Feasibility and validity of accelerometer measurements to assess physical activity in toddlers

Autor: De Bourdeaudhuij Ilse; Gubbels Jessica; Van Cauwenberghe Eveline; Cardon Greet
Sprache: Englisch
Veröffentlicht: 2011
Quelle: Directory of Open Access Journals: DOAJ Articles
Online Zugang: http://www.ijbnpa.org/content/8/1/67
https://doaj.org/toc/1479-5868
doi:10.1186/1479-5868-8-67
1479-5868
https://doaj.org/article/00cb2c4e17454475b1f4971fa02f6135
https://doi.org/10.1186/1479-5868-8-67
https://doaj.org/article/00cb2c4e17454475b1f4971fa02f6135
Erfassungsnummer: ftdoajarticles:oai:doaj.org/article:00cb2c4e17454475b1f4971fa02f6135

Zusammenfassung

Abstract Background Accelerometers are considered to be the most promising tool for measuring physical activity (PA) in free-living young children. So far, no studies have examined the feasibility and validity of accelerometer measurements in children under 3 years of age. Therefore, the purpose of the present study was to examine the feasibility and validity of accelerometer measurements in toddlers (1- to 3-year olds). Methods Forty-seven toddlers (25 boys; 20 ± 4 months) wore a GT1M ActiGraph accelerometer for 6 consecutive days and parental perceptions of the acceptability of wearing the monitor were assessed to examine feasibility. To investigate the validity of the ActiGraph and the predictive validity of three ActiGraph cut points, accelerometer measurements of 31 toddlers (17 boys; 20 ± 4 months) during free play at child care were compared to directly observed PA, using the Observational System for Recording Physical Activity in Children-Preschool (OSRAC-P). Validity was assessed using Pearson and Spearman correlations and predictive validity using area under the Receiver Operating Characteristic curve (ROC-AUC). Results The feasibility examination indicated that accelerometer measurements of 30 toddlers (63.8%) could be included with a mean registration time of 564 ± 62 min during weekdays and 595 ± 83 min during weekend days. According to the parental reports, 83% perceived wearing the accelerometer as 'not unpleasant and not pleasant' and none as 'unpleasant'. The validity evaluation showed that mean ActiGraph activity counts were significantly and positively associated with mean OSRAC-P activity intensity (r = 0.66; p < 0.001; n = 31). Further, the correlation among the ActiGraph activity counts and the OSRAC-P activity intensity level during each observation interval was significantly positive (ρ = 0.52; p < 0.001; n = 4218). Finally, the three sedentary cut points exhibited poor to fair classification accuracy (ROC-AUC: 0.56 to 0.71) while the three light PA (ROC-AUC: 0.51 to 0.62) and the ...