The hardest part of running training theory is that every athlete is their own N=1 study. We have individual inputs–training volume, intensity, etc. We have individual outputs–time to exhaustion, VO2 max, race results. But the line connecting input to output will always be a best guess, because we have way too many confounding variables (age, stress, training background, muscle fiber typology) and no way to look under the hood to see exactly what’s happening (humans prefer not to be dissected).
Even for a single athlete using different training interventions over time, there is not enough data to reliably infer what causes what. That training plan could have caused a world championship. Or it could have been the base built in that lower-intensity approach a year ago, mixed with better fueling, all combined with an optimal day in the menstrual cycle.
So how do we know what works?
Well, we do have a baseline understanding of physiology. While humans vary substantially, we’re all a similar assortment of semi-organized chemical interactions that love the show Ted Lasso. But that still raises the measurement problem–how can we actually know what training is best with this deluge of uncontrollable variables? It’s enough to make any coach pull their hair out, which may itself be an advantageous adaptation. Receding hairlines are built for aerodynamic speed, I tell myself.
While we may never be able to know the optimal approach for each individual, we can hypothesize what works across the population. N=1 sucks. But add up thousands of Ns, and we’re getting somewhere. Bad for scrabble, good for training theory. Nn-nn-nn-nnn, hey hey heyyy, good science.
And when you add up all those Ns in running, looking at training logs of top performers, you’ll notice something curious. Doubles. Lots and lots of doubles.
RELATED: 8-Minute Speed Legs
Doubles in Practice
A 2019 study in the Journal of Strength and Conditioning Research looked at 85 elite athletes over their first seven years of serious training to draw conclusions about the type of runs associated with top performance. Volume of easy runs had the highest correlation with performance, from 0.72 at 3 years to 0.68 at seven years. These findings were backed up by a 2020 study in the European Journal of Sports Science. Neither of those studies discuss doubles specifically, but I’d bet the dog and the car (as implied by the first thing, a Subaru) on some of that volume being accumulated with multiple runs in a day.
For example, a 2019 study on the Ingebrigtsen brothers indicated that they accumulated 150 to 160 kilometers a week on…get this…13-14 separate sessions. They sometimes do multiple harder workouts in a single day. And brother Jakob won the 1500 meter gold medal in Tokyo.
The same goes for athletes coached by Renato Canova, famous for his “block” training days of two hard workouts, plus countless easy doubles. Before medaling in the Olympic Marathon, Molly Seidel did 6 doubles most weeks. I bet there are more running Olympians who do triples (used in some East African training camps) than train solely with singles.
Last week’s article on cross-country ski training cited many studies (an excessive number of studies) about the high proportion of extremely easy training at the top end of that sport. Some of the greatest champions do 90%+ of their training at low heart rates! As outlined by a 2010 study in the Scandinavian Journal of Medicine & Science in Sports, skiers (and runners) accumulate aerobic volume to increase in mitochondrial content and capillaries around muscle fibers, improve metabolic function at both high and low intensities, and encourage a more economical use of oxygen to power performance at all intensities as well. All of that corresponds to faster race performances, even in events that are just a few minutes long.
But if it was just about volume, why aren’t we seeing top runners rely more on massive singles? That’s what bikers do for the most part: several tiny espressos, one big ride. Meanwhile, swimmers do tons of doubles, as do skiers. That’s interesting! So what connects running and swimming and skiing, but not biking? My guess is that it’s related to the biomechanical demand of the sport varying significantly at slow paces, plus injury risk. Biking is the same if you do it with low power or high enough power to mine Bitcoin in an environmentally responsible way (quads will save the rainforests?). For running, meanwhile, overload those biomechanical patterns and athletes may get inefficient, even if they are able to avoid injury.
So we’re trying to add up to a big aerobic number (adjusted for individual background), while balancing overall stress for long-term adaptation. But overemphasizing long singles may have diminishing returns due to the demands of running.
Doubles help solve the math equation.
But is it just about the extra volume? I doubt it.
Doubles may provide a hormonal stimulus that enhances adaptation, they may increase adaptation markers and protein expression associated with better performance, they may improve glycogen replenishment, they may even have some fascinating relationship to epigenetic signaling. A 2012 study on mice found that 3 x 10 minute runs in a single day led to the same adaptations as a 30 minute run, but possibly with slightly larger increases in expression of one protein (TSP-1). What would that type of study look like in humans? Unfortunately, finding out would require dissection. And that’s the type of lopsided trade that would only appeal to the New York Mets.
Whatever the exact reasoning, thousands of world-class athletes have come to the conclusion that doubles are key on the track and roads. However, they’re sometimes less common in trails and ultras. I have two theories for the cause of that offset.
