Twenty-seven trials tested desk stretches. Strengthening won

This article is general information about a body of research. It is not advice about your neck, and it cannot be, because we have not met you. If something hurts persistently, that is a conversation with a clinician.

One disclosure before the evidence: we make a timer that puts strengthening movements into work breaks, so we are an interested party here. Check the citation yourself. It is a single named review and it takes about a minute to find.

What the review actually did

Chen and colleagues published a systematic review in Physical Therapy in 2018, pooling 27 randomised controlled trials of interventions aimed at neck and shoulder discomfort in office workers. Randomised controlled trials are the design that can tell you whether a thing worked, rather than whether the people who did it were already the sort of people who feel better.

Four broad approaches had been tested across those trials. Here is how they came out.

• Strengthening exercise: it worked. Supported by moderate-quality evidence, which in this literature is about as good as it gets.

• Stretching: no better than not stretching. Not harmful. Just not different from the control condition.

• Ergonomic changes on their own: did not work. New chair, new monitor arm, new keyboard, nothing else. No demonstrated effect.

• Education on its own: did not work. Being told about posture, in the absence of anything else, changed nothing.

Three of the four things almost every workplace does about desk discomfort are in the “did not work” column. The fourth is the one nobody schedules.

Why stretching feels like it is working

The result is counterintuitive because stretching produces an immediate sensation of relief, and that sensation is real. It is just not the same thing as an outcome.

Stretch your upper traps and something genuinely changes for a few minutes. What the trials measured was different: whether people who stretched regularly over weeks reported less discomfort than people who did not. On that question, the two groups landed in the same place.

This is a common pattern in health research and it catches out good, careful people. Immediate relief is easy to feel and easy to publish content about. Durable change is slow, unglamorous, and only visible in a comparison you cannot run on yourself.

So keep stretching if you like it. It is pleasant, it is free and nothing in the review suggests it does harm. Just do not file it under “the thing I am doing about my neck”, because on this evidence it is not doing that job.

What “strengthening” meant in these trials

Not neck rolls. Not chin tucks held for eight seconds. Resistance exercise, loading the muscles of the shoulder girdle and upper back to the point where they had to adapt.

That distinction is the practical takeaway, and it explains why the finding is so rarely acted on. Stretching fits into an office. Loading does not, at least not obviously, which is why the advice that spreads is the advice that is easy rather than the advice that worked.

The good news is that bodyweight movements load those muscles perfectly well, and several of them fit next to a desk without changing clothes.

What that looks like at a desk

Not a prescription, and not aimed at anyone’s specific pain. This is simply what “strengthening for the upper back and shoulders” looks like when the only equipment is a room.

• Pulling of any kind. This is the one that matters most and the one almost nobody does. A row against a doorframe, or against a towel looped around a door handle, loads exactly the muscles that a day spent reaching forward for a keyboard does not.

• Pressing at an angle you can control. Hands on a desk edge, body straight, chest to the desk and back up. Lower the surface as it gets easier.

• Carrying something heavy for thirty seconds. A full kettle in each hand, walked around the room. Unfashionable and effective.

• Anything that makes you hold your ribs and shoulders in place under load. A wall sit does more for your upper back than it looks like it does, because you have to hold everything else still.

Twice a week is where the World Health Organization pitches its muscle-strengthening recommendation for adults. Alongside it sits the weekly aerobic target: 150 to 300 minutes of moderate effort, or somewhere between 75 and 150 minutes if the effort is vigorous. That is the target worth aiming at, and a short set twice daily on five working days clears the strengthening half of it without effort.

A full five-minute version, written out move by move, is in the between-meetings routine.

How often, and how much

Nothing in the review fixes a dose, so this is inference rather than instruction. The trials that worked ran supervised resistance sessions over weeks and months, not single heroic efforts. Two things follow. Consistency beats intensity, because the adaptation is slow and cumulative. And the load has to be genuinely hard by the end of a set, or nothing is being asked of the muscle at all.

The honest limits of this finding

We would rather state these than have you find them yourself later.

• Moderate-quality evidence is not certainty. It means the finding is reasonably trustworthy and could still move if better trials are run.

• It is about groups, not about you. A review tells you what happened on average across 27 trials. It cannot tell you what will happen to one person, and plenty of people have discomfort with a specific cause that a general finding does not touch.

• “No better than control” is not “does nothing at all”. It means stretching did not beat the comparison. In some trials the comparison was itself getting up and doing something.

• It does not mean ergonomics are pointless. It means ergonomic change on its own did not produce the outcome. A chair that fits you is still better than a chair that does not. It is just not an intervention.

• Persistent or worsening pain is a clinical question. Numbness, pain down an arm, or anything that wakes you at night is not something to solve with a blog post.

Why we build the product this way

This is the reason PomoFit’s breaks contain squats and push-ups rather than neck rolls, and it is the single clearest example of the evidence pointing somewhere the category does not go.

Look at what is on offer in this space. Desk-stretch videos, posture reminders, ergonomic checklists. Three interventions, and the 27-trial review put all three of them in the column that did not work. The one that did work is the one that requires you to do something slightly effortful for thirty seconds, which is a harder product to sell and a better one to own.

We are not claiming PomoFit treats anything. It is a timer that puts a strengthening movement in front of you at intervals, and the argument for that design is the review above rather than anything we have measured ourselves.

The timer is here and the free tier does not ask you to register. Pro is $4.99 monthly, which adds scheduling shaped around you, exercises you define yourself, ambient audio, deeper statistics and leaderboards. If you would rather follow something already built, the plans library has twelve. Full body strength and weekly progressive strength are the two that match this article. Back pain desk relief is the gentler entry point, and it is a movement plan rather than a treatment.

The wider pattern

This is not an isolated result. It fits a broader shape in the desk-health literature, which is that the interventions that feel like they should work often do not, and the ones that do work are narrow and specific.

Breaking up sitting is the clearest parallel. It does something for post-meal glucose, takes a few points off systolic readings, and cuts fatigue. What it does not do is improve cognition, however many articles insist otherwise. The evidence for that is set out in do exercise breaks improve concentration. Exercise snacks tell the same story from a different angle. Eleven trials, gathered into a 2025 review for the British Journal of Sports Medicine, showed them raising cardiorespiratory fitness while leaving body composition, blood pressure and blood lipids untouched. Real benefits, and narrow ones. The terminology is untangled in exercise snacks, movement snacks and deskercise.

Being told which narrow thing you are buying is more useful than being promised everything. It is also, going by the state of the competition, apparently rare.