Occupational health · Spine

Remote-working postural syndrome: what the literature actually shows

Since 2020 the kitchen table has become a workstation, and one explanation has taken hold without going through the evidence: if your back hurts, it is your posture. The available data say otherwise, and they point to other levers, less spectacular but better supported.

What exactly are we talking about?

The first job is to take the phrase apart. “Postural syndrome” gathers under one causal label presentations that each have their own diagnosis, prognosis and treatment. This chapter puts them back in their place, and sets out what remote working actually changes about how often they occur.

A causal label, not a clinical entity

No international classification recognises a “postural syndrome”. ICD-11 codes what the patient presents with: cervical spine pain (ME84.0), thoracic spine pain (ME84.1), low back pain (ME84.2). The word “postural” adds a presumed cause, and that cause is exactly what the following chapters examine.

This distinction is not pedantry. A patient told they have a “postural syndrome” hears that they hurt because they hold themselves badly, and that the answer is to sit up straight. They draw two false conclusions from it: that their back is fragile, and that the repair lies in a correction rather than in training.

What each presentation covers

The three presentations a physiotherapist meets in the remote worker are dealt with in their own right elsewhere on this site, and this article does not redo them. Non-specific neck pain covers the examination, the reasoning and the rehabilitation of the cervical spine. Thoracic spine pain covers the thoracic side, including what has to be ruled out before concluding. Non-specific low back pain covers the lumbar spine. This article deals with neither their diagnosis nor their rehabilitation: it deals with the attribution to the workstation, and with what the literature says about it.

What is workstation ergonomics really worth?

This is the chapter that surprises most people, experienced clinicians included. The question has been asked properly, in a Cochrane review restricted to randomised trials, and the answer is mostly negative.

What the Cochrane review examined

Hoe and colleagues updated in 2018 a review published in 2012, searching eight databases up to 10 October 2018. Strict inclusion criteria: randomised trials of ergonomic interventions for preventing musculoskeletal disorders of the upper limb or neck in office workers, with a baseline prevalence below 25 %. Fifteen trials retained, 2,165 workers. Only one trial was judged at low risk of bias; the other fourteen at high risk, because of small samples and a risk of selection bias¹.

The results, intervention by intervention

How to read this result without distorting it

Two faulty readings are waiting. The first is to conclude that ergonomics is useless: that is not what the review says, since it mainly documents an absence of evidence across few and methodologically fragile trials, and explicitly calls for better-quality studies. The second is to ignore the result and carry on attributing to the workstation whatever cannot otherwise be explained.

The tenable position is in between, and it is usable in a session: a comfortable workstation is a legitimate comfort goal, it is not a treatment for pain, and presenting it as one invites a disappointment that costs adherence to everything else.

A useful point about the sit-stand desk, often presented as a solution: the Cochrane review devoted to reducing sitting time at work concludes that the evidence is of very low to low quality for its effect on sitting time itself⁴, and Hoe's review finds it has no effect on upper limb pain¹. The device changes a habit, which is not nothing; it does not treat pain.

Does sitting really cause low back pain?

The question has been addressed with the right tool: not “is there an association?”, but “does that association satisfy the criteria for causality?”. The answer is clear, and it is not new.

Appraisal against the Bradford Hill criteria

Roffey and colleagues conducted a systematic review devoted to that single question, appraising each Bradford Hill criterion separately. Of 2,766 references screened, 24 studies met the inclusion criteria and five were of high quality (two case-control studies, three prospective cohorts). The result is stated unambiguously by the authors: strong, consistent evidence of no association between occupational sitting and low back pain, moderate evidence for the absence of a dose-response, non-significant risk estimates on the temporality criterion. Conclusion: it is unlikely that occupational sitting is independently causative of low back pain in the populations studied².

The review of reviews, ten years later

Swain and colleagues took the question up a level: a review of reviews, five databases searched from January 1990 to June 2018, 4,285 publications screened and 41 systematic reviews included, appraised with AMSTAR for quality and against the Bradford Hill criteria for causality. The finding is instructive in its mechanics: positive and null associations coexist across reviews, the meta-analyses lean more readily towards an association, and the reviews restricted to prospective studies are the least able to conclude. The authors' conclusion: despite the abundance of reviews, no consensus exists on causality for physical exposure in low back pain; the association is documented, it provides no causal explanation³.

So what about “text neck”?

The idea that cervical flexion in front of a screen produces neck pain has become a cultural given. It has been tested twice by the same team, cross-sectionally and then longitudinally, with an objective measurement of the angle. Both times, the result is negative.

