Aquatic vs. Land Exercise for Balance After 65

For steadier walking and easier everyday movement, older adults do not need to find the one exercise mode that defeats all others. The evidence points to a more useful question: what is limiting training now? Pain with land exercise, fear of falling, weak legs, poor balance during turns, and slow chair rises call for different emphasis.

Water exercise has a measurable edge over land exercise on some balance tests. Yet it has not beaten land exercise on the walking outcomes most people care about. Across the broader evidence, programs that combine resistance work with balance practice have the most consistent support for gait speed, chair-rise performance, and functional mobility.

What the direct aquatic-versus-land comparison found

The best evidence on exercise setting comes from a 2024 systematic review and meta-analysis by Deng et al. The review pooled 29 trials with 1,486 older adults and directly compared aquatic exercise with land-based exercise. It included participants who were healthy as well as people with musculoskeletal, cardiopulmonary, and nervous-system conditions.

Aquatic exercise produced higher Berg Balance Scale scores than land exercise: pooled standardized mean difference (SMD) 1.13, 95% confidence interval (CI) 0.25 to 2.00, p=0.01. It also produced an initial advantage on the 30-second chair-stand test, SMD 2.02, 95% CI 0.50 to 3.54, p=0.009. Those results suggest that water-based programs can improve some measured aspects of balance and repeated sit-to-stand capacity more than the land programs in these trials.

The result did not extend to every mobility endpoint. The six-minute walk test showed no statistically significant aquatic advantage: SMD 0.13, 95% CI -0.16 to 0.43, p=0.38. Timed Up and Go, which combines standing, walking, turning, and sitting, also did not reach statistical significance: SMD 0.44, 95% CI -0.04 to 0.91, p=0.07. Water may be a very good place to train; it is not a proven shortcut to better walking across the board.

The review's sensitivity analyses make restraint especially important. The Berg Balance Scale result remained statistically significant after the analysis, but fell to SMD 0.82, 95% CI 0.17 to 1.48, p=0.01. The chair-stand result fell to SMD 0.58, 95% CI -0.20 to 1.36, p=0.15, and no longer met the conventional threshold for statistical significance.

Between-study differences were large: I² was 94% for the Berg Balance Scale, 85% for Timed Up and Go, 96% for chair stand, and 62% for the six-minute walk. High heterogeneity means the studies did not produce neatly consistent estimates, likely because populations and programs differed. Seven aquatic studies also did not report enough detail about the pool setting to make dose-matched replication easy.

Why resistance alone is usually too narrow

Resistance training belongs in a mobility program. A 2021 systematic review by Keating et al. examined 12 randomized trials involving 499 participants assigned to resistance training. The trials repeatedly indicated improvements in gait and balance-related outcomes among older adults. But they used different loads, schedules, and endpoints, so the authors did not pool a single effect estimate. The review supports strength training as a foundation, not as proof that lifting alone is the best answer for every movement problem.

The more specific numbers favor combining strength work with balance training. In a 2023 network meta-analysis by Shen et al. of randomized trials in older adults with sarcopenia, resistance plus balance training increased usual gait speed by 0.16 m/s (95% CI 0.06 to 0.26). That exceeded the authors' minimally important difference of 0.1 m/s. The same combination improved Timed Up and Go by 1.85 seconds (95% CI 0.49 to 3.22) and improved five-repetition chair-stand time by about 1.72 to 2.28 seconds, depending on the comparison.

Those findings fit the task. Walking safely demands leg force, but also weight shifting, turning, single-leg control, and corrections after a wobble. Chair rises need strength, yet practice standing quickly and under control can matter as much as adding another machine exercise. In the same sarcopenia analysis, resistance plus balance plus nutrition improved handgrip strength by 4.19 kg (95% CI 2.55 to 5.83). That is useful physical-function evidence, but it does not prove fewer falls or longer life.

Population matters. Sarcopenia is a clinical subgroup, so these effect estimates do not automatically apply to every community-dwelling adult over 65. Still, they give the clearest quantitative support here for a multimodal approach when the goal is gait and day-to-day mobility.

Build training around the limiting factor

The wider literature supports structured, progressive exercise rather than loyalty to one modality. A 2026 systematic review of randomized trials in community-dwelling older adults concluded that both balance-based and strength-based interventions reduce fall-risk factors. A 2026 review of 95 controlled studies also found that exercise improves lower-body strength, mobility, and balance-related functional fitness, although substantial methodological differences prevented a meta-analysis.

A practical program should therefore include progressive resistance exercise and balance tasks that resemble the situations a person struggles with. If turning, uneven ground, or single-leg support causes instability, balance practice deserves deliberate time rather than a token warm-up. A 24-week randomized trial in healthy, community-dwelling older adults found additional functional-performance gains when unstable strength training was paired with cognitive training, according to Oliveira et al. The result supports task complexity as one option for capable adults; it does not mean everyone should begin on unstable equipment.

Aquatic exercise makes particular sense when joint loading, pain, or low confidence makes land training hard to sustain. The water changes loading and may let some people practice movement who would otherwise avoid it. A 2025 comparative study in elderly women found meaningful physical-fitness differences between aquatic and terrestrial environments, but its narrow population cannot establish a universal winner. The aquatic meta-analysis also found larger Berg Balance Scale gains in musculoskeletal-disease and nervous-system subgroups, but subgroup results are not a prescription for healthy adults.

Better test scores do not equal fewer falls

Most trials measured balance scales, gait speed, chair stands, walking distance, or Timed Up and Go. These are useful functional measures, but they are surrogate outcomes, not fall counts, fractures, hospitalization, disability, or mortality. In the sarcopenia network meta-analysis, multicomponent exercise plus nutrition did not reduce mortality: relative risk 1.23, 95% CI 0.74 to 2.06.

Falls also depend on medication effects, vision, home hazards, attention, prior falls, and behavior. The 2026 INITIATE prospective cohort study notes that clinicians still lack consensus on the best mobility and balance tests, and the cutoffs that predict falling. A performance score can guide training; it cannot forecast a person's future with the confidence of a weather app, let alone a crystal ball.

For people with repeated falls, a new neurologic symptom, chest pain, severe joint pain, or a condition that affects balance, individualized assessment belongs alongside exercise planning. Exercise is part of fall-risk management, not a replacement for it.

The practical verdict

Choose resistance plus balance training when the target is faster walking, easier chair rises, and better functional mobility. The strongest quantitative mobility findings come from that combination, though the clearest estimates come from adults with sarcopenia. Use balance-specific drills when balance is the bottleneck, because skill practice transfers best when it resembles the task.

Choose aquatic exercise when water makes regular training possible or safer-feeling than land exercise. It may outperform land programs on some balance measures, but current direct evidence does not show a statistically significant advantage for Timed Up and Go or six-minute walking performance. Consistent, progressive training that a person can actually do will beat a theoretically perfect program left on the calendar.