"There's only one hard and fast rule in running: sometimes you have to run one hard and fast."

Monday, January 24, 2011

Ultramarathon fueling revisited (warning: seriously dull stuff)

Every time someone decides to take up ultras, the same questions arise about what to eat and when; some bad ideas seem to arise just as often. The current one runs: one should train with little sugar in one's system to get used to that and then race with a lot of sugar in one's tank. This looks like a long-term variant of marathon carbohydrate loading, but it fails on a simple principle. Ancel Key's Minnesota Starvation Experiment showed that one can't make a radical change like that; the body, having adapted to one type of nutrition, doesn't immediately accept the "bonus" carbs (the biochemical basis is that the enzymes involved have been down-regulated). So the bottom line is: you can actually run an ultra on a high protein, low carb diet if you're used to it, but you can't then switch to high carb for the race without digestive distress; of course, most people opt for a higher carb diet and take in the majority of their calories in ultras as carbs.

A variant of the idea is more common: do your long runs without eating and get used to that, then eat during the race. This also fails, but takes some explanation.

When I first decided to do an ultra, I knew I'd have to eat during it and I'd never eaten in a race, so I practiced eating while running in training - and discovered that my body rejected all solid food, but I could down huge amounts of calories in liquid form (over time, this has changed somewhat). The question arose as to how many calories I should be ingesting and this question plagues every new ultrarunner. Those doing 100 Milers seem to have settled on about 6500 calories during the race, usually expressed as "2 gels per hour" by middle of the pack runners; this is about 2/3 rds of the calories burned during the race, the remainder coming from stored fat. Would more calories be better? Is there a limit? Tim Noakes did a study that has created a sort of "law" that the liver can only process about 1 gram of carbohydrate per minute (200-300 cal./hour, which works out to "2 or 3 gels" per hour), but this is probably variable and trainable, as Yiannis Kouros reportedly ate more calories than he burned in a 6 day race. Whether everyone ends up doing the same thing because it works or simply because it's become tradition is hard to say.

Let's take a look at the two different methods of long run training:

If you run until your muscles run out of glycogen, the body can adapt in three different ways. First, it can store more glycogen (there is a possible long-term adaptation of stopping overcompensatory glycogen storage, i.e. one can no longer carboload). Second, it can become more efficient, burning less sugar at training pace. Third, one can become slower and thus burn less sugar and more fat over the same distance run. That third one is not what one wants!

Alternatively, if you supply glucose to muscles during exercise by eating, the body adapts by becoming more efficient at that given concentration of glucose (because the body doesn't tap into glycogen reserves, there's the possibility of decreased glycogen storage). If the amount of glucose supplied is less than required, depletion still eventually occurs. If excess glucose is supplied, the excess ends up being stored as fat, as glycolysis and glycogen storage are incompatible.

Here's where my own bad idea occurred. What if you ran to depletion, then supplied glucose at a steady state? Would you get the benefits of both methods? As it happens, this turns out to be the worst possible strategy, though the reasons are obscure. Essentially, one "hits the wall" and the only pace possible afterward is a survival shuffle. One ends up training oneself to move as slowly as possible.

In the end, the answer is that one needs to do depletion runs and one needs to do very long runs with a constant supply of fuel, but... not at the same time. A fast mid-week run, depleting the muscles of glycogen, will give the benefits of that method, without having to go very far. Then, the weekend long run can be done with supplied glucose and one can simply manage to run further with fuel than without and that is what one's goal is in the long run: to go as far as possible.


nwgdc said...

Well done. Now I'm more confused than I was before.
It's especially hard for me to take in calories while training right now, as it's effing cold out and everything liquid seems to freeze, and I'm not thirsty.
I agree, though, (well, I THINK I'm agreeing with you...I could be disagreeing with you--better read it again) that one needs to train both glycogen depletion AND constant calorie supply.
Unfortunately, I'm bad at both.

As for my HR, I had them tested in '07, and my numbers were:
Aerobic: 150-163
"Tempo": 163-170
Anaerobic: 170-180
"Maximum Capacity":183-189

I'm really not sure what to make of those, but I DO know that keeping my HR under 150 seems like a nice pace I could run for hours.

Jim ... 50after40 said...

You call it seriuosly dull, I say great info and seriously interesting. Thanks for the post!

Stacey (aka Ultraprincess) said...

I think that everyone must figure out what their body needs... lots of trail and error!

Gotta run.