- Design
- Systematic review and meta-analysis of randomised trials
- Population
- 18 trials, 1,993 adults with cardiovascular disease in phase II or III rehabilitation
- Primary outcome
- Peak VO2, six-minute walk distance and daily steps
- Effect
- Peak VO2 +1.69 mL/kg/min (0.46–2.92); 6MWD +31 m (7.7–53.9); steps +1,063/day (384–1,825)
This meta-analysis pooled randomised trials from 2015 to August 2026 in which a wrist-worn wearable was built into fully remote or hybrid cardiac rehabilitation, started in phase II or III. It included 18 trials with 1,993 participants; trials focused mainly on heart failure were excluded.
Against comparators, wearable-guided programmes improved peak VO2 by 1.69 mL/kg/min (95% CI 0.46–2.92), six-minute walk distance by 31 m (7.7–53.9) and daily steps by about 1,060 (384–1,825). There was no clear effect on anaerobic threshold, BMI, waist, grip strength or quality of life.
The important caveat is that all three 95% prediction intervals included zero. On average these programmes help, but a new programme in a different setting might not. The interventions were also multicomponent, so the wearable itself cannot be credited with the effect.
Where centre-based rehabilitation is out of reach — the usual situation for many patients in India — a structured remote programme with activity tracking is a reasonable option, not a lesser substitute to be dismissed.
- Refer every eligible patient after an acute coronary event or revascularisation to cardiac rehabilitation, in whatever form is reachable.
- Where centre-based rehabilitation is not accessible, consider a structured remote or hybrid programme with activity monitoring.
- A wearable alone is not rehabilitation; the benefit came from programmes with feedback and behavioural support.
- Daily step count is a simple, trackable target to set and review at follow-up.
Why it matters
Low uptake of centre-based rehabilitation is the main gap, and this supports remote delivery as a workable alternative.
Don't overread it
The effect belongs to multicomponent programmes; it does not show that giving a patient a smartwatch improves fitness.
The statistics, in plain English
The confidence interval describes the average effect across trials, and it excludes zero. The prediction interval estimates the range of effects a new, similar programme might show, and it includes zero — so a single new programme could see no benefit. A 1.69 mL/kg/min rise in peak VO2 is modest but in the range considered meaningful.
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 cardiology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free