Yoga For Seniors: What The SAGE Trial Did Not Test

Yoga for seniors: the short answer

Yoga for seniors got a bad headline in 2025. A large Australian trial reported more falls in its yoga group, and the world was told that a popular kind of yoga was linked to falls in older people. That trial is called the SAGE study. The University of Sydney ran it. The Lancet Healthy Longevity published it in September 2025. People now quote it whenever fall prevention comes up. This article looks at what the SAGE trial actually tested, using the trial’s own published record.

The SAGE trial in short form

• The programme was labelled Iyengar yoga. It was not taught as Iyengar yoga. The poses, the order, the timing and the cues were written by university researchers who are not certified in the method. The teachers were handed a manual and asked to deliver it. Of the four teachers the trial named, no more than two held current Certified Iyengar Yoga Teacher credentials.
• The order was reversed. Standing poses come first in this method, because they build the leg strength and alertness that make everything else safe. The trial began each class with about 15 minutes of floor work, then asked people to stand on one leg in Week 1.
• The trial cannot say who taught whom. Four or more teachers from different yoga traditions taught 700 people over 80 classes, and the published record does not say which student had which teacher, or for how many classes.
• The numbers are thinner than the headline. Three tests were run on falls. One was just significant, one was significant, and one was not significant at all. Falls that caused injury did not increase.
• The people most at risk did better. Among the least physically active participants, the group falls prevention is actually aimed at, falls went down by 39%. That did not reach the summary.
• The published report leaves out things the agreed reporting standards ask for. Two international standards ask for who taught what, to whom, and how much of it was delivered as planned. The published paper does not supply that.
• The good news is the boring news. Five in-person studies of this method, run in rooms with teachers and props, found better balance, better walking and stronger legs. None of them found more falls.

Flametree Yoga Studio teaches this way across Darwin and LIVE online. Our companion article sets out the falls and yoga studies in detail.

Why does any of this matter?

Because falls kill more Australians than car crashes do, and a wrong signal about what helps is expensive.

The Australian Institute of Health and Welfare reports that 6,698 Australians died from falls in 2022-23. That is 43% of all injury deaths. There were about 248,000 fall-related hospitalisations in 2023-24. Almost two-thirds of those hospitalisations were women. By comparison, the Bureau of Infrastructure and Transport Research Economics recorded 1,300 road deaths in 2024.

So falls are roughly five times the road toll. Anything that genuinely reduces them is a serious public health measure. Anything wrongly blamed for causing them deserves a careful answer.

That is the reason for this article. A headline told several million people that a named yoga method makes older people fall over. If that is true, it should be said loudly. If it is not, the record should be corrected.

What did the SAGE trial actually test?

It tested an exercise programme that academics designed and that ran mostly over Zoom. A mixed group of yoga teachers delivered it to older adults who had almost no props and who often could not be seen. So it did not test yoga for seniors as this method actually teaches it.

The trial is called SAGE. Oliveira, Sherrington, Lord, Tiedemann and colleagues published it in Lancet Healthy Longevity in 2025, and it randomised 700 Australians aged 60 and over. The trial offered the yoga group 80 supervised classes across 12 months. Meanwhile the control group got two one-hour workshops of seated breathing and stretching. The authors also state that most classes ran online because of COVID.

Those are the study’s own published facts, and they matter more than any opinion about them. Above all, a programme written by people not trained in a method, and delivered in conditions the method does not accept, is not a test of that method. Instead, it is a test of that programme.

Why does the order of poses matter in yoga for seniors?

Because in this method balance is not something you practise. Rather, it appears once the legs, hips, spine and breath are ready.

A student stands in a wide, stable pose with the back foot turned and the front leg straight. The legs work, the hips open, the spine lengthens, the breathing settles. Weeks later, when that foundation is there, standing on one leg is a small step rather than a gamble. This is why standing poses come first, and why a teacher does not hurry them.

The SAGE programme did the opposite. Its published appendix shows single-leg balance in the very first block of classes. From Week 5 it adds a balance exercise that comes from physiotherapy rather than from yoga, and repeats it three times a class. In addition, each class opened with roughly 15 minutes of seated and lying work before anyone stood up.

