Performance · Evidence literacy

Does Higher BHB Hurt Athletic Performance?

A reasonable worry, and one that gets answered with a single number too often. The human trials on the Veech Ketone Ester have produced faster, unchanged and slower results at overlapping blood BHB levels — so the level alone is not the thing to read. Here is the record, sorted by the context in which each result was measured.

See the record Why there is no universal BHB target

Short answer

Not by itself. Human performance studies do not establish a universal blood-BHB cutoff above which athletic performance suddenly worsens. Performance has been positive, unchanged and negative at overlapping BHB levels under different exercise protocols.

This is not a page arguing that higher BHB is harmless, or that it helps. Acute short time-trial protocols include small decrements. Acute endurance protocols include both a reported gain and null findings. Repeated-use protocols include positive training-adaptation results as well as a null 5K result. The studies differ in dose, timing, fueling, population, event and co-ingested compounds, so BHB concentration alone is not a validated performance predictor.

Why a correlation is not enough

What it would take to show a performance cutoff — and what exists instead.

“Performance got worse when BHB was high” and “high BHB makes performance worse” are different claims. The second needs a kind of study the field has not run.

To validate a cutoff you would need

A convincing threshold test would deliberately create several BHB levels within the same molecule, exercise protocol and population while holding the rest of the protocol as constant as possible, then measure performance at each level and replicate the finding. The current literature does not provide that kind of validated dose-response threshold experiment.

No study in the indexed Veech Ketone Ester record has that design.

What the record actually contains

Separate trials that each chose one dose and one timing, measured BHB as it happened to land, and tested one event — a 20-minute time trial, a 10-km run, a three-hour race simulation, a three-week training block. Their BHB levels overlap; almost nothing else does. Lining them up by BHB and reading a trend across them is not an experiment. It is a scatter of unmatched protocols.

That is why the table below is grouped by context, not sorted by concentration.

The record

What happened across different Veech Ketone Ester performance protocols?

The pattern changes with how and when VKE was used — from repeated recovery dosing to acute endurance and short high-intensity efforts. These trials are selected to show why blood BHB concentration cannot be read as a cross-study dose-response curve. They are not presented as the complete VKE exercise literature. Same exact molecule in every row; read down each group before reading across groups.

