Introduction
Now that you are here, go ahead and put down your pitchfork and torch for a few minutes and hear me out.
If you’ve been an endurance athlete for longer than a day, you have probably been told how important carbohydrates are for performance. This is for good reason, as we’ve known for over 100 years now that carbohydrates play a crucial role in exercise performance. By the 1970s, thanks to the work of Scandinavian researchers, we understand a few crucial things about the relationship between glycogen (stored carbohydrate) availability and exercise performance: (1) muscle glycogen is depleted during exercise in an intensity dependent manner; (2) high carbohydrate diets increase muscle glycogen storage and subsequently improve exercise capacity and (3) muscle glycogen storage is acutely enhanced following prior glycogen depletion (i.e., the super-compensation effect). In more recent years, with the advancement of several molecular biology techniques, we know see that glycogen is more than a simple storage site for fuel. The presence, or lack thereof, of muscle and liver glycogen during training bouts also seems to influence the signaling pathways that drive the adaptive response to exercise. In other words, adaptations from an identical training session may vary based on the glycogen status (low or high) that the training session was completed under. This has caused some to ponder whether conducting training while on a low carb diet, or at least in a transient state of low glycogen availability, may be beneficial for overall endurance performance.
So, in this this article I hope to address the following:
- Why are carbohydrates important for endurance exercise in the first place?
- What does training with low glycogen availability do?
- Does training with low glycogen availability get us better results?
- Why is it so hard to get a clear answer?
I did a brief review in the below episode of the BIN Radio podcast, but this will be a more in-depth review.
BIN Radio: BONUS: Train Low, Compete High on Apple Podcasts
Why are carbohydrates important for endurance exercise in the first place?
The most basic factor that determines physical performance in any situation is the metabolic pathway that our body uses to create the energy to sustain that performance. For example, the body of an ultra-distance marathon runner who needs sustained sub-maximal performance for several hours is going to produce energy in a different way than a shot-putter who requires a few seconds of very high-intensity muscular effort. This is important to understand because knowledge of these systems and how they operate will give us an idea of how to train them, and how to support them through proper nutrition. While all three energy systems contribute energy for activity of all intensities, the magnitude in which each system contributes changes drastically as duration and intensity of exercise changes.
The body's primary source of energy for bodily processes is an organic compound named adenosine triphosphate (ATP). ATP is made up of adenosine (composed of adenine and a sugar molecule called ribose), and three phosphate groups. ATP produces energy through a process called hydrolysis (shown in the figure below), in which the ATP molecule is broken down by splitting off one of the phosphate groups, becoming adenosine diphosphate (ADP). This process releases energy that can be used for things like muscular contraction. The problem with this process is that the body has minimal ATP stores, so the ability to continuously replenish ATP is essential for sustaining athletic performance. There are three metabolic pathways through which the body accomplishes this task: phosphagen system, glycolytic system, and the oxidative system. The pathway and substrate used during exercise is determined by the availability of certain resources (like stored carbohydrates) and the intensity of exercise.
Phosphagen System
The phosphagen system provides the majority of the energy for very high-intensity exercises such as low repetition resistance training and sprinting. This energy pathway relies on the hydrolysis of existing ATP stores, with a limited ability to replenish those stores. This ability to replenish ATP is dependent on the presence of creatine phosphate(CP), which provides the free phosphate group needed to create new ATP from ADP. This limitation is one of the rationales for the use of creatine supplementation for athletes who need to perform frequent high-intensity exercise. The idea is that, with the chronic use of creatine supplementation, the athlete can increase their body's stores of CP, which will make it more effective at fueling high-intensity exercise. While this process is the quickest way for our bodies to replenish ATP and effectively fuel high-intensity activity, creatine phosphate is stored in very low quantities, and can only be the primary provider of energy for efforts lasting up to 5 seconds or so.
Glycolytic System (Anaerobic Glycolysis)
The glycolytic system, also known as anaerobic glycolysis, can be characterized as the breakdown of carbohydrate to replenish ATP without the presence of oxygen. This system becomes more involved when the duration of activity becomes longer than what the phosphagen system can support (approximately 6 seconds-2 minutes). By using carbohydrate, which is stored in the muscles or liver as glycogen or in the blood as glucose, ATP can be re-synthesized for more extended periods. Although the replenishment is not as fast as the phosphagen system, the ability to break down carbohydrates without the presence of oxygen allows for relatively high intensities to be maintained.
