ArticlePDF Available

Ketone ester supplementation blunts overreaching symptoms during endurance training overload

Authors:

Abstract and Figures

Key points: •Overload training is required for sustained performance gain in athletes (functional overreaching). However, excess overload may result in a catabolic state which causes performance decrements for weeks (non-functional overreaching) up to months (overtraining). •Blood ketone bodies can attenuate training- or fasting-induced catabolic events. Therefore, we investigated whether increasing blood ketone levels by oral ketone ester (KE) intake can protect against endurance training-induced overreaching. •We show for the first time that KE intake following exercise markedly blunts the development of physiological symptoms indicating overreaching, and at the same time significantly enhances endurance exercise performance. •We provide preliminary data to indicate that growth differentiation factor 15 (GDF15) may be a relevant hormonal marker to diagnose the development of overtraining. •Collectively, our data indicate that ketone ester intake is a potent nutritional strategy to prevent the development of non-functional overreaching and to stimulate endurance exercise performance. Abstract: It is well known that elevated blood ketones attenuate net muscle protein breakdown, as well as negate catabolic events, during energy deficit. Therefore, we hypothesized that oral ketones can blunt endurance training-induced overreaching. Fit male subjects participated in two daily training sessions (3 weeks, 6 days/week) while receiving either a ketone ester (KE, n = 9) or a control drink (CON, n = 9) following each session. Sustainable training load in week 3 as well as power output in the final 30 min of a 2-h standardized endurance session were 15% higher in KE than in CON (both P < 0.05). KE inhibited the training-induced increase in nocturnal adrenaline (P < 0.01) and noradrenaline (P < 0.01) excretion, as well as blunted the decrease in resting (CON: -6 ± 2 bpm; KE: +2 ± 3 bpm, P < 0.05), submaximal (CON: -15 ± 3 bpm; KE: -7 ± 2 bpm, P < 0.05) and maximal (CON: -17 ± 2 bpm; KE: -10 ± 2 bpm, P < 0.01) heart rate. Energy balance during the training period spontaneously turned negative in CON (-2135 kJ/day), but not in KE (+198 kJ/day). The training consistently increased growth differentiation factor 15 (GDF15), but ∼2-fold more in CON than in KE (P < 0.05). In addition, delta GDF15 correlated with the training-induced drop in maximal heart rate (r = 0.60, P < 0.001) and decrease in osteocalcin (r = 0.61, P < 0.01). Other measurements such as blood ACTH, cortisol, IL-6, leptin, ghrelin and lymphocyte count, and muscle glycogen content did not differentiate KE from CON. In conclusion, KE during strenuous endurance training attenuates the development of overreaching. We also identify GDF15 as a possible marker of overtraining.
Content may be subject to copyright.
J Physiol 597.12 (2019) pp 3009–3027 3009
The Journal of Physiology
Ketone ester supplementation blunts overreaching
symptoms during endurance training overload
Chiel Poff´
e1, Monique Ramaekers1, Ruud Van Thienen1and Peter Hespel1,2
1Exercise Physiology Research Group, Department of Movement Sciences, KU Leuven, Leuven, Belgium
2Bakala Academy-Athletic Performance Center, KU Leuven, Leuven, Belgium
Edited by: Michael Hogan & Bettina Mittendorfer
Key points
rOverload training is required for sustained performance gain in athletes (functional over-
reaching). However, excess overload may result in a catabolic state which causes performance
decrements for weeks (non-functional overreaching) up to months (overtraining).
rBlood ketone bodies can attenuate training- or fasting-induced catabolic events. Therefore, we
investigated whether increasing blood ketone levels by oral ketone ester (KE) intake can protect
against endurance training-induced overreaching.
rWe show for the first time that KE intake following exercise markedly blunts the development of
physiological symptoms indicating overreaching, and at the same time significantly enhances
endurance exercise performance.
rWe provide preliminary data to indicate that growth differentiation factor 15 (GDF15) may be
a relevant hormonal marker to diagnose the development of overtraining.
rCollectively, our data indicate that ketone ester intake is a potent nutritional strategy to
prevent the development of non-functional overreaching and to stimulate endurance exercise
performance.
Abstract It is well known that elevated blood ketones attenuate net muscle protein breakdown, as
well as negate catabolic events, during energy deficit. Therefore, we hypothesized that oral ketones
can blunt endurance training-induced overreaching. Fit male subjects participated in two daily
training sessions (3 weeks, 6 days/week) while receiving either a ketone ester (KE, n=9) or a
control drink (CON, n=9) following each session. Sustainable training load in week 3 as well as
power output in the final 30 min of a 2-h standardized endurance session were 15% higher in KE
than in CON (both P<0.05). KE inhibited the training-induced increase in nocturnal adrenaline
(P<0.01) and noradrenaline (P<0.01) excretion, as well as blunted the decrease in resting (CON:
6±2bpm;KE:+2±3bpm,P<0.05), submaximal (CON: 15 ±3bpm;KE:7±2bpm,
P<0.05) and maximal (CON: 17 ±2bpm;KE:10 ±2bpm,P<0.01) heart rate. Energy
Peter Hespel is full professor and Head of the Exercise Physiology Research Group and of the Bakala
Academy-Athletic Performance Center at KU Leuven, Belgium. His research mainly focuses on the
physiological adaptations to endurance exercise training and the interaction with nutrition. Chiel Poff´
e
is a doctoral candidate under the supervision of Peter Hespel and performed the current work as part of
his PhD thesis. His research is focused on the role of ketone bodies in the adaptive response to exercise
and training.
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society DOI: 10.1113/JP277831
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution
and reproduction in any medium, provided the original work is properly cited.
3010 C. Poff ´
e and others J Physiol 597.12
balance during the training period spontaneously turned negative in CON (2135 kJ/day), but
not in KE (+198 kJ/day). The training consistently increased growth differentiation factor 15
(GDF15), but 2-fold more in CON than in KE (P<0.05). In addition, delta GDF15 correlated
with the training-induced drop in maximal heart rate (r=0.60, P<0.001) and decrease in
osteocalcin (r=0.61, P<0.01). Other measurements such as blood ACTH, cortisol, IL-6,
leptin, ghrelin and lymphocyte count, and muscle glycogen content did not differentiate KE from
CON. In conclusion, KE during strenuous endurance training attenuates the development of
overreaching. We also identify GDF15 as a possible marker of overtraining.
(Resubmitted 13 March 2019; accepted after revision 25 April 2019; first published online 30 April 2019)
Corresponding author P. Hespel: Exercise Physiology Research Group, Department of Movement Sciences, KU Leuven,
Tervuursevest 101, Heverlee, 3001, Belgium. Email: peter.hespel@kuleuven.be
Introduction
Endurance athletes intermittently participate in over-
load training (e.g. training camps) or competition (e.g.
multi-stage cycling races) with the express purpose of
eliciting physiological responses that are crucial for
sustained performance gain. Training overload, however,
must be consistently well balanced with recovery, allowing
physiological repair mechanisms to produce beneficial
adaptations that eventually yield performance gains (e.g.
functional overreaching). Inadequate recovery, with or
without other stress factors, such as sleep deprivation,
negative energy balance, disease or mental fatigue,
eventually can result in a maladaptive catabolic state
requiring days to weeks (non-functional overreaching),
or even months (overtraining) to fully recover (Meeusen
et al. 2013). Because sufficient recovery will only result
in performance improvements during functional over-
reaching, prevention of non-functional overreaching is
pivotal in training management.
The primary indication of overreaching is a stagnation
or decrease in training-specific performance (Hooper et al.
1995; Urhausen & Kindermann, 2002). However, this
is often preceded by numerous other symptoms, such
as mood disturbances (Morgan et al. 1987; Killer et al.
2017) and dysregulation in various physiological systems,
including the autonomic nervous system (Lehmann et al.
1998), immunity (Fry et al. 1994b) and energy metabolism
(Lombardi et al. 2012). Nevertheless, previous studies
have failed to identify consistent physiological outcomes
(Urhausen & Kindermann, 2002; Cadegiani & Kater, 2017)
that can be used to predict overtraining. In fact, the
pathophysiology of overtraining remains poorly under-
stood, which impairs the design of optimal preventive
interventions (Armstrong & Vanheest, 2002; Kreher &
Schwartz, 2012).
Numerous studies have focused on the use of
post-exercise nutrition to facilitate recovery between
training sessions and thereby counteract overtraining
(Kreider et al. 2010; Hawley et al. 2011). Amongst
various nutritional interventions, protein–carbohydrate
co-ingestion is recognized to be the best strategy to
enhance recovery by stimulating both muscle glycogen
resynthesis (van Loon et al. 2000) and muscle repair
(Breen et al. 2011). However, protein plus carbohydrate
ingestion is insufficient to prevent overtraining (Achten
et al. 2004; Halson et al. 2004; Witard et al. 2011; D’Lugos
et al. 2016; Svendsen et al. 2016). Other nutritional inter-
ventions, such as antioxidant intake to protect against
exercise-induced oxidative stress, fail to negate over-
training (Gleeson & Bishop, 2000; Meeusen & Watson,
2007). In contrast, consistent antioxidant intake may even
inhibit beneficial training adaptations (Merry & Ristow,
2016).
Some recent publications have stimulated interest in
the use of exogenous ketone supplements as a novel
fuelling strategy to modulate metabolic responses both
during (Cox et al. 2016; Leckey et al. 2017) and after
exercise (Holdsworth et al. 2017; Vandoorne et al. 2017).
Ketone bodies, namely D-β-hydroxybutyrate (D-βHB),
acetoacetate (AcAc) and acetone, are fatty acid-derived
compounds that can serve as an alternative energy sub-
strate for active metabolic tissues including brain (Owen
et al. 1967), heart (Aubert et al. 2016) or skeletal muscles
(Balasse & F´
ery, 1989) under conditions of metabolic
stress (Johnson et al. 1969; Cahill, 1976). Aside from
their role in energy supply, ketone bodies also can
play a role in metabolic regulation by inhibiting muscle
proteolysis (Thomsen et al. 2018) and glucose depletion
(Robinson & Williamson, 1980), and stimulation of
muscle regeneration or remodelling by enhancing satellite
cell activation and differentiation (Zou et al. 2016).
Moreover, ketone bodies directly regulate a range of
putative factors involved in the development of over-
training, such as autonomic neural output, inflammation
and oxidative stress (Kimura et al. 2011; Puchalska &
Crawford, 2017). These actions clearly indicate a potential
role for ketone bodies in prevention of overtraining, but
the dietary conditions required to elevate blood ketone
levels, by a sustained low-carbohydrate high-fat diet, are
detrimental to endurance exercise performance (Cox &
Clarke, 2014).
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society
J Physiol 597.12 Ketone supplementation in endurance training 3011
However, exogenous ketone supplements recently
emerged as a novel approach to induce ketosis. Exogenous
supplements are available either in the form of ketone
salts or ketone esters (Cox et al. 2016; Brownlow et al.
2017; Leckey et al. 2017), but ketone esters allow
blood ketones to reach 3-fold higher levels than salts,
with less incidence of gastrointestinal problems (Stubbs
et al. 2017; Sansone et al. 2018). Recent studies in our
and other laboratories have shown that post-exercise
ingestion of the ketone monoester (R)-3-hydroxybutyl
(R)-3-hydroxybutyrate (KE) stimulates markers of protein
synthesis and potentially also muscle glycogen repletion
following exercise (Holdsworth et al. 2017; Vandoorne
et al. 2017). However, these experiments have only looked
at the acute effects following a single, high-intensity
exercise bout. The effects of long-term ketone ester
intake in training and recovery remain unknown.
Given the evidence that D-βHB infusion antagonizes
starvation-induced catabolic processes (Sherwin et al.
1975; Pawan & Semple, 1983), it is reasonable to post-
ulate that ketosis induced by consistent exogenous ketone
intake during strenuous training may alleviate detrimental
catabolic events.
Therefore, we performed a double-blind, placebo-
controlled study to assess the effect of KE during
strenuous endurance training with the express purpose
of producing a state of non-functional overreaching.
We hypothesized that KE can attenuate training-induced
pathophysiological effects leading up to non-functional
overreaching, and eventually enhance endurance exercise
performance.
Methods
Ethical approval and subjects
Twenty healthy, physically active males were recruited to
participate in this study, which was approved by the KU
Leuven Biomedical Ethics Committee (B322201733747),
and conforms to the Declaration of Helsinki.Potentialsub-
jects were screened using a medical questionnaire and a
physical examination, including a resting ECG, prior to
involvement in the study. From the initial 20 recruits, one
subject did not complete the study due to adverse reactions
to the protein–carbohydratedr inks prescribed bythe study
protocol, and another withdrew for reasons unrelated to
the study protocol. Eighteen subjects eventually completed
the full study protocol and were included in the final
data analyses (for subject characteristics see Table 1). All
subjects were regularly involved in sports and physical
activity at a rate of 4.8 ±0.4 h/week (mean ±SEM), but
none were consistently engaged in cycling. Throughout
the entire study period, subjects were instructed to refrain
from strenuous exercise other than prescribed by the study
protocol. All subjects were informed of the content and
Table 1. Subject characteristics
Control Ketone ester
Age (years) 21.2 ±2.9 21.4 ±2.4
Height (m) 1.81 ±0.04 1.80 ±0.04
Body mass (kg) 74.6 ±10.5 72.8 ±6.5
˙
VO2max (ml/kg/min) 55.3 ±6.1 55.9 ±5.5
Values are mean ±SD and represent baseline characteristics of
the subjects receiving either control (n=9) or ketone ester
(n=9).
potential risks involved with the experimental procedures
before providing their written consent.
Preliminary testing and subject randomization
Two weeks before baseline measurements, the sub-
jects completed three preliminary sessions with a 48-h
rest interval in between. During the first visit, sub-
jects performed a maximal incremental exercise test
on a bicycle ergometer (Avantronic Cyclus II, Leipzig,
Germany) to determine their ˙
VO2max. Initial workload was
set at 70 W, followed by 25-W increments per minute,
until volitional exhaustion. Respiratory gas exchange was
measured continuously during the test (Cortex Metalyzer
II, Leipzig, Germany) and the highest oxygen uptake
measured over a 30 s period was defined as the maximal
oxygen uptake rate ( ˙
VO2max). During the second and third
session, subjects were familiarized with the exercise testing
procedures to undergo in the experimental sessions. Sub-
jects started with 10 min of warming up (5 min at 100 W,
5 min at 150 W) followed by a 30 min simulated time-trial
(TT30min) on a cycling ergometer (Avantronic Cyclus II).
