Yoga for balance and preventing falls

Yoga for balance has good evidence behind the thing it measures most easily, and thin, contested evidence behind the thing you actually care about. The measurable gain is balance and mobility. The uncertain part is whether that gain turns into fewer falls. It is worth having both halves before you start, because most pages on this subject give you only the first.

The strongest single summary is the 2016 systematic review and meta-analysis in Age and Ageing, which pooled six trials and 307 people aged 60 and over. It found a small effect on balance performance, Hedges' g 0.40 with a confidence interval of 0.15 to 0.65, and a medium effect on physical mobility, g 0.50 across three trials. The reviewers were explicit that they could not say whether those improvements prevent falls, and they asked for the trials that would settle it.

Those trials have started arriving, and the largest of them did not go the way anybody expected. The SAGE trial, published in The Lancet Healthy Longevity in 2025, randomised 700 community dwelling people aged 60 and over to either an Iyengar yoga based exercise programme, 80 supervised hourly classes twice a week for a year, or to a seated yoga relaxation programme. The exercise group fell more: 0.87 falls per person year against 0.64, an incidence rate ratio of 1.33 with a confidence interval of 1.01 to 1.75. The authors concluded that that programme should not be recommended for fall prevention in its current form. There was no difference in serious or injurious falls, and the exercise group did gain about an hour a week of planned activity and reported better balance confidence.

You will also see a figure of a 48 percent reduction in falls from standing yoga quoted widely, including in a 2023 Innovation in Aging abstract on seated yoga and falls prevention. It comes from a small study in rural older adults, and it has been repeated far more often than it has been replicated. A 2019 systematic review of fall prevention outcomes in adults aged 50 and over, covering 11 studies, said plainly that more research was needed. A 2025 review of six trials in 170 older women found the balance effects inconsistent, with clear gains in only some groups.

So here is the honest position, and it is the one I teach from. Yoga for balance reliably improves how well you stand, how confidently you move and how strong your legs are, and every one of those is a known risk factor for falling. Whether a given class reduces your falls depends enormously on what is in the class, and one large trial of a fairly ambitious programme found the answer can be worse than nothing. It may support your balance alongside proper medical care. Practising in reach of a wall, and not competing with the person next to you, is not timidity here. It is the difference between the two results above.

Put the avoid list before the pose list, because on this page it is the more important half. A fall while learning to balance is a real risk in an older body, and the wrist and the hip do not care that the intention was preventive.

• Sirsasana and sarvangasana. Both put body weight through a neck that at 70 is stiffer than it was at 40, and the way out of both is an uncontrolled movement under load. In the published case reports of serious yoga injuries, headstand takes a disproportionate share. There is nothing in either pose that a person working on balance needs.
• Halasana, plough. Deep loaded spinal flexion. If bone density has not been measured, loaded forward bending is out, and after 65 in India it usually has not been measured. See the back pain page for the compression fracture case reports that make this a firm line rather than a caution.
• Natarajasana, and any one legged pose that also asks for a deep backbend or a grip behind you. The pose has two difficulties stacked on top of each other, and the falling happens in the transition, not the hold. Balance training does not need the shape to be interesting.
• Malasana, the deep squat. Not because squatting is bad, but because coming up out of a full squat on a tired knee is where students in my classes have actually lost their footing. If you want the hip opening, sit on two folded blankets. See yoga for knee pain for the rest of that argument.
• Eyes closed balance work, for the first two months. Closing the eyes removes the one input that is still reliable. It is a legitimate progression later. It is not week one.
• Balancing away from a wall, on a mat that slides, in socks. Three quarters of the falls I have seen in a hall in forty years had one of those three in them.

And the honest reading of the SAGE trial belongs here: an ambitious standing practice, twice a week for a year, taught by trained teachers, still produced more falls than sitting quietly did. That does not mean standing work is wrong. It means the pace at which you add it, and what you keep within reach, is not a detail.

The whole page in one table. Start with three from the practise column, not eight, and read the modification as the instruction rather than the concession.

Nothing in the avoid column is banned forever, except the first three. Those three have nothing to offer a person whose reason for practising is balance.

Balance is not one skill. It is at least four, and a practice that trains only the showy one leaves you exposed.