What this shows. Standing on one leg is the last step, not the first. Build the legs first, then ask them to balance. Ask first, and the person wobbles. In a falls study that is the last thing you want.

There is a further detail that any teacher will recognise. The trial’s manual told teachers to give no more than three cues per pose, because more might overwhelm people. Precise instruction is not a decoration in this method. It is the method. Capping it changes what is being taught.

What this shows. One pose, one chair, one precise instruction. Press the outer back heel down. That single cue is what turns a shape into a pose, because it sends the work back down into the leg. A teacher limited to three cues for the whole pose cannot teach this way. Christine Lalor at the Lost City, Litchfield.

The things that were different, in one list

Here is the whole case in one place. Each item below comes from the trial’s own published material, or from the standards this method sets for teaching yoga for seniors. Furthermore, no single item carries the argument on its own. Together they describe something else entirely.

1. People not certified in the method wrote the curriculum. The researchers set the poses, the order, the timing and the cue limit. The teachers simply received a manual.
2. Most of the teachers were not certified in the method either. Of the four teachers the trial named, no more than two held current Certified Iyengar Yoga Teacher credentials. The other two came from different traditions. Certification here is not a formality. It takes years, and the professional association awards it after assessment. A class taught by somebody who does not hold it is not a class in this method, whatever the manual on the cover says.
3. The order of the class was reversed. Floor work first, standing poses after. This method does it the other way, deliberately, and for older beginners especially.
4. Single leg balance arrived in Week 1. Before the legs had been built. A balance exercise borrowed from physiotherapy was then repeated three times a class from Week 5.
5. Some pose names are not from this method at all. The published appendix uses names that belong to other styles or to exercise science.
6. Instruction was capped at three cues per pose. Detailed correction is not a flourish in this method. It is the method.
7. Props were largely absent. Students used household objects. The wall, chair, blocks and belts that make a pose fit a body were mostly not there.
8. Most of it happened online, in poor conditions. Small spaces, cameras often off, phone-sized screens. The teacher frequently could not see the student.
9. The label moved around. The 2020 protocol called it a Hatha and Iyengar style. The 2025 paper called it Iyengar yoga-based. The press release and broadcast interviews called it Iyengar yoga.

Any one of these would be worth noting on its own. Taken together, however, they describe a different activity.

Who taught whom, and why the trial cannot say

The published record never says which teacher taught which student, or for how many of the 80 classes. Consequently, nobody can trace any part of the result to any one style of teaching yoga for seniors.

There is a plainer problem before you even reach the record keeping. Of the four teachers the trial named, no more than two held current Certified Iyengar Yoga Teacher credentials. The other two came from other traditions. So at least two of the four were not certified in the method at all. And because the record does not say who taught whom, there is no way to know how much of the teaching actually came from a certified teacher. It could have been most of it. It could have been very little.

That is not a technicality about paperwork. Certification in this method takes years. The professional association awards it after assessment, and it carries ongoing requirements afterwards. It is the thing that makes the name mean something.

Imagine a school comparing two ways of teaching mathematics. It uses four teachers, and no more than two of them are maths teachers. It keeps no record of which child sat with which teacher. At the end of the year it announces that the new maths method does not work. Any parent would ask the obvious question, and the school would have no answer.

What this shows. Four or more teachers taught 700 people, and of the four the trial named, no more than two were certified in the method. Nobody wrote down who taught whom. So the study cannot say whose teaching produced the result, and it cannot say how much of the teaching came from a certified Iyengar yoga teacher at all.

This also closes off the softer version of the claim. If the mix of teaching each student received was never recorded, then “partly Iyengar yoga” cannot be supported either. The honest description of what was tested is a mixed-style, mixed-teacher programme written by researchers.

It is worth adding that the trial’s own 2020 protocol planned for this. It said randomisation would be balanced by studio and instructor, which is the standard way of stopping teacher differences from contaminating a result. That sentence does not appear in the 2025 paper, and no replacement is described.

The 2013 pilot and the 2025 trial were not the same programme

This is the part that surprises people most. The same research group ran a smaller study of yoga for seniors in 2013, and it got the opposite result.

