Tuesday, August 20, 2013

Iron Rider's Recovery Based Training

I began a recent post on training with the statement "I don't get very many comments on my blog, but I love it when I do" and Iron Rider obliged me with the following friendly and helpful comment on that post:

How often are doing your MAF tests and training rides? ... What prompt[s] me to ask the question was your description of riding the MAF tests at a heart rate of 130-140 bpm for 45 min (not including the 30 45 min of warm up and cool down!) That seems to be a fairly strenuous aerobic workout. While I don't know your max heart or heart rate at lactate threshold, 130-140 bpm would almost always fall into the high aerobic zone for a man over the age of 45 ... doing 4-5 high aerobic zone workouts week is definitely working hard with very little recovery time. 

One possible reason why your 130-140 bpm sessions may seem "easy" (Zone 2) is that if you do not recover adequately then you start each day with an elevated resting heart rate. As a result, an easy workout gets the rate up to an extent that it would would not if you were fully rested. If this makes sense to you and you are concerned about recovery and overtraining, I suggest monitoring your resting heart rate first thing in the morning. If it increases, you are not recovering. If it decreases you are getting the result you want. 

(I have edited his comment to emphasize the points I am making and hope I have not thereby altered the meaning or tone.) Let me start with the following points:
  1. I really appreciate comments; even if I happen to disagree with a particular comment, I appreciate that someone has taken the time to talk to me. That said, I certainly did not disagree with this comment.
  2. I am a scientist by training, and one of the rules of thumb in science is that a fact is more valuable the more surprising it is; an unexpected fact is more likely to tell you something you did not already know. Iron Rider's comment was very surprising to me; I had been convinced that a training ride consisting of just a MAF test was not at all strenuous. Thus, his comment really made me reconsider what I am doing. And, to answer his question, his point makes a lot of sense to me and I am very concerned about recovery and overtraining.
It seems to me that there are two main points to Iron Rider's comment:
  1. Based on the standard calculations used to estimate the maximum heart rate for a man my age (64), my MAF test rides seem like they might be quite strenuous and that as a result, adding rest days to my schedule might be beneficial.
  2. By monitoring my resting heart rate, I might be able to determine when rest days are needed.
    Let me start with Iron Rider's first point, that my MAF tests are being ridden at a high heart rate for a man my age. I take this point very seriously, and here are my thoughts on this issue:
    1. I have previously posted on how I have estimated my maximum and lactate threshold heart rates and why I think my heart rate is higher than average for my age. These estimates are not as good as I would like them to be, but I do find them plausible. That said, I admit that the values I am estimating for my maximum and lactate threshold heart rates are very high, well outside most predictions.
    2. I have not yet made a direct determination of either my maximum heart rate or my heart rate at lactate threshold. Joe Friel argues that it is unnecessary and unwise to measure your maximum heart rate, that lactate threshold heart rate is just as useful while being easier and safer to measure, and I suspect my cardiologist would agree (though I will ask him at my next visit.) I hope to make a better measurement of my lactate threshold heart rate in the future, but until I do, there is not much more for me to say.
    3. In support of the heart rates I determined, I used those heart rates to define my heart rate zones for training, and the level of exertion I perceive match with the Joe Friel's perceived levels of exertion he associates with the various heart rate training zones. Iron Rider argues that this match could be misleading because it might result from training with inadequate recovery. However, because this match always exists, independent of whether this is a MAF test ridden after a day or two off the bike or is the fourth MAF test in a row, I am inclined to give it some weight. From Joe Friel's data, I have compiled a table showing the linkage between training zone, heart rate, and perceived effort. I cannot sing during my MAF tests, but I can talk in complete sentences.

    This brings me to Iron Rider's second point, that resting heart rate can be used to determine when one has recovered and thus when to plan the next strenuous training ride. He has built a training program around this idea which he named "Recovery Based Training". He provided a link to one of his blog posts on this program in his comment:


    ...and that post contains a link to the original post describing the plan:


    You should read about this plan on Iron Rider's blog, but to my eye, it is truly inspired and seems to have worked very well for him indeed. Fortuitously, I have been measuring my resting heart rate first thing each morning for over 16 months, since before my first brevet or blog post. My records are not perfect, there are many days when I forgot to measure my heart rate and the way I am measuring my heart rate (discussed below) may have problems, but there is enough data to do a preliminary analysis. Thus, I am in a position to ask retrospectively, if this same approach would have worked for me? In addition, I can ask the related question, am I riding MAF tests with too little time for recovery between them?