They were instructed to maintain their cadence between
80 and 100 rpm, and to adjust the resistance at 5 min
intervals from t5 to t25, and every min from t25 to t30 to
develop the highest possible mean power output (W) over
30 min. Following 15 min of active recovery by cycling
at 50 W, the subjects performed a 90 s all-out sprint
(90S) on a self-constructed isokinetic cycling ergometer
(Koninckx et al. 2008, 2010) with cadence fixed at 90 rpm.
Following familiarization, the subjects were pair-matched
to obtain two groups with similar distributions for ˙
VO2max,
mean power outputs in TT30min and 90S, training history
(hours/week), and body mass and height. The matched
pairs were then randomly split into two experimental
groups by an investigator who was otherwise not involved
in the trial.
General experimental design
After randomization the subjects were enrolled in a fully
controlled 3-week intensive cycling training programme
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society
3012 C. Poff ´
e and others J Physiol 597.12
which was designed with the express purpose of inducing
a state of non-functional overreaching. During the
training period the subjects received either a ketone ester
drink (KE, n=9) or a corresponding control drink
(CON, n=9). Following the training period, the sub-
jects were followed-up during a tapering week without
training.
Training intervention
A detailed overview of the training programme is shown
in Fig. 1. The 3-week overload-training programme
(28 training sessions) consisted of a combination of
high-intensity interval training (HIIT), intermittent end-
urance training (IMT) and constant-load endurance
training (ET) sessions. All training sessions were
performed in the laboratory on an electromagnetically
braked ergometer and under the careful supervision of
the investigators. The HIIT and IMT sessions were done
on calibrated cycling ergometers (Avantronic Cyclus II)
in order to be able to monitor the power output–heart
rate relationship (W). The ET sessions were alternately
performed on the same calibrated ergometers or on
routine cycling ergotrainers (Tacx Neo Smart, Wassenaar,
The Netherlands). The workload in ET sessions on the
routine ergotrainers was based on the heart rate–power
relationship obtained from the sessions on the calibrated
ergometers (2–3×/week). During the ET sessions, sub-
jects cycled at a continuous load (70–95% of mean power
output TT30min) for 60–150 min. HIIT sessions consisted
of 30 s maximal sprints at a cadence fixed at 100 rpm,
interspersed by 4 min 30 s active recovery intervals at
50 W. The number of sprints was increased from four
in week 1, five in week 2, to six in week 3. Subjects were
given verbal encouragement to perform maximally during
each sprint. IMT sessions consisted of five 6-min bouts at
100–110% of mean power output taken from TT30min,
separated by 8 min of lower intensity at 55–85%. During
the final week, the duration of the high-intensity bouts
was increased to 8 min, whilst decreased to 6 min for
the lower intensity bouts. Exercise intensities of IMT and
ET were calculated relative to the mean power output
effected during TT30min in the pre-test. The intensity
of each training session is given in Fig. 1. If a sub-
ject failed to maintain a cadence 70 rpm at the pre-
scribed workload, the workload was decreased to the
level that allowed the subjects to return and maintain a
cadence 70 rpm.
Post-exercise nutritional intervention
Subjects from both experimental groups received a
500 ml high-dose protein–carbohydrate drink (Table 2)
30 min after each exercise session. In addition,
immediately following each session and 30 min before
sleep, KE subjects received 25 g of ketone ester [96%
(R)-3-hydroxybutyl (R)-3-hydroxybutyrate] to elevate
post-exercise circulating plasma ketone concentrations, as
previously shown by our lab (Vandoorne et al. 2017). The
ketone ester supplements were purchased from TdeltaS
Day 1 Day 2 Day 3 Day 7Day 6Day 5Day 4
AMPMAMPMAMPM
Week 1Week 2Week 3
70 min
IMT
100/55%
Rest
70 min
IMT
100/65%
Rest
120 min
HIIT & ET
85%
Rest
Rest
Rest
Rest
Rest
Rest
Rest
30 min
HIIT
60 min
ET
70%
30 min
HIIT
90 min
ET
77.5%
70 min
IMT
110/80%
90 min
ET
90%
70 min
IMT
100/55%
60 min
ET
70%
70 min
IMT
105/65%
60 min
ET
85%
120 min
EPT
120min
85% - 30’ all-out
Rest
30 min
HIIT
60 min
ET
70%
30 min
HIIT
90 min
ET
80%
70 min
IMT
110/80%
120 min
ET
95%
70 min
IMT
100/55%
60 min
ET
70%
70 min
IMT
110/80%
60 min
ET
90%
70 min
IMT
110/85%
150 min
HIIT & ET
92.5%
Test Wk1
60 min
ET
77.5%
Test Wk2
90 min
ET
85%
Posttest
Tes t W k3
Rest
Figure 1. Overview of the training programme
HIIT, high-intensity interval training: 30 s all-out sprints interspersed by 4.5 min active recovery intervals (50 W). The
number of sprints was increased from 4 in week 1, to 5 in week 2, and 6 in week 3. IMT, intermittent endurance
sessions. Week 1 and 2: 5 ×6minwith8minrecovery.Week3:5×8 min with 6 min recovery, ET, constant-load
endurance training sessions. EPT120min, 120 min endurance performance test. Training intensities for IMT and ET
are expressed as a percentage of the mean power output effected during the 30 min time-trial (TT30min)inthe
pre-test.
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society
J Physiol 597.12 Ketone supplementation in endurance training 3013
Table 2. Composition of protein–carbohydrate recovery drinks
Protein-
carbohydrate
recovery
mixture
Total carbohydrates 60.6 g
Maltodextrin 30.6 g
Sucrose 20.0 g
Fructose 10.0 g
Total protein 31.0 g
Non-essential amino acids 15.9 g
Essential amino acids 15.1 g
Of which leucine 3.13 g
Composition of protein–carbohydrate drinks given 30 min
following each training session. Data represent grams of
ingredients per 500 ml drink.
Ltd (Thame, UK). Subjects in CON received an isocaloric
drink (CON) containing 16.4 g pure medium-chain
triglycerides (Now Foods, Bloomingdale, IL, USA). To
equalize the taste and appearance of CON and KE drinks,
1m
Mbitter sucrose octaacetate (Sigma-Aldrich, Bornem,
Belgium) was added to the CON, whilst a red colorant
(AVEVE Bloem, Merksem, Belgium) was added to both
the drinks to obtain similar appearance. Subjects received
50 ml of diet coke for mouth rinsing immediately following
the KE or CON drink to improve palatability (Leckey
et al. 2017). Capillary blood samples from the earlobe were
obtained before and after exercise and 30 min following
ingestion of the supplements to assess circulating
blood β-hydroxybutyrate concentrations (Glucomen Lx
plus-meter with Lx β-ketone strips, Menarini Diagnostics,
Firenze, Italy) during IMT sessions on days 6, 13 and 20
of the training period.
Experimental trials
Before (pre-test), during (days 7 and 14) and after
(post-test) the training period, as well as during the
recovery phase (3 and 7 days later), the subjects
participated in an experimental session involving a TT30min
and 90 s isokinetic sprint. On the evening before each
session, they received a standardized carbohydrate-rich
dinner (5400 kJ; 69% carbohydrate, 16% fat, 15%
protein). Next morning and after an overnight fast,
a blood sample was collected from both an earlobe
(capillary blood) and from an antecubital vein (Venoject,
Tokyo, Japan). The subjects then received a standardized
breakfast (2700 kJ; 71% carbohydrate, 15% fat, 14%
protein). Following a 1.5 h rest in a comfortable chair
they warmed up for 10 min at incremental workloads
corresponding to 70% (5 min) and 85% (5 min) of their
average power output recorded during TT30min in the
last familiarization session. Thereafter, they performed
aTT
30min in which they aimed for the highest possible
mean power output. During the first 5 min (t0–t5), the
workload was set equal to the average power obtained
during the last familiarization session. From t5 to t25, sub-
jectswereallowedtoadjusttheworkloadat5-minintervals
according to their subjective perception of fatigue. From
t25 to t30, 1-min adjustments were allowed to facilitate full
exhaustion by the end of TT30min.Subjectswereallowedto
drink water ad libitum and received online feedback about
the time to completion. On completion of the TT30min,
they recovered for 15 min by cycling at 50 W, followed by
the 90S. On day 18 of the training period, a 120 min end-
urance exercise performance test (EPT120min)wasincluded
to assess endurance performance (Jeukendrup et al. 1996).
This session consisted of a 90 min preload (cycling at 85%
of the mean power output effected during TT30min in the
pre-test) to induce fatigue, followed by an all-out 30 min
time trial. During these tests, heart rate was monitored
continuously (Polar RS800CX, Kempele, Finland), while
blood lactate concentration was measured (Lactate Pro2,
Arkray, Japan) in a capillary blood sample from the
earlobe at 5-min intervals during TT30min,and3minafter
completion of 90S. Ratings of perceived exertion (RPE,
6–20 Borg Scale; Borg, 1990) were recorded immediately
after completion of TT30min, 90S and EPT120min .Power
outputs were blinded to the subjects during all
tests. Standardized verbal encouragement was provided
only during 90S. During the pre- and post-test, an
additional blood sample was obtained from an antecubital
vein (Venoject) immediately before and after the
TT30min.
Body composition and anthropometry
Whole-body dual-energy X-ray absorptiometry (DXA)
scans (Discovery W, Hologic Inc., Bedford, MA, USA)
were made during both the pre-test and the post-test.
Scans were performed in the fasted state and at the
same time of the day during both sessions to mini-
mize measurement errors (Bone & Burke, 2016). Output
parameters considered were whole-body bone mineral
content (BMC) and density (BMD), and percentage body
fat and lean soft tissue mass. All scans were performed
by a single certified technician, and subject positions over
the different sessions were standardized according to the
manufacturer’s recommendations. The densitometer was
calibrated daily against a spinal phantom to account for
potential day-to-day variability. Following the DXA scan,
skinfold thickness was measured at 12 sites (biceps, triceps,
subscapular, supra-iliac, midaxillary, iliac-crest, abdomen,
chin, anterior thigh, posterior thigh, lateral calf and medial
calf) according to standard procedures.
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society
3014 C. Poff ´
e and others J Physiol 597.12
Subjective feelings of appetite, gastrointestinal
discomfort and recovery–stress state
During each experimental session, subjects completed
three questionnaires to assess their: (i) appetite sensations;
(ii) gastrointestinal discomfort; and (iii) recovery–stress
state. Appetite was assessed immediately before break-
fast using a validated 0–10 Likert visual analog scale
(VAS),adaptedfromWoodset al. (2018). Subjects
were provided with four appetite or satiety questions
(‘How hungry do you feel?’, ‘How full do you feel?’,
‘How satisfied do you feel?’, ‘How much do you think
you could eat now?’). Gastrointestinal discomfort was
rated after completion of breakfast by means of a 0–8
Likert scale questionnaire adapted from Pfeiffer et al.
(2009). The questionnaire comprised three sections, i.e.
upper abdominal problems (reflux, bloating, nausea,
vomiting); lower abdominal problems (cramps, flatulence,
abdominal pain, diarrhoea); and systemic problems
(dizziness, headache, muscle cramp, urge to urinate). Sub-
sequently, the Recovery-Stress Questionnaire for Athletes
(RESTQ-76 Sport questionnaire) was administered to
assess subjects’ recovery–stress state and toclassify whether
they became overreached or overtrained (Kellmann &
Kallus, 2001). In accordance with a previous study, the
recovery–stress balance was calculated by subtracting the
total recovery score (9 recovery subscales) from the total
stress score (10 stress subscales) (Coutts et al. 2007).
High scores in the recovery-associated scales represent
adequate recovery, while hig h scores for the stress-oriented
subscales represent intense subjective strain.
Nutritional control
Nutritional intake was monitored via an online dietary
platform (Mijn Eetmeter, Stichting Voedingscentrum
Nederland; https://mijn.voedingscentrum.nl). The user
modalities and dietary recording procedure were
explained in detail to the subjects following the last
familiarization session. Diaries were obtained during two
subsequent days at the beginning (days 3–4), mid (days
12–13) and at the end (days 19–20) of the training period.
Diaries were analysed for both macronutrient and energy
intake.
Muscle biopsy procedure and muscle glycogen
content
During the pre- and post-test, a percutaneous needle
biopsy (100–200 mg) was obtained under local anaesthesia
(2% xylocaine without adrenaline, 1 ml subcutaneously)
before and immediately after the TT30min from the m.
vastus lateralis using a 5-mm Bergstr¨
om-type needle.
Biopsies during the pre- and post-test were taken from the
left and right leg, respectively, while pre- and post-exercise
biopsies were obtained through the same incision, but
with the needle pointing distal vs. proximal, respectively
(Van Thienen et al. 2014). Part of the muscle sample was
immediatelyfrozeninliquidnitrogenandstoredat80°C
for assay of muscle glycogen content at a later date. Muscle
glycogen content was determined as glucose residues after
acid hydrolysis using a standard enzymatic fluorometric
assay (Lowry & Passoneau, 1972).
Urine sampling and analysis
Nocturnal urine was collected over 12 h (20.00–22.00 h
until 08.00–10.00 h) the night before each experimental
session in flasks prepared with 10 ml hydrochloric acid.
Total urinary output volume was noted, and urinary
ketone concentration was measured using ketone reagent
strips (Ketostix, Ascensia Diabetes Care). An aliquot of the
urine sample was stored at 80°C until assayed in a single
run for adrenaline and noradrenaline concentration using
a commercially available enzyme linked immunosorbent
assay (ELISA) (BA E-5400, LDN, Nordhorn, Germany).
Analysis of blood samples
Capillary blood samples from the earlobe were
immediately analysed for blood D-βHB (Glucomen
Lx plus-meter with Lx β-ketone strips, Menarini
Diagnostics). D-βHB measurements were performed by
an investigator who was otherwise not involved in
the experimental testing to ensure double-blindness.
