Validity Of Combined Heart Rate And Accelerometry To Predict Activity Energy Expenditure In Older Adults

Abstract

PURPOSE: The present study compared physical activity related energy expenditure (PAEE) as predicted by group calibrated algorithm combining accelerometry and heart rate. METHODS: N=45 older adults with and without cardiac risk performed a protocol consisting of simulated daily living activities (resting period, light, moderate and vigorous physical activity including walking in different speeds and stair climbing) and a cycle ergometer test (25W + 25W every 2 minutes). PAEE was concurrently assessed by indirect calorimetry (IC) and by combined heart rate (HR) plus uniaxial accelerometry (ACC). Raw HR and ACC data were transformed to PAEE using a published branched equation model. Estimated and measured PAEE were compared using Bland-Altman plots. SUMMARY OF RESULTS: Preliminary results of n=12 healthy older adults without cardiac risk (8 male and 4 female, 71.9 ± 5.2 years) showed lower ACC+HR-estimated PAEE for clustered activities (including cycle ergometer test) and resting period (clustered activities 2.17 ± 1.17 Kcal/min; resting period 0.19 ± 0.17 Kcal/min) than PAEE measured by IC (clustered activities 2.93 ± 1.27 Kcal/min; resting period 0.61 ± 0.40 Kcal/min). The Bland & Altman plots' limits of agreement were between -2.28 and +0.94 kcal/min. Visual inspection showed that the dispersion of the absolute differences between ACC+HR and IC seemed to increase with higher measured values. The correlation between combined ACC+HR-PAEE and IC-PAEE was rρ = 0.86. CONCLUSION: First results indicate that branched equation models based on combined ACC+HR registrations may underestimate PAEE in older people but appear sufficiently valid for the assessment of mean PAEE in groups. If the relatively large limits of agreement were to persist throughout this ongoing study, they would indicate a limited validity of existing ACC+HR branched equations for the individual assessment of PAEE in older adults. More research is needed to determine whether combined ACC+HR assessments are superior to simple ACC to measure physical activity in older adults.