Balance training and falls: the dose the prevention trials used

Exercise can make falls less common in older age, and the trials agree on that more than on most things. People who did balance and everyday-movement training fell about a quarter less often than people who did not, across dozens of studies and thousands of older adults.¹ Programs that added strength work on top did better still.

What the trials actually tested

Most of these trials sat inside one large Cochrane review of 108 studies and over 23,000 people, all living independently rather than in care.¹ Participants were split into an exercise group and a control group, then followed while researchers counted how many times each person fell.

The review sorted programs by what they actually did, not by their branding. "Balance and functional exercise" meant practicing the kind of movements a fall actually happens during. "Multiple types" meant balance work stacked with resistance training, usually with weights or resistance bands. Tai chi was tracked on its own.

A separate US government-funded review, drawing on 37 more recent trials and over 16,000 participants, found the same pattern using slightly different math.² Exercise lowered the rate of falls, lowered the odds any one person fell at all, and lowered injurious falls, though it did not clearly cut the risk of a serious injury once a fall happened.

That last distinction is worth sitting with. A fall that leads to a bruise and a fall that leads to a fracture are counted separately in these trials, and exercise's clearest win is in how often people go down at all, not necessarily in softening the landing when they do.

The size of the effect, by training type

The plain balance-and-function programs cut the rate of falls by 24%.¹ Programs that combined that balance work with resistance training did better, cutting the rate by 34%, roughly a third.¹ Tai chi, tracked separately across seven studies, cut the rate of falls by 19%, about a fifth.¹

How much each program type cut the rate of falls% fewer falls

Balance and functional exercise24%

Balance plus resistance training34%

Tai chi19%

Not every popular exercise made the list. The trials could not tell whether resistance training alone, walking programs, or dance changed how often people fell.¹ A later, dedicated review of dance found the same gap: the certainty of the evidence was rated very low, and the reviewers said outright that they remained uncertain whether dance reduces falls at all.⁴ Home resistance training done without supervision improved leg strength and one balance measure, but showed no measurable change in falls.⁵

That is not a small caveat. It means the ingredient that shows up across the effective programs is balance practice itself, not exercise or strength in general. A person could be getting stronger, walking further, or dancing weekly and still not be doing the specific thing the trials found protective. The exercises that worked practiced the skill of catching oneself, not just the muscles involved in it.

The dose that showed up in the effective trials

A separate review, built to compare preferences and pool findings by intervention type, looked at which fall-prevention programs had at least moderate certainty of benefit. The pattern among the effective exercise-based programs was consistent: supervised sessions, more than twice a week, kept up for over three months.³

That detail matters more than the headline percentage. The reviewers singled out long, supervised programs, and this evidence does not show what shorter or unsupervised ones do. The reduction in falls belongs to the sustained, coached version of this training, not to a few sessions of balance drills.

Duration also cuts the other way once training ends. In care-facility trials that followed people after their exercise program stopped, the benefit vanished. The rate of falls in people who had done the exercise, once they were no longer doing it, was statistically identical to people who never did it at all.⁶ Balance training does not appear to bank an effect for later. It has to keep happening.

Where the evidence is thinner

Most of this evidence comes from people living independently in the community, not from hospital wards or nursing homes, where different interventions and different results apply. It also comes overwhelmingly from supervised programs run by a trainer or physiotherapist, so it says less about the same exercises done alone from a video or an app.

The comparison between tai chi and combined balance-and-resistance training is also softer than the raw percentages suggest. Tai chi's 19% reduction came from lower-certainty evidence than the 24% and 34% figures for balance and multi-type training.¹ The numbers look close, but they are not resting on equally solid ground.

And almost none of these trials were built to isolate exactly which balance moves, or how many minutes of them, did the work. The label "balance and functional exercise" covers a range of specific programs, and the pooled result describes that range, not one recipe.

The people delivering the training differ across studies too. Some programs were run by physiotherapists, others by trained group instructors, and the reviews do not cleanly separate what a professional's involvement adds beyond the exercises themselves. Anyone comparing a clinic-based program to a community class is comparing across that gap without a clear answer for how much it matters.

What remains unknown

The biggest open question is what happens beyond a year or two, since most trials do not run that long. The second is what an unsupervised, home-only version of this training would do at the same dose, since almost all the strong evidence comes from coached sessions. The third is why resistance training and walking, both plausible tools for a stronger, steadier body, did not show the same fall reduction as targeted balance work.¹

What holds up is narrower than "exercise helps." It is: balance-specific movement, taught and supervised, kept up for months, cuts how often people fall by about a quarter, and adding strength training on top does better than that.¹