The Science Behind How Elite Runners Train
A wonderful new review article analyzed the literature and online resources to summarize the training that makes athletes into the best in the world. It's got everything: easy running, workout structure, season design, tapers, and so much more. Let’s dive into one of the most incredible articles ever.
A review article published on April 1 is SO GOOD that I couldn’t wait to tell you all about it, immediately outlining a summary. While I was reading the article with the care and attention that a normal person would give to handling an original copy of the Constitution, I thought about that publication date. April 1st… oh no. I am the fool in April Fool’s Day, getting totally surprised by a few posts each year. Was this study another example? Were the authors going to Rick Roll me in the conclusion?
You can breathe, because I think we’re safe. This study is so wonderful that it’s figuratively unbelievable, but not literally unbelievable. Are you ready for a deep dive into one of the coolest systematic summaries of training practices? Woohoo!
Published in Sports Medicine – Open by authors Thomas Haugen, Øyvind Sandbakk, Stephen Seiler, and Espen Tønnessen, the review article aimed to integrate scientific literature on training (studies) with results-proven practice (plans and logs) to present a training framework for elite long-distance performance. Imagine training theory as a complex song with dozens of inputs. A study might isolate the drums, a training schedule might show the vocals and a guitar, and even a coach might just hear the instruments they’re looking for. The review article brings it all together, saying “here’s a kickass song!” Yes, there are some discordant notes where things might not all line up perfectly, but that’s the nature of integrating studies with dozens of N=1 training schedules. In my eyes, the authors are badass and ambitious ROCK GODS.
I’m going to break down the article with the same headings as the authors, picking out the elements that I find the most interesting and important. Remember: I am one of those coaches who may hear what I want to hear, so read the whole article to draw your own conclusions.
RELATED: The Surprising Connection Between Training Intensity and Long-Term Growth
A skeptical reader might hear all of this complex analysis and think “Why? Don’t we have all of these studies that you talk about every damn week?” Yes, there are lots of studies, but contrasted with a field like medicine, the studies rarely drive the process of developing theory. As the authors say, “Sport scientists have historically found themselves testing hypotheses regarding why elite athletes train as they do rather than driving innovation around the how in the training process.”
The problem is that it’s nearly impossible to design studies with interventions over months and years, which is countered by the fact that it takes “8-to-10 years prior to reaching a high international standard.” Thus a study intervention of 6 weeks might be catching noise, or worse–thinking it finds signal but misattributing the mechanism in a way that undercuts long-term growth. It would be like measuring a stock price in a “pump-and-dump” scheme, concluding that the pump is what athletes should seek out while missing the dump that follows. As all morning runners know, you never want to miss the dump.
The leap forward in this article is that they stepped outside the confines of studies (including those that look at training characteristics over long time scales) and also examined publicly-available training data from other sources. Anyone who is interested in training theory might see an online training log and think, “How does this fit in? Is this what everyone is doing?” The authors sought to answer that question. And they did it without a single joke about dumps. HEROES.
RELATED: The Exciting, Rapid Adaptations From Speedwork in Studies
The authors started with a literature review. That alone would be helpful, compiling the training studies like has been done with cross-country skiing. But it’s the next step that blew my freaking mind. They searched for “non-scientific, publicly available, and English-language training information related to podium contestants from international championships.” It is a systematic look at the websites that many coaches have bookmarked, featuring training details of the best athletes in the world.
In total, that methodology led to training logs/information from 59 world-leading athletes and 16 coaches of podium contestants for 5k/10k on the track and road marathons, with long-term training information given more weight than shorter-term snapshots. The ambition of the study also opens it up to some potential issues–the underlying data is not peer-reviewed and it’s subject to numerous biases, everything is reported differently, and doping could make some of the information less like finding out how to optimize human physiology and more like finding out how to optimize centaur physiology. Most importantly for our purposes, the authors are confining their search to performance outliers on the track and in the marathon, and it’s key to be careful when interpolating from outliers to determine appropriate training for other athletes.
RELATED: The Science and Art of Pacing Easy Runs
Training Periodization and Competition Scheduling
Now let’s get into the sexy results. Throughout, I am going to try to provide quick takeaways for runners that may not be aiming for an international podium placing, but you can take or leave those editorial comments since they are coming from a biased coaching perspective.
