MLSS and training in swimmers

A quick thought for the morning, probably to be followed by something more involved later.  I've been trying to find a graph of MLSS (maximal lactate steady state) in different distance athletes, and in my search am running into articles that contain interesting tidbits.

Today's article looks at improving 500m swim times, which are around 2 minutes long, and in cycling this is a highly glycolytic anaerobic capacity interval.  So it's perhaps no surprise that this is figure 2:

Over the course of two months the athletes' times improved and the lactate concentration at MLSS did as well.  Despite this test being at anaerobic capacity, it also shows how much of an aerobic component there is at this exercise level.  Why does it show this?  Let's use Table 1 to guide us.

This looks like a traditional cycling plan to peak for early spring.  Let's note the two directions that MLSS moves in Figure 2.  During the lower intensity training the speed doesn't significantly improve, but the MLSS concentration goes up.  As training involves more time at threshold, the concentration doesn't go up again, but the athlete gets faster and gets to MLSS some seconds earlier.

To explain the December-January increase in MLSS, let's consider the effects of training slightly below and above MLSS, which in cycling we know as tempo and threshold, though due to the imprecision in swimming metrics, this might even be "sweet spot" and "low vo2max".  While these intensities stress the aerobic engine, they also burn a good deal of glucose and glycogen, engaging glycolytic muscles and increasing the rate at which they are capable of burning through their preferred fuel.  The increase in MLSS shows the improvement of the "glycolytic engine".

The January-February improvement in time and not another rise in MLSS could be explained by several things, but my guess is a larger increase of glycolytic enzymes and the building of type IIa/medium twitch fibers from a rest period in the training cycle.  I find myself wanting to see a 500m test performed in November, since low intensity training has different effects on MLSS than we see here.  More on that later, though.

This study encompassed both short and long distance swimmers.  No real distance specialties were mentioned except to group the two into different phenotypes, sprinters and endurance swimmers.

My first thought when I saw this graph: how cool is that?!  There's no explanation for it in the paper since its primary purpose was to investigate an inexpensive way to monitor training progress for swimmers.  I proffer two reasons behind why there should be different rates of improvement: lower intensities (November through January) lead to larger gains in endurance athletes who have greater type I fiber density, while higher intensities lead to larger gains in type II dense athletes.  My second and less likely explanation is that it takes the sprinters more time to see gains from their low intensity training, though you really can't be sure until it's tested.

The other thing I would like to see is Figure 2 separated by sprinters and endurance swimmers.  Hopefully I'll run across something like that soon.

Cadence and metabolism

A few days ago I was reading a physiology textbook and ran across this graph (from my phone, sorry about the quality):

This is the article referenced.

In case it's not obvious at first glance, these graphs say that the particular twitch (i.e. fast or slow) recruited for a muscle contraction depends both on its duration and force.  This makes sense from an economical standpoint: slow twitch fibers rely primarily on fat metabolism, and there's no need to get high force muscle fibers burning through glycogen for doing something as simple as typing on a keyboard.  (Don't know why that comes to mind.)

As in everything I read, I ask: how can I use this in cycling?  Maybe we can maintain more reliance on aerobic metabolism by pedaling at higher cadences.  Or maybe we can induce greater aerobic stress at high cadences.  This will be extremely useful in base training and general engine maintenance through the year.

Next I did a quick calculation, disregarding force evolution for a moment: if a muscle is engaged for 25% or 50% of a pedal stroke at 90rpm, how long does it contract for?  At 25%, 167ms, and at 50% it's 333ms.  At 120rpm we get 125 and 250ms.  It's looking okay right now, especially if the pedal force and total power output are kept low in addition to high cadence.  But the graph doesn't actually show us what motor units are engaged or even if this happens in leg muscles.  (First dorsal interosseus is in the hand.)  Let's dig deeper and see which muscles are engaged for how long in a pedal stroke.

Thanks for the fun colorful stuff, trainingpeaks.

I found a trainingpeaks article which led me to the above graphic, but I checked the references anyway.  This is one of the studies referenced.  Go to figure 6 for the meat and potatoes.  I suppose we can forgive the authors for working with triathletes, whose pedal strokes are very downward heavy due to their positions (different muscle recruitment is why you should always practice on your TT bike, kids).  Here is a second, most excellent article that reviews muscle engagement in relation to pedal and shoe conditions as well as seat height, and actually looks at two active and two former bike racers who can competently engage muscles entirely around the pedal stroke (look at subjects CDEF in figure 3, with cleats and normal saddle height).  Black regions of the graph represent muscle engagement.  In relation to fiber recruitment and aerobic stress, this shows that many muscles are engaged in half or more of the pedal stroke, indicating a good deal of fast twitch recruitment and glycolysis.

