Dance, Balance and Falls: A Physical Outcome, and Why That Changes the Argument

Falls are the most testable outcome in the whole of arts in health, because a fall is a counted event rather than a score on a questionnaire, and this is the corner where dance evidence is usually said to be strongest. The strongest version of the claim that can be defended, though, is a claim about exercise rather than about dance, and the distance between those two sentences is what this page is for.

I run singing sessions rather than dance classes, so my experience here is oblique and I want to be honest about that. What I have watched at close range is what a care home’s falls policy does to a room. In one of the two homes where I work, a run of falls led to a review, and after it the seated dance class that used the lounge before me stopped having anybody stand up at all. Not because the dance artist changed her mind, and not because anybody decided standing was bad for people. It was that the risk of a fall during an activity sat with the home, the risk of decline from never standing sat nowhere in particular, and only one of those two risks had a form attached to it. The residents least able to stand safely were the residents who most needed to practise standing, and the system had no way of holding both facts at once.

Where a claim about effect appears below it carries a tier (Tier 1 supported, Tier 2 promising but limited, Tier 3 practice and anecdote), explained on reading arts in health research.

Why a falls outcome is different from everything else on this site

Because it does not depend on how anybody felt about the class, and almost every other outcome in arts in health does.

Nearly all of this field’s research runs into the same wall: the participant cannot be blinded. Somebody who has spent ten weeks in a singing group knows they were in a singing group. When that person then fills in a wellbeing questionnaire, several things are mixed into the answer besides the intervention: politeness, the wish for the group to continue, the general lift of having been asked, and a perfectly human reluctance to say that a kind facilitator wasted their Tuesdays. None of that is dishonesty and all of it moves a score.

A fall does not work that way. It either happened or it did not. It can be counted prospectively, by somebody who does not know which group a person was in, and the count is the same whatever anybody thought of the class. That makes falls one of very few outcomes in this field that survives the blinding problem, and it is the main reason this corner is the most promising place in dance to run a serious trial. It is also, incidentally, why the site’s one unambiguous Tier 1 finding is a physical outcome as well: gait velocity improving by 11.34 m/min (95% CI 8.40 to 14.28) from 9 trials and 268 participants at moderate quality after rhythmic auditory stimulation¹. Measurable things get measured. See rhythmic auditory stimulation.

What the defensible claim actually is

That structured strength and balance exercise reduces falls, and that a dance class can be a way of delivering structured strength and balance exercise.

Read that twice, because the second half is doing something the first half is not. The falls prevention literature is about exercise: progressive strength work, balance challenged under changing conditions, sustained over months rather than weeks, at enough intensity to be a load. A dance class that does those things is delivering that intervention in a more sociable wrapper. A dance class that does not do those things, and plenty do not, is not delivering it at all, whatever it is called.

So the honest sentence for a reader is this: if a class has you shifting weight, changing direction, standing on one leg, reaching outside your base, and doing it week after week for months, it is plausibly doing the thing the exercise evidence supports. If it is an hour of pleasant seated movement to music, it is doing something else, and that something else may still be worth having. Nothing about that argument is specific to dance, which is exactly the point. The claim that requires a dance specific trial is the claim that dance does something the equivalent exercise class would not, and that is a head to head comparison, not a comparison against nothing.

Why no figure appears here

No dance and falls effect size is locked in this site’s source document, so none is printed on this page.

That rule is the reason the site exists. Almost every claim it spends its time correcting is a figure that was forwarded rather than opened, and the fastest way to become part of that problem is to quote a promising looking review from a summary of it. So the number is absent and the gap is named.

What would change this page: a current systematic review of dance interventions with falls or a validated balance measure as an outcome, reporting a pooled estimate, a confidence interval, a certainty rating, a sample size and a date, in a document a reader can open. On this site a Tier 1 label requires those five things printed next to the claim, which is the bar the rhythmic auditory stimulation finding clears and nothing in dance yet does.

It is worth adding that a large pile of studies would not by itself do it. The most useful demonstration on the site is the 2021 Cochrane review of music interventions in cancer care, which pooled 81 trials and 5,576 participants and still rated its findings very low certainty², which is Tier 2. Volume of studies is the least informative number in an abstract. And the one dance review that has been traced to source here, on depression, came out at 3 studies, 147 participants, SMD -0.67 (95% CI -1.40 to 0.05), very low quality, with its authors declining to draw firm conclusions³. See dance for depression for what that one turned out to contain.

