Blog post by Marco Altini
What's the relation between HRV & performance? Is higher HRV related to better performance?
In previous posts, we discussed user generated science and started exploring the relation between physiological variables such as HR & HRV and fitness, as derived from training load.
In this post, we take a step forward. By including in the analysis actual training data as measured using GPS sportwatches or other apps, and imported in HRV4Training via Strava, we can start looking not only at correlates of higher training load, but also of performance, as derived from workouts data.
Literature & recap of our previous findings
We covered the topic of HRV & fitness in the previous post. We've shown that higher training load was associated with lower heart rate, an indicator of better fitness, while heart rate variability was all over the place. Similarly, looking at my own data longitudinally, I've shown how increased fitness was associated with reduced heart rate, and no change in HRV over a period of 3 months.
The question we want to answer here is however different. Provided that you can (almost) always increase your training load, and potentially improve your fitness with proper aerobic workouts, is your baseline HRV associated to performance? In other terms, while training more was not associated with HRV, is performing better associated with higher HRV?
As I explained in our previous post, it could still be that regardless of training load, higher HRV might be associated to individuals with higher VO2max, due to genetics or other reasons, as reported by other researchers, when controlled by training load. Especially if we consider that baseline HRV does not seem to change that much within a person, genetics might be playing a substantial role.
An interesting recent study by Vesterinen et al. showed that individuals might be responding differently to training plans, based on their HRV baseline (albeit on the usual tiny sample size). In particular, individuals with higher HRV responded better to high intensity workouts, also suggesting potentially a genetic link between HRV and performance (you can find an overview of the paper on Runner's World and the actual paper here).
Dataset & assumptions
For this post we included all users that: 1) used the app for at least 2 weeks in order to gather baseline HR, HRV, training load and performance data 2) were runners 3) linked HRV4Training to Strava so that we could gather workouts data 4) trained mainly on flat routes (see later). We ended up with about 200 runners.
The main assumption behind this analysis is that we can quantify performance. We picked runners because it seems to be one of the simplest sports to analyze in terms of performance, as performing better means running faster. There are several oversimplifications here, however, as we are looking at a broad population with individuals of very different skills, we believe that the best pace during a run should be representative of individual differences in performance.
User generated data: challenges
These are the challenges of user generated data. While we are have little control on the app usage, collecting the right data points and the bigger sample size can provide unique insights.
During an actual study, you would simply ask all participants to perform a treadmill test and measure performance over a standardized protocol or (sub) maximal text. Here we have to be a bit more creative. An additional filter we had to include for example, was a limit on elevation gain, as it obviously has an impact on pace, and we cannot simply factor in this variable and still be able to look at relations across the population.
See for example below how most users running with high elevation gain per workout, ends up being on the bottom side of the plot. Being slow here with respect to others running on flat routes, doesn't mean performing worse, so to avoid the confounding effect of elevation gain on our performance metric, we simply filtered out these users.
We also performed other checks to make sure data can be trusted. For example, in the three plots below we can see the following:
Heart rate, heart rate variability & performance
Hopefully we have established that we have a good dataset of about 200 runners, including heart rate, heart rate variability and performance data, as derived using the best running pace during the period in which data was collected (2 weeks to several months, depending on the user).
Time to look at the data, what is the relation between HR and performance? We saw a strong relation between HR and training load, but do individuals with lower HR (most likely a result of a certain type of training, together with other - genetic or not - factors) also perform better?
We have a weak relation between HR and performance. Here is a quick link to the very strong relation we've previously reported between HR and training load. While there is a trend (a significant one if you are into p-values, with p = 0.0164), individuals with an average heart rate in our population (about 45-50 bpm) seem to be higher performers, possibly more than who is in the 30-35 bpm range. In general, HR seems a poor predictor of performance, in our dataset of highly trained individuals (a very important point).
What about heart rate variability?
The relation between HRV and performance seems stronger, with higher HRV being associated with better performance (p=0.004).
While these relations at the population level are informative and provide new insights, they are still not easily generalizable to each individual. Clearly, there is much spread in the data, and your baseline HRV might not be the best predictor of performance. However, our findings seem to back-up recent research highlighting how individuals with higher HRV might be better at high intensity workouts, potentially performing better, in the context of running. These data also supports the theory that a genetic link between HRV and performance might be explaining how these variables are related, more than training load.
More data. As we can always stratify once more, there's never enough data. Here are a few ideas for our future work. Breakdown by gender & age. Possible improvements in quantifying performance. Data to be analyzed longitudinally within a person to highlight possible changes in physiological parameters, training load and performance. Plenty to do, stay tuned.
1. Context & Time of the Day
3. Paced breathing
4. Orthostatic Test
5. Slides HRV overview
1a. Acute Changes in HRV
1b. Acute Changes in HRV (population level)
1c. Acute Changes in HRV & measurement consistency
1d. Acute Changes in HRV in endurance and power sports
2a. Interpreting HRV Trends
2b. HRV Baseline Trends & CV
3. Tags & Correlations
4. Ectopic beats & motion artifacts
5. HRV4Training Insights
6. HRV4Training & Sports Science
7. HRV & fitness / training load
8. HRV & performance
9. VO2max models
10. Repeated HRV measurements
11. VO2max and performance
12. HR, HRV and performance
13. Training intensity & performance
Camera & Sensors
1. ECG vs Polar & Mio Alpha
2a. Camera vs Polar
2b. Camera vs Polar iOS10
2c. iPhone 7+ vs Polar
2d. Comparison of PPG sensors
3. Camera measurement guidelines
1. Features and Recovery Points
2. Daily advice
3. HRV4Training insights
4. Sleep tracking
5. Training load analysis
6a. Integration with Strava
6b. Integration with TrainingPeaks
6c. TrainingPeaks update
6d. Integration with SportTracks
6e. Integration with Genetrainer
6f. Integration with Apple Health
7. HRV4T Coach advanced view
8. Acute HRV changes by sport
9. Remote tags in HRV4T Coach
10. VO2max Estimation
1. Intro to HRV
2. HRV normal values
3. HRV by sport
4. HRV, strength & power
5. AngelSensor & HRV
6. HRV 101: How to
7. Top 5 most read articles
8. HRV normalization by HR