Because I make things more complicated than they need to be, I went through all of the above.  And only now do I look for articles studying metabolism in relation to pedal stroke.  This only does 50 vs 100 rpm.  Not a lot of bike racers pedal at 50rpm, so while informative in a proof-of-concept way, isn't exactly what we're looking for.  I dismissed several more articles on their ridiculously low cadences (down to 15rpm) until I came across this gem, which looks at 60, 80, and 100rpm in female recreational cyclists.  As predicted, HR and VO2 were higher at 80rpm than 60rpm, and higher still at 100rpm.

What we're learning is that despite how much it sucks for a rider like me who prefers high torque situations, high cadence pedaling during endurance exercise will strain the aerobic system more than at lower cadences.  This aerobic stress is, after all, why we base train.  Now we can make it more effective.  In terms of practical application, you should pedal at a higher cadence (if you have that option) if want to save your precious glycogen stores for later in a race.  It's something we all kind of knew anyway, but now we know why.  That's important to me, anyway.

As an aside, I ran across this article showing the neuromuscular adaptations cyclists have in pedal stroke efficiency.  It shows that our muscles don't work against each other as much as compared to untrained individuals.  With the same overall power output, we use less oxygen.  Someday I'll get a pair of Vector pedals and really get to go nuts with pedal stroke analysis.

Crits and Cross: The simplest analysis

Following are four standard power outputs from two recent cross races and two crits from earlier this year.  Three are the same from the last post, but I swapped out White Park for Sucker Brook because its power profile involved a lot more sprinting out of corners followed by some power sections.  I'm including these in a separate analysis to show that you can't always rely on this view to properly assess the demands of a particular discipline.

Quad CX

Two things stand out about this.  First, the peaks.  Rarely are they under 500w, and oftentimes they're between 600 to 1000 or more.  If you want a sprint-out-of-corners race, this is the cat's pajamas.  Second is the blank space underneath the lines.  Those black spots that stand out among all the yellow are times of constant power output.  The biggest, most obvious and repeated one is the long uphill fire road.  There are a couple more that correspond to other long stretches of constant power output, but for the most part the graph is either fully on or fully off.

Sucker Brook CX

The graph for Sucker Brook tells the same story as Quad.  I count 22 peaks alone above 825w, and countless above 600.  Similarly there are black spaces at the bottom of the graph for constant power output sections.  Again, very few.

Wells Crit

Now would you look at all that black space under the trace!  Much more constant power output, but still a very spiky graph.  Even the biggest peaks have some black space under them (I go for long sprints).  Also in contrast to the cross races, there are a few times where there are no peaks to really tax the sprinting muscle fibers.

Keith Berger Crit

Now that looks more like a cross race.  The six corner office park crit does require a lot of power spikes to keep up with the group out of corners and move up when gaps appear.  As for differences with cross, refer to the last post for a rundown on how the physiological pedaling strain compares to cross.  And if you really want to see more gaps under the power trace, just blow it up to full size.  They're shorter and more frequent than in a cross race.  The largest black space under the trace is my last lap attempt to get away, which looks like a very rough version of a "race winning interval."

This type of looking at the graph analysis is the kind of thing that leads to the "crits are cross on pavement" adages.  But after these few posts you know some of the finer details on what makes the disciplines similar and different, and can plan your training to reflect the differences.

Crits and Cross: the same but different

"Crits are just cross on pavement."  Or vice versa.  Are they really?

In a manner of speaking this is true.  Both involve races of short duration with high aerobic and anaerobic components.  The following analysis will look at two crits and two cross races, each with different qualities.  I'll include four graphs with power and pedaling metrics for each, which I hope will illustrate the differences in race demands and therefore differences in the training required (aside from the obvious bike handling differences), especially for a perfectly average roadie like myself.  Jeremy Powers has a 400w(?) ftp and can hit 1400w out of every corner in a cross race, but few of us mortals can do the same. All graphs come from WKO4, build 173.

The first cross race (Quad) was a decent combination of grass crit and riding in the woods, complete with an uphill barrier, U-turn in sand, and a couple long power sections to balance out the turny/carving sections.  A pretty well rounded course.  The second cross race (White Park) had only one long and one medium power sections, and otherwise required very short bursts between sharp turns, both on flat ground and off camber.

The first crit (Wells Ave) is a long oval with one corner barely sharp enough to require not pedaling.  It's not technical at all (and is in fact the closest thing to a race in a straight line as you can get), but this day was full of primes, making and chasing breaks, and a (rare) race winning sprint.  The second crit (Keith Berger) is a classic six corner office park deal with 90ish degree corners, some crosswinds, and a 200m last-corner-to-finish-line sprint.

Here are the first four time-in-zones charts.

Quad CX

White Park CX

Wells Crit

Keith Berger Crit

All four graphs show the effective aerobic demands of cross and crits are similar based on time in zones.  However, compare the yellow line (time in zones) in zones 2 and 3 between the top two graphs and bottom two graphs.  They're a bit lower in the cross races, where you're either pedaling as hard as you can or you're negotiating a turn.  Otherwise the yellow line traces are fairly similar.  The bar graph portion (time in normalized power zones) is interesting, though.  Quad and Wells bear more resemblance to each other than they do to the other races of the same discipline.  Perhaps it's the courses.

But now let's look at the last metric on these graphs: average and normalized power for the races.  Both metrics for crits are significantly higher than they are for cross.  You may be asking why, with lower average and normalized power, I'm spending the same time in the same zones between crits and cross.  WKO4 models your effective FTP, and with a lack of long steady efforts, right now mine is modeled at a meager 250w (in June/July it was in the low 320s, but now with little maintenance since August, it's probably back around 300).  What this really says is that I'm sprinting out of each corner and then recovering on straightaways at mid to high tempo wattages.  So like I discussed in my last post, I may need more aerobic intervals in cross season in order to bring up these power numbers.  It may be different for you.

Next, to confirm, let's look at the mean maximal power (MMP) graphs.

Quad CX

White Park CX

Wells Crit

Keith Berger Crit

On the left side of the graphs, you'll notice the first two (cx) fall off pretty quickly.  These are always start, and then it's settling in, trying not to blow up in the first lap.  Mean wattages are higher all around with crits due to the recovery time.  Same goes for normalized too, but I'm not including those graphs because they're a bit redundant and boring to inspect.

The right side of the cross graphs show some different bumps, thanks to the terrain of the courses.  That kind of thing is rare to see in MMP graphs of road races, but it's here at Keith Berger thanks to a well executed but ill timed last lap attack.

Now we'll see some graphs that illustrate the different muscular demands.

Quad CX

White Park CX

Wells Crit

Keith Berger Crit

Now things are coming into sharper relief.  The first and most obvious aberration to the pattern is how little of Wells was spent coasting, but like I said it's basically a straight line so very little coasting through corners.  The cadence distribution in both crits has a median over 95rpm (highly aerobic), while for the cross races the median is around 85, and heavily skewed left.

Also notice the time spent not pedaling.  In Keith Berger, the more technical crit, there's only a few less minutes not pedaling than in either cross race.  Time in those super low cadence bins does add up.  Longer contractions are highly anaerobic, and create large demands on fast twitch fibers, which in turn creates a premium on mobilizing glycogen for these things.  Don't mistake this for saying cross isn't an aerobic sport, though.  Your heart rate is pinned for 40-60 minutes because you're trying to supply as much oxygen as you can to your muscles.  This still puts cross firmly in the endurance category, but it does blur the line between aerobic and muscular endurance.

Again let's confirm what we just discussed with some quadrant analysis.

Quad CX

White Park CX

Wells Crit

Keith Berger Crit

I think that's about as stark as it can get.  The first two graphs (cx) show significant distribution at low CPV, high pedal force, while the second two graphs (crits) show most of the pedaling done in the high CPV, low pedal force quadrant.

This hints at what I discussed in my last post.  In cross your aerobic engine is pinned through the whole race, and fibers heavily relying on anaerobic metabolism need to refuel with blood sugar, but such high muscular demands create more need for glycogen mobilization than a crit does (though some certainly can do the same).  Of course there are no hard and fast rules to any of this, though I hope I've illustrated some of the physiological differences between cyclocross and criteriums.