Do Weighted Blankets Help You Sleep?

Last checked 31 October 2026 · 14 min read

Yes — for a minority of the people who buy one, and not for the reason printed on the box. The best trial is unusually strong evidence that a heavy blanket changes how a night feels, and unusually weak evidence that it changes how a night is. In it, 120 patients with clinical insomnia slept four weeks under an 8 kg chain blanket; their insomnia scores fell from 21.7 to 9.2 out of 28. Their wrist monitors recorded no significant change in total sleep time or in time spent awake after falling asleep. Both of those results are real, and the gap between them is the entire subject: a weighted blanket restores one signal — steady pressure — that most people stop noticing within minutes of getting under it.

A large effect, and a smaller one

The study is the one every review cites, and it deserves the attention. One hundred and twenty patients with major depression, bipolar disorder, generalised anxiety or ADHD — all with insomnia severe enough to be treated — were randomly assigned to a heavy chain blanket or a light one for four weeks. The main measure was a questionnaire, the Insomnia Severity Index: seven questions, scored out of 28, where anything above 14 is clinical insomnia and 7 or below is remission.

The heavy-blanket group went from 21.7 to 9.2. The light-blanket group went from 21.2 to 18.8. Fifty-nine per cent of the heavy-blanket group improved by at least half, against five per cent of the control group. The reported effect size was 1.90, which is the sort of number you see in trials of drugs, not bedding, and it is why this study gets quoted in every article on the subject.

Read the first two rows together and the study becomes more interesting than its summary. The blanket changed the rating by twelve points and the recording by nothing measurable. That is not a contradiction to be explained away — it is the finding. A night can be experienced as much better without being longer, and for the question most people are actually asking — why does this feel so bad — the experience is not the lesser measure. It is the one that decides whether you get up feeling human.

But note the qualifier in the third row, because it is where the honest version of this page lives. In the whole group, the monitor saw nothing. In the subset who responded — and only about six in ten did — both the monitor and the rating saw better sleep maintenance. So the blanket does not lift everyone a little. It does something real to a minority and nothing to the rest, and the average is a blend of those two facts. Which is also why your own experience may have been exactly nothing while somebody else's was transformative, with no dishonesty on either side.

What the control group was sleeping under

Almost every summary of this study says it compared weighted blankets with ordinary blankets. The detail matters more than the summary: the control blanket was a plastic chain blanket, sewn to the same shape and size, weighing 1.5 kg. The researchers had weighed ordinary blankets sold in a Stockholm department store before designing it, and found they ran from 0.55 to 2.4 kg, with an average of 1.3 kg. In other words the comparison was not blanket against no blanket. It was 8 kg against the duvet you already own.

That changes the advice. If you sleep under a heavy winter duvet, or two layers, you may already be running a version of the intervention without having bought anything — and the trial's effect came from adding roughly six kilos on top of normal bedding, not from adding bedding to a bare bed. It also explains a result that otherwise looks strange: the control group improved for two weeks (21.2 to 15.5) and then drifted back to 18.8. A new blanket of any kind helps for a fortnight. What separated the groups was what happened in weeks three and four, when novelty had worn off and only the weight was left.

Heavier is not better, and the ten-per-cent rule is not from a trial

Nearly every buying guide repeats a rule: choose a blanket about ten per cent of your body weight. That rule does not come from the trials. In the study above, everybody started at 8 kg regardless of size, and ten of the 120 asked to swap down to 6 kg because 8 was too heavy for them. In the newer crossover work, the weights were fixed at 7, 9 and 14 kg for every participant. Nobody in either study was dosed by body weight.

The newer work is also the best test so far of whether more weight does more. Twenty healthy young adults napped for thirty minutes under four conditions — a sham blanket, 7 kg, 9 kg and 14 kg — with a week between each, measured by portable EEG. There were no significant differences in any sleep measure between any of the conditions. A second group of twenty slept overnight at home under the same four conditions. One statistical model suggested better sustained sleep efficiency at 9 and 14 kg than at the control; a sensitivity analysis of the same data, and a non-parametric test, did not reproduce it, and only seven participants had complete recordings. The authors' own conclusion is that the finding is exploratory and model-dependent.

Two other trials point the same way. Fifty patients sleeping in intensive care after major abdominal surgery were randomised to a weighted or a normal blanket; the sleep questionnaire came back 46 versus 56 out of 100, which is not a difference. Fourteen residents of a dementia unit slept two weeks under a weighted blanket and two weeks under usual bedding; there was no statistical evidence of better sleep, though fewer falls and fewer night wakings were recorded, and the wrist device used turned out not to measure sleep well enough to settle it.

What all of this adds up to is narrower than the marketing and more useful than the scepticism. The weight is not a dose. There is no evidence that going heavier helps more, some evidence that it helps less by making the thing too hot or too hard to move under, and a documented group of people — one in twelve in the largest trial — who ask for a lighter one.

What the pressure is standing in for

A bed at night is not one signal but four: sound, weight, heat and movement. They arrive together, they come from the same source, and they do the same job — they tell an old part of the brain that the room is occupied and nothing needs watching. When one of them disappears, the night changes shape, which is why an empty bed produces a three o'clock wake rather than a long lie-awake at eleven. A weighted blanket puts back exactly one of the four.

