Heart Rate Variability: A Longitudinal Comparison of Commercial Devices for Individual and Group Stress-Response
(0 User reviews)
673
99
1. Strong Lab, Air Force Research Laboratory, Wright Patterson Air Force Base, Fairborn, USA; Department of the Air Force, Washington D.C., USA; 3. 711 Human Performance Wing, Fairborn, USA
Advances in Applied Physiology (Science Publishing Group)
2022
7 : 2
26-33
10.11648/j.aap.20220702.12
English
Kaela Hierholzer, Robert Briggs, Michael Tolston, Nicholas Mackowski, Jason Eckerle, Maegan O’Connor, Kristyn Barrett, Roger Smith, Adam Strang. Heart Rate Variability: A Longitudinal Comparison of Commercial Devices for Individual and Group Stress-Response. Advances in Applied Physiology. Vol. 7, No. 2, 2022, pp. 26-33. doi: 10.11648/j.aap.20220702.12. Share Research.
Abstract
The collection of heart rate variability (HRV) for health and performance observations have become prominent. However, each wearable device has proprietary algorithms that govern methods and timing of HRV capture and subsequent analysis. The purpose of this study was to evaluate HRV metrics taken from three, commonly used commercial wearables, and identify reliability and relationships to one another over time. Methods: Twenty-five subjects (18 males; 7 females) with ages ranging from 23 to 41 years (32.70 ± 4.65 years) were included in this study. These subjects were participants in a 12-week exercise intervention study. Each subject was equipped with a Whoop Strap (v2.0), the Garmin Fenix 5 Smartwatch and chest strap, and the Omegawave chest strap and sensor. Statistical Analysis: Between and within-subject correlations were calculated as well as average correlations, descriptive and inferential statistics, and the resultant z-score, which was transformed back into a correlation. Intraclass correlation coefficients (ICC) were calculated. Finally, linear mixed models were used to evaluate trends in HRV. Results: Within-subject correlations (0.24 ± 0.27) were lower than between-subjects correlations (0.54 ± 0.43), t (35) = -4.02, p < 0.001. Garmin HRV Stress, Whoop RMSSD, Omegawave SDNN, and Omegawave RMSSD yielded an ICC between 0.65 and 0.75. Garmin All-day stress, Garmin prior all-day stress, and Omegawave LF/HF yielded an ICC of 0.30 and 0.37. To test the effects of day of the week on HRV, we fitted linear mixed models to HRV metrics from three of the identified communities related to ICC: Omegawave RMSSD (moderate to high ICC), Omegawave LF/HF (low to moderate ICC), and Whoop recovery score (very low ICC). There was a main effect of gender on Omegawave RMSSD (p = 0.020) and a negative effect of day of the week (p = 0.030). Day of the week was the only significant predictor of Whoop recovery score (p < 0.001). Conclusion: The correlations of HRV values remain more consistent when assessed at similar times of the day, rather than being device dependent. Regardless of which wearable device is considered, HRV measures should be collected at a specific time each day for the best reliability. When creating an individualized or group exercise program, the human performance specialist should be aware that fatigue may become increasingly evident during the course of each week (e.g. individuals demonstrably fatigued by Friday may exhibit physiological indicators of relative recovery by Monday).
Heart Rate Variability, Wearables, RMSSD, SDNN
The collection of heart rate variability (HRV) for health and performance observations have become prominent. However, each wearable device has proprietary algorithms that govern methods and timing of HRV capture and subsequent analysis. The purpose of this study was to evaluate HRV metrics taken from three, commonly used commercial wearables, and identify reliability and relationships to one another over time. Methods: Twenty-five subjects (18 males; 7 females) with ages ranging from 23 to 41 years (32.70 ± 4.65 years) were included in this study. These subjects were participants in a 12-week exercise intervention study. Each subject was equipped with a Whoop Strap (v2.0), the Garmin Fenix 5 Smartwatch and chest strap, and the Omegawave chest strap and sensor. Statistical Analysis: Between and within-subject correlations were calculated as well as average correlations, descriptive and inferential statistics, and the resultant z-score, which was transformed back into a correlation. Intraclass correlation coefficients (ICC) were calculated. Finally, linear mixed models were used to evaluate trends in HRV. Results: Within-subject correlations (0.24 ± 0.27) were lower than between-subjects correlations (0.54 ± 0.43), t (35) = -4.02, p < 0.001. Garmin HRV Stress, Whoop RMSSD, Omegawave SDNN, and Omegawave RMSSD yielded an ICC between 0.65 and 0.75. Garmin All-day stress, Garmin prior all-day stress, and Omegawave LF/HF yielded an ICC of 0.30 and 0.37. To test the effects of day of the week on HRV, we fitted linear mixed models to HRV metrics from three of the identified communities related to ICC: Omegawave RMSSD (moderate to high ICC), Omegawave LF/HF (low to moderate ICC), and Whoop recovery score (very low ICC). There was a main effect of gender on Omegawave RMSSD (p = 0.020) and a negative effect of day of the week (p = 0.030). Day of the week was the only significant predictor of Whoop recovery score (p < 0.001). Conclusion: The correlations of HRV values remain more consistent when assessed at similar times of the day, rather than being device dependent. Regardless of which wearable device is considered, HRV measures should be collected at a specific time each day for the best reliability. When creating an individualized or group exercise program, the human performance specialist should be aware that fatigue may become increasingly evident during the course of each week (e.g. individuals demonstrably fatigued by Friday may exhibit physiological indicators of relative recovery by Monday).
There are no reviews for this Publication.
There are no comments for this Publication.
You must log in to post a comment.
Log in
