Opinion, not a trial: in a debate article Noakes argues that after 4-6 weeks on either a low-carbohydrate, high-fat or a high-carbohydrate diet trained athletes perform alike (VO2max test, 5-km and 1.6-km treadmill time trials, 6 × 800 m, cycling to exhaustion at 70% VO2max), that about 10 g/h of carbohydrate during exercise is enough because low blood glucose, not muscle glycogen, ends prolonged effort, and that intakes of 30-90 g/h are unnecessary; he rests this mainly on his own group's trials, gives no sample sizes, and Burke's refutation in the same article disputes each point (level C)
Main caveat: An opinion piece in a debate format: no systematic search, no inclusion criteria, no risk-of-bias assessment, no pooled estimate.
This is the "No" side of a debate in The American Journal of Clinical Nutrition. Noakes, who says he promoted high-carbohydrate eating for over 35 years until his certainty unravelled from December 2010, argues that a habitual low-carbohydrate, high-fat (LCHF) diet does not impair exercise performance once athletes have adapted for at least 4 weeks (preferably 6 or more).
His case rests on trials from his own collaboration, which he describes as randomised controlled trials: athletes did equally well on LCHF and high-carbohydrate, low-fat (HCLF) diets in a VO2max test, 5-km and 1.6-km treadmill time trials, six 800-m repetitions at 86% VO2max and a ride to exhaustion at 70% VO2max.
In the last trial 10 g/h of carbohydrate during the ride, against a placebo, lengthened time to exhaustion equally after either diet, which he reads as proof that falling blood glucose (the "small glucose pool" in liver and blood), not muscle glycogen, causes fatigue in prolonged submaximal exercise.
From that he concludes that high in-exercise intakes (30-90 g/h) are unnecessary and about 10 g/h prevents the hypoglycaemia. He adds systematic reviews that found performance maintained on ketogenic diets, three further trials, and his group's Endocrine Reviews analysis.
Burke's refutation in the same article answers that glycogen depletion and hypoglycaemia are each not the sole cause of fatigue, that the 10 g/h trial was not tested against multiple strategies for raising carbohydrate availability, that fat yields less energy than carbohydrate for the same oxygen and so lowers power, and that her elite athletes' race performance was consistently impaired after LCHF.
Noakes's rebuttal says Burke's studies were not randomised and lasted 3.5 weeks, too short for adaptation. A disclaimer states that debate participants may argue positions that differ from their own views.
Participants
No participants: an opinion article.
The trials Noakes cites (from his reference titles; the article gives no numbers): competitive recreational athletes (5-km study), middle-aged men (1.6-km and 6 × 800-m study), triathletes (time-to-exhaustion study), plus runners and cyclists in other groups' trials and professional firefighters in one.
The studies Burke cites: elite athletes in her group's work, including, she says, Olympic and World Championship medallists; their sport and sample sizes are not given in the article.
Protocol
Format: Oxford-style debate; main argument (Noakes), refutation (Burke), rebuttal (Noakes); section editor named; no methods section.
Evidence Noakes selects: his collaboration's trials (refs 8-10, 13-14), his group's Endocrine Reviews analysis (ref 11; in the base as REF-476), systematic reviews of 17 and 13 studies and a meta-analysis of 10, a 28-day firefighter study, and three more trials (a 21-km race, a run to exhaustion with 60 g/h carbohydrate, a 5-h cycling time trial).
Evidence Burke selects: narrative reviews, a 2014 systematic review of carbohydrate during exercise, mouth-rinse and pre-exercise feeding work, her group's elite-athlete studies and an "LCHF sports dashboard".
Funding: no funding statement in the article.
Declared interests (full text): Noakes declares that he has written books on low-carbohydrate nutrition and holds shares in The Nutrition Network, and that all proceeds from these activities are donated to an NGO, The Noakes Foundation.Burke has been an unpaid, short-term member of scientific advisory boards for sports nutrition companies (Gatorade Sports Science Institute expert panel, Science in Sports), has received travel grants from organisations tied to both the carbohydrate and the LCHF industries, and has worked 45 years as a sports dietitian in elite sport, including periodised carbohydrate and specialised LCHF use for ultra-endurance events.The acknowledgements thank Prins, Koutnik, Buga, Volek and D'Agostino for a decade-long collaboration, and Koutnik for reviewing the manuscript.
Key Results
These are claims as the article states them, not findings of this article.
