Why Some Athletes May Get Slower With High-Volume Training

Two athletes. Same age, gender, body dimensions. Same speed, VO₂ max. Same job, stress levels. Together, they start the same high-volume, high-stress training approach.

What happens next?

One athlete improves massively, winning big races all over the world and the adulation of arenas full of dozens of fans. One athlete regresses massively, suffering health crises before quitting the sport entirely.

What the heck happened?

No one can know for sure. When I started coaching, that uncertainty haunted my dreams. It seems like we can know 1,000 variables, and that’s still not enough to trace how outcome follows intervention. The human body can sometimes seem like a black box, holding secrets inside (unless you cut it open, which some studies find may reduce performance).

One athlete improves massively, winning big races all over the world and the adulation of arenas full of dozens of fans. One athlete regresses massively, suffering health crises before quitting the sport entirely.

A new study on muscle-fiber typology and high-volume training examines one of the most important black-box physiology variables in a unique and exciting way. The results are striking, confirming suspicions that many coaches have had for decades, possibly explaining some of the big unanswered questions about individual variability with high-volume training.

… Or it could be single-variable noise obscuring some more complex signal from the 1,000s of variables we aren’t measuring. Physiology and performance are fun. Let’s do this.

Study Overview

Published on August 20, 2020 in the Journal of Applied Physiology, the study aimed to identify markers of training stress related to overload training. The study population was eight women and 16 men, all highly trained middle-distance runners with a median age of 21. Each participant did three weeks of normal training (54 km/week average for women; 74 km/week for men), followed by three weeks of overload training, classified as increases in volume of 10% in week 1, 20% in week 2 and 30% in week 3.

The cherry on top of the protocol was a taper week, classified as a 55% reduction in volume from week 3 of the overload period. Athletes completed an incremental treadmill running test at baseline, after the normal training three-week block, after the overload training three-week block, and after the taper week. At the same time as the running tests, researchers measured resting metabolic rate, subjective fatigue response and blood biomarkers.

That study design alone would be fascinating. But that is just the start. Because the researchers went a step further and estimated muscle-fiber typology of the calf muscle “by quantification of muscle carnosine using proton magnetic resonance spectroscopy and expressed as a z-score relative to a non-athlete control group.”

In a 2011 study in the journal PLoS One, muscle carnosine was associated with muscle-fiber type—whether an athlete is more slow-twitch or more fast-twitch, variables with a heavy genetic predisposition (see more here). Higher z-scores meant more fast-twitch, lower meant more slow-twitch.

With that, the table was set. Let’s feast on some delicious results.

Of the 24-athlete study population, 12 athletes had decreased time-to-exhaustion in the running tests after overload training, and worse responses to the taper period too. They increased training load, and they got slower, even after a taper window for supercompensation (at least temporarily). The authors classified those athletes as “functionally overreached.” The other 12 athletes had no decrease in performance, and a better supercompensation response after the taper.

If you skimmed to this point, PAY ATTENTION NOW, because here is the fascinating part. The 12 athletes that overreached had significantly higher muscle carnosine z-scores. In other words, they were estimated to be faster-twitch overall than the athletes that did not regress.

If you skimmed to this point, PAY ATTENTION NOW, because here is the fascinating part. The 12 athletes that overreached had significantly higher muscle carnosine z-scores. In other words, they were estimated to be faster-twitch overall than the athletes that did not regress.

That might not seem like an Earth-shaking conclusion, but my Earth definitely shook like one of those water cups in Jurassic Park. We are conditioned to think more is better, particularly when it comes to training volume.

Could that characterization of the conventional wisdom be wrong?

It depends.

If you’re keeping track at home, I have used that phrase at least 10,000 times, which means you get a free law-school education if you haven’t lost your punch-card. Other studies and coaches have theorized this relationship with faster-twitch/slower-twitch athletes for many decades. It likely explains how Olympian Bernard Lagat trained lower volume with a rest day to excel up to half-marathons, while Olympian Cam Levins trained up to 180 miles per week with three runs a day for similar outcomes.

Or in trail running, it might point to why UTMB winner Pau Capell can do massive training weeks, but 4-minute-miler and Western States 4th-place finisher Matt Daniels excels off 10 to 12 hours.

