Education for runners and endurance athletes. Learn to train smart, run fast, and be strong.
These articles are free.
Please help keep them alive by buying me a beer:
Buy me a beer.
This article is part of a series:
→ Part 1 — How much fuel is in your body?
→ Part 2 — How do your muscles burn fuel?
→ Part 3 — How long can you go?
→ Part 4 — Carboloading
→ Part 5 — Race day carb availability
→ Part 6 — Putting it into practice
→ Part 1 — How much fuel is in your body?
→ Part 2 — How do your muscles burn fuel?
→ Part 3 — How long can you go?
→ Part 4 — Carboloading
→ Part 5 — Race day carb availability
→ Part 6 — Putting it into practice
Performance nutrition: Part 5 of 6:
Why should you consider maintaining high carbohydrate availability during a race? And, will high carbohydrate availability improve your performance?
Thomas Solomon PhD.
13th June 2021.
You know how much fuel is stored in your body (if you don’t, read Part 1). You know how your muscles burn fuel during exercise (if you don’t, read Part 2). You know how long your stored fuel allows you to go (if you don’t, read Part 3). And, you know you can create a high carbohydrate availability before your race (if you don’t, read Part 4). But, can you maintain high carbohydrate availability during your race? And… Will doing so improve your race day performance? Stay with me today and I will take you on a deep-dive history lesson, starting with the 1924 Boston marathon, to find out how we know what we now know...
Reading time ~20-mins (4000-words).
Or listen to the Podcast version.
Or listen to the Podcast version.
So far you have learned that a high carbohydrate availability can be established before your race with a high-carb diet on the days before race day and a carb-full breakfast on the morning of your race. The obvious missing component of your race day performance nutrition equation is...
In 1924, yep 100 years ago, Sam Levine and colleagues at Harvard Medical School found that runners completing the Boston marathon had lower blood glucose levels than when they started and noted that the “condition” of the athletes at the end — symptoms of physical weakness, pallor, and collapse — was associated with their blood glucose levels. They described it as “a picture of shock not unlike that produced by an overdose of insulin”. Anyone who has hung out at the finish line of a marathon will relate to that description.
At the Boston marathon the following year, 17 athletes (including those who raced the previous year) were asked to eat a high-carbohydrate diet for 24-hours before the race and to start eating candy after about 24 kms (15 miles) into the race. They observed that the runners had a better “condition” at the finish line, their post-race blood glucose was higher than baseline, and their race times improved.
About 15-years later, in 1939, the same Erik Christensen and Ove Hansen who found that drinking 200 grams of “sugar water” before exercise, allowed untrained folks to ride for longer before hypoglycemia and fatigue occurred, also found that if the same “sugar water” was given at the point of fatigue, blood glucose levels were restored and subjects could ride for longer. (Note: you will need to learn German or have your German-speaking wife help you translate this paper)
These observations imply that a drop in blood glucose levels — hypoglycaemia — is a cause of fatigue but it took the Earth a few “runs” around the sun for anyone to delve deeper… In their 1967 paper showing that exercise depletes muscle glycogen and that the liver attempts to keep supplying glucose to the muscle by releasing more glucose into the blood, Jonas Bergström and Eric Hultman also found that intravenous glucose infusion helped reduce muscle glycogen use during exercise but that muscle glycogen was still responsible for the greater part of the energy production during exercise even when the blood sugar level is high. Subsequent work in endurance-trained athletes throughout the 80s, 90s, and 2000s showed that the decline in blood glucose contributes to fatigue during prolonged exercise, that glucose infusion during exercise can restore healthy blood glucose levels after hypoglycemia has developed, and that glucose infusion prolongs exercise time-to-exhaustion after muscle glycogen has been depleted during long-duration (2-3-hours) low-to-moderate intensity exercise (see here and here). But, the same studies showed that infusion doesn’t prevent muscle glycogen depletion during exercise (see here and here) and doesn’t improve high-intensity (75-80% VO2max, ~1-hour) time trial performance.
To put all that in plain English, we’ve known since the 60s that muscles keep munching through their glycogen during exercise even if glucose is abundantly available in the blood and that infusing glucose during exercise prevents hypoglycemia and delays fatigue.
This sounds wonderful but intravenous glucose infusion is neither allowed (it violates WADAs ethical code) nor is it practical during a race. So, the alternative is to eat carbohydrates like a normal human, just like Levine and colleagues examined in 1924.
Surprisingly, after Christensen and Hansen’s work in 1939, the role of carbohydrate ingestion on performance was largely ignored for a long ole time. But in 1961, using a carbon-14 radioisotope of glucose, Reichard and colleagues showed for the first time that “exogenous” glucose (that is glucose not produced in the body) can be “burned” during exercise to produce energy. Ten years later, using a similar method, Dave Costill showed that ingested glucose is “burned” 6.5-times faster during exercise than at rest while decreasing liver glucose output. This proved that eating glucose during exercise provides a fuel source to muscles to reduce the burden on the liver.
