Posture guide
Posture is the position your body is in as you sit, stand, move and sleep, and it changes all day. The idea that one correct posture protects your back and neck is everywhere, but the research doesn't back it up: physiotherapists don't agree on a best posture, reviews find no agreement that posture causes back pain, and in studies that followed people over time, posture didn't predict who got neck pain.
What helps is moving often, getting stronger and staying active. This page gives the short verdicts, with evidence grades: whether posture causes pain, how much exercise changes posture (a few degrees), and what to make of correctors, sleeping positions, school bags and power poses. The guides below take each shape in turn, with exercises and 3D demos.
Physiotherapists shown the same photos choose different best postures, and people with and without back pain have similar curves in their lower backs. The Chartered Society of Physiotherapy says "There is no perfect posture or right or wrong desk set up" and "Your best posture is your next posture".
Reviews find no agreement that sitting, standing, bending or lifting postures cause low back pain. When researchers followed people over time, posture didn't predict who got neck pain; poor sleep, low activity and doing the same task all day did.
Changing position through the day is what physiotherapists and the UK's workplace health regulator advise. For aches and pains, exercise helps, and strength training has the strongest evidence.
In trials, exercise made the head about 3 degrees more upright, and spine strengthening reduced older adults' upper-back curve by about 3 degrees. Pain can ease with no change in posture, so exercise for strength and comfort, not to reach a shape.
Posture correctors and posture shirts change how you hold yourself while you wear them, but there is no good evidence that they ease pain. NICE (the National Institute for Health and Care Excellence) advises against belts and corsets for low back pain.
A curve in a growing child, or new hunching, height loss or back pain with muscle spasms in later life, should be checked. The warning signs are listed below.
Each guide takes one shape or habit: how it's measured, what the evidence says, the exercises that help and when to get it checked.
Desk set-up from HSE and OSHA, why the best posture is your next one, and what trials say about standing desks, breaks and chairs. Desk moves with 3D demos.
Forward head postureForward head posture exercises from the trials: deep neck and upper-back strength, how much they move your head, and the neck symptoms that need a GP.
Rounded shouldersRounded shoulders exercises from the trials: why strength beats chest stretches, what rows can change, how long it takes, and what it means for shoulder pain.
KyphosisKyphosis exercises with 3D demos: what a hunched back is, how much exercise reduces the curve, safe moves with osteoporosis, and when a new hunch needs a GP.
Neck humpA neck hump can be a fat pad, a rounded upper back or normal bone. How to tell them apart, what exercise can and cannot change, and when to see a GP.
Anterior pelvic tiltMost people's pelvis tilts forward a little. What's normal, how it's tested, the hip and core exercises that help, and what they can and can't change.
Swayback and lordosisSwayback isn't the same as a deep low-back curve. What's normal, how both link to back pain, exercises with 3D demos, pregnancy, flat back and warning signs.
ScoliosisScoliosis exercises explained: what Schroth and other specific programmes change, safe general exercise for teens and adults, bracing, and warning signs.
Knock kneesMost knock knees in children straighten by age 7, and exercise can't reshape adult bones. But knees that cave in on squats and landings are trainable.
Posture is the position you hold for what you're doing, and it shifts with tiredness, comfort and mood. In 1,108 17-year-olds, those who sat slumped with the head forward were more likely to be depressed, and those who sat upright exercised more often. That was a snapshot, so it can't show which came first.
Source 23
Posture is measured from side-on photos, simple tools or X-rays. There are no agreed cut-offs between normal and not, and studies use different ones.
There is no agreed definition, so there is no reliable figure. When studies use their own definitions, large shares of ordinary people qualify: 20% to 40% of older adults have hyperkyphosis, for example. A shape that so many people have is better thought of as normal variation.
Source 24
A review of 41 reviews found no agreement that spine posture, sitting, standing, bending or awkward positions cause low back pain. Of 24 studies of sitting at work, the best found no link. On skin measures, people with and without back pain have similar lower-back curves; on X-rays, people with back pain tend to have slightly flatter lower backs, not more arched ones, and these snapshots can't show which came first.
