The usual advice about working from home stretches starts with the chair, because the chair is the visible difference between an office and a kitchen table. The chair is not the important difference.
What home working actually removes is incidental movement: the commute, the walk to a meeting room, the trip to the printer, standing in a queue for coffee. Those add up to a surprising amount of standing and walking that nobody planned and everybody lost. A routine that answers the chair and ignores that is answering the wrong problem.
The short answer
- The real loss is incidental movement, not lumbar support.
- Home-office surveys found most people working on laptops, on non-adjustable chairs, with more reported physical complaints.
- Sitting bouts get longer at home, and bout length is the part associated with mortality in the accelerometer data.
- The routine below rebuilds the movement rather than the furniture.
- One piece of equipment is worth buying before anything else, and it is not a chair.
Working from home stretches
Six movements, all needing nothing but a metre of floor and a doorframe. Run them once mid-morning and once mid-afternoon; the whole set is about five minutes.
1 · Standing shoulder circles. Six slow circles each way, biggest controlled range, no shrugging. 40 seconds.
2 · Chin tucks. Eight slow repetitions, one per breath. Chin straight back over the throat, not down. 40 seconds.
3 · Standing side bend. One arm overhead, reach long, lean away. 30 seconds each side.
4 · Tall rotation. Seated or standing, arms crossed, turn and hold at the end. 40 seconds each side. Thoracic rotation is the movement a screen day removes most completely.
5 · Doorway pec stretch. Forearm on the frame, elbow around shoulder height, step through. 30 seconds each side.
6 · Half-kneeling hip flexor stretch. At home you have floor, which the office denied you — use it. Pelvis tucked under, ribs down, glute on, then lean. 30 seconds each side. Without the tuck, the lower back arches and the hip barely moves.
Numbers 3 to 6 are the four directions a sitting day removes. Number 6 is the one you can finally do properly at home and the one people still skip.
What the home-office surveys found
The evidence base here is young and it is mostly cross-sectional survey work from 2020 and 2021, which is a real limitation and worth naming before the findings.
Gerding and colleagues surveyed the home workstations of university employees sent home in 2020, in Work in 2021, and found a large share working on laptops without external keyboards or monitors, and a large share using non-adjustable chairs. Xiao and colleagues, in the Journal of Occupational and Environmental Medicine in 2021, surveyed office workstation users working from home and found a majority reporting new physical health issues, with those who had adjusted their workstations reporting fewer.
Grade: weak. Self-selected online surveys, self-reported outcomes, no control group, conducted during a period when a great deal else changed at once. What they establish is a description of the setups people ended up with, not a causal account of what those setups did.
The more solid evidence is the general sedentary literature, and it applies with more force at home because the bouts get longer. Diaz and colleagues, in Annals of Internal Medicine in 2017, followed almost 8,000 adults with accelerometers and found both total sedentary time and the duration of individual sedentary bouts associated with mortality. Working from home lengthens bouts by removing the interruptions that used to be built into the day.
Rebuild the interruptions
This is the part that matters more than any stretch on this page.
- Take the meetings you can on foot. An audio call is a walk. This is the single largest recoverable block of movement in a home working day.
- Put something you need in another room — the water, the printer, the phone charger. The office's movement was accidental and yours has to be arranged.
- Break the long blocks deliberately. The metabolic trials used two minutes of light walking every 20 minutes and cut post-meal glucose and insulin substantially — Dunstan and colleagues in Diabetes Care in 2012. Peddie and colleagues, in the American Journal of Clinical Nutrition in 2013, found regular short breaks lowered post-meal glycaemia more than one continuous 30-minute walk with the same total activity.
- Keep a commute. A ten-minute walk before starting and after finishing replaces the one you lost and marks the boundary of the day, which is the other thing home working removed.
The micro-breaks research sets out exactly how strong each of those findings is, and it is more honest than the summary above has room for.
The one thing worth buying
Not a chair. An external keyboard, or a laptop stand — ideally both.
A laptop attaches the screen to the keyboard, so any height that suits your neck ruins your wrists and any height that suits your wrists puts your head down. It is a structural problem rather than a discipline problem, and no amount of posture attention solves it. Separating the two is the cheapest meaningful change in most home setups.