For track runners, there are generally three phases: preparation, competition, and transition. As has been standard practice since Arthur Lydiard coached a bunch of fast New Zealanders in the 1960s, the preparation period amounts to base building, with volume increasing in the general preparation phase at the start. As the preparation period gets more specific, there is a higher volume of race-paced intensity. One place where practice differs from some of the literature is in the prevalence of relatively high volume throughout the season. That underscores the importance of building a base and continually reinforcing it throughout the year, since aerobic adaptations take a long time and are ongoing over years.
The competition phase is a continuation of the specific preparation phase, with more races and recovery. The transition phase might be called an “off-season” in other contexts, and it can vary from 1 to 2 weeks of rest/low-intensity running to 4 weeks totally off. Off-seasons are not set-in-stone rules, they can be flexible based on the needs and background of the athlete, and can involve aerobic running and other activities.
A couple more interesting wrinkles for track runners: none of the athletes analyzed did any “over-distance” races (i.e. half marathons) in the 3-4 months leading up to 5k/10k championships, and their last race prior to championship competition was ~4 weeks before. Racing well beyond target distances may come with some drawbacks, particularly for very fast athletes.
Marathon runners approached it differently. They raced less, with just 6 competitions per year, with an average of 2 marathons per year separated by at least 3 months. That may be a lesson to trail and ultra runners to avoid an extremely high density of long races for peak performance.
While the general preparation phase/base building looks similar, marathon runners diverge from track runners with slower race-paced work in the specific preparation phase, and sometimes having their highest volumes of all directly preceding the taper. Athletes doing longer races may want to consider having their volume peak in the 2-4 weeks before their key races.
RELATED: Easy/Moderate Fartlek Workouts For Improving Endurance
As you can probably guess, an accumulation of low-intensity training was a key element in most training. Coaches had different terminology for this type of running–some call it “easy,” others “regeneration,” or “recovery.” But all likely rely on the same principles, increasing mitochondrial biogenesis and capillary density, leading to aerobic development over longer time horizons.
At least 80% of your training should likely be easy unless you are at very low volumes. The authors say that low intensity running should generally be 3-5 km/hour slower than marathon pace, usually lasting 40-70 minutes, with warm-ups and cool-downs lasting 10-30 minutes. Long runs are often faster, around 1-2 km/hour slower than marathon pace, with durations up to 2 hours for track runners and 1:15 to 2:45 for marathoners.
My co-coach Megan and I like the framework of 80% in Zone 1 in a 3-zone model, where the top end of zone 1 is the first ventilatory threshold when lactate production starts to increase–think easy up to steady runs in some cases. The authors frame long runs and progression runs within Zone 1, so it’s a wide range, rather than solely including slower running.
Higher-intensity training combines with easy training “to elicit a complex suite of overlapping and complementary adaptations.” The distribution of intense sessions depends on the athlete and their goals, but the authors include a fantastic principle from Coach Bill Bowerman and Olympic gold medalist Frank Shorter that they say applies to most training: “2–3 weekly interval sessions, a weekly long run, and fill the rest with as much [easy running] as you can handle.”
For many runners, 2-3 speed sessions a week may be overdoing it, particularly with uphills on trails often leading to higher heart rate. But the 2-3 session framework is a good way to think about runs when you let your heart rate get higher and the effort go up for a substantial portion of the run.
In the article, sessions included fartleks (30-60 minutes of running broken up with periods of intensity), threshold runs (20-50 minutes around half marathon effort), threshold intervals (3-15 minute intervals at half marathon effort totaling 30-75 minutes of intense effort), VO2 max intervals (2-4 minute duration at 3-10k effort with 2-3 minute recovery, totaling 15-20 minutes intervals), and hill repeats (30 seconds to 4 minutes duration at 5-10% grade). Athletes also did plenty of fast strides. Table 2 is worth a click–it’s like the ingredients list at the start of a kickass recipe for fitness.
RELATED: An Overview of the Norwegian Approach to Running Training
World-leading marathon runners train 500-700 hours per year, with track runners between 450 and 600 hours, spread out over 11-14 sessions (doubles are nearly ubiquitous). Interestingly, this is less time than sports like cross-country skiing, triathlon, and swimming, where athletes are often around 800-1200 hours per year. Running is weight-bearing, so the importance of injury prevention is key, rather than volume accumulation for its own sake.
