Get access to everything we publish when you sign up for Outside+.
In an intriguing new report, Louise Burke and colleagues from the Australian Institute of Sport have replicated and extended an important previous study. They have shown, for the second time, that a Low-Carb High-Fat Ketogenic-type diet significantly reduces oxygen economy and race performance (when compared to a High-Carb diet) among world class race walkers, even as it more than doubles fat-burning among the LC walkers. Burke is widely considered a world-leading endurance nutritionist, and last fall ran 3:38:56 (at age 60) in the NYC Marathon.
In late 2016, Burke and colleagues published a first study investigating how polar opposite diets might affect performance among 10K race walkers. That paper showed a 6.6% improvement among the High-Carb (HC) walkers, while the Low Carb (LC) walkers slowed by 1.6%. Subjects were tested before and after three weeks of intensified training.
Due to the interest in LC eating and endurance performance, the paper became one of the top-5 ranked in popularity among science papers. Many Keto and LC fans criticized the paper on social media, often citing their own success stories or stories they had heard from others. Burke refers to this as “anecdata” as opposed to data from evidence-based research.
Nevertheless, she evaluated the criticisms to determine how she might incorporate several in an experimental do-over. This is termed “replication” in research circles, and has been the focus of much recent discussion, because the results of many trials have not been replicated when repeated. The effect, some say, has been a “crisis of confidence” regarding scientific papers.
The New Paper:
Burke followed the same basic approach of her first study, with a couple of new wrinkles. For one, she included a half-dozen women walkers. She also added a new test race — a 20K — to the end of the trial. For this, the LC walkers were allowed two weeks to taper and switch to a “carb rich” diet.
Some Keto fans had previously argued that this was one way LC athletes might super-compensate and get faster. They suggested that the LC diet might behave like altitude training, causing an immediate reduction in performance but then achieving physiological adaptations that would produce enhanced performance with a carb-fortified taper.
Twenty-six world class race walkers attended a 3.5 week “intensified training” camp at the Australian Institute of Sport. All followed identical training regiments (twice-a-day workouts) and consumed the same number of calories daily at Institute-provided meals. Based on choice, eight walkers followed a HC diet and 10 chose to try a LC diet. (All had previously used a typical carb-centric diet.) The HC diet consisted of 60 to 65% carbs; the LC diet was 75 to 80% fats with fewer than 50 grams a day of carb calories (3.7% of total calories).
The diet assignments were not random or “blinded” as is customary in “gold standard” experiments. Such an approach would have been impossible, given the substantial taste differences between the two diets. Also, subjects with strong biases one way or another would have thought they were getting the “good” diet or the “bad” diet, and this would have influenced the results. Instead, subjects selected the diet they were most interested in, HC or LC, and presumably had equal confidence in their choice.
Subjects competed in a 10K race-walk competition on a 400-meter track at the beginning of the experiment, and again after the 3 weeks of hard training and strict diet adherence. Two weeks later, they entered an Aussie 20K road walk competition. During this de-adapt and taper time, all walkers, including the previously LC subjects, followed similar “carb rich” diets. (See the Quick Review below.)
Two weeks later, The Journal of Physiology published a free, 37-page Louise Burke historical review of ketogenic and low-carb studies involving endurance athletes. This paper offers an excellent summary of the field, including work by prominent low-carb scientists.
The HC walkers improved their 10K time by 2:14 (4.8%) after three weeks of intense training. The LC walkers slowed by 1:26 (1.6%). These results closely replicated what Burke had reported earlier, as did the findings that both groups improved their VO2 max but only the HC group achieved an increase in their relative economy (oxygen use for a given speed). Strikingly, in both studies, the HC group showed an increase in economy at the speeds typically needed to race in the events on the Olympic program — they were now able to walk at these speeds for a lower percentage of their maximal aerobic capacity.
Meanwhile, the LC group showed a reduction in economy and had to increase their oxygen use by 6- to 8-percent to walk at the same speeds. This was explained by the biochemistry of burning exercise fuels, where carbohydrate produces more ATP fuel units than fat for any given amount of oxygen that can be delivered to the muscle. The authors pointed out that this 6- to 8-percent is double the economy gain of new shoes that have caused great controversy for their 3- to 4-percent enhancement.
The poor economy of the LC walkers was surprising because they had lost 6 pounds in the three weeks of hard training. This might have been expected to improve economy, but didn’t, indicating the loss was probably connected to their high fat-burning. The HC walkers lost 2 pounds during the three weeks. This had little effect on their economy.
