The short version
- In a randomised crossover trial in 15 men, tolerance to the sedative effect of 50 mg diphenhydramine was complete within three days, with day-four sleepiness indistinguishable from placebo.
- An open-label study of 244 adults on prolonged-release melatonin, with 96 completing twelve months, found no evidence of tolerance, no rebound insomnia on discontinuation, and no suppression of the body's own melatonin production.
- A meta-analysis of 32 insomnia drug trials covering 3,969 participants found that 63.56% of the drug response was achieved in the placebo groups, including on polysomnographic measures.
- No herbal or mineral sleep supplement has been studied for tolerance over twelve months; the valerian and magnesium trials run for weeks.
- Stopping for a week is a free test: genuine tolerance to a sedative usually shows a discontinuation signature, while no change at all suggests the supplement had stopped contributing.
It is the most common message we get about anything in this category. It worked brilliantly for two weeks and then it stopped. People assume tolerance, the way it works with caffeine. The literature says something more awkward and more useful: for some sleep aids tolerance is real and fast, for others it does not appear at all, and for a large share of cases the honest explanation is that the effect was mostly placebo and placebo does not last.
Where tolerance is real, and how fast
The clearest example is the ingredient in almost every over-the-counter sleep aid on a pharmacy shelf. Richardson and colleagues gave 15 healthy men 50 mg of diphenhydramine or placebo twice a day for four days, randomised, double-blind, crossover, with objective and subjective sleepiness measures and psychomotor performance testing.
On day one, diphenhydramine produced significantly more sleepiness than placebo on both objective and subjective measures, and significant performance impairment. By day four, sleepiness on diphenhydramine was indistinguishable from placebo and the performance impairment was completely reversed. The authors describe tolerance as complete by the end of three days of administration, and call the speed of it remarkable.
Three days. If you have been taking a nightly antihistamine sleep aid for a month and it feels inert, that is not your imagination and it is not your fault. It is the pharmacology, and it was documented in 2002.
Where tolerance did not appear
Melatonin is the counter-example. Lemoine and colleagues followed 244 adults with primary insomnia on prolonged-release melatonin in an open-label study, of whom 112 completed six months and 96 completed twelve. There was no evidence of tolerance, discontinuation was not associated with rebound insomnia or withdrawal symptoms, and nocturnal urinary 6-sulfatoxymelatonin measurement showed no suppression of the body's own melatonin production after twelve months of nightly use.
Two honest caveats, because this is an unusually reassuring result. It was open-label, so everybody knew what they were taking, and the study population had already been through a placebo-controlled trial. But the physiological measure — endogenous melatonin production, unsuppressed after a year — is objective and does not care what anyone believed.
For botanicals and minerals, the truthful answer is that nobody has looked properly. The valerian trials are weeks long. The magnesium trials are weeks long. There is no twelve-month tolerance study of a herbal sleep blend, and if a brand tells you their formula does not lose effect over time, they are telling you something no one has measured.
The explanation people like least
Winkler and Rief meta-analysed the placebo arms of 32 randomised drug trials in primary insomnia — 82 treatment conditions, 3,969 participants, including polysomnography rather than just questionnaires. Placebo produced small-to-moderate significant improvements in objective sleep onset latency, total sleep time, wake after sleep onset and sleep efficiency, and in the matching subjective measures.
Their headline figure is the one to carry around: 63.56% of the drug response was achieved in the placebo groups. Not the reported response — the measured one, on a polysomnograph.
Put that next to how people actually buy sleep supplements. You order something during a genuinely bad stretch, which is the worst point in a fluctuating pattern. The first fortnight brings novelty, hope, a new bedtime routine and the ordinary drift of a bad patch back toward your average. Then the novelty wears off and your baseline reasserts itself. Nothing developed tolerance. The bad patch ended, and then a normal patch felt like failure.
