Send to

Choose Destination
Eur J Prev Cardiol. 2017 Dec;24(18):2009-2016. doi: 10.1177/2047487317738593. Epub 2017 Oct 25.

Can energy expenditure be accurately assessed using accelerometry-based wearable motion detectors for physical activity monitoring in post-stroke patients in the subacute phase?

Author information

1 HAVAE Laboratory (EA 6310), University of Limoges, France.
2 Univ.Grenoble Alpes, AGEIS, Grenoble, France.
3 Institut Universitaire de France, Paris, France.


Background In the subacute stroke phase, the monitoring of ambulatory activity and activities of daily life with wearable sensors may have relevant clinical applications. Do current commercially available wearable activity trackers allow us to objectively assess the energy expenditure of these activities? The objective of the present study was to compare the energy expenditure evaluated by indirect calorimetry during the course of a scenario consisting of everyday activities while estimating the energy expenditure using several commercialised wearable sensors in post-stroke patients (less than six months since stroke). Method Twenty-four patients (age 68.2 ± 13.9; post-stroke delay 34 ± 25 days) voluntarily participated in this study. Each patient underwent a scenario of various everyday tasks (transfer, walking, etc.). During the implementation, patients wore 14 wearable sensors (Armband, Actigraph GT3X, Actical, pedometer) to obtain an estimate of the energy expenditure. The actual energy expenditure was concurrently determined by indirect calorimetry. Results Except for the Armband worn on the non-plegic side, the results of our study show a significant difference between the energy expenditure values estimated by the various sensors and the actual energy expenditure when the scenario is considered as a whole. Conclusion The present results suggest that, for a series of everyday tasks, the wearable sensors underestimate the actual energy expenditure values in post-stroke patients in the subacute phase and are therefore not accurate. Several factors are likely to confound the results: types of activity, prediction equations, the position of the sensor and the hemiplegia side.


Wearable sensor; energy expenditure; monitoring; physical activity; stroke

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Atypon
Loading ...
Support Center