Learning to trust the drift test

  • Creator
  • #44997

    Fellow trainees,

    I’ve been trying to nail down my AeT for uphill hiking (my primary activity) and indoor cycling (secondary/backup/alternate activity).

    I’ve been ramping up level of effort (slope/speed on treadmill, targeted power output on exercise bike) and looking for cardiac drift. What I think I’m finding is that in order to achieve cardiac drift of ~5+% on a one hour effort, I’m putting out an effort that violates all of the subjective measures of being within AeT (breathing intensity, perceived effort, indefinite endurance). If I achieve a small cardiac drift, I’m usually completely spent at the end of the hour test, I’m definitely not able to nose breathe, and my heart rate is typically 80-85% of what I think my max heart rate is.

    I think from my reading that Cardiac Drift should be a more reliable measure than the subjective metrics, but I thought I’d ask here. Do someone people just not cardiac drift until further over Aerobic Threshold?

    I am not a well trained athlete – I am new to endurance training.

    I have significant data that I can post, just not sure what would be helpful yet.

  • Inactive
    Anonymous on #45479

    Another thought: It’s probably worthwhile to get a lab test done. If aerobic capacity is large enough that drift tests are very fatiguing, then (IMO) they have limited usefulness. At that point, I think it’s worth being more precise via lactate or a gas exchange test.

    danspam on #45482


    Thank you for taking a look. The posts above got a bit out of order. Some of the older posts that I mentioned that I thought maybe got caught in a spam filter suddenly appeared recently but were originally posted in response to Scott Johnston’s responses. You can see which ones I’m talking about by the out-of-sequence post #s.

    Almost all of the data is recorded via chest strap (Polar H10). On the specific test you noted, my watch lost pairing with the strap and fell back on its own internal (optical, wrist-based) meter. This does illustrate a possible vector for silent data corruption. I will turn the wrist-based meter all the way off for future tests to be sure.

    I am in possession of a lactate meter, but I’ll have to rely on solo testing and so far, my results have varied wildly with immediate consecutive tests, leading me to believe that my collection skills are not yet up to snuff.

    I have an exercise bike workout planned today and I will try testing during the workout to see if my skills have improved enough to get steady/reliable results without making today a rigorous AeT test.

    I live in a suburb of Atlanta, GA, and the closest place I can find to do a MET with a protocol that looks like it will find AeT is in Augusta, about a 2 hour drive. I will definitely go get a test if I keep getting confusing results at home.

    Thanks again for looking at the data

    Anonymous on #45527

    On the specific test you noted, my watch lost pairing with the strap and fell back on its own internal (optical, wrist-based) meter.

    I would shut off the optical monitor permanently so this can never happen.

    If you have your own lactate meter, you should be all set. If you have to pause between work intervals, that’s fine, just make sure to sample at the same time after each interval.

    My suggestion would be 5x5minutes with 60-90 seconds between repeats, starting at a moderate speed and increasing the speed by 10-15sec per mile until the last two repeats are faster than LT.
    ~ Steve Magness, The Science of Running, 2014

    I’ve tried different lengths of stages and three minutes seems to work just as well.

    danspam on #45541

    I may go through $50 in test strips before I get the skills to get solid readings down. I tried to take some readings during last night’s workout. 4 minutes into the warmup I got an sensible reading (1.4) but at 15 minutes in, I got 3.5, 2.1, 3.9, E, 1.4 on consecutive readings, whereupon I gave up for the night.

    as always thank you for your input.

    Anonymous on #45545

    Have you read this?

    Blood Lactate Test Protocol: Tips and Tricks

    danspam on #45551

    yes, I have; so, I was well prepared for failing a lot before I start to get the hang of it.

    I tried testing again some during tonight’s workout, with a few minor tweaks to my technique, and with one more day’s experience, and I think I got more sensible results.

    The workout is here

    I was aiming for constant 140W, originally planning to use this as a bit of a drift test, but as my lactate meter reading improved, I switched up a bit at the end.

    At the first dropout (around 0:10:00), I got readings of 1.4 and 1.4
    At the second dropout (around 0:50:00), I got readings of 1.7 and 1.9

    I decided to increase my target power to 150W for the final five minutes and see what lactate reading I would get. Immediately after finishing, I made a reading of 2.3. I did not have any more strips on hand so I could not repeat my reading.

    My TP premium has expired so I can’t easily run the numbers until I export the data and do my own spreadsheet; visually it looks like this test may have shown a little bit of cardiac drift, which would be somewhat consistent with the lactate readings, I think. Is that right?

    I’ll do a rigorous test on the treadmill soon.