Study	Dose · timing	Approx. BHB	Exercise protocol	Outcome	Direction
Repeated use · training and recovery protocolsDosed after sessions or daily for weeks; performance measured at the end of the protocol.
Poffé et al. 201918 physically active young men · not regular cyclists · double-blind parallel	25 g after each session + 25 g before sleep · 3 weeks of overload	≈ 2.6 mM30 min after each post-exercise dose; fasted mornings ~0.35 mM	28 supervised sessions; 2-h endurance test in week 3	Sustainable training load; final 30-min power	↑About 15% higher than control═ 30-min time trial not significantly different between groups
Robberechts et al. 202628 trained men · supervised training	25 g after each session + 25 g before sleep · 8 weeks	Raisedafter each dose; level not carried in this index	8 weeks of supervised cycling training; 30-min time trial at the end	30-min TT power; VO₂peak; muscle adaptations	↑302 vs 291 W — 4% higher than control↑ relative VO₂peak +12% vs +6%; citrate synthase 14% higher
Prins et al. 202618 recreational runners · single-blind parallel	30 g three times daily · 31 days	≈ 3 mMpost-dose	Usual training; 5-km time trial at the end	5-km time	═Unchanged↑ selected executive-function measures improved
Acute · endurance eventsSingle or staged dose; performance test after prolonged exercise.
Cox et al. 20168 endurance athletes in the performance arm · fasted protocol	573 mg/kg with carbohydrate · pre-exercise	Raisedlevel not recorded in this index	1 h submaximal cycling, then 30-min time trial for distance	Distance covered	↑About 2% furtherPoffé et al. later noted oral carbohydrate intake was about 40% lower in the ester condition than in control, and that testing was fasted
Evans et al. 20198 endurance-trained runners · double-blind crossover	573 mg/kg with an 8% carbohydrate-electrolyte solution · pre-exercise	≈ 1.0–1.3 mMduring exercise	1 h at ~65% VO₂max, immediately followed by a 10-km treadmill time trial	10-km time; cognitive tests	═2402 vs 2422 s, P = .483═ cognition no difference
Poffé et al. 202012 highly trained male cyclists · randomized crossover	25 + 20 + 20 g before and early in the event · 60 g/h carbohydrate	≈ 3 mM earlyrange 2.6–5.2; had fallen by the performance tests	3-h intermittent cycling, then 15-min time trial and sprint	Time-trial power; sprint time-to-exhaustion; glycogen breakdown	═273 vs 272 W; 59 vs 58 s; glycogen similar
Poffé et al. 2021 (hypoxia)14 highly trained male cyclists · with and without bicarbonate	75 g across the protocol	≈ 3 mMmaintained	Prolonged cycling in progressive normobaric hypoxia, then 15-min time trial and sprint	Time-trial and sprint performance; blood and muscle oxygenation	═Performance unchanged↑ oxygenation improved
Evans & Egan 201811 male team-sport athletes	750 mg/kg with a carbohydrate-electrolyte drink · pre-exercise	Raisedlevel not recorded in this index	Exhausting intermittent running; shuttle-run time to exhaustion; 15-m sprints	Time to exhaustion; sprint performance	═No significant improvement↑ executive-function decline attenuated
Acute · short, high-intensity effortsSingle dose before or during warm-up; performance test under an hour.
McCarthy et al. 202323 trained cyclists · triple-blind crossover	0.35 g/kg · 30 min pre-exercise	≈ 2.0 mMpre-exercise	15-min warm-up, then 20-min cycling time trial	Mean power	↓2.4% lower than placebo
Poffé et al. 202112 well-trained male cyclists · double-blind crossover · with or without bicarbonate	50 g in two aliquots during a 60-min warm-up · 60 g/h carbohydrate	≈ 3.5 mM · ≈ 4.5 mM with bicarbonatethroughout the time trial	30-min cycling time trial, then all-out sprint	Mean power; sprint time-to-exhaustion	↓About 1.5% lower in the ester conditions, with or without bicarbonate═ sprint unchanged by the ester

Cross-study BHB values are descriptive only. Different doses, sampling times, exercise protocols and populations prevent treating these rows as one dose-response experiment. “Raised — level not recorded in this index” means the trial reported elevated blood ketones but the value is not carried in our registry; the paper has it.

A Performance-related measures are not all the same endpoint. In eleven male endurance runners, 750 mg/kg of the ester without carbohydrate improved running economy by about 4.1% versus carbohydrate alone (blood BHB about 2.1 mM), while time to exhaustion was not significantly different — and there was no running-economy benefit when the ester was co-ingested with carbohydrate (about 1.8 mM). Within one study: an efficiency measure moved, a direct exhaustion endpoint did not, and the carbohydrate context changed the result. Improved running economy is not a proven race-performance gain. Brady & Egan · Med Sci Sports Exerc 2024

B Same molecule, same event family, different co-ingestion. In another prolonged cycling protocol from the same group — nine well-trained cyclists, 65 g of the ester with 60 g/h carbohydrate in every arm — the ester alone did not improve the final 15-minute time trial, while ester plus bicarbonate produced about 5% higher mean power than control. Blood BHB was about 2–3 mM early in the event and had fallen by the time trial. Protocol components can change the result; this is not ester-alone efficacy, and it does not mean bicarbonate universally unlocks the ester. Poffé et al. · Med Sci Sports Exerc 2021 (bicarbonate combination)

What the record shows

Faster, unchanged and slower — at overlapping BHB

Two acute short time-trial protocols produced small decrements. Acute endurance studies include one reported ~2% gain and several null findings. Repeated-use studies include positive training-adaptation results as well as a 31-day 5K null. Same molecule throughout.

What it does not show

A number above which performance drops

The lowest reading in the table sits in a null trial; the highest sits in a negative one; positive results sit in between. These cross-study results do not reveal a validated cutoff separating positive from null or negative performance outcomes.

What else differed

Six things that changed between those trials besides the BHB level.