In 1982, John Wahren’s lab found that when untrained regular folks drank “sugar water” (either 10 or 20 grams of glucose every 15-mins) during a ride-to-exhaustion at 60-65% VO2max following an overnight fast, hypoglycemia was prevented but time-to-exhaustion was unaffected when compared to water. But, as you are probably aware, studying untrained folks is not so useful for informing athletes’ knowledge. Fear not; many studies have examined endurance-trained athletes…
Later in the 80s, Ed Coyle and colleagues found that ingesting ~70-grams of maltodextrin (a glucose polymer) 30-mins into the ride and ~18-grams every ~30-mins thereafter, prolonged time-to-exhaustion during high-intensity cycling (70 to 80% VO2max) from 2-hours 14-mins to 2-hours 37-mins, on average, after an overnight fast, when compared to water. In a very similar follow-up study in 1986, they found that ~140-grams of maltodextrin at 20-mins and ~28-grams every 20-mins thereafter increased time-to-exhaustion from ~3 to ~4-hours, when compared to water, and that hypoglycemia was prevented, carbohydrate oxidation rates were maintained, but that muscle glycogen depletion was unaffected. To better simulate race-day conditions, in 1988 Tim Noakes and colleagues staged an outdoor 42 km marathon. Runners tapered their training and were fed a high carb diet for 2-days before the race, ate a carb-containing breakfast on the morning of the race, and drank (on average) either 10-grams of glucose, 38-grams of maltodextrin, or 44-grams of fructose every hour during the race. Similar to lab-based data, they found that blood glucose levels were maintained throughout the race but that neither the amount nor the type of carbohydrate prevented muscle glycogen depletion (nor was performance different between trials).
In the 90s and 2000s, stable isotopes were used to deeply study the flux (rate of movement into/out of tissues) and oxidation (“burning” to produce energy) of glucose in endurance-trained athletes during exercise, teaching us that glucose ingestion during prolonged exercise reduces liver glucose output but does not prevent muscle glycogen depletion. For example, data from Asker Jeukendrup in 2004, showed that ~97 to 99% of all glucose taken up into muscle during exercise is “burned” and that distributing low (~72 grams) or high (~360 grams) amounts of glucose throughout exercise suppresses liver glucose output and increases glucose uptake into muscle. In fact, ingesting 22-grams of glucose or sucrose (a glucose-fructose disaccharide) every 15-mins during exercise can completely suppress liver glucose output and prevents liver glycogen depletion, as shown using 13C-NMR methods. But, muscle glycogen is eventually depleted during exercise, even when as much as 360 grams of glucose is ingested.
So, in simple words:
Ingesting carbohydrate during exercise does not prevent muscle glycogen breakdown — muscle glycogen is always “burned” during exercise — instead, carbohydrate feeding simply maintains blood glucose levels, which reduces the need for liver glucose output, sparing liver glycogen for longer, all the while providing a continually high carbohydrate availability to the muscles.
With that knowledge in mind, since we know that faster runners burn glucose at higher rates (also see here, here, and here), faster runners will need to be more aware of their need for high carbohydrate availability during their race. Therefore, ingesting carbohydrate during a race allows glucose, the economical and rapid fuel, to be “burned” at the high rate required to provide energy to the working muscles.
For a visual representation of how all this fits together, check out my animation below:
So, yes, high carbohydrate availability can be maintained during a race. But, to make an informed decision as to whether this is a good idea, there is a very important question to consider...
Since short duration exercise cannot deplete liver or muscle glycogen, carbohydrate intake is perhaps unnecessary during short races. Although some studies (see Below et al. 1995 and Jeukendrup et al. 1997) have shown that carb intake during events as short as an 1-hour can assist performance, this only occurs in athletes who are fasted. When a high-carb diet is eaten on the days before trials and a high-carb breakfast is eaten on the morning of trials, no effect of “during exercise” carbohydrate intake is seen during performance tests lasting about 1-hour. Therefore, always consider nutrition in the context of your race duration — is it longer than your bodily stores of carbohydrate can last?
Note: The figure below will help you understand this point. To read about this in detail, read Part 3 of this series.
As I mentioned earlier, in 1924, while Europe was rationing sugar in the aftermath of WW1, Sam Levine and colleagues at Harvard Medical School were dishing it out to runners competing in the Boston marathon. This was to help prevent the signs of hypoglycemia they had observed in athletes at the end of the race the previous year. As you know already, they asked 17 athletes (including those who raced the previous year) to eat a high-carbohydrate diet for 24-hours before the race. But, they also asked them to start eating candy about 24 kms (15 miles) into the race. The result: athletes had a better “condition” at the finish line, higher post-race blood glucose levels than pre-race, and completed the marathon quicker than the previous year. As a fun side note, the chap who won the race in 1924 and came second in 1925 did not eat a high carb diet nor did he eat candy during the race — a fat-burning machine — but he also only ran 2:30 and 2:34 in each year, which means he would have been about half an hour slower than present-day carbohydrate-gobbling machines (even before the advent of supershoes).
Those investigators at Harvard hypothesised that by preventing hypoglycaemia during a prolonged, high-intensity race, they could prevent fatigue and improve performance. Of course, we cannot know that their intervention was the direct cause of better performance because they did not conduct a randomised controlled trial (RCT) and a lot can change in a year — better training, better taper, cooler weather…
Fortunately, many RCTs have been completed over the last ~50-years, permitting systematic reviews and meta-analyses of the evidence.
A 2011 meta-analysis of 88 RCTs providing carbohydrate before and/or during exercise-to-exhaustion or a time trial from Tom Vandenbogaerde and Will Hopkins, concluded that ingesting carbohydrate (vs. water) with an appropriate composition at an appropriate rate can improve endurance performance by ~2% to ~6%. Another 2011 meta-analysis from Temesi and colleagues found a small (2%) to moderate (7.5%) improvement in time trial or time-to-exhaustion performance when 30 to 80 grams per hour of carbs are ingested during exercise. And, a 2014 meta-regression by Trent Stellingwerff and Greg Cox found a relationship between increasing exercise time and the percent increase in performance with carb intake versus placebo (water), stating that multiple transportable carbohydrates (e.g. glucose and fructose) may be beneficial in prolonged exercise, but that athletes should tailor their needs based on tolerance (very important).