    Anonymous on #45575

    Sorry for the confusion… If you have a lactate meter, you don’t need to do a drift test.

    * Pick a power level, hold it for 3-5 minutes, take a sample.
    * Repeat with increases of 30-35 watts.
    * Continue until you get readings around ~2 and ~4.
    * Continue to MaxLASS if you’re curious.

    danspam on #45577

    Scott – no worries and nothing wrong on your part. I only switched up mid ride from a (sort of) drift test because I was (finally) able to get good readings on the lactate meter.

    I will follow your protocol on my next biking day. Today will be incline treadmill and I will do the same with increasing speeds at a constant slope.

    Thank you!

    danspam on #45580

    After reading the link on MaxLASS, I am a bit confused:

    1) They don’t seem to say how to determine MaxLASS
    2) They seem to imply that calculating V4 is more useful and discourage determining MaxLASS despite talking about it in the first module
    3) They seem to indicate that 7 minute segments are better for lactate testing than 3-5. Thoughts?

    Anonymous on #45601

    1) It’s not useful for training to be that precise. MaxLASS will change based on recovery and training.

    2) V4 is more useful if it’s used as a proxy for MaxLASS. Many misunderstand Jan Olbrecht when he suggests using V4, assuming he equates it with MaxLASS. He doesn’t. He suggests using V4 for tests because it’s easier to monitor changes in one variable (V) if you keep the other constant (~4 mM). That, and the power output between V4 and MaxLASS is rarely significant.

    3) I’ve never found longer stages useful. Besides, at 7′, you could do two tests at the same speed each 3′ if you wanted to.

    danspam on #45608

    Well, I did a lactate test this morning before your post. I did seven minute segments and took two (or in case 3) readings after each segment, which resulted in about 2 minutes of downtime between segments. This may have corrupted my test, but I still felt like I got reasonable looking data out of it.

    The whole test was on a Sole F63 treadmill @ 15% incline. Raw data below:

    speed End HR M1 M2 Avg M
    0 0.9
    1.8 105 0.8 1 0.9
    2 110 0.7 0.8 0.75
    2.2 117 1.6 1.5 1.55
    2.4 122 1.2 1.2 1.2
    2.6 126 1.6 1.5 1.55
    2.8 132 2 1.9 1.95
    3 144 2.7 2.6 2.65
    3.2 153 3.2 3.7 3.45
    3.4 160 5.4 4.7 5.05
    3.6 171 7.3 7.2 7.25

    on the 3.2 segment I got an E4 as my first reading, and then 3.2 and 3.7 on subsequent readings.

    It seems based on this that at 15% incline, 2.8mph and 132bpm is right up against my AeT.

    You must be logged in to view attached files.
    danspam on #45612

    forgot to add: speeds are in mph. The first (1.8mph) segment was 15 minutes warmup, all subsequent segments were 7 minutes. Also, I weighted 169 pounds at the end of the test, in case conversion to power is of any utility.

    Anonymous on #45726

    I haven’t tried or heard of a threshold test with multiple samples per stage, so I don’t have the experience to comment. My first thought is that it may have affected the readings. Within the two-minute rest interval, were the samples taken at the same time? For example, 45″ and 90″ into the interval?

    Assuming that the protocol didn’t affect the results, I would agree that the 132 bpm mark seems about AeT. The only question there is that lactate is continually rising from two stages before that point.

    The good thing about having your own meter is that you can learn more about this test as you test in the future. I suggest using a more standard protocol so it’s comparable to general accepted practices to reduce the unknowns.

    danspam on #45728

    Generally speaking, the tempo of the samples was consistent, EXCEPT that on one segment I got an E4 reading for my first sample, so the 3.7 was “late”

    Thank you for your comments and suggestions. Yes, next time I will do a single reading and five minute intervals. I’m currently waiting on a new batch of test strips to arrive. I’ll probably attempt a exercise-bike based text next.

    One tangential question: since I’m able to control effort/power applied directly, is it better to use power as my “knob” or heart rate when focusing workouts on building aerobic base? I would have guessed “power” based on the totality of what I’ve read here at this sight, but I’m not 100% sure, since you guys do often talk about HR, _and_ I have seen that at a given power level, my heart rate can vary considerably based on other factors.

    Many, many thanks for all of your input.

    Anonymous on #45767

    If they can be consistently measured, power (on a bike, not Stryd) and pace (on flat runs or on a treadmill) is better than HR. We use HR because consistent output in variable terrain isn’t measurable, and RPE is unreliable for 99% of people.

Viewing 15 replies - 16 through 30 (of 31 total)
  • You must be logged in to reply to this topic.