Posture correction exercises are sold on a chain of three claims: that your posture is bad, that the badness causes your pain, and that exercises will fix both. The first is a judgement, the second is much weaker than it sounds, and only the third has trials behind it — and what those trials measured was the angle, not the ache.
This page separates the three, because keeping them together is how a whole industry stays in business on evidence that would not support any one of them alone.
The short answer
- Exercise programmes can shift measured posture angles. Meta-analysed, modest, real.
- No trial has shown that shifting them reduces symptoms. That link is the missing one.
- Spinal curvature and pain are, repeatedly, not associated in the epidemiology.
- Trunk muscle endurance is associated with back pain, which points at capacity rather than shape.
- Posture braces and correctors have essentially nothing behind them.
Do posture correction exercises work
Split the question into the two things it hides.
Do they change posture? On the balance of evidence, yes, a little. Sheikhhoseini and colleagues meta-analysed therapeutic exercise for forward head posture in the Journal of Manipulative and Physiological Therapeutics in 2018 and found exercise programmes improved craniovertebral angle — the standard measurement of head position. Harman, Hubley-Kozey and Butler had run a ten-week randomised trial in the Journal of Manual & Manipulative Therapy in 2005 with 38 adults and found a combined stretching and strengthening programme improved forward head posture measures against a control.
Grade: moderate for the angle. Small samples, short programmes, measurements taken in standardised standing rather than in the wild, and no follow-up long enough to say whether the change outlives the programme.
Do they change how you feel? That is the question, and it is where the chain breaks. Neither of those studies was designed to answer it, and no trial has shown that moving an adult's habitual posture toward a textbook one reduces their pain. The outcome that gets measured is the outcome that is easy to measure.
The link that keeps failing
The reason nobody has demonstrated the symptom benefit is that the underlying association is far weaker than the advice assumes.
Richards and colleagues, in Physical Therapy in 2016, measured the sitting neck postures of more than 1,000 adolescents, sorted them into distinct clusters statistically rather than by eye, and looked for the association with neck pain. The clusters were real. The association was not.
Christensen and Hartvigsen reviewed the epidemiology of sagittal spinal curves and health in the Journal of Manipulative and Physiological Therapeutics in 2008 and found no consistent association between the shape of a spine and whether it hurt. Grob, Frauenfelder and Mannion, in the European Spine Journal in 2007, compared cervical spine curvature in people with and without neck pain and found no association there either. Swain and colleagues, reviewing the systematic reviews in the Journal of Biomechanics in 2020, concluded there is no consensus that spinal posture causes low back pain — the reviews disagree with one another.
Grade: reasonably solid for the negative. Different populations, different body regions, different research groups, the same answer.
State the counterweight at full strength. Mahmoud and colleagues, in Current Reviews in Musculoskeletal Medicine in 2019, meta-analysed forward head posture and neck pain and did find an association in adults — small to moderate, and absent in adolescents. It is cross-sectional, so it cannot say which came first, and a person whose neck hurts may hold their head differently because of it. That is the strongest finding the posture side has and it is not enough to carry the prescription.
What is associated with back pain
If shape is not the lever, something is. The best candidate in the same literature is capacity.
Nourbakhsh and Arab measured 600 people in the Journal of Orthopaedic & Sports Physical Therapy in 2002 and reported the comparison plainly: lumbar lordosis, pelvic tilt and leg length discrepancy showed no significant association with low back pain, while trunk muscle endurance and hip muscle tightness were strongly associated with it.
The office-worker trials point the same way. Andersen and colleagues, in Pain in 2011, gave workers with frequent neck and shoulder pain two minutes a day of progressive resistance training and measured a real reduction over ten weeks; their 2008 trial in Arthritis & Rheumatism found specific strength training beat general fitness training; Blangsted and colleagues' one-year randomised trial in the Scandinavian Journal of Work, Environment & Health in 2008 found specific resistance training reduced neck and shoulder symptoms where an all-round activity programme did not.
Grade: moderate, and it is the best evidence in this area. Randomised, workplace, clinical outcomes.
