Does Your Wearable Actually Track Breathing Rate While You Move?

This article covers what published validation research shows about wearable respiratory rate accuracy specifically during walking, running, and other exercise conditions. It does not cover resting or sleep accuracy in depth, and it does not address clinical deterioration monitoring, which is a separate question with separate evidence.

Different sensor types handle exercise-induced motion very differently, and the research shows this isn't a solved problem with one clear winner. Strain sensors placed on the chest or abdomen have held up reasonably well against reference airflow measurements during walking, while systems that rely on movement or vibration alone remain more vulnerable to the noise that exercise itself introduces.

The number on the screen versus what's actually happening in the sensor

There's a specific kind of doubt that shows up once someone starts paying attention to their wearable's breathing rate number during a run or a fast walk. Heart rate feels trustworthy by now, people have watched it climb and fall in ways that match how hard they're working. Breathing rate is newer territory, and it behaves oddly on screen sometimes: jumping around, lagging behind effort, or looking suspiciously smooth during exactly the moments when breathing should be ragged.

That suspicion isn't unfounded, and it maps onto something mechanical. Respiratory rate is harder to pull off a moving body than heart rate is, because the chest wall movement most sensors are trying to detect is the same chest wall movement that arm swing, foot strike, and torso rotation are also disturbing.

What changes when the body starts moving with intention

Static breathing and exercise breathing aren't just different speeds of the same thing. An inertial-sensor system built with three units, placed on the thorax, abdomen, and lower back, found statistically significant differences in respiratory rate readings between static postures and dynamic activities like slow walking, fast walking, running, and cycling. That's a useful confirmation that the sensing problem genuinely shifts once a body is in motion, not just that the numbers happen to look different.

The Zephyr BioHarness, a chest-worn device, was tested for test-retest reliability at rest, during a submaximal aerobic fitness test, and through recovery. Reliability during the aerobic test itself landed in a similar range to resting reliability. But the measurement error (SEM) was notably wider during exercise than at rest, which is the same gap most of the forum-style questions above are circling: the device isn't necessarily wrong, but it may be less precise exactly when someone wants precision most.

For anyone comparing a wrist-worn device against a chest strap for this specific purpose, the placement question is worth separating from the sensor-type question, since how wrist and chest placement affect respiratory rate accuracy turns out to be a distinct issue from which underlying technology is doing the sensing.

The bigger picture: PPG and ECG-derived breathing rate weren't built with exercise in mind

A lot of wrist and finger wearables estimate breathing rate indirectly, pulling it out of the same photoplethysmogram signal used for heart rate, rather than measuring chest movement directly. A review of algorithms for estimating breathing rate from ECG and PPG signals describes more than 100 proposed methods, and notes these algorithms create an opportunity for unobtrusive measurement in both healthcare and fitness contexts. The review itself, though, is framed around opportunity and methodology rather than confirmed exercise accuracy.

A related assessment that systematically compared these ECG- and PPG-derived algorithms against a reference measure did so under what it describes as ideal conditions, healthy participants at rest, comparing performance to impedance plethysmography as the clinical standard. That's a meaningfully different test condition than a treadmill or a trail. Extending confidence from ideal, resting comparisons to sweaty, arm-swinging exercise is exactly the kind of leap the underlying evidence doesn't make for us.

A broader review of contact-based respiratory rate methods lays out the working principles behind airflow, sound, chest-movement, and cardiac-modulation approaches, and frames respiratory rate as a vital sign of interest across clinical, occupational, and sporting settings, without claiming any single contact method has solved the exercise-motion problem outright. Readers curious about the wider accuracy question beyond exercise specifically may find whether wearables actually measure respiratory rate accurately a useful next stop, since it covers ground this article intentionally narrows away from.

Common questions