In 2018, Aird and colleagues systematically reviewed studies that compared fed vs. fasted‐state exercise on performance, concluding that eating carbohydrates (and carb-containing meals) before exercise improves aerobic performance for longer efforts (60-mins +) but not short-duration performance (less than 60-mins).
Useful info? Yes. But, another pitfall to these meta-analyses is that many of the included trials had a long-duration (1 to 2-hours) moderate-intensity exercise bout (presumably to deplete glycogen) immediately before a performance test. I’ve never understood what scientists are trying to model with this design because it doesn’t reflect many race scenarios (except maybe an often futile, solo breakaway effort in a cycling road race but even then, said cyclist, let’s call him “Jens Voigt” ,would have been well-fed up to the point he decided to give it large). So, what about “real life” performance studies?
The 2011 meta-analysis from Temesi and colleagues separated “performance test” from “pre-exercise + performance test” studies, finding a slightly smaller (2%) performance improvement with 30 to 80 grams/hour carb ingestion when subjects have not completed a pre-exercise bout, vs. a 7.5% improvement when they have. I.e. during-race feeding will enhance performance but more so when glycogen stores are low from the start (which is not an ideal scenario to put yourself in).
Also convincing, but one more nuance remains... Many studies examine “performance” with a time-to-exhaustion ride or run at a low to moderate intensity (60 to 70% VO2max) — what does that really tell us about performance? A time trial is more informative.
In 2013, Colombani and colleagues found only 17 RCTs (1 running, 1 soccer, 15 cycling) that mimicked “real life” — examining the effect of carbohydrate intake before or during a time trial in endurance-trained athletes who had eaten breakfast. But they found an unlikely effect with time trials up to ~70-minutes and a less-than-compelling ergogenic effect with time trials longer than ~70-mins. They used appropriately stringent inclusion criteria, excluding hundreds of papers, noting the general low-quality of studies... commenting that “The absence of clear evidence is, nevertheless, not clear evidence of an absent effect”. In 2016, when more RCTs were published, Pöchmüller and colleagues completed a similar meta-analysis of time trials in athletes who had eaten a pre-trial meal (breakfast). They found that ingesting carbs in a concentration range of 6–8% (6-8 g per 100 mL) before and/or while exercising longer than 90-minutes improves performance. But, due to the lack of sufficient RCTs, Pöchmüller et al. commented that findings cannot yet be confidently extrapolated to elite athletes or female athletes, and more work is needed on bouts lasting less than 90-minutes and in sports besides cycling.
This all sounds ace but if, like me, you have followed this body of evidence for nearly 20-years, it will be clear that, in general, we know a lot but there is a lack of randomised controlled trials in elite athletes, in female athletes, and in real race-like settings. This will keep scientists from gazing into blank spaces for a few more moons to come. Furthermore, most studies have examined cyclists. So, I was excited in 2016 to see Patrick Wilson publish a critical review of all known studies and meta-analyses to answer the question, “Does Carbohydrate Intake During Endurance Running Improve Performance?”. He concluded that running performance is most likely improved during events longer than 2-hours when 100–200 ml of a carbohydrate drink (5–8 grams per 100 mL; i.e. 5 to 16 grams of carbs) is drunk every 15–20 minutes, although several studies show benefits for tasks lasting 90 to 120-minutes. I was less excited to see how limited the scientific evidence base is in running — notably, there is a lack of studies in runners, including a lack of studies examining gels and foods in runners — therefore, in running, we largely apply cycling-based scientific evidence to generate running-based empirical evidence through observation in the field. And, in doing so… does carbohydrate intake during a (long) race improve performance?.
It did in the 1920s in Boston and it does now. Given what Levine and his colleagues learned at the Boston marathon 100-years ago, little has really changed — a high carbohydrate availability is key for successful race day performance.
In recent years, several governing bodies and reputable societies have updated their sports nutrition guidelines. In 2016, the ACSM published their new “Nutrition and Athletic Performance” position statement. In 2018, the International Society of Sports Nutrition (ISSN) published their “exercise & sports nutrition review update”. And, in 2019, the International Journal of Sports Nutrition and Exercise Metabolism dedicated an entire volume of 17 papers to the IAAF consensus statement on “Nutrition for Athletics”. Plus, there was a specific focus on “Nutrition for Ultramarathon Running” as part of the 2019 IAAF consensus statement. And, also in 2019, the ISSN published their position stand on “Nutritional considerations for single-stage ultra-marathon training and racing”.
To summarise what these guidelines say about carbohydrate intake...
For general endurance athletes:
And specifically for ultra-distance runners:
As you can see, there’s a little bit of variation between the various position statements, which is no surprise when different groups of experts evaluate evidence in its entirety, but, promisingly, the general sentiments are very similar and are “running down the same trails”.
Note: For a nicely written narrative on this topic, I can thoroughly recommend reading, “Toward a Common Understanding of Diet–Exercise Strategies to Manipulate Fuel Availability for Training and Competition Preparation in Endurance Sport” by Louise Burke and colleagues.
This cartoon represents a summary of the guidelines and position stands from ACSM (2016), ISSN (2018), IAAF (2019), ISSN (2019) and IAAF (2019).
On your “A” race day, you will need to line up with a high carbohydrate availability and, after the B of the bang, if your race is long, you should be planning to maintain a high carbohydrate availability throughout the race. Doing so will help delay fatigue for as long as possible by sparing liver glycogen, keeping blood glucose within the normal range, and supplying glucose to the hungry muscles. As Renato Canova’s Kenyan athletes say, “when I finish the fuel, I stop” — the best distance runners on Earth understand their metabolic limitations very well.
Establishing and maintaining high carbohydrate availability during your race can be achieved with three simple steps:
Muscle glycogen “supercompensation” (aka carbo-loading) is probably best achieved by eating a high-carbohydrate diet for the 24 to 48-hours before race day while tapering your training or resting altogether. “Carbo-loading” in this way is especially important if you typically eat a low-carb diet.
