Comparison of Heart Rate Variability Measurement Methods in Panels of Older Adults from the Outdoor Physical Activity and Health Study

摘要

Background: Between May 2013 and August 2015, participants in 5 separate Outdoor Physical Activity and Health study panels were asked to undergo endothelial function testing (EndoPAT, Itamar Medical) and ambulatory electrocardiogram (Hotter) monitoring for heart rate variability (HRV) analysis. In research mode, the EndoPAT system also provides HRV data. Our objective was to compare HRV results from Holter and EndoPAT to determine if the two methods were comparable in a field study. Methods: 221 healthy, current non-smokers, aged 55-84 exercised outdoors daily for 30 minutes over the 10 week study period. Participants underwent weekly Holter monitoring (～5hrs; GE SEER), and EndoPAT testing before and after exercise. 5-minute windows of Holter data were analysed for HRV by trained technicians. Time matched EndoPAT HRV results were also collected. Pearson correlations between log-transformed Holter and Endopat HRV data were calculated, accounting for repeated measurements among participants, for all panels and participants, and by age, panel, body mass index (BMI) and sex. Results: Over the 5 panels, HRV was measured 3748 times using Holter monitors and 3680 times using EndoPATs. Participant characteristics varied between panels: mean age, 63-65 years, percent female, 56-71, and percent normal BMI, 14-46. For the 5 panels combined Holter and Endopat HRV time domain measures were strongly and significantly correlated (0.75 to 0.83, p < 0.0001), while Holter and Endopat HRV frequency domain measures were less strongly correlated (0.28 to 0.63, p < 0.0001). Correlations were consistent between panels, sexes, and BMI groups. Conclusion: Time domain HRV measures from Holter monitor and EndoPAT were strongly correlated. Where both endothelial function and HRV time domain are outcomes of interest, a single instrument to measure both parameters could reduce cost, analysis time, and participant burden.