Back, neck and shoulder pain from desk work

It usually starts on a Thursday. An ache between the shoulder blades that was not there on Monday, a neck that clicks when you turn to a colleague, a lower back that complains when you stand up after the fourth meeting in a row.

And the advice is always the same. Sit up straight. Raise the screen. Change the chair. Get a standing desk. It sounds reasonable, it is easy to sell, and it rests on a story almost everyone believes: that the hours in the chair are breaking your back down piece by piece.

That story holds up worse than you would think. And the main reason it is worth abandoning is not that it is wrong. It is that it sends you to the furniture shop when you should be going to the floor.

The chair is not the villain in your lower back

In 2010 a Canadian research group did something that is rarely done: they tested the causal claim itself instead of repeating it. The review was published in The Spine Journal and looked specifically for evidence that occupational sitting causes low back pain.

The search returned 2,766 citations. Twenty-four studies met the criteria, five of them high quality: two case-control studies and three prospective cohorts. The result was not the hedged "we cannot say" that usually ends this kind of work.

They found strong, consistent evidence of no association between occupational sitting and low back pain. They found moderate evidence for the absence of any dose-response trend, meaning that more hours in the chair did not produce more trouble. Their conclusion was that occupational sitting is unlikely to be independently causative of low back pain in the working populations studied.

Read that carefully, because there are two ways to overstate it. It does not say that sitting is good for you. It does not say your back is not hurting. It says that the number of hours in the chair is the wrong variable to attack, which is a different thing entirely.

The neck and shoulders are a different case

Here the evidence diverges, and most pages online lump together what actually comes apart.

The NHS lists bad posture, giving sitting at a desk for a long time as its example, among the most common causes of neck pain. Sweden's national health guidance service, 1177, says much the same: a common reason for neck and shoulder pain is stiffness in the muscles after sitting for a long time or working in an awkward position.

The difference from the lower back is not arbitrary. Your lower back carries you whatever you do. The neck and shoulder girdle, by contrast, hold up a head and two arms in a position that does not change for hours, while the eyes are locked to a screen and the hands to a keyboard. That is static work, muscles holding tension without ever being allowed to let go.

1177 adds one more factor that rarely appears in ergonomics leaflets: the pain can be caused by stress. They also write that the risk of neck pain falls when you enjoy your work. That is an uncomfortable sentence to read on a training page, because it points at something no exercise in the world can fix. If most of your load sits in how the week feels rather than in how you sit, that is the problem to solve, and we take it up in the guide to stress management.

What actually prevents back pain

If the chair is not the lever, what is? That question has an unusually clear answer.

In 2016 JAMA Internal Medicine published a systematic review and meta-analysis of everything that had been tested to prevent low back pain. The material was 21 randomised trials with 30,850 participants between them. Six different strategies were set against each other.

The numbers are relative risks. A value of 1.00 means unchanged risk. Exercise combined with education landed at 0.55, roughly halving the risk of an episode of back pain. Exercise on its own landed at 0.65.

Then look at the bottom three rows, because that is where the real news sits. Education alone: 1.03. Back belts: 1.01. Shoe insoles: 1.01. Three measures all sold as back care, and three results that are unchanged risk in practice.

The pattern is hard to miss. What worked was the body doing work. What did not work was somebody telling it something, strapping a belt around it or putting something underneath it.

The limits of what training does

A page that stopped here would be an advert. There is a study that pulls the other way, and it is worth knowing about.

In 2021 the Journal of Orthopaedic & Sports Physical Therapy published a randomised trial under the heading that exercise is medicine, but perhaps not for preventing low back pain. The participants had recently recovered from an episode of back pain. They received either 12 weeks of exercise and education with supervised sessions, or an education booklet. After a year, 63 percent of the exercise group had a recurrence against 57 percent of the controls. The hazard ratio was 1.09 with a confidence interval from 0.7 to 1.8, meaning no statistically significant difference.

Two things stop it from overturning the picture. The trial planned for 160 participants but recruited 111, which leaves it too small to reliably detect a moderate effect. And more importantly, it asks a different question: stopping a relapse in someone who has just been in pain is not the same as preventing trouble broadly across a working population.

