The long run is a central piece of marathon training for everyone from recreational marathoners all the way up to national-level competitors. There are a lot of reasons why distance runners, and marathoners in particular, do long runs; these include increasing mitochondrial and capillary density, improving leg strength and resilience to the pounding of a two, three, or four-hour race, and, most pertinent to this post, to put in mileage in a low glycogen availability state.
If you aren't up on the latest training jargon, that last point just means that one purported benefit of long runs is to deplete the stored carbohydrates in your body so you are forced to burn more fat. When your body has run out of carbohydrates, you have "hit the wall," colloquially speaking—fat is much less efficient to burn than carbohydrates, so you end up slowing down significantly. When this occurs in a race, it's sometimes called "bonking." This is obviously undesirable during a marathon, and one way to avoid it should be to train your body to parse out its carbohydrate stores more efficiently, sparing more for the end of the race, hence the long run as a staple of marathon training. The long run takes on such importance in a typical marathon training plan that you'll often see runners doing as little as 30 or 40 miles a week putting in a 16-22-mile long run every week to prepare for the marathon. Moreover, this run is virtually always done at an easy pace, or with only small segments at marathon pace.
There are a number of problems with this situation, but the one I'd like to focus on today is the carbohydrate-burning aspect. In short, doing long easy runs is not going to deplete most runners' carbohydrate stores! Let's take a look at why.
Physiological models of running predict that the caloric cost of running per unit of distance covered (i.e. per mile) depends chiefly on how much you weigh. Running six-minute mile pace is obviously more calorie-intensive than running ten-minute mile pace, but this is counterbalanced by the fact that you cover ground a lot faster at six-minute pace. I suspect that this relationship breaks down at very high speeds, simply because most engines and machines (even biological ones) are less efficient when operating near their limit, but that's beside the point: most people run the marathon at moderate to slow speeds. The model I use predicts that a 140-pound runner needs about 105 Calories to run one mile, regardless of the pace.
The critical piece of information that's lost on most marathoners is that not all of these calories come from carbohydrates. The higher the energetic demand of whatever activity you're doing, the greater the proportion of your energy comes from carbohydrates. While out for a leisurely walk, your body is burning mostly fat. But if you are running near your VO2 max, like you would in a 5k race, nearly all (~95%) of your energy comes from carbohydrates.
As the graph below (taken from a 2010 mathematical model on marathon running),1 the fraction of carbohydrates that you burn is proportional to the square of your fractional oxygen utilization, or your % VO2.
So, our 140-pound runner still requires ~105 Calories per mile, but how much of this must be carbs will depend on his pace. At a reasonable marathon pace (80-85% of 5k pace), roughly 70% of his calories will be coming from carbohydrates. Assuming typical carbohydrate storage values in his leg muscles, he'd hit the wall in a marathon somewhere in the vicinity of 22 miles into the race if he ingested no additional calories.
But what if you're doing an easy long run? When most people run "easy" or "moderate," they're only going 55-70% of their VO2 max. At this pace, only about half of your calories come from carbohydrates, meaning you deplete your carb stores much slower. Even a very conservative estimate predicts that a 16:00 5k runner going seven-minute mile pace won't hit the wall until almost 25 miles of running! More reasonable estimates peg the number at more like 29 miles.
As an interesting aside, heavier runners do indeed require more energy to run a mile, but they aren't predicted to run out of carbohydrates any earlier during a marathon if you assume similar body composition. This is because greater muscle mass in your legs allows you to store more carbohydrates, so a 140-pound and 160-pound runner with a similar build and running at the same relative intensity should hit the wall at about the same point. Heavier runners can still struggle with fueling issues because they need to consume more calories to push back the wall, but that's a topic for another day.
So if you don't actually run out of carbs during a long easy run, why do you get tired on a 20-miler? First, there's the obvious reason of muscular fatigue—if you don't usually go that far, your legs will surely be hurting. Second, there appears to be an "anticipatory" effect with glycogen depletion. Your brain, sensing that your carbohydrate stores are dwindling, amplifies your sense of fatigue and slows you down before you crash outright. This is why drinking a sports drink appears to aid performance in races like a half marathon, which is far too short for real glycogen depletion to have an effect.2
If you do want to deplete your body's carbohydrate reserves in training, you need to do one of two things. You can either run in a partially-carbohydrate-depleted state, or you can run faster. The easiest way to run in a partially-depleted state is to get up and run right away in the morning, without consuming any carbohydrates. A lot of people do this anyways, since it's easy on the stomach!
