- Design
- single-blind three-arm randomised controlled trial over 12 weeks
- Population
- 117 long-term care residents in Chongqing, China; mean age 75.5 years, 50% female, 39 per arm
- Primary outcome
- Fried frailty phenotype at 12 weeks
- Effect
- significant time, group (F=3.46, P=0.035) and interaction (F=45.99, P<0.01) effects; augmented reality better than control at 12 weeks (P<0.01); adherence 91% and 89%
One hundred and seventeen residents of long-term care facilities in Chongqing, mean age 75.5, were randomised in equal numbers to Baduanjin - a traditional eight-movement Chinese exercise set - delivered either with an augmented reality system providing real-time kinematic tracking and visual feedback, by video instruction, or to a control group. Assessment was single-blind at baseline, six and twelve weeks, with the Fried frailty phenotype as the primary outcome.
Adherence was high in both exercise arms (91% augmented reality, 89% video). Frailty showed significant effects of time, group and their interaction, with the augmented reality group improving more than controls at 12 weeks. The augmented reality arm also outperformed both the video arm and controls on physical function, activities of daily living, nutrition, cognition and mood.
The comparison that matters is augmented reality against video, not against control - everyone knows exercise beats no exercise. That the feedback arm beat the video arm on every secondary domain, with adherence nearly identical in both, suggests the benefit came from doing the movements better rather than from doing them more often. Thirty-nine residents per arm is small, the outcomes are numerous and the report gives F statistics rather than effect sizes, so the size of any of these differences is unknown. The transferable point is not the hardware: it is that supervised, corrected exercise outperformed unsupervised exercise in frail residents, which argues for a person in the room before it argues for a screen.
- Prioritise supervised or corrected exercise over handing residents a video - that was the comparison that separated
- Use the Fried phenotype or an equivalent to measure frailty before and after any exercise programme, so you know whether it worked
- Note adherence was around 90% in both arms - engagement was not the differentiator
- Do not read this as a case for buying augmented reality equipment; read it as a case for staffing the session
- Baduanjin is low-intensity, seated-adaptable and needs no equipment, which makes it feasible where space and staff are limited
Why it matters
It isolates supervision, rather than the exercise itself, as what made the difference in frail residents.
Don't overread it
One hundred and seventeen residents at one site, many outcomes tested, and results reported without effect sizes.
The statistics, in plain English
The results are reported as F statistics from repeated-measures analysis of variance with P values, and no effect sizes or confidence intervals - so we know differences reached significance but not how large they were, which is the number a reader needs. With 39 residents per arm and at least six outcome domains tested, some of the secondary findings would be expected to separate by chance alone. The primary frailty outcome and the high adherence are the parts to trust.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for geriatrics, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free