Oxidative System
The phosphagen and glycolytic systems operate in an anaerobic (without oxygen) environment and can only re-synthesize ATP through the use of creatine phosphate or carbohydrates, respectively. Once exercise duration exceeds the ability of these two systems to sustain activity, the body is forced to decrease the exercise intensity to accommodate the use of the oxidative system. The oxidative system takes advantage of oxygen's presence to oxidize (break down) carbohydrates and fats to replenish ATP. Although protein can contribute small amounts of energy, carbohydrates and fats remain the preferred fuel source for almost all activity. While the oxidative system is much slower, it can provide continual energy for very long periods of time. Due to this, the oxidative system is the primary source of energy both at rest and during low-intensity activities.
Fuel Sources
Alright, so both carbs and fats can be used to provide the energy necessary to provide that constant stream of ATP to the working muscles, with carbs being a bit more versatile with its ability to enter both anaerobic and aerobic pathways. But there’s a catch. Carbohydrates are stored as glycogen in body in two major locations in the body: muscles and the liver. However, we can only store about 2000-3000 calories worth of glycogen at those sites, whereas we have a practically unlimited amount of energy available from fat, even in lean individuals. Now you may be thinking “Pssh, 3000 calories is PLENTY to support most endurance activities. Shirley, I am not going to burn that many calories on my 10k race!”. And I understand why you’d say that. However, there are two assumptions hidden in that statement that are not necessarily true. The first is that glycogen and performance are like gas in a car, in that as long as you are not 100% on E, performance continues unhindered. In reality, our bodies are smart and really do not want to completely empty glycogen stores, and so performance starts to decrease even with modest reductions in glycogen availability. For example, if you are 40% glycogen depleted, your effort will begin to feel subjectively harder at the same objective intensity (pace) than if you were only 10% depleted. In addition to just feeling harder, the amount of force your muscles can produce will be lowered involuntarily the more glycogen depleted you become (i.e. this is not something you can overcome with pure willpower). Read more about this concept here. The second piece is that glycogen is not stored or used in a uniform manner during exercise. For example, if you are performing a set of bench press to failure, the glycogen in your chest and arm muscles are going to deplete pretty fast, but the glycogen in your legs will be largely unaffected. Our 2000-3000 calorie number we were so excited about earlier refers to whole body glycogen, so when we are talking about something like running, we can only really the glycogen of the legs.
This is where you say: “Ok, so I’ll just use fats since I have a never ending supply!”. The problem there is that it takes considerably more oxygen to get the same amount of ATP from fat compared to carbs, meaning your exercise intensity is going to be more limited.
Ok, so we’ve got carbs which are good at fueling performance at higher intensities but are limited by storage capacity, and fats which have tons of storage capacity but are not awesome at fueling higher intensity exercise. If only we could train our body in a way that improves its ability to use fat at higher intensities, that way we can delay the point where glycogen depletion becomes an issue for performance! Enter: today’s main topic.
What does training with low glycogen availability do?
In the intro, we established that maximizing glycogen availability gives us the best performance, particularly during high-intensity efforts, so why would we purposefully undermine that by training with low glycogen? Depending on where an athlete is in a training cycle, the goal may not be to necessarily do sessions at the highest possible intensities. For example, it is very typical for runners to do mostly slow and easy runs in the first part of a training cycle in order to maximize the efficiency of their aerobic energy production. During these sessions, intensity is purposefully limited so that they can spend the most training time producing energy with their aerobic system. Benefits of this type of training include mitochondrial biogenesis (increasing the number of mitochondria in the muscle, which is where aerobic metabolism takes place), shifting of fiber-types towards slower-twitch phenotypes, and increased capillary density (allowing for more efficient delivery of blood to working muscles). Sure, having higher glycogen status could probably increase the running speed that feels “easy”, but that is not the goal here. An easy run at a heart rate of 130bpm and a pace of 9min/mile is not necessarily going to be more beneficial than an easy run at a heart rate of 130bpm and at a 10min/mile pace. Therefore, during these training sessions the main goal is not to achieve the fastest pace for a given distance, but to maximize those adaptations of the aerobic system. So, the question now becomes: how does glycogen status affect those adaptations?
