Author(s): MOMMA, R., TANIMURA, Y., GOTO, H., KAGEYUKI, T., YAMAGISHI, T., HAKAMADA, N., MATSUBAYASHI, T., Institution: JAPAN INSTITUTE OF SPORTS SCIENCES, Country: JAPAN, Abstract-ID: 1568

INTRODUCTION:
In para-athletes, heart rate (HR) measurement during exercise is essential to monitor exercise intensity and ensure safety of the athletes. Traditionally, chest band HR sensors, which can measure HR accurately using electrocardiography (ECG), are utilized in many sports fields. Furthermore, optical heart rate (OHR) sensors, which measure pulse rate on forearm or upper arm, have been widely used in recent years and are expected to replace the chest band HR sensors. While armband sensors measure pulse rate using changes in the intensity of light reflected from the skin, body movement and muscle contractions can interfere with the intensity of the light, making it difficult to accurately capture the pulse in such cases [1]. However, few studies have examined the accuracy of armband sensors during exercise in wheelchair racing athletes where strong upper body motions repeatedly occur. Therefore, this study sought to investigate the accuracy of armband sensors during exercise in wheelchair racing athletes by comparing with chest band sensors.
METHODS:
Subjects were 6 middle- and/or long-distance track and/or marathon athletes with cervical and thoracic spinal cord injuries. We analyzed 8 trials of measurement data performed in 2023 (2 of them were measured once on different days). The athletes performed a progressive incremental speed test on a wheelchair roller ergometer (ESSEDA). During the test, HR was simultaneously measured using a chest band sensor (Polar H10) and armband sensor (Polar verity sense). Armband sensors were measured on the forearm for 4 of the 8 trials and on the upper arm for the other 4 trials. HR data were averaged over the last minute of each stage, and a single regression analysis was performed for each athlete to examine changes of HR as a function of speed increment with the two sensors. Moreover, all data derived from the 8 trials were pooled, and then the same analysis was performed. The coefficient of determination was interpreted as R2>0.9 (good), R2>0.6-0.89 (moderate), and R2<0.59 (poor), respectively.
RESULTS:
Armband sensor could measure the changes of HR with exercise speed in most of 8 trials, with good in 5, moderate in 1, and poor in 2 trials, respectively. In 1 trial with moderate, HR did not rise with exercise speed only in the last stage, whereas it showed linear increase with speed in the preceding 4 stages. In 3 trials with good and 1 with moderate, the measurement site was the forearm; in 2 poor trials, on the upper arm. Chest band sensor could capture changes in HR with exercise speed in 4 of 8 trials (R2>0.9), while it did not increase linearly with exercise speed in the other 4 trials (R2<0.59). When pooled all data, R2=0.78 for armband and R2 =0.26 for chest band sensors.
CONCLUSION:
Armband sensors more accurately captured the changes of HR during the incremental tests in wheelchair athletes, especially when it was measured on the forearm.

Reference: [1] Horton et al., 2017