The LC walkers also reported a greater relative perceived exertion when exercising at the same effort as the HC walkers, and a higher heart rate. It appeared they had to work harder at any given pace.
In the 20K road walk, the HC walkers performed at about the same level as in the second 10K. The LC subjects rebounded from their poor second 10K performance, and returned to the level of their first 10K. In other words, after two weeks of carb-rich eating, the LC walkers recovered the performance they had lost in 3 weeks of LC eating. They didn’t, however, attain a new level of super-compensated improvement.
(The research project included a third group that followed a Polarized Carb diet-and-exercise program. They generally consumed a HC diet, but sometimes restricted carbs before a morning long run, or added them to a difficult high-intensity workout. This approach has gained followers in recent years, with some research supporting it. In the Australian studies, the PC walkers regularly finished behind the HC walkers but ahead of the LC walkers.)
Apparently, the HC walkers got faster due to the hard training and a diet that could sustain such training. The LC walkers did the same training, but reported significant difficulties in the first two weeks. In the third week, they felt better. While their fat-burning improved significantly — it more than doubled — their economy deteriorated. That is, they got less speed/power from the oxygen they burned. This difference in oxygen burning efficiency of carbs vs fats has been known for a century or so, and replicated often.
As Burke et al write: “The increase in fat oxidation associated with keto-adaptation reduces the speed achieved at any given percentage of maximum aerobic capacity.”
What Remains Unknown:
LC adherents often claim that subjects in experiments such as this one don’t have enough time to fully adapt to the LC diet. Burke and her co-workers agree that longer, well-controlled studies of LC adaptation need to be undertaken. However, they point out that subjects in their prior study achieved fat burning rates that are “the highest ever recorded in the literature,” including other studies that lasted more than six months.
Of course, 10K and 20K race walkers aren’t marathoners. (The walkers take 42 minutes to 1:27 to complete the two distances.) In particular, given the duration of their events, they aren’t limited by the size of their muscle glycogen stores for optimal performance. But they are also not that different. They tend to race at an oxygen-consumption level close to that of elite marathon runners, and share the need to obtain the highest speed for their muscle oxygen delivery.
At any rate, a diet that lowers performance economy according to classic rules of nutrition and exercise physiology (i.e., fats don’t burn as efficiently as carbs) is likely to have similar results for all endurance exercise where high-intensity workloads are the hallmark of success. Burke and colleagues acknowledge this could change in ultra-marathon races lasting far longer where intensity is less important, and can be supported by the less economical but greater pool of body fat stores.
Most endurance athletes seem to train and race best on healthy high-carb diets. There will always be exceptions to this rule, but it’s a good general guideline. Quoting the paper: “Adaptation to a ketogenic, low-carb, high-fat diet reduced exercise economy and impaired performance of a real-life endurance event in elite athletes.”
Also Worth Considering:
Little is known about the long-term effects of a low-carb diet on immunity, the bones, the microbiome, heart health, and more. Several months ago, Burke et al reported that the LC walkers in their research exhibited changes in blood markers of bone turnover that suggest poor bone health.
Feel free to experiment if you want. There’s no indication that LC diets are dangerous in the short term. But proceed with caution, and don’t hesitate to turn back if you don’t like what you experience. (You’ll have to expect a rough, two- to three-week adaptation period — nearly everyone encounters this — so try to hang in long enough.) And remember: Long-term effects are unknown.
Quick Review of the Aussie Experiment:
- Twenty-six world class race walkers agreed to participate; eight women and 18 men.
- They raced a track 10K, and then did hard training for three weeks.
- Eight followed a HC diet, 10 a LC diet, and eight a PC diet.
- After three weeks, they raced another 10K. The HC walkers improved by an average of 2:14, while the LC walkers slowed by 1:26.
- All athletes de-adapted and tapered two weeks before a 20K road competition. During this time, they consumed carb-rich diets.
- In the 20K race, the HC walkers maintained the level they had reached in the second 10K. Depending on the way the results were analysed, the LC walkers improved back to their first-10K level or slightly more, but did not show any super-compensation effect.
- Overall, the most consistent improvements in performance from research studies like this come from what is known as “the training camp effect.” When elite athletes come together and push each other in a dedicated training block, they are likely to achieve new levels of performance. These outcomes are best supported by dietary practices that provide carbohydrates to meet the fuel needs of training, particularly during high-quality workouts.