Physical activity assessment by accelerometry in people with heart failure
Dibben, 2020

Description

Thresholds were created for two acceleration-based metrics (sum of vector magnitudes and mean amplitude deviation) for GENEActiv monitors worn at the hip and each wrist in patients with heart failure. Proportion of correct predictions were 96 (right wrist), 99 (left wrist), and 95% (waist) for sum of vector magnitudes and 69 (right wrist), 64 (left wrist), and 87% (waist) for mean amplitude deviation.

Development/Validation

Sample: 22 (17 males) participants with heart failure, 70.7±14.1 years of age

Setting: Laboratory

Activities: Activities of daily living, overground walking, stationary

Criterion: Oxycon Mobile (VO2)

Accelerometer(s): GENEActiv on left hip and both wrists

Validation approach: Leave-one-out cross-validation

Phase Designation

(What's this?)

This model is in Phase 1.

Instructions

Thresholds are available for two acceleration-based metrics, sum of vector magnitudes (SVM) and mean amplitude deviation (MAD). For SVM, vector magnitude (square root of the sum of the squared acceleration in each axis) minus 1000 mg is calculated for each epoch. Mean amplitude deviation is the typical distance of points around the mean (this can also be done in R using a package like acc or GGIR). Threshold were derived using mixed effects regression and receiver operator characteristic for each wear location. The authors recommend use of the regression-derived SVM cut-points.

Source Information

Reference:

Dibben, G. O., Gandhi, M. M., Taylor, R. S., Dalal, H. M., Metcalf, B., Doherty, P., ... & Hillsdon, M. (2020). Physical activity assessment by accelerometry in people with heart failure. BMC Sports Science, Medicine and Rehabilitation, 12(1), 1-12. https://doi.org/10.1186/s13102-020-00196-7 Link to Paper

Corresponding author: Grace Dibben, Gd318@exeter.ac.uk


Contact

Kimberly Clevenger at accelerometerrepository@gmail.com