To zoom out for a second, it’s helpful to think about one of the shortcomings of looking at the training of the best in the world, rather than a broader cohort of athletes. With 59 world-leading athletes included, I think it’s a conservative estimate that 59,000 athletes trained as hard as they could with the same big goals. We aren’t seeing those other 58,941 training logs, which is likely explained by genetics and random chance mixed together to lead to earlier stagnation or injury. If you throw 59,000 eggs at a wall and 59 don’t break, it doesn’t mean we should emulate every aspect of what those 59 did. That’s especially important when it comes to very high training volumes.
In addition, many of the athletes accumulate volume on dirt roads and forest paths, which stands in contrast to some studies that say that running on softer surfaces does not decrease loading rate. Whenever I see Kenyan training camps on the rutted-out dirt roads of Iten, I salivate at the idea of one of those teams showing up at the start line of Western States.
Athletes increase volume steadily in the preparation phase. They start by increasing frequency (including through doubles), before increasing the volume of runs. That’s a terrific lesson for all runners, aiming for a higher density of runs prior to emphasizing longer and longer runs, with even the weekly long run potentially having diminishing returns beyond 2-3 hours.
Typical training volumes add up to 160-220 km/week for marathon runners and 130-190 km/week for track runners. Those athletes at the lower volumes are usually doing a slightly higher proportion of intensity. My theory is that this equalization of stress points toward an inflection point where even genetic outliers who rarely get injured and adapt readily start to be overstressed and lose potential adaptations. The authors indicate that athletes who have tried even more strenuous programs with higher volume or intensity have often struggled with injury management and fatigue.
RELATED: The Performance Impacts of Periodization and Training Intensity Distribution
Intensity Zones and Intensity Distribution
The authors casually drop some of my favorite knowledge bombs when it comes to intensity quantification. After discussing the lack of consensus on whether to use heart rate, blood lactate, pace, or other variables, they say: “We would argue that this lack of consensus is consistent with an uncomfortable truth; no single intensity parameter performs satisfactorily in isolation as an intensity guide due to (1) intensity–duration interactions and uncoupling of internal and external workload, (2) individual and day-to-day variation, and (3) strain responses that can carry over from preceding workouts and transiently disrupt these relationships.” In other words, a well-developed and calibrated sense of feel may be what matters most, but it’s hard to quantify calibrated feel in study settings. Their “intensity scale” in Table 3 is fantastic–check it out, print it out, put it above your bed like it’s a childhood poster.
So we know that 80% or so of training is low intensity. What is the periodized distribution of medium (threshold) and high (Critical velocity/VO2+) intensity? There is no set standard, with pyramidal design (more Zone 2 threshold, less Zone 3 VO2) in preparation periods, and a move toward more polarized design for some track runners pre-competition. The specific answers are unfolding at the finish lines of major races around the world. Important note: 1% of training volume is spent doing sprints/strides, where athletes do 60-120 meter faster bursts. Never, ever forget your strides.
RELATED: The Case for Sustainable Training Levels to Support Long Term Growth
One of the most fascinating offsets between the literature and results-driven practice is in tapering philosophy. The research generally supports 2-3 week tapers up to 40-60% of volume. Meanwhile, “most long-distance runners do not report a substantial decrease in training volume until the last 7–10 days prior to competition.” The last intense session is usually 3-5 days before the goal event, and it’s not overly stressful.
For all athletes, tapering matters, but don’t have your volume and intensity drop off a cliff unless you have proof that works for you. I like athletes to have a progressive 10-14 day taper, with an intense session 10 days out, shorter aerobic long run 7 days out, and light speed session 4 days out (sometimes keeping that session aerobic before ultra events that are 8+ hours long).
RELATED: Understanding Long Run Training For Ultras
There are so many more amazing nuggets of wisdom in the review article! Read it, read it again, read it a third time until your dreams are about intensity distributions. It’s a wonderful template for combining literature with non-peer-reviewed information, and I can’t wait to see where the field goes from here.
Run easy, do your strides, include some intensity but don’t overdo it. Repeat for 8-10 years. And heck, repeat it for a few more sets of 8-10 years after that. Incredible, FAST adventures await!
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 they wrote a book called The Happy Runner.