So the honest summary is this. Exercise is the best lever we have, and not by a narrow margin. It is not a guarantee, and anyone promising you that is selling something.

The dose: one hour a week

Now for the most useful number of all, because it was tested on exactly the group this page is about.

Danish researchers took 447 office workers with and without neck and shoulder trouble. They were randomised at cluster level into four groups for 20 weeks. Three of the groups received exactly one hour of supervised strength training for the neck and shoulders per week, distributed in different ways. The fourth group did no training at all.

The figures cover the 256 participants who had a pain rating of at least 3 at baseline, where the starting value in the neck was 3.2. All three training groups reduced neck pain compared with the group that did not train, and the reductions were judged clinically significant.

The interesting part is the last column. Between the three schedules there was no statistically significant difference in how much neck pain fell. But the measure of function in the arms, shoulders and hands improved only in the two groups that trained in longer sessions. Nine short bursts a week also lowered the pain, but they did not improve function.

In practice that means two things. The hour is the dose, not the number of occasions. Put it where your week has room. But do not grind it into the smallest possible pieces just because the calendar looks easier that way.

The four movements a desk leaves unloaded

A full-body programme with six movement patterns and a ready-made eight-week schedule is in the guide to strength training at home. This is something narrower: the four movements a working day at a desk never asks for, which is exactly why they are missing.

Think about what an office day looks like to the body. Everything happens in front of you. The arms are extended towards a keyboard, never pulling towards the body. The shoulders carry without lifting. The hip is bent at ninety degrees for eight hours without ever being extended. The muscles of the back resist nothing at all, because the backrest is doing it for them.

1. Table row

The most important of the four, because the pulling pattern is the one entirely absent from a working day. Lie on your back under a sturdy table, grip the edge with your hands roughly shoulder-width apart, brace your body into a straight line from shoulder to knee and pull your chest up towards the edge. Lower slowly. Bent knees with your feet flat on the floor make it easier, straight legs make it harder. Check that the table will hold before you hang from it.

2. Lateral raise in the plane of the shoulder blade

One of the five exercises in the Danish trial. Stand on an elastic band, or hold two water bottles. Lift straight arms out to the side and about thirty degrees forward, so not straight out sideways but diagonally in front of the body, up to shoulder height. Lower slowly. In the trial the movement was done with a dumbbell.

3. Bird dog

Get on all fours with your hands under your shoulders and your knees under your hips. Extend your right arm forward while your left leg goes back, into a straight line. Hold for three seconds without letting the hip rotate, then switch sides. Swedish health guidance lists exercises that train the back and abdominal muscles as useful for back trouble, as long as they do not make the pain worse.

4. Hip bridge

Lie on your back with bent knees and your feet on the floor. Press through your heels and lift your hips until your body forms a straight line from knee to shoulder. Hold for a moment at the top, then lower under control. This is the only one of the four that extends the hip, the exact opposite of what the chair has been doing to it all day.

The whole circuit takes around fifteen minutes with rest between sets. Three to four rounds a week add up to roughly the hour the Danish trial used. If you only manage two, you are still meeting the adult recommendation of muscle-strengthening activity on at least two days a week. The load should be heavy enough that the last repetitions are hard while your form holds all the way through.

Breaks: what they do and what they do not

Both the NHS and 1177 tell you to keep moving rather than to rest up. The NHS advises against a neck collar precisely because it is better to keep the neck moving, and its back pain advice is to stay active and not to stay in bed for long periods. 1177 recommends moving at regular intervals if you have sedentary work, and taking breaks so the muscles get a proper chance to relax. That is good advice. Just be clear about what kind of claim it is.

Breaks make the day more bearable. They interrupt the static work in the neck and shoulders before it has time to set. What they do not have behind them is the same kind of evidence that strength training has in the table further up. Take the breaks, but do not let them replace the hour.

The same logic applies to stretching. It is not the treatment, whatever it feels like. What stretching actually gives you is in the guide to stretching before or after a workout, along with what it demonstrably does not.

