If you train with power, what is your current TSB? My HRV tends to follow the trends of my TSB.
Based on your “I’m feeling a bit fatigued” comment it may be time to take an easy week. If you are following a TR plan they seem to be built in every 3rd or 4th week. Or possibly your body is working to fight off a pending illness.
I don’t track TSB. My training load hasn’t increased in the last couple of weeks though.
Could be the case. A few people on my team at work were off last week with illness, so I might be fighting off whatever they had.
I ended up doing a low-intensity run yesterday, instead of the higher intensity workout I had planned. My reason for tracking HRV in the first place was to try and pick up on the times when my body was under more stress than usual, but I still felt ok, so thought I’d actually listen to the advice 
Question to HRV4T users. In simple terms, can anyone kindly explain how to read their CTL/ATL graph in the insights section? In the app it shows MAX CTL, no idea what ATL figure is, there is no scale to judge by… MAX CTL, is it good or bad?
hi there, just wanted to clarify that we are not fitting anything here. On the contrary, the point we are making is that the relationship between rMSSD and performance is really weak, and as a matter of fact we do not use HRV in our “fitness” estimates (VO2max, lactate threshold, FTP). The plot includes a line of best fit simply to show how the relationship is not really there (basically flat in the middle, and huge variability, cloud of points). Hope this clarifies, this is certainly not an attempt to derive a model to predict one from the other.
take care,
Marco
MAX CTL means that for the selected period of time, this is the maximum load you have sustained. It’s good because it means you trained consistently for a while, but it’s also uncharted territory, hence caution is a good idea (in my opinion). In Pro we provide a few higher level indicators that might be helpful, you can check them out at the website (HRV4Training Pro) - note also that training load is just training load, it does not account how you are responding physiologically to a specific block, which is the whole point of HRV. I discuss these aspects in this post if you are interested: Serena’s sub-4 marathon. How to use HRV4Training to monitor… | by Marco Altini | Medium
always individual. The best way to look at the data is to ignore the absolute value and take the time to build a good historical background, up to 2 months. From there, you will be able to see how baseline changes outside of your normal values are aligned with periods of higher physiological stress. This is when holding back can provide performance advantage in the long term, as shown in recent studies in both runners and cyclists. Check this out for more information on HRV-guided training: Training Prescription Guided by Heart Rate Variability
this is a bit of a misconception. I’ll try to explain why.
HRV is affected by sleep stage, this is why it is not really a good idea to compare random periods of the night. How does this system (or any other) report sleep stages back to you? Computing them from HRV So yes, either you average a few hours over the night (as done in research for nocturnal measurements) and get an HRV score or you should use only periods of deep sleep to check HRV during the night.
On the other hand, in the morning you are awake, and as long as you do things properly (relaxed, same context and body position, before eating and drinking, etc.) - data is valid and very well representative of chronic physiological stress.
There is also research showing how data during deep sleep for example correlates with morning measurements much better than “whole night data” which speaks to the previous point of HRV plotted as shown here not being really useful as it is highly affected by sleep stage (“Heart Rate Variability During Deep Sleep Offers a Time-Efficient Alternative to Morning Supine Measurements - A Study in World Class Alpine Skiers”).
I cover some of these aspects in our blog as we now integrated HRV4Training with the Oura ring, which provides the same data you see in other sleep trackers: Oura ring integration: read sleep data, whole night heart rate and HRV in HRV4Training
Hope this helps making sense of the data, always remember that it’s about context and confounding factor, and there is nothing that always works or never works, everything needs to be carefully analyzed and most importantly properly used (instead of being misused).
This reminds be a bit of 5 years ago when nobody believed we could measure HRV with the camera, and now all sort of sensors do. I used to read that chest straps were the only way to do it, and they were perfect, but of course that is not the case and now we all know there are many chest straps that are good for HR but not for HRV and they are all potentially affected by artifacts that need to be properly dealt with
hi there,
Marco here (scientist behind HRV4Training). I’ve read a lot of good words about HRV4Training here and I would like to thank everyone that found it useful and supports our work.
This forum came recommended by a friend. I tried to reply to quite a few comments above to provide some help to the ones that had questions about our app. There is quite a bit of information on the Blog, and some recent case studies that I think can help to understand how to use the data, check them out as well.