Venous blood samples were collected into vacuum tubes
containing either EDTA or lithium heparin or Silica
Clot Activator (BD Vacutainer). Tubes were centrifuged
(1500 rpm for 10 min at 4°C) and the supernatant
was stored at 20°C until later analysis. Commercially
available ELISAs were performed to determine leptin,
total ghrelin, growth differentiation factor 15 (GDF15)
and total osteocalcin in serum, while IL-6 levels were
determined in EDTA plasma (Leptin: BMS2039INST,
Thermo Fisher Scientific, Waltham, MA, USA; Total
Ghrelin: EZGRT-89K, Merck, Darmstadt, Germany;
GDF15: DGD150, R&D, Minneapolis, MN, USA; Total
osteocalcin: KAQ1381, Thermo Fisher Scientific; IL-6:
HS-600B, R&D). Cortisol and ACTH levels were assayed
using electrochemiluminescence immunoassays (ECLIAs)
in serum and EDTA plasma samples, respectively. TRAP5b
activitywasmeasuredinserumbyadirectcapture
enzyme-immunoassay (no. 8036, TECOmedical, Sissach,
Switzerland). Fasted whole blood samples obtained
during the pre- and post-test were analysed for the
proportion of lymphocyte subset ty pes, i.e. T-cell (CD3+),
T-helper/inducer cell (CD4+), T-suppressor/cytotoxic
cell (CD8+) and CD4+/CD8+cell count by flow
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society
J Physiol 597.12 Ketone supplementation in endurance training 3015
cytometry. Plasma glucose levels were determined using
a standard enzymatic fluorometric assay (Lowry &
Passoneau, 1972). All assays were run in a single batch,
which included all samples, according to the respective
protocols supplied by the manufacturer.
Statistical analysis
Statistical analyses were performed in R version 3.3.3
using the nlme package for mixed-effects models and
the stats package for unpaired ttests (R Development
Core Team, Vienna, Austria). Differences between mean
values over time and between conditions were analysed
using a two-way repeated-measures analysis of variance
(group ×time). D-βHB levels during the training
sessions, as well as muscle glycogen content and blood
glucoselevelsbeforeandafterTT
30min were assessed
by a three-way repeated-measures analysis of variance
(group ×time ×number of training/experimental
session). Pearson correlation coefficients were calculated
using percentage change scores unless otherwise stated.
Holm–Sidak’s multiple comparison test was used for post
hoc analysis, when appropriate. Where relevant, Cohen’s
dwas calculated as index of effect size (ES). Outliers were
identified using the ROUT method in GraphPad Prism
version 8.00 (GraphPad Software, La Jolla, CA, USA). If
outliers were detected, statistics were performed both with
and without the outlier and both results are included in
the text. Statistical significance was defined as P<0.05.
Data are presented as mean ±SEM.
Results
Blood D-βHB and urinary ketone excretion
In the pre-test, fasted blood D-βHB levels were similar
between the groups at 0.1 mM. During the training
period, fasted D-βHB levels gradually increased, reaching
peak levels at 0.35 mMin the post-test (P<0.01;
Fig. 2A). During the recovery week, blood D-βHB levels
returned towards baseline within 3 days. There were no
differences between the groups at any time. Ketone bodies
were undetectable (<0.05 g/l) in urine during the pre-test
(Fig. 2B), yet transiently increased during weeks 1 and
2(P<0.01 and P<0.001, respectively), irrespective of
the experimental condition. During the training sessions,
blood D-βHB levels were low (0.1–0.3 mM)inboththe
groups before and immediately after exercise (Fig. 2C).
However, ketone ester intake immediately after exercise
increased blood D-βHB levels to 2.6 ±0.2 mMwithin
30 min (30’ post-ex in Fig. 2C), whilst values were
unchanged in CON (P<0.001 vs. KE).
Training load
Total weekly training load progressively increased from
4600 kJ in week 1, to 6400 kJ in week 2 and 9600 kJ
in week 3 (Fig. 3A). Compared to their normal training
volume (4.8 ±0.4 h/week), this corresponds to a 120%
increment in week 1, 165% in week 2 and a 3-fold
increment in week 3. Training load was similar between
the groups in weeks 1 and 2, but was 15% higher in KE
Pre-ex
Post-ex
30' post-ex
Pre-ex
Post-ex
30' post-ex
Pre-ex
Post-ex
30' post-ex
0
1
2
3
4
Blood D-ßHB (mM)
Day 6 Day 13 Day 20
*§*§
*§
Pre
Week 1
Week 2
Week 3
Day +3
Day +7
0.0
0.1
0.2
0.3
0.4
0.5
Blood D-ßHB (mM)
##
Pre
Week 1
Week 2
Week 3
Day +3
Day +7
0.0
0.1
0.2
0.3
0.4
Urinary ketone (g)
KE
CON
#
#
A
B
C
Figure 2. Effect of ketone ester
supplementation on blood d-βHB
concentrations and urinary ketone
excretion
Data are mean ±SEM for fasted morning
blood D-βHB concentration (A)and
nocturnal urinary ketone excretion (B)
before (Pre) and at the end of weeks 1
(Week 1), 2 (Week 2) and 3 (Week 3) of
the training period, and after 3 (Day +3)
and7(Day+7) recovery days after
training. During the training period the
subjects received either control (,n=9)
or ketone ester supplements (,n=9)
immediately after each training session. C,
blood D-βHB concentrations before
(Pre-ex), and immediately (Post-ex) and
30 min after (30’ post-ex) the IMT sessions
on days 6, 13 and 20. P<0.05 KE vs.
CON at time points indicated; #P<0.05
vs. PRE for both KE and CON; §P<0.05
vs. pre-ex for the indicated group.
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society
3016 C. Poff ´
e and others J Physiol 597.12
than in CON at week 3 [KE: 10266 ±321 kJ vs. CON:
8962 ±646 kJ, 95% confidence interval (CI): +237 to
+2372 kJ, P<0.05, d=0.80]. Differences in work output
between the groups were most explicit for the prolonged
endurance training sessions (ET) at the end of the training
period, with no significant differences for the HIIT and
IMT sessions (Fig. 3B).
Exercise performance and peak lactate concentration
Mean power outputs in TT30min and 90S in the pre-test
were similar between the groups (TT30min: 216 ±7vs.
215 ±8 W; 90S: 492 ±20 vs. 499 ±17 W for KE and
CON, respectively). In KE, compared to the pre-test mean
power output in TT30min was 4.9 ±1.5% higher (95% CI:
+3to+18 W, P<0.05, d=0.47) in the post-test, and
7.5 ±1.7% (95% CI: +8to+24 W, P<0.001, d=0.71)
and 8.3 ±2.1% (95% CI: +8to+28 W, P<0.001,
d=0.83) higher on day +3 and day +7, respectively
(Fig. 4A). Conversely, training did not improve TT30min
performance in CON (95% CI pre- vs. post-test: 2to
+14 W, P>0.05), except on day +7(+7.0 ±2.0%, 95% CI:
+5to+25 W, P<0.001, d=0.52). However, TT30min mean
power outputs were not significantly different between
A
123456789
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
0
500
1000
1500
2000
Training session number
Total work output (kJ)
3 keeW2 keeW1 keeW
*
*CON
KE
0
5000
10000
15000
Total work output (kJ)
KE
CON
3 keeW2 keeW1 keeW
*
#
#
B
Figure 3. Effect of ketone ester supplementation on training workload
A, individual data points together with means ±SEM representing total work output per week. B, means ±SEM
for work output per training session in subjects receiving either control (/bars, n=9) or ketone ester (/bars,
n=9) supplements. The subjects performed 28 training sessions over a 3-week training period. P<0.05 KE vs.
CON at time points indicated; #P<0.05 vs. PRE for both KE and CON.
Pre
Week 1
Week 2
Week 3
Day +3
Day+7
0
200
210
220
230
240
250
Power output (W)
§§#
CON KE
0
150
175
200
225
250
Power output (W)
*
KE
CON
Pre
Week1
Week 2
Week 3
Day+3
Day+7
0
440
470
500
530
560
Power output (W)
#
CBA
Figure 4. Effect of ketone ester supplementation on exercise performance
Data are means ±SEM. Aand B, mean power output during the 30 min simulated time-trial (TT30min)(A)andina
90 s all-out cycling bout (90S) (B) before (Pre) and at the end of weeks 1 (Week 1), 2 (Week 2) and 3 (Week 3) of
the training period, and after 3 (Day +3) and 7 (Day +7) days of recovery. C, mean power output in the final half
an hour of a 120 min endurance performance test (EPT120min) on day 18 of the training period. Subjects received
either control (/open bars, n=9) or ketone ester (/filled bars, n=9) during each training session. P<0.05 KE
vs. CON; #P<0.05 vs. PRE for both KE and CON; §P<0.05 vs. PRE for indicated group.
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society
J Physiol 597.12 Ketone supplementation in endurance training 3017
CON and KE at any time (95% CI for KE vs. CON during
post-test: 33 to +14 W, P>0.05). Furthermore, mean
power output during EPT120min on day 18 was 15%
higher in KE than in CON (KE: 216 ±8Wvs. CON:
188 ±14 W, 95% CI for KE vs. CON: +5to+52 W,
P<0.05, d=0.77; Fig. 4C). Compared to the pre-test,
mean power output effected in 90S was stable in weeks
1 and 2, but decreased by 5.5 ±1.4% in the post-test
(P<0.05, d=0.30; Fig. 4B). Yet during the recovery
period 90S power outputs returned to baseline by day +3.
Performance in 90S was not significantly different between
the groups at any time. Blood lactate levels following 90S
peaked at 13–15 mMin the pre-test as well as in weeks 1
and 2. However, in the post-test peak blood lactate levels
decreased by 5m
M(P<0.001) to 9.8 ±1.0 mMin CON
and 8.2 ±0.8 mMin KE. During the recovery period,
blood lactate concentrations rapidly returned to baseline
in CON (day +3: 13.2 ±1.9 mM,P=0.29 vs. PRE; day
+7: 12.6 ±1.5 mM,P=0.10 vs. PRE), but not in the KE
group (day +3: 10.9 ±0.8 mM,P<0.05 vs. PRE; day +7:
9.6 ±1.1 mM,P<0.01 vs. PRE).
Heart rate and blood pressure
Resting heart rate (HR) was 62 ±4bpminKEvs.
66 ±2bpminCON(P=0.73). In CON the training inter-
vention gradually decreased resting HR to 60 ±1bpmin
the post-test (P<0.05), and even further to 55 ±2bpmat
day +3(P<0.001; Fig. 5A). At day +7,restingHRwasstill
6bpmlowerthaninthepre-test(P<0.05). In contrast,
in KE resting HR was stable throughout the full training
period, and consistently was 2–5 bpm higher than in
CON (P<0.05). Only on day +3,restingHRinKEwas
slightly lower than at baseline (P<0.05). The training
also substantially decreased HR during submaximal and
maximal exercise. In CON submaximal and maximal HR
on average decreased by 16 bpm (range: 5to–28bpm)
from the pre-test to the post-test (P<0.001; Fig. 5B,C).
Thus, submaximal HR during TT30min was 167 ±4in
the pre-test, decreasing to 152 ±4 bpm in the post-test.
By analogy, maximal HR taken from 90S dropped from
189 ±2 to 172 ±3bpm(P<0.001). In KE submaximal
and maximal HRs also dropped during the training period,
yet the drop was markedly smaller than in CON (P<0.05
and P<0.01 vs. CON, respectively). Submaximal HR
dropped from 163 ±3 in the pre-test to 157 ±3bpm
in the post-test (P<0.05). Corresponding maximal HRs
were 189 ±3 and 179 ±2bpm(P<0.001). In both the
groups the training-induced suppression of submaximal
and maximal HR was rapidly inverted during the recovery
week. Within a week, submaximal HR returned to base-
line in KE, but not in CON (P<0.001). Clearly, ketone
ester intake consistently blunted (over)training-induced
bradycardia both at rest and during submaximal and
maximal exercise. Furthermore, the decrement of resting
HR effected by the training intervention was associated
with a drop of diastolic blood pressure from 68 ±1to
57 ±3 mmHg in CON (P<0.001), but not in KE (pre-test
67 ±2, post-test 62 ±2 mmHg, P=0.20). Systolic blood
pressures on average were 126 mmHg and were not
altered by either training or ketone ester intake.
Body composition and bone mineralization
Baseline values for body composition and bone
mineralization were similar between the two experimental
HRRest
**
Pre
Week 1
Week 2
Week 3
Day +3
Day +7
Pre
Week 1
Week 2
Week 3
Day +3
Day +7
Pre
Week 1
Week 2
Week3
Day +3
Day +7
-20
-10
0
10
HR (bpm)
HRMax
HRSubmax
#
#
* #
*
* #
KE
CON
§
§
§
§
§
§
Figure 5. Effect of ketone ester supplementation on heart rate
Data are means ±SEM and represent changes in resting (HRRest), submaximal (HRSubmax) and maximal (HRMax )
heart rate before (Pre), and at the end of weeks 1 (Week 1), 2 (Week 2) and 3 (Week 3) of the training period,
and after 3 (Day +3)and7(Day+7) recovery days. Subjects received either control (,n=9) or ketone ester
supplements (,n=9) following each training session. P<0.05 KE vs. CON; #P<0.05 vs. PRE for both KE and
CON; §P<0.05 vs. PRE for indicated group.
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society
3018 C. Poff ´
e and others J Physiol 597.12
Table 3. Effect of ketone ester supplementation on body composition and bone mineralization
Control Ketone ester
Pre Post Pre Post
Body weight (kg) 72.9 ±3.6 73.1 ±3.2 72.1 ±2.1 71.9 ±2.2
Lean mass (kg) 60.4 ±2.6 61.6 ±2.4 61.1 ±1.9 61.8 ±1.9
Percent body fat (%) 13.1 ±1.3 11.6 ±1.1#11.4 ±0.9 10.1 ±0.8#
Sum skinfolds (mm) 103.6 ±10.6 95.3 ±8.9#87.1 ±7.7 80.8 ±7.2#
BMC (g) 2749 ±133 2744 ±130 2758 ±106 2782 ±112#,§
BMD (g/cm2)1.19±0.04 1.19 ±0.04 1.23 ±0.03 1.24 ±0.03
Data are mean ±SEM for body weight, body composition and bone mineralization parameters measured by DXA scan, in subjects
receiving either ketone ester supplements (n=9) or control (n=9). BMC, bone mineral content; BMD, bone mineral density. #P<0.05
pre vs. post; §P<0.05 group ×time interaction.