Your liver glycogen store, which is depleted after a night of sleep, will be maximised if you eat a high-carbohydrate breakfast on race day, 2- to 4-hours before gun time.
Consuming carbohydrates at regular intervals from the start of a race lasting around 60-mins or longer will help delay glycogen depletion and maintain blood glucose levels for as long as possible.
So, if you haven’t already guessed it, what Sam Levine and colleagues discovered in 1924 at the Boston marathon — to eat lots of carbs for the 24-hours before a race, on the morning of the race, and during the race — is about as simple and accurate a message as anyone can put in a bottle. Some folks are just ahead of their time.
Sports nutrition guidelines for establishing and maintaining high carb availability on race day are pretty clear and very useful. They also confirm the notion that high carbohydrate availability is a key facet of your endurance performance. The various guidelines and position statements are also built on scientific evidence, which is good, but scientific evidence must always be balanced with empirical evidence from coaches and feedback from athletes. Why? Well, because there is no “one size fits all” approach. In their utility, the specifics of sports nutrition guidelines — the grams per day and the grams per hour — are rather generalised and should not be used in a cookie-cutter type way. Consequently, one key question remains... “How can you maintain high carbohydrate availability on race day?”. Stay tuned and find out in the final part of this series...
Until that time, keep training smart...
Can you maintain high carbohydrate availability during your race?
Unlike our canine companions, who can rapidly replenish muscle glycogen levels during prolonged exercise even when carbohydrate intake is low (see here and here), we are not blessed with glycogen replenishing powers when on the go — when we move for prolonged periods, we deplete glycogen. Although some evidence in endurance-trained athletes shows that, during prolonged (~3-hours) low-intensity cycling, glycogen resynthesis can occur in inactive type 2 (aka “fast-twitch”) muscle fibres when carbohydrate is consumed, the amount is small and resynthesis does not occur in the type 1 (aka “slow-twitch”) fibres being used. Consequently, if you want to increase carbohydrate availability during a race, you need to add some carbohydrate into the mix.In 1924, yep 100 years ago, Sam Levine and colleagues at Harvard Medical School found that runners completing the Boston marathon had lower blood glucose levels than when they started and noted that the “condition” of the athletes at the end — symptoms of physical weakness, pallor, and collapse — was associated with their blood glucose levels. They described it as “a picture of shock not unlike that produced by an overdose of insulin”. Anyone who has hung out at the finish line of a marathon will relate to that description.
At the Boston marathon the following year, 17 athletes (including those who raced the previous year) were asked to eat a high-carbohydrate diet for 24-hours before the race and to start eating candy after about 24 kms (15 miles) into the race. They observed that the runners had a better “condition” at the finish line, their post-race blood glucose was higher than baseline, and their race times improved.
About 15-years later, in 1939, the same Erik Christensen and Ove Hansen who found that drinking 200 grams of “sugar water” before exercise, allowed untrained folks to ride for longer before hypoglycemia and fatigue occurred, also found that if the same “sugar water” was given at the point of fatigue, blood glucose levels were restored and subjects could ride for longer. (Note: you will need to learn German or have your German-speaking wife help you translate this paper)
These observations imply that a drop in blood glucose levels — hypoglycaemia — is a cause of fatigue but it took the Earth a few “runs” around the sun for anyone to delve deeper… In their 1967 paper showing that exercise depletes muscle glycogen and that the liver attempts to keep supplying glucose to the muscle by releasing more glucose into the blood, Jonas Bergström and Eric Hultman also found that intravenous glucose infusion helped reduce muscle glycogen use during exercise but that muscle glycogen was still responsible for the greater part of the energy production during exercise even when the blood sugar level is high. Subsequent work in endurance-trained athletes throughout the 80s, 90s, and 2000s showed that the decline in blood glucose contributes to fatigue during prolonged exercise, that glucose infusion during exercise can restore healthy blood glucose levels after hypoglycemia has developed, and that glucose infusion prolongs exercise time-to-exhaustion after muscle glycogen has been depleted during long-duration (2-3-hours) low-to-moderate intensity exercise (see here and here). But, the same studies showed that infusion doesn’t prevent muscle glycogen depletion during exercise (see here and here) and doesn’t improve high-intensity (75-80% VO2max, ~1-hour) time trial performance.
To put all that in plain English, we’ve known since the 60s that muscles keep munching through their glycogen during exercise even if glucose is abundantly available in the blood and that infusing glucose during exercise prevents hypoglycemia and delays fatigue.
This sounds wonderful but intravenous glucose infusion is neither allowed (it violates WADAs ethical code) nor is it practical during a race. So, the alternative is to eat carbohydrates like a normal human, just like Levine and colleagues examined in 1924.
Surprisingly, after Christensen and Hansen’s work in 1939, the role of carbohydrate ingestion on performance was largely ignored for a long ole time. But in 1961, using a carbon-14 radioisotope of glucose, Reichard and colleagues showed for the first time that “exogenous” glucose (that is glucose not produced in the body) can be “burned” during exercise to produce energy. Ten years later, using a similar method, Dave Costill showed that ingested glucose is “burned” 6.5-times faster during exercise than at rest while decreasing liver glucose output. This proved that eating glucose during exercise provides a fuel source to muscles to reduce the burden on the liver.