The 2013 pilot improved balance, chair stands and walking speed. The 2025 trial reported more falls. The two are often described as if the second were simply a bigger version of the first. They are not. Almost every feature that made the first one Iyengar yoga was changed.

Feature	2013 pilot, 54 people	2025 SAGE trial, 700 people
Who designed the classes	A certified teacher, with the researchers	University researchers, none certified in the method
Who taught the classes	Certified teachers, in the room	Four or more teachers, no more than two certified in the method
Order of the class	Standing poses first	About 15 minutes of floor work first
Pose names used	Standard names from the method	Includes names not used in the method at all
Instruction	Normal detailed correction	Capped at three cues per pose
Standing on one leg	After the foundation was built	Week 1
Where it happened	In a studio, in person	Mostly online, small spaces, few props
Result	Balance and mobility improved	33% higher fall rate

Read that table again. Six things changed, and then the result changed. Therefore this is not a mystery that needs an explanation about yoga. It is simply a different programme.

What do the study’s own numbers show?

They show a result that is real but fragile. They also show several findings about yoga for seniors that never reached the headline. In other words, the summary and the data tell different stories.

The researchers ran three separate tests on falls from the same data. The rate of falls gave p=0.044, which sits just inside the usual threshold. The number of people who fell gave p=0.020. However, the time until a person’s first fall gave p=0.096, which is not significant. If yoga genuinely made people unsteady, the time to a first fall is exactly where you would expect to see it.

What this shows. The study produced five findings about falls. One was alarming, and it became the headline. Four were reassuring, and they did not. All five come from the same study and the same people.

Two further findings deserve more attention than they received. Falls that caused injury did not increase, and the yoga group’s rate was slightly lower. And among the least physically active participants, falls fell by 39%. That group is the one falls prevention programmes are designed for. The finding did not quite reach significance, at p=0.05, and we are not going to overstate it. But it belongs in any fair summary, and it was not in one.

One more thing is worth knowing. The control group had four times the rate of missing fall records, and contributed about 19% fewer person-years of follow-up. In a study that counts falls per person-year, that difference pushes in one direction.

Confidence went up. Steadiness did not.

What happens when belief outruns capacity

There is one finding in the trial that explains a great deal, and it has had almost no attention.

The yoga group’s confidence in their own balance improved significantly, at p=0.014. Their planned activity went up by about an hour a week, at p<0.0001. Both of those results are statistically stronger than the headline finding about falls.

Think about what that combination does. People felt steadier. So they did more. But the underlying capacity had not been built, because the programme never built it. More exposure plus unchanged capacity is a recipe for more stumbles, and stumbles that do not cause injury.

That is exactly the pattern the data shows. Falls up slightly. Injurious falls not up at all. Confidence up. Activity up.

This is not a reason to think yoga is dangerous. It is a reason to think this programme gave people belief without giving them the legs to go with it. Teaching the foundation is precisely what closes that gap.

What do the reporting rules require?

There are agreed international standards for describing what a trial actually did. They exist because a result nobody can reproduce is not much use. Moreover, for yoga for seniors these standards ask specifically about the style taught and the teacher’s training in it.

The TIDieR checklist, published in the BMJ in 2014, is the general standard for describing any health intervention. CLARIFY, published in BMJ Open in 2021, is the standard written specifically for yoga research. Between them they ask for the qualifications of whoever delivered the intervention, where it took place, how much each person actually received, how far it was delivered as planned, and for yoga specifically the style and the instructor’s training in that style.

The published SAGE reporting does not provide several of those items. It does not say how many teachers there were in total, or who taught whom, or how much each participant received from each teacher. It reports no independent check that teachers delivered the same thing, resting instead on a statement that they followed a written protocol.

This is a separate point from everything else in this article, and it stands on its own. It is also the kind of point that journals act on, because it is about reporting rather than about yoga.

Can yoga for seniors be taught properly online?

Yes, and certified teachers do it every day. However, it requires conditions that the trial did not have.

Online teaching in this method works when the teacher can actually see the student, which means cameras on and positioned so the whole body is visible, on a screen big enough to be useful. It works when the student has the ordinary props, a wall to work against, and enough floor space to open the arms. It works when the teacher can name a correction precisely and the student knows what the words mean.