Table 4. Effect of ketone ester supplementation on total energy and macronutrient intake
Control Ketone ester
Week 1 Week 2 Week 3 Week 1 Week 2 Week 3
Energy intake 15,439 ±741 15,055 ±493 15,282 ±808 14,688 ±1042 16,655 ±854#17,568 ±1121#
Carbohydrate 7211 ±311 7412 ±329 7417 ±397§6934 ±354 8647 ±464#9014 ±614#,§
Protein 2655 ±228 2673 ±65 2613 ±163. 2807 ±230 3019 ±217 3273 ±235
Fat 3362 ±221 3293 ±205 3394 ±380 2878 ±502 3240 ±328 3413 ±578
Data are mean ±SEM (kJ/day) for total energy and macronutrient intake, in subjects receiving either ketone ester supplements (n=9)
or control (n=9). Macronutrient data are exclusive of the KE and CON drinks. #P<0.05 vs. week 1; §P<0.05 KE vs. CON at indicated
time point.
groups (Table 3). The training period decreased body fat
percentage by 1.4 ±0.2% (P<0.001, d=0.36) and sum
of skinfolds by 7.3 ±1.4 mm (P<0.001, d=0.28), while
both lean mass and body weight remained unchanged
irrespective of the experimental conditions. In KE, BMC
from the pre-test to the post-test increased by 24 ±10 g
(P<0.01, d=0.07), whilst it was stable in CON (6±8g,
P=0.53). BMD was unaffected by the experimental
conditions.
Total energy and macronutrient intake and appetite
sensations
With the exception of the ketone ester/control drink
and the post-exercise recovery shakes prescribed by
the study protocol, food intake during the full study
period was ad libitum. Concurrent with the increase in
training workload from training week 1 to 3, KE sub-
jects spontaneously increased their total energy intake by
20% from 14 700 to 17 600 kJ/day. Total energy
intake increased proportionately (+1966 ±826 kJ/day at
week 2 and +2880 ±489 kJ/day at week 3, P<0.01
and P<0.001 vs. week 1, respectively) with the training
load and the concomitant increase in energy expenditure
[+1089 ±135 kJ/day at week 2 and +3362 ±104 kJ/day
at week 3; assuming a mechanical efficiency of 23.8%
(Ettema & Loras, 2009)] in KE, while it remained stable
in CON (energy intake: 384 ±794 kJ/day at week 2
and 157 ±751 kJ/day at week 3, P=0.86) (Table 4).
The increasing energy intake in KE was largely effected by
greater amounts of carbohydrate intake (+25.5 ±6.3% at
week 2 and +29.9 ±5.9% at week 3, both P<0.001) at
fairly constant fat and protein intake. Subjective ratings of
appetite were similar between the groups and over time
(data not shown).
Hormonal parameters
We measured ‘energy homeostasis and appetite hormones’
(GDF15, leptin, ghrelin), urinary catecholamines and the
ACTH–cortisol hypothalamic pituitary axis. All hormonal
levels were similar between the groups at baseline (Figs 6
and 7). Serum GDF15 gradually increased during the
training period in all subjects (P<0.001), yet the rise
was greater in CON than in KE (P<0.05). Thus, in
the post-test serum GDF15 levels were markedly lower
in KE (361 ±19 pg/ml) than in CON (435 ±29 pg/ml,
95% CI of KE vs. CON: 17 to 133 pg/ml, P<0.05,
d=0.91). However, values rapidly returned to base-
line during the recovery week. Compared to baseline,
the 3-week training intervention decreased serum leptin
3-fold in CON (P<0.001), but not in KE (P=0.26),
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society
J Physiol 597.12 Ketone supplementation in endurance training 3019
returning to baseline within 1 week after training. Non-
etheless, serum leptin levels were not significantly different
between the groups at any time. Serum ghrelin was not
affected by the training programme or by ketone ester
intake (KEPre: 486 ±38 pg/ml; KEPos t: 471 ±31 pg/ml;
CONPre: 466 ±31 pg/ml; CONPo st : 461 ±30 pg/ml;
group effect: P=0.74; time-effect: P=0.32). Nocturnal
urinary catecholamine excretions were stable throughout
the full intervention period in KE. In contrast, in CON
both the adrenaline and the noradrenaline excretions
increased about 2-fold from the pre-test to the post-test
(main group effect for adrenaline, P<0.01; interaction
effect for noradrenaline, P<0.01). Thus, in the post-test
urinary noradrenaline output was 108 ±20 nmol in CON
(P<0.01 vs. pre-test) and 56 ±17 nmol in KE (P=0.82 vs.
pre-test). Corresponding values for adrenaline excretion
were 41 ±17 nmol in CON vs. 17 ±7nmolinKE
(group effect: P<0.01). Plasma ACTH levels tended to
decrease in both groups from the pre-test to the post-test
(pre-test 54.4 ±6.4 ng/l vs. post-test 44.1 ±5.5 ng/l,
P=0.08), while serum cortisol levels were unchanged
(pre-test 184.5 ±8.0 μg/l vs. post-test 182.2 ±7.8 μg/l,
P=0.83). ACTH/cortisol ratio tended to decrease in CON
(pre-test 340 ±91 ×106vs. post-test 196 ±22 ×106,
P=0.07), but not in KE (pre-test 306 ±46 ×106vs.
post-test 278 ±55 ×106,P=0.72).
Moderate to strong negative correlations were found
between both absolute and delta GDF15 levels and
changes in maximal heart rate during the training
period (r=−0.64 and r=−0.59, respectively, both
P<0.001). One outlier was identified in both data sets,
but its removal only marginally affected the correlation
coefficients (r=−0.64 and r=−0.60, respectively, both
P<0.001, Fig. 8). Changes in leptin levels were positively
correlated with alterations in BMC (r=0.56, P<0.05).
Cytokine and immune response
Plasma IL-6 levels were similar between both groups in
the pre-test (KE: 0.68 ±0.06 vs. CON: 0.85 ±0.12 pg/ml;
Pre
Week 1
Week 2
Week 3
Day +3
Day +7
0
20
40
60
80
Urinary adrenaline (nmol)
*
Pre
Week 1
Week 2
Week 3
Day +3
Day +7
0
50
100
150
Urinary noradrenaline (nmol)
KE
CON
§
A
B
Figure 6. Effect of ketone ester supplementation on urinary catecholamine excretion
Data are means ±SEM for urinary (A) adrenaline and (B) noradrenaline excretion before (Pre) and at the end of
weeks 1 (Week 1), 2 (Week 2) and 3 (Week 3) of the training period, and after 3 (Day +3)and7(Day+7) days of
recovery after training. Subjects received either control (,n=9) or ketone ester supplements (,n=9) following
each training session. P<0.05 KE vs. CON; #P<0.05 vs. PRE for both KE and CON; §P<0.05 vs. PRE for
indicated group.
Pre
Week 1
Week 2
Week 3
Day +3
Day +7
0
200
250
300
350
400
450
500
GDF15 (pg/ml)
§§
*
#
Pre
Week 1
Week 2
Week 3
Day +3
Day +7
0
100
200
300
400
500
Leptin (pg/ml)
§§
§
KE
CON
A
B
Figure 7. Effect of ketone ester supplementation on ‘appetite’ hormones involved in energy balance
regulation
Data are means ±SEM and represent changes in (A)GDF15and(B) leptin concentration before (Pre), after 1
(Week 1), 2 (Week 2) and 3 (Week 3) weeks of training and following 3 (Day +3) and 7 (Day +7) days of recovery.
Subjects received either control (,n=9) or ketone ester supplements (,n=9) during the training period.
P<0.05 KE vs. CON; #P<0.05 vs. PRE for both KE and CON; §P<0.05 vs. PRE for indicated group.
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society
3020 C. Poff ´
e and others J Physiol 597.12
P=0.71) and were stable throughout the study in
both groups (data not shown). Baseline values for
lymphocyte subtype counts, i.e. CD3+,CD4
+and CD8+,
were similar between the groups (Table 5). Compared
to the pre-test, in the post-test CD3+and CD8+cell
counts were decreased by 5 and 14%, respectively
(P<0.001), whereas CD4+increased by 6% (P<0.001).
As a result CD4+/CD8+ratio increased by 25%
(P<0.001). Lymphocyte changes were similar between the
groups.
Bone metabolism markers
The training period decreased total serum osteocalcin
concentration by 30% (pre-test: 19.5 ±0.6 ng/ml vs.
post-test: 13.1 ±0.6 ng/ml, P<0.05), irrespective of
the experimental condition. Levels returned to baseline
within the 1-week recovery period. TRAP5b activity was
not modified by the training or by ketone ester intake
(KEPre: 4.25 ±0.25 U/l; KEPo st : 4.23 ±0.23 U/l; CONPre:
4.06 ±0.36 U/l; CONPost : 3.92 ±0.26 U/l; group effect:
P=0.30; time effect: P=0.18). Alterations in osteocalcin
showed a strong inverse correlation with changes in
GDF15 (r=−0.61, P<0.01).
Blood glucose levels and muscle glycogen content
Initial muscle glycogen content was 140 ±8 mmol/kg wet
weight in CON vs. 122 ±9 mmol/kg in KE (P=0.41).
The training period slightly decreased resting muscle
glycogen content in CON (28 ±9 mmol/kg, P<0.05)
but not in KE (+2.4 ±13.2 mmol/kg, P=0.99),
yielding identical muscle glycogen contents between CON
(111 ±10 mmol/kg) and KE (124 ±9 mmol/kg) in
the post-test. The TT30min decreased muscle glycogen
content in both groups by 44% during the pre-test
(KE: 59 ±10 mmol/kg; CON: 69 ±12 mmol/kg;
both P<0.001 compared to pre-exercise), while muscle
glycogen was unaltered by the TT30min in the post-test
(KE: 92 ±9 mmol/kg; CON: 86 ±8 mmol/kg; P=0.10
and P=0.19 vs. pre-exercise, respectively) for both groups.
Plasma glucose levels immediately before the TT30min were
similar between the groups during both the pre- and the
post-test (KEPre:4.1±0.3 mM;KE
Post :4.9±0.2 mM;
CONPre:3.9±0.2 mM;CON
Post :4.2±0.4 mM;group
effect: P=0.45; time effect: P=0.37). At the end of
the TT30min, glucose levels were increased by 1.5 mM
(P<0.001 vs. pre-exercise) during both the pre- and
the post-test independent of KE supplementation (KEPre:
5.6 ±0.3 mM;KE
Post :5.9±0.2 mM;CON
Pre:6.0±0.4 mM;
CONPost :5.3±0.6 mM).
Stress–recovery state
The training period substantially decreased the
stress–recovery state in both groups (pre-test: 80 ±5vs.
post-test: 6±7, P<0.001). This was due to an increase in
‘total stress’ scores (pre-test: 49 ±3vs. post-test: 110 ±6,
P<0.001) together with a decrease in ‘total recovery’
scores (pre-test: 129 ±4vs. post-test: 104 ±4, P<0.001).
Stress-recovery state rapidly improved during the recovery
week, yet values remained below baseline even at day 7
(69 ±6, P<0.05 vs. pre-test). There were no differences
between the groups at any time.
Gastrointestinal tolerance
Gastrointestinal discomfort scores were slightly higher
in the post-test than in the pre-test (pre-test: 7 ±1vs.
post-test: 11 ±2 out of a maximum of 96, P<0.05).
This small increase was due to a higher incidence of lower
abdominal symptoms (pre-test: 2 ±1vs. post-test: 4 ±1
out of a maximum of 32, P<0.05) and more systemic
discomfort (pre-test: 3 ±1vs. post-test: 4 ±1outof
150 200 250 300 350 400 450 500
-30
-20
-10
0
10
GDF15 (pg/ml)
HRmax (bpm)
r = -0.64; p < 0.001
-50 0 50 100 150 200
-30
-20
-10
0
10
GDF15 (pg/ml)
HRmax (bpm)
r = -0.60; p < 0.001 Week 1
Week 2
Week 3
AB
Figure 8. Relationship between GDF15 and maximal heart rate
Correlation analyses showing a strong negative correlation between alterations in maximal heart rate (HRmax)
and both (A) absolute and (B) delta GDF15 during the training period. Individual data points represent changes
after 1 (Week 1), 2 (Week 2) and and 3 (Week 3) weeks of training, compared to pre-test values.
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society
J Physiol 597.12 Ketone supplementation in endurance training 3021
Table 5. Effect of ketone ester supplementation on lymphocyte cell counts
Control Ketone ester
Pre Post Pre Post
Mature T (CD3+)1.13±0.02 1.07 ±0.01#1.11 ±0.02 1.06 ±0.01#
T-helper (CD4+)0.57±0.02 0.59 ±0.02#0.56 ±0.04 0.60 ±0.03#
T-cytotoxic (CD8+)0.48±0.03 0.41 ±0.02#0.43 ±0.03 0.37 ±0.02#
CD4+/CD8+1.23 ±0.09 1.49 ±0.12#1.35 ±0.14 1.67 ±0.15#
Data are means ±SEM (n×109cells/l) for lymphocyte subtype counts in subjects receiving either ketone ester supplements (n=9)
or control (n=9). #P<0.05 pre vs. post.
a maximum of 32, P<0.05). Upper abdominal distress
was stable between the pre-test and post-test (P=0.73).
Gastrointestinal discomfort scores returned to baseline
within the 7-day recovery period. No differences were
observed between the experimental conditions at any time
point.
Identification of experimental condition
Upon completion of the study the subjects were asked
to identify their presumed group allocation. Nobody was
certain of his experimental condition, but 10 out of
18 (5 in KE and 5 in CON) subjects correctly guessed
the condition based on subjective perceptions, indicating
successful blinding of the treatments.
