In response to NYT “Can Ketones Rev Up Our Workouts?”

It's an exciting time for the field of human performance, physiology, and metabolism. There's tremendous interest from academics, elite athletes, and keen observers on whether ketones can be utilized as a nutritional substrate to improve performance.

Usually ketones are only present as a fuel source when the body is in a low carbohydrate state (ketogenic diet or fasted). The theory behind why exogenous ketones would improve performance is that one can have full carbohydrate reserves and ketones at the same time, thereby providing fuel reserves not seen in physiology normally. There's a rapidly growing body of data in this area, and the role of ketones as a metabolic substrate and signalling metabolite is an active area of research, so some of the confusion / hype / exaggeration / skepticism is understandable. However, we have sufficient basic scientific understanding to have a clear discussion without imprecision and conflation. My goal in this response is to clarify and add perspective to the New York Times piece "Can Ketones Rev Up Our Workouts?".

The reason why people take exogenous ketones is to elevate blood ketone levels, mainly d-beta-hydroxybutyrate (BHB). The key mistake is conflating all "ketone supplements" and "exogenous ketones." There are different types of ketone precursors (e.g. coconut oils, MCT oils, etc). There are different types of ketone salts (e.g. sodium BHB salts, potassium BHB salts, etc). There are different types of ketone esters (D-BHB monoester, acetoacetate diester, etc). All these different compounds have different pharmacokinetics and characteristics; they raise blood BHB (and other metabolites) to different levels and each are being studied independently for potential in different applications. This is critical to understand. To assume they are all the same is as serious an error as assuming that all computers (desktop, tablet, smartphone, quantum, etc.) are the same, all social networks (Facebook, LinkedIn, Twitter, Snapchat, etc.) are the same, or all types of engines (steam, diesel, electric, etc.) are the same.

Many lay commentators point to Dr. Louise Burke and her study on cyclists supplementing with acetoacetate diester as proof that "ketones don't work." It's not clear to me why this paper was a focal point, because the Burke study was not the first study showing worse performance results when athletes used an exogenous ketone,. Again, as this is novel area of inquiry, it is understandable that there is confusion about the multiple types of exogenous ketones and the different streams of work in this field.

Let's first discuss the ketone supplement Burke investigated.

Burke studied acetoacetate diester, a type of ketone ester with no published pharmacokinetic data in humans and mainly investigated for its potential to manage oxygen toxicity in animals. In Burke's study using acetoacetate diester, the peak levels of BHB in athletes were 1.1 mM. Several previous athletic performance studies looking at ketone salts elevated BHB levels between 0.3 - 1.2 mM. The reduction in performance seen with acetoacetate diester is consistent with the negative results reported with athletes using ketone salts. Thus, one should not necessarily be surprised by Burke’s negative time trial result given the precedent set with BHB salts. They both exogenously raised BHB levels only to the ~1 mM range, and both failed to observe performance improvements.

The University of Oxford studied cyclists using d-BHB monoester (exclusively in H.V.M.N. Ketone) . In these studies, athletes hit peak levels of BHB of 3 mM and achieved a statistically significant 2+% improvement in time trial performance. The difference in available BHB provided by the different exogenous ketone supplement may account for the polar conclusions between negative results with acetoacetate diester or ketone salts and Oxford’s positive result with d-BHB monoester.

To underline this point, this is like comparing high-octane fuel for Formula 1 racecars to diesel fuel for long-haul trucks. Yes, they're both petroleum-based fuels. But we understand combustion, and we expect different performance characteristics. Burke studied a compound designed for oxygen toxicity. Oxford studied a compound designed for enhanced soldier performance. Yes, they're both "ketone supplements." But we understand physiology and metabolism, and we expect different performance characteristics.

Now, let's analyze the NYT discussion of the study designs.

"Last year, a much-discussed study conducted by researchers at Oxford University found that a proprietary type of ketone supplement [D-BHB monoester] improved cycling performance in a group of trained riders. But that study looked at relatively moderate exercise, not the kind of intense exertion required during competition."

The study design was 1 hour pre-fatigue at 75% max wattage, and followed immediately with an all-out, full intensity, 30 minute time trial. Working closely and directly with the authors of the Oxford study, I know it is factually incorrect to characterize the Oxford study's time trial as "relatively moderate."

"So for the new study, which was published last month in Frontiers in Physiology, scientists at the Australian Institute of Sport and other institutions decided to create a more real-world test of ketone supplementation... The experiment the scientists concocted was simple. About an hour before a training session, they gave each rider a drink that contained either ketones or a placebo."

According to the primary source, Burke’s protocol was to give each cyclist acetoacetate diester or placebo twice with 200 ml of diet cola, once 50 minutes before and once 30 minutes before the time trial. I'm dumbfounded that an experimental design claiming to be "more real-world" would require elite athletes to drink more than a full can's (which is 355 ml) worth of fizzy, carbonated soda minutes before a time trial. Moreover, it was reported that the tested cyclists had major palatability issues with acetoacetate diester and almost all of them experienced intense GI distress and some vomited. With no pharmacokinetics data, we have no idea whether the cyclists were under or overdosed for performance benefit let alone dosed safely! In fact, the authors themselves admit that their results may have been driven primarily by GI distress itself as opposed to measuring any effect on the metabolic efficacy of acetoacetate diester.

I recommend that Burke omit the diet cola, investigate the palatability and pharmacokinetics of acetoacetate diester, and/or work with a palatable exogenous ketone with more comprehensive, published human data to better mimic real-world conditions in future study.

I can agree with one of the parting thoughts of the NYT piece: “Everyone wants simple answers,” [Burke] says, but the body’s mechanisms are too complex to allow for that, especially during sports." Yes, human performance is indeed complicated, so let’s make the conversation around it as clear as possible. Let’s compare apples with apples, oranges with oranges.

UPDATE: Read this peer-reviewed response to Dr Burke's study, written by ketone experts, H.V.M.N.'s Dr Brianna Stubbs and USF's Dr Dominic D'Agostino. The response was published in March 2018 in the journal Frontiers in Physiology.

Disclaimer: As CEO of H.V.M.N., the company commercializing and marketing D-BHB monoester as H.V.M.N. Ketone, the world's first ketone ester drink, I have a vested interest in this topic.