Fat vs. Carbohydrate Utilization At Different Power Zones

Does anyone know of articles or research that try to describe the relationship between fat vs. carb utilization at different power zones?

Currently, I mostly can find things using heart rate or % of VO2 max (and they are usually researched around FatMax). I’m however curious if there exists anything that gives a clear overview on the other zones (e.g. something that says most people utilize 90% of their energy/caloric needs coming from fat and 10% from carbs when cycling in Zone 1, to give a rough example).

This question has been studied extensively since the mid/late 1800s (no, not a typo). You will therefore be able to find graphs of the relationship between estimated fat and carbohydrate oxidation and exercise intensity in practically any undergraduate exercise physiology textbook and all over the web.

Generally speaking, however, in the fasted state (i.e., no recent food intake) we use mostly fat at rest, with a bit of a jump up to closer to a 50/50 mixture during low intensity exercise. From there the balance shifts non-linearly more and more toward carbohydrate, reaching 100% (apparently - see my previous posts about this particular issue) as the intensity approaches VO2max.

Although that is the general picture, exercise intensity is not the only determinant of the rates of fat and carbohydrate oxidation. As indicated above, nutritional state is important, with recent (i.e., previous 6-8 h…yes, we studied this time course back in my UT Austin days: Exercise metabolism at different time intervals after a meal - PubMed) carbohydrate intake increasing carbohydrate and decreasing fat utilization. A person’s habitual diet, fitness, age, sex, and probably genetics also all play a role.

What the above means is that it is very difficult to accurately estimate how much fat/carbohydrate a given individual oxidizes during exercise under a particular condition. As well, there’s not much “actionable intelligence” in such data for an athlete. Thus, although software exists that attempts to provide such estimates (and I was encouraged to develop such a chart for WKO4), this is really more of a sales/snow job than anything else.


I haven’t seen anything based off FTP rather than VO2max, no.

Consider though that percent fat utilization even at a given percent VO2max is highly variable between individuals at low intensities (say tempo downwards). We are talking about like +/- 20% in either direction when you’re in z2.

As a result, you can’t generalize population averages to yourself in any useful way.

1 Like

Such a study doesn’t need to be done to know what the result would be. Substrate oxidation is more closely related to muscular metabolic fitness than to cardiovascular fitness. If your maximal metabolic steady state (for which FTP is a surrogate) is higher or lower relative to VO2max than average, then the “bend” in the curvilinear relationship I verbally described above will be as well. This of course has been understood ever since the lactic threshold concept emerged and overtook VO2max as THE determinant of performance ca. 40-50 y ago.

1 Like

Would something like that, if accurately possible in the future, give a ‘big breakthrough’ in sports performance or are the current general guidelines on daily recommended carbohydrate intake more than good enough that it would only give marginal gains (or perhaps non at all or even negative)? Or is that what you meant with “not much actionable intelligence”?

I can give you an N=1 set of data from some lab testing I did last year. At ~60% FTP I’m burning about 90% fat. My break even point is ~75% FTP I.e. I’m burning 50/50 fat and carbs. By the time I’m at at threshold it’s nearly all carbs, fat only contributing ~10%.

Data taken from a ramp test. So doesn’t provide much insight into how these numbers are affected with time, fatigue, fueling, etc. I can make some educated guesses e.g. If I go up a hill at threshold and then settle back into Z2 I’m guessing my body doesn’t immediately revert to using 90% fat but carries on burning a high proportion of carbs for a while. But overall fits fairly well with my experience e.g. I can ride around at ~200W which is low Z2 for me for a long time without really having to worry much about fueling, but if I’m doing a lot of sweetspot work or higher then need to be taking on a lot of carbs.


One factor to keep in mind when attempting to interpret such data is that VCO2 kinetics are much slower than VO2 kinetics. Fat oxidation will therefore initially be overestimated and carbohydrate oxidation will be underestimated after each increase in intensity, until RER has sufficient time to “catch up”. If I were designing an incremental exercise test where this was the primary focus, I would use 8-10 min stages to overcome this effect.


WKO4 (and now 5) largely exists because I developed a novel model of the relationship between training and performance (for which I have retained the intellectual property rights). In keeping with my long-standing saying “it’s your glycogen budget, spend it wisely”, it would be useful to more accurately estimate carbohydrate utilization during training sessions as an input to this model. However, whether such an approach would make a large difference or none at all is impossible to say.

1 Like

I’m not sure what your opinion is, if any, of INSCYD testing, but I went through one of their metabolic tests ~4 years ago and part of the output was the chart below.

Not sure if it’s accurate, or representative, but it was actionable in that it set my FatMax range. I believe there is some controversy or misunderstanding(s) on this board as to what that term means … but the coach I worked with at the time used the middle of my FatMax range (I believe it was 185W) as the number I should be around when I went on long (4-6 hour) endurance rides to maximize benefit while minimizing recovery needs.



I also received this table. Again, unsure of its utility other than how to apply FatMax to training. I suppose it could directionally help you understand/visualize how much you should be fueling during a race at an anticipated intensity.

1 Like
  1. Incyd’s conceptualization of physiology is incorrect.

  2. There is no evidence that basing training off of the maximal rate of fat oxidation is beneficial, nor is there really any reason to think it might be.

Sorry, but you asked!


No need to apologize. I’m looking for accurate information, not what I want to hear.

1 Like

Yup, ramp test is probably the wrong term, from memory the increments were about every 5 minutes. So maybe still a bit of lag but interesting to me. Primary purpose of the testing was for a research study on different fueling including ketones (I was one of the lab rats!). Haven’t got the data for subsequent sessions unfortunately as the study hasn’t been published yet.