What a serious falls trial has to do, and why dance projects rarely can

Three requirements, and community dance provision generally cannot meet any of them.

Numbers. Falls are relatively infrequent events, so detecting a change in their rate needs a lot of participants. A study of thirty people over eight weeks will record a handful of falls in total, and any difference between arms will be noise. This is the requirement that rules out most of the sector immediately.

Duration. Follow up has to be months, and preferably a year. A programme that improves a balance score at week 8 and is abandoned at week 12 has not prevented anything. Falls prevention is an adherence problem as much as a physiology problem, and short trials measure the wrong end of it.

Prospective recording. Falls have to be recorded as they happen, usually in a diary or on a monthly card returned to the study, because asking somebody at the end of a year how many times they fell produces an undercount. People genuinely forget falls that did not hurt them, and the ones that did not hurt them are the ones that predict the ones that will.

Now put those against how community dance is funded: short grants, small cohorts, evaluation budgets that stretch to a wellbeing questionnaire at the start and the end. That is not a criticism of the organisations, which are spending their money on delivery, and it is the structural reason this literature looks the way it does. Who pays for arts on prescription covers the funding cycles behind it.

The adherence argument, which is real and is not an effect size

The best argument for dance in falls prevention is that people keep doing it, and that argument should be made as what it is.

Structured home exercise programmes are effective and widely under used, for the ordinary reason that doing leg strengthening exercises alone in a kitchen twice a week is dull. A class with music, company and somebody expecting you is a different proposition, and if it produces a year of weekly attendance where a leaflet produces six weeks, the dance version may deliver more actual exercise even if its content is no better.

That is a serious argument and it is not an effect size. It would be tested by comparing adherence between a dance delivered programme and a conventional one over a year, which is a different trial from any of the ones usually cited. Making the adherence case explicitly is more honest than dressing it up as a physiological claim, and it is also more persuasive to anybody who has ever tried to get somebody to do their exercises.

Seated classes, and the group most at risk

Seated dance cannot train standing balance, and saying so plainly is more use to a reader than implying otherwise.

Standing balance is trained by standing, by being slightly unstable and recovering, by reaching beyond your base of support. A seated class does not do that, and no amount of enthusiasm changes it. What seated work does do is keep joints moving, maintain strength in the legs and trunk, hold on to coordination and rhythm, and provide a reason to move at all for somebody who otherwise would not. Those are worth having and they are different claims.

The uncomfortable consequence is the one I described from the care home. The population at highest risk of falling is often the population least able to practise the thing that would reduce the risk, and institutional risk policies push in the same direction, because a fall during an activity generates paperwork and a slow decline does not. Seated and chair based dance covers the practicalities, and arts in care homes covers who provides this work and what is expected of them.

How large the literature is, and why that does not settle it

A great deal has been written, and the size of the writing is regularly used to imply a strength it does not have.

The WHO Europe scoping review covers over 900 publications, comprising 200 plus reviews and 700 plus individual studies, with the reviews between them covering over 3,000 studies⁴. Those counts nest inside one another rather than adding, and a scoping review maps a literature rather than pooling it, so it carries no tier and supports no claim about effect. The sector’s own reports and case studies, gathered in one place⁵, should be read the same way: as a description of what is happening, not as evidence that it works. That distinction is the subject of creative health policy and reports.

If you are unsteady and thinking about a class

Ask the clinician first, then ask the class.

Anybody who has fallen in the past year should be assessed rather than simply signed up, because falls have causes a dance class cannot touch: blood pressure that drops on standing, medication effects, poor vision, foot problems, and conditions that need treating in their own right. That assessment is the intervention with the best evidence behind it, and a class is an addition to it rather than a replacement for it.

Then the practical questions. What training does the leader have, and have they worked with older people or with your condition. What happens if somebody becomes unsteady or falls, and is anybody present first aid trained. Is there a seated version of everything. Is the floor even and the room uncluttered. Are they insured. And how long is the class funded for, because consistency over months is the thing most likely to matter and short grants are the norm. Finding a dance for health class goes through the rest of it, and dance and health covers why exercise, therapy and a social night out get conflated in the first place.

Nothing on this page is a reason to change, delay or decline any part of anybody’s treatment, and no dance class substitutes for a falls assessment.