PURPOSE: Walking for health is recommended by health agencies, partly based on epidemiological studies of self-reported behaviors. Accelerometers are now replacing survey data, but it is not clear that intensity-based cut points reflect the behaviors previously reported. New computational techniques can help classify raw accelerometer data into behaviors meaningful for public health. METHODS: Five hundred twenty days of triaxial 30-Hz accelerometer data from three studies (n = 78) were employed as training data. Study 1 included prescribed activities completed in natural settings. The other two studies included multiple days of free-living data with SenseCam-annotated ground truth. The two populations in the free-living data sets were demographically and physical different. Random forest classifiers were trained on each data set, and the classification accuracy on the training data set and that applied to the other available data sets were assessed. Accelerometer cut points were also compared with the ground truth from the three data sets. RESULTS: The random forest classified all behaviors with over 80% accuracy. Classifiers developed on the prescribed data performed with higher accuracy than the free-living data classifier, but these did not perform as well on the free-living data sets. Many of the observed behaviors occurred at different intensities compared with those identified by existing cut points. CONCLUSIONS: New machine learning classifiers developed from prescribed activities (study 1) were considerably less accurate when applied to free-living populations or to a functionally different population (studies 2 and 3). These classifiers, developed on free-living data, may have value when applied to large cohort studies with existing hip accelerometer data.
PURPOSE: Walking for health is recommended by health agencies, partly based on epidemiological studies of self-reported behaviors. Accelerometers are now replacing survey data, but it is not clear that intensity-based cut points reflect the behaviors previously reported. New computational techniques can help classify raw accelerometer data into behaviors meaningful for public health. METHODS: Five hundred twenty days of triaxial 30-Hz accelerometer data from three studies (n = 78) were employed as training data. Study 1 included prescribed activities completed in natural settings. The other two studies included multiple days of free-living data with SenseCam-annotated ground truth. The two populations in the free-living data sets were demographically and physical different. Random forest classifiers were trained on each data set, and the classification accuracy on the training data set and that applied to the other available data sets were assessed. Accelerometer cut points were also compared with the ground truth from the three data sets. RESULTS: The random forest classified all behaviors with over 80% accuracy. Classifiers developed on the prescribed data performed with higher accuracy than the free-living data classifier, but these did not perform as well on the free-living data sets. Many of the observed behaviors occurred at different intensities compared with those identified by existing cut points. CONCLUSIONS: New machine learning classifiers developed from prescribed activities (study 1) were considerably less accurate when applied to free-living populations or to a functionally different population (studies 2 and 3). These classifiers, developed on free-living data, may have value when applied to large cohort studies with existing hip accelerometer data.
Authors: Charles E Matthews; Kong Y Chen; Patty S Freedson; Maciej S Buchowski; Bettina M Beech; Russell R Pate; Richard P Troiano Journal: Am J Epidemiol Date: 2008-02-25 Impact factor: 4.897
Authors: Tiago V Barreira; Theodore W Zderic; John M Schuna; Marc T Hamilton; Catrine Tudor-Locke Journal: Gait Posture Date: 2015-04-24 Impact factor: 2.840
Authors: Richard P Troiano; David Berrigan; Kevin W Dodd; Louise C Mâsse; Timothy Tilert; Margaret McDowell Journal: Med Sci Sports Exerc Date: 2008-01 Impact factor: 5.411
Authors: Carlos A Celis-Morales; Francisco Perez-Bravo; Luis Ibañez; Carlos Salas; Mark E S Bailey; Jason M R Gill Journal: PLoS One Date: 2012-05-09 Impact factor: 3.240
Authors: Chirag J Patel; Jacqueline Kerr; Duncan C Thomas; Bhramar Mukherjee; Beate Ritz; Nilanjan Chatterjee; Marta Jankowska; Juliette Madan; Margaret R Karagas; Kimberly A McAllister; Leah E Mechanic; M Daniele Fallin; Christine Ladd-Acosta; Ian A Blair; Susan L Teitelbaum; Christopher I Amos Journal: Cancer Epidemiol Biomarkers Prev Date: 2017-07-14 Impact factor: 4.254
Authors: Purva Jain; John Bellettiere; Nicole Glass; Michael J LaMonte; Chongzhi Di; Robert A Wild; Kelly R Evenson; Andrea Z LaCroix Journal: J Gerontol A Biol Sci Med Sci Date: 2021-01-01 Impact factor: 6.053
Authors: Supun Nakandala; Marta M Jankowska; Fatima Tuz-Zahra; John Bellettiere; Jordan A Carlson; Andrea Z LaCroix; Sheri J Hartman; Dori E Rosenberg; Jingjing Zou; Arun Kumar; Loki Natarajan Journal: J Meas Phys Behav Date: 2021-02-25