How addressing sleep, not just detecting it, is helping senior living communities reduce falls and shift dementia care from reactive to preventive.
A collaborative piece from Withings Health Solutions and Composure.
Senior living has spent decades optimizing the response to a fall: faster alarms, better sensors, more frequent rounding, tighter documentation after an incident. Those investments matter, but by the time an alarm sounds, the opportunity to prevent the fall may have already passed.
What if care teams could see risk building hours earlier?
Sleep disruption, nighttime agitation, and unusual bed exits can provide a window into what is happening before an incident, particularly for residents living with dementia. That is why Composure and Withings Health Solutions are exploring a different model: using passive sleep and movement data not simply to document what happened, but to understand risk earlier and measure whether interventions are actually working.
Composure's network audio solution delivers personalized sound environments designed to improve sleep quality for residents living with dementia. Withings Health Solutions' under-mattress sleep mat measures changes in sleep and nighttime behavior over time, objectively and without disrupting a resident's routine. Paired together, the two point toward a model of dementia care built around what happens before a crisis, not just how fast staff can respond to one.
Why fall risks look different in dementia
Falls are not a uniform problem in senior living. For residents living with dementia, cognitive decline, disrupted circadian rhythms, nighttime agitation, and shifting sleep-wake cycles compound the risk. A resident who wakes disoriented at 2am, unsure of where they are or why, faces meaningfully higher odds of an unassisted, unsafe bed exit than a cognitively intact resident waking for the same reason. Medications commonly used to manage dementia-related symptoms add another layer of complexity.
That's why falls deserve to be treated as a dementia-specific problem, not a general senior-living one. And prevention pays off beyond the incident count: calmer nights, less disorientation, and fewer residents living through the fear, confusion, and physical harm that so often follow a fall, on top of the disruption it brings to someone already struggling to make sense of their surroundings.
The warning signs may start at night
Falls are typically measured after the fact: an incident report, a liability entry, a line in a quality-of-care review. But data from sleep and nighttime movement monitoring tells a different story: falls are often the visible endpoint of a pattern that began well before someone got out of bed. Unplanned bed exits, especially those outside typical sleep-wake windows or following a night of poor, fragmented sleep, are strongly associated with next-day fall risk, a connection that's especially pronounced in residents with dementia, whose sleep is already more fragile, and more predictive.
A recent 90-day pilot in a memory care unit at a New York-based continuing care retirement community put that link to the test. Among 47 residents, median nightly bed exits dropped 30% and fall incidents fell 45%; more than half of residents flagged as high fall-risk had zero falls during the pilot. Sleep quality moved too: sleep efficiency, how effectively residents actually sleep during the time they spend in bed, improved for 62% of participants, with an average gain of about 5%, at the high end of the natural variation typical for this population, and considerably larger for several residents.
Seen this way, fall prevention stops being only about how many people are watching, and becomes about noticing the right signal early enough to act on it. The question shifts from "how quickly can we respond?" to "what could tell us to intervene sooner?"
A new way of proving what works
"The technology to monitor sleep passively, every night, has existed for years. What senior living has lacked is proof. For non-drug interventions, operators have mostly had to rely on hope. Pairing Composure's approach with objective nightly data changes that: now they can measure whether it's working. And in healthcare, what you can measure is what gets adopted" said Antoine Pivron, VP, Health Solutions, Withings.
That kind of measurement is what makes Composure's model different from most non-drug interventions in senior living, which have long been considered promising but hard to validate at scale. A baseline period establishes each resident's sleep patterns before Sound Blanket is introduced; the same measures are tracked afterward to see what changes. It's the kind of evidence operators will increasingly expect from any approach that claims to help residents with dementia.
Once installed and tuned to a resident's care plan, Sound Blanket runs on its own, every night: no app to open, no device to charge, nothing for staff to switch on or check. Nightly sleep and movement data feeds into regular reports Composure shares with each community's care team, turning nightly patterns into something the team can act on. Rolling it out takes minimal construction or IT lift: rooms are outfitted, a short baseline period runs before the intervention begins, and staff get a quick walkthrough rather than a training course. From there, Composure stays engaged with the community to make sure it's getting full value from what's been deployed.Turning sleep data into actionable signalsFor residents with dementia, how data is collected matters as much as the data itself. Wearables and technologies that require charging, remembering, or active participation are difficult to use consistently with this population. The Withings sleep sensor sits unobtrusively under the mattress, capturing nighttime data without requiring a resident to wear anything or change their routine.That data can provide insight into several aspects of nighttime behavior, including:
- Bed exits: frequent or unusual bed exits are often a sign of disrupted sleep itself, and understanding their frequency and timing can help care teams identify changes in nighttime behavior and potential periods of increased risk.
- Sleep efficiency and sleep latency: tracking how easily residents settle into sleep and how consistently they remain asleep can help demonstrate whether nighttime patterns are changing.
- Sleep stages: changes in light, deep, and REM sleep provide additional context on overall sleep architecture, an area Composure and Withings continue to study as measurement approaches for this population mature.
The goal isn't to overwhelm care teams with more data. It's to surface a small number of meaningful signals that show whether something is changing, and whether an intervention is having its intended effect. That data feeds back in two directions: at the resident level, it informs periodic adjustments to how each Sound Blanket experience is calibrated to that person's shifting sleep patterns; at the product level, aggregate data across communities shapes how Composure refines Sound Blanket itself, its sound design, delivery model, and measurement approach improving as the dataset grows.
The opportunity goes beyond fall prevention
Fall reduction understandably gets significant attention from senior living operators; falls carry substantial consequences for residents, staff, families, and communities. But better understanding what happens overnight has implications beyond falls. As sleep and nighttime agitation are better understood and addressed, communities may also see changes in psychotropic medication use, nighttime behavior, staff burden, and overall resident well-being.
At one community that has used Sound Blanket for over 18 months, longer-term data shows antipsychotic use down 19% and antianxiety/hypnotic use down 34%, on top of the community's standard Gradual Dose Reduction practices, with behavioral incident rates holding steady. Across Composure's partner communities more broadly, care teams have consistently reported minimal disruption to existing routines: no added training, no new protocols, just a calmer nighttime environment.
If these patterns hold as more communities adopt this approach, they point to something bigger: sleep may be one of the most overlooked windows into the health and well-being of residents living with dementia, one that could help address not just fall risk, but some of the broader challenges that make dementia care so complex.
Moving care upstream
"Poor sleep in dementia care is one of those things that has been a big problem for so long that, for most people, it no longer looks like a problem that can be solved, just something to work around. We're seeing that's simply not true, and that addressing it head-on non-pharmacologically is one of the few clinical levers that can produce numerous positive outcomes and provide a better quality of life for people living with dementia." – Jeff McSpadden, CEO, Composure
As passive monitoring becomes easier to deploy and integrate into everyday care, senior living has a chance to move further upstream: from documenting incidents to understanding patterns, from responding to risk to catching it earlier, and from assuming an intervention works to proving that it does.