Does melatonin work for insomnia? What the guidelines actually say

For chronic insomnia, the evidence does not support melatonin, and sleep medicine's own guideline recommends against it. That surprises most people, because it is the supplement the whole category is built on. The reason is not that melatonin does nothing. It is that melatonin does something other than what people are buying it for.

What the guidelines say

In 2017 the American Academy of Sleep Medicine published a clinical practice guideline on medications for chronic insomnia in adults. It reviewed each drug and supplement in turn and issued a recommendation for or against.

Melatonin got a recommendation against — for both sleep-onset insomnia and sleep-maintenance insomnia. It was a weak recommendation, which in guideline language means the evidence was low quality and the balance of benefit and burden was unfavourable, not that the substance is dangerous.

The year before, the American College of Physicians had recommended that all adults with chronic insomnia receive CBT-I as the initial treatment, with medication considered only afterwards and only as a shared decision with a clinician.

So the two major guidelines in this area agree: the first thing to try is not a pill or a supplement. It is the behavioural approach.

What melatonin actually does

Melatonin is not a sedative. It is a signal.

Your pineal gland releases it as evening light fades, and the rise tells the rest of the body that night has started. It is the body's way of saying when, not how much. It marks the time; it does not push you under.

That distinction explains most of the confusion. Someone whose problem is that their body clock is running late — a teenager who cannot fall asleep before 2am and cannot wake before 11, or a traveller five time zones from home — has a timing problem, and a timing signal can help. Someone who is in bed at a reasonable hour, tired, and lying awake because their mind will not stop has a different problem entirely, and no amount of clock signal addresses it.

Where melatonin does have a role

• Jet lag. Shifting the clock deliberately after crossing time zones is what it is good at.

• Delayed sleep-wake phase disorder. A body clock genuinely running hours late, most commonly in adolescents and young adults.

• Some shift-work schedules, under guidance, where the aim is to move the clock rather than to sedate.

In each of those, timing and dose matter more than most people realise — taken at the wrong hour it can shift the clock the wrong way. That is a conversation with a clinician, not a guess.

The size of the effect

Pooled analyses of melatonin trials find it shortens the time taken to fall asleep by roughly seven minutes on average, and increases total sleep time by a similarly modest amount.

Seven minutes is a real effect. It is also not what somebody lying awake for ninety minutes is looking for. And for comparison, the meta-analysis behind the CBT-I guidelines found the behavioural approach producing changes several times that size in the same measures — with the difference that those improvements were still there a year later, because what changed was behaviour rather than chemistry.

The bottle problem

There is a second issue that has nothing to do with the biology.

In the United States melatonin is sold as a dietary supplement, not a medicine, which means the manufacturing is not held to pharmaceutical standards. An analysis of melatonin supplements sold across North America found actual melatonin content varying enormously from what the label claimed — some products containing a small fraction of the stated dose, others several times it, with variation between lots of the same brand. A number of samples also contained serotonin, which has no business being in a sleep supplement bought off a shelf.

So the honest position is that if you take a 5mg melatonin tablet, you do not reliably know what you took. That uncertainty makes it very hard to know whether it helped, which in turn makes it easy to keep taking something that is not doing much.

Melatonin and CBT-I side by side

The row worth staring at is the last-but-two. CBT-I is slower and, in the first fortnight, actively unpleasant. What you get for that is change that stays when you stop doing it, because nothing is being propped up.

If you are already taking it

Nothing here is a reason to stop something abruptly, and melatonin is not a drug people need to taper off in the way some prescription sleep medications are. But two things are worth doing.

Ask what problem you are treating. If you are waking at 3am and lying there, melatonin is aimed at the wrong thing, and continuing it is mostly a way of feeling like something is being done.

And raise it with your own doctor if you have been taking it nightly for months, are pregnant, are taking other medications, or are giving it to a child. Long-term safety in adults is not well established, and "available without a prescription" is not the same as "studied for years of nightly use".

When something else is going on

Some sleep problems are not insomnia, and no supplement or behavioural programme addresses them. See a clinician rather than experimenting if you snore loudly, wake gasping, or have been told you stop breathing; if your legs feel restless in the evening; if you fall asleep during the day in situations where you would not expect to; or if this began suddenly or alongside a new medication.

Somnera is a self-guided education programme built on CBT-I. It is not a diagnosis and does not replace care from your own clinician.

Citations

• Sateia MJ, et al. Clinical practice guideline for the pharmacologic treatment of chronic insomnia in adults: an American Academy of Sleep Medicine clinical practice guideline. J Clin Sleep Med. 2017;13(2):307–349. doi:10.5664/jcsm.6470

• Qaseem A, et al. Management of chronic insomnia disorder in adults: a clinical practice guideline from the American College of Physicians. Ann Intern Med. 2016;165(2):125–133. doi:10.7326/M15-2175

• Ferracioli-Oda E, Qawasmi A, Bloch MH. Meta-analysis: melatonin for the treatment of primary sleep disorders. PLoS One. 2013;8(5):e63773. doi:10.1371/journal.pone.0063773

• 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:10.5664/jcsm.6462

• Trauer JM, et al. Cognitive behavioral therapy for chronic insomnia: a systematic review and meta-analysis. Ann Intern Med. 2015;163(3):191–204. doi:10.7326/M14-2841