skip to main content
Idioma:

0322 Field-Based Sleep Measurement: Concordance Between Commercial Activity Monitors and an Actigraph

Kubala, A G ; Santos, E C ; Barone Gibbs, B ; Buysse, D J ; Patel, S R ; Hall, M H ; Kline, C E

Sleep (New York, N.Y.), 2018-04, Vol.41 (suppl_1), p.A123-A124 [Periódico revisado por pares]

US: Oxford University Press

Texto completo disponível

Citações Citado por
  • Título:
    0322 Field-Based Sleep Measurement: Concordance Between Commercial Activity Monitors and an Actigraph
  • Autor: Kubala, A G ; Santos, E C ; Barone Gibbs, B ; Buysse, D J ; Patel, S R ; Hall, M H ; Kline, C E
  • Assuntos: Sleep
  • É parte de: Sleep (New York, N.Y.), 2018-04, Vol.41 (suppl_1), p.A123-A124
  • Descrição: Abstract Introduction Many commercial activity monitors (CAM) now measure sleep-wake patterns. However, it is unknown whether these CAMs provide sleep data that are similar to research-grade actigraphy. The purpose of this study was to examine the concordance between multiple CAMs and a validated research-grade actigraph. Methods On the same wrist, 28 healthy adults (50% female, 25.0 ± 4.2 y) wore an Actiwatch Spectrum (AW) and alternated wearing 6 CAMs for one 24-h period (Samsung Gear Fit2, Fitbit Alta, Polar A360, Jawbone Up3, Xiaomi Mi Band 2, Mistfit Shine 2). A daily sleep diary was also completed. Sample sizes for comparisons ranged from 18–27 due to CAM device error. Unedited AW (default setting) and CAM outputs were used for comparisons. Comparison variables included total sleep time (TST) and wake after sleep onset (WASO). Direct comparisons between single-night AW and CAM data were made via paired t-tests, mean absolute percent error (MAPE) calculations and intra-class correlations (ICC). Results On average, Fitbit, Jawbone, Misfit, and Xiaomi Mi significantly overestimated TST relative to AW (54.4–76.7 min, p≤.007), while Polar underestimated TST (67.3 min, p≤.002). Samsung TST did not statistically differ relative to AW (-31.6 min, p=.22). WASO was significantly underestimated by all devices (17.2–29.0 min, p≤.005) except for the Polar which overestimated (29.8 min, p≤.001) and the Jawbone which did not statistically differ from AW (2.3 min, p=.78). MAPEs ranged from 13.9% (Samsung) to 24.5% (Misfit) for TST and 38.5% (Fitbit) to 127.7% (Polar) for WASO. TST ICCs ranged from .17 (Polar) to .83 (Fitbit), while WASO ICCs ranged from .26 (Samsung) to .45 (Fitbit). Conclusion MAPE ranges indicated a high magnitude of error relative to AW, while ICCs indicated small-moderate agreement between CAMs and AW. Samsung and Jawbone were the only CAMs that did not statistically differ from AW for average TST or WASO, respectively. These data suggest that the sleep tracking ability of CAMs is questionable in comparison to an Actiwatch. Future research will evaluate whether concordance is better between expert-edited AW and CAM data. Support (If Any) NIH K23 HL118318 (PI: Kline).
  • Editor: US: Oxford University Press
  • Idioma: Inglês
Voltar para lista de resultados

  • Realização: ABCD-USP USP   Logos de Redes Sociais: Freepik

Buscando em bases de dados remotas. Favor aguardar.