Discussion
Early detection and prevention of overtraining is pivotal
in athlete training management. From this perspective
we investigated whether post-exercise ketone ester intake
(KE) can prevent non-functional overreaching and
performance impairment during an episode of excessive
training load. The endurance training programme used
induced explicit cardiovascular, hormonal and perceptual
symptoms of overreaching in all subjects. Interestingly,
KE markedly inhibited the appearance of these symptoms,
whilst enhancing tolerable training load, increasing energy
intake and stimulating endurance exercise performance.
Our data indicate that KE is a potent strategy to
prevent overtraining and stimulate endurance training
adaptation. In addition, we provide preliminary evidence
that GDF15 may be a valid hormonal marker of
overtraining.
Although the complex pathophysiology is still poorly
understood, it is the prevailing opinion that auto-
nomic neural imbalance plays an important role in the
development of overtraining (Fry et al. 1991; Lehmann
et al. 1998). From this perspective two distinct over-
training types were defined. ‘Sympathetic’ overtraining
is characterized by elevated basal sympathetic tone,
whilst sympathetic drive is abnormally decreased in the
‘parasympathetic’ overtraining form (Israel, 1976). The
former has primarily been associated with high-intensity
anaerobic exercise, versus the latter with endurance
training activities (Kuipers & Keizer, 1988), but both forms
probably rather exist as a continuum (Fry et al. 1994a,
2006). Initially sympathetic overtraining develops as a
stress response attempting to maintain functional status.
However, in a later stage, parasympathetic symptoms
gradually predominate due to fatigue at the site of
the sympathetic neuroendocrine system. In the CON
conditions of the current study, the subjects exhibited
the initial basal sympathetic stress response as evidenced
by elevated nocturnal catecholamine excretion, pre-
dominantly noradrenaline, which reflects spillover from
the sympathetic nervous system (Esler et al. 1988; Fry
et al. 1994a). Nonetheless, against the face of this
elevated sympathetic activity, resting heart rate decreased
significantly, which indicates that at the cardiac site
the elevated central sympathetic drive was probably
overruledbyexcessvagaloutput(Hedelinet al. 2000;
Pichot et al. 2002) either or not in conjunction with
downregulation of cardiac β2-adrenoceptor sensitivity
(Gleeson, 2002; Fry et al. 2006). However, these auto-
nomic responses were substantially altered by KE during
training. KE fully counteracted the initial training-induced
sympathetic overdrive, as evidenced by stable nocturnal
catecholamine excretion throughout the full training
period. Nonetheless resting heart rate slightly increased in
weeks 1 and 2, probably indicating transient sympathetic
dominance at the site of the heart. KE intake also markedly
impacted exercise tachycardia, which also depends on
sympathetic–parasympathetic interactions. In CON the
training markedly suppressed the exercise-induced rise
in heart rate, including a substantial drop in maximal
heart rate by 10–28 bpm. KE administration clearly
counteracted this effect, probably due to sympathetic
regulation (Lehmann et al. 1991; Stanojevic et al. 2013).
However, we did not measure plasma catecholamine
levels during exercise. Nevertheless, central autonomic
command of exercise heart rate is also controlled by input
from mechano- and metaboreceptors in active muscles
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society
3022 C. Poff ´
e and others J Physiol 597.12
(Fisher, 2014), which might also be changed by KE intake.
Furthermore, regulation might also occur at the site
of cardiac adrenergic sensitivity or G protein-coupled
receptor 41 (GPR41) activity. GPR41 is present in
sympathetic ganglions and is thereby directly involved
in sympathetic control (Kimura et al. 2011). D-βHB
suppresses GPR41 activity in mice, resulting in a depressed
sympathetic tone and heart rate (Kimura et al. 2011).
Apart from the KE or CON supplements, the subjects
received a standardized protein–carbohydrate solution to
stimulate recovery immediately after each training session.
Intakeofotherfoodsanddrinkswasad libitum because we
wanted to evaluate the effect of KE on appetite regulation
and spontaneous food intake. It is well known that over-
training in endurance athletes often results in appetite
suppression (Fry et al. 1992). Furthermore, decreased
appetite during the ketogenic diet has been linked to
elevated plasma ketone levels (Paoli et al. 2015). However,
in the current study, neither the training overload nor KE
decreased appetite/hunger perception. Stubbs et al. (2018)
have demonstrated that acute KE intake, which elevated
plasma D-βHB levels to 3–4 mMcompared to isocaloric
glucose ingestion, suppressed appetite in healthy sub-
jects. However, appetite scoring in the current study was
done in the early morning in the fasted state when blood
D-βHB levels were basal in both experimental groups
(<0.4 mM). Nonetheless, actual food intake pattern was
significantly different between CON and KE. Spontaneous
energy intake in CON was constant throughout the study
(15 000 kJ per day) despite the gradual increase in
training load from week 1 to week 3 (see Fig. 3), resulting
in an energetic deficiency of 1470 and 2800 kJ per
day during weeks 2 and 3, respectively. Conversely, KE
gradually increased energy intake to 17 600 kJ per day,
predominantly via extra carbohydrate intake, resulting
in an energetic balance during both week 2 and week 3
(average energy surplus of 198 kJ per day). It is important
to note that the effect of KE to blunt the appearance of
overreaching symptoms such as decreased heart rates,
elevated urinary noradrenaline excretion and elevated
serum GDF15 level was obvious before any significant
difference in daily energy intake occurred. This suggests
that modulation of food intake per se was not the primary
mechanism of action of KE. Aiming to elucidate the
underlying mechanism, we measured serum levels of the
‘appetite hormones’ leptin, ghrelin and GDF15. GDF15
is a peptide which has only recently been discovered
to act as a stress-induced hormone that is involved in
appetite regulation by decreasing food intake (Johnen
et al. 2007; Macia et al. 2012; Wang-Wei Tsai et al.
2013; Patel et al. 2019). Here, for the first time, we
demonstrate that training overload increased systemic
GDF15 level. Interestingly, this effect was negated by KE
intake, which might at least partly explain the higher
energy intake during training in the latter group. This
observation together with literature data (Patel et al.
2019) indicates that the GDF15 increment in CON
reflects training-induced physiological stress, rather than a
moderate energy deficit. Additional support for this comes
from a recent study in our laboratory showing that 4 weeks
of hypocaloric diet (30% energy deficit), aimed to induce
body weight loss (minus 2–4 kg) in fit lean females, did
not alter systemic GDF15 levels either in the presence or
in the absence of KE supplementation (Hiroux et al.,
unpublished observations). Note that this suppression
of GDF15 occurred in the absence of elevated plasma
D-βHB concentration, indicating adaptation of GDF15
secretion by short-term KE intake. Conversely, neither
leptin nor ghrelin were directly involved in food intake
regulation during the intervention period. Serum leptin,
an appetite suppressor, was even higher during training in
KE than in CON, while ghrelin was unaffected. However,
it is well established that leptin during episodes of energy
deficit operates as a ‘starvation signal’ by decreasing basal
energy expenditure through suppression of heart rate,
blood pressure, thyroid hormone levels and sympathetic
nervous system activity (Flier, 1998; Pandit et al. 2017).
Thus, lower basal metabolic rate effected by a decrease in
leptin (Woods et al. 2017, 2018) might explain the absence
of body weight drop in CON despite energy expenditure
in training exceeding energy intake during the later stage
of the training period.
It is also well known that the dysregulation of
hormonal and energy balance in overtraining, especially in
non-weight-bearing sports such as cycling and swimming,
can stimulate bone demineralization and thereby impair
long-term bone health (Nagle & Brooks, 2011; Olmedillas
et al. 2012). In both experimental groups, we observed a
decrease of the bone formation marker osteocalcin against
stable activity of the osteoclast marker TRAP5b. This
may indicate net bone resorption (Crockett et al. 2011;
Ferreira et al. 2015). These results are in accordance with
a study showing similar symptoms of net bone resorption
occurring in professional cyclists during a 3-week cycling
race, i.e. Giro d’Italia (Lombardi et al. 2012). Non-
etheless, the training intervention did not significantly
alter BMD or BMC measured by a whole-body DXA
scan. This may be due to the short duration of the inter-
vention, because months rather than weeks of strenuous
cycling training are needed to induce a measurable degree
of bone demineralization (Barry & Kohrt, 2008). In
addition, consistent ingestion of a protein–carbohydrate
mixture may be sufficient to maintain bone mineral status
(Townsend et al. 2017). Other studies also have suggested
a link between bone metabolism and energy balance
regulation (Confavreux et al. 2009; Lombardi et al. 2012).
Along such interaction, leptin during training was
positively correlated with BMC (r=0.56). In addition,
a training-induced rise of GDF15 was associated with a
decrease in osteocalcin (r=−0.61), w hich for the first time
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society
J Physiol 597.12 Ketone supplementation in endurance training 3023
indicates that GDF15 may be implicated in the regulation
of bone metabolic activity. Taken together, the above
observations support a tight link between maintenance of
energy balance and bone metabolism during intensified
training. In fact, the beneficial effect of KE on bone
mineral content during the overtraining period might be
largely explained by better matching of energy intake to
energy expenditure in training driven by the concerted
actions of appetite hormones, most prominently
GDF15.
It is well established that low muscle glycogen levels
in endurance athletes impair the capacity to sustain
strenuous training. Based on longitudinal observations
in endurance (over)trained individuals it has been post-
ulated that glycogen depletion may be involved in the
development of overtraining (Costill et al. 1971, 1988).
Against such an opinion, our current observations show
that despite a multiplicity of other symptoms of over-
reaching, resting glycogen levels remained within the
normal range in both groups, confirming previous studies
showing that glycogen depletion is not a prerequisite for
overtraining to develop (Snyder et al. 1995; Halson et al.
2004). In addition, pre- and post-exercise muscleglycogen
contents were similar between CON and KE. Furthermore,
in line with earlier findings (Achten et al. 2004), the
training overload markedly blunted exercise-induced net
muscle glycogen breakdown, which has been attributed
to decreased adrenergic sensitivity (Lehmann et al. 1993;
Jeukendrup & Hesselink, 1994). Again, KE did not impact
exercise-induced net glycogen breakdown during a 30 min
maximal exercise bout. However, the training inter-
vention also substantially reduced peak blood lactate levels
produced by a 90 s all-out exercise bout. This effect faded
within 3 days of recovery in CON, whilst persisting in KE
until day +7.
The overload training programme elicited clear physio-
logical dysregulations as well as negated positive training
adaptations, which clearly indicates a state of functional
overreaching in some, versus non-functional overreaching
in others. Sprint performance decreased in all subjects
(range: 1to12%), whilst endurance performance
in TT30min on average was stable (range: 4to+12%),
which corroborates earlier findings (Woods et al. 2018).
Sprint performance restored to baseline within 3 days of
recovery, while performance improvements in TT30min
only occurred by day +7. Nonetheless, in some sub-
jects performance impairments persisted until the end
of the recovery period (90S: 5to+13%; TT30min :
5to+14%). KE administration did not alter sprint
power output, but clearly stimulated endurance exercise
performance during the final week of the overload period,
as evidenced by a 15% increase in both training load
and EPT120min compared to CON. Note that the observed
increase in endurance performance during the final stage
of the training period coincided with higher rate of daily
carbohydrate intake in KE than in CON. However, this
caused neither different blood glucose levels nor different
muscle glycogen contents between the groups. Non-
etheless, it cannot be excluded that beneficial regulation
of blood glucose turnover during exercise contributed to
explaining the ergogenic effect of KE in EPT120min.
The training overload programme induced many
responses indicating the development of a physiological
overtraining status. Concomitantly, mental well-being
consistently deteriorated. Results from the RESTQ-76
Sport questionnaire indicate impaired general well-being,
attention issues, and both physical and emotionally
exhaustion. The decrement in RESTQ-76 scores in fact
was similar to the effect of a 4-week overload training
period in well-trained triathletes (Coutts et al. 2007), but
was independent of the experimental group. This shows
that in the conditions of the current study RESTQ-76 Sport
scoring was less sensitive than the physiological measures
in identifying the development of overtraining.
The primary aim of the study was to assess the
effect of KE in overtraining. At the same time, this
is the first longitudinal study to explore such a wide
spectrum of physiological responses induced by a period
of deliberate and well-controlled endurance training
overload. Our observations corroborate the prevailing
opinion that overtraining is a complex integrative
response involving dysregulations in multiple systems,
including the autonomous nervous system, energy balance
and hormonal status (Lehmann et al. 1998; Kreher
& Schwartz, 2012), which eventually also result in
immunological impairment. Therefore, identification of
a harmful degree of overreaching by a single biomarker
is conceivably irrelevant (Urhausen & Kindermann,
2002; Hecksteden et al. 2016; Greenham et al. 2018).
Attempts to use hormonal markers of overtraining, such as
ACTH, cortisol, growth hormone, thyroid hormones and
prolactin, have generally failed. Nonetheless, here we show
GDF15 to consistently increase with training overload.
Furthermore, GDF15 changes clearly discriminated the
less (KE) versus themoreovertrained(CON)experimental
group, and both delta and absolute GDF15 was highly
correlated with the training-induced drop of peak exercise
heart rate (r=0.60 and r=0.64, respectively). Moreover,
our data indicate that GDF15 is not only implicated in
the regulation of energy balance during training, but
is probably also implicated in the regulation of bone
metabolism. Clearly GDF15 deserves further attention
as a potential sensitive hormonal marker of overtraining
development.
Previous studies have raised concerns about the
implementation of ketone supplementation because of a
high incidence of gastrointestinal symptoms due to acute
ketone ingestion (Leckey et al. 2017; Vandoorne et al.
2017). However, compared to ketone ester intake, gastro-
intestinal symptoms are more explicit with ingestion of
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society
3024 C. Poff ´
e and others J Physiol 597.12
ketone salts (Veech, 2004). In the conditions of the current
study, involving KE three times daily, the incidence of
gastrointestinal complaints was low and similar between
KE and CON. This suggests that gastrointestinal distress
was due to the overtraining per se (De Oliveira & Burini,
2009) rather than to the supplements.
In conclusion, here we demonstrate that an oral ketone
ester is a potent nutritional strategy that prevents the
development of physiological overtraining symptoms and
non-functional overreaching. In addition, we provide pre-
liminary observations to indicate that GDF15 may be an
adequate hormonal marker for the development of over-
reaching/overtraining.