On the account of a teacher involved in the trial, students practised in spaces as small as a cupboard, used household objects instead of props, and frequently kept their cameras off. That is not an argument against online yoga for seniors. Rather, it describes online teaching stripped of the conditions it needs.

Flametree runs LIVE online classes alongside its Darwin studio classes, with cameras on and students in the picture. It is the same teaching, delivered a different way.

What this shows. A student at home in a live online class, with her teacher on the screen in front of her. She is in Downward Facing Dog with her head resting on a block. Her camera is on and her whole body is in the frame, so the teacher can see what her legs are actually doing and correct her by name. She has room to move and props within reach. That is what online teaching looks like in this method when the conditions are right. It is a long way from practising in a cupboard with the camera off.

Two different ideas of balance in yoga for seniors

Underneath the argument about labels sits a genuine difference of view. It is worth setting out plainly, because both sides are reasonable and they are not the same.

Conventional balance training treats balance as a skill. You find the edge of a person’s steadiness, you work at that edge, and you make it harder over time. Importantly, this approach has good evidence behind it, and we are not disputing that.

This method treats balance as a by-product. You build the legs, open the hips, lengthen the spine and steady the breath. Steadiness follows. Nobody drills wobbling.

What this shows. Warrior 2 taught at the wall. The hand on the wall is not there to stop a fall. It is there so the legs can work hard while the student stays steady enough to listen. Build the legs first and steadiness follows. This is the ordinary middle step in yoga for seniors. It is what makes the one-legged poses possible later on.

	Iyengar yoga	Conventional balance training
What is trained	Strength, alignment, breath, attention	Balance itself, directly
Order	Foundation first, challenge later	Graded challenge from the start
How long a position is held	30 seconds to several minutes	Short repetitions
Props	Wall, chair, blocks and belts, routinely	Usually none, sometimes a chair nearby
Who teaches	A teacher certified in the method over years	A physiotherapist, exercise physiologist or trainer
Correction	Individual, in the moment, body by body	Standard instructions, standard progression
Balance is	A result	A target

Now the point. The SAGE programme was built on the right-hand column and given the name from the left. That is why its own authors noted that their programme differed from most successful falls programmes. They were correct. But the outcome was then attributed to the name rather than to the design.

What do the props in yoga for seniors actually do?

They let a person hold a pose correctly today, in the body they have today, rather than approximately. In yoga for seniors this is the whole game.

A hand to the wall in a standing pose is not a concession. It tells the student where upright is, so the legs can work without the fear of tipping. A chair under the lower hand in a triangle pose lets a stiff 75-year-old make the same shape, with the same effect on the legs and hips, as someone half their age makes without one. A block under a hand does the same thing at a smaller scale.

What this shows. Mountain Pose, the pose everything else is built on, with one hand resting on a chair. Nothing here is watered down. The chair simply lets the legs and spine do their work without the student bracing against a fear of tipping. Christine Lalor at the Lost City, Litchfield.

This is the part that is hardest to reproduce without training, and it is the part the trial had least of. As a result, a programme given to people with no props is not a gentler version of this method. It is a different activity altogether.

What does the wider evidence say about yoga for seniors?

It is consistently positive, and the 2025 trial is the exception rather than the rule.

Tiedemann and colleagues ran a 12-week randomised pilot of 54 community-dwelling adults, mean age 68. The Journals of Gerontology Series A published it in 2013. Teachers ran it in person, and a certified teacher helped design the curriculum. The yoga group improved on one-leg standing time, chair stands, walking speed and closed-eye balance. No serious adverse events occurred. Notably, that pilot shares an author with the 2025 trial.

 

The other four studies

DiBenedetto and colleagues, in Archives of Physical Medicine and Rehabilitation in 2005, studied 23 older adults over eight weeks and found longer stride and greater hip extension, both of which are linked to fall risk. Schmid and colleagues, in the same journal in 2010, studied 14 older adults with a fear of falling and found balance and confidence improved. Ni and colleagues in 2014 compared yoga, tai chi and standard balance training in 39 older people with a history of falls, and found all three improved balance with no clear winner. Zhu and colleagues, in the Journal of Sports Medicine and Physical Fitness in 2021, randomised 48 adults aged 60 and over and found gains in strength, one-leg balance and confidence.