After that, the specific ergonomic rules have thinner evidence than their confidence suggests. Hoe and colleagues' Cochrane review, updated in 2018, found low-quality evidence and no clear demonstration that ergonomic training or workstation adjustment prevents neck and upper-limb disorders — with one signal, for forearm support used with an alternative pointing device. That came from Rempel and colleagues' year-long randomised trial in Occupational and Environmental Medicine in 2006, which found a forearm support board reduced the risk of developing neck and shoulder disorders in 182 workers.
So: separate the screen from the keyboard, get your forearms supported, and stop optimising. Desk setup and mobility works through the rest of the furniture question.
The sofa, and other bad surfaces
Working from a sofa or a bed is not going to injure you, and it is worth being clear about that rather than joining the alarm.
What it does do is make position changes rarer, because a soft surface offers fewer of them, and it makes the laptop problem worse because there is no surface to raise it on. If it is what you have, the answer is not guilt — it is to shorten the blocks. An hour on the sofa broken by a two-minute walk is a different exposure from three hours on the sofa.
What this routine will not do
It will not correct your posture. Richards and colleagues, in Physical Therapy in 2016, found no association between measured sitting neck posture clusters and neck pain in over 1,000 adolescents, and Swain and colleagues, in the Journal of Biomechanics in 2020, found no consensus that spinal posture causes low back pain.
It will not replace activity. Ekelund and colleagues' pooled analysis of over a million adults in The Lancet in 2016 found the sitting–mortality association largely attenuated by 60 to 75 minutes a day of moderate activity. Stretching is not that, and the walk is the more important half of this page.
And it will not answer pain. Back and neck pain are common and usually settle. Pain radiating into an arm or a leg, numbness, weakness, pain that followed a fall, or pain with fever or unexplained weight loss needs a clinician who can examine you. Nothing here diagnoses anything and nothing here is treatment.
Measure it, since you are at home anyway
Home is where the battery is easiest to run: you have floor, a wall, a doorframe and privacy, which the office denied you.
The toe touch test reads the posterior chain in five bands in about a minute. The Range Score composes whatever tests you have taken into a single 0–100 reading, restates every left–right gap it finds rather than averaging your sides into one flattering number, and draws the ground you have not surveyed as unsurveyed rather than filling it in.
Take a baseline, work for three weeks, take it again under the same conditions. Limber's default re-test interval is 21 days, adjustable to 14 or 28, and the app keeps your first reading as a datum and measures every later one from it. On the web nothing is stored at all — the survey lives in the address bar, so a reading you want to keep is a link you save.
The cluster is desk and posture.
Questions
What is the best stretching routine for working from home?
Six movements covering the four directions a sitting day removes: shoulder circles, chin tucks, a standing side bend, a tall rotation, a doorway pec stretch and a half-kneeling hip flexor stretch. About five minutes, twice a day. The last one is the movement home finally makes possible and the one most people still skip.
Why does working from home make me stiffer than the office did?
Most likely because of what it removed rather than what it added. The commute, the walk to a meeting and the trip to the printer were unplanned movement that nobody counted. Sedentary bouts get longer at home, and bout length is the feature associated with mortality in the accelerometer studies.
Is working from a sofa bad for you?
It will not injure you, and the alarm about it is overdone. What it does is reduce how often you change position and make the laptop height problem worse. If the sofa is what you have, shorten the blocks — an hour interrupted by a walk is a different exposure from three hours uninterrupted.
What should I buy first for a home office?
A laptop stand and an external keyboard, so the screen and the keyboard can be at different heights. That is a structural fix for a structural problem. After that, forearm support has the one piece of good trial evidence in this area; the rest of the ergonomic rules are weakly supported.
Do I need an expensive chair to work from home?
No evidence suggests price prevents pain. A chair you can change position in is more defensible than a chair that holds you in one, because the exposure that matters is time in a single shape. The money is better spent on making it easy to stand up and walk.
How often should I get up when working from home?
More often than the day naturally makes you, which at home is rarely. The controlled trials that found metabolic effects used two minutes of light walking every 20 minutes; the discomfort studies find small benefits anywhere from 20 minutes to an hour. Two minutes each hour is a defensible default.