Copyright © 2018 mysportscience. Mysportscience Ltd is a limited company (registered in England & Wales 9443573)

© 2023 by "This Just In". Proudly created with

Ketogenic diets for athletes

December 1, 2016

Earlier in the year, I attended the Symposium that celebrated the 10th anniversary of the IOC Diploma in Sports Nutrition. One of the speakers was Louise Burke from the Australian Institute of Sport. Her talk was to discuss pros and cons of the much talked about Low Carbohydrate High Fat (LCHF) diet. Last week I spoke with Dr Paoli from the University of Padova at a conference in Bologna. He presented some of his outstanding work on ketogenic diets. Both talks were a balanced view of the literature... so are these just fad diets? Or is it the next best thing for athletes? 


Let’s start with the current advice to athletes. This advice is to periodize nutrition, to stress carbohydrate in training and competition when this is required and to reduce carbohydrate intake for specific purposes. Especially in the popular press but even in some scientific publications this picture is frequently distorted by LCHF or ketogenic diet advocates. It is sometimes said that the advice is to eats lots of carbohydrates at all times.  This is not the advice! Another distortion that is often used is the statement that “scientists think that fat is not used at higher intensities of exercise”. That is also not true. Of course fat is used at higher intensities. Even at 85%VO2max fat is used, but at the same time carbohydrate is the unrefuted dominant fuel at those intensities. Patients who cannot use fat as a fuel have a good exercise capacity but cannot sustain this for long. Patients with McArdles disease who cannot use carbohydrate as a fuel have very poor exercise capacity but can go longer at low intensities.   


LCHF and ketogenic diet advocates often refer to one study from the 1980s as THE evidence that this works. Every reference to improved performance with ketogenic diets goes back to this study. Who actually takes the time to read the paper will discover that the paper isn’t actually supporting some of the claims that are often made. The paper shows that a ketogenic diet results in no changes in endurance capacity at a low intensity. There was huge individual variation in endurance capacity between individuals, which is to be expected with open ended exercise at that (low) intensity. In fact there is one individual who has an abnormally large variation in endurance capacity that may have skewed this data. Regardless, the data of this study do not support a performance benefit of a ketogenic diet and this was not the conclusion the authors drew. Also, the exercise is at such low intensity that the relevance for athletes should be questioned anyway.


There are a number of studies that used a LCHF or a ketogenic diet and show no differences in performance or a decrement in high intensity performance. An example was discussed in this previous blog: "Low carb diet v high carb diet and cycling performance". If studies have demonstrated a tendency towards performance improvement, this has always been at low exercise intensities, where the type of fuel used probably doesn’t make a lot of difference, intensities that can be sustained by both carbohydrate or fat. The majority of short term studies show that LCHF diets may be detrimental to performance and LCHF advocates are usually quick to point out that there may not have been enough time to adapt. Therefore we will look at the few studies that were 4 weeks or longer:


Below are some studies (of 4 weeks or longer) that measured actual performance or endurance capacity:


Phinney et al 1983 (4 weeks)

No difference in endurance capacity at low intensity (62-64%VO2max) in 5 subjects on a ketogenic diet versus a mixed diet.


Helge et al 1996 (7 weeks)

Smaller training adaptations after 7 weeks with high-fat versus high-carbohydrate diet. Diet was not ketogenic, but carbohydrate intake was very low in individuals who were training 3-4 times a week.


Zajac et al 2014 (4 weeks)

Ketogenic diet or mixed diet for 4 weeks in a cross-over design in off road cyclists. Reductions in peak power were observed after ketogenic diet.

This study is sometimes quoted for improvements in performance because they also observed increased VO2 and lactate threshold, but these are mere artefacts of increases in fat metabolism. Glycogen depletion results in increased LT, this is not the same as a training effect and it also increases oxygen uptake at the same workload (decreased economy).


Fleming et al 2003 (6 weeks)

Authors report small decrements in peak power output and endurance performance in the high fat diet group (not ketogenic) versus the high carbohydrate group.


If we have missed a study that is 4 weeks or longer, and has measured endurance capacity of performance please email us at and the study will be added.



There is currently no evidence in favour of a ketogenic diet for exercise performance. There is little evidence in general, but evidence is especially lacking in well-trained athletes, who train on a daily basis and compete at high intensity.  


It is sometimes argued that the absence of evidence is not evidence of absence and that is correct but at the same time without evidence we are only talking about a theory and not scientific facts. All we can conclude at this point is that several studies have been conducted and no study has yet provided any evidence and if anything studies seem to suggest a reduction in high intensity performance.


In my discussion with Dr Paoli in Bologna, he suggested that the ketogenic diet is perhaps not something for athletes during the competitive season but could be a good way to prepare for the season. Although, again, we don’t have evidence to support these ideas, this is possible and intriguing. It is also completely in line with current guidelines, where you plan your nutrition according to your goals, sometimes high and sometimes low carbohydrate. Not one diet that is best for everyone and all purposes. 


There is a need for more research and less cherry picking, more evidence and less “belief”. More actual measurements and less surrogate measurements. More discussions about well performed studies and facts and less about anecdotes, feelings and emotions.








Please reload

About me
  • Facebook App Icon
  • Twitter App Icon
  • LinkedIn App Icon
Please reload

Related posts
Recent posts
Please reload

Search By Tags