A 2026 lab trial found that cyclists burned less of their race-day carbs when they added a ketone drink. Exogenous carbohydrate oxidation (the carbs you swallow and then burn) fell from 1.50 to 1.35 grams per minute. But the riders didn’t last any shorter, and carbs still beat placebo by a mile.
Translation: the headline is real, the sample is tiny, and your gels are safe.
What the 2026 ketone esters trial measured
Researchers led by Martyn tested 8 trained men (average VO2max 66) in a lab. They carb-loaded for 36 hours first (12 g per kg of body weight). Then the men rode for 3 hours at 95% of lactate threshold. That’s the effort where lactate starts to pile up fast.
Each rider took 120 g of carbs per hour. Some also took 75 g of a ketone monoester, a drink that raises ketone levels in your blood within minutes. Others got a placebo with no carbs.
Here’s what the tracer-labeled carbs showed.
Oxidation efficiency also slipped, from 75% to 67%. Efficiency here means how much of the carbs you swallowed got burned. Blood glucose was lower too: 4.9 mmol/L with carbs, 4.4 with carbs plus ester, and 3.7 with placebo.
In short: the ester seemed to change how the body handled the carbs. It didn’t show up as a faster or slower rider.
What it did not show
Time to exhaustion at 150% of threshold was 349 seconds with carbs and 319 with carbs plus ester. That gap wasn’t meaningful. Placebo managed only 75 seconds.
Think of your body as a car with two fuel lines. Carbs and ketones both flow to the engine. The ester may let the engine draw from the ketone line a bit, so less carb gets used. That’s a plausible story, not a proven one.
This was one lab, 8 men, and 120 g/h. Most athletes eat 60 to 90 g/h. We can’t say what happens at those doses. A Maastricht trial on the same question was due to finish in 2025. We couldn’t find published results.
Small trial. Big caution.
Why the hype started, and why it faded
A 2016 study in elite cyclists reported a roughly 2% time-trial gain with a ketone ester plus carbs. That paper fueled most of the buzz. It used tightly controlled conditions that look little like a real race.
Later work didn’t match it. A 2020 trial by Poffe tested 12 trained cyclists and triathletes. They rode 3 hours with 60 g/h of carbs and took 65 g of ester in 3 doses. Then they did a 15-minute time trial.
| Study | Form tested | Who | Main result |
|---|---|---|---|
| Cox 2016 | Monoester + carbs | Small elite group | Roughly 2% time-trial gain |
| Leckey 2017 | Diester | 10 elite male cyclists | Worse 31 km time trial, poor tolerability |
| Poffe 2020 | Monoester, 65 g | 12 trained athletes, 60 g/h carbs | No gain: 273 W vs 272 W |
| Martyn 2026 | Monoester, 75 g | 8 trained men, 120 g/h carbs | Less carb burned, capacity unchanged |
The Leckey diester is a different compound from the monoester in the other three. Don’t lump them together.
Poffe also checked muscle glycogen, your stored carbs. The ester didn’t spare it. Net breakdown over 3 hours was 78 mmol/kg wet weight with ester and 60 with the control (P=0.08). The trend ran toward more breakdown, not less.
Early attempts to pool the small trials point the same way, though none we found is a peer-reviewed meta-analysis yet. Most performance outcomes come out flat, and results vary a lot from study to study. That’s not proof of zero effect. It does mean nobody should promise you a faster split.
Side effects worth knowing
Poffe’s riders saw blood pH drop from 7.42 to 7.36. Bicarbonate, your blood’s built-in acid buffer, fell from 26.0 to 21.6 mM. That’s a mild acid load, like adding a little lemon to a glass of water.
Upper-abdominal discomfort scored 4 with ester versus 2 for control, out of 32 (P=0.03). Total gut discomfort was about the same, 13 versus 12 out of 96. So the ester didn’t wreck every stomach. It did nudge the upper gut.
Ghrelin, the hunger hormone, was about 17% lower. Hunger after the ride was lower too, 3.6 versus 4.9. Urine output was also lower, 300 versus 522 mL.
Translation: you might not feel like eating. In a long race, that’s a problem. If you skip gels, you lose the one fuel with proven benefits.
Runners and triathletes should read this the same way cyclists do. The labs used bikes, but the fuel logic is the same on foot. Gut tolerance is often the first limit on a marathon or a 70.3 run leg.
What to do on race day
Keep your plan boring. Eat the carb dose you’ve practiced. Drink what you’ve practiced. The research gives you no reason to swap gels for a ketone drink.
If you’re curious, run a trial well before your race. Pick a long session that looks like race day. Take your normal carbs. Add the ester only on one of two matching days.
Write down gut symptoms, appetite, and effort. Note the weather and your sleep. Those things move your numbers as much as any supplement.
Then decide with a cool head. If you felt flat or queasy, drop it. If nothing changed, you’ve learned that cheaply.
A fair way to test it yourself
Say Dana, a 41-year-old half-marathoner, wants to try an ester on her long runs. Here’s an example, with made-up numbers for illustration. She does two similar 2-hour runs at the same easy effort, 3 weeks apart. Her fueling is 75 g/h of carbs both times.
Her drift ratio is the change in heart-rate-to-pace efficiency between the first and second half of a workout. Pros call it Pa:Hr. In run one it’s 4.8%. In run two, with the ester, it’s 5.1%. Her notes say she felt bloated in the second run.
That’s no real gap. One pair of runs can’t prove anything, since sleep, heat, and mood all move the number. Dana would need several repeats before she trusts it. If the answer stays “no difference,” she saves her money and her stomach.
AthleteOS can’t see which fuel your muscles burn. No watch can. It tracks your aerobic decoupling and per-session outcomes. Pair that with a simple fueling log and you can compare race-day plans against your own history. You can try AthleteOS and see your own numbers.
What to do with this
Keep the carbs. Both carb conditions beat placebo in 2026. Nothing here replaces them.
Expect a small effect, if any. The only gains seen so far are small and haven’t been repeated. The pattern across studies is mostly null.
Fix the basics first. A solid Zone 2 base and heat prep, like heat acclimation, have stronger evidence than any ester.
If your carbs don’t work, a ketone drink won’t rescue you.