I have been looking for ways to automatically track training status (have you recovered enough for a v02max interval session?) and fitness. One thing I could easily track is “heart rate recovery”. Basically how quickly your HR drops after a hard effort if you stop pedalling and coast for a minute or so.
There are a number of ways I could extract this data
highest 60s HRR in an activity
highest from a bpm of at least 90% of HR max
highest 60s HRR in 1st hour of ride (to avoid impact of drift, fatigue on later efforts)
highest 60s HRR while freewheeling (if cadence avail) … so no number if you didn’t actually try test this to avoid clouding the data over time
Comparing these curves from many rides on the same plot might be useful. Also HRR over time.
Very useful, but it is likely to reveal different types of recovery. For example recovery during a ride will depend on what is happening during that ride, recovery after the ride will depend on users wearing HR monitor at least 1 easy minute after stopping the main effort. So I think you will generate a lot of messy data BUT I still think it is worth doing and I don’t see any other platform offering this.
I think it would be nice to have HRR based on in-workout data, even if not freewheeling, as long as the “recovery” is below some percentage of ftp, say below 60%? 50%? I would likely never free wheel during a training session, even if to get this metric, which I value.
That would be great. Even better would be the ability to analyze HR drops from rides without power. That way, one could come home from a commute and know where to set the pain dial for that evening’s trainer session.
I have made some progress on this. The algorithm looks for the biggest HR drop over a 60s window in the activity starting from a HR of at least Z5 (SuperThreshold). It throws out sections that have a drop of more than 8 bpm in 1 second (likely HR glitch), those with peaks more than the starting HR or troughs less than the ending HR.
I have run it over my own data and get “good” numbers for some rides without having been specifically trying to measure this. Every now and then I do a hill fast and freewheel for long enough on the other side. And in the past when I have been in great shape the “good” numbers are higher so it might have some use as a fitness indicator.
I would like to automatically eliminate the irrelevant numbers. If the activity also has power I could look at that and only use data where the avg power is 25% of FTP or something.
This looks very good !!!
I’ve been experimenting during my ride to push my HR up to above threshold, during a VO2max interval, then at the end of the intervals, just spinning my legs something like 50 to 80 W… this way a get a measurement of HRRc during the ride and should be close real recovery since at the end of the VO2max interval, I really recovering by doing almost no exertion.
HOWEVER… i get a measurement for the ride, but how do I track this across the season? How can i get a trend or check if the HRRc is improving as a result of the training I’m doing?
Another idea would be to be able to plot’ zones and show what is good vs and in terms of HRRc, so people know what to tagret… form my reading, and good number would be 30+, but would be great to see a classification in terms of 'low, average, superior, elite, etc… sort of like the classification done for W/KG for power
Heart rate is very individual so I am not sure that it would be a good idea to compare HRRc to others. Some people are hamsters and hit high heart rates, others much lower, but that doesn’t translate to on-bike performance.
Totally agree… it’s very personal. However, this is like FTP… the absolute figure does not mean much unless taken into consideration with weight, so W/Kg instead of just W
What I’m wondering is about HRRc as a ratio of the LTHR, or HRmax, or maybe even HRR (Hear Rate Reserve).
So, for example, wouldn’t diving HRRc by LTHR, give a % of how fast HR goes down after being at LTHR… this would translate to a relative value that would be measurable across athletes, as it takes into account the LTHR, and not just an absolute value?
HRRc as a % of LTHR might work. I would need to look at that for some athletes with otherwise similar fitness (maybe similar eFTP and weight) and see if it they had similar numbers or something.
I did think about % of HRR but the problem there is that your resting HR goes down as you get fitter (to a point) so your HRR gets bigger which would then reduce HRRc%.
Tx I hadn’t seen that paper before. So if I am reading it correctly measuring HRRc from a lower starting HR (not LTHR which for me is 90% of max) is a better way of detecting functional overreaching (higher HRRc)? And if HRRc is high compared to before your training block you should consider taking it easy for a week?
What would be a good % of max HR to use for the starting point? The paper talks of 60-65% of MAS but I am not sure how to relate that to a % of max HR or a HR zone.
My opinion… the HRR difference would be clearer at lower intensities (60-70% of MAS/MAP) and when a faster HRR is associated to an increased RPE (very important point) suggests f-OR.
The Lamberts and Lambert Submaximal Cycle Test (LSCT) is very similar but adds other parameters (http://www.scienceandcycling.com/lsct/basic-interpretation-lsct/). Personally I like to do the LSCT as a warm up before a moderate or high intensity workout. If I find signs of fatigue I will postpone the moderate/high intensity workout and do some low intensity training. Or you could use something like this paper (https://pubmed.ncbi.nlm.nih.gov/31490431/)
If you want to check HRR at lower intensities I would say a good starting point is 70-75% of max HR. The problem is that a faster HRR could be present on f-OR or on a athlete that has an improved training status and you need the RPE information to differentiate between these situations. One parameter that works really well for me is not being able to quickly reach 90% of my max HR on the third stage of the LSCT. If that happens I know it is time to rest.
Tx. I am going to have to add RPE into Intervals.icu soon! Though it’s not helpful if Intervals.icu says you should have gone easy after the fact, would need to try predict training status instead.
