Best carbohydrate for performance: does the evidence support the hype?

Carbohydrate has been the cornerstone of endurance fuelling for decades. Carbohydrate ingestion during exercise works, this is not the question but the question that is often asked what is the best carbohydrate for performance. Whether newer carbohydrate products, such as modified starches and hydrogels, and low-carbohydrate protocols improve endurance performance beyond well-established carbohydrate nutrition strategies, is the subject of discussion. The marketing claims suggest that the various new products and approaches work. This blog by Dr Mike Ormsbee will explore the data around carbohydrate fuelling to separate marketing and hype from mechanisms and evidence.

What are the best carbohydrates for performance?

Can modified carbohydrates change metabolism? Hydrothermally modified starches are not "fast" or "slow" carbohydrates in the textbook amylose-to-amylopectin sense. They are instead engineered in the lab. These modified carbohydrates were originally developed for patients with glycogen storage disease, where a slow, steady glucose release was useful from a medical perspective. Athletes saw the description "slow release of glucose" and asked the obvious question: can modified carbohydrates help endurance performance?

In one study, Roberts and colleagues (1), gave nine trained cyclists 1 g/kg of either modified carbohydrate or maltodextrin 30 minutes before a 2.5-hour ride at 70% VO₂max, followed by a time-to-exhaustion bout at 100% VO₂max. The modified carbohydrate condition produced the expected metabolic profile: lower insulin response, slightly higher circulating free fatty acids, and a trend toward higher fat oxidation late in exercise although this did not reach statistical significance . In short, participants saw a modest increase in fat oxidation.

Can modified carbohydrates improve performance?

In the study by Roberts et al. (1), the protocol was not designed to capture how athletes actually compete on race day. In order to address this gap in the literature, our research group conducted a study to better reflect race-day demands. Ten well-trained male cyclists (VO₂max ~59 mL/kg/min) took 60 g modified carbohydrate, 30 g modified carbohydrate, 60 g sucrose (Gatorade), or placebo 30 minutes pre-exercise (3). To simulate feeding during a race, participants then consumed 15 g every 15 minutes during a 180-minute protocol that finished with all-out sprints at 100% of watt max. We replicated the Roberts result on substrate use: fat oxidation was higher with modified carbohydrate.

However we did not see was a performance benefit. At 60 g/h, performance was unchanged, but GI distress and nausea were higher. At 30 g/h, the gut issues were reduced, but performance was actually worse. We then ran the experiment again in elite runners with microdialysis probes inserted into the abdominal fat to measure glycerol release directly (4). This gives us some insight in what happens to the mobilistion of fat duirng exercise. Modified carbohydrate did not change interstitial glycerol, did not change plasma glycerol, and did not change 5 km time-trial performance compared with sucrose or placebo.

Modified carbohydrate intake before and during exercise did not enhance performance and was accompanied by greater GI distress and nausea

Across approximately 20 studies in this area the picture is consistent. Modified starches can shift fuel selection towards greater fat oxidation. However, modified carbohydrates do not consistently improve performance and they can increase nausea at higher doses. For elite athletes trying to win a race, modified carbohydrates do not outperform a well-formulated glucose-fructose drink.

Can carbohydrate hydrogels reduce gastrointestinal distress?

Carbohydrate hydrogels add a gelling agent - typically pectin and alginate - to a normal carbohydrate solution. In the stomach they form a gel that is supposed to slow gastric emptying, encapsulate the sugar, and release it more gradually in the small intestine. The marketing claim is that hydrogels can deliver more carbohydrate to the muscle with less GI distress. Considering GI distress can impair athletic performance, this is an interesting alternative.

In a review paper, King and colleagues (5) identified studies providing carbohydrate hydrogel compared to typical carbohydrate or placebo across various endurance events lasting >1 hour. The review found no clear benefit of carbohydrate hydrogel on carbohydrate delivery to muscle, GI distress, or exercise performance during cycling, running, or skiing.

A recent experimental paper compared a carbohydrate hydrogel to a carbohydrate control or placebo in runners doing a 5 km time trial (6). The authors reported ~2% improvement in 5 km running time. Upon closer inspection, the participants were moderately trained, with 5 km times ranging from approximately 21 to 28 minutes. Whether these same observations hold true in elite athletes running sub-14 minute 5 km remains to be determined.

