Using the glycogen equation

I’ve referenced this a few times. Jonathan Savage crafted an equation, out of 1993 research from Dutch professor JA Romijn, that estimates the percentage of energy you use from glycogen, vs fat, based on your effort relative to your VO2max.

y = .0021x² + .7896x – 21.031

… where X is the percentage of your VO2max and y is the percentage of caloric energy that comes from glycogen.

So, for a simple example, if I do a run at an effort equal to 70% of my VO2max:

(.0021 * 70 * 70) + (.7896 * 70) – 21.031


10.290 + 55.272 – 21.031


44.531 % from glycogen

Note: Yes, this is a massively reductionist estimate, ignoring all sorts of factors, not to mention the VO2max percentage is itself a guesstimated assumption based on what you believe is your VO2max, with your given x-percentage probably based solely on your running pace and/or heart rate. But correlative observation and evidence indicates that it’s more accurate than other methods, so here we are.

Anyway… at 70% effort, if I burn 123 calories per mile, then 54.8 of those calories will come from glycogen, with the rest coming from fat.

If (based on other quantitative assumptions) we can say I have 1208 calories of glycogen stored in my lower body, I can run about 22.0 miles at this effort before my lower body runs out of glycogen.


Now, this number probably seems low to many experts, who are used to assuming that the glycogen rate is closer to 65-70%.

At the same time, most coaches and runners are used to pushing harder on regular runs (even recovery runs) than they should. Many run their marathons harder than their bodies probably should, and it figures they burn out of glycogen before mile 20. So it would figure that most would assume a 65ish% glycogen burn rate.

Even Daniels‘ metrics flattens the slide on his similar VDOT metric, where as you drop below an 80% effort it flattens the change in the percentage of max until it assumes a rough minimum of about 80%. I would posit you can (and many do) their comfortable easy runs at far less than 80% of VO2max.

More than anything the Romijn/Savage Equation, while counterintuitive, is probably closer to an accurate estimate of glycogen use than other estimates. It’s one I go by.

Where am I going with this? Basically, this equation has some uses.

  • Let’s say you are trying to eat low carb, or at the least carb cycle to improve your insulin sensitivity and in turn how well your body utilizes ingested carbs. If you’re doing serious endurance training, you probably do need carbs. If you know your training volume and intensity, you can figure out how many carbs you can eat on a given day to fuel or offset the glycogen your body will burn that day.
  • This can help you get an idea of what pre/post workout or race nutrition you need to fuel or recover from your workout. Your protein needs should remain fairly static, and of course you need not worry too much about recovering fat since we all carry so much. But your carb needs can vary dramatically.
  • Similarly, this can help you figure out an accurate fueling plan for a marathon or longer race. By knowing exactly how much glycogen you need, relative to your estimated glycogen body storage, you now have a hard figure of carbohydrate calories you know you need to ingest. You can plan accordingly.
  • This can also help address the larger question of how many carbs you need on average. Perhaps you can set a low baseline, like 100-150g a day, and then add carbs based on how many you may burn in a typical week. The equation can give you a fairly accurate number.

There are I’m sure other examples. I’ll leave it to you to find possible ways to utilize this data.

Tagged , , , , ,

2 thoughts on “Using the glycogen equation

  1. […] Race Eve Pasta Gorge is a favorite runner ritual, and while you may not substantially improve your glycogen reserves, you at least won’t go to bed […]

  2. […] the glycogen usage rate of the running I plan to do. The less intense the running, the more of those calories will come […]

Leave a Reply

%d bloggers like this: