Lab tests, 20 and 90 minute time trials

There's frighteningly little literature on how lab tested variables relate to the standard field tests cyclists use.  There's more than I'll ever be able to read when it comes to comparing in-lab tests to other in-lab tests.  VO2max and peak wattage in a ramp test?  Got it.  Comparing MLSS, Dmax, CP, OBLA, or a certain % of VO2max?  Got loads of those.  I've been digging a lot to find some good studies that actually use things that cyclists do on a daily basis.

Enter exhibit 1.  20 and 90 minute time trials from elite and internationally competing cyclists were used to test a variety of previously defined in-lab LT markers.  It's got some very, very interesting data.  First, 20 and 90 minute TTs were not well correlated: r =0.66, p=0.54.  I would rather they had done a 60 minute test because of the relationship we always hear between a 20 minute maximal effort being 105% of your FTP, which is ostensibly the power you can hold for an hour.  But it's actually fine that they didn't use a 60 minute test, because a 90 minute test is going to be pretty close too, and I'll examine why.

We've seen of TTs that are well over an hour, like this year's Giro Stage 14 that was won by Vasil Kiriyenka (spoiler alert: he just took the world ITT title).  Second on the Giro stage, LL Sanchez, who also competed in worlds, was down 2:45 at worlds and only 0:12 down in the Giro.  I know, worlds and grand tour ITTs are different, but because Sanchez and Kiriyenka competed in the Giro, Vuelta, and worlds this year, so we can feel comfortable in comparing their performances.  As a side note, many other top finishers in the Giro TT were well known breakaway riders and long TTers, such as Sylvain Chavanel and dare I say Alberto Contador on the rare chances he can get away, like Stage 17 of the 2012 Vuelta.

Why'd I go over that?  Oh yeah, 20 minute power doesn't correlate well with 60 minute power, and most likely doesn't correlate well with 90 minute power.  In the graphic above, the x-axis is self explanatory, but the y-axis test went like this: 3 minutes at 50% VO2max (previously determined), then increasing the workload by 5% VO2max until voluntary exhaustion.  Highest wattage reached during this test was the Wpeak.  And how well did this correlate?  Compared to most other variables tested, very well.

All the variables tested here are lab things, nothing cyclists can generally do on their own.  They all require equipment that measures VO2max or blood lactate (though there are some semi-affordable ones of the latter) and occasionally a little graphical or mathematical inspection to find inflection points and tangential lines.

What I find particularly fascinating is the low correlation of VO2max and the TT powers.  This kind of thing shows that over longer periods of time (20+ minutes according to this study) the maximal oxygen transport that a body is capable of producing is limited by other physiological factors.  For instance, over a 2:10 marathon, an average of 80% of the fuel used is carbohydrates (source).  So we can conclude that a 60 or 90 minute time trial, though creating great aerobic stress, are still primarily glycolytic in nature.  A rider's muscle fiber makeup, VO2max, and other factors like fatigue will influence performance, and just how aerobic a 60 or 90 minute TT really is.  That's a fascinating paper as well as a classic, so I'll get into it more sometime later.

Now what can we conclude about a 20 minute time trial?  It uses a very, very large proportion of glycolysis as the primary fuel source.  I was always skeptical of the standard 20 minute test being preceded by a 5 minute maximal effort, but now it's a bit more obvious why it's there.  Burning through your first hard effort when you're fresh is a good way to get your 20 minute time trial to behave more like something longer.  For riders with high functional reserves and VO2max power, this is important.  This can set training zones at more reasonable targets so an athlete is able to spend more time in the target zones and reap greater training adaptations, especially since power from 20 minutes may fall off rapidly as it approaches the 60 minute mark, but again, this depends on the individual.

Postscript:
I spend a lot of time thinking whether we really should care about our MLSS values, VO2max, or such things to set training zones.  How long can we really spend at MLSS?  Can you really spend an hour at the FTP as set by a 20 minute time trial?  (In most cases no.)  What I've been seeing is that it doesn't matter that much, every rider comes with a different set of strengths and weaknesses at all power levels.

Although here's a study that offers a clue: these TdF riders (figures 2A and 3A) did way over 4.0mmol/L blood lactate in a 30 minute time trial, and several were well under.  Unfortunately this study was from 2006 and may very well be EPO tainted, hence I won't hold it in high regard.