Fitness Tracker Heart-Rate Sensors: Why Wrist-Based PPG Struggles, By Design
Peer-reviewed testing against ECG and chest straps consistently finds wrist heart-rate sensors less accurate during exercise — and the reason is physical, not a software bug.
How PPG measures a pulse without touching the blood
Nearly every consumer wearable's heart-rate sensor uses PPG (photoplethysmography): an LED shines light into the skin, and a photodetector measures how much of that light is reflected back. Blood volume in the microvascular tissue changes slightly with each heartbeat, which changes how much light gets absorbed versus reflected — and that subtle, rhythmic fluctuation is what the sensor interprets as a pulse. It's a genuinely clever, non-invasive technique, and the underlying principle is sound.
Why the wrist specifically struggles
The problem isn't the PPG technique itself — it's that the wrist is a particularly difficult location to apply it. Wrist skin sits over a combination of tendon, bone, and comparatively thin, mobile tissue, which makes the reflected-light signal more vulnerable to motion artifacts (the sensor picking up light changes from the wrist physically moving, not just blood volume changing) and to variable skin contact as the band shifts during activity. A chest strap, by contrast, uses direct ECG (electrocardiogram) electrodes reading the heart's actual electrical signal at the source, with stable skin contact against a flatter, less mobile part of the body — a fundamentally more direct and stable measurement.
What peer-reviewed testing actually found
Multiple independent, peer-reviewed studies comparing wrist PPG devices against ECG or chest-strap references have found consistent patterns:
- A Cleveland Clinic study testing four popular wrist devices (Apple Watch, Fitbit Blaze, Garmin Forerunner 235, TomTom Spark Cardio) against a continuous 4-lead ECG and chest monitor across treadmill, stationary bike, and elliptical activities found that wrist-worn monitors were less accurate than a standard chest strap, and that the devices could both over- and underestimate heart rate — not a consistent one-directional bias.
- A systematic review of chest-worn versus wrist-worn sensors found that chest ECG devices (specifically citing the Polar H10) maintain accuracy across different exercise intensities, while wrist PPG devices show larger errors, particularly in beat-to-beat analysis and heart rate variability estimation.
- A study specifically testing accuracy across different physical activities found accuracy dropping to its lowest levels (a concordance correlation coefficient under 0.50 — meaningfully poor agreement) during short, high-intensity, full-body movements like burpees, attributing the drop specifically to motion artifacts.
These findings held consistently across multiple studies and multiple brands (Apple, Fitbit, Garmin, TomTom, Polar, Whoop all appear across the cited research) — the pattern is a property of wrist-based PPG measurement generally, not a defect specific to any one manufacturer's implementation.
Placement matters as much as the device itself
One of the more actionable findings across this research: anatomical placement significantly affects accuracy, independent of the specific device. Testing identical sensor units at different body positions found the upper arm consistently outperforming the wrist and forearm, with one study finding a forearm-mounted sensor (Verity Sense) and an upper-arm-worn device (Whoop) showing the closest agreement with a chest-strap reference among all positions tested. The upper arm has more stable, less mobile tissue and more consistent band pressure during activity than the wrist — the same fundamental issue that makes the wrist harder to measure accurately in the first place.
When wrist PPG is actually reliable
The research is consistent on this point too: wrist PPG performs well in a specific, narrower set of conditions:
- At rest — with minimal motion, wrist PPG has been found accurate and reliable, with mean absolute error as low as ~3 bpm against a chest-strap reference in some testing.
- During steady-state, moderate exercise (a consistent pace on a bike or treadmill, no sudden intensity changes) — accuracy holds up meaningfully better than during variable or high-intensity, full-body movement.
- Data coverage (the percentage of time a device reports a reading within 10 bpm of a reference) has been measured as high as 92–98% across various activity types in some studies — meaning the device is usually close, with accuracy problems concentrated in specific harder conditions rather than being wrong all the time.
The clinical verdict, from the people who need accuracy most
For most fitness tracking — pacing a casual workout, monitoring general activity trends — the wrist's accuracy limitations during intense exercise are a minor inconvenience. For anyone managing a diagnosed heart condition, the clinical guidance is more direct: as one study's lead cardiologist put it, if you need to know your heart rate with precision — training within a doctor-set safe limit due to coronary artery disease or heart failure, for example — wrist-worn monitors are less accurate than a standard chest strap, and shouldn't be relied on as a substitute for one in that specific context.
Practical guidance for anyone tracking heart rate
- Trust wrist PPG more at rest and during steady, moderate exercise than during high-intensity, full-body, or rapidly-changing activity.
- Consider an upper-arm band over a wrist device specifically if accuracy during exercise matters to you — the research consistently favors that placement.
- Use a chest strap for anything safety-critical — training near a medically set heart-rate limit, or any use case where a real-time accurate number matters more than a general trend.
- Don't over-interpret single readings during erratic movement (burpees, boxing, rapid direction changes) — this is precisely where every device type in the research showed its largest errors, regardless of brand or price.
Bottom line
Wrist-based heart-rate tracking isn't inaccurate because of a specific brand cutting corners — the physical difficulty of getting a stable optical reading from a mobile, bony part of the body is a consistent limitation found across every major manufacturer in peer-reviewed testing. It's genuinely good at rest and during steady exercise, and genuinely less reliable during sudden, high-intensity, full-body movement — and if precise, medically relevant heart-rate data is what you need, the research points consistently toward a chest strap, not a firmware update, as the actual fix.