That is the whole case for it, and it is a good case, because pressure is the one you can buy. But it also predicts who it will not help, and the prediction is testable. If what your night is missing is sound — if the house is too quiet, if every small noise arrives as a separate event — then a fan will do more for you than eight kilos of glass beads, and no amount of weight will replace a signal you were not missing. People arrive at this distinction without being taught it: the ones who reach for a box fan and the ones who reach for a heavy blanket are usually describing two different holes in the same night.

There is a second clue in how the pressure behaves, and it is the most interesting thing in this whole subject. Pressure is a signal that disappears. Within minutes of getting under a heavy blanket you stop being able to feel it; the sensation returns when you move, and it returns most sharply when you take the blanket off. One long-term user put it exactly that way — that folded double it felt no different from a heavy quilt while he was asleep, and that it was only noticeable when he tried to remove it.

If that is right, then the blanket cannot be working by being felt all night, because it is not felt all night. Either it does its work in the ten minutes of settling, when the pressure is still news, or it works as a boundary — something you meet only when you turn over, which is precisely when a half-waking brain is deciding whether the environment is normal. Nobody has tested either version. It is worth holding both in mind, because they lead to different advice: the first says wear it to fall asleep and take it off later; the second says the weight only matters at the edges of the night, and a lighter blanket that stays put may beat a heavy one you fight.

Why it works the first week and fades in the second

Search for experiences and you will find a pattern the trial does not show. One person slept through the night four nights running and called it a saviour. Another said it helped for exactly one night, and that the help was placebo. A third lasted about a week, then tried a heavier one at 20 pounds and got the same nothing. A fourth said the first two nights felt strange and then the body adjusted — and went on sleeping well.

The trial, by contrast, shows the effect arriving fast and then holding: most of the improvement came in week one (21.7 to 13.2), the rest accumulated slowly, and at twelve months, with everyone on a weighted blanket, 92 per cent of the original participants were responders. So in a controlled study the effect grows, and in self-purchased use it often evaporates within days.

Three differences could explain that, and this page cannot tell you which is true. The trial's participants had clinical insomnia and a psychiatric diagnosis, so there was more room to improve. They were blinded — told they would receive one of two chain blankets, with no hint that weight was the variable — which removes the specific disappointment of a purchase that is not working. And they had a protocol and a follow-up call, which is a mild intervention in itself.

The practical version: a first night that is dramatically better is weak evidence, in either direction. A fourth week that is better than the first is strong evidence. If the effect is gone by night five and you have not changed anything else, the honest reading is that you got the fortnight that any new bedding buys, and the weight was not the active ingredient for you.

How it fails

The failures are consistent enough to list, and they are mostly physical rather than psychological.

Heat. This is the most common complaint by a distance. Glass beads hold coolness for the first hour and then radiate back what you have given them, so the blanket that felt wonderful at eleven can be unbearable at three. People who run hot describe exactly this curve. In the largest trial it barely appears — one participant stopped for anxiety, and no other side effects were reported — but that study ran in Sweden, in winter, in a clinical population, which is close to the best-case conditions for a heavy object.

Not being able to move. A side sleeper bruised a wrist and hurt her back in a single night under a 15-pound blanket, with a fitness tracker on the same arm. Another woke with sore ankles because the weight kept her toes pointed all night. Someone else described it as too hot most of the time and a pain to move, and stopped. These are not sensitive people being fussy; they are the predictable cost of adding weight to a body that turns over thirty times a night.

Claustrophobia. It is real, it is not rare, and it is the one failure the trial's authors name directly: one participant withdrew because of it, and they note that people who feel it usually manage with a lighter blanket. If you feel pinned rather than held, the experiment is over. Do not persist to prove a point.

Being the wrong size for it. A 120-pound person bought a 15-pound blanket, found it too heavy to be comfortable, and it has been in a closet for years. The weight that feels like being held at 80 kilos feels like being trapped at 55. Since no trial individualises the dose, the only calibration available is your own first ten minutes.

Anyone who cannot get out from under it. This is the one place where the answer is not "try it and see". For infants, for small children, and for anyone with dementia or limited mobility, a heavy blanket is a risk of overheating and of being unable to free themselves, and it should not be used at all. That is not a cautious footnote; it is the reason several manufacturers put an age floor on the label.

The one-week test that costs nothing

Before buying anything, run the experiment the trial ran, at home, with what you already own. The point is to find out whether the pressure channel is the one you are missing — not to find out whether a specific product is good.

If the folded duvet does something, buy the lightest blanket that produces the feeling and not the heaviest you can lift. If it does nothing, spend the money on the other three channels instead: a fan for sound, a cooler room for heat, and a mattress that does not move when you do. Those are cheaper, and two of them have better evidence behind them than the blanket does.

Where this fits

The four-signal frame this page leans on is set out properly in why you sleep worse when your partner is away, which is the involuntary version of the same experiment: one person leaves and three of the four signals go with them, which is why the fix people invent there is usually a pile of pillows on the empty side — pressure, bought cheaply. If the suspect is the surface rather than what is on top of you, whether a new mattress is the answer covers the expensive version of the same mistake. If the night breaks at a fixed hour rather than at the start, waking at three is the closer page. And if you were hoping a blanket would act like a sedative, it does not: that distinction, and why the clock matters more than the dose, is in why melatonin does not work.