References
Achten J, Halson SL, Moseley L, Rayson MP, Casey A &
Jeukendrup AE (2004). Higher dietary carbohydrate content
during intensified running training results in better
maintenance of performance and mood state. JApplPhysiol
96, 1331–1340.
Armstrong LE & Vanheest JL (2002). The unknown mechanism
of the overtraining syndrome: clues from depression and
psychoneuroimmunology. Sport Med 32, 185–209.
Aubert G, Martin OJ, Horton JL, Lai L, Vega RB, Leone TC,
Kove s T, G ard ell SJ, K r ¨
uger M, Hoppel CL, Lewandowski
ED, Crawford PA, Muoio DM & Kelly DP (2016). The failing
heart relies on ketone bodies as a fuel. Circulation 133,
698–705.
Balasse EO & F´
ery F (1989). Ketone body production and
disposal: effects of fasting, diabetes, and exercise. Diabetes
Metab Rev 5, 247–270.
Barry DW & Kohrt WM (2008). BMD decreases over the
course of a year in competitive male cyclists. JBoneMiner
Res 23, 484–491.
Bone J & Burke LM (2016). DXA estimates of body
composition and carbohydrate loading. Ann Nutr Metab 68,
228–229.
Borg G (1990). Psychophysical scaling with applications in
physical work and the perception of exertion. Scand J Work
Environ Heal 16, 55–58.
BreenL,PhilpA,WitardOC,JackmanSR,SelbyA,SmithK,
Baar K & Tipton KD (2011). The influence of
carbohydrate–protein co-ingestion following endurance
exercise on myofibrillar and mitochondrial protein
synthesis. JPhysiol589, 4011–4025.
Brownlow ML, Jung SH, Moore RJ, Bechmann N & Jankord R
(2017). Nutritional ketosis affects metabolism and behavior
in Sprague-Dawley rats in both control and chronic stress
environments. Front Mol Neurosci 10, 1–17.
Cadegiani FA & Kater CE (2017). Hormonal aspects of
overtraining syndrome: a systematic review. BMC Sports Sci
Med Rehabil 9, 1–15.
Cahill GF (1976). Starvation in man. Clin Endocrinol Metab 5,
397–415.
Confavreux CB, Levine R & Karsenty G (2009). A paradigm of
integrative physiology, the crosstalk between bone and
energy metabolisms. Mol Cell Endocrinol 310, 21–29.
Costill DL, Bowers R, Branam G & Sparks K (1971). Muscle
glycogen utilization exercise on successive days. JAppl
Physiol 31, 834–838.
Costill DL, Flynn MG, Kirwan JP, Houmard JA, Mitchell JB,
Thomas R & Park SH (1988). Effects of repeated days of
intensified training on muscle glycogen and swimming
performance. Med Sci Sport Exerc 20, 249–254.
Coutts AJ, Wallace LK & Slattery KM (2007). Monitoring
changes in performance, physiology, biochemistry, and
psychology during overreaching and recovery in triathletes.
Int J Sports Med 28, 125–134.
Cox PJ & Clarke K (2014). Acute nutritional ketosis:
implications for exercise performance and metabolism.
Extrem Physiol Med 3, 17.
Cox PJ, Kirk T, Ashmore T, Willerton K, Evans R, Smith A,
MurrayAJ,StubbsB,WestJ,McLureSW,KingMT,Dodd
MS, Holloway C, Neubauer S, Drawer S, Veech RL, Griffin JL
& Clarke K (2016). Nutritional ketosis alters fuel preference
and thereby endurance performance in athletes. Cell Metab
24, 256–268.
CrockettJC,RogersMJ,CoxonFP,HockingLJ&HelfrichMH
(2011). Bone remodelling at a glance. JCellSci124, 991–998.
D’lugos AC, Luden ND, Faller JM, Akers JD, Mckenzie AI &
Saunders MJ (2016). Supplemental protein during heavy
cycling training and recovery impacts skeletal muscle and
heart rate responses but not performance. Nutrients 8, 1–19.
De Oliveira EP & Burini RC (2009). The impact of physical
exercise on the gastrointestinal tract. Curr Opin Clin Nutr
Metab Care 12, 533–538.
EslerM,JenningsG,KornerP,WillettIAN,DudleyF,Hasking
G & Anderson W (1988). Assessment of human sympathetic
nervous system activity from measurements of
norepinephrine spillover. Hypertension 11, 3–20.
Ettema G & Loras HW (2009). Efficiency in cycling: a review.
Eur J Appl Physiol 106, 1–14.
Ferreira A, Alho I, Casimiro S & Costa L (2015). Bone
remodeling markers and bone metastases: from cancer
research to clinical implications. Bonekey Rep 4, 1–9.
Fisher JP (2014). Autonomic control of the heart during
exercise in humans: role of skeletal muscle afferents. Exp
Physiol 99, 300–305.
Flier JS (1998). What’s in a name? In search of leptin’s
physiological role. J Clin Endocrinol Metab 83, 1407–1413.
Fry RW, Grove JR, Morton AR, Zeronit PM, Gaudieri S & Keast
D (1994b). Psychological and immunological correlates of
acute overtraining. Br J Sp Med 28, 241–246.
Fry AC, Kraemer WJ, Van Borselen F, Lynch JM, Triplett NT,
Koziris LP, Fleck SJ & Fleck Cate SJ (1994a). Catecholamine
responses to short-term high-intensity resistance exercise
overtraining. JApplPhysiol77, 941–946.
Fry RW, Morton AR, Garcia-Webb P, Crawford GPM & Keast
D (1992). Biological responses to overload training in
endurance sports. Eur J Appl Physiol Occup Physiol 64,
335–344.
Fry RW, Morton AR & Keast D (1991). Overtraining in
athletes. Sport Med 12, 32–65.
Fry AC, Schilling BK, Weiss LW & Chiu LZF (2006).
B2-Adrenergic receptor downregulation and performance
decrements during high-intensity resistance exercise
overtraining. JApplPhysiol101, 1664–1672.
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society
J Physiol 597.12 Ketone supplementation in endurance training 3025
Gleeson M (2002). Biochemical and immunological responses
to overtraining. JSportSciMed1, 31–41.
Gleeson M & Bishop NC (2000). Modification of immune
responses to exercise by carbohydrate, glutamine and
anti-oxidant supplements. ImmunolCellBiol78, 554–561.
Greenham G, Buckley JD, Garrett J, Eston R & Norton K
(2018). Biomarkers of physiological responses to periods of
intensified, non-resistance-based exercise training in
well-trained male athletes: a systematic review and
meta-analysis. Sport Med 48, 2517–2548.
Halson SL, Lancaster GI, Achten J, Gleeson M & Jeukendrup
AE (2004). Effects of carbohydrate supplementation on
performance and carbohydrate oxidation after intensified
cycling training. JApplPhysiol97, 1245–1253.
Hawley JA, Burke LM, Phillips SM, Spriet LL & Hawley JA
(2011). Nutritional modulation of training-induced skeletal
muscle adaptations. JApplPhysiol110, 834–845.
Hecksteden A, Skorski S, Schwindling S, Hammes D, Pfeiffer
M, Kellmann M, Ferrauti A, Meyer T & Lucia A (2016).
Blood-borne markers of fatigue in competitive athletes –
results from simulated training camps. PLoS One 11, 1–13.
Hedelin R, Wiklund U, Bjerle P & Henriksson-Larsen K (2000).
Cardiac autonomic imbalance in an overtrained athlete. Med
Sci Sports Exerc 32, 1531–1533.
HoldsworthDA,CoxPJ,KirkT,StradlingH,ImpeySG&
Clarke K (2017). A ketone ester drink increases postexercise
muscle glycogen synthesis in humans. Med Sci Sport Exerc
49, 1789–1795.
Hooper SL, Mackinnon LT, Howard A, Gordon RD &
Bachmann AW (1995). Markers for monitoring overtraining
and recovery. Med Sci Sport Exerc 27, 106–112.
Israel S (1976). The problems of overtraining with reference to
performance physiology and internal medicine. Med Sport
16, 1–12 [in German].
Jeukendrup AE & Hesselink MKC (1994). Overtraining – what
do lactate curves tell us? Br J Sp Med 28, 239–240.
Jeukendrup A, Saris WHM, Brouns F & Kester ADM (1996). A
new validated endurance performance test. Med Sci Sports
Exerc 2, 266–270.
Johnen H et al. (2007). Tumor-induced anorexia and weight
loss are mediated by the TGF-βsuperfamily cytokine MIC-1.
Nat Med 13, 1333–1340.
Johnson RH, Walton JL, Krebs HA & Williamson DH (1969).
Metabolic fuels during and after severe exercise in athletes
and non-athletes. Lancet 2, 452–455.
Kellmann M & Kallus KW (2001). Recovery-Stress Questionnaire
for Athletes: User Manual. Human Kinetics, Champaign, IL.
Killer SC, Svendsen IS, Jeukendrup AE & Gleeson M (2017).
Evidence of disturbed sleep and mood state in well-trained
athletes during short-term intensified training with and
without a high carbohydrate nutritional intervention. J
Sports Sci 35, 1402–1410.
Kimura I, Inoue D, Maeda T, Hara T, Ichimura A, Miyauchi S,
Kobayashi M, Hirasawa A, Tsujimoto G & Lefkowitz RJ
(2011). Short-chain fatty acids and ketones directly regulate
sympathetic nervous system via G protein-coupled receptor
41 (GPR41). Proc Natl Acad Sci U S A 108,
8030–8035.
Koninckx E, Van Leemputte M & Hespel P (2008). Effect of a
novel pedal design on maximal power output and
mechanical efficiency in well-trained cyclists. JSportSci26,
1015–1023.
Koninckx E, Van Leemputte M & Hespel P (2010). Effect of
isokinetic cycling versus weight training on maximal power
output and endurance performance in cycling. Eur J Appl
Physiol 109, 699–708.
Kreher JB & Schwartz JB (2012). Overtraining syndrome: a
practical guide. Sports Health 4, 128–138.
Kreider RB, Wilborn CD, Taylor L, Campbell B, Almada AL,
Collins R, Cooke M, Earnest CP, Greenwood M, Kalman DS,
Kerksick CM, Kleiner SM, Leutholtz B, Lopez H, Lowery
LM, Mendel R, Smith A, Spano M, Wildman R, Willoughby
DS, Ziegenfuss TN & Antonio J (2010). ISSN exercise &
sport nutrition review: research & recommendations. JInt
Soc Sports Nutr 7, 1–43.
Kuipers H & Keizer HA (1988). Overtraining in elite athletes,
review and directions for the future. Sport Med 6,
79–92.
Leckey JJ, Ross ML, Quod M, Hawley JA & Burke LM (2017).
Ketone diester ingestion impairs time-trial performance in
professional cyclists. Front Physiol 8, 1–10.
Lehmann M, Dickhuth H, Gendrisch G, Lazar W, Thum M,
Kaminski R, Aramendi J, Peterke E, Wieland W & Keul J
(1991). Training - overtraining. A prospective, experimental
study with experienced middle- and long-distance runners.
Int J Sports Med 12, 444–452.
Lehmann M, Foster C, Dickhuth H-H & Gastmann U (1998).
Autonomic imbalance hypothesis and overtraining
syndrome. Med Sci Sport Exerc 30, 1140–1145.
Lehmann M, Foster C & Keul J (1993). Overtraining in
endurance athletes: a brief review. Med Sci Sport Exerc 25,
854–862.
Lombardi G, Lanteri P, Graziani R, Colombini A, Banfi G,
Corsetti R & Lucia A (2012). Bone and energy metabolism
parameters in professional cyclists during the Giro d’Italia
3-weeks stage race. PLoS One 7, 1–9.
Lowry OH & Passoneau JV (1972). A flexible system of
enzymatic analysis. Academic Press, New York, NY.
Macia L, Wang V, Tsai W, Nguyen AD, Johnen H, Kuffner T,
Shi Y-C, Lin S, Herzog H, Brown DA, Breit SN, Sainsbury A
& Aguila MB (2012). Macrophage inhibitory cytokine 1
(MIC-1/GDF15) decreases food intake, body weight and
improves glucose tolerance in mice on normal & obesogenic
diets. PLoS One 7, 1–8.
MeeusenR,DuclosM,FosterC,FryA,GleesonM,NiemanD,
Raglin J, Rietjens G, Steinacker J & Urhausen A (2013).
Prevention, diagnosis, and treatment of the overtraining
syndrome: joint consensus statement of the european college
of sport science and the American College of Sports
Medicine. Med Sci Sports Exerc 45, 186–205.
Meeusen R & Watson P (2007). Amino acids and the brain: do
they play a role in “central fatigue”? IntJSportNutrExerc
Metab 17, 37–46.
Merry TL & Ristow M (2016). Do antioxidant supplements
interfere with skeletal muscle adaptation to exercise training?
JPhysiol594, 5135–5147.
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society
3026 C. Poff ´
e and others J Physiol 597.12
Morgan WP, Brown DR, Raglin JS, O’Connor PJ & Ellickson
KA (1987). Psychological monitoring of overtraining and
staleness. Br J Sports Med 21, 107–114.
Nagle KB & Brooks MA (2011). A systematic review of bone
health in cyclists. Sports Health 3, 235–243.
Olmedillas H, Gonz´
alez-Ag¨
uero A, Moreno LA, Casajus JA &
Vicente-Rodr´
ıguez G (2012). Cycling and bone health: a
systematic review. BMC Med 10, 1–10.
Owen OE, Morgan AP, Kemp HG, Sullivan JM, Herrera MG &
Cahill GF (1967). Brain metabolism during fasting. JClin
Invest 46, 1589–1595.
Pandit R, Beerens S & Adan RAH (2017). Role of leptin in
energy expenditure: the hypothalamic perspective. Am J
Physiol Regul Integr Comp Physiol 312,
938–947.
Paoli A, Bosco G, Camporesi EM & Mangar D (2015). Ketosis,
ketogenic diet and food intake control: a complex
relationship. Front Psychol 6, 27.
Patel S, Alvarez-guaita A, Melvin A, Coll AP, Savage DB &
Rahilly SO (2019). GDF15 provides an endocrine signal of
nutritional stress in mice and humans. Cell Metab 29,
707–718.