How to tell the difference on yourself
There is a simple test and it costs nothing. Stop for a week. True pharmacological tolerance to a sedative tends to come with a discontinuation signature — a few worse nights before things settle. If you stop and nothing changes at all, the honest reading is that it had stopped contributing some time ago, and you have been paying a subscription for a habit.
The other diagnostic is what changed around it. If the supplement "stopped working" the same month you started drinking more in the evenings, a nightcap is a far better suspect than tolerance, and it acts on the exact half of the night people describe. Shift work, a new schedule, or a warmer season will all do the same thing.
What we'd actually tell you
On our own shelf: UNWIND contains melatonin and botanicals, and CALM is magnesium glycinate. Neither is a sedative, neither has been studied for twelve-month tolerance, and we are not going to claim otherwise. What the melatonin evidence suggests is that the melatonin component specifically is unlikely to be the thing that fades on you.
Three rules we would give a friend. Do not escalate the dose when something stops working — that is the reflex that turns a 1 mg melatonin habit into a 10 mg one with no evidence behind it. Take a week off every couple of months, both to check whether it is still doing anything and to reset the expectation. And if you have cycled through three or four sleep supplements in a year, the pattern is the point: that is chronic insomnia behaving like chronic insomnia, and it is a clinician's problem rather than a shopping problem. We would rather say that than sell you the fifth bottle.
Good questions
Do you build a tolerance to melatonin?
The best available evidence says no. In an open-label study following 244 adults, with 96 completing twelve months of nightly prolonged-release melatonin, there was no evidence of tolerance, no rebound insomnia when it was stopped, and no suppression of the body's own melatonin production. The study was not blinded, but the hormone measurement is objective and it did not budge.
Why did my sleep aid stop working after two weeks?
Most often because it was doing less than it appeared to. In pooled insomnia trials, 63.56% of the drug response also showed up in the placebo arm on polysomnography. People buy during a bad patch, and bad patches end on their own. If it was an antihistamine sleep aid, though, this is real pharmacological tolerance and it completes in about three days.
Is it safe to take a sleep supplement every night?
That depends entirely on which one, and for most botanicals nobody has run the study. Melatonin has twelve-month data with no tolerance and no withdrawal. Valerian, magnesium and herbal blends have been tested for weeks, not years, so any brand claiming long-term safety data for those is describing something that has not been measured.
Should I increase the dose if my sleep supplement stops working?
No. Escalating is how a 1 mg melatonin habit becomes a 10 mg one with no evidence supporting the jump, and with antihistamines a higher dose does not defeat tolerance that completed in three days. A better move is a week off, then an honest look at what else changed — alcohol, schedule, season — before you change anything you are swallowing.
Should I cycle off my sleep supplement?
A week off every couple of months is a reasonable habit, mainly as a test rather than a detox. It tells you whether the product is still contributing and resets your expectations. If you feel no difference at all during that week, you have your answer and you can stop paying for it. If several bad nights follow, that is worth noting too.
I have tried four sleep supplements this year. What now?
Stop buying the fifth. Cycling through products is what chronic insomnia looks like from the inside, and it is a pattern that responds to cognitive behavioural therapy for insomnia far better than to another formulation. Ask a clinician, and get screened for the physical causes — snoring, breathing pauses, restless legs, thyroid — before assuming the problem is which capsule you chose.
Sources
- Richardson GS, Roehrs TA, Rosenthal L, Koshorek G, Roth T. Tolerance to daytime sedative effects of H1 antihistamines. J Clin Psychopharmacol, 2002. View study
- Lemoine P, Garfinkel D, Laudon M, Nir T, Zisapel N. Prolonged-release melatonin for insomnia - an open-label long-term study of efficacy, safety, and withdrawal. Ther Clin Risk Manag, 2011. View study
- Winkler A, Rief W. Effect of Placebo Conditions on Polysomnographic Parameters in Primary Insomnia: A Meta-Analysis. Sleep, 2015. View study
These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.
