Wearable Sensors Reveal How Sedentary Older Patients Really Are in Rehab

Older adults recovering in geriatric rehabilitation are supposed to be moving, exercising, and rebuilding the strength they need to live independently. Yet a new study from the Netherlands suggests that many of them spend more than 90 percent of their waking hours sitting, reclining, or lying down — even while enrolled in active rehabilitation programmes. The research, published in European Geriatric Medicine, also tested whether a wearable sensor system that gives patients and therapists feedback on their sedentary behaviour could help change that picture. The verdict was a nuanced one: the technology proved feasible enough that patients wore it almost every day, but its usability fell short, undermined by inaccurate measurements, unclear goals, and a lack of obvious added value for the people it was meant to help.

The study, led by Suzanne M. Debeij of Leiden University Medical Center and colleagues, is part of a larger multicentre cohort project called Better@Home, which followed 110 patients across eight Dutch geriatric rehabilitation centres. For the sensor sub-study, the researchers recruited 26 patients from a single rehabilitation centre in an urban area, along with nine care professionals, most of them physical therapists. The participants were typical of the geriatric rehabilitation population: the median age was 77 years, roughly six in ten were women, and most lived alone. Their reasons for rehabilitation ranged from elective orthopaedic surgery and trauma to respiratory disease, and they spent a median of about 25 days as inpatients followed by roughly 46 days of home-based rehabilitation.

The technology at the heart of the study was the Hipper sensor monitoring system, a hip-worn device containing a three-dimensional accelerometer that samples movement 12.5 times per second and produces an activity output every minute. That output is expressed as a Physical Activity Measure, or PAM score, which reflects the amount of body movement recorded and can be converted into metabolic equivalents for analysis. Data travel wirelessly from the sensor to a Raspberry Pi-based transmitter and then to web dashboards accessible to both patients and care professionals. Crucially, the system was not just a measurement tool: therapists were expected to use the data to coach patients about their movement behaviour, often during regular therapy sessions. Three versions of the sensor were used during the study period, including an updated prototype that provided minute-by-minute rather than 15-minute output and required redesigned dashboards.

The researchers assessed usability and feasibility using a mixed-methods design. A focus group with the care professionals was structured around the User Experience Honeycomb model and analysed with the qualitative Framework method, while patients completed interviewer-administered surveys at admission and after discharge. Quantitative adherence came from the sensors themselves, with wear days counted as a measure of how consistently the system was actually used. This triangulation allowed the team to compare what professionals and patients said about the technology with what the devices recorded in practice.

The adherence numbers were striking. Twenty-three patients wore the sensor during inpatient rehabilitation for a median of 11 days, corresponding to 100 percent of the days the sensor was offered. During home-based rehabilitation, 19 patients used the device, wearing it on 96.6 percent of the days it was available. For a population characterised by frailty, multimorbidity, and often some degree of cognitive impairment, those adherence rates are remarkable and suggest that, at a basic level, the system worked as intended. The researchers concluded that feasibility — the extent to which the digital health system functions as intended — was sufficient, though they caution that high adherence may partly reflect compliance with clinician instructions rather than genuine engagement with the technology.

What the sensors recorded, however, was sobering. During inpatient rehabilitation, patients spent a median of 93 percent of their waking day — about 666 minutes, or more than 11 hours — in sedentary behaviour. At home, the proportion dipped slightly to 90.5 percent, but because total wear time increased, the absolute sedentary time actually rose to a median of about 704 minutes per day. Time spent in moderate-to-vigorous physical activity was essentially zero in both settings, at 0.0 percent of waking time, while light physical activity accounted for only about 7 percent in hospital and roughly 10 percent at home. The pattern of sedentary bouts did shift between settings: patients had more frequent but shorter sedentary episodes at home, and time spent in uninterrupted bouts longer than an hour decreased by nearly 48 minutes.

These findings matter because prolonged sedentary behaviour is not a benign habit. It is associated with increased risk of non-communicable diseases and all-cause mortality, even in older adults who meet recommended physical activity guidelines, and it may delay functional recovery, worsen deconditioning, and increase dependency in daily activities. Both total sedentary time and long, uninterrupted sedentary bouts contribute to the harm. For frail older patients, simply breaking up sitting with short, low-effort movements may be a more realistic starting point than pushing them toward higher-intensity exercise — which is precisely the kind of behavioural nudge that monitoring-and-feedback systems are designed to deliver.

Yet the qualitative data revealed why the system struggled to fulfil that promise. Care professionals valued the sensor as a conversation starter that opened meaningful, motivational discussions and offered insight into when and how intensely patients moved — one therapist described being able to see whether a patient practised all morning and then did nothing in the afternoon. But they were troubled by the opacity of the PAM score, which patients could not interpret, and by perceived measurement inaccuracies that eroded trust in the data. One physiotherapist recounted a patient who was offended when the sensor failed to record her climbing the stairs, even though she had been working hard. The denser data streams from the updated sensors made the dashboard graphs more complex and harder to interpret, and practical frictions — the risk of losing the device, logistical hurdles, and the many steps required to enrol and transfer patients between settings — further impeded integration into clinical workflows.

Patients echoed many of these concerns. Thirteen of the fifteen who answered the sensor-related survey questions said the system offered them no added value, and only seven said they fully or partially trusted the sensor. Some were unsure why they were wearing it at all, with answers ranging from “for research” to “I don’t know, I’m in a wheelchair.” Nearly half reported that their care professionals did not provide adequate support, and several wished they could review the results together with a therapist. Still, the picture was not uniformly negative: eight patients found the system easy to use, five said they would recommend it, and one noted that it created a sense of external accountability for staying active.

When asked what a better intervention would look like, the professionals were clear: it should be easy to use, have a clearly defined objective, incorporate technology that provides direct visual or tactile feedback, and show patterns, intensity, type, and timing of activity over time. They also stressed preconditions that go beyond the device itself — educating patients and informal caregivers about the risks of sedentary behaviour, conducting a pre-implementation analysis of needs and available technology, involving the entire multidisciplinary team, and cultivating a rehabilitation climate that invites and makes it safe to be physically active, with staff acting as role models. The study’s authors conclude that sensor-based feedback in geriatric rehabilitation can indeed be feasible and valuable, but only if measurement accuracy improves, the technology is tailored to individual patient goals, its added value becomes evident to users, and support is provided to both patients and professionals. In other words, the hardware may be ready; the harder engineering challenge is human — designing eHealth that fits the routines, abilities, and motivations of the older adults and clinicians it is meant to serve.

Subject of Research: Usability and feasibility of wearable sensor-based feedback on sedentary behaviour in geriatric rehabilitation

Article Title: Usability and feasibility of a wearable sensor-based monitoring system providing feedback on sedentary behaviour during inpatient and home-based geriatric rehabilitation

Article References: Debeij, S. M., Haaksma, M. L., de Waal, M. W. M., van Haastregt, J. C. M., Pol, M. C., Kasteleyn, M. J., & van Dam van Isselt, E. F. (2026). Usability and feasibility of a wearable sensor-based monitoring system providing feedback on sedentary behaviour during inpatient and home-based geriatric rehabilitation. European Geriatric Medicine. https://doi.org/10.1007/s41999-026-01616-6

Image Credits: AI Generated

DOI: 10.1007/s41999-026-01616-6

Keywords: wearable sensors, sedentary behaviour, geriatric rehabilitation, eHealth, physical activity, older adults, feasibility, usability, behaviour change, digital health, home-based rehabilitation, patient adherence