Rounding your lower back as you lift has not been shown to be a risk factor for back pain; in 7 studies, people with back pain bent their lower backs less when lifting, not more. Heavy, frequent lifting at work raises the risk a little: the odds rose by about 11% for every 10 kg lifted and by 9% for every 10 lifts a day.
In 7 trials of 19,317 workers, those taught how to lift had as much back pain as those who weren't. Most physiotherapists and manual handling advisers still believe a straight back is safest, though there is no evidence that any one lifting posture prevents back pain.
In studies done at one point in time, adults with neck pain had slightly more forward head posture than adults without; teenagers showed no difference. When people were followed, posture didn't predict neck pain: sitting posture at 17 didn't predict persistent neck pain at 22, and girls who sat upright were more likely to have it than girls who slumped.
In 396 adults, the angle of the neck while texting didn't predict new neck pain over a year, but poor sleep and low activity did. Our tech neck guide covers neck pain from phones and screens.
In studies that followed office workers, the risk factors for new neck pain were low satisfaction with the workplace, a keyboard close to the body, little variety in tasks and feeling tense. Posture was studied but wasn't among them. The Chartered Society of Physiotherapy now says posture plays a much smaller role in neck problems than previously thought.
Shown nine sitting postures, 295 physiotherapists in four countries mostly chose one of two very different ones. In a survey of 544, three different sitting postures were picked as best, though there is no strong evidence that any posture leads to better health. Healthy people asked to copy one "ideal" sitting curve needed feedback or hands-on help to produce it.
The Chartered Society of Physiotherapy says "There is no perfect posture or right or wrong desk set up", "The key is to keep moving" and "Your best posture is your next posture". NHS inform says your posture doesn't need to be perfect, so move often rather than staying in one position for too long.
The US workplace safety agency, OSHA, says working in the same posture or sitting still for long periods is not healthy, however good the posture.
Some NHS pages still say sit upright and list bad posture as a common cause of neck pain, and some hospital leaflets teach one correct way to sit. This guide follows the newer research and the physiotherapists' advice.
Short, frequent breaks are better than long, rare ones: the Health and Safety Executive (HSE), the UK's workplace health regulator, says 5 to 10 minutes every hour beats 20 minutes every 2 hours. In office workers, little variety in tasks went with more neck pain.
Our editorial starting point: stand up and move for a minute or two every 30 to 60 minutes, and change position in between. The sitting guide has a 3-minute desk break.
In a review of 249 trials, exercise reduced long-lasting low back pain by about 15 points out of 100 compared with no treatment or usual care, and no type was clearly best. NICE says to consider a group exercise programme for low back pain. The UK guidelines ask for muscle-strengthening activity on at least 2 days a week; any is better than none.
For long-lasting neck pain, strengthening, motor control exercise (training control of the deep neck muscles), and yoga, Pilates or tai chi helped about equally. In office workers with neck pain, neck and shoulder strengthening and general fitness training both reduced pain, and in 198 workers, 2 minutes a day of band exercise, 5 days a week, cut neck and shoulder pain by 1.4 points out of 10 in 10 weeks.
In a review of 10 trials of people with forward head posture and neck pain, 3 to 8 weeks of exercise made the head about 3.4 degrees more upright than in comparison groups. In 99 older adults with a rounded upper back, 6 months of spine-strengthening classes 3 times a week reduced the curve by 3 degrees on X-ray, and across 24 studies of age-related hyperkyphosis the effect was small. Smaller trials report bigger changes, but they are more open to bias.
The programmes that worked usually combined strengthening, such as rows and back extensions, with a chest stretch and chin tucks.
In a taping trial, pain eased while head posture didn't change, and the two weren't linked. Exercise that doesn't aim at posture eases neck pain as much as exercise that does. When spine strengthening reduced older adults' curves, their everyday physical function didn't change, and a review of forward head posture says the link between posture change and pain change is not established. Exercise for strength, comfort and confidence, not to reach a shape.
