Accuracy of the Polar FT40M heart rate monitor in estimating energy expenditure during treadmill running

Estimation of energy expenditure (EE) from heart rate (HR) monitors has been reported to provide inaccurate predictions of EE during exercise by some studies, whilst others have reported good predictive accuracy. Some of the discrepancies reported could be related to the exercise intensity used where accuracy could be influenced by the upper/lower HR ranges utilised, and whether male or female participants were used as gender has been reported to influence predictive accuracy. Therefore, the purpose of this study was to examine the accuracy of the Polar FT40M heart rate monitor (HRM) in estimating EE during treadmill running at four exercise intensities in males and females. Following institutional ethics approval, six male (mean ± SD: age = 24.33 ± 3.82 years, height = 180.67 ± 7.54 cm, weight = 81.67 ± 13.16 kg) and six female (mean ± SD: age = 22.17± 3.13 years, height = 162.67 ± 4.82 cm, weight = 59 ± 3.16 kg) participants conducted a preliminary trial to determine treadmill running speeds which correspond to 55%, 65%, 75% and 85% of estimated HRmax. The participants then performed four steady-state submaximal trials at the speeds determined during the preliminary trials. EE was measured using indirect calorimetry (IC) (Cortex Metalyzer 3b) and estimated using HRM simultaneously for 5 min at each intensity. A three-way ANOVA (measuring device [IC/HRM] × intensity × gender) was conducted to compare the mean EE. There was no significant difference (P = 403) in average gross EE (all intensities) between IC (45.3 ± 2.9 kcal) compared to HRM (44.1 ± 2.0 kcal). Measured and predicted EE differed most (16.4%) at 55% HRmax and least (1.4%) at 75% HRmax, but intensity did not have a significant effect (P = 0.068) on the accuracy of HRM estimation of EE (IC: 29.7 ± 15.7, 41.2 ± 16.5, 50.9 ± 15.4, 59.5 ± 14.4 vs. PHRM: 24.8 ± 5.4, 38.3 ± 8.0, 51.7 ± 9.3 and 61.6 ± 10.2 kcal for 55%, 65%, 75% and 85% HRmax, respectively). There was also no significant gender effect at each intensity (P = 0.178), but when comparing average gross EE, gender significantly affected the accuracy of the HRM (P = 0.012) with an overestimation of EE in females (IC: 34.2 ± 13.1 vs. HRM: 40.6 ± 15.2 kcal, P < 0.01, effect size = 0.45) and underestimation in males (IC: 56.5 ± 12.7 vs. HRM: 47.6 ± 16.9 kcal, P < 0.01, effect size = −0.59). The results show that the Polar FT40M HRM provides good predictive accuracy of EE across different exercise intensities, but gross average EE is influenced by gender with an overestimation of EE in females and underestimation in males