So the exercises worth doing are not wrong — they are usually mislabelled. A programme of scapular and neck strengthening plus some chest and upper-back mobility is a reasonable thing to do. Calling it correction attaches it to a mechanism that has not held up.
What about posture braces
Very little. Devices that pull the shoulders back have no serious trial evidence for changing symptoms, and their proposed mechanism is the same one the epidemiology keeps failing to support. A brace also does the work the muscles would otherwise do, which is the opposite of the capacity finding above.
The honest position is that they are a comfort item, not a treatment, and they should be described as one.
A programme, described accurately
Here is what a defensible version looks like, with the claim attached to each part.
- Move often. Two minutes an hour. This has weak but consistent evidence for the discomfort of a held position and none for correction. See the micro-breaks research.
- Restore the directions a screen day removes — thoracic extension and rotation, shoulder external rotation, hip extension. Eight desk-viable movements with their holds are in desk stretches.
- Add a small dose of resistance training for the neck, shoulder blades and trunk. This is the part with randomised trials behind it. Two minutes a day was enough in the trial that measured it.
- Do not chase an angle. People are poor judges of their own posture, the measurements clinicians use have limited reliability, and no target position has been shown to be the right one.
That programme is worth doing. It is worth doing because it adds movement variety and capacity to a day short of both, not because your spine is in the wrong shape.
Where pain belongs
Posture is the most over-claimed explanation in this area, and the cost of over-claiming it is that people spend months adjusting themselves instead of getting examined.
Back and neck pain are common and usually settle. Pain that radiates down an arm or a leg, numbness or pins and needles, weakness, pain that followed a fall or an injury, pain that wakes you at night, unexplained weight loss, fever, or any change in bladder or bowel control needs a clinician who can examine you. Nothing on this page diagnoses anything and nothing on it is treatment.
What we will not measure for you
Limber does not score your posture, and it is worth saying why on a page like this.
There is no posture reading in the battery. The five tests measure range at specific joints — posterior chain, ankle dorsiflexion, shoulder reach, hip flexor length, squat depth — and a range reading is not a posture judgement. The wall angel test, which is the closest thing in the design to a posture measure, is not built, in the app or here. The thoracic spine has no reading at all yet, and the neck score is inferred from the shoulder reach as a named proxy.
What you can take today is the shoulder mobility test, both sides, five bands, and the Range Score, which composes what you have measured and draws the rest as unmeasured ground. If a product offers to score your posture from a photograph, ask what the score was validated against. Forward head posture covers the measurement problem in detail, and the cluster is desk and posture.
Questions
Do posture correction exercises work?
They can change measured posture angles by a modest amount — that much is meta-analysed. What no trial has shown is that changing the angle changes symptoms. So the honest answer is that they work on the thing being measured and have not been shown to work on the thing people want.
Can you actually fix bad posture?
You can change how you habitually stand and sit, with effort and over months. Whether you should is a separate question, because the epidemiology repeatedly finds no consistent association between spinal shape and pain. There is also no established correct posture to aim at, which makes "fixing" a strange word for it.
Does bad posture cause back pain?
The systematic reviews do not support that claim. A review of reviews on spinal posture and low back pain found no consensus, epidemiology of spinal curvature and health found no consistent association, and a study of 600 people found lumbar lordosis unrelated to back pain while trunk muscle endurance was strongly related.
How long does it take to improve posture?
The trials that measured angle changes ran eight to sixteen weeks. Whether the change persists after the programme stops is largely unstudied. Given the weak link to symptoms, the more useful question is how long it takes to build the capacity and movement variety a desk day removes — and the answer to that is a few minutes most days, indefinitely.
Are posture correctors or braces worth it?
There is no serious trial evidence that they change symptoms, and their proposed mechanism depends on the same posture–pain link that keeps failing. They also do work your muscles would otherwise do, which runs against the one finding here with good support: capacity matters.
What should I do instead if my back or neck hurts?
Move position often, restore the directions a desk day removes, and add a small dose of resistance training — that last part is what the randomised office-worker trials support. And if the pain radiates into a limb, comes with numbness or weakness, or followed an injury, see a clinician who can examine you rather than adjusting your chair again.