New research questions why carbs improve exercise performance
A sweeping review asks whether we slow down because muscles run low on glycogen, or because the brain is protecting itself from falling blood sugar.
The Study
"Carbohydrate Ingestion on Exercise Metabolism and Physical Performance"
Where: Endocrine Reviews, January 2026
Authors: Timothy Noakes, Philip Prins, Alex Buga, Dominic D'Agostino, Jeff Volek, Andrew Koutnik
For decades, the standard story about endurance fatigue has gone like this: during long or hard efforts, your muscles burn through glycogen---their stored form of carbohydrate---and once those glycogen stores are depleted, performance inevitably falls. In that model, carbs help mainly by refilling glycogen stores before you start and by supplying extra fuel to the working muscle during exercise.
This review takes a fresh look at that assumption. The authors comb through more than 160 studies and argue that a different pattern stands out. Across many of these experiments, athletes tend to fail not when muscle appears truly out of fuel, but when blood glucose---the sugar circulating in the bloodstream---drops too low. In other words, the key question they're really asking is:
When performance falls apart in long exercise, is it because the muscle is out of glycogen, or because the body can no longer keep blood glucose in a safe range?
They suggest the evidence leans toward the second explanation. In their view, carbohydrates still help performance, but primarily because they prevent or reverse exercise‑induced hypoglycemia (EIH)---a drop in blood sugar during hard, sustained work---rather than because they simply top off muscle glycogen.
They also introduce a brain‑centered idea: that the nervous system monitors blood glucose and starts dialing back how hard you can push before the muscle reaches a true energy crisis. In that framing, carbs are powerful because they keep the brain comfortable enough to keep letting you work.
What the Review Looked At
To build this case, the authors pull together several lines of evidence.
They revisit early Scandinavian work from the 1930s, in which researchers had runners and cyclists go to exhaustion while tracking blood sugar levels. Often, by the time participants felt they "couldn't go on," their blood glucose had fallen into what we would now consider a hypoglycemic range. When the researchers gave them sugar and asked them to continue, the athletes' symptoms eased, and they could keep going for another 30--60 minutes. Notably, this extension happened without an immediate, dramatic shift in the measured respiratory markers of muscle metabolism. Those early papers explicitly floated the idea that the brain, starved of glucose, was driving the sensation of exhaustion.
The review then moves to the famous glycogen‑depletion work from the late 1960s. With the advent of the muscle needle biopsy, scientists could measure glycogen stored in muscle before and after exercise. The landmark 1967 study by Bergström and colleagues showed that athletes who started a 75%-of‑VO₂max cycling test with higher muscle glycogen could ride significantly longer. Those findings launched carb‑loading and cemented the belief that running low on muscle glycogen is what ultimately forces you to stop.
But if you look back at the data tables, blood glucose was also falling, especially in the low‑carb conditions. That part received far less emphasis in the interpretation.
On top of that historical foundation, the authors review a large body of carbohydrate‑feeding studies where athletes:
- Rode or ran for 1.5--4 hours with and without carbs during exercise.
- Had blood glucose, whole‑body carbohydrate and fat oxidation, and sometimes muscle glycogen tracked.
- Finished with either a time‑to‑exhaustion test or a time trial.
These are the experiments that underpin modern guidelines suggesting endurance athletes consume 60--90 grams of carbohydrate per hour, sometimes more, in long events.
More recently, low‑carb, high‑fat (LCHF) adaptation studies have added another layer. In these trials, athletes spend weeks on either a high‑carb or low‑carb diet, then complete standardized exercise tests, sometimes with small carbohydrate doses during the effort. These studies are useful for disentangling how much performance depends on glycogen and carbohydrate burning versus the ability to lean on fat.
Finally, the review incorporates animal work. In mice and rats, researchers can manipulate glycogen storage or block liver glucose production and see whether endurance fails when muscle glycogen is low, or when blood glucose can no longer be maintained, even if muscle glycogen is still available.
What the Review Actually Found
Fatigue often shows up alongside falling blood glucose
Across many human studies, a simple pattern recurs: in prolonged exercise without carbohydrate intake, blood glucose tends to drift down over time. When that drop is large enough, athletes slow or stop. When those same athletes repeat the protocol with carbohydrates during exercise, several things typically happen:
- Blood glucose stays closer to its starting value or rises modestly.
- Time to exhaustion is extended, or time‑trial performance improves.
- Muscle glycogen at the point of fatigue often isn't dramatically higher than in the no‑carb condition.