Or in trail running, it might point to why UTMB winner Pau Capell can do massive training weeks, but 4-minute-miler and Western States 4th-place finisher Matt Daniels excels off 10 to 12 hours. Steve Magness’ amazing book Science Of Running delves into this area, as does the training section of the book my wife and I wrote. It’s not a secret that muscle-fiber physiology likely influences adaptation responses to training. However, it’s somewhat shocking to see it laid out so emphatically in an ingenious study design.

Now is a good time for a major shout-out to the study authors for asking an important question in a unique way. When contextualizing how the findings may influence training methodology, there are four important points to think about. Well, there are actually thousands of points possibly worth talking about, but I don’t want you to have to carb-load just to finish my articles.

First, functional overreaching is not necessarily a bad thing.

After a recovery window, it can lead to adaptation and higher performance. However, overreaching is also associated with worse health variables in other studies (including hormonal balance and inflammation), though that was not seen in this instance. Plus, it’s hard to isolate training overreaching from confounding variables like inadequate calorie intake and life stress, although the authors had great protocols in place.

Second, athlete muscles are usually a mix of Type II fast-twitch and Type I slow-twitch.

Muscle-fiber typology falls on spectrums that also include intermediate muscle fibers that demonstrate properties of both. There can also be variance within the same athlete based on the specific muscle studied. In general, long-distance running chooses for slower-twitch athletes, since those muscles are packed with mitochondria and are made for oxygen processing, leading to less fatigue with sustained exercise.

Meanwhile, the fatigability of Type II fast-twitch fibers may lead to longer-term muscle damage from the same stimulus. In this study, for example, 20 of the 24 athletes had negative z-scores, indicating they are slower twitch than the general population. And that’s for middle distance runners—I imagine elite marathoners or nordic skiers would be even slower twitch. Meanwhile, basketball or football chooses for faster-twitch athletes, allowing for explosive movement with the trade-off of more fatigability in aerobic exercise.

It’s possible for an athlete to optimize their physiology with long-term training. For example, I went to college to play football as a sprinting powerlifter, while I can now barely lift an extra-large calzone without a spotter. A 2018 study in the European Journal of Applied Physiology looked at a set of 52-year-old identical twins, one of which trained and one of which didn’t, and found a 55% greater expression of slow-twitch fibers in the trained twin. Long-term, higher-volume training is indispensable to cause those year-over-year adaptations.

But even with anecdotes like those, genetic predispositions matter. Training is about dialing in training for your genetic wiring, while also trying to incorporate the thousands of other variables that are relevant in training response.

Third, the study was conducted over seven weeks.

That time horizon is not adequate to draw conclusions on long-term training trajectory. That may be especially relevant to fast-twitch muscle fibers, which may take longer to recover and may respond to more extended tapers. In addition, training volume is just one of many variables influencing training outcomes, so it’s possible that if some of the dials were shifted on different schedules, the results could have been different.

Physiology can respond in non-linear ways, so isolating causation relationships with specific variables can feel a bit like walking barefoot through a dog park. You control what you can see, but you can’t see everything.

Physiology can respond in non-linear ways, so isolating causation relationships with specific variables can feel a bit like walking barefoot through a dog park. You control what you can see, but you can’t see everything.

Fourth, the entire athlete population reported worse weekly wellness scores during high-volume training.

Running is fun, I promise. However, only the overreached group reported reduced sleep quality and upper respiratory infection symptoms. And there were no observed differences in blood biomarkers or resting metabolic rate. Reading the findings together, it may seem to indicate that the responses in the overreached athletes were not related to systemic, cellular-level fatigue. Instead, it was more confined to muscle-fiber contractibility and other factors about how the musculoskeletal system transfers power and fatigues during exercise.

Disclaimers are like old running shoes—I have dozens of them lying around at any time. 

Middle-distance, college-aged elite athletes may respond to training differently than a broader population. Perhaps the more fast-twitch athletes just needed to take more down weeks during the training build, or do the big training over more years.

Disclaimers noted, the study confirms what I see in the training of elite athletes and beginners alike.

Maybe adaptation to high-volume over a running career prevents this type of overreaching. On top of all that, this is just one study and some of the findings are not mirrored elsewhere, particularly related to biomarkers and fatigue. Disclaimers noted, the study confirms what I see in the training of elite athletes and beginners alike.

I think the study has four takeaways for your training.

One: It may help to have a general feel for whether you are more slow-twitch, or more fast-twitch.