In healthy adults, 12 weeks of Pilates twice a week didn't change posture. A review of 9 studies (643 people), which measured scoliosis curves and posture, reported improvements, but the studies were of mixed quality. In 118 older adults with a rounded upper back, 6 months of yoga improved one measure of the curve by 4.4% but not another. Both help neck pain about as much as other exercise.
Braces for forward head posture changed the head angle by 3 to 5 degrees in small studies, mostly measured while people were wearing them. A review of posture shirts found 6 studies, all in people without pain and mostly of poor quality, and no good evidence that the shirts help pain. If a corrector reminds you to move, use it for short spells, not as a fix (our editorial view).
Lumbar supports don't prevent low back pain (moderate-quality evidence from 7 trials of 14,437 people), and NICE says belts and corsets should not be offered for low back pain.
Kinesiology tape changes posture briefly at best. In older adults with a rounded back, stretched and unstretched tape reduced the measured curve by the same amount, which points to something other than the tape, and across trials taping didn't reduce older adults' upper-back curve. In a trial of 60 people, exercise did better than tape for forward head posture.
Wearable sensors measure posture accurately. But in a trial of 130 care workers, a device that beeped when they bent forward didn't change how often they did. In one pilot trial, movement retraining with sensor feedback, alongside usual care, helped back pain; the device maker paid for half of it. We found no trial of a posture app on its own.
Hyperkyphosis affects about 20% to 40% of older adults. It goes with difficulty rising from a chair, poorer balance and slower walking, and a bigger curve went with more falls, including in a study that followed older people for a year; those studies can't show that the curve causes the falls.
In adults aged 45 and over with hyperkyphosis, exercise reduced the curve a little and improved quality of life slightly; its effect on falls is uncertain. The kyphosis guide has the exercises, and the older adults guide covers strength and balance.
Unexplained back pain with muscle spasms, height loss, or a curved spine or change in posture can be signs of spinal fractures from osteoporosis. See a GP, especially if you are 50 or over, have been through the menopause or take steroid medicines.
There isn't enough research to name a best position. A review found only 4 studies, in which lying on your side was linked with fewer morning aches, and people who woke with neck symptoms spent more time in awkward positions. Sleep in the position that feels comfortable, and try a change if you wake stiff (our editorial suggestion).
In a trial of 313 people with long-lasting low back pain, a medium-firm mattress gave less pain in bed and less disability after 90 days than a firm one, and a review of 39 studies favoured medium-firm for comfort and sleep. No type of pillow has been shown to be clearly better for neck pain.
A review of 69 studies found no convincing evidence that a schoolbag's weight, design or the way it's carried raises the risk of back pain in children, and bags heavier than 10% of body weight weren't linked with more low back pain. Suggested limits range from 5% to 20% of body weight, so no single figure is evidence-based. Two straps and heavy items packed close to the back make a bag more comfortable, and heavy backpacks don't cause scoliosis.
The claim that power posing raises testosterone came from a study of 42 people. A repeat study in 200 people found no effect on testosterone, cortisol (a stress hormone) or risk-taking, just a small rise in feeling powerful, and the original study's first author has said she doesn't believe the effect is real.
In lab studies, sitting upright through a stressful task went with better mood and self-esteem than slumping, including in people with mild to moderate depression. Across 73 experiments, the difference came from slumped postures lowering mood, not from expansive poses raising it. Slumping also lowers lung-function readings while you slump.
Use these whenever you've been still for a while: 8 to 10 sit-to-stands, 30 to 45 seconds of marching, 6 rotations and 8 to 10 open books each side, 8 slow cat-cows, 5 to 10 slow chin tucks, 3 neck circles each way (start with small turns if your neck is sore) and chest stretches of about 30 seconds, with walking building towards 150 minutes a week (walking from the UK guidelines; the other amounts are the moves' own or our editorial starting points).