When exercise protocols are matched for work done, carbohydrate restriction augments AMPK and p38MAPK activation, both of which play important roles in the adaptation process and getting us those benefits we discussed above. In short, these signaling pathways that are activated from training with low glycogen availability seem to increase some of the adaptive responses from a training session, including an increased ability to metabolize fat for energy. This is important for endurance performance, because the higher the intensity that we can use fat for fuel, the more we can spare our limited stores of glycogen. As we discussed earlier, as glycogen gets lower during exercise we start seeing commensurate decrements to performance, so it may be beneficial to delay our usage of glycogen until later in a race.
Does training with low glycogen availability get us better results?
We are FINALLY properly set up to take a stab at answering the question! We’ve gone over some of the unique adaptations that can come from training with low glycogen, and they seem to be adaptations that would benefit endurance performance. However, whenever we see a training strategy like this, where the potential benefits are coming from a handful of mechanistic processes, it is always useful to delineate how many additional layers of assumptions lie between the mechanisms (what we’ve covered so far) and the actual outcome we care about (endurance performance).
The below diagram shows us where we stand so far:
Each arrow represents a logical step that requires an additional layer of evidence to support it. A mistake I often see is athletes and coaches seeing evidence for the first claim and assuming it is also direct evidence for the third claim. The mechanistic research in steps one and two are very important for our understanding of how this concept may work and allow us to control for most confounding variables, but there is a limit to how much we can extrapolate those results to actual performance. This last step is where the applied research comes in. Basically, when we actually try this nutrition strategy in endurance athletes, does their performance improve?
There are three main strategies we’ll talk about that can be implemented to put this idea into practice:
- Chronically low carb intake
- Train low, compete high
- Train low sometimes, compete high
As we look at each of these strategies, we will do so through the lens of a fictional athlete, Faust Runnar. To be clear, the fictional outcomes of Faust’s endeavors I will discuss are based on results of the research in each of these strategies, but this seems like more fun than just regurgitating study results. Faust is currently training for the marathon(current PR of 2:30, and can consistently hit the 2:35-2:45 range) and has a daily carb intake of 500g and wants to know what it would look like if he wanted to try some of these strategies.
Chronically low carb intake. In this strategy, Faust lowered his overall daily carb intake to something in the 100g range (increasing fat/protein to achieve caloric balance) with the hopes of becoming a fat-burning machine. Indeed, after several months of doing this, muscle biopsies are showing increases in mitochondrial and capillary density in his leg muscles. While this seems promising to Faust, he can’t help but notice that his tempo runs at race pace are getting increasingly difficult and he can’t seem to hit that extra gear. He also notices he is getting sick a little more often. On race day, he hits a time of 2:55 which is a good bit slower than his normal range.
Train low, compete high. Faust is convinced that his suboptimal performance on race day in the previous scenario was due to not being able to push his pace higher on race day. But what if he trained the same way (with low carbs), but on race day he used higher carbs before and during the race so that he can get the best of both worlds. So he embarks on his next training cycle, notices all the same things in training. On race day, he eats a high carb breakfast and slurps down some Gatorade at gels at every opportunity. He achieves a time of 2:45. Better than before, but at the higher end of his normal range when he had a more traditional diet. Then it hits him. He realizes that just like fat metabolism can improve by training for it specifically, so can carbohydrate metabolism! He figures his race didn’t go as planned because his body just wasn’t used to metabolizing carbs for fuel at that point, in addition to not being used to the faster paces since his higher intensity days in training were compromised by low glycogen availability.