First, trail running relies heavily on resilience to fatigue from variable musculoskeletal loading patterns. The track is an aerobic system contest, the trails involve the same aerobic pathways with a wrinkle–it doesn’t matter how strong your aerobic system is if your legs are Jello. Musculoskeletal damage (and thus, potential adaptation after recovery) accrues over longer runs, so perhaps the long singles create resilient monsters (assuming an athlete doesn’t break first).
Two, the margins in trail running are not as narrow as on the roads and track. Our sport is messy, full of rocks and roots and airplane arms, so the champions don’t need to find every single possible advantage. A 0.1% improvement in aerobic power might be swallowed up by a 10% improvement from not eating sh*t on a descent. (That also gives me hope that doping is less common in trail running than in a sport like cycling).
How To Add Doubles
While doubles work for pro athletes, I have seen in coaching that they can work for almost anyone, subject to a few disclaimers. First, the body knows stress, not miles. A double can be counterproductive if it adds even an ounce too much to the stress scales. Only double if you have the time, energy, and life force to spare.
Second, increasing volume increases injury risk. It’s hard to run a PR with an achilles that sounds like a creaky doorway.
Third, adaptation is a high-stakes game. Overloading stress in moderation followed by recovery can lead to breakthroughs. But overloading a bit too much can lead to stagnation and regression. Some world-class athletes are likely chosen partially because they are genetic anomalies with adaptation under high chronic stress loads. So make sure you’re always listening to your body.
Given those risks, my co-coach Megan and I introduce doubles with five guidelines. This will be the next topic for “Sexy Science Corner” on our podcast, so listen when that comes out for more info.
One: Keep it very easy–up to 2x your 5k pace
If the goal is the aerobic stimulus while balancing stress, almost no pace is too slow. I personally do my doubles without a GPS watch, partially because it may auto-pause due to how slow I go. If your normal easy pace is 8 minutes per mile, you can make it 9 or 10 minutes per mile, especially to start. Molly Seidel often does them at 8+ minute pace, with a marathon pace in the mid-5s. I have seen some pro runners in Boulder doing them at what looks like 10 minute pace. Glorious prancy ponies! Just focus on good form–light on your feet with quick strides.
Two: Keep it short–20 to 30 minutes is plenty
There is some evidence that the productive hormonal stimulus of running rises most rapidly in the first half-hour, before leveling off (and sometimes reversing, though it’s debated and individual-dependent). Many athletes describe feeling refreshed after a quick afternoon shake-out. For trail runners, we really love doing some of these sessions on the “treadhill” to reduce impact and get climbing-specific biomechanical loading.
Three: At least a few hours after your first activity
Glycogen replenishment is a key element in doubles, so some fun food and a few hours is plenty. Canova blocks sometimes involve tinkering with glycogen levels for elite male athletes, but we have seen that backfire.
Four: Add them on workout days first, aerobic days next
Easy doubles on workout days may maximize adaptation benefits, plus there is a greater endurance stimulus. Once an athlete is adapted to the approach, we’ll occasionally have them run more moderately on some workout-day doubles or treadhills (sometimes even with structured workouts like Canova blocks, but that’s a training element you should only add at the direction of a coach due to the high risk of injury and overtraining). Don’t double on long run days, which may overwhelm glycogen replenishment and increase breakdown rates. Any double could be replaced by easy cross training as well, which should accumulate aerobic adaptations at similar rates, with lower injury risk.
Five: All doubles are optional
There is no such thing as a mandatory double for our athletes. The main training session is what matters most, and tons of our team members have won some of the biggest races in the world without any doubles at all. So listen to your body. Are you dreading it? Skip. Do you have any niggles, even the smallest whisper from a gnat’s ass? Chill. Affect sleep, family, work? Bag it. Does it slow down your recovery for the next day? It’s OK to give it a couple weeks for adaptation, but if that persists, nix doubles until you wake up the next day feeling as strong or stronger than you would otherwise.
Many athletes we coach will see this general weekly structure on their peak build weeks, when life stress is low, and health is perfect. Peak ultra builds usually involve fewer (if any) doubles, in order to maximize the musculoskeletal adaptation stimulus, as described above.
Monday: rest and recovery
Tuesday: easy run and hill strides (6-12 miles)
Wednesday: workout (8-13 miles) and optional double/treadhill (2-4 miles)
Thursday: easy run (6-12 miles) and optional double/treadhill (2-4 miles)
Friday: easy run and optional hill strides (3-8 miles) or x-train/rest
Saturday: Long run (10-25 miles)
Sunday: easy run and hill strides (6-15 miles) and optional double/treadhill (2-4 miles)
The big thing to remember: doubles are 100% not necessary. But then again, running isn’t really necessary either, in the big scheme of the universe. Doubles, like run training in the first place, is all about exploring the limits of your potential by doing something that seems moderately unreasonable.
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.