Each belongs in the interpretation. None of them is captured by the BHB number alone.

1Dose and timing.From 0.35 g/kg thirty minutes out, to 50 g across a warm-up, to 25 g after training for weeks. The two negative short time-trial results followed acute pre-effort dosing. Other acute endurance protocols produced either a reported gain or null findings, while repeated post-exercise protocols also include positive results. Timing and use case are part of the evidence.

2Event duration and intensity.Both decrements came in all-out efforts of 20–30 minutes. Acute endurance results were mixed: Cox 2016 reported a small gain, while several later running and cycling protocols reported no significant performance change. Event duration and intensity are part of the protocol, not a BHB threshold.

3Carbohydrate availability.Cox 2016 was a fasted protocol in which later authors note the control condition received more carbohydrate. Poffé 2020 and 2021 supplied 60 g/h to every condition. In Poffé 2019 the ester group spontaneously ate more carbohydrate by week three. Fuelling differed across protocols and belongs in the interpretation. These studies do not isolate carbohydrate availability as the sole cause of the differing outcomes.

4GI tolerance.Poffé 2021 recorded slightly higher GI distress with the ester — dizziness, bloating, nausea — at 50 g before an all-out effort, and the authors say they cannot exclude it as a contributor. Tolerability rides along with dose.

5Population and training status.Trained and highly trained cyclists, recreational runners, team-sport athletes, and physically active men who were not regular cyclists. Results do not automatically generalize from one to another.

6When BHB was measured.In Poffé 2020 the ester raised BHB to about 3 mM early, but the level had fallen by the time the performance tests ran. A trial's headline BHB is not always the BHB that was present during the effort.

Proposed thresholds and proposed causes

Hypotheses from particular protocols — not validated cutoffs.

On thresholds

Some exercise papers have proposed favorable BHB ranges under specific conditions. Those proposals are hypotheses derived from particular protocols, not validated universal cutoffs. Poffé et al. 2021 discussed a suggested lower bound above roughly 1–2 mM and an upper bound around 3 mM — and in the same discussion noted that an earlier benefit had occurred when BHB had fallen to about 0.5 mM, while in its own time trial higher BHB correlated with greater impairment. Those are observations from two protocols, reported here as the paper reported them.

A range proposed from one exercise context is not a performance cutoff for every other.

On causes

In Poffé 2021, correcting the acid-base disturbance did not remove the performance decrement, so acidosis alone did not explain the result. The paper discusses several possible mechanisms, but the trial does not establish one universal cause. This page does not say high BHB itself caused the negative findings, that acidosis explains them, that bicarbonate fixes them, or that any single mechanism is settled.

Where a paper says the mechanism is unresolved, so does this page.

Timing and use case are part of the evidence. A repeated post-exercise protocol does not show that a pre-race dose helps a short time trial — and an acute 20-minute decrement does not show that post-exercise dosing impairs training. Read each result in its own context.

A BHB level ≠ a performance prediction

Why there is no universal target Read a performance trial in ten checks

Study-reading checklist

Five questions before you read a BHB level as a performance verdict.

1Acute or repeated?A single pre-effort dose and a weeks-long post-exercise protocol are different uses. Do not read one as the other.

2How long and how hard?A 20-minute all-out trial and a three-hour event are different physiological questions.

3When was BHB measured?Before the effort, during it, or after? Was ketosis still present when performance was tested?

4What was fuelling?Fasted or fed; carbohydrate matched between conditions or not. It changes what the ester was doing.

5Was it the level — or the protocol?If the paper cannot separate them, neither can you. Report the result with its protocol attached, and do not assign a cause the paper did not.

The absence of a validated threshold does not mean the ester always helps. It means the BHB number alone did not predict the direction of the result — so read the trial, not the meter.

Every indexed VKE study, with its result

From evidence to the bottle

For athletes, follow the protocol — not just the BHB number.

KetoneAid uses Veech Ketone Ester, the molecule tested in the VKE performance studies on this page. Those trials include positive, null and negative findings under different protocols.

That is why the product choice, timing and use case should stay attached to the actual evidence instead of a universal blood-ketone target.

