Melatonin: Effective Sleep Aid or Public Health Threat?

A recent JAMA Network report questions the safety of melatonin gummies, but is there a role for melatonin in managing insomnia?

Sleep is one of the essential elements of human life, and insomnia one of its great curses. Few patients are more desperate for help than those suffering from severe insomnia. Given that virtually all of our pharmaceutical options for sleep aids have their problems — from concerns for addiction (the Valium-type meds) to erratic sleep behavior and a concerning mortality signal (Ambien and its ilk) to noisome side effects (dry mouth, dizziness, urinary retention, etc. for antihistamines and many sedating psychiatric medications) — it’s no surprise that a supplement like melatonin, which promises to be “safe and natural,” has found a huge niche among consumers. Should we encourage the use of melatonin or steer clear of the stuff?

Recent headlines might incline one towards the latter:

The study cited was recently published by JAMA Network, and found concerning irregularities in the actual contents of multiple “melatonin” gummy products. Twenty-five were studied, 5 of which also included CBD as an ingredient. One of those 5 with CBD had no detectable melatonin at all (!), and 22 of the 25 total products were off on the amount of melatonin by more than 10%, with a range from 74% to 347% of the labeled dose.

That sounds pretty terrible... I will say, though, that only 3 of the 25 were off by more than 3mg, and this sort of modest variation is unlikely to have the sort of major clinical effects, like, say, putting 10mg of the blood pressure medication, amlodipine, into a 2.5mg tablet might — that would be a problem!

However, this misdosing would start to concern me if small children were taking melatonin (and they are). Even then, a study from Japan exploring a possible role in seizure prevention for high-dose melatonin in 12 children ages 2-19 gave doses in the 10-40mg range without any lab abnormalities or reported side effects (beyond daytime sleepiness in 3 children). In this context, last year’s CDC MMWR on the subject of pediatric melatonin overdoses struck me as surprisingly strident; they found that among the nearly 28,000 children who sought care at a health care facility from 2012-2021 for melatonin exposure, 1% required ICU admission. I suppose that’s possible with children taking down entire containers of melatonin gummies like candy; but I also wonder if other circumstances and/or medications were involved (the study did not corroborate hospital records with poison control reports).

In any case, I think there is a simple solution to this concern: don’t buy supplements in gummy form. They are not good for your teeth. We eat too much candy as it is. Just don’t buy gummies — especially if small children live in your house.

The real question worth answering is: does melatonin have any value as a sleep aid?

Biological plausibility is on its side. Melatonin is a hormone secreted by our pineal glands (think, “third eye” in eastern spiritual traditions) when not blocked by the recognition of blue light wavelengths by our visual system. As the “hormone of darkness,” melatonin retains an appropriate aura of mystery; many species without pineal glands secrete melatonin, and melatonin is presumed to have properties beyond maintenance of circadian rhythms, leading to its study for immune modulating, anti-epileptic and antioxidant qualities, among others. However, no one denies that melatonin plays a key role in initiating the sleep process in humans; and in our brightly-lit, iPhone-dominated society, it would be no surprise that the timely release of melatonin has been thrown off for many.

Literally hundreds of studies have been run on melatonin’s efficacy as a sleep aid, and the results have been mixed. Perhaps this is no surprise given that someone with minimal pre-bedtime exposures to blue light and plentiful physiologic melatonin secretion might not have need of more, while a hospital worker fresh from the 7 PM to 3AM shift under bright lights could see substantial benefit. In any case, meta-analyses of both pediatric and mixed adult and pediatric studies have been modestly positive.

In children and adolescents, subjects averaged falling asleep about 20 minutes earlier (“sleep latency”), and slept about that much longer, as well. In the meta-analysis of all trials (predominantly adults), there was a statistically significant improvement in sleep latency of about 7 minutes, and overall sleep of roughly 8 minutes. Are these modest improvements clinically significant? That’s a stretch, especially for the adult trials. However, buried within that noise some people likely experience real benefit — probably people with low levels of endogenous melatonin. howeer – possibly due to detectable melatonin being a poor surrogate for biologically active melatonin – this has not been shown to be the case when studied.

I would argue that there is little risk to a trial of melatonin for most adults, especially since there is no controversy that poor sleep is bad for our health, and pharmaceutical options to treat insomnia are thought to impact sleep quality if not outright increase mortality (with the possible exception of $100/mo ramelteon/Rozerem, which, of course, works by binding melatonin receptors and has not been tested against melatonin in human trials).

The optimal dose of melatonin is a matter of some controversy. Some subscribe to taking a “physiologic” dose, or something roughly akin to what the body makes on its own. This is estimated to be in the 0.3mg ballpark — not an easy dose to find in the realm of melatonin supplements, which often start at 3mg. On the other extreme are the ScienceBros (my usual go-to source for medical information) advocating for melatonin mega-doses, in the 20-100mg range.

Personally, I am agnostic about dosing. I was taught many years ago that the risk of taking more than 3 mg of melatonin is that, like most endocrine glands, the pineal might start reducing its own production of its signature hormone over time if enough is being taken in supplement form, and/or melatonin receptors could be downregulated, making supplementers insensitive to their own melatonin upon stopping. I could not find any actual evidence of this phenomenon, however, and a case reports exist like this tiny (2 subject) study in which over 3 months of 5mg melatonin did not lead to a disrupted Circadian rhythm with abrupt cessation of supplemental melatonin.

It strikes me as plausible that high doses, which would take longer to metabolize, might serve to keep dropping someone prone to frequent waking back into their sleep program. Or perhaps at high enough doses melatonin truly has a sedative property that some have proposed, but even 100mg of IV melatonin did not appear to induce sedation in healthy adult subjects, so I am dubious of this claim — and therefore of modest irregularities in supplement doses posing a health risk to the general public.

While the theoretical potential for melatonin supplementation in kids to affect neural and sexual development prompts me to give the usual disclaimer — “discuss with your doctor before trying!” — I think it’s okay for adults to experiment with melatonin for sleep if they struggle with insomnia, and not just of the shift work or jet lag variety. I would start with low doses in the 1-3mg range, but not be afraid to move up the scale if high doses are helpful, just with the caveat that there is some theoretical risk of a rough return to life without supplemental melatonin. Timing of the dose is also worth adjusting, as some recommend taking melatonin 2-3 hours before sleep (more closely approximating the natural production of melatonin after sunset and well before bedtime at most times of year), while others suggest only 30 minutes before bedtime. Given that it takes at least that long to reach its peak, taking melatonin right at bedtime like true sedative/hypnotics is not generally recommended.

Few of us have access to a sleep lab to assess our nightly sleep quality, or draw bloodwork every morning to assess inflammatory and immune health biomarkers. The alternative? Keep a sleep diary before, during, and after trialing different melatonin doses. How long did it take to fall asleep? How many hours did you sleep (these numbers can be measured with various smart-watches and biosensors these days, of course, with variable accuracy)? Most importantly, did you feel refreshed upon waking? Did your energy hold up through the day? It’s hard for me to believe that there is an untoward effect on health from taking a supplement like melatonin if sleep hours improve and one feels better for it.

For that reason, I think melatonin is a perfectly reasonable first choice for adults dealing with insomnia. As with most sleep medications, I would keep expectations low but hopes high, since placebo effect is our best friend in sleep medicine. Just don’t buy the gummies, okay?