Heart Rate Variability (HRV)
Heart Rate Variability (HRV) can be used to measure stress, either to evaluate recovery status or exercise intensity.
1 What is HRV?
Heart Rate Variability (HRV) is a measure of the irregularity of the Heart Rate. The time between heartbeats varies slightly, even when the average Heart Rate is steady. For example, a Heart Rate of 60 BPM is an average of one beat per second. However, the actual time between heartbeats could vary so that some beats occur after 0.8 seconds, and some after 1.2 seconds. In the context of HRV, this irregularity is a good thing, and lower HRV indicates an increased level of stress.
2 HRV to Measure Recovery Status
- HRV can be measured during exercise or at rest.
- There are various ways of analyzing HRV that provide different values, and these methods have different benefits.
- Resting HRV tends to decline with training stress, but there are wide variations between individuals and there are other factors that can influence HRV on a daily basis.
- There is evidence that HRV can be used to detect Overtraining Syndrome, but only by comparison with prior HRV data.
- Generally, HRV is greatest at rest and the variability declines as the heart rate rises. Therefore, looking at HRV to Heart Rate ratios is important rather than looking at raw HRV values.
- HRV is linked to aerobic fitness, with the fittest individuals having the greatest variability, and this can be used to predict V̇O2max .
- Lower HRV is associated with greater risk of death after heart attacks.
- Some Running Watches can record or display HRV, and some have software to use HRV to predict recovery or V̇O2max.
3 HRV and Overtraining Syndrome
Overtraining Syndrome is a serious long term problem for athletes. The science around HRV and Overtraining Syndrome is tricky to interpret for several reasons:
- Many of the studies evaluate the change in HRV with increasing training load (overload). This overload is quite different from Overtraining Syndrome and the results do not necessarily transfer. By comparison, few studies look at large groups of athletes to see what happens as some of them suffer Overtraining Syndrome.
- Differing HRV metrics (see below) are used in different studies, making comparison difficult.
- The HRV is often measured while resting but awake, and HRV can be sensitive to changes in mood or stress which are more variable while conscious.
- Relatively short time periods are used, and Overtraining Syndrome typically requires a longer study period.
4 HRV Metrics
There are a number of mathematical approaches to evaluating HRV. Most of these metrics do not adjust for heart rate, so HRV appears disproportionately higher at lower heart rates, confounding analysis. These include:
- rMSSD. This is the square root of the mean sum of the squared differences between R–R intervals. Using rMSSD typically has less measurement error and is less influenced by breathing rate than other metrics. It is also used as the basis of the next two metrics.
- Ln rMSSD. This is the natural logarithm rMSSD, and this produces a smaller number which tends to be in the range 3.0-8.0.
- Ln rMSSD to R-R Interval Ratio. Using the ratio of Ln rMSSD to the heart rate (interval between beats or R-R Interval) adjusts for changes in Resting Heart Rate (RHH). An athlete could have a reduced HRV purely due to a slightly elevated RHH.
- SDNN. The standard deviation of R-R intervals. The problem with SDNN is that if the heart rate is changing, (going up or down steadily), then the SDNN will be inappropriately high.
- High Frequency Power (HF). Spectral analysis can provide the power in the high frequencies, typically 0.15 to 0.4 Hz (high frequency here is relative.)
- Low Frequency Power (LF). Like HF but for the low frequencies, typically 0.04 to 0.15 Hz,
- Normalized LF power (LFn). This is LF/(LF+HF).
- pNN50. The percentage of R-R intervals that differ by more than 50ms. I find this is far too sensitive to heart rate to be of much use.
5 Watches with HRV Recording
There are a number of watches that will record HRV, or more accurately, will record the beat-to-beat time for later HRV analysis.
- Recent Garmin Watches. require you to download enable_hrv_settings_file.fit that you copy onto the watch. You must connect the watch to a computer and copy the file to the folder "GARMIN\NEWFILES", which on Windows may require you to show hidden folders. Simply disconnect and the watch will restart, processing the FIT file. You can disable HRV with this file disable_hrv_settings_file.fit. The watches include Garmin Epix, Garmin 920XT, Garmin 620, Garmin 235, Garmin Fenix 3, Garmin 920XT.
- Garmin 910XT. This requires you to cycle power off and then on again, then hit the up button, then the down button, repeating 10 times until you get the diagnostic menu.
- Fenix 5X. Garmin Fenix 5X has a menu option to enable and disable HRV.
- Suunto Watches. These simply record HRV data automatically.
- Polar V800. The Polar V800 will display HRV, though the details of the calculation are not provided. You can use the V800 to record HRV data, but not as part of a normal workout which limits the value.
6 Software to Analyze HRV
There are a number of ways you can use HRV as an athlete.
- There are a number of HRV Apps for smartphones that are cheap and easy to use.
- Firstbeat has a system that measures HRV overnight and includes analysis software. This is probably the best solution, but it's also rather expensive for the recreational athlete, costing over $1,000.
- Some Running Watches can record HRV for use in Firstbeat algorithms or other analysis.
- A number of running watches have the Firstbeat software built in for calculating aerobic training load and recovery time.
- Running watches also include algorithms for estimating aerobic fitness or training intensities based on HRV.
- K. Hottenrott, O. Hoos, HD. Esperer, [Heart rate variability and physical exercise. Current status]., Herz, volume 31, issue 6, pages 544-52, Sep 2006, doi 10.1007/s00059-006-2855-1, PMID 17036185
- F. Lombardi, Chaos Theory, Heart Rate Variability, and Arrhythmic Mortality, Circulation, volume 101, issue 1, 2000, pages 8–10, ISSN 0009-7322, doi 10.1161/01.CIR.101.1.8