Fitness tracking devices vs conventional methods of evaluating sleep
| Device | Comparator | N | Study population | Results |
|---|---|---|---|---|
| Fitbit13 | Polysomnography and actigraphy | 24 | Healthy adults with no history of symptoms of sleep disorders Mean age 26.1 | Sensitivity for sleep 97.8% Specificity for wakefulness 19.8% Overestimated total sleep time and sleep efficiency and underestimated wake time after sleep onset |
| Fitbit Ultra14 | Polysomnography and actigraphy | 63 | Children and adolescents undergoing overnight clinical polysomnography Mean age 9.7 | ‘Normal’ mode overestimated total sleep time and sleep efficiency and underestimated wake time after sleep onset ‘Sensitive’ mode underestimated total sleep time and sleep efficiency and overestimated wake time after sleep onset |
| Jawbone UP15 | Polysomnography | 65 | Healthy adolescents Mean age 15.8 | Overestimated total sleep time and sleep efficiency and underestimated wake time after sleep onset, no difference in sleep onset latency No clear correlation between ‘light’ and ‘deep’ sleep and conventional polysomnographic sleep stages |
| Jawbone UP16 | Polysomnography | 28 | Midlife women Mean age 50.1 | Sensitivity for sleep 96% Specificity for wakefulness 37% Overestimated total sleep time and sleep onset latency and underestimated wake time after sleep onset |
| Jawbone UP17 | Polysomnography and actigraphy | 64 | Children and adolescents with suspected sleep-disordered breathing Mean age 8.4 | Sensitivity for sleep 92% Specificity for wakefulness 66% No difference from polysomnography on total sleep time, sleep efficiency, sleep onset latency, and wake time after sleep onset Compared with actinography, overestimated sleep onset latency and underestimated wake time after sleep onset |