Description

A total of six sets of cut-points were generated using linear regression (four sets) or receiver operating characteristics (ROC) analysis (two sets) for children wearing an ActiGraph on the right hip. Instead of the proprietary counts output by ActiLife software, these cut-points are derived for counts generated using Brønd’s (2017) open-source method using three different band pass filters (AG: 0.29 to 1.66 Hz, AC4: 0.29 to 4 Hz, AC10: 0.29 to 10 Hz). Coefficient of determination (R²) was 0.61 to 0.89 for the linear regression cut-points and sensitivity and specificity ranged from 73.4 to 100.0% and 85.6 to 100.0%, respectively, for the ROC cut-points.

Development/Validation

Sample: 36 3^rd-4^th grade children in the calibration portion

Setting: Inside and outside of school

Activities: Bicycling, overground/self-paced walking and running, playground activity, sports, stationary

Criterion: MetaMax 3b (VO₂)

Accelerometer(s): ActiGraph GT3X+ on right hip

Validation approach: Comparison to criterion

Phase Designation

(What's this?)

This model is in Phase 0 and 2.

Source Information

Reference:

Brønd, J. C., Aadland, E., Andersen, L. B., Resaland, G. K., Andersen, S. A., & Arvidsson, D. (2019). The ActiGraph counts processing and the assessment of vigorous activity. Clinical Physiology and Functional Imaging, 39(4), 276-283. https://doi.org/10.1111/cpf.12571 Link to Paper

Corresponding author: Jan Brønd, brond@health.sdu.dk

Contact

Kimberly Clevenger at accelerometerrepository@gmail.com