These are small studies, and we will say so plainly. They measured the things that predict falls rather than falls themselves. Even so, they all point one way. Teachers ran every one of them in a room with the students, and none of them reported an increase in falls.

 

What runs against us

It is also fair to record the part that does not help our case. For preventing falls specifically, tai chi has the larger and more consistent evidence base, and the major reviews name it as effective. Yoga is not yet established in the same way. And no form of exercise replaces an eyesight check, a medication review with your doctor, or getting rid of the loose rugs.

Why teachers say a pose is not an exercise

This section is for readers who want the reasoning from inside the tradition. It is also, we think, where the misunderstanding starts.

B.K.S. Iyengar was explicit that an asana is not a physical exercise that happens to look like one. In Light on Yoga, published in 1966, he set out the difference. An exercise repeats a movement to produce an effect on muscle. A pose is held, and while it is held the practitioner works on it with attention, adjusting the placement of the foot, the action of the leg, the length of the spine, the evenness of the breath.

Patanjali’s definition, which Iyengar translated and taught from, is two words: steady and comfortable. A pose is correct when both are true at once. Not braced. Not collapsed.

 

Why this matters for a prop

That definition is not mysticism. Instead, it has a directly practical consequence for an older student. If a pose must be both steady and comfortable, then a person who cannot yet achieve both gets a prop, so that they can. The wall, the chair and the block exist to make the definition reachable. They are not a lesser version of the pose.

This also explains why the cue cap matters so much. If the work of a pose is the attention brought to it, then the instructions are not commentary on the exercise. They are the exercise. A teacher restricted to three cues is being asked to teach something else.

Geeta Iyengar made a related point about older practitioners in Yoga: A Gem for Women. The practice regenerates energy at exactly the stage of life when the body’s own recovery is slowing. That is an argument for teaching the method properly to older people, not for watering it down.

What good yoga for seniors looks like

First, look for a teacher who holds certification in a method, not just a few weeks of training. Certification in this tradition takes years, and an association awards it after assessment. Nobody simply claims it. The term itself is a protected certification mark, which is why accuracy about who holds it matters.

Look for props in the room. Walls, chairs, blocks, belts and bolsters mean the teacher intends to make the pose fit you.

Look for standing poses near the start, and for a teacher who is in no hurry to get you onto one leg.

What this shows. This is where the work goes, eventually. Warrior 3 is a full single leg balance, and it is taught with a chair within reach. The chair is not there because the student is failing. It is there so the standing leg can be asked to work properly instead of gripping for dear life. This is the pose the SAGE programme introduced in Week 21. The foundation for it starts on day one.

Look for someone watching you. In a studio that means a teacher moving around the room. Online it means your camera on and you in the picture.

Flametree teaches at Woolner near Parap, Coolalinga, Casuarina, Palmerston and the Darwin CBD, and LIVE online across Australia. If you want to start, the half price 10 class pass with 2 classes free covers Darwin and online.

Yoga for seniors and the SAGE trial: questions people ask

Was the SAGE trial really testing Iyengar yoga?

In our view as certified teachers of the method, no. The sequence was reversed, the cues were capped, the props were largely absent, and the curriculum was written by people who are not certified in it. Of the four teachers the trial named, no more than two held current Certified Iyengar Yoga Teacher credentials. Because the record does not say who taught whom, there is no way to know how much of the teaching came from a certified teacher.

The trial’s own published record also does not say which student was taught by which teacher, so no part of the result can be attributed to any one method in any proportion.

Should older people avoid yoga because of this study?

We do not think the study supports that, and the study itself did not find that falls causing injury increased.

The wider evidence from classes taught in person points the other way. The sensible reading is that how yoga is taught matters enormously, which is the opposite of a reason to avoid it.

Can Iyengar yoga be taught well online?

Yes, when the conditions are right: cameras on and positioned so the teacher can see you, adequate space, a wall, ordinary props, and a teacher who can correct you precisely by name.

The trial had almost none of those conditions, which is a fact about that trial rather than about online teaching.

