Performance of Widely used Consumer Wearables for Sleep Tracking

Sleep tracking by consumers is becoming increasingly prevalent, yet few studies have evaluated the accuracy of such devices. We sought to evaluate the accuracy of three devices (Oura Ring Gen3, FitBit Sense 2, and Apple Watch Series 8) compared to the gold standard sleep assessment (polysomnography [PSG]). Thirty-five participants (aged 20-50 years) without a sleep disorder were enrolled in a single-night inpatient study during which they wore the Oura Ring, FitBit, Apple Watch, and were monitored with PSG. For detecting sleep vs. wake, sensitivity was <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{\geq 9 5 \%}$</tex> for all devices. For discriminating between sleep stages, sensitivity ranged from <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{5 0 \% - 8 6 \%}$</tex>, as follows: Oura ring sensitivity <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{7 6 . 0 \% - 7 9 . 5 \%}$</tex> and precision <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$77.0 \%{-} \mathbf{7 9 . 5 \%}$</tex>; FitBit sensitivity <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$61.7 \%-78.0 \%$</tex> and precision <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$72.8 \% \mathbf{7 3 . 2 \%}$</tex>; and Apple sensitivity 50.5\%-86.1\% and precision 72.7\%<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{8 7 . 8 \%}$</tex>. Oura ring was not different from PSG in wake, light sleep, deep sleep, or REM sleep estimation. Fitbit overestimated light (18 minutes; <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{p}&lt;\mathbf{0 . 0 0 1}$</tex>) and underestimated deep (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{1 5}$</tex> minutes; <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{p}&lt;\mathbf{0 . 0 0 1}$</tex>) sleep. Apple underestimated duration of wake (7 minutes; <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{p}&lt;\mathbf{0 . 0 1}$</tex>) and deep (43 minutes; <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$p&lt;0.001$</tex>) sleep and overestimated light (45 minutes; <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{p}&lt;\mathbf{0 . 0 0 1}$</tex>) sleep. In adults with healthy sleep, all devices were similar to PSG in estimation of sleep duration, with the devices also showing moderate to substantial agreement with PSGderived sleep stages.