Ultramarathons. What , why and how are the questions I'm most commonly asked. Hopefully you find some answers here from my own personal perspective. My other blog at www.mile27.com.au/blog is full of information on running and health and fitness in general.

Sunday, 2 January 2011

Training Low and Racing High – Part 1

The idea of training when the body is low in glycogen and then racing when glycogen stores are full is gaining popularity with endurance athletes particularly ultrarunners. The idea makes sense, training with low levels of glycogen teaches your body to burn fat more efficiently therefore preserve your very limited glycogen stores. Even a very lean athlete of 5% body fat and weighing 60kg will still have 3kg of fat which equates to 27000 calories compared with approx 2000 calories of glycogen.

Whilst this makes sense it doesn’t necessarily follow that this will improve our race times. The human body is extraordinarily complex and some questions need to be answered before we can determine if this concept actually improves performance in ultra distance racing

• Is running out of glycogen a limiting factor in ultra races or do most runners manage to consume enough glycogen during the race?
• Are there any benefits in being able to consume less glycogen during a race?
• Does training with low glycogen levels or in a fasted state have any positive or negative effects on performance?
• If it does improve our fat burning abilities and preserve our glycogen stores dose this allow us to run faster?
• Does an increased usage of fat mean that during competition less carbohydrate is necessary and if so could it decrease the chances of stomach problems during competition?
• Does consuming carbohydrates during a race negate the fat burning benefits gained in training when not consuming carbohydrates?
• Does training in a fasted or low glycogen state have any negative affects on the body?
• Are there other factors affecting performance more important than fat usage?
• Would training in a low glycogen or fasted state negatively affect these other factors?

There has been number of studies done trying to shed some light on the answers to these questions but before we get too in depth there are a few concepts to understand.

Low Glycogen vs Fasted State

Some studies test athletes in a fasted state , normally an overnight 12 hour fast whereas other studies use exercise to deplete glycogen stores and then perform an additional training session in a low glycogen state to see the effect.
Both tend to restrict the intake of carbohydrates during exercise, however there is a major difference between the two. Being in a fasted state doesn’t mean you have low glycogen levels, your muscle glycogen levels can be very high in a fasted state because if you haven’t exercised since your last meal then there is no need for the body to use any of its stored muscle glycogen. In a fasted state your blood glucose levels and liver glycogen levels will be low as this is what the brain uses for energy and even if you are asleep the brain still needs energy. A low glycogen state means your muscles don’t have much glycogen left in them.

Anecdotal Evidence

Supporters of the train low race high strategy point out that this technique is commonly used by Kenyan runners who apparently perform a lot of training in a fasted state. Remember being in a fasted state doesn’t mean your glycogen levels are low. In fact the Kenyans typically have a very high carbohydrate diet which would mean their glycogen levels are likely to be relatively full by the end of a day. First thing next morning when they train they will have low blood glucose levels and low liver glycogen levels not low muscle glycogen levels.

In fact since one of the adaptions to endurance training is the ability to store more glycogen in an athletes muscle it means highly trained athletes will have the ability to run 20 miles or so in a fasted state at a good pace quite comfortably. In fact the great Gebrselassie ran his first marathon in 2.05 without consuming any sports drink. (For his world record he did consume sports drink).Coincidentally it is thought that 2 hours is about the maximum an athlete can last without additional carbohydrates when running at that intensity.

So just because the Kenyans may train in a fasted state doesn’t mean you should. If you can run 20 miles or so at a comfortable pace in around 2 hours then training in a fasted state wont slow you down at all, otherwise you will struggle.

The question I can hear you asking is that in an ultra the pace is slower so therefore you should be able to run for longer in a fasted state. This is correct but then you have to ask is this of any benefit? Does training in a fasted state improve your ability to run an ultra. There is little doubt that your body will adapt and you will be able to run at a faster pace in a fasted state than when you first begin training like that. Once again you have to ask is this of any benefit for an ultra?

Research Limitations

When people like myself are trying to argue a point we often quote research that appears to support our viewpoint. Unfortunately in this case there is no specific research we can use. The longest exercise period in any of the studies I could find was 2 hours. Clearly running for 24 hours is very different to riding a bike or running for two.

The most commonly used technique used in research to put an athlete in a low glycogen state is to have them undergo two training sessions per day with no carbohydrate ingested either between or during the sessions. The second session is therefore completed in a low glycogen state.

In reality this doesn’t occur. Most athletes who train twice a day eat between the two sessions. Endurance athletes do perform low glycogen training when they do back to back long runs, eg 20 miles Friday evening and 20 miles Saturday morning. Researchers have not used this method in any of their studies though.