The authors went study by study and categorized the results. In about two‑thirds of the trials where performance was measured, taking carbs during exercise improved endurance. In that subset, roughly 88% of the placebo or control trials also showed a clear drop in blood glucose. In trials where blood glucose in the control condition stayed relatively stable, carbs were much less likely to show a performance boost.
That doesn't prove blood sugar is the only factor, but it does line up with the idea that carbs seem to matter most when they are preventing an otherwise inevitable slide in blood glucose.
Muscle energy isn't "empty" at the point of exhaustion
The classic story is: as long exercise goes on, muscles chew through glycogen, ATP production can't keep up, and the muscle runs out of usable energy. The review questions whether that's what's actually happening at the moment athletes stop.
When researchers have taken muscle biopsies at or near exhaustion---both after prolonged submaximal work and intense efforts---they generally find that ATP levels are maintained within a narrow range. If ATP truly collapsed, we'd expect something more like rigor (locked, non‑relaxing muscles) rather than the familiar feeling of global fatigue and an inability to maintain pace.
To the authors, this points toward some upstream regulation---likely in the brain and central nervous system---that reduces drive to the muscle before the muscle experiences catastrophic energy failure. From that perspective, sensing a fall in blood glucose and "pulling the plug" is a plausible protective strategy.
Carbs during exercise don't consistently save muscle glycogen
If carbohydrates during exercise mainly helped by sparing muscle glycogen, we'd expect to see clear, consistent glycogen preservation whenever people took in carbs. The actual picture is mixed.
In many of the studies where muscle biopsies were done:
- Carbohydrate intake during exercise did not significantly reduce the rate of muscle glycogen breakdown compared to placebo.
- In some cases, especially with higher carbohydrate availability---either from frequent feeding or intravenous glucose---muscle glycogen use was actually higher, because the whole system shifted further toward carbohydrate as fuel while fat oxidation was suppressed.
At the same time, carbohydrate feeding clearly affected liver metabolism:
- Liver glycogen breakdown and liver glucose output generally fell when blood glucose and insulin were elevated by carbohydrate intake.
- In some tracer work, high carbohydrate intake almost completely shut down the liver's own glucose production during exercise.
The overall pattern looks less like "carbs are preserving the big glycogen tank in muscle" and more like "carbs are helping preserve the small but critical glucose pool in blood and liver."
More carbs in doesn't always mean better performance
The review also asks how much carbohydrate is actually needed during exercise to get most of the benefit.
When athletes consume carbs at increasing rates---say from 10--15 grams per hour up to 90--120 grams per hour---labs see predictable metabolic changes:
- The body uses more of the carbohydrate you're taking in, especially when the drink contains more than one type of sugar (for example, glucose plus fructose).
- It relies less on fat, often almost gram‑for‑gram in energy terms, to keep overall energy output about the same.
What's less consistent is any extra performance benefit beyond the smallest effective dose. In several long‑ride or long‑run protocols followed by a time trial:
- Modest carb intake---around 15--30 grams per hour---was enough to prevent blood glucose from falling and to improve performance compared with no carbs.
- Raising intake beyond that clearly changed what fuels were being burned, but did not reliably make the athletes faster in the time trial.
The authors' interpretation is that once you've done enough to keep blood glucose from sliding downward, the additional gains from pushing carb intake higher may be smaller than we often assume, at least outside of very high‑intensity or elite race situations.
Low‑carb athletes can still perform, if blood glucose is protected
Low‑carb, high‑fat diets provide a natural test of how much performance really depends on carbohydrate.
In several trials:
- After several weeks on a low‑carb diet, athletes showed very high rates of fat oxidation, sometimes over 1.5 grams per minute, even at intensities above 80% of VO₂max.
- Their muscle glycogen stores and overall carbohydrate burning were lower than on high‑carb diets.
- Yet in prolonged, submaximal tests, their time‑to‑exhaustion or time‑trial results were similar to high‑carb conditions, as long as hypoglycemia was avoided.
One study the review highlights used a simple intervention during a long submaximal ride: athletes on both diets took in around 10 grams of carbohydrate per hour. That relatively small amount:
- Eliminated exercise‑induced hypoglycemia.
- Improved time‑to‑exhaustion by about 12--20%.
- Did so without fully "restoring" carbohydrate oxidation to high‑carb levels.
For the authors, this is a key piece: even when glycogen and carb use are generally lower, a small amount of carbohydrate during long exercise can still deliver a meaningful performance bump, mainly by stabilizing blood glucose.
Animal studies point to liver glycogen as a key regulator
In rodents, researchers can do things we obviously can't in humans: knock out specific enzymes, over‑express certain proteins, or block particular pathways.
The review points to a few telling findings:
- Mice engineered to store more glycogen in the liver, but not necessarily in muscle, are able to sustain exercise longer and keep blood glucose stable compared to controls.
