin Contributes to the Seasonality of Multiple Sclerosis Relapses.Cell · 2015 · 162(6):1338–1352 · the opposite direction, with the mechanism

    doi.org/10.1016/j.cell.2015.08.025 · PMID 26359987
11. MacDonald IJ, Huang C-C, Liu S-C, Tang C-H. Reconsidering the Role of Melatonin in Rheumatoid Arthritis.Int J Mol Sci · 2020 · 21(8):2877 · the review that calls the question unsettled

    doi.org/10.3390/ijms21082877 · PMID 32326031