Most of the research looks at performing high intensity interval training in a low glycogen state whereas most ultrarunners perform a long run when in a low glycogen state.

Anybody who is using this research to say that the training low racing high will improve performance in ultrarunning is making a giant leap of faith.

Research Conclusions

For those of you who don’t like trawling through research papers here is a summary of what I have found. Those of you who would like to see the research part 2 of this blog will summarise a number of articles.

The main findings are as follows
• Training in a fasted or low glycogen state improves fat oxidation in exercise of up to 2 hours
• No research article has shown any performance benefits in running or cycling in exercise tests of up to 2 hours long
• Performing interval sessions in a low glycogen state reduces the intensity of the workout but appears not to reduce the benefit of the workout compared with using carbohydrates

Some conclusions made by authors include

• “The “train low, compete high” paradigm of glycogen levels has been challenged by a new study that shows no benefit of training in a low-glycogen state. Although some genes involved in training adaptations were enhanced after training in the low glycogen state, this does not mean athletes should adopt the practice. It could have deleterious effects on the ability to train hard and recovery from training, could be a possible risk factor for overtraining and could impact the immune system.” (1)

• “Training with low muscle glycogen reduced training intensity and, in performance, was no more effective than training with high muscle glycogen. However, fat oxidation was increased after training with low muscle glycogen, which may have been due to the enhanced metabolic adaptations in skeletal muscle.” (2)

• “training under conditions of reduced carbohydrate availability from both endogenous and exogenous sources provides an enhanced stimulus for inducing oxidative enzyme adaptations of skeletal muscle, although this does not translate to improved performance during high-intensity exercise." (3)

• “Although there was a decrease in exercise-induced glycogen breakdown and an increase in proteins involved in fat handling after fasting training, fat oxidation during exercise with carbohydrate intake was not changed.” (4)

• "The main findings of the present study were that: training in an overnight-fasted state enhances storage of muscle glycogen compared to training in the fed state;.. and peak VO2 and peak power improved more when training in the fasted state compared to the fed state."(5)

So where does that leave ultrarunners?

To put it simply we don’t know!

It appears that training in a low glucogen state improves the ability of the body to oxidise fat but this has no effect on high intensity exercise of up to 2 hours (although at least one study claims that once exercise is performed with carbohydrate there is no difference in fat oxidation(4) and another suggested that training in a fasted state improved peak VO2max and peak power (5)

Whether or not it has any effect on exercise lasting 10 or more hours is unproven and open to debate until any scientific study looks at this specifically

There appear to be few if any studies on the benefits of a low glycogen or fasted state training on exercising for longer than 2 hours. Extrapolating the results from a one or two hour time trial to a 24 hour Ultramarathon is not possible.
Some of you may be thinking that if it appears that there is very little difference between training in a low glycogen or fasted state compared with using carbohydrates then why not train that way.

Some researchers felt that immune function may be compromised training in a low glycogen or fasted state which could to leading to higher risk of illness and injury.
The question could be asked that instead of looking at how to conserve our carbohydrates by burning more fat why don’t we look and developing a bigger storage of carbohydrates. It is well documented that we can exercise at a higher intensity if we can burn carbohydrates and as the goal is to finish quicker not go for longer does this make more sense?

The body adapts to the specific training stimulus so whilst training in a low glycogen state does improve fat metabolism, training using carbohydates will increase the amount of glycogen that can be stored in the muscles.

Not enough Glycogen or too much?

One theory I have and I stress theory as there is no research to back me up (primarily because there has been so few studies done specific to ultrarunners) is that ultrarunners do suffer from low glycogen towards the end of a race but the answer to the problem is to consume less carbohydrates when racing not more.

Let me explain, I think that many ultrarunners consume too much food during a race. We simply do not have enough glycogen in our muscles to run an Ultramarathon, (you may be able to walk it) and if training can preserve our muscle glycogen and increase our ability to burn fat then that has to be a good thing. However given that we need to consume some carbohydrates in a race otherwise our performance will be impeded we need to determine the optimal amount.

If you are better at burning fat then obviously you will need less which has to be a good thing. Remember that high intensity interval training also improves the ability of the body to use fat for energy as does training in a fed state using carbohydrates but the research is showing that training low glycogen will promote more energy to be burnt from fat.

If we ingest too much carbohydrate during a race our stomachs cant digest it and deliver the glucose to our blood so in effect we are depleting our bodies glycogen levels. When this happens we start to feel weak so we ingest more carbohydrates which makes the problem worse until we are forced to walk. The reduced intensity gives the stomach a chance to digest all the carbs and deliver some much needed glucose to the muscles and off we go again.

