Cannabinoids cut insomnia severity and added 34 minutes of sleep across 10 trials

Thirty-four minutes is the number people will remember, and it is not the most useful one in the paper.

Asim and colleagues, writing in Sleep Medicine, pooled 10 RCTs involving 2134 participants to ask whether cannabinoids do what a large and growing number of people already use them for. Total sleep time increased by 33.99 min, and insomnia severity fell by nearly four points on the standard scale.

The comparison was not against nothing. The trials compared cannabinoids with placebo or melatonin, which means part of this evidence is against an active sleep aid rather than a sugar pill.

Four outcomes, all moving together

Insomnia is a common sleep disorder that presents with difficulty falling asleep or staying asleep, and the four things this analysis measured map onto that description.

Cannabinoids significantly reduced insomnia severity and improved sleep quality, by 3.73 and 3.94 points. Sleep efficiency, the share of time in bed actually spent asleep, improved by 4.31%. Sleep onset latency was modestly reduced, by about twelve minutes between lights out and sleep.

Four outcomes pointing the same way is more persuasive than any one of them. A drug that only shortened the wait to fall asleep might be doing nothing but sedating; one that also lengthens total sleep and raises efficiency is changing the shape of the night.

Where 3.73 points sits on the scale

The Insomnia Severity Index runs from 0 to 28. Roughly, 8 to 14 is subthreshold insomnia, 15 to 21 is moderate clinical insomnia, and above that is severe.

A drop of 3.73 moves someone a useful distance down that range without necessarily moving them out of a category. The figure usually cited as a clinically important difference on this index is around 6 points, so this result is real, consistent, and smaller than what a patient would describe as their problem being solved.

That gap between statistically clear and clinically transformative is the honest summary of the whole paper. Something is happening. It is not a cure.

The side effect count nobody puts in the headline

Adverse events, mostly mild dizziness and dry mouth, were more frequent with cannabinoids, at a rate nearly four times higher than the comparators.

No serious harms were observed, and that qualification matters. Dizziness and a dry mouth are not reasons to avoid a treatment that works. But a four-fold increase is not a rounding error either, and in a chronic condition where people take something nightly for months, mild-but-constant is exactly the kind of side effect that ends adherence.

The range around that figure is also enormous, running from under two to nine. Ten trials produce imprecise safety estimates, and safety is where imprecision costs the most.

What was in the capsule, and what was not

The word cannabinoids covers a great deal of ground, and this is the limitation with the most practical bite.

Trials use defined products: a measured dose of cannabidiol, or a fixed ratio of cannabidiol to tetrahydrocannabinol, taken by mouth at a set time. What people buy varies in strength by an order of magnitude, arrives in forms with wildly different absorption, and is frequently smoked, which changes the pharmacology entirely.

Marijuana can be habit-forming, and its legal status differs by country and by state. None of the trials pooled here tested what somebody does with a vape at eleven at night, and the results should not be read as if they had.

What to do about a bad night, repeatedly

Insomnia is treatable and the first-line treatment is not a drug. Cognitive behavioral therapy for insomnia has better long-term evidence than any sleep medication, and unlike a nightly capsule it does not stop working when you stop taking it.

The case for cannabinoids sits after that, alongside the other pharmacological options, and it now has more evidence behind it than it did. The authors are appropriately unfinished about it, calling for larger, long-term trials to define optimal dosing and sustained efficacy. Tolerance is the standard failure mode for sleep drugs, and none of these trials ran long enough to find out whether it applies here.

If the ten-day fix becomes the ten-month habit, this analysis has nothing to say about what happens next. That is the question worth asking a doctor before starting, rather than after.