In part one, I explained the problem I've had in falling apart in marathons. In part two, I discussed how part of the problem might be "not having enough fuel in the tank." The next thing to consider is how to run faster, using less glycogen... if possible.
Let's say that I fall apart at 17 miles at marathon pace. One way to deal with that is to run that in training and then do another mile or two at whatever pace I can manage. The next time out, I might be able to run a little further at that pace. The problem with this is that it is so close to racing for me that it takes weeks to recover and I can't do just one hard run every few weeks and hope to improve.
Two other ideas have become common. One is to run long and then add miles at marathon pace just at the end, to get a feel for being able to run that pace when tired. Another is to alternate miles at half-marathon pace with miles at marathon pace; the faster miles should produce fatigue and then marathon pace would feel like a recovery pace. Neither of these work well for me.
One way to increase the per centage of fat being burned is to do long runs first thing in the morning, without fueling. Not having any liver glycogen (which happens early in such runs) has been shown to force the body to adapt to running more on fat. I've always done this - it's a comfort thing for me - but it obviously hasn't been enough.
A completely different approach
There's been a number of runners having success running on extremely low carb diets and this can circumvent the whole running-out-of-glycogen problem. If you don't eat any carbs, for about three days your body converts proteins into glucose (and ketone bodies), to supply fuel to cells that absolutely require glucose, such as red blood cells. After that, to spare muscle losses, the body starts to convert fats into ketone bodies and the brain slowly switches from using solely glucose to using about 65% ketone bodies after six weeks. If you're training during this, there's no glucose to form glycogen in the muscles, so they have to run on fat.
It's easy enough to come up with a 1600-1700 calorie diet that's healthy and matches these requirements, but it gets progressively harder with the more one ingests. If you're eating 3000 calories, you end up having to eat bizarre things that I don't recognize as food (or you end up with way too much of some minerals, which can have serious consequences over time).
I'd assumed that, because there were measurable adaptations to a low carb diet still happening at six weeks, that this diet had to be maintained indefinitely. However, a paper from Tim Noakes' lab (Goedeke, Christie, et. al. [rats, I lost the reference]) showed that runners increased the per centage of fat burned after only 5 days of a very low carb diet. Muscle adaptations happen at a different rate than others; I hadn't thought of that. Also, the improvement appears to be "robust" - it doesn't just go away.
So it should be possible to divide a standard healthy balanced diet into low carb and high carb foods and eat twice as much of the low carb foods for a week, then spend a week eating twice as much of the high carb foods. This might be enough to get enough of the effect to last for a marathon or 50K (though maybe not longer races).
But... it should be possible not to alter one's diet at all and get the same effect, as well. If one simply depletes muscles of glycogen, then run hard enough to continue depleting them for five days, it should be the same as not eating carbs; either way, the muscles can't store glycogen over that time. This is an approach called "crash training," which I covered on this blog before. After five hard days, one would need extensive recovery, which would allow the muscles to restock glycogen, just like in carbohydrate-loading.
One can't take that much time off, though, without losing some fitness, so it would be necessary to supplement training with low-intensity running or with cross-training.
This is an interesting possibility, one I may have to try.
Lost in Space
4 days ago