| Claim (who) | Number given | Source the author names |
|---|---|---|
| Equal performance on LCHF and HCLF after 4-6 wk (Noakes) | VO2max test; 5-km and 1.6-km treadmill time trials; 6 × 800 m; time to exhaustion at 70% VO2max | His own trials (refs 8-10) |
| Fat-to-carbohydrate crossover moved by LCHF (Noakes) | from 50% to 85% VO2max | Refs 13-14 |
| Peak fat oxidation in 6 × 800 m at 86% VO2max (Noakes) | > 2.1 g/min in the text; "2 g/min" in the abstract | Ref 9 |
| 10 g/h carbohydrate vs placebo during the ride (Noakes) | time to exhaustion up 12-22% in one paragraph, 12-20% in another, 22% in the abstract, equally after both diets | Ref 10 |
| In-exercise intake needed to prevent hypoglycaemia (Noakes) | about 10 g/h; 30-90 g/h called unnecessary | Refs 11, 21, 23 |
| Studies where carbohydrate helped vs did not (Noakes) | blood glucose fell in up to 88% of placebo groups vs 30%; carbohydrate helped about 2.7 times more often when glucose fell | Ref 11 (his group's review) |
| Run to exhaustion with 60 g/h carbohydrate (Noakes, citing another trial) | about 50 km run; final carbohydrate oxidation 3.1 vs 1.2 g/min, fat 0.4 vs 1.2 g/min (normal vs low-carbohydrate diet) | Ref 30 |
| Fat-adapted elite athletes' fat oxidation (Burke) | about 2 g/min in individuals during sustained high-intensity work | Ref 46 |
| In-exercise carbohydrate dose-response (Burke) | benefit up to 90 g/h, growing with duration | Ref 45 (systematic review) |
| Burke's own trials (Noakes's critique) | not randomised; 3.5 weeks; in the latest, LCHF performance still improving at 3 weeks | Ref 50 |
Noakes's own concession: no evidence that a low-carbohydrate diet improves performance; he claims equally that none shows it impairs any activity studied after adequate adaptation.
Noakes concedes that preventing hypoglycaemia does not let exercise continue indefinitely and that other, presumably non-metabolic, factors dominate beyond 3-4 h.
Burke's list of conditions that disfavour LCHF: sustained intensity above 85% VO2max; and confounders in some studies: large weight loss, order effects, many dropouts.
Practical Application
This is a dissenting position against three guides of the base, [SYNTH-50], [SYNTH-02] and [SYNTH-69], and their defaults do not change on it. SYNTH-50 tells a rider to train the gut toward the race dose, SYNTH-02 sets up to 90 g/h for an amateur's 3-6 h event, and SYNTH-69 sets daily carbohydrate by the day's load and fuels every key session; Noakes argues that about 10 g/h is enough to prevent the low blood glucose he holds responsible for fatigue, and that the habitual diet makes no difference to performance after 4-6 weeks of adaptation.The trial behind the 10 g/h figure compared it only with a placebo, so it cannot show that 10 g/h does as well as 60-90 g/h.
What would change if he is right: the gut-training ladder and the higher per-hour rows of SYNTH-02 would buy little for performance, the daily carbohydrate-by-load advice in SYNTH-69 would matter for health and comfort rather than speed, and a fat-adapted athlete could race on small carbohydrate doses. None of that is tested against the base's doses in the evidence he brings.
What he brings and what he does not. He brings trials from his own collaboration in recreational and middle-aged athletes, with laboratory tests (a VO2max test, short treadmill time trials, an interval session, time to exhaustion at a fixed 70% VO2max), and reviews concluding that performance can be maintained.He brings no head-to-head of 10 g/h against higher doses, no race result from his own trials (the one race he cites, a 21-km run, is another group's trial), no sample sizes in the article, and no answer to Burke's elite-athlete trials (one of that kind is in the base as REF-486) beyond calling them too short and not randomised.
Burke's counter, in the same article: fat yields less energy than carbohydrate for the same oxygen, so power or speed falls; her elite athletes' race results were consistently worse after LCHF; carbohydrate during exercise shows a dose-response up to 90 g/h; and sustained work above 85% VO2max disfavours LCHF.A meta-analysis in the base found repeated high-intensity work tended to be impaired and lactate lower on low-carbohydrate diets (REF-1702, pooled performance estimates not significant).
Inference, not tested here: where the two sides agree is that fat oxidation rises a lot with adaptation and that some carbohydrate during a long effort helps; an athlete who chooses a low-carbohydrate diet for other reasons should still carry carbohydrate on long rides and expect the hardest, most glycolytic sessions to be where any cost shows.
Limitations
An opinion piece in a debate format: no systematic search, no inclusion criteria, no risk-of-bias assessment, no pooled estimate; the evidence is chosen by each debater, and the disclaimer says debaters may argue positions that are not their own.
Self-citation: the core trials and the review behind the hypoglycaemia argument come from the author's collaboration; the acknowledgements name the same collaborators, one of whom reviewed the manuscript.
Declared interests: Noakes writes low-carbohydrate nutrition books and holds shares in The Nutrition Network (all proceeds stated as donated to the NGO The Noakes Foundation); Burke has been an unpaid, short-term member of sports nutrition companies' scientific advisory boards (Gatorade Sports Science Institute expert panel, Science in Sports), has received travel grants from organisations related to both the carbohydrate and the LCHF industries, and has worked 45 years as a sports dietitian in elite sport, including specialised LCHF use for ultra-endurance events.
Numbers that do not agree inside the article: the 10 g/h benefit is 22% in the abstract, 12-22% and 12-20% in the text; peak fat oxidation is "2 g/min" in the abstract and above 2.1 g/min in the text; the abstract calls it the highest ever reported, while Burke reports about 2 g/min in her elite athletes.
What the cited trials can and cannot show (from the article's own description): in his own trials, time to exhaustion at a fixed intensity and short laboratory time trials, not races (the 21-km race he cites is another group's); 10 g/h tested against a placebo, not against higher doses; Noakes calls the trials randomised, but their size and design are not given here.
Not read: the cited trials, the companion "Yes" article and the debate consensus article.
How this entry was checked and changed
Why this level
expert opinion by the rubric: one side of a published debate, built on the author's selection of trials, mostly his own group's; it reports no new data