Activity monitors after stroke have not yet shown a reliable boost in daily movement
Physical inactivity remains one of the most stubborn barriers in stroke rehabilitation. Survivors often leave the acute phase with hemiparesis, fatigue, gait impairment, and a reduced confidence in walking, then settle into routines that reinforce sedentary behavior rather than recovery. Wearable activity monitors and smartphone-based tools were designed to interrupt that pattern by making movement visible, measurable, and more actionable at the bedside and in the community.
The evidence assembled in this Cochrane review does not yet support that promise with confidence. Across the randomized evidence base, activity monitors were feasible to use, but their effect on physical activity after stroke remained uncertain and, in the outcomes that mattered most, inconsistent.
Four small trials left the central question unresolved
Investigators identified 4 randomized controlled trials, reported in 11 publications, from an initial pool of 28,098 references. Together, those studies enrolled 245 participants, with individual trial sizes ranging from 27 to 135. The populations were heterogeneous in timing after stroke: one study enrolled people less than one week from the event, two enrolled participants at one to three months, and one included people at a median of 51 months poststroke. Stroke severity also varied, with reported National Institutes of Health Stroke Scale medians ranging from 1 to 6.
Three studies took place in inpatient rehabilitation, and one was conducted in a university laboratory. In every trial, the intervention was not a monitor alone. Participants received an activity monitor alongside another intervention, such as walking retraining or usual inpatient rehabilitation, and were compared with the same intervention without a monitor. That matters clinically because the device was never tested as a stand-alone solution; it was used as a behavioral adjunct layered onto rehabilitation already in motion.
The primary outcomes were steps per day and time spent in moderate-to-vigorous physical activity. Secondary outcomes included sedentary time, light-intensity activity, walking duration, fatigue, mood, quality of life, community participation, and adverse events. None of the trials reported the main combined outcome of moderate-to-vigorous physical activity as prespecified. One trial did separate moderate and vigorous activity, which offers a narrow window into how the devices may have shifted behavior in rehab settings.
Step counts did not move in a clinically convincing direction
Daily step counts are the most concrete expression of whether a stroke survivor is becoming more active. They reflect not just exercise tolerance but also confidence, motor recovery, and the practical ability to translate therapy into everyday mobility. In this review, that signal was weak.
In a community setting, one randomized trial with 27 participants found no clear effect of an activity monitor added to another intervention on steps per day. The mean difference was -1930 steps, with a 95% confidence interval from -4410 to 550. In inpatient rehabilitation, two randomized trials with 83 participants together also failed to show a clear benefit; the mean difference was 1400 steps, with a 95% confidence interval from -40 to 2840. Both findings were graded as very low-quality evidence.
The confidence intervals tell the story. They are wide enough to include a meaningful improvement, no change, and, in the community trial, possible harm or regression. In rehabilitation research, that range usually means the intervention has not been isolated well enough from background variability to justify routine adoption as an evidence-based strategy for increasing daily ambulation.
Moderate activity rose in one study, but the signal was narrow
One trial reported a different pattern. When an activity monitor was added to usual inpatient rehabilitation, moderate-intensity physical activity increased by 4.4 minutes per day, with a 95% confidence interval from 0.28 to 8.52. The evidence for that estimate was rated low quality. The same study did not show a clear increase in vigorous-intensity activity, with a mean difference of 2.6 minutes per day and a 95% confidence interval from -0.8 to 6.
A gain in moderate-intensity activity may still matter biologically. After stroke, even modest increases in sustained walking and active standing can reduce deconditioning, support cardiovascular conditioning, and train more efficient motor patterns through repetition. But a shift of only a few minutes a day, without parallel evidence for higher step counts or broader functional gains, is not enough to establish a reliable rehabilitation tool. In practice, it is a signal worth watching, not a basis for routine implementation.
Feasibility was better than efficacy
The review found low overall risk of bias aside from unavoidable problems with blinding of participants and personnel. That limitation is familiar in behavioral rehabilitation trials, where patients and therapists usually know whether a monitor is being worn. Still, the dominant issue was not methodological theater; it was the scarcity and small size of the underlying evidence.
None of the included trials reported adverse effects. In a rehabilitation context, absence of harm reporting is not the same as proof of safety. Wearables are usually low-risk, but they can still provoke frustration, false reassurance, or disengagement if patients perceive the device as a substitute for structured therapy rather than a support to it. Without explicit reporting, those practical effects remain invisible in the published data.
What these studies do show is that activity monitors can be integrated into poststroke rehabilitation workflows. Therapists and patients were able to use them in inpatient and experimental settings. The gap lies between feasibility and a reproducible improvement in movement outcomes.
Why the biology is plausible but the trial signal is thin
The appeal of activity monitors after stroke is straightforward. Stroke recovery depends on repetition, feedback, and gradual expansion of tolerated activity. A wearable device can make movement immediately measurable, provide reinforcement, and create a target for goal-setting. In theory, that feedback loop should help patients accumulate more steps, sit less, and spend more time in productive mobility.
Real-world recovery is more complicated. Early poststroke fatigue, balance fear, hemiplegia, aphasia, cognitive deficits, and depression can all interfere with adherence, even when patients can see their activity data. Inpatient rehabilitation also includes many competing goals, from transfers and dressing to gait retraining and safety training, so step count alone may not capture the broader work of recovery. If the intervention does not change how therapists coach mobility or how patients respond to the monitor’s feedback, it may never produce a large enough behavioral shift to alter the outcome curve.
The evidence base is still too small for routine recommendation
Only four small randomized trials have examined this question, and three were conducted in inpatient rehabilitation rather than in typical long-term community care. That leaves a major evidence gap for the phase of recovery when people with stroke must manage their own activity without direct supervision. The review also searched broadly across multiple databases and trial registries through 3 March 2018, which strengthens confidence that the small evidence base reflects the field rather than an incomplete search.
For now, the most defensible conclusion is restrained. Activity monitors may fit into stroke rehabilitation practice, but current evidence does not show that they reliably increase physical activity after stroke. Larger randomized trials, with clear reporting of daily steps, moderate-to-vigorous activity, sedentary time, and patient-centered outcomes, are still needed before these devices can be considered a proven intervention for poststroke mobility recovery.