The other option, going faster, is most definitely more challenging, but it has the added bonus of being more specific to your marathon preparations. The centerpiece of Kenyan marathon training, and an element conspicuously missing in many American marathon training plans, is the "long-fast run"—a long run of up to 25 miles done every few weeks at 87-95% of marathon pace (faster paces for shorter distances, obviously). Not only is this great metabolic training for the marathon, but it actively pushes the limits of the body's carbohydrate stores. Renato Canova, an Italian who coaches many of the elite Kenyan marathoners who are currently tearing up the roads, describes how the long fast run stresses the body's energy usage in a 2009 post on LetsRun.com:
for PUSHING the athlete to chose,among the fatty acids, the ones having higher energetic power, WE HAVE TO PUT IN CRISIS THE MUSCULAR FIBRES AND THE BODY.
We start running FAST, for example, 28 km at 3:10 pace, waiting that the tank of glycogen becomes empty. After 28, suddenly you run 3:30 and after 3:45. THE REAL TRAINING ARE ONLY THE LAST 2 KM, THE FIRST 28 HAD THE GOAL TO USE THE FUEL YOU HAD IN THE TANK.
When the body is in difficulty, our fibres pay more attention to the ability in chosing and selecting and using the fats with more energetic power. The next long run fast, our athlete is able to last, at the same speed, not 28, but 30 km, going in crisis after that mileage.
A long easy run can accomplish a lot. But one thing that it's not going to do is fully deplete the glycogen stores of most experienced runners. Keep this in mind when planning out your training for the marathon.
Appendix: Rapoport's mathematical model
For basic calculations of when a runner is expected to run out of glycogen, I use an excel spreadsheet I designed based on a semi-empirical mathematical model described in a 2010 paper by Benjamin Rapoport of Harvard Medical School. Raporport's model uses observations from the scientific literature on fractional carbohydrate use at varying fractions of VO2 max, glycogen storage densities, and anthropometric data to construct an equation describing the distance of glycogen depletion as a function of running intensity. The model certainly isn't perfect, and you could quibble about certain parameters, but it allows us to get reasonable estimates of if and when a given runner will hit the wall. If you require an abnormally high amount of calories per mile of running, or if your body stores an abnormally low amount of carbs, the model will be inaccurate. If you'd like to play around with the spreadsheet I made, you can check it out at RunnersConnect.net here.
1. Rapoport, B. I., Metabolic Factors Limiting Performance in Marathon Runners. PLoS Computational Biology 2010, 6 (10), e1000960.
2. Painelli, V. S.; Nicastro, H.; Lancha, A., Carbohydrate mouth rinse: does it improve endurance exercise performance. Nutrition Journal 2010, 9 (33).
Great article. Thank you. I really appreciate the quality of thought and analysis that you put into your posts. I've shared this to a couple of the running communities on G+.
John, nice article, I liked the statement on starting the run in a partially- depleted carbo state. Here is what happened to me 12 years ago.
I was over weight (214lbs) and got into a university diet study.
I was running 50 miles a week.
The study had me (and 30 others) eating 1700 calories a day (basic pryimid eating-19 servings of food groups)
I weighted 214lbs at start. at end I weighted 182, lost 32 lbs in 8 weeks, 4 lbs a week.
I ran a 50K ultra 3 weeks in bonking big time the last 10 miles.
3 weeks after that I ran my fastest marathon in 24 years, it was not until 24 miles did I slow down even a little, maintaining 9 min pace.
3 weeks after that I ran my best trail 50K ultra ever in 5h10min (avg 10 min miles) and I NEVER walked and ran ALL the hills. I never got tired or hungry.
My theory is I got very efficient at burning my excess fat as the 1700cal got burned up just with regular day to day work/life. I also have heartrate data that showed my HR got very low at faster running paces.
My longest training runs were 2 -14 milers, of course, running races of 2-31milers and 1- 26 miler was just as good or better than long training runs.
Love your site.
jeff