Fat Burns. Carbs Win.Two scientists. One question. Does going low-carb hurt endurance performance?
https://www.sciencedirect.com/science/article/pii/S0002916526000808
https://www.sciencedirect.com/science/article/pii/S0002916526000778
https://www.sciencedirect.com/science/article/pii/S000291652600078X
Louise Burke says yes. Tim Noakes says no. Theyx.com/i/article/2060751339299221509
https://www.sciencedirect.com/science/article/pii/S0002916526000808
https://www.sciencedirect.com/science/article/pii/S0002916526000778
https://www.sciencedirect.com/science/article/pii/S000291652600078X
Louise Burke says yes. Tim Noakes says no. Theyx.com/i/article/2060751339299221509
The manifesto: academic.oup.com/edrv/article/4…
1. The Strawman of "Obligatory Fuel" and Muscle Rigor
Noakes builds his entire case against the Anaplerotic Theory (TAT) by claiming that if muscle glycogen depletion truly caused fatigue, ATP would have to fall low enough to produce muscle rigor. Since rigor never happens, TAT must be wrong.
This is a textbook strawman. Nobody in modern exercise physiology argues that fatigue equals ATP depletion to rigor levels. The actual mechanisms by which low glycogen impairs performance are well documented and have nothing to do with reaching rigor. They include reduced calcium release from the sarcoplasmic reticulum, altered excitation-contraction coupling, changes in Na+/K+ ATPase activity, and reduced economy of ATP production due to lower CHO contribution per litre of O2. Noakes knocks down a position no current researcher holds, then declares the field disproven. That is not how science works.
1. The Strawman of "Obligatory Fuel" and Muscle Rigor
Noakes builds his entire case against the Anaplerotic Theory (TAT) by claiming that if muscle glycogen depletion truly caused fatigue, ATP would have to fall low enough to produce muscle rigor. Since rigor never happens, TAT must be wrong.
This is a textbook strawman. Nobody in modern exercise physiology argues that fatigue equals ATP depletion to rigor levels. The actual mechanisms by which low glycogen impairs performance are well documented and have nothing to do with reaching rigor. They include reduced calcium release from the sarcoplasmic reticulum, altered excitation-contraction coupling, changes in Na+/K+ ATPase activity, and reduced economy of ATP production due to lower CHO contribution per litre of O2. Noakes knocks down a position no current researcher holds, then declares the field disproven. That is not how science works.
2. The "Having Your Cake and Eating It" Result
Noakes' showcase finding, repeated throughout the review, is that 10 g/h of carbs during exercise rescued performance equally on both low-carb and high-carb diets in the Prins 2025 study. He treats this as the death blow to the high-carb model.
It is actually the death blow to his own model. The original Volek, Phinney and Noakes pitch for LCHF was that fat adaptation removes the need for exogenous carbs. If you require ingested carbohydrate during exercise to perform, then CHO is not optional; it is essential. Noakes has shifted the question from "do you need carbs" to "how few can you get away with," which is a completely different debate. The 10 g/h number is also conveniently tiny, but you can only spin it as a win if you ignore the fact that, without those carbs, his keto-adapted athletes underperform.
Noakes' showcase finding, repeated throughout the review, is that 10 g/h of carbs during exercise rescued performance equally on both low-carb and high-carb diets in the Prins 2025 study. He treats this as the death blow to the high-carb model.
It is actually the death blow to his own model. The original Volek, Phinney and Noakes pitch for LCHF was that fat adaptation removes the need for exogenous carbs. If you require ingested carbohydrate during exercise to perform, then CHO is not optional; it is essential. Noakes has shifted the question from "do you need carbs" to "how few can you get away with," which is a completely different debate. The 10 g/h number is also conveniently tiny, but you can only spin it as a win if you ignore the fact that, without those carbs, his keto-adapted athletes underperform.
3. Capacity vs Performance, and Recreational vs Elite
Almost every RCT he leans on measures time to exhaustion, Wingate output, treadmill ramp tests, or 1RM lifts in recreational athletes. Capacity is not performance. Time to exhaustion at 70 percent VO2max is not the Tour de France. A 5 km treadmill run with a metabolic cart attached is not a 5 km race.
Where it matters most, elite athletes at race intensities in ecologically valid settings, the LCHF diet costs measurable time. The Supernova series in elite race walkers showed this consistently. Noakes dismisses these as non-RCTs and excludes them from his analysis. He then leans on his own non-randomized observations from athletes who were never going to win an Olympic medal. The findings he most needs to address are the ones he most aggressively excludes.
