Exergames for Knee OA: Helpful at Home, Not Clinic-Equivalent

Exergames can make home exercise more usable. They have not replaced physiotherapy.

For someone with knee osteoarthritis, the practical appeal is obvious: exercise at home, receive feedback, and have a reason to do the next session. The evidence supports part of that promise. Exergames—digital games that require physical movement—consistently beat passive or no-treatment controls in the small osteoarthritis trials published so far. They have not, however, been tested against supervised physiotherapy in this evidence base. That distinction decides what patients and clinicians can honestly claim.

The best current summary, a 2025 systematic review by Guede-Rojas et al., included eight randomized controlled trials with 401 people with osteoarthritis. Seven studies enrolled people with knee OA. Every trial reported improvement in at least one outcome with a non-immersive exergame versus its control. The outcomes included pain, disability, balance, strength, gait, proprioception, range of motion, quality of life, depression, and fear of movement.

That is a consistent positive signal. It is not a pooled estimate of how much an exergame improves pain or walking. The trials used different programs, outcomes, and protocols, so the reviewers could not combine their results in a meta-analysis. A collection of positive studies can tell us that the intervention merits attention; it cannot supply a reliable average effect size when the studies measure different things.

What the strongest practical trial actually tested

A single-blind randomized trial of the Rak Kao mobile app offers a useful real-world example. Over four weeks, knee OA patients around age 62 to 63 used either app-guided exercise or a static educational handout. Among 82 completers, participants using the app performed 76% of repetitions with correct form, compared with 53% in the handout group; the difference was statistically significant (p<0.05).

That is a 23-percentage-point absolute difference, or a 43% relative increase in correctly performed repetitions. The app group also had lower dropout: 4.5% versus 11.1%. Secondary KOOS and Knee Society Score outcomes favored the app group.

The primary endpoint matters here. The trial was powered to assess exercise accuracy, not clinical outcomes such as pain or physical function. Better form is a plausible route to better rehabilitation, but it is not the same thing as proving a meaningful, lasting improvement in symptoms. And a static handout is a passive comparator. The study shows that interactive guidance can outperform printed instruction; it does not show that an app outperforms a physical therapist.

Why the early results may look larger than routine practice

The most important caution comes from a 2022 meta-analysis by Hai et al. in healthy older adults, not people with OA. Across 48 studies and 2,197 participants, exergame training improved balance (Hedges' g=0.59, 95% CI 0.46 to 0.71), lower-body strength (g=0.51, 95% CI 0.37 to 0.65), and gait (g=0.33, 95% CI 0.08 to 0.59). These are useful benchmarks, but they cannot be transferred directly to knee OA.

Two moderator findings make that limitation more consequential. Higher BMI was associated with smaller effects (β=-0.01, p<0.01), and larger studies also reported smaller effects (β=-0.004, p<0.01). The OA trials average about 50 participants each, while OA populations often have higher BMI than the healthy volunteers in the meta-analysis. This does not prove that exergames fail in people with higher BMI, nor does it quantify the reduction for knee OA. It does mean the small positive OA trials are better treated as upper-bound estimates than as a forecast of what every home user will experience.

The same meta-analysis found no reliable improvement in lower-body flexibility: g=0.10, with a 95% CI from -0.45 to 0.67. Some OA trials reported range-of-motion gains, but the broader result remains uncertain because its confidence interval crosses zero.

Digital rehabilitation can work without proving that games are the active ingredient

Technology-assisted home rehabilitation has stronger footing than the narrower claim that gamification itself matches clinic care. In a 2025 randomized trial by Plavoukou et al., people with knee OA received remotely delivered physiotherapy using biomechanical sensor feedback or in-person physiotherapy. Pain, muscle strength, and neuromuscular outcomes were comparable between groups. That result supports remote, feedback-guided delivery as a viable model. It did not test an exergame, so it cannot establish that a game is interchangeable with supervised care.

A 2025 randomized trial of exergames as complementary therapy adds directional support in older adults with knee or hip OA, but its mixed-joint population prevents a knee-specific conclusion. A separate knee OA trial using semi-immersive VR reported improvements in pain, disability, function, and balance. VR glasses are a substantially different intervention from a phone-based, non-immersive program, so those findings should not be treated as proof for mobile apps.

The evidence has a methodological ceiling

Seven of the eight OA trials in the systematic review had high risk of bias on the Cochrane RoB-2 tool. Their mean PEDro score was 6.1, which indicates moderate methodological quality. Participants cannot be blinded to whether they are playing an interactive game or receiving a handout or no intervention. That can change effort, expectations, and self-reported outcomes. Unblinded assessors can add another source of bias.

These problems do not erase every positive result. They limit confidence in the size of the effect and make comparisons with active rehabilitation especially important. None of the OA exergame trials reviewed by Guede-Rojas et al. used supervised physiotherapy as the comparator. No non-inferiority or equivalence trial in this evidence base shows that home exergames match clinical rehabilitation outcomes.

Duration is another gap. Four weeks was the modal intervention length across the OA studies, including Rak Kao. No trial has established whether benefits persist after the program ends or whether longer programs produce better results. Cost is also unresolved: the available cost-effectiveness analysis of rehabilitation gaming studied upper-limb rehabilitation in multiple sclerosis, not knee OA. It cannot answer whether a knee OA exergame saves money or improves value for patients.

Who should consider an exergame?

An exergame is a reasonable adjunct for people with knee OA who struggle to follow a paper home-exercise plan, need feedback on form, or want a more engaging way to exercise between clinic visits. Evidence from healthy older adults also suggests that pairing exergames with conventional exercise may help motor performance and quality of life; a 12-week randomized trial found benefits for a hybrid program compared with traditional exercise alone. That study did not enroll people with OA, so it supports the adjunct concept rather than a knee-OA treatment claim.

Home use needs judgment. A study of home-based exergames in older adults raised safety concerns for vulnerable people exercising without supervision and noted dropout above 20% in this type of trial. The low dropout in a structured four-week app study may not carry over to independent use at home. People with poor balance, substantial mobility limits, or fall risk should ask a clinician or physical therapist whether monitoring is appropriate before starting.

Verdict: Exergames are credible tools for making home exercise more accurate and engaging, and small OA trials consistently beat passive controls. The evidence does not show that they replace supervised physiotherapy, predict durable benefits, or deliver the published effect sizes in routine higher-BMI OA populations.