    Simply looking at my resting heart rate data, I see no evidence of it changing in any organized way; it has varied from 54 to 76 bps, mostly between 65 and 70 bpm, but not with any pattern I can fathom; it is not obviously lower after a rest day or higher after a hard day. At a first approximation, this answers the question; if I cannot see obvious changes in my resting heart rate, I cannot use my resting heart rate to decide when I have recovered and when I have not. However, I wanted to dig a little deeper, so I did a statistical analysis of the data to see if I could uncover trends in the data not obvious to visual inspection. What I did was to define 6 groups of heart rate values:
    1. Those measured the morning after one full day or more entirely off the bike. This should be a fully rested state.
    2. Those measured the morning after one day of MAF tests (the previous day being off the bike.)
    3. Those measured the morning after two days of MAF tests.
    4. Those measured the morning after three days of MAF tests.
    5. Those measured the morning after a 200K brevet. (There are only two of those.)
    6. Those measured the morning after a long training ride, one 60 or more miles long. I selected this distance because it is at about 60 miles I start finding these rides difficult to complete.
    For these comparisons to be valid, it is important that my resting heart rate not be changing over the time that I rode these MAF tests. That turns out to be the case. For my measurement of average resting heart rate after rest days and MAF tests I used data from 2013. Here is all my heart rate measurements from 2013, and although there is considerable day to day variability, the data is not trending up or down, so averaging the data over time makes sense. 

    "Resting" heart rate each day I measured it in 2013. The darker line through the data is the least squares trend line through the data. The fact it is close to level indicates that my average "resting" heart rate has been constant over the time interval in which I am interested.

    The following table gives the averages of heart rates taken the day after the above six kinds of training days described above. The first column lists the kind of training day, the second the average of the heart rates measured the day after each of these kinds of rides, the third the standard deviation (a measure of how much individual measurements vary from each other), and the last the number of heart rates averaged.

    Average Heart RateSDN
    Rest Day
    1st Day MAF 
    2nd Day MAF
    3rd Day MAF
    200K Brevet
    Long Ride

    Iron Rider defines three ranges of heart rates. The first, called "Rested", is a heart rate equal to or less than his resting heart rate. The third, called "Elevated", is a heart rate 10% or more higher than his resting heart rate. The second, called "Tired", is anything in between. I am a bit confused by this division because any measurement is going to have some variability and normal variation will mean that, even if one is never tired, one's heart rate will be above one's average resting heart rate half the time. I will assume that Iron Rider has a very stable heart rate and he simply doesn't see much variation, so this is not an issue for him. Unfortunately, that is not true for me, and therein lies a tale.

    If I round my normal resting heart rate to 66 beats per minute, anything above 73 beats per minute would be "Elevated", and anything between 67 and 72 beats per minute would be "Tired". The problem is, given the observed variation in my heart rate (shown in the SD column in the table above), even if I am never tired, I would expect to experience an "Elevated" heart rate every week or two and a "Tired" heart rate every other day. This probably makes it impossible for me to follow a "Recovery Based Training" regimen.

    Even though I cannot use the "Recovery Based Training" regime on a day to day basis, by averaging the data from multiple rides, I might be able to reduce the variability in my heart rate data enough to ask the question "are MAF tests leaving me tired?" Simply looking at the average heart rates would suggest that they might be; my average heart rate increases with every sequential MAF test. However, once again variability rears it ugly head; when I used a standard statistical test (the t-test) to compare the average heart rate after rested days to that after 2 or 3 days of MAF tests, the difference was not significant. That is, the difference in average heart rates could well be due to chance. The only comparison which was significant1 was that between my average resting heart rate and my heart rate after a 200K brevet. In fact, my average heart rate after a brevet qualifies as an "Elevated" heart rate. I definitely need to rest after a brevet!

    An obvious question at this point is am I measuring my resting heart rate correctly; can I reduce day to day variability by changing how I do that? The way I currently measure my heart rate is based on convenience. I get up, brush my teeth, get dressed, walk down stairs, and use my blood pressure monitor to measure both my blood pressure and heart rate. A few days ago, I tried something different. I took my Polar heart rate monitor upstairs when I went to bed, and then put it on when I woke up, while remaining still and in bed. What I found is that my heart rate varied between 63 and 69, all while lying quietly in bed. I left the Polar on as I went through my morning routine and saw heart rates up into the 90s. When I sat down to measure my blood pressure, it dropped to 75 while the heart rate measured by the blood pressure meter was 74. The lesser problem was that the variability in heart rate before I got out of bed was at least as great as what I am doing now and thus changing my routine will not solve my variability problem. The greater problem was that my wife was not at all pleased by this change in routine. She is used to sleeping in for a few minutes after I get up and her sleep was disturbed by my heart rate experiment. Today, at her suggestion, I came downstairs as usual, laid down on the couch, and used my blood pressure monitor to measure my heart rate five times in a row while lying quietly. I measured heart rates of 72, 70, 69, 67, and 71, in order; variability continues to be an issue.

