Melatonin

Melatonin is one of the most misunderstood things sold in a drugstore. People take it the way they’d take Benadryl, like a sedative, like a thing that knocks you out. That isn’t what it does. It’s a circadian signal. Your pineal gland makes it in tiny amounts starting in the early evening as the light dies down, and the signal it sends to the rest of your brain is roughly: it’s getting to be nighttime, start winding things down.

That’s the whole job. No sedation. No respiratory depression. No binding at GABA receptors. It tells the master clock in your hypothalamus what time it is. If your clock is already on time and you’re just tossing because of stress or caffeine or your kid woke up at 3 AM, melatonin won’t do much for you. If your clock is genuinely shifted, melatonin can be one of the most useful things in the cabinet.

The other piece nobody talks about is the dose on the bottle. The 5mg gummy at CVS, the 10mg “extra strength” at Costco, the 12mg sleep stack from whatever Instagram brand. Your pineal gland produces something on the order of 0.1 to 0.3mg at peak. Pharmacologic doses don’t make the signal louder in a useful way. They flood the receptors, cause downregulation, give you weird dreams, and leave you groggy at 9 AM. More is worse.

The four situations where it earns its keep

Four specific situations, in roughly the order of how well the data holds up.

Jet lag. You flew from New York to Paris and your body still thinks it’s 6 PM when the Parisians are going to bed. A small dose of melatonin a few hours before local bedtime for two or three nights helps drag your circadian rhythm forward. This is the use case where the evidence is cleanest and the effect is easiest to feel.

Delayed sleep phase syndrome. The teenager or twenty-something whose body genuinely doesn’t want to sleep until 3 AM and doesn’t want to wake until 11. This is an actual circadian phenotype, separate from laziness or screens. Small-dose melatonin given several hours before the target bedtime, paired with morning bright light, can shift the whole rhythm earlier over a few weeks. It’s slow, and you have to be consistent, but it works.

Shift work, occasionally, in patients who are stuck on rotating schedules and need to convince their brain that 10 AM is bedtime. It’s never a clean fix because the rest of the world is loud at 10 AM, but it can take the edge off.

REM sleep behavior disorder. This one sits in its own category. RBD is the disorder where people physically act out their dreams, punching, kicking, falling out of bed. It tends to show up in older men and it’s a meaningful neurological finding because it predicts Parkinson’s disease and Lewy body dementia downstream. Melatonin at moderate doses, 3 to 6mg at bedtime, reduces the muscle activity during REM and lets the bed partner sleep without getting punched. Clonazepam used to be first-line for RBD. Melatonin has quietly displaced it in a lot of sleep clinics because the side effect profile is better.

The dose on the bottle is wrong, the timing on the label is wrong, and the use case in most people’s heads is wrong. Three for three.

Where the marketing falls apart

Chronic insomnia in middle-aged adults. The classic patient is the 50-year-old executive who can’t fall asleep because her brain is still finishing tomorrow’s meeting at midnight. Her circadian rhythm is fine. The problem is hyperarousal, and melatonin doesn’t touch hyperarousal. She’ll come back in three weeks frustrated that the bottle she spent $24 on at Whole Foods didn’t help. What helps her is CBT-I, maybe a low-dose trazodone, sleep restriction, taking caffeine out after noon.

Anxiety. There’s a tiny anxiolytic effect at the receptor level that some studies have chased, but in practice nobody should be using melatonin for anxiety. It doesn’t reach a useful dose for that purpose, and there are actual anxiolytics if anxiety is the problem.

Kids who won’t fall asleep. This is the use case where I have the strongest opinion and the smallest amount of grace. Parents are giving toddlers melatonin gummies because the pediatrician’s office mentioned it in passing or a friend on a group chat said it worked. Melatonin in kids isn’t well studied long-term. It’s an active hormone with effects on reproductive timing in animal studies. The kid who can’t fall asleep almost always has a behavioral or environmental reason: screens in the bedroom, no consistent bedtime routine, a 6 PM nap. Fix the environment before you medicate the child with a hormone we don’t fully understand the developmental effects of.

I had a mom in last spring with a 7-year-old who’d been on 5mg melatonin nightly for two years. The kid was waking at 4 AM every morning wired. We pulled the melatonin over ten days, tightened up the bedtime routine, killed the iPad at 7 PM, and within three weeks he was sleeping through the night with no supplement. I see some version of that story constantly.

Timing matters more than dose

Take melatonin at bedtime and you’re using it as a sedative, which it does poorly. Take it three to five hours before your desired bedtime and you’re using it as a phase shifter, which is what it actually does. For a delayed-phase patient trying to fall asleep at 11 PM instead of 2 AM, that means dosing around 6 or 7 PM, not at 10:45 while brushing teeth.

Dose of 0.3 to 1mg is what the chronobiology labs use. You can usually find 0.5mg or 1mg tablets if you look. The 3mg version is the smallest dose most US retailers carry, which is already higher than what produces the cleanest phase shift. The 5mg and 10mg gummies that dominate the shelves are pharmacologic doses being marketed as if more were better. They produce supraphysiologic blood levels that persist into the morning and leave you groggy.

The supplement industry mess

Melatonin is sold as a dietary supplement in the US, which means it doesn’t go through the FDA approval process and the dose on the bottle is whatever the manufacturer decided to print. A study from JAMA a few years back tested melatonin gummies sold to children and found actual melatonin content ranging from 74% to 347% of what the label claimed. One product had 0% melatonin and was just sugar. Another had CBD in it that wasn’t listed.

If you’re going to use it, get a brand with USP verification on the label, which means a third party has actually tested what’s in the bottle. That cuts out most of the worst offenders. And if your pharmacy can compound a 0.5mg tablet, that’s probably more useful than anything you’ll find on a shelf.

How I think about it in clinic

Melatonin is a niche tool. Useful for jet lag, useful for genuinely delayed-phase patients, useful for RBD, occasionally useful for shift work. For most of the people who walk into a psychiatrist’s office complaining about sleep, melatonin isn’t the answer, and a lot of them have already been taking it for months at the wrong dose at the wrong time and getting nothing from it except weird dreams. When I take it off the table I usually have to explain why for fifteen minutes, because the cultural assumption that melatonin is a sleep aid is deep and the supplement industry has spent two decades reinforcing it.

The patients who do well on it are the ones who use it the way the chronobiology literature uses it. Small dose. Right time of evening. Specific phase-shift goal. Paired with morning light if there’s a real circadian shift to chase. Used for a finite stretch, not indefinitely. That’s the actual playbook, and almost nobody arrives at the clinic already running it.