If the problem is that the quarter of an hour never fits anywhere, that is a scheduling problem rather than a training problem. There are solutions for it in the guide to what to do when you have no time to exercise.

If you are already in pain

The most common instinct is to protect whatever hurts. That instinct is usually wrong here.

1177 writes that physical activity reduces the risk of back and neck pain, so it is good to move as usual even when your neck and shoulders hurt. They also write that muscle strength can decline if you rest too much, that the muscles then work less well than usual and that balance can suffer too. For back trouble, they note that it takes longer to recover if you lie down and rest more than usual. The NHS gives the same instruction for back pain in its own words: stay active and try to continue with your daily activities.

In practice that means lowering the load rather than removing it. Do the table row with bent knees, halve the number of sets, drop the exercise that hurts and keep the other three. There is a difference between discomfort that fades once you get going and pain that grows the more you do. The second one is a stop sign.

What this means on a Thursday

The summary fits in four lines.

The hours in the chair are not what is breaking your lower back down, so stop spending your energy there. The neck and shoulders are affected by the static position, but in their case stress is as real a factor as posture. What has the strongest support in both directions is loading the body rather than equipping it. And the dose tested on people with exactly your job is one hour a week.

That is four exercises, fifteen minutes, three to four times a week. A missed session is data, not a failure. What decides how your back feels in five years is how many times you came back, not how perfect any single week looked.

Frequently asked questions

Can a desk job cause back pain?

Not in the way most people think. A systematic review in The Spine Journal in 2010 screened 2,766 citations, included 24 studies and found strong, consistent evidence of no association between occupational sitting and low back pain. It also found moderate evidence that there is no dose-response trend, meaning more hours in the chair did not mean more trouble. The authors concluded that occupational sitting is unlikely to be independently causative of low back pain in the working populations studied. That does not mean your back is not hurting. It means the number of hours in the chair is the wrong variable to attack.

What exercises help neck and shoulder pain?

Loaded strength work for the neck and shoulders has the strongest support. In a Danish trial of 447 office workers, one hour of specific strength training per week for 20 weeks produced a clinically significant reduction in neck pain among those who had pain at baseline. One of the five dumbbell exercises in that trial was the lateral raise, lifting the arms out to the side to shoulder height. At home you can do the same movement with an elastic band or a pair of water bottles. The NHS makes the same point from the other direction for the neck: it advises against a neck collar because it is better to keep the neck moving.

How often should I train to avoid the aching?

One hour a week is the dose that has actually been tested on office workers. In the Danish trial, 447 participants were split into one 60-minute session a week, three 20-minute sessions, or nine 7-minute sessions. All three groups reduced neck pain compared with the group that did not train, with no statistically significant difference between the schedules. But the measure of arm, shoulder and hand function improved only in the two groups that trained in longer sessions. The hour is the dose. Distribute it to fit your week, but do not grind it into the smallest possible crumbs.

Will a standing desk or an ergonomic chair help?

We cannot tell you from the research gathered here, because neither standing desks nor chairs were among the strategies tested. What we do know is the pattern that keeps repeating. In the 2016 meta-analysis in JAMA Internal Medicine, exercise produced a clear reduction in risk, while back belts landed at 1.01, shoe insoles at 1.01 and education on its own at 1.03. In other words, no effect at all from the three passive measures. Standing for part of the day may well feel better, and there is no reason to avoid it. Just do not expect the furniture to do the job that loading the body is supposed to do.

Can I train the pain away completely?

Exercise is the best lever we have, but it is not a guarantee. In the 2016 meta-analysis, exercise combined with education lowered the risk of an episode of low back pain to 0.55, roughly halving it. At the same time, a 2021 randomised trial of 111 people who had just recovered from an episode of back pain found that 12 weeks of exercise and education did not reduce the risk of recurrence compared with an education booklet. 63 percent of the exercise group had a recurrence against 57 percent of the controls. Preventing trouble broadly across a working population is not the same question as stopping a relapse in someone who has just been in pain.