3D demoTreadmill walkingBeginner · Cardio machine
3D demoSit to stand from a chairBeginner · A bench or box
3D demoStanding marchBeginner · No equipment
3D demoThoracic rotationBeginner · No equipment
3D demoSide-lying Open Book RotationBeginner · No equipment
3D demoCat-cowBeginner · No equipment
3D demoChin tuckBeginner · No equipment
3D demoNeck CARsBeginner · No equipment
3D demoDoorway chest openerBeginner · A wall
3D demoElbows backBeginner · No equipmentOn 2 or 3 days a week, start with 2 to 3 sets of 10 to 15 band rows or pull-aparts, 5 to 8 of each Y, T and W, 8 slow wall angels, back lifts of 3 to 5 reps held 3 to 5 seconds, 3 rounds of 6 bird dogs each side, 3 sets of 12 bridges and 3 carries of 20 to 30 metres, and add load as it gets easy (our editorial starting points; the back-lift dose is the one the seated lift's guide takes from the UK osteoporosis consensus).
3D demoBand rowBeginner · A band + An anchor point
3D demoBand pull-apartIntermediate · A band
3D demoProne Y-T-W raisesBeginner · No equipment
3D demoWall angelBeginner · A wall
3D demoSeated upper-back lift (sit tall, shoulders back)Beginner · A bench or box
3D demoLow cobra (Bhujangasana)Beginner · No equipment
3D demoBird dogBeginner · No equipment
3D demoGlute bridgeBeginner · No equipment
3D demoFarmer's carryBeginner · A dumbbell + A kettlebell
3D demoSupported single-arm dumbbell rowIntermediate · A dumbbell + A bench or boxPick one you'll keep doing: the upper-back routine and the band and home workouts for strength, Pilates or yoga if you prefer a class style, and the chair workout if standing is hard.
A 12-minute routine for a stiff upper back and rounded shoulders: chin tucks, wall angels, Y-T-W raises and open books.
12 minBeginner8 movesA foam roller, A wallA full-body resistance band workout: squats, presses, pull-aparts, band walks, good mornings and flyes.
25 minBeginner7 movesA band, An anchor pointAn upper body workout at home with no gym: push-ups, band rows, pike push-ups, face pulls, close-grip push-ups and Y-T-W raises, with dumbbell options.
30 minIntermediate13 movesA band, An anchor pointA 15-minute beginner core workout on the floor: dead bugs, bird dogs, knee planks, side planks and curl-ups.
15 minBeginner6 movesNo equipmentEight gentle neck and shoulder stretches for desk stiffness, done standing or seated in ten minutes.
10 minBeginner8 movesA wallA 20-minute beginner wall Pilates workout: standing and floor moves that use a wall for support and feedback. No equipment.
20 minBeginner9 movesA wallA 20-minute beginner yoga sequence: mountain, chair, warrior II, triangle, tree, down dog, cobra and child's pose.
20 minBeginner9 movesA stepA 20-minute seated workout for older adults who cannot stand for long: marching, leg and arm strength, posture, and sit-to-stands that lead back to standing.
20 minBeginner11 movesA bench or box, Something with weightFor a whole week of training with strength work built in.
Your first four weeks: two, then three short strength sessions a week at home plus walks, building gently to a full routine.
2 training days20 minBeginner3 days a weekA free 4-week, 3-day Pilates plan for complete beginners. Two mat days and one wall day, with clear rules for when to move each exercise to its harder version.
3 training days25 minBeginner3 days a weekA weekly plan for women over 50: strength, bone-loading and balance on 3 days plus walking, built on UK osteoporosis consensus advice.
3 training days45 minBeginnerGet emergency help for any of these:
Ask for an urgent GP appointment or contact NHS 111 if you have back pain and you feel hot, cold, shivery or generally unwell, or the pain is severe, started suddenly or is getting worse quickly.
Source 77
Book an appointment for any of these:
A GP, or in many areas NHS physiotherapy, which you can contact without seeing a GP first. Our neck pain guide lists the neck symptoms that need urgent care.
Reviews find no agreement that posture causes low back pain, sitting at work on its own is unlikely to, and people with and without back pain have similar lower-back curves.
Girls who sat upright at 17 were more likely to have persistent neck pain at 22 than girls who slumped, and the angle of the neck while texting didn't predict new neck pain.