In 1982, John Wahren’s lab found that when untrained regular folks drank “sugar water” (either 10 or 20 grams of glucose every 15-mins) during a ride-to-exhaustion at 60-65% VO2max following an overnight fast, hypoglycemia was prevented but time-to-exhaustion was unaffected when compared to water. But, as you are probably aware, studying untrained folks is not so useful for informing athletes’ knowledge. Fear not; many studies have examined endurance-trained athletes…
Later in the 80s, Ed Coyle and colleagues found that ingesting ~70-grams of maltodextrin (a glucose polymer) 30-mins into the ride and ~18-grams every ~30-mins thereafter, prolonged time-to-exhaustion during high-intensity cycling (70 to 80% VO2max) from 2-hours 14-mins to 2-hours 37-mins, on average, after an overnight fast, when compared to water. In a very similar follow-up study in 1986, they found that ~140-grams of maltodextrin at 20-mins and ~28-grams every 20-mins thereafter increased time-to-exhaustion from ~3 to ~4-hours, when compared to water, and that hypoglycemia was prevented, carbohydrate oxidation rates were maintained, but that muscle glycogen depletion was unaffected. To better simulate race-day conditions, in 1988 Tim Noakes and colleagues staged an outdoor 42 km marathon. Runners tapered their training and were fed a high carb diet for 2-days before the race, ate a carb-containing breakfast on the morning of the race, and drank (on average) either 10-grams of glucose, 38-grams of maltodextrin, or 44-grams of fructose every hour during the race. Similar to lab-based data, they found that blood glucose levels were maintained throughout the race but that neither the amount nor the type of carbohydrate prevented muscle glycogen depletion (nor was performance different between trials).
In the 90s and 2000s, stable isotopes were used to deeply study the flux (rate of movement into/out of tissues) and oxidation (“burning” to produce energy) of glucose in endurance-trained athletes during exercise, teaching us that glucose ingestion during prolonged exercise reduces liver glucose output but does not prevent muscle glycogen depletion. For example, data from Asker Jeukendrup in 2004, showed that ~97 to 99% of all glucose taken up into muscle during exercise is “burned” and that distributing low (~72 grams) or high (~360 grams) amounts of glucose throughout exercise suppresses liver glucose output and increases glucose uptake into muscle. In fact, ingesting 22-grams of glucose or sucrose (a glucose-fructose disaccharide) every 15-mins during exercise can completely suppress liver glucose output and prevents liver glycogen depletion, as shown using 13C-NMR methods. But, muscle glycogen is eventually depleted during exercise, even when as much as 360 grams of glucose is ingested.
So, in simple words:
Ingesting carbohydrate during exercise does not prevent muscle glycogen breakdown — muscle glycogen is always “burned” during exercise — instead, carbohydrate feeding simply maintains blood glucose levels, which reduces the need for liver glucose output, sparing liver glycogen for longer, all the while providing a continually high carbohydrate availability to the muscles.
For a visual representation of how all this fits together, check out my animation below:
So, yes, high carbohydrate availability can be maintained during a race. But, to make an informed decision as to whether this is a good idea, there is a very important question to consider...
Does carbohydrate intake during a race improve performance?
The answer to this question is best answered by first considering the duration of your race.Since short duration exercise cannot deplete liver or muscle glycogen, carbohydrate intake is perhaps unnecessary during short races. Although some studies (see Below et al. 1995 and Jeukendrup et al. 1997) have shown that carb intake during events as short as an 1-hour can assist performance, this only occurs in athletes who are fasted. When a high-carb diet is eaten on the days before trials and a high-carb breakfast is eaten on the morning of trials, no effect of “during exercise” carbohydrate intake is seen during performance tests lasting about 1-hour. Therefore, always consider nutrition in the context of your race duration — is it longer than your bodily stores of carbohydrate can last?
Note: The figure below will help you understand this point. To read about this in detail, read Part 3 of this series.
×
![]()
To examine this question in detail, let's jump back into John DeLorean’s time-travelling car and head back to Boston…
As I mentioned earlier, in 1924, while Europe was rationing sugar in the aftermath of WW1, Sam Levine and colleagues at Harvard Medical School were dishing it out to runners competing in the Boston marathon. This was to help prevent the signs of hypoglycemia they had observed in athletes at the end of the race the previous year. As you know already, they asked 17 athletes (including those who raced the previous year) to eat a high-carbohydrate diet for 24-hours before the race. But, they also asked them to start eating candy about 24 kms (15 miles) into the race. The result: athletes had a better “condition” at the finish line, higher post-race blood glucose levels than pre-race, and completed the marathon quicker than the previous year. As a fun side note, the chap who won the race in 1924 and came second in 1925 did not eat a high carb diet nor did he eat candy during the race — a fat-burning machine — but he also only ran 2:30 and 2:34 in each year, which means he would have been about half an hour slower than present-day carbohydrate-gobbling machines (even before the advent of supershoes).
Those investigators at Harvard hypothesised that by preventing hypoglycaemia during a prolonged, high-intensity race, they could prevent fatigue and improve performance. Of course, we cannot know that their intervention was the direct cause of better performance because they did not conduct a randomised controlled trial (RCT) and a lot can change in a year — better training, better taper, cooler weather…
Fortunately, many RCTs have been completed over the last ~50-years, permitting systematic reviews and meta-analyses of the evidence.
A 2011 meta-analysis of 88 RCTs providing carbohydrate before and/or during exercise-to-exhaustion or a time trial from Tom Vandenbogaerde and Will Hopkins, concluded that ingesting carbohydrate (vs. water) with an appropriate composition at an appropriate rate can improve endurance performance by ~2% to ~6%. Another 2011 meta-analysis from Temesi and colleagues found a small (2%) to moderate (7.5%) improvement in time trial or time-to-exhaustion performance when 30 to 80 grams per hour of carbs are ingested during exercise. And, a 2014 meta-regression by Trent Stellingwerff and Greg Cox found a relationship between increasing exercise time and the percent increase in performance with carb intake versus placebo (water), stating that multiple transportable carbohydrates (e.g. glucose and fructose) may be beneficial in prolonged exercise, but that athletes should tailor their needs based on tolerance (very important).