What makes someone a certified Iyengar yoga teacher?

A formal assessment process run by the professional association, taking years rather than weeks, with ongoing requirements after it.

It is not a self-description. If you are unsure, ask a teacher directly whether they hold current certification, and in what.

Is chair yoga for seniors the same as Iyengar yoga?

No. Chair yoga usually means practising while seated. In this method a chair is a prop used to support a standing pose, which is a different thing entirely.

Sitting down does not train standing up. A chair is most useful as something to hold while you are on your feet.

Why do Iyengar teachers use walls, chairs and blocks?

To make the correct shape available to the body in front of them, rather than a rough approximation of it.

A prop is not a sign that a student is struggling. It is how a pose is made to fit a person, and it is how older students get the full benefit of a pose safely.

What should I look for in a yoga class if I am over 60?

A certified teacher, props in the room, standing poses taught early and carefully, and someone actually watching you work.

Ask how long the teacher has taught older beginners. It is a fair question and a good teacher will welcome it.

What was the SAGE study and who ran it?

The University of Sydney ran the SAGE study. The Lancet Healthy Longevity published it in September 2025. Oliveira, Sherrington, Lord, Tiedemann and colleagues randomised 700 Australians aged 60 and over. The yoga group got 80 supervised classes across 12 months. Most of those classes ran online because of COVID. The control group got two one-hour workshops. The research team wrote the programme. Certified Iyengar yoga teachers did not. Nor did they teach all of it. Of the four teachers the trial named, no more than two held current Certified Iyengar Yoga Teacher credentials. That is the heart of the argument on this page.

Does yoga work for fall prevention in older adults?

Most of the research on yoga for fall prevention is encouraging. Five earlier studies found gains in one-leg standing time, chair stands, walking speed, stride length and confidence. None of them found serious harm. The 2025 SAGE trial is the odd one out. Its programme differed from taught Iyengar yoga in order, in timing, in props, in cueing and in who wrote it. So look at how a class is taught. Do not go by the label on the door.

Where can I find yoga for seniors near me in Darwin or online?

Flametree Yoga Studio teaches at Woolner near Parap, Coolalinga, Casuarina, Palmerston and the Darwin CBD, plus LIVE online across Australia. See the timetable and book a class here.

More on this from Flametree

This article is the detailed companion to our post on falls and yoga studies, which covers the research on balance and falls more broadly.

Strong bones and steady legs are the same project. Flametree has also written on yoga for bones and osteoporosis and on how a yoga pose builds muscle and bone at the same time. There is also a page on activities for seniors in Darwin.

Sources

1. Oliveira JS, Sherrington C, Lord SR, Tiedemann A, et al. The effect of an Iyengar yoga-based exercise programme versus a seated yoga relaxation programme on falls in people aged 60 years and older (SAGE). Lancet Healthy Longevity 2025;6(9):100749. DOI 10.1016/j.lanhl.2025.100749. 700 participants.
2. Tiedemann A, et al. A 12-week Iyengar yoga randomised pilot in community-dwelling older adults. Journals of Gerontology Series A 2013;68(9):1068. 54 participants.
3. DiBenedetto M, et al. Effect of a gentle Iyengar yoga program on gait in the elderly. Archives of Physical Medicine and Rehabilitation 2005. 23 participants.
4. Schmid AA, et al. Effect of a 12-week yoga intervention on fear of falling and balance in older adults. Archives of Physical Medicine and Rehabilitation 2010. 14 participants.
5. Ni M, et al. Comparison of yoga, tai chi and conventional balance training in older adults with a history of falls. 2014. 39 participants.
6. Zhu R, et al. Iyengar yoga versus tai chi in adults aged 60 and over. Journal of Sports Medicine and Physical Fitness 2021. 48 participants.
7. Australian Institute of Health and Welfare. Falls in older Australians. aihw.gov.au.

Written by Christine Lalor, Certified Senior Iyengar Yoga Teacher and certified Yoga Therapist, Flametree Yoga Studio, Darwin. This article is general information about a yoga method and the published research record. It is not medical advice, and it is not advice about your own fall risk. Speak to your doctor about that.