Where Yoga Helps in Rehabilitation, and Where It Does Not

A physiotherapist suggests yoga for back pain. A friend says yoga is what caused theirs. Both experiences are represented in the published trials. Yoga is a form of exercise, and in the trials it behaves like one. Against doing no exercise it produces small improvements in chronic low back pain and function, most of them below the threshold the reviewers themselves call clinically important. Against other active exercise it is about level. In older adults it improves balance and mobility. And it carries a measurable rate of adverse events, mostly more back pain. A useful answer contains all four of those parts.

Chronic low back pain: what the Cochrane review found

The 2022 Cochrane review by Wieland and colleagues included 21 randomised trials covering 2,223 participants, mostly women in their forties and fifties, using iyengar, hatha or viniyoga styles. Its findings are unusually precise about size, and the precision is the point.

• Compared with no exercise at three months, there may be a small improvement in back-specific function and probably a small improvement in pain. The review describes both as clinically unimportant, because each falls short of the minimal clinically important difference it names for that scale.
• There is probably a small improvement in physical and mental quality of life, and little to no improvement in depression.
• Compared with back-focused exercise such as physical therapy, there is probably little or no difference in back-specific function.
• Every included trial was at high risk of performance and detection bias, because neither participants nor providers can be blinded to whether they are doing yoga, and the outcomes were self-assessed.

A separate 2020 systematic review by Zhu and colleagues, pooling eighteen trials, reached the same place by a different route: yoga beat non-exercise controls on pain and disability, and showed no significant difference against physical therapy exercise.

What the guideline actually recommends

NICE guideline NG59 recommends considering a group exercise programme for people with low back pain with or without sciatica, and names the options as biomechanical, aerobic, mind-body, or a combination. Yoga is one kind of mind-body exercise, so it sits inside that recommendation rather than being singled out by it. The guideline's own sentence continues: take people's specific needs, preferences and capabilities into account when choosing the type of exercise.

That is the honest shape of the evidence. The recommendation is for exercise. Which kind is largely a question of what a particular person will keep doing.

Balance and mobility in older adults

Here the evidence is more favourable to yoga specifically. Youkhana and colleagues pooled six trials of relatively high methodological quality covering 307 participants aged 60 and over, and reported a small effect on balance performance and a medium effect on physical mobility compared with control conditions. The authors stop short of the next claim: whether those improvements translate into fewer actual falls is, in their words, still a question for further research.

The harms are in the literature too

The Cochrane review found low-certainty evidence that yoga increased the risk of adverse events relative to no exercise at six to twelve months, the most common being increased back pain. That is not a reason to avoid yoga. It is a reason to treat it as an intervention with a risk profile rather than as something categorically gentle.

Cramer and colleagues reviewed published case reports and case series of yoga-associated adverse events. The musculoskeletal system was the most commonly affected, followed by the nervous system and the eyes, and headstand, shoulder stand, lotus position and forceful breathing were the techniques most often implicated. Ten of the reported cases involved a pre-existing condition, mainly glaucoma or osteopenia. The authors are explicit that a review of case reports cannot estimate how often any of this happens, and nothing here should be read as a rate.

What this means for you

Yoga is exercise, with an exercise's ordinary trade-offs: real but modest benefit for chronic low back pain, roughly on par with other supervised exercise, better evidence for balance and mobility after 60, and a measurable rate of soreness or worse. None of that makes it a bad choice or a superior one. It makes it one option among the several supervised, back-focused programmes NICE recommends choosing between based on what a person will actually keep doing.

Anyone with glaucoma, osteopenia, uncontrolled hypertension or a recent surgical repair, or with back pain that has not yet been assessed, is the specific case the case-report literature says needs a modified approach or an assessment first, not this article.

Where it does not belong

• As a substitute for assessment. Back pain that has not been assessed is not a candidate for any exercise class, yoga included.
• As a substitute for measured, progressive loading where that is the goal. A rehabilitation target expressed as a strength or symmetry criterion needs a programme that can be progressed against a measurement, which a general class is not built to do.
• Unmodified, for a person with glaucoma, osteopenia, uncontrolled hypertension or a recent surgical repair. The case-report literature clusters around specific postures for specific reasons.
• As a claim of superiority. On the strongest available comparison, yoga and back-focused physical therapy exercise are probably about the same for function in chronic low back pain.

Quick reference

Mind-body exercise

The category NICE guideline NG59 uses for exercise combining movement with attention to breath and body awareness. Yoga and tai chi are examples.

Hatha

A broad, generally slower style of yoga built around held postures and breathing. One of the styles most used in the trials the Cochrane review pooled.

Ashtanga

A vigorous style in which a set sequence of postures is performed in a fixed order, linked to the breath.

Minimal clinically important difference

The smallest change on a scale that a patient would actually notice as meaningful. A statistically significant result below this threshold is real but not clinically important, which is how the Cochrane review describes yoga's effect on pain and function.

Adverse event

Any unfavourable outcome occurring during or after an intervention, whether or not it was caused by it. In these trials the most common was increased back pain.

The training behind this article

Ludmila Voroncova holds 400 hours of Yoga Alliance training, 200 hours in Hatha and 200 in Ashtanga, and taught group classes and retreats over seven years. Breathing, mobility and controlled-load work from that practice are used in home programmes and graded return to activity. They are used as exercise, chosen and progressed for a given person, and the evidence above is why the claim stops there.