Your Smartwatch May Undercount Steps If You Use a Cane or Walker
Two studies asked the same question with different assistive devices, and got meaningfully different answers depending on which one.
This piece covers step-count accuracy specifically during the use of assistive walking devices (walkers, canes, oxygen carts), distinct from step-count accuracy in people with movement disorders as a broader category. It covers wrist-worn consumer devices in adult populations.
Two recent studies tested whether assistive walking devices throw off a wearable's step count, and found different degrees of the same underlying problem. A cross-sectional study of adults using a rolling walker or pushing an oxygen cart found dramatic step undercounting on a wrist-worn Apple Watch, severe enough that the authors describe real potential for clinical misinterpretation. A separate, smaller study testing a standard cane found accuracy held up well during cane use, with the more noticeable disruption coming from a two-wheeled walker specifically. The device matters. A cane is not the same test as a rolling walker or an oxygen cart, and the two studies together suggest the wrist accelerometer's confusion is tied to how much the assistive device changes the arm's normal walking motion.
The undercounting problem, and why it's not just about movement disorders
Step counting on a wrist-worn wearable works by detecting a specific, repeated arm-swing pattern associated with walking. Anything that changes how the arms move during walking has the potential to break that detection, independent of whether the person has any underlying movement disorder at all. Someone recovering from a broken ankle and using a cane temporarily, or an older adult who relies on a rolling walker day to day, both introduce that same kind of motion change.
That's a genuinely different question from step-count accuracy in people with diagnosed movement disorders, which involves the disorder's own effect on gait and motor control. This is specifically about the assistive device itself, independent of any underlying condition, a distinction with real practical relevance since assistive-device use is common for reasons that have nothing to do with a chronic movement disorder.
2 studies
- In 42 community-dwelling adults aged 51 to 80, an Apple Watch Series 8 showed dramatically impaired step detection and significant undercounting when participants used a rolling walker or pushed an oxygen cart, with the authors warning of real potential for clinical misinterpretation of the resulting activity data.
- In 11 participants navigating an obstacle course, Fitbit step counts recorded at the wrist, hip, and ankle were significantly correlated with true step counts (captured via video) during both a standard cane trial and a no-device control trial, with the more disruptive result specifically tied to a two-wheeled walker rather than the cane.
Why a walker breaks the pattern more than a cane does
A standard cane is typically used in one hand, alternating with the opposite leg's stride, close to a natural walking rhythm. A rolling walker requires both hands on a frame that moves forward with the whole body, changing the arm swing far more fundamentally, and an oxygen cart adds another item to push or pull with at least one arm. That's a plausible mechanical explanation for why the PeerJ study's walker-and-cart population saw dramatic undercounting while the cane trial in the smaller study held up.
The gap between the two studies is also a reminder that 'assistive device' isn't one uniform category for a wrist accelerometer. A rolling walker changes arm motion in a way a standard cane may not, at least based on what these two specific comparisons found. Someone using a cane occasionally may see reasonably accurate counts; someone relying on a rolling walker or pushing an oxygen cart daily has real reason to expect their wrist-worn device is missing a meaningful share of their actual steps.
Neither study tested every assistive-device type, and the cane-trial study had only 11 participants. The direction of the finding, walker use disrupting wrist detection more than cane use, is a real, specific result from these two studies, not a comprehensive map of every mobility aid's effect.
How this differs from the existing movement-disorders question
The site's existing coverage of whether fitness trackers are accurate for people with movement disorders addresses a broader population where the disorder itself, not an assistive device, is the primary variable affecting gait. These two new studies isolate a narrower, device-specific mechanism: even someone with typical gait mechanics, using a walker or cart for an unrelated reason (recovery from injury, general stability support, portable oxygen), can see the same category of undercounting.
For anyone relying on a wrist-worn device's step count while using a rolling walker or oxygen cart specifically, mostly, the practical takeaway from the PeerJ study is to treat that number with real skepticism rather than as a reliable activity record, since the reported undercounting was severe enough that the researchers flagged genuine potential for clinical misinterpretation if that data were used to guide care decisions.
Common questions
Does using a cane throw off a smartwatch's step count?
One study found step counts stayed significantly correlated with actual steps during standard cane use, similar to a no-device control trial. The bigger disruption in that same study came from a two-wheeled walker, not the cane.
Does a rolling walker or oxygen cart affect step tracking more than a cane?
Based on these two studies, yes. A cross-sectional study of adults using rolling walkers and oxygen carts found dramatic step undercounting on a wrist-worn device, more severe than what a separate study found for standard cane use.
Is this the same issue as inaccurate tracking in people with movement disorders?
No. These studies isolate the assistive device itself as the variable, independent of any underlying movement disorder, testing whether the device's effect on arm motion disrupts wrist-based step detection regardless of why someone is using it.
Should I trust my step count while using a rolling walker?
The evidence here suggests real caution is warranted. The study testing rolling walkers and oxygen carts found the undercounting severe enough to raise concern about clinical misinterpretation if that data were used to inform care decisions.