A major review published in January 2026 in Endocrine Reviews challenges the dogma of muscle glycogen and carbo-loading. It is not the depletion of reserves that causes endurance fatigue, but the inability to maintain stable blood sugar levels. A paradigm shift that could transform the way endurance athletes eat.
The dogma: fill the tank before the race
Since the Scandinavian studies of the 1960s, sports nutrition has been based on a simple equation: the more muscle glycogen you store before the event, the longer you'll last. Hence the pre-marathon pasta parties and the recommendations of 60-90 grams of carbohydrate per hour during the race.
The muscle would be a reservoir to be filled. The «marathon wall»? An empty tank. This vision has shaped decades of products - gels, drinks, bars - all calibrated to maximise carbohydrate intake through carbo-loading.
Proofreading: it's not the muscle, it's the brain
The review published in Endocrine Reviews - more than 100 years of data under the microscope - is the first of its kind in the world. tells a different story.
By re-analysing the original studies, the authors found that the subjects who collapsed at the end of the test not only had low muscle reserves. They also showed marked falls in blood sugar - blood sugar levels. And that's where it gets interesting.
Unlike muscles, the brain cannot function without glucose. When blood glucose levels fall beyond a certain threshold - known as exercise-induced hypoglycaemia (EIH) - the central nervous system triggers a protective response: reduced motor signal, increased sensation of fatigue, subjective inability to maintain intensity.
In other words: it's not the muscle that fails, it's the brain that pulls the handbrake to protect itself. And this even though muscle glycogen reserves are not necessarily depleted.
The liver, an underestimated key player
If the brain depends on blood glucose, where does this glucose come from during exercise? PThe main source of glucose is the liver, which releases its own glycogen reserves and manufactures glucose from other substrates (gluconeogenesis).
The problem is that the liver stores only 50 to 100 grams of glycogen - far less than the muscles. And if its production no longer keeps pace with demand, blood sugar levels fall.
The authors of the review argue that it is this liver failure - and not the depletion of muscle stores - that constitutes the real bottleneck in endurance performance. Protecting blood sugar levels then becomes the priority objective, rather than filling up the muscles.
Is 10 grams per hour enough?
This is the data that is the subject of most debate. A randomised crossover trial carried out on competitive triathletes tested the effect of a minimum intake of 10 grams of carbohydrate per hour during an endurance test.
Result: time to exhaustion increased by 22 %. Above all, exercise-induced hypoglycaemia disappeared - in both groups, whether they were on a high-carbohydrate or low-carbohydrate diet on a daily basis.
Ten grams per hour is a far cry from the 60 to 90 grams recommended by current guidelines and carbo-loading. That's barely half an energy gel. And yet.., this is enough to maintain blood sugar levels and stave off fatigue.
The authors of the review are not saying that high intakes are unnecessary. But they question their systematic necessity and their cost-benefit ratio - particularly in terms of digestive tolerance, a recurring problem among ultra-endurance athletes who try to absorb 90 to 120 grams of carbohydrate per hour.
Low-carb and performance: reconciling the two?
The study on triathletes provides another piece of information. After four weeks of adaptation on a very low-carbohydrate diet (less than 50g/day), endurance performance was comparable to that obtained on a high-carbohydrate diet - provided that this minimum supplement of 10 g/h is provided during exercise.
Keto-adapted« athletes develop remarkable fat oxidation capacities - in excess of 1.5 g/min, even at high intensity. This does not mean that the low-carb diet is superior. But absolute dependence on carbohydrates during exercise is less rigid than previously thought.
What this means for sportspeople
- Less pressure on pre-competition carbo-loading. Replenishing your reserves is still useful, but it may not be the limiting factor we thought.
- More modest and better tolerated intakes during exercise. Rather than forcing 60-90 g/h - with the digestive problems that entails - a regular intake of 10-20 g/h could maintain blood sugar levels without overloading the digestive system.
- A new focus on blood sugar levels. Continuous glucose sensors could become management tools - aiming for glycaemic stability rather than maximising intake.
- A place for low-carb strategies. Athletes who prefer a low-carbohydrate diet are not condemned to under-performance, as long as they keep their intake to a minimum during exercise.
What we don't know yet
The review proposes a change of interpretative framework, but interventional studies remain limited. The trial on triathletes involved a small sample.
The classic recommendations (30-90 g/h) are not out of the blue: they work for many athletes. The challenge is not to abandon them, but to understand when lower intakes may be sufficient.
The question of intensity remains open. At very high intensities, fat oxidation does not fully compensate for carbohydrate demand. The high recommendations are probably still relevant for short, intense efforts.
The debate on carbo-loading has only just begun
What's at stake here goes beyond the question of grams of carbohydrate per hour. It's about challenging the mental model that has dominated for the past 60 years: moving from «muscle-reservoir» to «blood sugar-brain protector».
The data is solid but still patchy. One thing is certain: the debate on carbo-loading has been relaunched, and the next generation of recommendations is unlikely to resemble the previous one.
FAQ
Has carbo-loading become unnecessary? No. The review does not recommend abandoning carbohydrate supercompensation, but suggests that fatigue is linked more to falling blood sugar levels than to muscle glycogen depletion. Pre-competition strategies can be reconsidered, not abandoned.
Are 10g of carbohydrates per hour really enough? In the study cited, yes - for triathletes tested to exhaustion. But the classic recommendations (30-90 g/h) remain relevant for certain intensities, durations and athlete profiles. The optimum intake depends on the context.
Do low-carb diets harm endurance performance? After adaptation (around 4 weeks), endurance performance can be maintained on a low-carbohydrate diet, as long as minimal carbohydrate intake is provided during exercise. Keto-adapted athletes develop a strong capacity for fat oxidation.
How can I tell if I'm hypoglycaemic during exercise? Typical symptoms include sudden and disproportionate fatigue, difficulty concentrating, trembling and a feeling of weakness. Continuous glucose sensors allow objective monitoring, but are still not accessible to most amateurs.
Sources
- Noakes T. et al. Carbohydrate Ingestion on Exercise Metabolism and Physical Performance. Endocrine Reviews, January 2026. academic.oup.com
- Prins P.J. et al. Carbohydrate ingestion eliminates hypoglycemia and improves time to exhaustion in triathletes. 2025. pubmed.ncbi.nlm.nih.gov
- Gatorade Sports Science Institute - Dietary Carbohydrate and the Endurance Athlete. gssiweb.org
Are you interested in this subject?
→ Do you have a specific question? Our AI assistant can give you quick guidance:Talking to the assistant
→ Need structured support? Find out more about our formats - from one-shots to editorial partnerships: See the offers
→ Would you like to discuss it in person? 30 minutes to define your needs, with no obligation: Reserve a slot