•  
  •  
 

ACCURACY OF ACTIVITY MONITORS IN MEASURING ENERGY EXPENDITURE AND HEART RATE DURING A GYM-BASED ROUTINE

Abstract

Veronika Pribyslavska1, Jennifer L. Caputo2, Dana K. Fuller2, & Vaughn W. Barry2

1Arkansas State University, Jonesboro, Arkansas; 2Middle Tennessee State University, Murfreesboro, Tennessee

PURPOSE: Two wrist-worn monitors (FBS and GVF) and an armband (SWA) were evaluated in measuring energy expenditure (EE) and heart rate (HR) during a gym-based routine. METHODS: Men (n = 21) and women (n = 16) completed a gym-based routine comprised of 15-min stationary cycling (SC), 15-min treadmill running (TR), and 35-min resistance training (RTR) at self-selected intensities while wearing the monitors. All monitors were compared to a portable metabolic analyzer (OM) for EE. The FBS and GVF were compared to a chest HR monitor (PM) for maximal HR (HRmax) and average HR (HRavg). RESULTS: Compared to the OM, the FBS and GVF produced higher EE estimates while the SWA overestimated EE during TR and underestimated EE for the rest (Table 1). Equivalency testing determined that no monitor was equivalent to the OM, although the SWA yielded the most favorable agreement for whole session as the 90% CI (410.9-500.1 kcal) overlapped the higher end of the equivalency zone (423.6-517.7 kcal) by only 12.7 kcal. Acceptable measurement error (≤ 20%) for whole session was produced by the SWA and for TR by the SWA and GVF, while for RTR the GVF had the highest measurement error (Table 1). Significant bias was observed for the FBS and GVF during SC (10.1 ± 39.2 kcal and 18.6 ± 33.7 kcal), RTR (47.4 ± 52.7 kcal and 82.0 ± 79.2 kcal), and whole session (83.2 ± 93.7 kcal and 104.4 ± 131.9 kcal). In regards to HRavg and HRmax, both monitors’ 90% CIs fell in the equivalency zones with the exception of HRavg measured by the FBS for RTR. The GVF had an improved accuracy over the FBS as indicated by lower error rates for HRavg (SC: 2.0% vs 5.9%; TR: 2.4% vs 3.1%; and RTR: 2.4% vs 9.5%) and HRmax (SC: 0.8% vs 1.4%; TR: 0.9% vs 1.5%; and RTR: 0.9% vs 7.0%). CONCLUSION: The study protocol simulated real-world conditions to facilitate naturalistic application of the findings. No monitor accurately estimated EE, however, the SWA had the most favorable estimates. The FBS and GVF demonstrated comparable performance for both EE and HR estimates.

Pribyslavska Table 1.docx (1206 kB)
Pribyslavska Table 1

This document is currently not available here.

Share

COinS