The cross-sectional measurement

Correia and colleagues studied 582 volunteers aged 18 to 65. The cervical flexion angle was measured with a cervical range of motion device (CROM), standing and then sitting, while the participant typed a message on their smartphone. After multiple logistic and linear regression, no link: neither with the prevalence of neck pain (OR 1.00; 0.98 to 1.02), nor with its frequency, nor with its maximum intensity, whether measured standing or sitting¹².

The prospective check

The legitimate objection to a cross-sectional study is the direction of the arrow. The same team therefore followed 457 volunteers without neck pain for twelve months, with the same objective CROM measurement at baseline. At one year, 396 participants were reviewed (87 %) and 40 new cases of neck pain had appeared (10 %).

Result: the cervical flexion angle did not increase the risk of neck pain, neither standing (OR 1.0; 0.97 to 1.04) nor sitting (OR 1.01; 0.98 to 1.04). Two factors did increase that risk: poor sleep quality (OR 1.76; 1.17 to 2.63) and an insufficient level of physical activity (OR 2.41; 1.03 to 5.65)⁵.

What did remote working actually change?

The previous chapters dismantle presumed causes. This one describes what actually changed, because the discomfort of remote workers is itself well documented.

Degraded equipment, and measured

The most cited survey covers 843 faculty, staff and managers at an American university sent to work from home. The figures describe an improvised set-up: 85 % always or often used a laptop, and 55 % of all respondents were working on its screen alone, without an external monitor. Fewer than 45 % of chairs had adjustable armrests. And more than 40 % reported moderate to severe discomfort: severe low back and mid-back pain, moderate discomfort of the eyes, neck and head, discomfort of the upper back and shoulders⁶.

A second survey, of 51 respondents at a Canadian university, measures the change rather than the state: 51 % reported a worsening of discomfort they already had, and 24 % entirely new discomfort since moving home⁷.

How to reconcile these figures with the previous chapters

There is no contradiction, and that is the point to hold. These surveys are cross-sectional and self-reported : they describe an association between a situation and discomfort, without establishing that the workstation is its cause. What remote working changed is not, in any case, limited to the furniture: commuting disappeared, so did social breaks, the day stretched out, the boundary between work and rest blurred, and the postural variety of an office (walking to the printer, changing room, walking to the canteen) was replaced by the stillness of a single room.

In other words, the variable that changed most is probably not posture, but the amount of movement. That is consistent with the prospective risk factor identified for neck pain: insufficient physical activity.

Movement and load: what is actually established?

After four chapters of demolition, this one builds. The interventions that work are known and quantified, and they have one thing in common: they increase capacity rather than correct a position.

Preventing neck pain

The most recent meta-analysis on preventing episodes of neck pain screened 4,703 references and retained 5 randomised trials, 1,722 participants, of whom 80 % were office workers. Participants were free of neck pain at baseline, and exercise was compared with no intervention, a placebo, an attention control or a minimal intervention. Result: moderate certainty evidence that exercise reduces the risk of a new episode of neck pain, OR 0.49 (95 % CI 0.31 to 0.76) in the short term, that is up to twelve months. The authors set out two limits: the result was not robust to sensitivity analysis on missing data, and the clinical significance of the effect remains uncertain. Two trials had methodological reservations, three a high risk of bias⁸.

Preventing low back pain

The benchmark remains the meta-analysis published in JAMA Internal Medicine, which retained 23 publications covering 21 randomised trials and 30,850 participants. The results are unusually instructive, because they set what works against what does not within the same analysis.

Two practical lessons follow. First, explaining is not enough : a relative risk of 1.03 for education alone is a null result, however good the explanation. Second, education added to exercise does better than exercise alone (0.55 against 0.65), which suggests it works as a vehicle for adherence rather than as a treatment.

Treating established neck pain

In the office worker already in pain, a meta-analysis of 8 randomised trials examined exercise outside the workplace context alone. Seven of the eight trials reported a significant fall in pain intensity on a visual analogue scale and five a fall in the Neck Disability Index after strengthening. The authors conclude that there is low certainty that strengthening the neck, shoulder and scapular muscles reduces pain and disability, while stressing that all eight trials were at high risk of bias¹⁰.

Dose: what really decides the result

This is the most useful study for organising a programme, and it is little known. One hundred and twenty-five female office workers followed a specific strengthening programme for 16 weeks, recording training volume in a logbook (sets × repetitions × load). Regression analysis shows a dose-response relationship: the total training volume was negatively correlated with the change in the pain index (β = −0.16; p = 0.045), and so was the volume per session as well (β = −0.20; p = 0.034).

The result that matters is this: attendance, by contrast, was not correlated with improvement (1.69 sessions a week on average). The authors recommend optimising the volume per session by ensuring a load at 10 to 15 repetition maximum and applying progression¹¹.

How do you run a remote worker's consultation?