In any case I’m happy to help anyone that is trying to understand how to use the tools. I realize there is a bit of a steep learning curve, we are not used to work with data with much day to day variability and where positive outcomes are often highlighted by a stable condition (for example baseline within normal values or low coefficient of variation with respect to our historical data, instead of a “higher” score).
This is a good entry point: The Big Picture covering these aspects, it can help you to use the data - it’s quite simple once you get to understand how things work at an acute (day to day) and chronic (baseline) level, and how you can make adjustments based on that information.
A case study where you can see responses to both training and lifestyle can probably clarify a lot: Heart Rate Variability (HRV) response to training and lifestyle: a case study | by Marco Altini | Medium
Thank you and happy training.
@mcneese.chad - @marco_alt would be a great contributor to have on here. Can his comments be un-flagged? He knows way more about HRV than anyone else commenting in this thread.
hi Hugo, I would say indeed that there are broad population level correlations between these parameters, similarly to just heart rate and HRV. However, this of course does not mean that they are interchangeable or that one predicts the other. Blood pressure for example requires both ECG and PPG data to be estimated, as you need to figure out how much time it takes for the blood to travel, which is linked to pressure, I’m saying this just to clarify that PPG-based apps claiming to measure blood pressure are something to be skeptical about (and currently banned from the apple store), as the physiological principles are just not there (while for HRV you just have to measure the time between beats, which can be done using blood flow / PPG, or ECG interchangeably).
This is a very good point. I’ll try to provide some input. The view that training should cause a dip in HRV is a bit simplistic, HRV is a measure of physiological stress, and a high acute stressor is typically highlighted by a reduction in HRV, but chronic changes are hardly simply linked to TSS or training intensity, let alone the fact that in a situation of positive adaptation to training, your HRV should increase (or at least be stable) when consistently increasing training load.
Here is a case study where these aspects are covered: Heart Rate Variability (HRV) response to training and lifestyle: a case study | by Marco Altini | Medium
This is of course all independent of the tool you use (Emfit or morning measurement with an app like HRV4Training).
I would recommend focusing on rMSSD, which is a clear marker of parasympathetic activity (the only aspect that can be reliably measured, and which is linked to your recovery - the “rest and digest” system).
Note also that most HRV features are highly correlated, and they keyword is really “variability” - that’s what matters more than one feature or the other. This being said, using rMSSD or any feature derived from it (as recovery points in HRV4Training) makes it easier to focus on what matters without getting lost with all the different metrics. Some additional considerations here: Heart Rate Variability (HRV) features: can we use SDNN instead of rMSSD? A data-driven perspective on short term variability analysis
rMSSD is also what is used in more than 90% of the recent literature, including HRV-guided training: Training Prescription Guided by Heart Rate Variability
I think they’re still here Marco – just needs somebody with more moderation capabilities to unflag them
Done.
thanks! Hopefully the information is useful. Appreciate your quick help.
make sure always to use the Breathe app in the morning, random data point will reflect acute changes related to any random acute stressor (think working out, but even just having coffee, eating, or just reading an upsetting post online). Context is key, and establishing the morning routine is really important.
In two ways:
You can see this visually here: The Big Picture - hope this helps!
Hi Marco,
Thanks for that. I still think this is a bit misleading:
You could draw any line through this data. I assumed that the think black line is intended to be best fit and the grey range is a 1-sigma variation or something. Certainly, if your point is that there is no underlying relationship, I don’t see how the data presentation helps show that.
But thanks again for your comment, and I don’t mean this to be knocking in any way. I just think the data points by themselves make it clear that there’s no correlation in the data. The thin line and shaded region imply there is.
Good point. I agree with you, this plot probably comes from a series in which the same visualization was used for different variables (including some with a strong relationship), hence the potential confusion. Appreciate your feedback!
Marco - could you please elaborate a little bit more on SWC/normal values. Is this calculated from all rMSSD in the entire time series or only from the 7 days making up the moving average for a given day?
Edit: just noticed in the chart that your normal values are for 30 days. Is this correct? So you take the standard deviation for the last 30 days? And construct the green band around the mean for the 30days for each day?
Edit2: and the 0.5 or 0.75 std is only to one side of the 30d mean or is it 0.5/0.75 for both sides, e.g. 0.25/0.375 for each side to the mean?