Almost every RCT he leans on measures time to exhaustion, Wingate output, treadmill ramp tests, or 1RM lifts in recreational athletes. Capacity is not performance. Time to exhaustion at 70 percent VO2max is not the Tour de France. A 5 km treadmill run with a metabolic cart attached is not a 5 km race.
Where it matters most, elite athletes at race intensities in ecologically valid settings, the LCHF diet costs measurable time. The Supernova series in elite race walkers showed this consistently. Noakes dismisses these as non-RCTs and excludes them from his analysis. He then leans on his own non-randomized observations from athletes who were never going to win an Olympic medal. The findings he most needs to address are the ones he most aggressively excludes.
4. Selective Inclusion and the Double Standard on Evidence
Noakes makes a great show of rigour by citing 160 plus studies and his "16 lines of evidence." Look closer, and the standard he applies to his own evidence is dramatically softer than the one he applies to Burke's.
He throws out the Supernova studies as non-RCTs but leans heavily on rat studies (the PTG overexpression mice), a single 1936 case study by Boje, Christensen and Hansen from 1939, and retrospective reinterpretations of studies whose original authors drew the opposite conclusion. He repeatedly accuses other authors of reaching unwarranted conclusions, then, in the same paragraph, offers his own reinterpretation as the "more logical explanation," with no new data. This is not a reanalysis; it is rewriting. And his core empirical claim, that 88 percent of studies showing CHO benefit had falling blood glucose in controls, is the kind of post hoc data mining that needs independent replication before it gets treated as an established fact. He treats it as proven within his own paper.
Noakes makes a great show of rigour by citing 160 plus studies and his "16 lines of evidence." Look closer, and the standard he applies to his own evidence is dramatically softer than the one he applies to Burke's.
He throws out the Supernova studies as non-RCTs but leans heavily on rat studies (the PTG overexpression mice), a single 1936 case study by Boje, Christensen and Hansen from 1939, and retrospective reinterpretations of studies whose original authors drew the opposite conclusion. He repeatedly accuses other authors of reaching unwarranted conclusions, then, in the same paragraph, offers his own reinterpretation as the "more logical explanation," with no new data. This is not a reanalysis; it is rewriting. And his core empirical claim, that 88 percent of studies showing CHO benefit had falling blood glucose in controls, is the kind of post hoc data mining that needs independent replication before it gets treated as an established fact. He treats it as proven within his own paper.
5. The Crossover Point and the Fat Oxidation Sleight of Hand
Noakes claims that keto-adapted athletes can hit fat oxidation rates of 2 g/min even at 85 to 90 percent VO2max, and uses this to argue that fat can fuel essentially any endurance event, including a sub-2-hour marathon. He then constructs Figures 25 and 26 showing that fat plus modest CHO oxidation can theoretically cover the energy cost of world records from 42 km to 500 km.
The biochemistry he never engages with is that fat oxidation yields roughly 5 to 7 percent less ATP per litre of oxygen than carbohydrate oxidation. This has been known since Krogh and Lindhard in 1920, which Noakes himself cites. At the oxidative ceiling that determines elite endurance pace, that deficit is the whole game. A 5 to 7 percent loss of economy at threshold is the difference between winning and missing the podium. Noakes acknowledges fat oxidation supports the energy, but never reckons with the fact that producing the same power output costs more oxygen on fat. He also leans on the idea that respiratory exchange ratios above 1.0 overestimate CHO oxidation, which is a real measurement caveat, but he applies it asymmetrically to inflate his fat oxidation numbers while not adjusting his CHO numbers in the studies he criticizes. And his sub-2-hour marathon argument requires a 52 kg runner. Real elite marathoners are heavier and run faster than his model assumes, and Kipchoge's actual nutrition strategy was high-carb gel feeding throughout. The biochemistry won.
Noakes claims that keto-adapted athletes can hit fat oxidation rates of 2 g/min even at 85 to 90 percent VO2max, and uses this to argue that fat can fuel essentially any endurance event, including a sub-2-hour marathon. He then constructs Figures 25 and 26 showing that fat plus modest CHO oxidation can theoretically cover the energy cost of world records from 42 km to 500 km.
