Exogenous ketone bodies and endurance performance
Physiology review (Evans 2017) + the overreaching RCT (Poffé 2019)
Key Findings (evidence summary)
Ketone bodies (acetoacetate, acetone, β-hydroxybutyrate) are produced when carbohydrate availability is low and serve as an alternative fuel for brain, heart and skeletal muscle
They are oxidised during exercise, markedly elevated in post-exercise recovery, and used better by exercise-trained muscle
Metabolic actions: attenuate glucose utilisation in peripheral tissues, anti-lipolytic in adipose tissue, attenuate proteolysis in muscle; βHB also acts as a signalling metabolite (histone deacetylase inhibitor) relevant to training adaptation
Ketone esters allow acute nutritional ketosis without dietary restriction; initial reports suggest altered metabolic response and improved performance, other evidence suggests a role in recovery — the review states no effect sizes
Evidence Summary — overreaching trial (Poffé 2019)
Fit male subjects, two training sessions per day, 3 weeks, 6 days/week; ketone ester (KE, n = 9) or control drink (CON, n = 9) after each session
Sustainable training load in week 3 and power output in the final 30 min of a 2-h standardized endurance session were 15% higher in KE than CON (both P < 0.05)
KE inhibited the training-induced rise in nocturnal adrenaline and noradrenaline excretion (P < 0.01)
KE blunted the fall in resting HR (CON −6 ± 2 vs KE +2 ± 3 bpm), submaximal HR (−15 ± 3 vs −7 ± 2 bpm) and maximal HR (−17 ± 2 vs −10 ± 2 bpm)
Energy balance turned negative in CON (−2135 kJ/day) but not in KE (+198 kJ/day)
GDF15 rose with training, ~2-fold more in CON than KE; ΔGDF15 correlated with the drop in maximal HR (r = 0.60) and in osteocalcin (r = 0.61) → GDF15 proposed as an overtraining marker
ACTH, cortisol, IL-6, leptin, ghrelin, lymphocytes and muscle glycogen did not differ between groups
Practical Application
In 18 fit men training twice a day for 3 weeks, a ketone ester after each session kept sustainable training load and late-session power 15% higher than control, blunted the catecholamine and heart-rate signs of overreaching, and prevented the spontaneous slide into negative energy balance (Poffé 2019); the acute race-day performance case rests only on "initial reports" (Evans 2017). Ketones are a recovery tool for very heavy blocks, not a proven ergogen.
Consider only for: deliberate overload blocks or camps where recovery between two daily sessions is the limiter — and only if tolerated and affordable
Not first-line — carbohydrate periodization and training design are higher-return; ketones complement CHO, they do not replace it (REF-331/332/333)
Not from this paper; the base's own practice: dose and timing follow the product's protocol and are tested in training; neither abstract states a gram dose or a βHB target. Avoid for short glycolytic events — Evans notes ketones attenuate glucose utilisation, so high-intensity output could suffer (inference, not tested)
Limitations
One small controlled trial (9 per group) carries the recovery claim; replication needed
Poffé's design confounds ketone ester with energy intake (KE prevented the negative energy balance) — part of the effect may be calories
GI intolerance common at higher doses (not quantified in the abstracts)
Cost prohibitive for most; race-day feasibility questionable
Long-term safety / training-response effects unclear; commercial claims exceed the evidence
Audit note (2026-07-06)
DOI: 10.1113/JP273185 | PMID: 27861911 [DOI CORRECTED 2026-07-09]
Level B+ → B: the acute performance effects are contradictory (some studies ergolytic); the main evidentiary strength lies in a single overreaching RCT with small n (n=9/group).
How this entry was checked and changed
Content correction 2026-09-27
"β-HB 2-5 mmol/L", "useful when CHO already ≥100 g/h", "dose 15-25 g ketone ester 30 min pre-session" -> not in either abstract; removed.
"Some studies show small ergogenic effect in TT; others neutral or ergolytic" -> Evans 2017's abstract says only "initial reports suggest ... improves exercise performance"; reworded; the ergolytic claim stays only in the audit note as the earlier regrade rationale.
Added Poffé 2019's full abstract results (HR, catecholamines, energy balance, GDF15, null markers) and the energy-intake confound. Sport "running" -> general.
2026-09-28: added Journal line from the Europe PMC record (Europe PMC/PubMed PMID 27861911).