Our bodies can only absorb approx 40-60 grams of carbohydrates per hour when we exercise, any more than that will slow down gastric emptying. This equates to 1.5 gels per hour. Food such as porridge, pasta, rice pudding, sandwiches etc that are often eaten during ultras can contain far more than 40-60 grams.

Limiting factors in an Ultra

Why do we slow done towards the end of an ultra? Without doubt running low in muscle glycogen is one factor but I believe a more important factor is muscular fatigue. Running stresses our muscles and causes small micro-tears in the fibres. The longer we run for the more damage occurs until eventually our legs are so sore that we slow down and eventually are forced to walk.

Looking at the ability of our body to burn fat and discussing train low race high strategies ignores this completely and no research I found even mentioned this ( primarily because very few articles mention ultramarathon runners).
Is it possible that any potential benefits gained in training low could be offset by negative affects on muscular resistance to damage it suffers when we run?

Quality of training in a low glycogen or fasted vs fed state

Research showed that when performing high intensity interval training in a fasted or low glycogen state intensity was less than training in a fed state however this did not reduce the training effect.

How that relates to ultras is open to interpretation. It appears that if you do perform high intensity interval training (and you should!) in a low glycogen or fasted state even though it will feel harder , you may not completed as many intervals and they will be slower it will still have the same training benefit in terms of a 1-2 hour time trial performance. Does this mean it will have the same benefit for an ultra?

I would argue that the ability to perform less repetitions of intervals would be detrimental to performance in an ultra in terms of the muscular overload to our legs. Even if the aerobic benefit is the same I question if the muscles themselves will have the same load put on them with less intervals. Increased tolerance to muscular damage is a big part of ultrarunning.

No studies looked at longer training sessions. It would have been very interesting to look at performing one run per week of 3 hours plus in a fasted , low glycogen and fed state with only the fed state consuming additional carbohydrates and seeing any chance in a 50-100k time trial.Unfortunately that hasn’t bed done yet so we need to decide for ourselves which approach to take.

Many athletes get fasted and low glycogen states mixed up , as mentioned earlier they are vastly different. I believe both have there merits but more benefit is gained when consuming additional carbohydrates when running. Here are my reasons why

Fasted State Training

Personally I believe that performing high quality long runs in a high glycogen state with additional carbohydrates has more benefits than in a low glycogen state ( whether you have breakfast or not doesn’t matter as your stores will be topped up from yesterdays eating).

It allows you to run faster for longer. For example you may be able to run for 3 hours at 9 minute mile pace in a low glycogen state. If you performed the training in a low glycogen state but consumed carbohydrates during your run you may be able to run for 4 hours at 9 minute miles and if your muscles weren’t low in glycogen and you used carbohydrates during your run you may manage 4 hours at 8.5 minute miles.
I believe the extra distance covered and faster speed puts more load on your legs making them more resistant to fatigue in an ultra.

Low Glycogen training

Many ultra runners perform back to back long runs and I am a big believer in this. Doing a long run Friday afternoon followed by another long run Saturday morning ensures that the Saturday run is done in a low glycogen state as it is hard to eat enough in one meal to top the stores up after a long run.

I prefer to use additional carbohydrates when doing the second run since once again it allows me to run further and faster than if I did it without carbohydrates. I believe the extra speed and miles puts better condition into my legs making them better able to handle the damage that running long distances does.

Whilst I would undoubtably improve my ability to run without additional carbohydrates if I trained that way I cant see a reason to do so. Why go through that process if there is no benefit other than the ability to run for longer without additional carbohydrates? It doesnt mean I’ll be a faster ultrarunner.


Training in a low glycogen state or fasted state does improve the bodys ability to metabolise fat in exercise lasting up to two hours (it may improve it for exercise lasting up to 24 hours but no research supports that either way) but whether that helps in performance is open to debate.

Personal Summary

I believe that consuming additional carbohydrates when we train allows us to run faster and for longer which although may reduce my fat burning potential it will improve the ability of my legs to resist the damage that occurs in an Ultramarathon.
As long as I am able to consume (and digest) sufficient carbohydrates on a regular basis when racing then I cant see any benefit in having to make do with less.


(1) Summary from IAAF 2nd International Consensus Conference on “Nutrition for Athletics“ held in Monaco from April 18-20, 2007.)

(2) Training with low muscle glycogen enhances fat metabolism in well-trained cyclists.
Hulston CJ, Venables MC, Mann CH, Martin C, Philp A, Baar K, Jeukendrup AE.
Med Sci Sports Exerc. 2010 Nov;42(11):2046-55.)

(3) Promoting Endurance Training Adaptations with Nutritional Interventions: The Potential Benefits of “Low Carbohydrate” Training. Drust B, Morton JP. Kinesiology 2009; 41: 19–24.