Train low sometimes, compete high. Learning from his past two races, Faust knew he had to address two main limitations. First, he needed to be better fueled for his high-intensity days in training. Second, he wanted his body accustomed to utilizing both carbs and fats for fuel. To achieve this, he went back to a daily carb intake of 500g, but organized his training and nutrition timing so that his easy runs would be completed under low glycogen availability. He would complete a high-intensity session in the afternoons (depleting his glycogen), not refuel in the evening meal, then complete the easy run the following morning. Not discouraged by past results, Faust was sure that this would be the winning strategy. After all his biker friend, Ryder, had told him of a study where this strategy improved certain measures of cycling performance. On race day, he fueled with high carb intakes before and during the race and achieved a time of 2:32. While Faust is happy to be within a few minutes of his all-time PR, he is not too sure that this new approach is the sole reason for his good performance.
If you didn’t follow the storyline, the “train low sometimes, compete high” strategy shows the most promise, but it is still unclear whether it provides a clear benefit over a more traditional “Train high, compete high” strategy. Studies on this specific variation of the strategy are sparse, but show either a positive or neutral effect. The key to this strategy is that daily carbohydrate intakes are still high to support performance, but the timing of that intake is manipulated so that certain sessions are conducted with low glycogen availability.
Could this whole article have just been that last paragraph? Sure, but where is the fun in that? However, I do feel bad that I have not provided the solid answer you are looking for, so let me at least explain WHY I am full of empty promises and disappointment.
Why is it so hard to get a clear answer?
Before we get frustrated with the lack of clear answers in this research, let’s take a second to think about why it is so hard to figure this stuff out and prove it empirically.
-Small sample sizes. Exercise science is notorious for having low sample sizes due to a general lack of funding combined with the logistical difficulty of conducting long-term studies where participants are required to be seen several times a week like they are in training studies. Also, elite athletes are….elite. Which means there just are not very many of them, and even fewer of them are willing to drastically change their nutritional approach for the sake of science when it could mean the difference between them making a living or not.
- Small effect sizes. When we are comparing different training strategies, we are not expecting huge differences in performance. Especially in elite athletics, we are just trying to find things that are 1-2% better. Sure, we could do some stuff with untrained subjects to get over that piece, but with untrained subjects pretty much ANYTHING is going to give us a pretty big effect, and so the differences in effects between two training strategies is still going to be small by comparison. This amplifies the sample size issue because the smaller the effect is that you are trying to find, the larger the sample size that is needed to actually detect that effect.
- Impossible to blind subjects. In research, we typically want to do randomized control trials that are blinded (subjects don’t know whether they are getting the control or intervention treatment) and placebo-controlled. This is challenging in exercise science, because people can usually tell if they are exercising or not. In this example, an endurance athlete is going to know they are in the intervention group once you tell them to stop eating bagels before training sessions. One way to get past this is to find athletes who are already doing one of these strategies and compare them to athletes who have more traditional nutrition practices. However, you then lose the all-important “randomized” part of the research which opens the door for all kinds of confounding variables.
- Athletic performance is complicated. Variation in performance even within the same individual can vary drastically from day-to-day. Let’s say we test an athlete’s 5K time before and after an 8-week intervention and they get 16:00 and 15:50 respectively on those tests. Can we confidently say that the intervention shaved 10 seconds off their 5K time? Probably not. It is likely that 10 seconds is within their normal fluctuation in performance from day-to-day. There are things we can do to try and standardize these variations (controlled diets, ensuring sleep is similar, completing pre and post tests at the same point in a training cycle, etc.) but it is impossible to control for certain things Any effect we are looking for would have to push through a ton of noise, which is especially difficult when the effect is expected to be small in the first place.
In conclusion, the “train low sometimes, compete high” strategy might be a useful one for improving endurance performance long-term, but for now it should be viewed as experimental. One thing is for sure, it should not even be considered if you have not already nailed some of the lower hanging fruit in nutrition. When it comes to nutrition, there is a hierarchy of priorities (see figure below), and strategies like this are far less important than things like overall energy balance, macronutrient distribution, and hydration.
Further Reading
I am taking a new approach with my citations, where instead of citing a billion things in the article itself and providing, which bogs down the flow of the article, and a traditional reference list (that nobody reads), I am just going to start providing links to the sources I got the information from. So here you go:
Periodized Nutrition for Athletes - PubMed (nih.gov)
The central governor model of exercise regulation applied to the marathon - PubMed (nih.gov)
