Balance training

Standing upright is an active process. The body sways continuously by small amounts, and the brainstem and cerebellum correct that sway many times a second by combining information from three sensory systems and sending corrections to the muscles.

The first system is vision, which gives the position of the horizon and the movement of the surroundings. The second is the vestibular system of the inner ear: three fluid-filled semicircular canals detect head rotation, and two small organs called otoliths detect gravity and straight-line acceleration. The third is proprioception (the sense of body position), which comes from stretch sensors in muscles, receptors in tendons and joints, and pressure sensors in the soles of the feet.

When one input is unreliable, the brain relies more on the others. This is why balance is harder in the dark, on soft ground or with the eyes closed, and why many balance exercises remove one input on purpose.

Corrections follow a rough order. Small sway is handled at the ankles, larger sway by bending at the hips and trunk, and when the centre of mass moves beyond the base of support, the only correction left is a quick step. A recovery step depends on fast, forceful leg muscles, which links balance to strength training and muscle mass.

Each part of the balance system changes with age. The inner ear loses some of its sensory hair cells, sensation in the feet can decline (much more so with diabetes or nerve damage), and vision is affected by conditions such as cataracts. Reaction time lengthens, so corrections arrive later. On the output side, muscle power (the ability to produce force quickly) tends to decline, and loss of muscle with age, called sarcopenia, reduces the strength available for a recovery step. Sedating medicines and alcohol affect the whole system at once.

The authors of the Brazilian one-leg stance study describe balance as declining quickly after the mid-50s [1]. The consequence most studied is falling: at least one-third of community-dwelling people over 65 fall each year [2].

Balance is partly a learned skill. The weighting of sensory inputs and the speed and strength of corrective movements both respond to practice, which is the basis for the exercise programmes tested in fall-prevention trials.

The best-known balance test with survival data asks a person to stand on one leg for 10 seconds, eyes open and without support, with up to three attempts. In a Brazilian study of 1,702 adults aged 51 to 75 (68% men), 20.4% could not complete it. Over a median of seven years, 17.5% of those who failed died, compared with 4.6% of those who passed [1].

After adjusting for age, sex, body mass index and existing illnesses, failing the test was linked to 84% higher all-cause mortality: a hazard ratio (the risk in one group relative to another; 1.84 means 84% higher) of 1.84, 95% CI 1.23 to 2.78. Adding the test to standard risk factors improved the prediction of death [1]. The authors note that they could not account for recent falls or physical activity.

An earlier review of objective physical capability tests in older adults found that all five studies of standing balance showed some link between poorer balance and higher mortality, but not every result was statistically significant, and the tests differed too much between studies to be combined into one estimate [5]. Balance is one of several such tests covered in functional fitness.

The largest review of exercise for fall prevention, a 2019 Cochrane review, pooled 108 randomised trials with 23,407 people aged 60 and over living in the community in 25 countries (average age 76, 77% women). Balance and functional exercises, meaning exercises done standing that challenge balance and resemble daily tasks, reduced the rate of falls by 24%: a rate ratio (the number of falls in one group relative to another; 0.76 means 24% fewer) of 0.76, 95% CI 0.70 to 0.81, from 39 trials and 7,920 participants, rated high-certainty evidence. The same exercises reduced the number of people who fell at least once by 13% [2].

Programmes combining several types of exercise, most often balance and functional exercises plus resistance training, probably reduced the rate of falls by 34% (rate ratio 0.66, 95% CI 0.50 to 0.88; 11 trials, moderate-certainty evidence). Tai chi may reduce the rate of falls by 19% (rate ratio 0.81, 95% CI 0.67 to 0.99; low-certainty evidence). The effects of programmes consisting mainly of resistance training, dance or walking were uncertain, and no trials compared flexibility or endurance exercise alone with a control group [2].

Reported side effects were mostly minor muscle and joint problems. One trial reported two serious adverse events, a pelvic stress fracture and a hernia operation [2].

A 2017 meta-analysis of 88 randomised trials (99 comparisons, 19,478 participants) examined which features of exercise programmes predicted larger effects. Overall, exercise reduced the rate of falls in community-dwelling older people by 21% (rate ratio 0.79, 95% CI 0.73 to 0.85). Programmes that challenged balance and involved more than 3 hours of exercise a week had the largest effects; together these two features explained 76% of the differences between trials and were linked to a 39% reduction in falls (rate ratio 0.61, 95% CI 0.53 to 0.72) [3].

The same analysis found fall reductions in people with Parkinson's disease (rate ratio 0.47, 95% CI 0.30 to 0.73) and with cognitive impairment (rate ratio 0.55, 95% CI 0.37 to 0.83), each from a small number of comparisons. It found no evidence of a fall-prevention effect in residential care, among stroke survivors or in people recently discharged from hospital [3].

In trial programmes, "challenging balance" generally meant exercises done standing that reduce the base of support, move the centre of mass and reduce reliance on the hands, progressed over time as participants improved.

The trunk muscles contribute to balance, particularly in hip and trunk corrections to larger sway. Core stability training in the narrow sense, meaning isolated trunk exercises performed lying or kneeling on a mat, is widely promoted, but it has not been evaluated as a separate exercise category in the major fall-prevention reviews, and its effect on falls or mortality is unknown [2].

The programmes that reduced falls in trials were done standing and moving, with the trunk working as part of whole-body balance tasks rather than in isolation [2].

The balance and mortality data are observational. People who cannot stand on one leg for 10 seconds are more likely to be unfit, frail or already ill, so the test partly measures overall health, and illness causing poor balance (reverse causation) is a plausible explanation for part of the link. The one-leg stance study is a single study from one research group [1].

The randomised trials measured falls, not deaths. No trial has randomly assigned people to balance training and followed them for all-cause mortality, so any effect on lifespan would run through fewer falls and fall injuries, described in fall prevention. Most trials in the Cochrane review had unclear or high risk of bias on at least one item [2]. On the balance of this evidence, the US Preventive Services Task Force recommends exercise to prevent falls in adults 65 and over at increased risk of falling [4].