Thursday, June 07, 2012

Noninvasive measuring of LT threshold and Calibration Runs

I mentioned a few posts ago very interesting paper I found about the title line subjects..let me discuss that today.

Back when I started running 5 years ago I didn't have a Garmin. I ran for one full year with a very simple pedometer to measure my distance and didn't even measure my time. Of course, I weight 228 lbs at the start  and I knew I had to take it easy.

After a year, the pedometer wore out, ;), and I did NOT wear out. I was thin and running much better and so I treated myself to a Garmin 405 with HR monitor. Geek heaven!

Within a few months of Garmin running I noticed that there seemed to be a pretty straightforward relationship between HR and pace. I the relationship linear? It seemed so...but googling I could not find ONE article or posting about the idea. 

So I got the notion of going to the track and running N laps with each lap slightly faster and then doing a plot of the speed vs HR. I've already shown you these plots a few times but here's a link to one.  The relationship is, in fact, linear for the speeds I was running. Cool! The slope tells you how many BPM you need to increase to move up your pace. The intercept tells you the overall shift of the line up and down (basic fitness). 

About 2 weeks ago 2Slow4Boston had a posting where he talked about getting his Lactate Threshold (aka "LT") measured. This involved a skin prick to get a drop of blood and a lot of hard running.

That caused me to get interested in ways of finding LT without a blood sample..that would be awesome, methinks. So I did some googling for that idea and found this:
When I saw that title my bpm went way up! 8). To my surprise and further excitement the method involved doing exactly the calibration run idea!! Bravo!

Here is a link to a copy of the paper.

There is one difference from my calibration graphs: you can see immediately ..there is a little curved part at the top of the graph.

At the point marked "Vd" which is called the 'deflection velocity' the pace starts increasing but the HR does not. You are into anaerobic running at that point so your HR does not can't help you bu t the anaerobic running can keep you going (for a while).

So, why didn't I see "Vd" in my calibration runs? In reading the paper it's clear that you need enough data points at very high speeds to see it. I increased pace per 400m lap.  They used 200m to get more points and get to higher velocities before they were too tired.  

Once you get a clear signal, you can measure the LT as shown in the next graph.
This graph shows the relationship between the blood method and and the "Vd" measurement. You can see the points start curving over at the same point lactate starts shooting up. It makes perfect sense, since the anaerobic metabolism is just kicking in right there.

How well correlated is this relationship? They measured that too in the next graph shown here. This shows a 10 measurements of Vd vs LT. 

Basically what this is saying is that if you can get a good measurement of Vd you can get a very very good idea of LT.

In the paper there is more detail, but these experiments were done with a large number of runners of various abilities. All were pretty seasoned runners though.
The graph above shows two runners and how their calibration runs vary as they get trained. The line moves down, the slope gets a teeny bit less steep and the Vd point moves to higher speed. 

As in interesting side note..notice that for the right graph (runner "B") the 3rd graph is actually much worse then the previous two...he got worse! If you read the comment they say that this degradation in performance happened about 2 days before the guy was diagnosed with mononucleosis! He probably felt like crap during that run!

Also in the paper are graphs that show some marathon paces people ran successfully vs their LT pace tested just before. We all have our ways of estimating our pace and this would add another tool!

Calling all volunteers!

Michael was kind enough to give me some garmin tcx files to shake out my software I have for doing these kinds of plots...thanks Michael!

Now I'm hoping that I can find 10 more strong runners out there to do  calibration runs and send me your tcx files if you want to get a shiny graph right back 8)

How I  do it: 

1) First: I am not an expert at anything nor a doctor. Running really fast is could injure yourself any number of ways or even die. If you do this it is your free choice, I accept no responsibility for your actions!

2) I use a running track on a day that below 60F and less than a few mph of wind....find two points marked on the track you can remember that are a half lap apart.

3) Warm up at a SLOW pace for about  mile...keep your HR low...the idea of this is to gradually increase your HRs over time so if you warm up too fast you'll mess up the result. Try to practice keeping your HR as steady as possible for the warm up. 

4) Trigger your lap button at one of the points on the track after your warm up and run the next half lap at the same HR as your warm up... Then, speed up very slightly by 3-5bpm or so and run the next 200m and hit the lap button again.

5) Continue speeding up every 200m and trying to keep your HR steady during that time. Don't worry if you aren't will still work, but try your best. Don't forget to hit the lap buttons!

6) Eventually you will be going really fast and getting pretty tired. If you can, try to keep going as close to your max HR as you can. This is where you will produce the data for finding Vd. If you can't hold, thats ok too you'll still see your slope and intercept to compare with another time.

7) After your last 200m split stop your watch and save the tcx and send it to me! (you can either find it on your computer or you can download it from Garmin Connect online after your upload..assuming you use that)

I will send you back a graph of your data along with many thanks! If you do one of these runs every 6 weeks or so during heavy training you will get to see your improvements.

Send your tcx files to 

paul50isthenew30 at gmail dot com

If you don't get a reply it means I've been buried! 8)

..if this works out I'll  set up a web engine so anybody can do it...upload tcx, get back graph.


  1. I am NOT volunteering, but I had to comment that this post made me smile. Done like a true running geek!

  2. This is super cool! I'm also not volunteering at the moment, but I'm curious for the results. :)

  3. I am totally interested...but what is your deadline? I am working on getting back into speed.

  4. Wow, thanks Paul! I like this method better than the one I read on I will do it and plot my points. I have to wait a couple weeks because I have a race coming up and I don't want to injure myself. Again, thank you.

  5. Paul - awesome stuff. I would love to help. I just need to know two things: 1) can slow people like me help? 2) I may be delayed in helping - is there a deadline? Please let me know.

    I have a md appt this or next week to discuss my foot; I think that I have a stress fracture or strain.

    1. THis is going to be an ongoing thing..(I hope), so don't worry about it.

      Slow is just fine (and btw you are not slow). If we don't have enough leg muscles/endurance to see the deflection point that is still useful: says you have other things to work on first! 8)

      Hope you get some good advice from Dr.


Feel free to leave a comment!