The patient nearly always arrives with a hypothesis already formed, often the right question asked back to front. This chapter offers an approach that takes their complaint seriously without endorsing their explanation.

What the patient brings

The opening sentence is nearly always a variant of “I think it's my position”. Two replies are bad ones. Confirming it without reservation installs a belief the data do not support and leaves the patient dependent on a piece of furniture. Contradicting it head-on, in the first session, breaks the alliance before it has begun.

The wording that works acknowledges the observation and shifts the explanation: the pain is real and the context genuinely has changed; what changed is not only the position, it is the amount of movement, the sleep, the hours and the workload. That reopens the field of levers instead of narrowing it to a chair.

What to measure

• The amount of movement, above all: step count, structured weekly physical activity, the commuting that disappeared when remote working began.
• Sleep, since it is the documented prospective risk factor for neck pain. Duration, waking, bedtimes, screens.
• Workload and how it is organised : screen hours, back-to-back meetings, no break, spilling into the evenings.
• Capacity : grip strength, endurance of the cervical extensors, the ability to sustain effort. That is the determinant we can act on.
• Beliefs : what the patient thinks about their back, and what they avoid. Established kinesiophobia weighs more than screen height.

What to offer, and in what order

1. An exercise programme with an explicit dose. It is the best-supported intervention, and its dose matters more than its frequency. A load allowing 10 to 15 repetition maximum, progression, volume per session tracked.
2. General physical activity. The risk factor identified in the prospective cohort is insufficient activity, not posture. Replacing the commuting that disappeared is a concrete goal.
3. Sleep, treated as a goal of care and not as a remark at the end of the session.
4. Setting up the workstation, last and for what it is : comfort. An external monitor and a separate keyboard for someone working eight hours on a laptop are reasonable, without any promise of treatment attached.
5. Postural variety rather than the right posture. The instruction “change position often” is modest, and that is precisely why it does not betray the data.

A model programme, and what it aims at

No protocol is validated for the “remote worker”. What follows is built from the interventions whose effect has been measured, and each line says where it comes from.

A methodological remark is called for on this table: three of its six rows are built on risk factors identified in cohorts, and not on intervention trials targeting those factors. Does acting on sleep reduce neck pain? Nobody has shown that it does. What we know is that poor sleep precedes the pain in a cohort of 396 people followed for a year. That is reason enough to attend to it, it is not proof of efficacy, and the patient has a right to know the difference.

Frequently asked questions

Should the screen height still be adjusted?

Yes, for comfort, and no, to treat the pain. The Cochrane review found no effect of workstation adjustment on upper limb pain. A screen at eye level avoids eye strain and discomfort, which is enough to justify it without crediting it with therapeutic power.

Is the sit-stand desk useless?

It has no demonstrated effect on upper limb pain, and the evidence for its effect on sitting time itself is of very low to low quality. It changes a habit, which has value for postural variety; selling it as a treatment for low back pain is not supported by the data.

So posture does not matter at all?

That is not what the data say. They say that no particular posture has been shown to cause low back pain or neck pain, and that no postural correction has shown a therapeutic effect. Posture remains a variable of comfort and fatigue, and postural variety is probably more useful than the search for an ideal position.

And if my patient gets better after changing chair?

That happens, and it is not to be disputed. An individual improvement after a change of workstation does not establish causality at group level, and the reverse is also true: a negative meta-analysis does not forbid a given patient from benefiting from an adjustment. What we cannot do is promise that benefit.

Should you talk to the patient about “text neck”?

It is worth taking the term apart if they use it. Cervical flexion during smartphone use has been measured objectively in 582 and then 457 people, without any association or any prospective risk appearing. The conversation does, however, open naturally onto sleep and physical activity, which do stand out.

How long before an effect is seen?

The prevention trials measure their results at twelve months, and the dose-response study covers sixteen weeks of strengthening. The order of magnitude to give is therefore that of a quarter, not of a week.

And regular breaks?

The Cochrane review does find an effect of additional breaks on neck discomfort, of the order of 0.25 points, with very low qualityevidence, in data entry operators. It is little, but it is one of the few organisational interventions to stand out, and it costs nothing.

References

Sixteen references, resolved one by one through the PubMed E-utilities API. For each of them the abstract was read in full and the figures cited in the article checked at source. The year kept is that of the issue, re-read on the XML record: the first “year” field of a PubMed record is sometimes a revision date, and shifts the citation by a year.

The three clinical presentations this article has not covered each have their own page. Non-specific neck pain for the cervical spine, thoracic spine pain for the thoracic side, and non-specific low back pain for the lumbar spine. Two diagnoses not to miss in the screen worker complete the picture: cervicogenic tension-type headache and carpal tunnel syndrome.