Why breathing signals might help wearables stage sleep
Sleep stages differ in breathing patterns: breathing becomes irregular and shallow during REM sleep, and more regular during deep non-REM sleep. Wrist devices that only track movement and pulse can miss these breathing changes, so adding a respiratory sensor could give the algorithm another clue for distinguishing sleep stages. One early study estimated respiratory rate from wrist actigraphy and used it to detect REM sleep, showing that breathing information from a wrist device can align with polysomnography—the gold-standard sleep lab test [1].
However, that study is abstract-only and does not report how much the respiratory signal improved staging over movement alone. The idea is plausible, but the size of the benefit remains unclear.
What wrist wearables can already do without breathing sensors
Wrist devices that combine movement (accelerometry) and pulse (photoplethysmography, or PPG) can stage sleep with moderate agreement to polysomnography. In a clinical population of 244 adults and 48 children with various sleep disorders, a wrist PPG-plus-accelerometry algorithm achieved 76.4% accuracy and a kappa of 0.62 for four sleep stages—where kappa measures agreement beyond chance, with 0.6 often considered substantial [2]. Performance was higher in children and in people with non-REM parasomnias, and lower in those with REM parasomnias [2].
A scoping review of 62 wearable setups found that devices using only accelerometer data are good at telling sleep from wake but poor at identifying multiple sleep stages, while adding PPG improves staging [4]. That review did not include respiratory sensors, so it shows the baseline that respiratory measurement would need to beat.
What is still missing to answer the question
No study in this set directly compares a wrist or finger wearable with and without added respiratory measurement for sleep staging or disorder detection. The closest evidence is indirect: wrist actigraphy with estimated respiratory rate can detect REM sleep [1], wrist PPG plus accelerometry stages sleep moderately well without breathing sensors [2], and a wrist pulse oximeter can measure oxygen saturation relevant to apnea [5]. Each piece supports a possible role for respiratory signals, but none measures the added value.
The scoping review also notes that wearable sleep staging reliability depends on validation across diverse populations and consistent reporting, and it recommends comparing commercial algorithms across multiple sleep stages [4]. Until studies test respiratory-enhanced wearables head-to-head against the same devices without respiratory sensing, the improvement remains suggested rather than demonstrated.
Sources used in this answer
Actigraphic detection of REM sleep based on respiratory rate estimation
Wrist actigraphy with estimated respiratory rate showed potential for detecting REM sleep compared with polysomnography, but the abstract does not quantify the improvement over movement alone.
It is All in the Wrist: Wearable Sleep Staging in a Clinical Population versus Reference Polysomnography
A wrist-worn PPG and accelerometry algorithm achieved 76.4% accuracy and kappa 0.62 for four-class sleep staging in a clinical population, with better performance in children and non-REM parasomnias and worse in REM parasomnias.
Estimating the severity of obstructive sleep apnea using ECG, respiratory effort and neural networks
A neural network using ECG and respiratory effort estimated obstructive sleep apnea severity and performed sleep staging, but the study did not use wrist or finger wearables for respiratory measurement.
Evaluating reliability in wearable devices for sleep staging
A scoping review of 62 wearable setups found that accelerometer-only devices are effective for sleep/wake detection but not multiple sleep stages, while adding PPG improves staging; respiratory sensors were not included.
Performance evaluation of a wrist-worn reflectance pulse oximeter during sleep
A wrist-worn reflectance pulse oximeter was evaluated during sleep against polysomnography for oxygen saturation and apnea-related indices, but the abstract does not report how closely it matched the reference.