×
![]()
These analyses are convincing but these systematic reviews did not use very stringent inclusion criteria and also included studies in which subjects were fasted. Because, as you already know, it is not particularly wise for an endurance athlete to line up with low liver glycogen levels on race day, it is more appropriate to examine studies in which subjects have eaten breakfast before testing.
In 2018, Aird and colleagues systematically reviewed studies that compared fed vs. fasted‐state exercise on performance, concluding that eating carbohydrates (and carb-containing meals) before exercise improves aerobic performance for longer efforts (60-mins +) but not short-duration performance (less than 60-mins).
Useful info? Yes. But, another pitfall to these meta-analyses is that many of the included trials had a long-duration (1 to 2-hours) moderate-intensity exercise bout (presumably to deplete glycogen) immediately before a performance test. I’ve never understood what scientists are trying to model with this design because it doesn’t reflect many race scenarios (except maybe an often futile, solo breakaway effort in a cycling road race but even then, said cyclist, let’s call him “Jens Voigt” ,would have been well-fed up to the point he decided to give it large). So, what about “real life” performance studies?
The 2011 meta-analysis from Temesi and colleagues separated “performance test” from “pre-exercise + performance test” studies, finding a slightly smaller (2%) performance improvement with 30 to 80 grams/hour carb ingestion when subjects have not completed a pre-exercise bout, vs. a 7.5% improvement when they have. I.e. during-race feeding will enhance performance but more so when glycogen stores are low from the start (which is not an ideal scenario to put yourself in).
Also convincing, but one more nuance remains... Many studies examine “performance” with a time-to-exhaustion ride or run at a low to moderate intensity (60 to 70% VO2max) — what does that really tell us about performance? A time trial is more informative.
In 2013, Colombani and colleagues found only 17 RCTs (1 running, 1 soccer, 15 cycling) that mimicked “real life” — examining the effect of carbohydrate intake before or during a time trial in endurance-trained athletes who had eaten breakfast. But they found an unlikely effect with time trials up to ~70-minutes and a less-than-compelling ergogenic effect with time trials longer than ~70-mins. They used appropriately stringent inclusion criteria, excluding hundreds of papers, noting the general low-quality of studies... commenting that “The absence of clear evidence is, nevertheless, not clear evidence of an absent effect”. In 2016, when more RCTs were published, Pöchmüller and colleagues completed a similar meta-analysis of time trials in athletes who had eaten a pre-trial meal (breakfast). They found that ingesting carbs in a concentration range of 6–8% (6-8 g per 100 mL) before and/or while exercising longer than 90-minutes improves performance. But, due to the lack of sufficient RCTs, Pöchmüller et al. commented that findings cannot yet be confidently extrapolated to elite athletes or female athletes, and more work is needed on bouts lasting less than 90-minutes and in sports besides cycling.
This all sounds ace but if, like me, you have followed this body of evidence for nearly 20-years, it will be clear that, in general, we know a lot but there is a lack of randomised controlled trials in elite athletes, in female athletes, and in real race-like settings. This will keep scientists from gazing into blank spaces for a few more moons to come. Furthermore, most studies have examined cyclists. So, I was excited in 2016 to see Patrick Wilson publish a critical review of all known studies and meta-analyses to answer the question, “Does Carbohydrate Intake During Endurance Running Improve Performance?”. He concluded that running performance is most likely improved during events longer than 2-hours when 100–200 ml of a carbohydrate drink (5–8 grams per 100 mL; i.e. 5 to 16 grams of carbs) is drunk every 15–20 minutes, although several studies show benefits for tasks lasting 90 to 120-minutes. I was less excited to see how limited the scientific evidence base is in running — notably, there is a lack of studies in runners, including a lack of studies examining gels and foods in runners — therefore, in running, we largely apply cycling-based scientific evidence to generate running-based empirical evidence through observation in the field. And, in doing so… does carbohydrate intake during a (long) race improve performance?.
It did in the 1920s in Boston and it does now. Given what Levine and his colleagues learned at the Boston marathon 100-years ago, little has really changed — a high carbohydrate availability is key for successful race day performance.
×
![]()
So far in this series, I have made a deep dive into the individual scientific studies and the systematic reviews and meta-analyses that have brought us to where our knowledge is today. But, to relay the synthesis of evidence to the target audience — to coaches and athletes — governing bodies and professional societies publish guidelines. So...
What do sports nutrition guidelines say?
Older guidelines from the 1990s recommended that endurance athletes should consume a total of about 600 g every day during heavy training and to consume carbohydrates during exercise, generally in the form of solutions containing glucose, sucrose, or maltodextrins, at a rate of 30 to 60 grams per hour. The old (and now "retired") American College of Sports Medicine (ACSM) position stand from 2009 gave similar advice: daily intake of 6 to 10 grams of carbs per kg body weight (~400-650 grams per day) and 30-60 grams of carbs per hour during exercise. With their ears-to-the-ground of the emerging evidence from science and practice, in 2011, Louise Burke, John Hawley, and Asker Jeukendrup proposed an update to the recommendations, stating that “carbohydrate intake during exercise should be scaled according to the characteristics of the event”. They suggested that athletes could fluctuate daily carbohydrate intake relative to their needs — the emergence of “carbohydrate periodisation” — and that, during exercise, smaller amounts of carbs (30 grams/h) could be ingested during shorter events with higher rates of intake (60...90...120 grams/h) during longer events.In recent years, several governing bodies and reputable societies have updated their sports nutrition guidelines. In 2016, the ACSM published their new “Nutrition and Athletic Performance” position statement. In 2018, the International Society of Sports Nutrition (ISSN) published their “exercise & sports nutrition review update”. And, in 2019, the International Journal of Sports Nutrition and Exercise Metabolism dedicated an entire volume of 17 papers to the IAAF consensus statement on “Nutrition for Athletics”. Plus, there was a specific focus on “Nutrition for Ultramarathon Running” as part of the 2019 IAAF consensus statement. And, also in 2019, the ISSN published their position stand on “Nutritional considerations for single-stage ultra-marathon training and racing”.