Pawan GLS & Semple SJG (1983). Effect of 3-hydroxybutyrate
in obese subjects on very-low-energy diets and during
therapeutic starvation. Lancet 1, 15–17.
Pfeiffer B, Cotterill A, Grathwohl D, Stellingwerff T &
Jeukendrup AE (2009). The effect of carbohydrate gels on
gastrointestinal tolerance during a 16-km run. Int J Sport
Nutr Exerc Metab 19, 485–503.
Pichot V, Busso T, Roche F, Garet M, Costes F, Duverney D,
Lacour JR & Barth´
el´
emy JC (2002). Autonomic adaptations
to intensive and overload training periods: a laboratory
study. Med Sci Sports Exerc 34, 1660–1666.
Puchalska P & Crawford PA (2017). Multi-dimensional roles of
ketone bodies in fuel metabolism, signaling, and
therapeutics. Cell Metab 25, 262–284.
Robinson AM & Williamson DH (1980). Physiological roles of
ketone bodies as substrates and signals in mammalian
tissues. Physiol Rev 60, 143–187.
Sansone M, Sansone A, Borrione P, Romanelli F, Di Luigi L &
Sgro P (2018). Effects of ketone bodies on endurance
exercise. Curr Sports Med Rep 17, 444–453.
Sherwin RS, Hendler RG & Felig P (1975). Effect of ketone
infusions on amino acid and nitrogen metabolism in man. J
Clin Invest 55, 1382–1390.
Snyder AC, Kuipers H, Cheng B, Servais R & Fransen E (1995).
Overtraining following intensified training with normal
muscle glycogen. Med Sci Sport Exerc 22,
1063–1070.
Stanojevic D, Stojanovic Tosic J & Djorjevic D (2013). Heart
rate modulations in overtraining syndrome. Serbian J Exp
Clin Res 14, 23–28.
Stubbs BJ, Cox PJ, Evans RD, Cyranka M, Clarke K & de Wet H
(2018). A ketone ester drink lowers human ghrelin and
appetite. Obesity 26, 269–273.
StubbsBJ,CoxPJ,EvansRD,SanterP,MillerJJ,FaullOK,
Magor-Elliott S, Hiyama S, Stirling M & Clarke K (2017). On
the metabolism of exogenous ketones in humans. Front
Physiol 8, 848.
Svendsen IS, Killer SC, Carter JM, Randell RK, Jeukendrup AE
& Gleeson M (2016). Impact of intensified training and
carbohydrate supplementation on immunity and markers of
overreaching in highly trained cyclists. Eur J Appl Physiol
116, 867–877.
Thomsen HH, Rittig N, Johannsen M, Møller AB, Jørgensen
JO, Jessen N & Møller N (2018). Effects of 3-hydroxybutyrate
and free fatty acids on muscle protein kinetics and signaling
during LPS-induced inflammation in humans: anticatabolic
impact of ketone bodies. Am J Clin Nutr 108, 1–11.
Townsend R, Elliott-Sale KJ, Currell K, Tang J, Fraser WD &
Sale C (2017). The effect of postexercise carbohydrate and
protein ingestion on bone metabolism. Med Sci Sports Exerc
49, 1209–1218.
Urhausen A & Kindermann W (2002). Diagnosis of
overtraining what tools do we have? Sport Med 32,
95–102.
van Loon LJ, Saris WH, Kruijshoop M & Wagenmakers AJ
(2000). Maximizing postexercise muscle glycogen synthesis:
carbohydrate supplementation and the application of amino
acid or protein hydrolysate mixtures. Am J Clin Nutr 72,
106–111.
Van Thienen R, D’Hulst G, Deldicque L & Hespel P (2014).
Biochemical artifacts in experiments involving repeated
biopsies in the same muscle. Physiol Rep 2, e00286.
Vandoorne T, De Smet S, Ramaekers M, Van Thienen R, De
Bock K, Clarke K & Hespel P (2017). Intake of a ketone ester
drink during recovery from exercise promotes mTORC1
signaling but not glycogen resynthesis in human muscle.
Front Physiol 8, 1–12.
Veech RL (2004). The therapeutic implications of ketone
bodies: the effects of ketone bodies in pathological
conditions: ketosis, ketogenic diet, redox states, insulin
resistance, and mitochondrial metabolism. Prostaglandins,
Leukot Essent Fat Acids 70, 309–319.
Wang-Wei Tsai V, Macia L, Johnen H, Kuffner T, Manadhar R,
BeckJørgensenS,KiMichelleLee-NgK,PingZhangH,Wu
L, Peter Marquis C, Jiang L, Husaini Y, Lin S, Herzog H,
Brown DA, Sainsbury A, Breit SN & Morrison C (2013).
TGF-βsuperfamily cytokine MIC-1/GDF15 is a
physiological appetite and body weight regulator. PLoS One
8, 1–10.
Witard OC, Jackman SR, Kies AK, Jeukendrup AE & Tipton
KD (2011). Effect of increased dietary protein on tolerance
to intensified training. Med Sci Sports Exerc 43,
598–607.
Woods AL, Garvican-Lewis LA, Lundy B, Rice AJ & Thompson
KG (2017). New approaches to determine fatigue in elite
athletes during intensified training: resting metabolic rate
and pacing profile. PLoS One 12, 1–17.
Woods AL, Rice AJ, Garvican-Lewis LA, Wallett AM, Lundy B,
Rogers MA, Welvaert M, Halson S, McKune A & Thompson
KG (2018). The effects of intensified training on resting
metabolic rate (RMR), body composition and performance
in trained cyclists. PLoS One 13, e0191644.
ZouX,MengJ,LiL,HanW,LiC,ZhongR,MiaoX,CaiJ,
Zhang Y & Zhu D (2016). Acetoacetate accelerates muscle
regeneration and ameliorates muscular dystrophy in mice. J
Biol Chem 291, 2181–2195.
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society
J Physiol 597.12 Ketone supplementation in endurance training 3027
Additional information
Competing interests
The authors declare that they have no competing interests.
Author contributions
Conception and design of the study: CP and PH. Biopsy
sampling: RVT. Data collection and biochemical analyses: CP
and MR. Analysis and interpretation of the data and manuscript
drafting: CP and PH. All authors critically evaluated the
manuscript and approved it for submission. All authors agree
to be accountable for all aspects of the work in ensuring that
questions related to the accuracy or integrity of any part of the
work are appropriately investigated and resolved. All persons
designated as authors qualify for authorship, and all those who
qualify for authorship are listed.
Funding
This study was funded by Research Fund Flanders (Fonds voor
Wetenschappelijk Onderzoek – Vlaanderen; research grant no.
G080117N).
Acknowledgements
The authors wish to thank all subjects for their dedicated
cooperation in this demanding trial.
C
2019 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society
... Here, we made within-(n = 9 studies) and betweengroup (n = 12 studies) comparisons to demonstrate potential LEA where EEE has been reported or where training loads have otherwise been characterized (as increase/decrease compared to control condition). Two papers reported actual EA data [19,22], while another four studies [78][79][80][81] reported EEE (along with EI and FFM; thus EA could be directly calculated) and two papers reported total daily EE [82,83]. In four studies [84][85][86][87], we were able to estimate EEE using a metabolic equivalent of task (MET) approach [88] or for running data, utilizing the conversion factor 1 kcal/kg BM/km of running [89]. ...
... no between-group difference) while no REDS or OTS symptoms were shown in the last two studies [81,92]. Only eight studies reported performance outcomes, with a performance impairment (relative to baseline or control group) evident in five studies [21,22,78,86,97], with the rest [79,80,85] showing either no change or no difference between groups. ...
... Nevertheless, the prevalence rates of at least one RED-S component range from ~ 30 to 90% (depending on the sport, type and level of athlete, sex of the athlete, and diagnostic tool [148][149][150][151]). Our analysis showed that 84% (n = 18 studies; Table 2) of training-overload/OTS studies show indications of either LEA and/or low CHO availability with subsequent OTS/RED-S symptoms (n = 14 or 67% of papers; [19, 21, 22, 78, 83-86, 90, 91, 93, 95, 97, 98]) or impaired sports performance (n = 9 or 43% of papers; [21,22,78,86,[93][94][95][96][97]). These findings are not surprising as they are in line with the RED-S prevalence rates in the literature. ...
Article
Full-text available
The symptom similarities between training-overload (with or without an Overtraining Syndrome (OTS) diagnosis) and Relative Energy Deficiency in Sport (RED-S) are significant, with both initiating from a hypothalamic–pituitary origin, that can be influenced by low carbohydrate (CHO) and energy availability (EA). In this narrative review we wish to showcase that many of the negative outcomes of training-overload (with, or without an OTS diagnosis) may be primarily due to misdiagnosed under-fueling, or RED-S, via low EA and/or low CHO availability. Accordingly, we undertook an analysis of training-overload/OTS type studies that have also collected and analyzed for energy intake (EI), CHO, exercise energy expenditure (EEE) and/or EA. Eighteen of the 21 studies (86%) that met our criteria showed indications of an EA decrease or difference between two cohorts within a given study (n = 14 studies) or CHO availability decrease (n = 4 studies) during the training-overload/OTS period, resulting in both training-overload/OTS and RED-S symptom outcomes compared to control conditions. Furthermore, we demonstrate significantly similar symptom overlaps across much of the OTS (n = 57 studies) and RED-S/Female Athlete Triad (n = 88 studies) literature. It is important to note that the prevention of under-recovery is multi-factorial, but many aspects are based around EA and CHO availability. Herein we have demonstrated that OTS and RED-S have many shared pathways, symptoms, and diagnostic complexities. Substantial attention is required to increase the knowledge and awareness of RED-S, and to enhance the diagnostic accuracy of both OTS and RED-S, to allow clinicians to more accurately exclude LEA/RED-S from OTS diagnoses.
... respectively), while average Wingate power output also favored both ketone groups, but not significantly (p = .082). Previous studies evaluating the effects of ketone supplements on exercise performance have been inconsistent, with some showing positive effects (Cox et al., 2016;Kackley et al., 2020;Poffé et al., 2019), some negative effects (Leckey et al., 2017;O'Malley et al., 2017), while others observed no differences (Evans et al., 2019;James & Kjerulf Greer, 2019;Rodger et al., 2017;Waldman et al., 2018). These conflicting results may be caused by differing methodological practices among studies (Margolis & O'fallon, 2020). ...
... It has been suggested that a minimum blood ketone concentration of ≥1 mM is important to elicit positive effects (Margolis & O'fallon, 2020). For instance, some studies have used ketone esters (Cox et al., 2016;Evans et al., 2019;Leckey et al., 2017;Poffé et al., 2019) with doses ranging from 330-573 mg/kg body mass, while others have used ketone salts (James & Kjerulf Greer, 2019;Kackley et al., 2020;O'Malley et al., 2017;Rodger et al., 2017;Waldman et al., 2018) with doses ranging from 88 to 300 mg/kg body mass. The present study used ketone salts (approximately 92 mg/kg body mass) and elevated blood ketones (peaked at 0.7 mM), but did not attain the proposed minimum 1 mM concentration. ...
Article
Full-text available
Coingestion of ketone salts, caffeine and the amino acids, taurine, and leucine improves endurance exercise performance. However, there is no study comparing this coingestion to the same nutrients without caffeine. We assessed whether ketone salts–caffeine–taurine–leucine (KCT) supplementation was superior to caffeine-free ketone salts–taurine–leucine supplementation (KT), or to an isoenergetic carbohydrate placebo (CHO-PLAC). Thirteen recreationally active men (mean ± SD : 177.5 ± 6.1 cm, 75.9 ± 4.6 kg, 23 ± 3 years, 12.0 ± 5.1% body fat) completed a best effort 20-km cycling time-trial, followed 15 min later by a Wingate power cycle test, after supplementing with either KCT (approximately 7 g of beta-hydroxybutyrate, approximately 120 mg of caffeine, 2.1 g of leucine, and 2.7 g of taurine), KT (i.e., same supplement without caffeine), or isoenergetic CHO-PLAC (11 g of dextrose). Blood ketones were elevated ( p < .001) after ingestion of both KCT (0.65 ± 0.12 mmol/L) and KT (0.72 ± 0.31 mmol/L) relative to CHO-PLAC (0.06 ± 0.05 mmol/L). Moreover, KCT improved ( p < .003) 20-km cycling time-trial performance (37.80 ± 2.28 min), compared with CHO-PLAC (39.40 ± 3.33 min) but not versus KT (38.75 ± 2.87 min; p < .09). 20-km cycling time-trial average power output was greater with KCT (power output = 180.5 ± 28.7 W) versus both KT (170.9 ± 31.7 W; p = .049) and CHO-PLAC (164.8 ± 34.7 W; p = .001). Wingate peak power output was also greater for both KCT (1,134 ± 137 W; p = .031) and KT (1,132 ± 128 W; p = .039) versus CHO-PLAC (1,068 ± 127 W). These data suggest that the observed improved exercise performance effects of this multi-ingredient supplement containing beta-hydroxybutyrate salts, taurine, and leucine are attributed partially to the addition of caffeine.
... The combination of hyperketonemia with replete CHO stores may enhance acute endurance exercise capacity (Cox et al., 2016;Poffé et al., 2021), although both null (Dearlove et al., 2019;Evans & Egan, 2018;Evans et al., 2019;Poffé et al., 2020) and negative effects (Leckey et al., 2017) have also been reported. Moreover, prolonged supplementation of ketones during high-intensity exercise training may prevent the deleterious effects of overreaching (Poffé et al., 2019), and has been hypothesized to influence the adaptive response to exercise training (Evans et al., 2017). ...
... Calorie intake during race tended to be increased compared with participants habitual intake with Ex Ket only (~700 Kcal increase with Ex Ket). This finding is consistent with previous work reporting dietary intake in free-living individuals supplemented with the same ketone ester during endurance exercise training (Poffé et al., 2019), but is contrary to F I G U R E 6 Skeletal muscle protein and abdominal subcutaneous adipose tissue gene transcripts. (a) Skeletal muscle pyruvate dehydrogenase kinase 4 (PDK4) fold changes from pre-to post-intervention. ...
