Kate Lyden1, Dinesh John, Philippa Dall, Malcolm H Granat. 1. 1Division of Endocrinology, Metabolism, and Diabetes, University of Colorado, Anschutz Medical Campus, Aurora, CO; 2School of Health Sciences, Northeastern University, Boston, MA; 3Department of Health Sciences, Glasgow Caledonian University, Glasgow, UNITED KINGDOM; 4School of Health and Life Sciences, University of Salford, Manchester, UNITED KINGDOM.
Abstract
UNLABELLED: A triaxial accelerometer worn on the thigh can provide information on the angle of rotation of the thigh. These data may be used to estimate periods of lying versus sitting. PURPOSE: To develop and test a classification algorithm to identify sedentary events as either lying or sitting events using a thigh-worn, triaxial accelerometer. METHODS: Seven-day free-living activity from 14 sedentary workers was recorded using the activPAL3™ monitor. Participants recorded when they went to bed and when they got up in a diary. All "in-bed" sedentary events were assumed to be lying and all "not-in-bed" sedentary events as sitting. An algorithm computed the angle of rotation of the y-axis, which was used to detect orientation of the thigh. Crossing a rotational threshold in the transverse plane of ±65° was used to classify a sedentary event as lying. The classification accuracy of the algorithm was compared with self-reported classification from the diary. RESULTS: The algorithm classified 96.7% of the sedentary time "in bed" (sensitivity) as lying and 92.9% of the time "not in bed" as not lying (specificity). CONCLUSIONS: Triaxial accelerometer data recorded from a single site on the thigh can be used to classify sedentary events as sitting and lying. The automated method developed in this study will allow objective measurement of diurnal lying time and that while sleeping without relying on self-report. This will help advance the understanding of the impact of different types of sedentary behaviors on various health outcomes.
UNLABELLED: A triaxial accelerometer worn on the thigh can provide information on the angle of rotation of the thigh. These data may be used to estimate periods of lying versus sitting. PURPOSE: To develop and test a classification algorithm to identify sedentary events as either lying or sitting events using a thigh-worn, triaxial accelerometer. METHODS: Seven-day free-living activity from 14 sedentary workers was recorded using the activPAL3™ monitor. Participants recorded when they went to bed and when they got up in a diary. All "in-bed" sedentary events were assumed to be lying and all "not-in-bed" sedentary events as sitting. An algorithm computed the angle of rotation of the y-axis, which was used to detect orientation of the thigh. Crossing a rotational threshold in the transverse plane of ±65° was used to classify a sedentary event as lying. The classification accuracy of the algorithm was compared with self-reported classification from the diary. RESULTS: The algorithm classified 96.7% of the sedentary time "in bed" (sensitivity) as lying and 92.9% of the time "not in bed" as not lying (specificity). CONCLUSIONS: Triaxial accelerometer data recorded from a single site on the thigh can be used to classify sedentary events as sitting and lying. The automated method developed in this study will allow objective measurement of diurnal lying time and that while sleeping without relying on self-report. This will help advance the understanding of the impact of different types of sedentary behaviors on various health outcomes.
Authors: P M Dall; D A Skelton; M L Dontje; E H Coulter; S Stewart; S R Cox; R J Shaw; I Čukić; C F Fitzsimons; C A Greig; M H Granat; G Der; I J Deary; Sfm Chastin Journal: J Meas Phys Behav Date: 2018-03
Authors: Natasha Reid; Robin M Daly; Elisabeth A H Winkler; Paul A Gardiner; Elizabeth G Eakin; Neville Owen; David W Dunstan; Genevieve N Healy Journal: PLoS One Date: 2016-04-13 Impact factor: 3.240
Authors: S F M Chastin; M L Dontje; D A Skelton; I Čukić; R J Shaw; J M R Gill; C A Greig; C R Gale; I J Deary; G Der; P M Dall Journal: Int J Behav Nutr Phys Act Date: 2018-02-26 Impact factor: 6.457