That is difficult to estimate, and a lab test would be optimal. Short of that, consider the whole of your athletic history together. I was a football player that could bench press a small horse and sprint like a happy pony, so that one is easy. Matt Daniels is a sub-4-minute miler with a fast 400 time, so he may be in the intermediate range. Other ultrarunners just get stronger as the distance goes longer, while struggling to hit high speeds. In general, it may help to look at your expected PRs at shorter distances relative to longer ones, accounting for training approach, though that is not validated. Another approach outlined in our book involves comparing stride-speed on a 30 seconds fast/30 seconds easy workout to tempo speed on a 20-to-30-minute sustained effort, but that’s purely based on our data and is likely subject to 100 confounding variables.

There are other methods you can search online, but they don’t involve adding $1 to our royalty check, and daddy needs a new pair of running shoes.

Two: Responses to increases in training stress are highly variable, and more is not better.

The fast-twitch/slow-twitch delineation is cool and relevant, and there are likely many other variables that work similarly. Olympic marathoners are likely selected partially for their slow-twitch genetics, thus it makes sense that training philosophy grounded in the approaches of the top pros calls for high volume and tons of workouts. Interpolating from Olympic outliers (or ultra champions) can cause athletes with different genetics to get slower. If an athlete is training longer and harder and feeling worse and worse, it’s key to adapt the training approach. You see this problem manifested all the time in college teams or other groups that all do the same workouts.

The same principles apply to workouts. In general, faster-twitch athletes seem to excel off more sustainable workout efforts, likely for similar physiological reasoning. Our athletes that are likely faster-twitch do more hill workouts, where muscle damage is usually lessened.

Three: The body doesn’t know miles, it knows stress.

Accumulation of stress varies based on muscle-fiber type, along with many other training and life variables. The point is not to maximize stress, since as the study showed, maximizing stress can lead to reduced performance in athletes that can’t absorb it. Instead, the point is to sustainably adapt to the stress you introduce, accounting for the complexity of life too. Clare Gallagher did around 50 miles a week before winning the Western States 100, Scotty Hawker under 50 miles a week before finishing 3rd at UTMB. But Jim Walmsley did 130+ miles per week before winning Western States, and Courtney Dauwalter does big training weeks to become an unstoppable superhuman full of speed and smiles.

Different things work for every athlete. Meanwhile our training logs don’t say anything about that. They say weekly miles, badass workouts, Strava kudos-inducing shreds that take all day. Remember that miles and paces don’t account for the complex interplay between your genetics, environment, and background.

Four: Vert could change the game.

Anecdotally, our internal data on our running team seems to indicate that athletes we estimate to be faster-twitch respond slightly worse to training volume increases, but slightly better to training vert increases. To summarize it excessively, faster-twitch athletes on our team require higher levels of vert before big events, whereas slower-twitch athlete responses are more variable.

Anecdotally, our internal data on our running team seems to indicate that athletes we estimate to be faster-twitch respond slightly worse to training volume increases, but slightly better to training vert increases. To summarize it excessively, faster-twitch athletes on our team require higher levels of vert before big events, whereas slower-twitch athlete responses are more variable. That could be all noise, rather than signal. Or it could be something that is getting at a new study question—what happens when faster-twitch muscles are subject to eccentric muscle contractions of steep downhills? Or perhaps it’s related to central fatigue? Those are questions for another day.

Our internal data also shows faster-twitch athletes may respond better to cross training. But I don’t have credible theories for the mechanism there.

Takeaway message

You may have heard me say something like this: Throw 100 eggs at a wall, and one might not break. That doesn’t mean we should copy the approach of that egg, throwing ourselves into a wall and hoping for the best.

When it comes to muscle-fiber typology, perhaps a sample size of the general population responding to overload training might mean 10 don’t break, or 25, or 50. But breaking or not breaking is not a choice each of those eggs is making. That’s why it’s key to pay attention to how you feel and how you respond, rather than comparing to other eggs.

Because we all have our breaking points. Where those points lie isn’t just based on how bad we want it or our past training. It’s not even based on muscle fiber typology or VO₂ max. No, it’s way more complex than that.

Training response is based on the interaction of millions of genetic variables with environment and long-term adaptation that all integrate together to mean that every athlete has a different optimal training approach (including in the same athlete across time).

In other words … it depends. I told you that physiology and performance were fun.

David Roche partners with runners of all abilities through his coaching service, Some Work, All Play. With Megan Roche, M.D., he hosts a weekly, 30-minute podcast on running (and other things), and they wrote a book called The Happy Runner.