Fall Prevention in LMICs: Balance Improves, Falls Uncertain

Exercise improves balance in lower-resource settings. Whether it prevents falls remains unanswered.

Fall-prevention advice often travels farther than the research behind it. Clinical guidelines largely grew out of trials in high-income countries, where older adults may have better access to primary care, physical therapy, safer housing, medication review, and follow-up. A 2026 systematic review of low- and middle-income countries (LMICs) finds that exercise programs improve balance in community-dwelling older adults, but it could not confirm a statistically significant reduction in documented falls.

That distinction matters. Better balance is a meaningful fall-risk marker, not a fall event. The evidence supports structured balance and strength exercise as a practical action now. It does not yet prove that exercise alone reduces fall rates across LMIC settings.

What the LMIC review found

Chellapillai et al.'s 2026 systematic review and meta-analysis included 30 studies from 12 LMICs, involving 9,817 community-dwelling adults aged 60 to 94; 65% were women. Interventions included Otago-style exercise, Thai traditional dance, senior dance, and education programs. Only nine studies provided enough compatible data for meta-analysis, and each outcome drew on just two to four studies.

For exercise-based interventions, the pooled balance result was large and statistically significant: standardized mean difference (SMD) 1.18, 95% confidence interval 0.30 to 2.06, p=0.008. The wide interval means the estimated size remains imprecise, but it excludes no effect. The result supports the authors' conclusion that these interventions reduce fall risk through improved balance.

Balance has a credible connection to falls. In a retrospective study of 300 community-dwelling older adults, Yin et al. in 2025 found that people classified as at higher fall risk had slower cadence, shorter steps, wider step width, greater gait variability, and more sway. Their analysis identified several gait and balance components that predicted fall-risk group. The study supports balance and gait assessment as risk markers; it does not show that changing one balance score will automatically prevent a future fall.

The clinical endpoint in the LMIC review was less certain. For documented fall incidence, the pooled estimate was SMD -0.40 (95% CI -1.08 to 0.29; p=0.18). The confidence interval includes no effect, so this result was not statistically significant. It also includes a potentially substantial reduction, which is why calling exercise ineffective would overread the data.

Mobility showed a similar non-significant pattern: SMD -0.50, 95% CI -1.13 to 0.12, p=0.11. Fear of falling barely moved: SMD -0.04, 95% CI -0.69 to 0.61, p=0.90. These are separate outcomes. A program can improve balance while leaving confidence unchanged, and a balance gain can fail to produce a measurable fall-rate change in a small, heterogeneous set of studies.

Why the fall-rate question is still open

The LMIC meta-analysis has too little pooled evidence to settle the question. Two to four studies per outcome leave plenty of room for chance, differences in program delivery, and inconsistent fall measurement. Self-reported falls also create recall bias, especially where ongoing surveillance is limited. The fall-rate confidence interval spans both benefit and no benefit.

The studies also combined interventions with different mechanisms and practical demands. Thai traditional dance, senior dance, Otago-style balance training, and education-only programs do not ask the same thing of participants or target the same pathways to a fall. Pooling them can provide a broad answer, but it can hide whether one format works better in a particular community.

The review's narrative synthesis identified Thai traditional dance, senior dance, and structured education as potentially effective approaches, but too few studies of each type could be pooled to confirm a result. Group programs remain plausible and practical. A 2022 umbrella review by Neil-Sztramko and colleagues found that community-based group physical activity and nutrition programs can improve mobility-related outcomes in older adults. Mobility improvement still differs from fewer falls, but group delivery may make regular participation more achievable.

Structured exercise also has a broader functional evidence base. A 2024 systematic review of randomized trials of the Otago Exercise Program found improved physical function in older adults. It cannot provide an LMIC-specific estimate for actual fall rates, so it should not be used to fill the gap left by the LMIC review.

Exercise addresses one part of a multicausal problem

Falls rarely have a single cause. Uneven walking surfaces, poor lighting, missing handrails, unsuitable footwear, medication effects, weakness, and gait impairment can all matter. A review in adults with mild to moderate cognitive impairment cataloged environmental hazards, lower-limb weakness, gait problems, and medications among the modifiable fall-risk factors in that population. Racey et al.'s 2022 review is not evidence that every older adult has the same risk profile, but it illustrates why balance training cannot remove every route to a fall.

The most consequential gap is program design. Chellapillai and colleagues found no LMIC trial of a multifactorial community program that combined exercise with home-hazard modification and medication review. Every included study tested a single modality or education alone. That means LMIC research has not tested whether a broader approach can produce the fall-rate reductions reported in more resourced settings.

For comparison, a 2026 nationwide randomized trial in the Netherlands found that a fall-prevention intervention reduced falls and fall-related injuries among community-dwelling older adults at increased risk of falling. This does not contradict the LMIC review as cleanly as it first appears. The populations, healthcare environments, and program models differ, and the LMIC analysis did not yield a precise null result. The Dutch trial shows what a comprehensive program can achieve in that setting; it does not establish that the same result will transfer unchanged elsewhere.

Assessment and technology do not transfer automatically either

Tools such as the Short Physical Performance Battery and STEADI were developed for settings with clinical staff and follow-up capacity. In a prospective cohort of 417 older adults recruited through nine primary care practices, Welch et al. found that the SPPB helped stratify fall risk. The SPPB includes gait time, chair stands, and balance testing. It may be useful where trained personnel can administer it, but a tool validated in primary care cannot be assumed ready for every low-resource community.

High-income research is also testing interventions that most LMIC community programs cannot readily use. A 2024 double-blind randomized trial in 54 people with chronic stroke used specialized treadmill force-field training for 24 sessions over 12 weeks. Krause et al. found that repeated mediolateral leg perturbations improved an aspect of walking-balance control, step-width modulation. The population and equipment make it a poor substitute for evidence in generally healthy community-dwelling older adults.

Technology is not automatically a shortcut. In a multicenter randomized trial, 461 people with multiple sclerosis were assigned to usual care or six months of unsupervised home-based step exergaming, prescribed at 120 minutes weekly. Hoang et al. reported no difference in fall rates between groups; average training was 70 minutes per week. The study involved a disease-specific population and a custom device, so it says little about general LMIC fall prevention. It does show that a screen and a stepping game do not make the endpoint disappear.

What older adults and communities can do now

For community-dwelling older adults in LMIC settings, the best-supported option is structured exercise that trains balance and strength. Group programs, including culturally appropriate dance or community exercise formats, may be especially workable when they improve attendance and access. The evidence justifies this recommendation because balance improved significantly; it does not justify promising a known reduction in fall rates.

Programs should also look beyond the exercise session. Ask about obvious household and walking hazards, footwear, lighting, assistive needs, and medicines that may affect balance. Those measures need LMIC-specific trials, especially in combination with exercise, before anyone can state their fall-rate effect with confidence. But the current research gap is an argument for testing that model, not for waiting until every sidewalk behaves itself.

Verdict: Balance and strength exercise is a sensible, evidence-backed starting point for older adults in LMICs. The large balance improvement in the 2026 review is real. Statistically confirmed reduction in documented falls is still an unanswered clinical question because the available pooled evidence is too small and too imprecise.