Why can a wrist or fingertip sensor give a different reading while its wearer is walking? The device is measuring through a moving contact between sensor and body. Motion changes pressure, alignment, and the signals reaching its detector. Separating those changes from genuine physiology is a measurement problem, not simply a matter of adding more data.
A sensor sees a signal, not its meaning
Wearables use sensors to collect information such as heart rate, skin electrical characteristics, temperature, and movement. Digital health systems can transmit those measurements for monitoring or self-management [1]. But the sensor directly records a physical signal; deciding what that signal means requires processing and context.
For a pulse oximeter, red and near-infrared light pass through a fingertip to a detector. Oxygenated and deoxygenated blood absorb those wavelengths differently, and the device estimates oxygen saturation from their ratio [2]. If the finger shifts, the light path and contact can change at the same time as the blood signal.
Motion creates a competing pattern
A sensor’s output during movement combines the intended physiological signal with effects from motion. In pulse oximetry this is called motion artifact, and it can disturb readings [2]. Imagine trying to measure a small wave in a bowl while someone jostles the table: the detector receives water motion, but some comes from the disturbance rather than the wave you care about.
This interference can be especially consequential when a person’s oxygen level changes during activity. A device that performs acceptably at rest may have difficulty distinguishing the oxygen-related signal from movement-related variation [2]. A stable-looking number is not automatically proof of a stable underlying condition, just as a noisy number does not by itself identify the cause of the noise.
The body and device both shape accuracy
Motion is one source of error, but not the only one. Poor fingertip blood flow and nail polish can reduce pulse-oximeter accuracy; blood loss and constricted vessels can also interfere [2]. Skin pigmentation matters because melanin absorbs some red and near-infrared light, and the cited source describes a tendency for pulse oximeters to overestimate oxygen saturation in darker skin tones [2].
Different sensors face different physical constraints. A wearable lung patch described by NIBIB uses a microchip sensor to detect tiny vibrations associated with lung sounds, while a learning model helps classify them; this research approach does not mean every wearable diagnoses disease [1]. The measurement method, placement, signal processing, and validation all shape what a device can responsibly report.
Better monitoring needs tested measurements
One proposed pulse-oximeter research direction uses green light reflected from the skin rather than red and near-infrared light transmitted through the finger. Researchers described investigating whether this approach and calibration for skin pigment could improve measurement across patients [2]. It is a research effort, not evidence that all motion or accuracy problems are solved.
Continuous data can reveal patterns that a single spot measurement misses, and digital systems may share information with health professionals [1]. But duration does not erase measurement error: repeated biased or motion-disturbed readings can still mislead. Reliable interpretation depends on understanding what was measured, under what conditions, and how the device was evaluated. Movement makes wearable sensing difficult because the body is both the source of the signal and part of the measurement path.
Explore the physics: an explanatory diagram
Sources & further reading
- Digital Health NIBIB
- Updating Pulse Oximeters NHLBI
Prepared for the independent PSI Center reading library. About edition dates and source checking.
This explainer describes scientific principles. It does not assess individual treatment choices.
