The magic carbo-loading calculator


As of September 2017, new Sweat Science columns are being published at Check out my bestselling new book on the science of endurance, ENDURE: Mind, Body, and the Curiously Elastic Limits of Human Performance, published in February 2018 with a foreword by Malcolm Gladwell.

- Alex Hutchinson (@sweatscience)


Okay, I’m a bit late to the party on this one. A couple of weeks ago, the newswires were buzzing with the news of MIT/Harvard MD-PhD student Benjamin Rapoport’s new calculator that allows you to determine exactly how much carbohydrate you need to load up on before running a marathon. Putting aside my initial skepticism (surely how much carb you need to load up on is well known by now?!), I finally had a chance to check out both the calculator and the PLoS Computational Biology paper it’s based on (the full text is freely available at the link).

The quick summary: the calculator is just a toy, and should not guide your fuelling decisions; the paper, on the other hand, I found surprisingly interesting — though I still wouldn’t use it to plan my fuelling strategy.

Essentially, what Rapoport does is crunch the numbers on exactly how much carbohydrate is available within the body: circulating in the blood stream, stored in the liver, and loaded in the muscles. He brings in a bunch of other well-known information about things like the proportion of carbs versus fat burned at various exercise intensities, adds a few simplifying assumptions, and produces a model of exactly how much carbohydrate a typical runner can store under normal and “carbo-loaded” conditions, and how fast they can run a marathon without having those stores run out.

His summary and synthesis of the science is great — an excellent primer for the interested amateur on how marathon nutrition works. I might quibble with a few of his assumptions — he cites a couple of studies from the 1960s and 70s, for instance, to claim that “the total energetic cost of running depends only on the distance run and not on running speed,” whereas more recent studies have disputed this. But overall, it’s interesting stuff.

The calculator, though, is silly, at least insofar as it purports to tell you how many carbohydrates you need to run at a given marathon pace. For one thing, it’s dramatically simplified relative to the journal paper. The journal acknowledges that you need to know quantities like your VO2max and the total muscle mass in your legs to make a reasonable estimate. The calculator, on the other hand, only asks for your weight (from which it makes a guess about how big your legs are), your resting heart rate (from which it estimates your VO2max), and your age (from which it estimates your maximum heart rate in order to make a further guess about your running intensity at a given pace). All of these estimates have such huge ranges of uncertainty that a single one of them would make the calculator useless for practical purposes. With all of these estimates deployed simultaneously, it becomes a toy.

But I also object on a more fundamental level. Even if the calculator was a perfect reflection of the paper (i.e. even if you fed it all sorts of personal performance and morphological details), I still don’t believe it tells you anything useful. Rapoport sets up the problem as follows:

Analyses of endurance physiology have often either used coarse approximations to suggest that human glycogen reserves are insufficient to fuel a marathon (making ‘hitting the wall’ seem inevitable), or implied that maximal glycogen loading is required in order to complete a marathon without ‘hitting the wall.’

In other words, the problem is that people either take in too few or too many carbohydrates. But the parameters of carbo-loading are well-established — Rapoport himself uses them in his model. As Asker Jeukendrup’s recent book Sports Nutrition: From Lab to Kitchen says, trained athletes can pretty much max out glycogen stores by taking in an extra 10 g of carbs per kilogram of bodyweight for just one day before competition. So why would you take in significantly more (or less) than that? If you’re not doing that (or some related protocol designed to fill your glycogen stores as high as they’ll go), it’s because you got bad advice, not because you didn’t have a calculator telling you your precise needs.

The other parameter is pace: the calculator offers to tell what pace you can run without exhausting carbohydrate stores, based only on your weight, age and resting heart rate. Seriously? Show me any 2:30 marathoner, and I’ll find you a 4:30 marathoner with identical weight, age and resting heart rate.

As I said at the top, I enjoyed the paper — it’s a great synthesis of what’s going on inside your body during a marathon. It offers some interesting insights about the point at which carbohydrate stores become a theoretical limiting factor to ultimate performance. But as far as practical relevance, I don’t think it has any. Choose your pace based on what your training says you can do, and choose your carbo-loading strategy to start with a full tank — it’s as simple as that.