    Why is my heart rate both so variable and so high? I think this is just an example of the fact that we are all individuals and that as a result, we are all different one from another. That is why the medical community is so excited at the prospect of personalizing medical care ("precision medicine" nĂ© "personalized medicine") and why the exercise community has advised us for years to personalize our exercise routines. Thus, I completely believe that  Iron Rider's "Recovery Based Training" works for him and suspect it will work for most of you. If it does works for you, I highly recommend you give it a try. I would if I could, but I can't. That said, I certainly found Iron Rider's comment useful and I haven't completely given up on using his approach. I have been re-motivated to directly measure my lactate threshold heart rate and I will continue to experiment to see if I can find a way to measure my resting heart rate that yields less variability.


    1) I am aware that doing multiple t tests on the same data can make data seem more significant than it is. I did so because I knew I was going to have trouble finding any significant differences at all and I wanted to maximize sensitivity. Because the results I got were NOT significant, this concern does not arise.

    2) I decided not to post MAF test results this week because household duties cut into my training last week so, compared to last week, I have nothing interesting to report.


    1. That was a very interesting and detailed post. Thanks very much for the feedback and your thoughts. It's clear that you have thoroughly read my posts on this subject and I appreciate that since, as I said, this theory has been an experiment of one (as far as I know); therefore, getting other perspectives is very interesting and valuable to me.

      I would like to address your important question and concern about the variability of your heart rate and finding a reliable basis for measurement of resting heart rate. Those are valid concerns and I wanted to share how I address them.

      First of all, let me say that my heart rate is also very variable by which I mean it reacts quickly to changes in my physical state. From wearing my polar monitor I know that my heart rate quickly changes in response to sitting, standing, walking and yes brushing my teeth, even stressful thoughts.

      As a result, my resting heart rate test protocol is very strict because I know that very small actions can cause different readings. So here are my protocol considerations and the thoughts behind them:

      First, I check my heart rate before getting out bed or doing anything else – even sitting up. (It only takes a minute or two since I either use a monitor or check for 15 secs and multiply by 4. My wife doesn’t even notice when I do it.)

      Although I may check my heart rate several times over a few minutes, I only use the lowest number from my morning test. I do not average the varying numbers I might get because any number higher than the lowest number arguably reflects “some stimulus” that has raised the HR. Think of this way, if you wanted to find your max (achievable) heart rate, you would not average your high numbers from testing since any number less than max HR is by definition not the max. Only the highest number would count.

      If I could digress for a section for a point of clarification - when I say max heart rate, I really mean the max heart rate that a person can achieve in an activity. Experience has shown me that that “max heart rate” can be sport specific, for example I can achieve a higher max rate running that I can cycling (maybe because cycling is partially non-weight bearing?) Similarly when fitness changes the “max” heart rate can change as well, for example if you are now fit enough to push even harder you can have a higher “max” (So maybe I should refer to functional max heart rate?) In any event, back to the testing . . .

      Similarly, you cannot (under normal circumstances) lower your heart rate beyond its lowest number for your condition on any given day. Therefore any number above the lowest of the morning I discard. That gives me just one number to work with.

      Using your numbers then – on the day when you recorded a 63 and 69, I would call that a 63.

      Another consideration is that one or even two days of rest may not be enough time off to show your lowest “rested” heart rate especially if you’ve been training 4-5 days a week for one or more weeks. In my case, my initial rested HR was determined after a week off from any strenuous exercising.

      A final side consideration - lowering your resting heart rate through exercise is a very long term process and is not the goal of my theory. The goal of the theory was to find a way to train efficiently, recover fully and participate in this taxing sport while working, raising a family and doing that balancing act we all have to do, without overtraining or driving myself into sickness. If there is a better way to measure recovery (especially for endurance exercise), I am not aware of it and would be very interested to learn about it but I am happy to say that that by using my theory I have achieved that goal over the last few years and continue to do so. I hope you find success as well.

    2. And to add one more consideration: after re-reading your post, here's how I would interpret your numbers using my theory.
      First - Your heart rate when fully rested is 54.
      Second - Your MAF rides/tests shown in the chart were being perfomed while you were in a "tired" (possibly "elevated") HR condition and the level of underlying or carryover fatigue increased from day 1 to day 3.
      Third - After the brevet your were probably in an "elevated" state.
      Fourth - Since May 14 your morning readings have largely been above 60 bpm and since April 21 your lowest morning readings have risen by almost 10 BPM (from 54 to 64/65).

      Based on that intepretation, my plan would called for:
      Taking some rest recovery time to get the morning HR back down at least to the 50's.
      Decreasing the frequency of the MAF rides to allow for (more) recovery between them.
      Then after the Brevet it would've called for a long enough rest period for full recovery before resuming hard training.

      Again this is not to suggest that my program is "right" for you but just to clarify how it works. Basically it's more "food for thought."

      Good luck with the training and remember to enjoy the ride.