Bending your back when you lift hasn't been shown to cause back pain, and safe-lifting training doesn't prevent it. Physiotherapists' advice is not to fear bending and to keep moving.
Correctors change posture while you wear them in small, short studies, and there's no good evidence that they ease pain.
The NHS says the most common type of scoliosis isn't thought to be linked to posture, and heavy backpacks don't cause it.
Exercise changes posture measures by a few degrees on average. The NHS says exercises for kyphosis won't improve the curve, though trials in older adults show small reductions. Expect a little change, not a straight spine.
A repeat study in 200 people found no effect on testosterone or cortisol, and the original study's first author no longer believes the effect is real.
Published limits range from 5% to 20%, and heavier bags weren't linked with more back pain. A lighter bag is more comfortable, but the exact figure isn't evidence-based.
Probably not on its own. Reviews of the research find no agreement that sitting, standing, bending or lifting postures cause low back pain, and people with and without back pain have similar curves in their lower backs. Heavy, frequent lifting at work raises the risk a little, so how much you lift matters more than how you bend.
Adults with neck pain have slightly more forward head posture on average, but that kind of study can't show which came first, and teenagers show no difference. When researchers followed people over time, posture didn't predict who got neck pain, while poor sleep and low activity did. Holding one position for a long time can make a sore neck ache, so changing position helps.
There isn't one correct posture. Physiotherapists choose different best postures, and the Chartered Society of Physiotherapy says your best posture is your next posture. A good posture is one that feels comfortable and that you change often.
They change how you hold yourself while you wear them, but the studies are small and short, and there's no good evidence that they ease pain. NICE advises against belts and corsets for low back pain. If one reminds you to move, use it for short spells alongside exercise, not as a fix.
Exercise changes posture measures, but only by a few degrees on average: about 3 degrees in head position after a few weeks, and about 3 degrees in the upper-back curve after 6 months of back strengthening in older adults. Whether those degrees matter for your health is much less clear than adverts suggest. Pain and stiffness can improve whether or not your posture changes.
In trials, programmes ran from 3 weeks to 8 months, often 3 sessions a week, and in a recent review of forward head posture, programmes of 8 weeks or more tended to do better. Expect weeks to months, not days. Judge progress by how you feel and move as much as by photos.
No single exercise is best. Programmes that changed posture usually combined strengthening, such as rows and back extensions, with a chest stretch and chin tucks. For aches, strength training has the strongest evidence, and the best routine is the one you'll keep doing.
No. The NHS says the most common type of scoliosis isn't thought to be linked to bad posture, and orthopaedic surgeons say heavy backpacks don't cause it. See a GP if you notice signs such as uneven shoulders, one hip sticking out or a curve that's getting worse.
The hormone claim didn't hold up: a study of 200 people found no effect on testosterone, cortisol or risk-taking, and the original study's first author has said she doesn't believe the effect is real. People may feel a little more powerful afterwards, and slumping may dampen mood slightly compared with sitting normally.
There isn't enough research to name one. Small studies link lying on your side with fewer morning aches, and in a trial of 313 people with long-lasting low back pain, a medium-firm mattress helped more than a firm one. No type of pillow has been shown to be clearly better for neck pain.
Studies haven't shown that a schoolbag's weight, design or the way it's carried causes back pain in children, and bags over 10% of body weight weren't linked with more back pain. Two straps and heavy items packed close to the back make a bag more comfortable. A child whose back pain doesn't settle, or who has a curve you can see, should see a GP.
Slumping lowers lung-function readings while you slump, but there's no evidence that it harms your lungs. Don't put chest pain down to posture: call 999 for chest pain that comes on suddenly and doesn't go away, or that spreads to your arm, neck, jaw or back, or comes with sweating, feeling sick or breathlessness.
Exercise education, not a diagnosis or treatment plan, and clinical review of these guides is pending. Get a curve or hunch that is new or getting worse checked, above all in a growing child or after a fall, and get urgent help for numbness or weakness in the legs or a change in bladder or bowel control. Health disclaimer.