The biochemistry he never engages with is that fat oxidation yields roughly 5 to 7 percent less ATP per litre of oxygen than carbohydrate oxidation. This has been known since Krogh and Lindhard in 1920, which Noakes himself cites. At the oxidative ceiling that determines elite endurance pace, that deficit is the whole game. A 5 to 7 percent loss of economy at threshold is the difference between winning and missing the podium. Noakes acknowledges fat oxidation supports the energy, but never reckons with the fact that producing the same power output costs more oxygen on fat. He also leans on the idea that respiratory exchange ratios above 1.0 overestimate CHO oxidation, which is a real measurement caveat, but he applies it asymmetrically to inflate his fat oxidation numbers while not adjusting his CHO numbers in the studies he criticizes. And his sub-2-hour marathon argument requires a 52 kg runner. Real elite marathoners are heavier and run faster than his model assumes, and Kipchoge's actual nutrition strategy was high-carb gel feeding throughout. The biochemistry won.
Conclusion
The message is the same as before. Carbohydrates win, and fat is not an efficient fuel for high-performance exercise. Yes, you can run a long way at a slow pace on a high-fat diet. For the 0.00001% of people whose entire sporting goal is to finish a 200 km ultramarathon at 5 km/h, the LCHF approach can technically work. For the rest of endurance sport, where pace at threshold determines outcomes and where the difference between fuels shows up in seconds per kilometre, the answer has not changed in a hundred years. Carbohydrate oxidation produces more ATP per litre of oxygen consumed. More ATP per litre of oxygen means more power at the same VO2. More power at the same VO2 means you finish ahead of the keto guy. Noakes can rewrite the historical literature, exclude inconvenient elite-athlete data, and lean on a 10 g/h carbohydrate dose that itself proves his own original LCHF claim wrong, but the stoichiometry does not negotiate.
Carbs win.
The message is the same as before. Carbohydrates win, and fat is not an efficient fuel for high-performance exercise. Yes, you can run a long way at a slow pace on a high-fat diet. For the 0.00001% of people whose entire sporting goal is to finish a 200 km ultramarathon at 5 km/h, the LCHF approach can technically work. For the rest of endurance sport, where pace at threshold determines outcomes and where the difference between fuels shows up in seconds per kilometre, the answer has not changed in a hundred years. Carbohydrate oxidation produces more ATP per litre of oxygen consumed. More ATP per litre of oxygen means more power at the same VO2. More power at the same VO2 means you finish ahead of the keto guy. Noakes can rewrite the historical literature, exclude inconvenient elite-athlete data, and lean on a 10 g/h carbohydrate dose that itself proves his own original LCHF claim wrong, but the stoichiometry does not negotiate.
Carbs win.
Noakes' Evidence 5 and Table 8, where he reports that in 105 of 125 studies (88 percent) in which blood glucose fell in the control group, CHO ingestion improved performance. The 88 percent figure is roughly defensible from Noakes' own tables (my replication landed at 87 percent, and his Table 8 arithmetic actually gives 84 percent, not the 88 he prints), but three caveats undermine how he uses it.
First, the number is correlational, not causal. Long exercise bouts that drop blood glucose are also bouts where CHO helps for many other reasons (gut training, CNS effects, muscle substrate), so the same data fits half a dozen mechanisms. Second, the denominator is cherry-picked. Flip the question and ask how often BG fell in placebo across all studies where CHO improved performance, and his own data gives 53 percent, not 88. Third, the entire calculation rests on Noakes and his co-authors classifying 300-plus studies themselves, with no independent audit and several ambiguous entries that another reviewer would code differently. The pattern is real, but the "88 percent" is suggestive evidence dressed up as a smoking gun.
First, the number is correlational, not causal. Long exercise bouts that drop blood glucose are also bouts where CHO helps for many other reasons (gut training, CNS effects, muscle substrate), so the same data fits half a dozen mechanisms. Second, the denominator is cherry-picked. Flip the question and ask how often BG fell in placebo across all studies where CHO improved performance, and his own data gives 53 percent, not 88. Third, the entire calculation rests on Noakes and his co-authors classifying 300-plus studies themselves, with no independent audit and several ambiguous entries that another reviewer would code differently. The pattern is real, but the "88 percent" is suggestive evidence dressed up as a smoking gun.
It took me a long friggin time to read this , and it's a painful read at points, but at the end it's underwhelming as a manifesto for low-carb and exercise performance!academic.oup.com/edrv/article/4…
• • •
Missing some Tweet in this thread? You can try to
force a refresh