To summarise what these guidelines say about carbohydrate intake...
For general endurance athletes:
The ACSM (2016) recommends
basing daily carb intake on training load:
… 3 to 5 grams of carbs per kg body weight for light training;
… 5 to 7 g/kg for a moderate training load (1 h/day);
… 6 to 10 g/kg for a high load (1-3 h/day);
… 8 to 12 g/kg for a very high load (more than 4-5 h/day).
Pre-race carb-loading with a diet containing 10 to 12 g/kg/day of carbs for 36–48 hours, for events longer than 90-mins.
Pre-race feeding 1 to 4-hours before the race with 1 to 4 g/kg of carbs.
During race carb intake based on duration:
... up to 45 mins → no carbs needed;
... 45 to 75 mins → a mouth rinse or small carbohydrate amount;
… 1 to 2½ h → 30–60 g/h of carbs;
... longer than 2½ h → up to 90 g/h of carbs.
The IAAF (2019) recommends
pre-race carb-loading with 10 to 12 g/kg/day of carbs for 36 to 48-hours before the race,
a pre-race meal 1 to 4-hours before the race containing 1 to 4 g/kg of carbs,
during-race nutrition depending on race duration:
... 45 to 75 min → a mouth rinse or small carbohydrate amount;
… 1 to 2½ h → 30–60 g/h of carbs;
... longer than 2½ h → up to 90 g/h of carbs.
And the ISSN (2018) recommends
consuming 30 to 60 g/hour of carbohydrate in a 6–8% carbohydrate-electrolyte solution every 10 to 15 min throughout high-intensity exercise longer than 90-minutes.
basing daily carb intake on training load: … 3 to 5 grams of carbs per kg body weight for light training;
… 5 to 7 g/kg for a moderate training load (1 h/day);
… 6 to 10 g/kg for a high load (1-3 h/day);
… 8 to 12 g/kg for a very high load (more than 4-5 h/day).
Pre-race carb-loading with a diet containing 10 to 12 g/kg/day of carbs for 36–48 hours, for events longer than 90-mins.
Pre-race feeding 1 to 4-hours before the race with 1 to 4 g/kg of carbs.
During race carb intake based on duration: ... up to 45 mins → no carbs needed;
... 45 to 75 mins → a mouth rinse or small carbohydrate amount;
… 1 to 2½ h → 30–60 g/h of carbs;
... longer than 2½ h → up to 90 g/h of carbs.
The IAAF (2019) recommends
pre-race carb-loading with 10 to 12 g/kg/day of carbs for 36 to 48-hours before the race,
a pre-race meal 1 to 4-hours before the race containing 1 to 4 g/kg of carbs,
during-race nutrition depending on race duration: ... 45 to 75 min → a mouth rinse or small carbohydrate amount;
… 1 to 2½ h → 30–60 g/h of carbs;
... longer than 2½ h → up to 90 g/h of carbs.
And the ISSN (2018) recommends
consuming 30 to 60 g/hour of carbohydrate in a 6–8% carbohydrate-electrolyte solution every 10 to 15 min throughout high-intensity exercise longer than 90-minutes.
And specifically for ultra-distance runners:
The ISSN (2019) recommends
a general moderate-to-high carbohydrate diet of 5 to 8 g/kg/day during training and during a race, to prevent caloric deficits, to aim to consume 150–400 kcals/hour to include 30–50 grams/hour of carbs and 5–10 grams/hour of protein from a variety of calorie-dense foods while considering food palatability, tolerance, and savoury vs. sweet preference in longer races.
While the IAAF (2019) recommends
consuming ~0.8 to 1 g/kg/hour carbohydrate during exercise in a 6–10% weight to volume solution.
a general moderate-to-high carbohydrate diet of 5 to 8 g/kg/day during training and during a race, to prevent caloric deficits, to aim to consume 150–400 kcals/hour to include 30–50 grams/hour of carbs and 5–10 grams/hour of protein from a variety of calorie-dense foods while considering food palatability, tolerance, and savoury vs. sweet preference in longer races.
While the IAAF (2019) recommends
consuming ~0.8 to 1 g/kg/hour carbohydrate during exercise in a 6–10% weight to volume solution.
As you can see, there’s a little bit of variation between the various position statements, which is no surprise when different groups of experts evaluate evidence in its entirety, but, promisingly, the general sentiments are very similar and are “running down the same trails”.
Note: For a nicely written narrative on this topic, I can thoroughly recommend reading, “Toward a Common Understanding of Diet–Exercise Strategies to Manipulate Fuel Availability for Training and Competition Preparation in Endurance Sport” by Louise Burke and colleagues.
×
![]()
What can you put in your performance nutrition toolbox?
