Deutsche Zeitschrift für Sportmedizin (Feb 2016)

Nutrition & Sports

  • Thiel C,
  • Gabrys L,
  • Vogt L

DOI
https://doi.org/10.5960/dzsm.2016.220
Journal volume & issue
Vol. 67, no. 2

Abstract

Read online

Due to the growing importance of habitual physical activity, accelerometry is increasingly playing a significant role in various preventive and rehabilitative contexts. The registration of accelerations of body segments like hip or wrist has become established as an objective method to measure physical activity and sedentary behavior and estimate activity-related energy expenditure in the field. To ensure a valid measurement, wearing position and protocol, instrument settings, as well as data cleaning and analysis are chosen according to validation studies in the same population using the same accelerometer model. Measurements of habitual physical activity of healthy children and adults are usually performed over 7 days, 8-10 hours each, using the highest resolution possible (10-100Hz). After eliminating nonwear periods, the relevant outcomes are calculated from the preprocessed, device-specific counts per minute, which refer mostly to epochs of 2 to 60 seconds, using cut-points and regression models. This procedure is based on simple assumptions and models, notably the representativeness of movements registered in body segments, as well as the invariant relation between the acceleration registered and the physical activity or energy expenditure. Scheduling and preparation,data collection, signal processing and interpretation require sound knowledge of the usage requirements, methodological characteristics, and limitations of accelerometry. Due to varying cognition, compliance, body composition and movement economy, the observation of younger, older or chronically ill persons may require specific preparations, device settings, and analysis protocols. For publication, besides results, the method of data collection and its limitations should be presented carefully and thoroughly.KEY WORDS: Physical Activity, Objective Monitoring, Wearable Sensors, Methodological Standard