Comparison of daily step counts among wearable devices in free-living adults: A systematic review and network meta-analysis
8 affiliations
- 1Department of Kinesiology, University of Massachusetts Amherst, Amherst, MA 01003, USA
- 2Institute for Applied Life Sciences, University of Massachusetts Amherst, Amherst, MA 01003 USA
- 3Gillings School of Global Public Health, Department of Epidemiology, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA
- 4Hospitals Project Development Ltd./LLC, Graz 8010, Austria
- 5The Research Institute, Cumberland Heights Foundation, Nashville, TN 37209, USA
- 6Department of Population Science, American Cancer Society, Atlanta, GA 30303, USA
- 7Department of Clinical and Applied Movement Sciences, University of North Florida, Jacksonville, FL 32224, USA
- 8Department of Kinesiology, Recreation, and Sport Studies (Emeritus), The University of Tennessee, Knoxville, TN 37996, USA
Abstract
Background: Wearable activity monitors are widely used to measure daily step counts, a simple and health-relevant metric of physical activity. However, differences in sensor technologies, algorithms, and wear locations across devices can result in substantial variability in step count estimates, creating challenges for comparing and harmonizing data across studies and populations. This study aimed to compare free-living daily step counts across commonly used activity monitoring devices in adults using network meta-analysis.
Methods: We systematically searched PubMed, Scopus, and PsycINFO through November 2025 for studies reporting concurrent daily step measurements using at least 2 devices from our target list of 13 manufacturers of research- and consumer-grade accelerometers and pedometers. Network meta-analysis estimated the ratio of means (ROMs) and mean differences in daily steps between devices.
Results: Across 59 studies, most consumer devices, pedometers, and the thigh (ActivPAL) and hip-worn research devices (ActiGraph GT, Axivity, and Actical) were within 10% ROM and <700 steps/day. Wear location was the primary source of variation, with wrist-worn and ankle-worn (StepWatch) research-grade accelerometers estimating 15%‒30% more daily steps (+1000 to 2200 steps/day) than hip-worn devices, while consumer-grade wrist devices showed smaller differences (3%‒15% higher or +200 to 1000 steps/day).
Conclusion: These findings provide researchers with evidence-based guidance for selecting appropriate step devices, interpreting step count data across studies, and considerations when combining data from multiple device types and wear locations.
Related
- Clinical applications of wearable devices for individuals with at‐risk substance use and substance use disorders: A systematic review
- The use of artificial intelligence and wearable devices in enhanced recovery after surgery pathways following radical cystectomy: A systematic review
- Wearable biofeedback devices for remote rehabilitation of elderly patients with arthritis and fibromyalgia: a randomized controlled trial protocol
- Assessing the Skin Sensitizing Potential of Constituents and Leachables from Wearable Devices Using GARDskin and Associated Applications
- Digital mobility assessment using wearable devices in people with Parkinson’s disease: a feasibility and implementation study in Ghana (TraPCAf) compared with a UK cohort
BibTeX
@article{paluch2026comparison,
title = {{Comparison of daily step counts among wearable devices in free-living adults: A systematic review and network meta-analysis}},
author = {Amanda E. Paluch and Kelly R. Evenson and Eric M. Eberl and Susan P. Marcotte and Pardis Parvizi and Erika Rees-Punia and Lindsay Toth and David R. Bassett},
journal = {Journal of Sport and Health Science},
year = {2026},
month = jul,
doi = {10.1016/j.jshs.2026.101162},
url = {https://doi.org/10.1016/j.jshs.2026.101162},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - JOUR
TI - Comparison of daily step counts among wearable devices in free-living adults: A systematic review and network meta-analysis
AU - Amanda E. Paluch
AU - Kelly R. Evenson
AU - Eric M. Eberl
AU - Susan P. Marcotte
AU - Pardis Parvizi
AU - Erika Rees-Punia
AU - Lindsay Toth
AU - David R. Bassett
PY - 2026
DA - 2026/07/23
JO - Journal of Sport and Health Science
DO - 10.1016/j.jshs.2026.101162
UR - https://doi.org/10.1016/j.jshs.2026.101162
AB - Background: Wearable activity monitors are widely used to measure daily step counts, a simple and health-relevant metric of physical activity. However, differences in sensor technologies, algorithms, and wear locations across devices can result in substantial variability in step count estimates, creating challenges for comparing and harmonizing data across studies and populations. This study aimed to compare free-living daily step counts across commonly used activity monitoring devices in adults using network meta-analysis. Methods: We systematically searched PubMed, Scopus, and PsycINFO through November 2025 for studies reporting concurrent daily step measurements using at least 2 devices from our target list of 13 manufacturers of research- and consumer-grade accelerometers and pedometers. Network meta-analysis estimated the ratio of means (ROMs) and mean differences in daily steps between devices. Results: Across 59 studies, most consumer devices, pedometers, and the thigh (ActivPAL) and hip-worn research devices (ActiGraph GT, Axivity, and Actical) were within 10% ROM and <700 steps/day. Wear location was the primary source of variation, with wrist-worn and ankle-worn (StepWatch) research-grade accelerometers estimating 15%‒30% more daily steps (+1000 to 2200 steps/day) than hip-worn devices, while consumer-grade wrist devices showed smaller differences (3%‒15% higher or +200 to 1000 steps/day). Conclusion: These findings provide researchers with evidence-based guidance for selecting appropriate step devices, interpreting step count data across studies, and considerations when combining data from multiple device types and wear locations.
SN - 2095-2546
SN - 2213-2961
ER -