Sleeping position and shoulder pain is the thinnest-evidence subject in this cluster, and this page is going to say so before it says anything else. Three papers make up close to the whole literature. One of them is explicitly a hypothesis rather than a study. One is a cross-sectional look at 83 patients. The third is a review whose finding was that the field is small and of low quality.
That is not a preamble to a stronger claim. It is the claim. Everything below is written at the strength the evidence supports, which is low, and the practical suggestions at the end are labelled as comfort measures with nothing behind them rather than as advice.
The short answer
- The whole evidence base is three papers, and one of them is a hypothesis article.
- Nobody has established which way the causation runs between how you lie and whether your shoulder hurts.
- Night pain and pain on lying on the shoulder are common complaints in shoulder problems.
- They are also on the referral list. Pain that wakes you consistently is a reason to be examined.
- Grade: thin, and that is the finding.
Sleeping position and shoulder pain
The claim in circulation is straightforward: lying on one shoulder for hours compresses it, and doing that most nights over years produces or worsens shoulder pain. It has an obvious mechanism, it matches a lot of people's experience, and it sells pillows.
It is also close to untested. There is no trial in which people were assigned to sleep one way or another and followed for shoulder pain. There is no prospective study in which sleeping side was recorded first and shoulder pain arrived later. What exists is a proposal, a snapshot and a review of a small field, and the honest reader should know which is which.
The whole evidence base
Named, dated, and graded individually rather than blended into a paragraph that sounds more solid than any of them.
| Paper | What it is | What it supports |
|---|---|---|
| Zenian, Medical Hypotheses, 2010 | A proposed mechanism | Nothing on its own |
| Kempf & Kongsted, JMPT, 2012 | 83 patients, cross-sectional | An association at one moment |
| Cary, Briffa & McKenna, BMJ Open, 2019 | Scoping review | That the field is small and weak |
Grade: thin, and that is the finding.
Why a hypothesis is not evidence
Zenian's 2010 paper set out the idea that sleep position contributes to shoulder pain. It is cited constantly, and it is worth knowing what it is.
Medical Hypotheses is a journal that exists to publish ideas which have not yet been tested. That is its stated purpose, not a criticism of it — a place to put a plausible mechanism into print so that somebody can go and test it is a useful thing for a field to have. The journal's own name tells you what a paper in it is.
So a citation to it establishes that somebody proposed this, in print, in 2010. It does not establish that anybody has since demonstrated it. A proposal is where an investigation starts, and in this case the investigation has barely happened.
Grade: absent. A hypothesis paper is by design not evidence, and no amount of citation converts it into any.
Which way does the arrow point
Kempf and Kongsted, in the Journal of Manipulative and Physiological Therapeutics in 2012, looked at 83 patients with one-sided shoulder pain and compared the painful side against the side they preferred to sleep on.
Eighty-three people is a small study, and the design is the more important limit. It is cross-sectional: sleeping side and shoulder pain were recorded at the same moment. From a snapshot, "lying on that shoulder made it hurt" and "that shoulder hurts, so I have stopped lying on it" look identical in the data. People with a painful shoulder change how they sleep — that is one of the commonest things they report doing — so the reverse explanation is not a technicality but the obvious rival.
Cary, Briffa and McKenna took the broader question in BMJ Open in 2019, scoping the literature on sleep posture and spinal symptoms. Their finding was about the field rather than the body: the studies are few and of low quality, and the measurement of sleep posture itself is inconsistent between them. It is difficult to establish what happens to a body across seven unconscious hours, and the research reflects that difficulty.
Grade: weak for the association, absent for the direction.
Night pain is a referral matter
Here is the part of this subject that is well established, and it changes the shape of the advice.
Night pain, and pain on lying on the affected shoulder, are among the commonest complaints people with shoulder problems describe. That is a clinical observation with wide agreement behind it, and it is why sleeping position feels so central to anyone in the middle of one.
But it cuts the other way from how it is usually used. Because night pain that wakes you is also on the referral list. Pain that persists, wakes you at night, radiates into a limb, comes with numbness or weakness, or followed a fall or a dislocation is a matter for a clinician who can examine you. So is an arm you cannot lift at all, sudden weakness, or shoulder pain arriving with fever or a general feeling of being unwell. Nothing here diagnoses anything and nothing here is treatment.