Shop KetoneAid See the VKE study index

50% VKEKe4Most concentrated · best-sellingShop Ke4 →

25% VKEKe2More diluted · balance of taste and valueShop Ke2 →

8% VKEKetone ShotReady to drink · easiest way to try VKEShop Ketone Shot →

Go deeper

From the level to the trial.

The number

Back to Ketone Science Explore Veech Ketone Ester science

References cited on this page

McCarthy DG, Bone J, Fong M, et al. Acute Ketone Monoester Supplementation Impairs 20-min Time-Trial Performance in Trained Cyclists: A Randomized, Crossover Trial. Int J Sport Nutr Exerc Metab. 2023;33(4):181–188. PubMed ↗ · Study index
Poffé C, Wyns F, Ramaekers M, Hespel P. Exogenous Ketosis Impairs 30-min Time-Trial Performance Independent of Bicarbonate Supplementation. Med Sci Sports Exerc. 2021;53(5):1068–1078. PubMed ↗ · Study indexn = 12 analysed per the full text (the abstract states 14). Source for the proposed-range discussion, the mechanism caveat and the GI note.
Cox PJ, Kirk T, Ashmore T, et al. Nutritional Ketosis Alters Fuel Preference and Thereby Endurance Performance in Athletes. Cell Metab. 2016;24(2):256–268. PubMed ↗ · Study indexThe observation that carbohydrate intake was about 40% lower in the ester condition, and that testing was fasted, is as discussed in Poffé et al. 2021.
Evans M, McSwiney FT, Brady AJ, Egan B. No Benefit of Ingestion of a Ketone Monoester Supplement on 10-km Running Performance. Med Sci Sports Exerc. 2019;51(12):2506–2515. PubMed ↗n = 8; 573 mg/kg with an 8% carbohydrate-electrolyte solution; plasma BHB ~1.0–1.3 mM during exercise; 10-km TT 2402 vs 2422 s, P = .483.
Poffé C, Ramaekers M, Bogaerts S, Hespel P. Exogenous ketosis impacts neither performance nor muscle glycogen breakdown in prolonged endurance exercise. J Appl Physiol. 2020;128(6):1643–1653. PubMed ↗
Evans M, Egan B. Intermittent Running and Cognitive Performance after Ketone Ester Ingestion. Med Sci Sports Exerc. 2018;50(11):2330–2338. PubMed ↗ · Study index
Poffé C, Robberechts R, Podlogar T, et al. Exogenous ketosis increases blood and muscle oxygenation but not performance during exercise in hypoxia. Am J Physiol Regul Integr Comp Physiol. 2021;321:R844–R857. PubMed ↗ · Study index
Poffé C, Ramaekers M, Van Thienen R, Hespel P. Ketone ester supplementation blunts overreaching symptoms during endurance training overload. J Physiol. 2019;597(12):3009–3027. PubMed ↗ · Study index
Robberechts R, Bekhuis Y, Stalmans M, et al. Post-exercise ketone supplementation improves endurance performance and mitochondrial adaptations during an 8-week endurance training intervention. J Physiol. 2026. PubMed ↗ · Study index
Prins PJ, Buga A, Storoschuk K, et al. The Effects of 31-Day Exogenous Ketone Consumption on Running Performance, Cognitive Function, Metabolism, Body Composition, Hemodynamics, and Mood in Recreational Runners: A Randomized-Control Trial. J Am Nutr Assoc. 2026;45(6):545–560. PubMed ↗ · Study index
Brady AJ, Egan B. Acute Ingestion of a Ketone Monoester without Co-ingestion of Carbohydrate Improves Running Economy in Male Endurance Runners. Med Sci Sports Exerc. 2024;56(1):134–142. PubMed ↗ · Study index
Poffé C, Ramaekers M, Bogaerts S, Hespel P. Bicarbonate Unlocks the Ergogenic Action of Ketone Monoester Intake in Endurance Exercise. Med Sci Sports Exerc. 2021;53(2):431–441. PubMed ↗ · Study indexNot the same study as the 30-minute time-trial paper above. Four conditions; 65 g ester; 300 mg/kg bicarbonate; 60 g/h carbohydrate in every arm.