Today, let’s have fun with training theory! One of the coolest things about coaching athletes long-term is seeing how progress unfolds with all different types of input variables. This article is based on a provocative idea: long races and ultras can actually make athletes faster over multiple years.
To most of you, that statement probably seems innocent enough. But there are probably some readers who did incredulous spit takes, covering their computer in research-backed hydration drink. “Slower events make slower athletes,” they might say. “It’s a trade-off, and there is a long history of examples.”
That’s generally true, particularly when looking at the history of road and track runners coming to ultras. Optimizing performance at longer races often rewards the type of training that corresponds with reduced top-end output. But what if an athlete focuses heavily on constant speed reinforcement, particularly in the recovery period after long races?
Based on what my co-coach Megan and I have seen, if speed is reincorporated as soon as possible after the recovery window, athletes can actually use the supercompensation opportunity to get faster than ever (we talk about it on our podcast here). Let’s get down and dirty with some speculative, salacious training theory.
The Speed Paradox
In some places online, like the message boards that make up the Internet’s putrid bowels, ultra runners in particular are derided for not being fast. A comment may say that athletes only move up in distance when they can no longer compete at shorter events. I think that’s a crock of BS.
Many studies show that there is correlation between running economy at short distances and long distances, and between uphills, flats, and downhills. Training approach changes how these physiological traits express themselves, but at 5 miles and 50 miles, there is usually a somewhat similar physiological origin point in top athletes pushing the margins of performance. The best endurance runners should be the best endurance runners.
But there’s a problem–let’s call it the speed paradox. The short-distance athlete can excel at long distances with some specific training, but the long-distance athlete may not excel in the same way at shorter distances with specific training. Athletes may start longer races with a speed background, and they have immediate success. But over the course of a few years, they get a bit slower. Yeah, they might excel in longer races, but the same athletes that started out with the range to rock a 10k or a 100k get better and better at longer distances while losing their edge in the short stuff. Meanwhile, it’s next-athlete-up, with a new generation of fast athletes coming into the sport and rocking it, before following a similar trajectory to longer distances. It can seem like a conga line to longer distances and slower paces.
The question Megan and I had when we started out coaching: Why? If there is correlation in speed at all distances, and similar aerobic processes are being rewarded, is it inevitable that athletes get slower as they move to longer distances?
Our initial theory: Athletes don’t get slower because of the races, but because speed development takes year-round reinforcement.
It all gets back to the physiological systems involved. Aerobically, long events are massive supercompensation stimuli, which could be a good thing for all distances. Musculoskeletally, there is some breakdown, but nothing that won’t be cleared from bloodwork within a week or two. So why do athletes seem to come back slower, and slower, and slower (at shorter distances) as they stack up these stimuli over time?
You don’t see this same problem in cycling. The athletes that put out the most power on a single 30-minute climb will generally be the same athletes that excel in 6-hour mountainous races, which makes sense given the correlation in aerobic speeds. Meanwhile, in running, the 5-mile runners and the 50-mile runners are usually totally different. That provides a helpful clue–what differentiates runners and cyclists, and may explain why runners get slower when racing longer, but cyclists don’t?
We think it has to do with the neuromuscular and biomechanical systems. Given the unique mechanical demands of running, putting out lots of power is a much greater strain on musculotendinous units and the nervous system than a sport like cycling. You can probably feel that in your hamstrings by imagining sprinting fast after a week off, whereas doing a cycling sprint after a week off would not end in any ruptured tendons, and the offset with normal peak power wouldn’t be so great.
Unlike the aerobic system, the neuromuscular and biomechanical systems adapt rapidly, but they also de-adapt rapidly. That becomes a huge issue when they aren’t reinforced throughout the year, which comes to a head as athletes do specific preparation for longer races, then recover after them.
In the build for long events, as training gets more specific in the final month or two, substantial amounts of top end speed training largely get removed in most training approaches. That’s probably necessary to optimize fatigue resistance and muscle fiber recruitment. The biggest problem we see comes after the event.
Most athletes use the post-race window to recover, build back with easy activity, and work back into higher volume. They have a major breakdown stimulus, and they build back off a slower base. The wild part of the speed paradox is that this process may even be rewarded. Particularly in 100 milers, there may be some benefit to approaches that neglect intermediate muscle fibers and top-end speed, sacrificing running economy at VO2 max for endurance and running economy at and below aerobic threshold.
