Related Experiment Video
Updated: Dec 21, 2025
08:45
Assessment of Physical Activity Intensity with Accelerometers and Oxygen Consumption
Published on: June 20, 2025
387
Accelerometry analysis options produce large differences in lifestyle physical activity measurement
Marissa Mueller1, Ruth Chimenti1, Shannon Merkle2
1Department of Physical Therapy & Rehabilitation Science, University of Iowa, Iowa City, IA, United States of America.
Physiological Measurement
|May 21, 2020
Summary
Accelerometer-based physical activity (PA) measurements vary significantly based on processing algorithms and filters. These differences impact estimates of energy expenditure and moderate-to-vigorous PA, highlighting the need for standardized reporting in research.
Area of Science:
- Exercise Science
- Biomedical Engineering
- Public Health
Background:
- Objective measurement of physical activity (PA) using accelerometers is crucial in research and clinical settings.
- The impact of various processing algorithms, filters, and corrections on PA measurement variability is often underestimated.
Purpose of the Study:
- To investigate how different scoring methods influence lifestyle PA estimates.
- To analyze the effects of specific processing techniques on accelerometer-derived PA metrics.
Main Methods:
- 132 adults wore wrist-worn accelerometers for one week to assess lifestyle PA across four domains: total energy expenditure (EE), active EE, moderate-to-vigorous PA (MVPA), and steps.
- 1-5 algorithms per domain were applied, with and without wrist correction and low-frequency-extension (LFE).
- Objective PA estimates were compared to self-reported data from the International Physical Activity Questionnaire.
Main Results:
- Significant differences in PA estimates were observed across algorithms, with large effect sizes (d = 0.08–1.88).
- Wrist correction consistently reduced PA estimates (p < 0.05, d = 0.26–3.04), except for step counts.
- Low-frequency-extension (LFE) notably increased step counts (d = 1.44, p < 0.05) but had minimal impact on other outcomes.
Conclusions:
- Accelerometer-based PA measurement is highly sensitive to algorithm and filter selection, rendering previously validated methods non-interchangeable.
- Caution is advised when interpreting absolute PA estimates due to algorithm dependency.
- Standardized reporting of detailed methodologies is essential for cross-study comparisons.
