INTRODUCTION: In many sports, heart rate (HR) monitoring is used to quantify internal training load and to control and steer exercise intensity [1]. In swimming, the standard method of ECG-based HR recording is error-prone, as chest straps may slip during swimming and wall push-offs, especially when no swim suit is worn. The optical HR sensor Verity Sense (VS) by Polar may overcome the practical limitations of chest straps due to fixation on the swimming goggles. In this study [2], we evaluate the agreement between the VS and Polar H10 chest strap sensors during swimming. METHODS: 36 competitive swimmers (12-31 yrs) were equipped with VS and H10 HR sensors during regular swimming practice (various intensities, rest periods, swimming styles and drills). The VS and H10 were worn as recommended by Polar, at the right temple and at the chest covered by a swim suit. Agreement between HR data series (stored with 1 HR value per second and manually synchronised for optimal overlap) was analysed at the individual level using Bland-Altman analysis [3]. Group summaries are presented as medians (interquartile ranges). In addition, qualitative analyses were performed by visual inspection of individual Bland-Altman plots and HR time series to identify potential sources of training content-related measurement error. RESULTS: First analyses included datasets with 3199-9139 data pairs per athlete. The median standard deviation of differences (SDdiff) between sensors was 5.4bpm (interquartile range 4.2, 7.3) with median 95% limits of agreement from -10bpm (-15.4, -7.8) to 10bpm (8, 13.7). The median bias of the VS was -0.2bpm (-0.8, 0.1) compared to the H10. Intra-individual results varied substantially between swimmers with agreement ranging from excellent to unacceptable (95% limits of agreement minimum -4 to 4bpm, maximum -28 to 18bpm). Visual analysis indicated that larger errors tend to be present at low to moderate HR values, e.g. typically during rest periods or at the start and end of exercise sets. CONCLUSION: Our preliminary analyses revealed an acceptable median agreement of the VS during regular swimming practice (SDdiff interquartile range =7bpm), with a substantial variability in results between swimmers. Qualitative analyses further indicated that a notable proportion of the observed errors occurred during specific exercise or rest periods. In summary, optical HR monitoring using the VS represents a practical alternative to the error-prone use of chest strap sensors (when used without a swim suit) for monitoring internal load and controlling exercise intensity in swimming. Nevertheless, users must be aware that measurement error can be unacceptably high in some athletes or during specific training periods, calling for familiarisation and an evaluation of whether the measured values are physiologically plausible.
© Copyright 2022 27th Annual Congress of the European College of Sport Science (ECSS), Sevilla, 30. Aug - 2. Sep 2022. Veröffentlicht von Faculty of Sport Science - Universidad Pablo de Olavide. Alle Rechte vorbehalten.