- Mice that lack the ability to synthesize muscle glycogen do not show dramatic drops in endurance, as long as liver glycogen and gluconeogenesis are intact.
- When liver gluconeogenesis is blocked in rats, endurance plummets and blood glucose falls, even if muscle glycogen stores are still available.
These studies support the idea that hepatic glycogen and the liver's ongoing glucose production are central to sustaining blood glucose and, by extension, endurance, while muscle glycogen, although important, may not be the single decisive factor we often make it out to be.
Why This Matters for Endurance Athletes
For years, the takeaway from sports science has been straightforward: hard racing relies heavily on carbohydrate; glycogen is finite; therefore, you should maximize muscle glycogen before the event and push carbohydrate intake as high as your gut will tolerate during it.
This review doesn't overturn those practices, but it does suggest a shift in emphasis.
In the framework Noakes and colleagues propose:
- Glycogen is still a major fuel. Early in exercise, especially at higher intensities, muscles rely heavily on it.
- But many athletes appear to hit the wall when the system can no longer keep blood glucose in a comfortable range, not when muscle energy is truly "empty."
- The brain seems to respond to falling blood sugar by reducing how much muscle it's willing to recruit, long before ATP levels in the muscle collapse.
In that view, the most important job of carbohydrate during long exercise is to support blood glucose and keep the brain out of perceived danger, not simply to stuff as much fuel as possible into the muscle.
Practically, that suggests a few things:
- For long, submaximal efforts, you may get most of the benefit of carbohydrate with a relatively modest, steady intake, as long as it's enough to prevent a steady drift downward in blood glucose.
- If you train or live on a lower‑carb diet and are well adapted, you may be able to rely more on fat for much of the effort, but a small amount of carbohydrate during very long sessions can still serve as an important backstop against hypoglycemia.
- Rather than asking, "What's the maximum number of grams per hour I can handle?", it may be more useful to ask, "Am I doing enough to keep my blood sugar from crashing late in the session or race?"
None of this means high‑carb strategies are wrong, or that carb‑loading is useless. The review is careful to point out that context matters: event duration, intensity, your training status, and your usual diet all influence what's optimal. But it does argue that we should think of carbs less as a one‑way ticket to "more energy" and more as a tool for controlling when and whether your brain ever sees a drop in blood glucose big enough to pull back on the reins.
What the Review Does Not Settle
Because this is a narrative review rather than a pooled meta‑analysis, there are important caveats.
- Many of the underlying studies are small, with highly trained but not necessarily elite athletes.
- Time‑to‑exhaustion tests at fixed workloads, a common design in this literature, don't always mimic real‑world racing, where people choose their own pace and tactics.
- There are well‑documented cases where very high carbohydrate intakes do seem to help, particularly in elite settings or races done close to the limits of sustainable intensity.
The review also doesn't claim that muscle glycogen is unimportant. Depletion clearly changes which fibers are available and how they function, and there are contexts---such as repeated high‑intensity surges---where local muscle factors may dominate.
What it does do is pull together a lot of disparate data and ask whether the simplest possible story---that you slow down only when the muscle runs out of glycogen---really fits what we see. The authors argue that a model in which the brain is watching blood glucose and acting conservatively matches just as much of the evidence, especially in very long efforts.
Bottom Line for the Reader
Stripped down, the message of this review is:
- When you "hit the wall" late in a long workout, it may be less about your muscles literally running out of glycogen, and more about your body failing to keep blood glucose in the zone your brain considers safe.
- Carbohydrates help because they shore up that blood‑and‑liver glucose pool, giving the brain permission to keep letting you push.
- For many people in long, steady efforts, you may not need to max out your hourly carbohydrate intake to get most of that benefit. A modest, consistent intake that prevents blood sugar from sagging may be enough.
- Different background diets---higher‑carb or lower‑carb---change which fuels you lean on, but in both cases, a small amount of carbohydrate during very long sessions can still make a noticeable difference, mainly by protecting blood glucose.
The review doesn't offer a one‑size‑fits‑all fueling rule. But it does suggest a shift in what you pay attention to. Instead of thinking only in terms of "how much glycogen do I have?", it may be worth asking a more specific question: "In the back half of my longest efforts, is my blood sugar being defended?" Carbs are still a key tool for endurance. This paper just argues that their most important job may be in the bloodstream and the brain, not only in the muscle.
Take control of your health with expert guidance
Levels pairs real-time glucose data and comprehensive lab testing with clinician analysis and personalized support—everything you need to turn insights into real health improvements. Click here to get started with Levels.