Consistent night pain is a reason to be examined, not a reason to buy a different pillow. That is the single most useful sentence on this page, and it is the opposite of what most pages on this subject conclude.
Comfort measures, unsupported
If you want to try something anyway, here are the usual suggestions, offered as comfort measures with no evidence behind them and labelled that way rather than dressed up.
- A pillow under the top arm when side-lying, so the upper arm is supported rather than hanging across the body.
- Not sleeping with the arm overhead, which some people find leaves the shoulder aching in the morning.
- Changing sides, if you have lain on the same one for years.
None of those has been tested. None is a treatment for anything. If one of them makes your night more comfortable, that is a fact about your night and a perfectly good reason to keep doing it — it is not a finding, and it does not mean the mechanism behind it is real.
What would be dishonest is to present that list with the confidence of a protocol, which is how it is normally presented. The suggestions cost nothing and risk nothing, and that is the whole of their case.
Morning stiffness
A shoulder that feels stiff on waking and loosens within a few minutes of moving is ordinary. Joints are stiffer after hours of stillness, everywhere in the body, and the stiffness easing as you move is the expected pattern rather than a sign of anything.
If you want something gentle for the first minutes of the day, two moves from Limber's catalogue suit it: M034 Cross-Body Shoulder Hold, 30 seconds per side, drawing the arm across the chest; and M036 Standing Shoulder CARs, 40 seconds per side, taking the joint slowly around its own outline. Together they take under three minutes.
That is movement, not treatment. It is offered because moving a stiff joint in the morning is pleasant and low risk, and for no stronger reason than that. Stiffness that lasts well over half an hour, or that comes with swelling, heat or fever, is a different matter and belongs with a clinician.
What we would rather measure
We publish what can be measured, and how you slept is not on that list.
The shoulder mobility test reads the Apley scratch reach on each arm in five bands — 10, 30, 55, 75, 95 — which tells you what the chain can do today. Its own limit is worth knowing: the smallest gap it can express between your arms is 20 points, and Limber calls 12 points an imbalance, so any difference at all between your sides is flagged, and a one-band gap means measure again rather than act. The Range Score folds whatever you have read into one number and leaves the unmeasured ground blank rather than filling it in.
None of that says anything about your night. A reach reading is a reading of range, taken awake, in a known position — and the honest gap between that and seven unconscious hours is exactly the gap this whole page is about.
The joint itself is covered in rotator cuff mobility, the pain literature in shoulder impingement and mobility work, the top of the same chain in neck stretches, and the region as a whole in shoulders and upper back.
Questions
Does sleeping on your side cause shoulder pain?
Nobody has shown that it does. The literature is a hypothesis paper from 2010, a cross-sectional study of 83 patients from 2012, and a 2019 scoping review that found the whole field small and low quality. An association at a single moment cannot establish direction, and a painful shoulder is a strong reason to change how you lie.
What is the best sleeping position for shoulders?
There is no evidence-based answer, and any page that gives you one with confidence is going beyond what has been studied. Some people find a pillow supporting the top arm more comfortable when side-lying, and some find sleeping with the arm overhead leaves the shoulder aching. Those are comfort measures rather than findings.
Why does my shoulder hurt more at night?
Night pain is one of the commonest complaints in shoulder problems, which is well observed even though the reason is not settled. It is also on the list of reasons to be examined rather than to read further. Pain that wakes you consistently, or that comes with weakness or numbness down the arm, needs a clinician who can examine you.
Will a different pillow help?
It may make you more comfortable, and no study establishes anything beyond that. Buying one is cheap and harmless, so there is no reason not to try it — but if the pain wakes you regularly, replacing the pillow is not the response the situation calls for, and the delay is the real cost.
Is morning shoulder stiffness normal?
Stiffness on waking that eases within a few minutes of moving is ordinary and happens at joints all over the body. A gentle cross-body hold and a slow circle at the joint are pleasant ways to spend the first minutes, and are offered on that basis rather than as treatment. Stiffness lasting well over half an hour, or with swelling, heat or fever, is a different matter.
Should I stop sleeping on the painful side?
Most people do it without being told, which is part of why the research on this is so hard to read. If lying on that side hurts, avoiding it is a reasonable thing to do for comfort, and it is not a correction of anything. What matters more is that a shoulder painful enough to disturb your sleep gets looked at by someone who can examine it.