Therein lies the toughest part of deciding what comes next in training. Training specifically for the attributes that make an athlete great at long races may lead to fantastic performances for a while. Specificity wins.
But there’s a catch–because top-end speed is a relevant variable in ultra speed, once the athletes get far enough away from their previous speed training, they may get slower at everything (particularly with age, when the inertia is already pushing toward reduced peak outputs). Specificity wins until the thing you want to be specifically good at is an event that is slow relative to your maximal physiological capacities.
When we first started coaching, we saw something fascinating. If instead of starting slow after recovery, athletes almost immediately began reinforcing speed development, they could actually get faster within a month or two in some instances. That feeds back into performance at all distances, and athletes don’t need to be confined to short-distance or long-distance: they’re just all-around badass athletes. Their general fitness attributes improve, and with specific training, they can be ready for almost anything.
The supercompensation stimulus of the race causes breakdown that approximates overreaching, and the science of overreaches indicates that there may be a chance to bounce back stronger. We think the problem is that in the window of post-race opportunity, athletes neglect the nervous and biomechanical systems. So while their bodies WANT to get faster on a cellular level (where the aerobic system operates), on the systems level connecting muscles and joints to put out power (where the neuromuscular and biomechanical systems operate), their ability erodes away. You’re left with a strong, aerobic monster, and each of those efficient mitochondria translate to worse power output. Their internal metrics may even be stronger, as their output declines.
Our way around this paradox was to try introducing fast running very rapidly after big endurance events. The total stress remains low to allow the endocrine system to rebound, but we want to ride the wave of supercompensation to higher outputs before it crests and goes back out to sea.
In practice, the general rule we apply is 1 day fully resting for every 10 miles raced, with wiggle room depending on the athlete and the endocrine demands of the training block. So a 50 miler is around 5 days off, and once athletes go farther than that, it’s usually about a week. They return to easy running, and within a few days (usually a week at most), they are already doing strides. It starts on hills, before moving to flats after a few days, with lower-volume VO2 max workouts within a week or two. A typical schedule after a peak 50 miler for a pro athlete maxing out at 80-90 miles per week may look like this:
- Day 1: race
- Day 2: PIZZA
- Day 3: BURGERS
- Day 4: TACOS
- Day 5: BEEF STEW
- Day 6: rest or easy jog
- Day 7: 5 miles easy
- Day 8: 6 miles easy
- Day 9: 7 miles easy plus 6 x 20 second hills fast
- Day 10: rest
- Day 11: 8 miles easy plus 6 x 30 second hills
- Day 12: 3 miles easy, 6 x 2 minute hills mod/hard around 5k effort, 5 x 20 seconds fast/2 min easy, 2-3 miles easy
- Day 13: 8 miles easy
- Day 14: 5 miles easy with 5 x 20 seconds fast/2 min easy
- Day 15: 13 miles easy
- Day 16: 10 miles easy with 6 x 30 seconds fast/2 min easy
Afterward, they return to more normal training volume, with a heavy emphasis on speed for 1 to 4 more weeks, even if their later goals are long ultras.
There is a ton of variance based on the person, but there is a universal general principle: before rebuilding the aerobic base, make sure that the neuromuscular and biomechanical systems are tuned up to go fast (and not “fast” in the sense of tempo-run speed, but “fast” in the sense of near-max output).
This is just one theory, mixing objective data from our team with subjective opinions about what works. Tons of different things work for different athletes, and you could probably find dozens of world-class athletes who excel over many years without giving a single crap about any of this.
But for most athletes, I think that it’s key to constantly reinforce speed development almost year-round, particularly when aging past 30, when the inertia may be pointing toward reduced top-end speed. When competing in longer, break-down heavy races, there is a bonus opportunity, and a bonus curse.
Opportunity: The race could be a major supercompensation stimulus.
Curse: The specific training and the race itself are slower, and could lead to top-end regression due to neglecting the neuromuscular and biomechanical demands of higher-output running.
A theory about how to seize the opportunity: Try to get that race supercompensation into a speed context as quickly as possible.
Or not! The wickedly fun part of training theory for ultras is that the science is still uncertain, so theories will be tested in the real world over the coming years. I can’t wait to update you on what we learn along the way.
David Roche partners with runners of all abilities through his coaching service, Some Work, All Play. With Megan Roche, M.D., he hosts the Some Work, All Play podcast on running (and other things), and you can find more of their work (AND PLAY) on their Patreon page starting at $5 a month.