Since the evidence shows that carbohydrate ingestion allows exercise to continue even if glycogen levels are depleted, then theoretically, you could start the race with low liver and muscle glycogen if enough carbohydrate can be ingested during a race to meet the metabolic demand of your muscles. But, this is a bloody big risk. The evidence also shows that low muscle glycogen reduces power output possibly due to a direct blunting of muscle contractility. Furthermore, if you cannot ingest enough carbohydrate during the race (due to poor logistics, aversion, or sickness etc) then you will very quickly encounter Darth Fader, the Sith Lord of fatigue. Don’t “run” that risk, especially when considering how oh-so-simple it is to start a race with high liver and muscle glycogen levels...On your “A” race day, you will need to line up with a high carbohydrate availability and, after the B of the bang, if your race is long, you should be planning to maintain a high carbohydrate availability throughout the race. Doing so will help delay fatigue for as long as possible by sparing liver glycogen, keeping blood glucose within the normal range, and supplying glucose to the hungry muscles. As Renato Canova’s Kenyan athletes say, “when I finish the fuel, I stop” — the best distance runners on Earth understand their metabolic limitations very well.
Establishing and maintaining high carbohydrate availability during your race can be achieved with three simple steps:
Muscle glycogen “supercompensation” (aka carbo-loading) is probably best achieved by eating a high-carbohydrate diet for the 24 to 48-hours before race day while tapering your training or resting altogether. “Carbo-loading” in this way is especially important if you typically eat a low-carb diet.
Your liver glycogen store, which is depleted after a night of sleep, will be maximised if you eat a high-carbohydrate breakfast on race day, 2- to 4-hours before gun time.
Consuming carbohydrates at regular intervals from the start of a race lasting around 60-mins or longer will help delay glycogen depletion and maintain blood glucose levels for as long as possible.
Sports nutrition guidelines for establishing and maintaining high carb availability on race day are pretty clear and very useful. They also confirm the notion that high carbohydrate availability is a key facet of your endurance performance. The various guidelines and position statements are also built on scientific evidence, which is good, but scientific evidence must always be balanced with empirical evidence from coaches and feedback from athletes. Why? Well, because there is no “one size fits all” approach. In their utility, the specifics of sports nutrition guidelines — the grams per day and the grams per hour — are rather generalised and should not be used in a cookie-cutter type way. Consequently, one key question remains... “How can you maintain high carbohydrate availability on race day?”. Stay tuned and find out in the final part of this series...
Until that time, keep training smart...
Disclaimer: I occasionally mention brands and products but it is important to know that I am not affiliated with, sponsored by, an ambassador for, or receiving advertisement royalties from any brands. I have conducted biomedical research for which I have received research money from publicly-funded national research councils and medical charities, and also from private companies, including Novo Nordisk Foundation, AstraZeneca, Amylin, A.P. Møller Foundation, and Augustinus Foundation. I’ve also consulted for Boost Treadmills and Gu Energy on their research and innovation grant applications and I’ve provided research and science writing services for Examine — some of my articles contain links to information provided by Examine but I do not receive any royalties or bonuses from those links. These companies had no control over the research design, data analysis, or publication outcomes of my work. Any recommendations I make are, and always will be, based on my own views and opinions shaped by the evidence available. My recommendations have never and will never be influenced by affiliations, sponsorships, advertisement royalties, etc. The information I provide is not medical advice. Before making any changes to your habits of daily living based on any information I provide, always ensure it is safe for you to do so and consult your doctor if you are unsure.
If you find value in this free content, please help keep it alive and buy me a beer:
Buy me a beer.
Share this post on your social media:
Want free exercise science education delivered to your inbox? Join the 100s of other athletes, coaches, students, scientists, & clinicians and sign up here:
About the author:
I am Thomas Solomon and I'm passionate about relaying accurate and clear scientific information to the masses to help folks meet their fitness and performance goals. I hold a BSc in Biochemistry and a PhD in Exercise Science and am an ACSM-certified Exercise Physiologist and Personal Trainer, a VDOT-certified Distance running coach, and a Registered Nutritionist. Since 2002, I have conducted biomedical research in exercise and nutrition and have taught and led university courses in exercise physiology, nutrition, biochemistry, and molecular medicine. My work is published in over 80 peer-reviewed medical journal publications and I have delivered more than 50 conference presentations & invited talks at universities and medical societies. I have coached and provided training plans for truck-loads of athletes, have competed at a high level in running, cycling, and obstacle course racing, and continue to run, ride, ski, hike, lift, and climb as much as my ageing body will allow. To stay on top of scientific developments, I consult for scientists, participate in journal clubs, peer-review papers for medical journals, and I invest every Friday in reading what new delights have spawned onto PubMed. In my spare time, I hunt for phenomenal mountain views to capture through the lens, boulder problems to solve, and for new craft beers to drink with the goal of sending my gustatory system into a hullabaloo.
Copyright © Thomas Solomon. All rights reserved.
I am Thomas Solomon and I'm passionate about relaying accurate and clear scientific information to the masses to help folks meet their fitness and performance goals. I hold a BSc in Biochemistry and a PhD in Exercise Science and am an ACSM-certified Exercise Physiologist and Personal Trainer, a VDOT-certified Distance running coach, and a Registered Nutritionist. Since 2002, I have conducted biomedical research in exercise and nutrition and have taught and led university courses in exercise physiology, nutrition, biochemistry, and molecular medicine. My work is published in over 80 peer-reviewed medical journal publications and I have delivered more than 50 conference presentations & invited talks at universities and medical societies. I have coached and provided training plans for truck-loads of athletes, have competed at a high level in running, cycling, and obstacle course racing, and continue to run, ride, ski, hike, lift, and climb as much as my ageing body will allow. To stay on top of scientific developments, I consult for scientists, participate in journal clubs, peer-review papers for medical journals, and I invest every Friday in reading what new delights have spawned onto PubMed. In my spare time, I hunt for phenomenal mountain views to capture through the lens, boulder problems to solve, and for new craft beers to drink with the goal of sending my gustatory system into a hullabaloo.
Copyright © Thomas Solomon. All rights reserved.