Validity of a Smartphone Application and Chest Strap for Recording RR Intervals at Rest in Athletes

Abstract

Objective: To evaluate the accuracy of the smartphone application (app) HRV Expert (CardioMood) and a chest strap (H10 Polar) for recording R-R intervals compared with electrocardiogram (ECG). Methods: A total of 31 male recreational runners (age 36.1 [6.3] y) volunteered for this study. R-R intervals were recorded simultaneously by the smartphone app and ECG for 5 minutes to analyze heart-rate variability in both the supine and sitting positions. Time-domain indexes (heart rate, mean R-R, SD of RR intervals, count of successive normal R-R intervals differing by more than 50 ms, percentage of successive normal R-R intervals differing by more than 50 ms, and root mean square of successive differences between normal R-R intervals), frequency-domain indexes (low frequency, normalized low frequency, high frequency, normalized high frequency, low-frequency to high-frequency ratio, and very low frequency), and nonlinear indexes (SD of instantaneous beat-to-beat variability and long-term SD of continuous R-R intervals) were compared by unpaired t test, Pearson correlation, simple linear regression, and Bland–Altman plot to evaluate the agreement between the devices. Results: High similarity with P value varying between .97 and 1.0 in both positions was found. The correlation coefficient of the heart-rate-variability indexes was perfect (r = 1.0; P = .00) for all variables. The constant error, standard error of estimation, and limits of agreement between ECG and the smartphone app were considered small. Conclusion: The smartphone app and chest strap provide excellent ECG compliance for all variables in the time domain, frequency domain, and nonlinear indexes, regardless of the assessed position. Therefore, the smartphone app replaces ECG for any heart-rate-variability analysis in runners.

Utilização de cardiofrequencímetros para mensuração da Variabilidade da Frequência Cardíaca no repouso: uma revisão de literatura

“…Os modelos de cardiofrequencímetros utilizados nos estudos foram: Polar T31 (Radespiel-Troger et al, 2003), Polar S810 (Barbosa et al, 2016;Gamelin et al, 2006Gamelin et al, , 2008Nunan et al, 2009;Porto & Junqueira Jr, 2009;Vanderlei et al, 2008), Polar RS800 (Barbosa et al, 2016;Vasconcellos et al, 2015;Wallén et al, 2012), Polar V800 (Castro et al, 2021;Cilhoroz et al, 2020;Giles et al, 2016;Huang et al, 2021), Polar H7 (Plews et al, 2017) e Polar H10 (Pereira et al, 2020). Bosquet (2006) utilizaram o cardiofrequencímetro Polar S810 e o ECG para a análise de índices de VFC no domínio do tempo (SDNN, RMSSD, NN50, PNN50, SD1 e SD2) e da frequência (LF ms2, HF ms2, LF n.u., HF n.u., LF/HF e VLF), nas posições ortostática e supina, em 18 homens (idade média de 27,1 anos).…”

Different acquisition systems for heart rate variability analysis may lead to diverse outcomes

“…A decrease in the correlation values of the parameters in the frequency domain (LF and HF) was found between the devices used. A recent comparative study (24) demonstrated excellent compatibility between the smartphone APP and the ECG. In addition, agreement was observed for all parameters in the time domain, frequency domain, and nonlinear indices.…”

“…Although the reproducibility of HRV devices has been evaluated in different situations (21)(22)(23), whether their performance is equivalent to the gold standard remains controversial. In some validation studies, mobile devices synchronized with the transmitters are used only for signal acquisition (22,24,25). In these studies, HRV analysis is often performed by an external software.…”

Wavelet Analysis and Self-Similarity of Photoplethysmography Signals for HRV Estimation and Quality Assessment

“…The Polar H10 device was chosen as a source of ground truth labels since it uses ECG for RR interval detection and has demonstrated good accuracy in several studies [ 38 , 39 , 40 , 41 ]. In particular, the study [ 38 ] reports that the difference between ECG RR intervals and Polar H10 RR intervals has a mean of ms and limits of agreement of ms at rest, and therefore recommends it as the “gold standard for RR interval assessments”.…”