• Ankle strategy, which is the small continuous correcting that happens under a standing foot. Tadasana trains it, and only if you stand still long enough to feel it working. Six breaths is long enough. Two is not.
• Single leg standing, which is what most people mean by balance. Vrikshasana, and the honest progression is toes down for a fortnight, then ankle on the calf, then foot on the inner thigh, and never the arms overhead until the foot is quiet.
• Leg strength, which is the part that actually catches you. Virabhadrasana 2 and utkatasana to a shallow depth. A person who can stand up from a low chair five times without hands rarely goes down from a stumble. A person who cannot usually does.
• Weight shift and rotation, which is the mechanism in the real falls. Trikonasana for the sideways shift, garudasana arms for crossing the midline. Falls happen when you turn, reach or step sideways, almost never while standing on one leg on purpose.

Then the two that are not about the legs at all. Savasana, eight minutes, because fear of falling makes people shuffle, shorten their stride and look at the floor, and every one of those makes a fall more likely. And anulom vilom, eight rounds, for the same reason: a person who can lengthen an exhale at will has one thing to do at the top of a flight of stairs other than tense up.

Add one practice a week, in that order. Three practices held properly beat eight rushed, and this is the page on the site where that is least negotiable. If you want a ready made starting structure, the seven day beginner plan is built at this pace, and the ten minute morning sequence is the one my older students actually keep doing.

Balance that has changed recently is a symptom, and yoga is not the first thing to do about it. See a doctor before you start if any of this is true.

• The balance got worse over weeks or months rather than years.
• There is dizziness, spinning, or a feeling of the room moving, particularly when you turn your head or sit up in bed.
• You have fallen twice or more in the past year, or fallen once and could not get up. That is a formal threshold for a falls assessment, not a judgement.
• There is numbness, pins and needles or a loss of feeling in the feet. Yoga does not fix that and it changes what is safe.
• Your eyesight has changed, or you are wearing new varifocal lenses. Varifocals and stairs are a well known combination.
• You are taking four or more prescribed medicines, or any drug for blood pressure, sleep or mood. A medication review often does more for falls than an exercise class does. See yoga for high blood pressure for the postural drop that matters here.
• There is known osteoporosis, or a previous fracture from a minor fall.

None of that rules out practice. It changes the order. Get the assessment, then bring its findings to a class, and the class gets built around them.

Week one to two: nothing improves and something feels worse, because standing still on two feet with attention is harder than it sounds and you will notice a wobble you had been unconsciously managing for years. That noticing is the first gain, though it does not feel like one.

Week three to six: single leg standing time roughly doubles for most people, from a wobbly five seconds to a steady ten or fifteen. Leg strength lags behind and improves in week five or six rather than week three.

Week eight to twelve: this is where the trial evidence sits, and where the confidence changes. Students stop putting a hand out for the wall before they need it. In the twelve week studies, reduced fear of falling shows up about here, and fear is a genuine outcome rather than a soft one.

Beyond three months: strength and confidence hold as long as the practice holds, and the SAGE result is a caution against thinking that more is automatically better. Twice a week, forever, at a depth you can control, is the shape that ages well. A heroic eight weeks followed by nothing is worse than useless.

Write these in a notebook on the first day and again every four weeks. They take five minutes and they are more use than any sensation.

1. Single leg standing time, eyes open, next to a wall, best of three on each side. Under ten seconds is worth mentioning to a doctor. Note both sides, because a large difference between them is itself information.
2. Sit to stands from a dining chair in thirty seconds, arms folded across the chest. If you need the hands, use them and record that you did.
3. Stumbles and near misses this month, with where each one happened. Bathroom, kerb, stairs, dark corridor. The pattern tells you more than the count, and half of what it tells you is about the house rather than the body.
4. How afraid you are of falling, nought to ten, answered quickly. This is the number that moves first, and the one that changes whether you leave the house.

Four numbers. Not a diary. The students who keep this up are the ones who can tell in month three that something is working, and that is the month most people stop.

Balance is the one thing on this site I would not ask a beginner over 65 to learn alone from a screen without at least someone else in the house. Not because the poses are hard, but because the useful correction is invisible from inside: you cannot see that you have gripped the standing foot, dropped the opposite hip, or started holding your breath the moment the foot leaves the floor, and all three make the wobble worse rather than better.

The classes here are traditional yoga taught slowly by one teacher across about forty years, at a residents' association hall, to students of all ages. If you have been searching for yoga for balance or fall prevention near you, the in person classes run in north Bangalore, in the hall at MSR Nagar, with students coming from Mathikere, Malleswaram and Yeshwanthpur, and a chair for anybody who wants one. If getting to a hall is itself the risk, the live online classes suit this better than you would expect: the camera set side on, in your own room, next to your own furniture, shows a standing leg more clearly than a teacher at the front of a hall can, and you are practising in the place where you will actually be walking about. The first class is free. Come and stand on one leg badly in front of somebody, which is how everybody starts.