Validation of cosinuss° in-ear sensor to estimate energy expenditure using heart rate, acceleration, and mobile spirometry
Document type:
Konferenzbeitrag
Contribution type:
Poster
Author(s):
David Camargo, Henning Wackerhage & Martin Schönfelder
Pages contribution:
459
Abstract:
Introduction
A growing number of portable devices has become available to quantify activities of daily living monitoring heart-rate (HR) and/or acceleration (ACC). Using these data, it is possible to estimate energy expenditure (EE). A linear relationship has been reported between EE and body acceleration in walking. Although there are different kinds of studies related to HR, ACC and EE, no one used an ear sensor before. This study assessed the validity of novel ear sensors and compared the outcome to mobile spirometry estimating EE.
Methods
Nineteen healthy male and female participants (age 26.4 ± 3.4, body mass 69.5 ± 9.8) performed one lab test consisted of a standard treadmill test (Modified Bruce Protocol) and an outdoor walking test (1 kilometre) at moderate and maximal walking speeds up- and downhill. As reference, EE was measured by a mobile spiroergometry device (MetaMax 3B-3X, Cortex) based on the Bouwer Equation (using O2 and CO2), whereas HR and ACC, were measured continuously by cosinuss° ear sensors one and two. Validity of the methods used to estimate EE was compared using Pearson correlations, and a regression model was conducted based on the lab session, and root mean squared error (RMSE)
from cross-validation at the individual and population levels.
Results
A mixed-model analysis identified HR and ACC as factors that best predicted the relationship between HR, ACC, and EE. The model (with the highest likelihood ratio) was used to estimate energy expenditure. The correlation coefficient (r) between the measured and estimated energy expenditure was 0.789 whilst the (r) between magnitude vector and energy expenditure was 0.971. The model therefore accounted for 98.4% (r2) of the variance in EE in this sample. The linear regression equation for changes in HR showed the highest coefficient of determination (r2) of 0.984 with an F < 0.05, while the linear regression equation for changes in magnitude vector showed a coefficient of determination (r2) of 0.941 with an F < 0.05. When the HR signal was included, the (r2) value (> 0.984) between the ACC outputs and EE improved.
Discussion
The novel ear sensor cosinuss° One and Two can be used to estimate EE in physical activities, but it presents a small percentage of overestimation. The use of HR and ACC increases the prediction of EE. The sensor is well tolerated for medium or light intensity training but increases its variability at high intensities.
Keywords:
Energy Expenditure, Physical Activity, Heart Rate, Accelerometery, Prediction Equation