Half marathon and full marathon time estimation in HRV4Training Pro

Blog post by Marco Altini

​We have just released our latest feature in HRV4Training Pro: half marathon and full marathon running time estimates. In this post we go over how these prediction models work.

In particular, this work is an extension of our previously published analysis (see "Estimating running performance combining non-invasive physiological measurements and training patterns in free-living” which was accepted for publication at the 40th International Engineering in Medicine and Biology Conference, full text here, while another blog post explaining the paper can be found here).

In the published work we built models able to estimate running performance (10 km time) using 2 years of real world data from more than 2000 individuals, including morning physiological measurements obtained using HRV4Training, workouts acquired from Strava and TrainingPeaks, anthropometrics and training patterns.

What did we learn?

We provided insights on the relationship between training and performance, including further evidence of the importance of training volume and a polarized training approach to improve performance.

While no causal link can be established, as users did not participate in an intervention, it is of interest to determine the impact of features representative of training patterns as derived from workouts, for example training polarization, a hot topic these days. In our analysis, age, BMI, resting HR, speed to HR ratio and time spent at moderate HR intensity entered the model with a positive sign, meaning that a lower value for these predictors is associated with a faster 10 km. On the other hand, HRV (rMSSD), average distance and speed, percentage of workouts performed 5% faster or 5% slower than the average training entered the model with a negative coefficient. Thus, according to our dataset and analysis, a more polarized training regime, with a higher percentage of workouts preformed either faster or slower than the average workout, as well as a lower percentage of workouts performed at moderate HR intensity, is associated with improved performance.

Lactate threshold estimation in HRV4Training ​

​Following our analysis, earlier this year we have released a new feature in HRV4Training, lactate threshold estimation (more on this at this link), basically turning around the modeling detailed in this post and published in the paper.

In practical terms, the lactate or anaerobic threshold, is approximately the pace you should be able to hold for a distance between 10 and 15 km. This is the criteria used in HRV4Training, which should help you making sense of the app estimation.

Intuitively, knowing your lactate threshold can help you defining pacing strategies for racing events between the 5 km and the half marathon (or longer, but in that case, other factors such as training volume start to play a more important role), as well as determining training pace for intervals and tempo runs. Some useful insights in this context are provided by Greg McMillan in this article, that I’d recommend checking out.

Estimating running time

​As mentioned at the beginning of this article, we have now added new models to estimate half marathon and full marathon times, based on a similar analysis, and the following parameters:

• Resting measurements of cardiac activity (heart rate and heart rate variability).
• Physiological data during training (for example the running pace to heart rate ratio, that we described in our VO2max estimation paper).
• Training pace and volume.
• Training patterns (for example training polarization, or time spent at low or high intensities vs time spent at moderate intensities, as well as time spent at different speeds).
• Previous running performance (automatically derived from racing events or hard intervals workouts).
• Estimated VO2max.
• Long runs (number and frequency) .

You will see a different degree of confidence on the models, as we try to move away from ‘exact’ estimates, as there is no such a thing and all models include an error, to provide you with a range of most likely values that can help you, given your knowledge of your workouts and physical condition, getting a realistic understanding of what racing times could be possible based on the available data (R2 was above 0.85 for both models). This is what you see below as optimistic and pessimistic values.

In addition, the models differ as the full marathon estimation model highly relies on the presence long runs among your workouts, an aspect often forgotten by other estimators, and which we believe is key to performance in long distance events.

Below you can see another example, we hope you’ll find the new feature useful.