BACKGROUND: Activity monitors are widely used in research, and are currently being used to study physical activity (PA) trends in the US and Canada. The purpose of this study was to determine if body mass index (BMI) affects the step count accuracy of commonly used accelerometer-based activity monitors during treadmill walking. METHODS: Participants were classified into BMI categories and instructed to walk on a treadmill at 3 different speeds (40, 67, and 94 m·min(-1)) while wearing 4 accelerometer-based activity monitors (ActiGraph GT1M, ActiCal, NL-2000, and StepWatch). RESULTS: There was no significant main effect of BMI on pedometer accuracy. At the slowest speed, all waist-mounted devices significantly underestimated actual steps (P < .001), with the NL-2000 recording the greatest percentage (72%). At the intermediate speed, the ActiGraph was the least accurate, recording only 80% of actual steps. At the fastest speed, all of the activity monitors demonstrated a high level of accuracy. CONCLUSION: Our data suggest that BMI does not greatly affect the step-counting accuracy of accelerometer-based activity monitors. However, the accuracy of the ActiGraph, ActiCal, and NL-2000 decreases at slower speeds. The ankle-mounted StepWatch was the most accurate device across a wide range of walking speeds.
BACKGROUND: Activity monitors are widely used in research, and are currently being used to study physical activity (PA) trends in the US and Canada. The purpose of this study was to determine if body mass index (BMI) affects the step count accuracy of commonly used accelerometer-based activity monitors during treadmill walking. METHODS:Participants were classified into BMI categories and instructed to walk on a treadmill at 3 different speeds (40, 67, and 94 m·min(-1)) while wearing 4 accelerometer-based activity monitors (ActiGraph GT1M, ActiCal, NL-2000, and StepWatch). RESULTS: There was no significant main effect of BMI on pedometer accuracy. At the slowest speed, all waist-mounted devices significantly underestimated actual steps (P < .001), with the NL-2000 recording the greatest percentage (72%). At the intermediate speed, the ActiGraph was the least accurate, recording only 80% of actual steps. At the fastest speed, all of the activity monitors demonstrated a high level of accuracy. CONCLUSION: Our data suggest that BMI does not greatly affect the step-counting accuracy of accelerometer-based activity monitors. However, the accuracy of the ActiGraph, ActiCal, and NL-2000 decreases at slower speeds. The ankle-mounted StepWatch was the most accurate device across a wide range of walking speeds.
Authors: Grace E Ahn; Joan S Chmiel; Dorothy D Dunlop; Irene B Helenowski; Pamela A Semanik; Jing Song; Barbara Ainsworth; Rowland W Chang; Rosalind Ramsey-Goldman Journal: Arthritis Care Res (Hoboken) Date: 2015-05 Impact factor: 4.794
Authors: Britt Elin Øiestad; Emily Quinn; Daniel White; Frank Roemer; Ali Guermazi; Michael Nevitt; Neil A Segal; Cora E Lewis; David T Felson Journal: J Rheumatol Date: 2015-06-15 Impact factor: 4.666
Authors: Ville Stenbäck; Juhani Leppäluoto; Rosanna Juustila; Laura Niiranen; Dominique Gagnon; Mikko Tulppo; Karl-Heinz Herzig Journal: J Clin Med Date: 2022-06-07 Impact factor: 4.964
Authors: Daniel K White; Catrine Tudor-Locke; David T Felson; K Doug Gross; Jingbo Niu; Michael Nevitt; Cora E Lewis; James Torner; Tuhina Neogi Journal: Arch Phys Med Rehabil Date: 2012-12-07 Impact factor: 3.966
Authors: Renee M Jeffreys; Thomas H Inge; Todd M Jenkins; Wendy C King; Vedran Oruc; Andrew D Douglas; Molly S Bray Journal: J Phys Act Health Date: 2014-09-10