(4) Effect of training in the fasted state on metabolic responses during exercise with carbohydrate intake. De Bock K, Derave W, Eijnde BO, Hesselink MK, Koninckx E, Rose AJ, Schrauwen P, Bonen A, Richter EA, Hespel P J Appl Physiol. 2009 Jun;106(6):1757-8.

(5). Adaptations to skeletal muscle with endurance exercise training in the acutely fed versus overnight-fasted state. Stannard SR, Buckley AJ, Edge JA, Thompson MW.
J Sci Med Sport. 2010 Jul;13(4):465-9. Epub 2010 May 7.


John Kynaston said...

Thanks Andy.

I've read through once and will do so a few more times to really understand what you've written.

I wanted to thank you for taking the time to read all the research, summarise it and share it with us.

All the best for 2011.

Chris said...

Andy, interesting stuff and an interesting analysis. I am not an ultrarunner nor am I any kind of expert, just someone with an interest in nutrition.

Have you ever looked at ketogenic diets? In an ultra it strikes me that you need to be good at burning fat - you are going to use up your glycogen - and that is what a keto diet trains you to do. In many ways a ketogenic diet mimics fasting. I just wonder why you would mess up the adaptation to fat fuelling by eatimg carbs for the run

Anyway there is an interesting article by Phinney you might want to read through. A previous study is here

Chris said...

Jsut spotted the Phinney link is no working. Try this:


or this link

GaryB said...

Hi Andy

I havce just had a read through and found it very interesting. I remember many years ago reading a study paper done on the effect on athletes training regularly in a low glycogen state. This Study mainly concentrated on the bodies ability to convert muscle protein to carbohydrate when carbohydrates were no longer available. I think if i remember right it had said that this normally started to happen 3 to 5 hrs into the exercise period if no carbohydrate was available within the body or had not been consumed through food or drink. I cant remember the details but basically they were concentrating on the long term damage to the body if this type of training was of a regular nature( what ever that ment)I seem to remember there was a reference to a study paper published on heart muscle damage caused by muscle protein conversion in a ultra cyclist but i never read it all. The study didnt give any recommendations or conclusions as it was open to further study.

I thought that might be of interest to you but it was so long ago it might be outdated by now?

Andy Cole said...

Hi Andy, Really interesting stuff, but you lose me occasionally in the science and nomenclature. Looking at it from my decidedly non-expert point of view the dilemma seems to be:

When running an ultra you use maybe 500 to 1000 calories an hour (around 200-250 per mile). You can absorb 60g of carbs so I guess that's about 250 cals an hour you can usefully put back in. When your initial carb stores have gone (say 4-5 hours or so?) you are left with a deficit which has to be made up by burning fat.

You need to get efficient at burning this fat but the question is do you do this by training with no carbs (sort of "stress-training"), or with the 250 cals an hour of carbs going in(which mimics racing conditions).

You are saying that there is no proof either way but your instinct is that your fat burning efficiency will improve by the the same amount whichever path you take, so you might as well take the carbs route and get a faster training session.

Do you think this is likely to be as true for the "average" runner (who might say expect to complete a TMB in 40+ hours)as for someone like yourself who will be nearer to 24?

SteveQ said...

This is the second time I've stumbled here (I'll have to bookmark it). I've been trying to go through the same material, starting from a biochemist's point of view. I think that one of the keys is the number of days one is glycogen-deprived; for example, one's brain switches from using glucose to ketone bodies, with the most rapid change occuring on three day's fast, but at the expense of burning protein.

There's just too many variables!

There are ultrarunners who live in a glycogen-deprived state and do fairly well (Monica Scholz); they are able to run long frequently, but not at the best possible speed (IMHO). There are runners who live on a very high carb diet (Yiannis Kouros) and are never glycogen-deprived. The question becomes: are either of these methods better than a happy median of the two?

William Sichel said...

This is a fascinating subject. We know that ultra runners will run low on muscle glycogen during races. I think it can only be a good thing to do some long training runs (as long as possible), after an over-night fast and then taking only water during the run. This puts you under an acute metabolic stress which is highly specific to the race situation and will promote favorable adaptations.
On the subject of back to back 20 milers - I feel that this is second best to actually doing a really long run, say 40 miles.

Andy said...

Thanks for your feedback William.

I agree it seems to make sense that doing a long Run in a fasted state will help. Unfortunately there is no research to support this theory.

The question is once you start consuming carbs ( ie during a race ) does that negate the fat burning effects gained in training. Research tends to say yes but there has been no research on exercising for long periods of time as in an ultra.

In my opinion a 40 miler will be of better benefit than back to back 20's but some people may not be able to handle a 40 but they can handle 2x20.