In summary, carbohydrate hydrogels are not worse than standard carbohydrate drinks, but the data do not support the claim that they deliver more carbohydrate to muscle, reduce GI distress, or meaningfully improve performance compared with a properly dosed glucose-fructose mix. We need higher-intensity, longer-duration studies in athletes who actually experience GI problems before drawing stronger conclusions. Until then, hydrogels are a user-preference call, not an evidence-based fix-all solution.

Is there a role for low carbohydrate pre-exercise?

In the endurance space, moderate to high carbohydrate intakes are advocated to support performance. However, a recent paper made the case that exercise performance is largely about preventing brain hypoglycaemia, and that as little as ~10 g of carbohydrate may be enough to do that (7). The interpretation, picked up enthusiastically online, was that endurance athletes have been over-fuelling on carbohydrate for decades.

However, a critical look at the rationale and at the evidence is required to interpret such claims. The authors proposed that only a small amount of carbohydrate is required to protect the central nervous system glucose supply. However, the conclusion that elite endurance performance can be sustained on minimal carbohydrate is not supported by the broader literature. Studies that infused fat under controlled research conditions decades ago showed no performance benefit, and a heavy reliance on carbohydrate persisted regardless. At intensities approaching VO₂ max, the muscle uses carbohydrate. There is no current evidence that any chronic adaptation removes that requirement for carbohydrate at the very high end of exercise intensity; an intensity that is required for performance in most sports.

Where low-carbohydrate approaches make more sense is in recreational athletes who do not race at very high intensities, and who may have reasons, such as general health, glucose regulation, to prefer lower glycaemic loads. The same recreational athlete can perform very well on far less carbohydrate intakes than the elite. That is a real and useful insight. It important though to separate this from the claim that elites do not need carbohydrate.

What carbohydrate strategy should you use?

The answer depends on your goal. If you are racing to win, or trying to perform at your absolute best, the evidence still points to traditional glucose-fructose mixtures, dosed to your gut tolerance, in the 60–90 g/h range during exercise (with the higher end requiring gut training). Pushing carbohydrate intakes above 90–120 g/h is not yet supported by performance data and some studies show that instead of sparing glycogen we would be sparing fat and the benefits of that are questionable.

If your goal is finishing a race rather than optimising performance, or you are training for general fitness and health, you have more room to experiment. Modified carbohydrates and hydrogels are not better than glucose-fructose mixes, but they are also not worse for performance. If you tolerate them and prefer lower glucose excursions, they are reasonable choices. Having said that you might get the same effect by ingesting a small amount of a glucose-fructose drink. Lower carbohydrate intakes during exercise are also an option for a recreational athlete, especially in events at low to moderate intensities.

A carbohydrate glucose-fructose mix at the right dose, taken by an athlete with a trained gut, remains the most evidence-based strategy

Across various carbohydrate products, including modified carbohydrates and hydrogels, the marketing has consistently outpaced the research. Personal preference is another story and is a very important part of a good plan. A carbohydrate glucose-fructose mix at the right dose, taken by an athlete with a trained gut, remains the most evidence-based strategy.

References

1. Roberts MD, Lockwood C, Dalbo VJ, Volek J, Kerksick CM. Ingestion of a high-molecular-weight hydrothermally modified waxy maize starch alters metabolic responses to prolonged exercise in trained cyclists. Nutrition. 2011;27(6):659-65.

2. Baur DA, Vargas FdeC, Bach CW, Garvey JA, Ormsbee MJ. Slow-absorbing modified starch before and during prolonged cycling increases fat oxidation and gastrointestinal distress without changing performance. Nutrients. 2016;8(7):392.

3. Baur DA, Willingham BD, Smith KA, Kisiolek JN, Morrissey MC, Saracino PG, et al. Adipose lipolysis unchanged by preexercise carbohydrate regardless of glycemic index. Med Sci Sports Exerc. 2018;50(4):827-36.

4. King AJ, Rowe JT, Burke LM. Carbohydrate hydrogel products do not improve performance or gastrointestinal distress during moderate-intensity endurance exercise. Int J Sport Nutr Exerc Metab. 2020;30(5):305-14.

5. Rowe JT, King RFGJ, King AJ, Morrison DJ, Preston T, Wilson OJ, et al. Glucose and fructose hydrogel enhances running performance, exogenous carbohydrate oxidation, and gastrointestinal tolerance. Med Sci Sports Exerc. 2022;54(1):129-40.

6. Noakes TD, Prins PJ, Buga A, D’Agostino DP, Volek JS, Koutnik AP. Carbohydrate ingestion on exercise metabolism and physical performance. Endocr Rev. 2026;47(2):191-243.