Article
Full-text available
Elevating blood ketones may enhance exercise capacity and modulate adaptations to exercise training; however, these effects may depend on whether hyperketonemia is induced endogenously through dietary carbohydrate restriction, or exogenously through ketone supplementation. To determine this, we compared the effects of endogenously‐ and exogenously‐induced hyperketonemia on exercise capacity and adaptation. Trained endurance athletes undertook 6 days of laboratory based cycling (“race”) whilst following either: a carbohydrate‐rich control diet (n = 7; CHO); a carbohydrate‐rich diet + ketone drink four‐times daily (n = 7; Ex Ket); or a ketogenic diet (n = 7; End Ket). Exercise capacity was measured daily, and adaptations in exercise metabolism, exercise physiology and postprandial insulin sensitivity (via an oral glucose tolerance test) were measured before and after dietary interventions. Urinary β‐hydroxybutyrate increased by ⁓150‐fold and ⁓650‐fold versus CHO with Ex Ket and End Ket, respectively. Exercise capacity was increased versus pre‐intervention by ~5% on race day 1 with CHO (p < 0.05), by 6%–8% on days 1, 4, and 6 (all p < 0.05) with Ex Ket and decreased by 48%–57% on all race days (all p > 0.05) with End Ket. There was an ⁓3‐fold increase in fat oxidation from pre‐ to post‐intervention (p < 0.05) with End Ket and increased perceived exercise exertion (p < 0.05). No changes in exercise substrate metabolism occurred with Ex Ket, but participants had blunted postprandial insulin sensitivity (p < 0.05). Dietary carbohydrate restriction and ketone supplementation both induce hyperketonemia; however, these are distinct physiological conditions with contrasting effects on exercise capacity and adaptation to exercise training. Exercise performance and adaptive responses to an endogenously‐ and exogenously‐induced hyperketonemia are markedly different.
... GDF15 has recently been proposed to act as an "exerkine" which is implicated in regulation of acute exercise responses such as the activation of lipolysis, inflammation, and the unfolded protein response (77,78). In this regard, our present and earlier (42,79) reports indicate that the independent or concerted actions of KE and BIC potentially might provide an interesting nutritional tool to modulate training adaptation. Along such perspective, we previously demonstrated consistent KE intake to attenuate the gradual increase in serum GDF15 during training overload, and to blunt the development of overreaching symptoms resulting in enhanced endurance exercise performance (79). ...
... In this regard, our present and earlier (42,79) reports indicate that the independent or concerted actions of KE and BIC potentially might provide an interesting nutritional tool to modulate training adaptation. Along such perspective, we previously demonstrated consistent KE intake to attenuate the gradual increase in serum GDF15 during training overload, and to blunt the development of overreaching symptoms resulting in enhanced endurance exercise performance (79). ...
Article
Full-text available
Available evidence indicates that elevated blood ketones are associated with improved hypoxic tolerance in rodents. From this perspective, we hypothesized that exogenous ketosis by oral intake of the ketone ester (R)-3-hydroxybutyl (R)-3-hydroxybutyrate (KE) may induce beneficial physiological effects during prolonged exercise in acute hypoxia. As we recently demonstrated KE to deplete blood bicarbonate, which per se may alter the physiological response to hypoxia, we evaluated the effect of KE both in the presence and absence of bicarbonate intake (BIC). Fourteen highly trained male cyclists performed a simulated cycling race (RACE) consisting of 3h intermittent cycling (IMT 180' ) followed by a 15-min time-trial (TT 15' ) and an all-out sprint at 175% of lactate threshold (SPRINT). During RACE, fraction of inspired oxygen (F i O 2 ) was gradually decreased from 18.6 to 14.5%. Before and during RACE, participants received either i) 75g ketone ester (KE), ii) 300 mg/kg body mass bicarbonate (BIC), iii) KE+BIC or iv) a control drink in addition to 60g carbohydrates per h in a randomized, crossover design. KE counteracted the hypoxia-induced drop in blood (SpO 2 ) and muscle oxygenation by ~3%. In contrast, BIC decreased SpO 2 by ~2% without impacting muscle oxygenation. Performance during TT 15' and SPRINT were similar between all conditions. In conclusion, KE slightly elevated the degree of blood and muscle oxygenation during prolonged exercise in moderate hypoxia without impacting exercise performance. Our data warrant to further investigate the potential of exogenous ketosis to improve muscular and cerebral oxygenation status, and exercise tolerance in extreme hypoxia.
... A potentially safer alternative for ketone salts is the ketone ester 3-hydroxybutyl-3-hydroxybutanoate (3HHB, Fig. 1a) which is endogenously converted to two 3HB − molecules 11 . Sustained nutritional ketosis through oral ingestion of 3HHB has previously shown to be safe in healthy adults 12 and has been extensively studied in elite athletes for its potential to improve exercise performance and endurance [13][14][15] . ...
Article
Full-text available
In septic mice, 3-hydroxybutyrate-sodium-salt has shown to partially prevent sepsis-induced muscle weakness. Although effective, the excessive sodium load was toxic. We here investigated whether ketone ester 3-hydroxybutyl-3-hydroxybutanoate (3HHB) was a safer alternative. In a mouse model of abdominal sepsis, the effects of increasing bolus doses of 3HHB enantiomers on mortality, morbidity and muscle force were investigated (n = 376). Next, plasma 3HB- clearance after bolus d-3HHB was investigated (n = 27). Subsequently, in septic mice, the effect on mortality and muscle force of a continuous d,l-3HHB infusion was investigated (n = 72). In septic mice, as compared with placebo, muscle force was increased at 20 mmol/kg/day l-3HHB and at 40 mmol/kg/day d- and d,l-3HHB. However, severity of illness and mortality was increased by doubling the effective bolus doses. Bolus 3HHB caused a higher 3HB− plasma peak and slower clearance with sepsis. Unlike bolus injections, continuous infusion of d,l-3HHB did not increase severity of illness or mortality, while remaining effective in improving muscle force. Treatment of septic mice with the ketone ester 3HHB partly prevented muscle weakness. Toxicity of 3HHB administered as bolus was completely avoided by continuous infusion of the same dose. Whether continuous infusion of ketone esters represents a promising intervention to also prevent ICU-acquired weakness in human patients should be investigated.
... # Difference between groups within intervention point, p < 0.05, ## p < 0.01 has also been proposed as a marker of overtraining, indicated by the bivariate correlation between circulating GDF15 and a decrease in maximal heart rate after a 3-week exercise intervention. 34 In the present study, we made a similar observation, as the linear regression between change in circulating GDF15 (CPH -PMO + 2) and change in maximal heart rate (CPH -PMO + 2) was robust (r 2 = 0.64, p = 0.001, data not shown). Together, these data show that despite an energy deficit (avg. ...
Article
Full-text available
Aim & methods: Extreme endurance exercise provides a valuable research model for understanding the adaptive metabolic response of older and younger individuals to intense physical activity. Here, we compare a wide range of metabolic and physiologic parameters in two cohorts of seven trained men, age 30±5 years or age 65±6 years, before and after the participants travelled ≈3000 km by bicycle over 15 days. Results: Over the 15-day exercise intervention, participants lost 2-3 kg fat mass with no significant change in body weight. V̇O2 max did not change in younger cyclists, but decreased (p=0.06) in the older cohort. The resting plasma FFA concentration decreased markedly in both groups, and plasma glucose increased in the younger group. In the older cohort, plasma LDL-cholesterol and plasma triglyceride decreased. In skeletal muscle, fat transporters CD36 and FABPm remained unchanged. The glucose handling proteins GLUT4 and SNAP23 increased in both groups. Mitochondrial ROS production decreased in both groups and ADP sensitivity increased in skeletal muscle in the older but not in the younger cohort. Conclusion: In summary, these data suggest that older but not younger individuals experience a negative adaptive response affecting cardiovascular function in response to extreme endurance exercise, while a positive response to the same exercise intervention is observed in peripheral tissues in younger and older men. The results also suggest that the adaptive thresholds differ in younger and old men, and this difference primarily affects central cardiovascular functions in older men after extreme endurance exercise.
Article
We have previously demonstrated that exogenous ketosis reduces urine production during exercise. However, the underlying physiological mechanism of this anti-diuretic effect remained unclear. Therefore, we investigated whether acute exogenous ketosis by oral ingestion of ketone ester (KE) during a simulated cycling race (RACE) affects the hormonal pathways implicated in fluid balance regulation during exercise. In a double-blind crossover design, 11 well-trained male cyclists participated in RACE consisting of a 3-h submaximal intermittent cycling (IMT 180' ) bout followed by a 15-minute time trial (TT 15' ) in an environmental chamber set at 28 °C and 60 % relative humidity. Fluid intake was adjusted to maintain euhydration. Before and during RACE, the subjects received either a control drink (CON) or the ketone ester (R)-3-hydroxybutyl (R)-3-hydroxybutyrate (KE), which elevated blood β-hydroxybutyrate to ~2-4 mM. Urine output during IMT 180' was ~20% lower in KE (1172 ± 557 ml) than in CON (1431 ± 548 ml, p < 0.05). Compared with CON, N-terminal pro-atrial natriuretic peptide (NT-pro ANP) concentration during RACE was ~20% lower in KE (p < 0.05). KE also raised plasma noradrenaline concentrations during RACE. Performance in TT 15' was similar between CON and KE. In conclusion, exogenous ketosis suppresses diuresis and downregulates α-natriuretic peptide activity during exercise.
Article
Objectives Exogenous ketone (monoester or salt) supplements are increasingly being employed for a variety of research purposes and marketed amongst the general public for their ability to raise blood beta-hydroxybutyrate (β-OHB). Emerging research suggests a blood glucose-lowering effect of exogenous ketones. Here, we systematically review and meta-analyze the available evidence of trials reporting on exogenous ketones and blood glucose. Methods We searched 6 electronic databases on December 13, 2021 for trials of any length that reported on the use of exogenous ketones compared to a placebo. We pooled raw mean differences (MD) in (i) blood β-OHB and (ii) blood glucose using random-effects models, and explored differences in the effects of ketone salts compared to ketone monoesters. Publication bias and risk of bias were examined using funnel plots and Cochrane's risk-of-bias tool, respectively. Results Twenty-eight trials including a total of 332 participants met inclusion criteria. There was no evidence for publication bias. Four trials were judged to be at low risk of bias with some concern for risk of bias in the remaining trials. Compared to placebo, consumption of exogenous ketones raised blood β-OHB (MD = 1.98 mM; 95% CI: 1.52 mM, 2.45 mM; P < 0.001) and decreased blood glucose (MD = −0.47 mM; 95% CI: −0.57 mM, −0.36 mM; P < 0.001) across the post-supplementation period of up to 300 minutes. Across both analyses, significantly greater effects were found following ingestion of ketone monoesters compared to ketone salts (P < 0.001). Conclusions Consumption of exogenous ketone supplements leads to acutely increased blood β-OHB and decreased blood glucose. Ketone monoesters exert a more potent β-OHB-raising and glucose-lowering effect as compared to ketone salts. Funding Sources Michael Smith Foundation for Health Research (MSFHR) Scholar Award.
Article
Recently developed ketone (monoester or salt) supplements acutely elevate blood β-hydroxybutyrate (BHB) exogenously without prolonged periods of fasting or carbohydrate restriction. Previous (small-scale) studies have found a blood glucose-lowering effect of exogenous ketones. This study aimed to systematically review available evidence and conduct meta-analyses of studies reporting on exogenous ketones and blood glucose. We searched 6 electronic databases on December 13, 2021 for randomized and non-randomized trials of any length that reported on the use of exogenous ketones. We calculated raw mean differences (MD) in blood BHB and glucose in two main analyses: (I) after compared to before acute ingestion of exogenous ketones, and (II) following acute ingestion of exogenous ketones compared to a comparator supplement. We pooled effect sizes using random-effects models and performed prespecified subgroup analyses to examine the effect of potential explanatory factors, including study population, exercise, blood BHB, and supplement type, dosing, and timing. Risk of bias was examined using Cochrane's risk-of-bias tools. Studies that could not be meta-analyzed were summarized narratively. Forty-three trials including 586 participants are summarized in this review. Following ingestion, exogenous ketones increased blood BHB (MD = 1.73 mM, 95% CI: 1.26 mM to 2.21 mM, P < 0.001) and decreased mean blood glucose (MD = -0.54 mM, 95% CI: -0.68 mM to -0.40 mM, P < 0.001). Similarly, when compared to placebo, blood BHB increased (MD = 1.98 mM, 95% CI: 1.52 mM to 2.45 mM, P < 0.001) and blood glucose decreased (MD = -0.47 mM, 95% CI: -0.57 mM to -0.36 mM, P < 0.001). Across both analyses, significantly greater effects were seen with ketone monoesters compared to salts (P < 0.001). The available evidence indicates that acute ingestion of exogenous ketones leads to increased blood BHB and decreased blood glucose. Limited evidence on prolonged ketone supplementation was found.
Article
Endurance athletes may implement rigid dietary strategies, such as the ketogenic diet (KD), to improve performance. The effect of the KD on appetite remains unclear in endurance athletes. This study analyzed the effects of a KD, a high-carbohydrate diet (HCD), and habitual diet (HD) on objective and subjective measures of appetite in trained cyclists and triathletes, and hypothesized that the KD would result in greater objective and subjective appetite suppression. Six participants consumed the KD and HCD for two-weeks each, in a random order, following their HD. Fasting appetite measures were collected after two-weeks on each diet. Postprandial appetite measures were collected following consumption of a ketogenic meal after the KD, high-carbohydrate meal after the HCD, and standard American/Western meal after the HD. Fasting total ghrelin (GHR) was lower and glucagon-like peptide-1 (GLP-1) and hunger were higher following the KD versus HD and HCD. Fasting insulin was not different. Mixed-effects model repeated measures analysis and effect sizes and 95% confidence intervals showed that postprandial GHR and insulin were lower and GLP-1 was higher following the ketogenic versus the standard and high-carbohydrate meals. Postprandial appetite ratings were not different across test meals. In conclusion, both fasting and postprandial concentrations of GHR were lower and GLP-1 were higher following the KD than the HC and HD, and postprandial insulin was lower on the KD. Subjective ratings of appetite did not correspond with the objective measures of appetite in trained competitive endurance athlete. More research is needed to confirm our findings.
Article
Full-text available
GDF15 is an established biomarker of cellular stress. The fact that it signals via a specific hindbrain receptor, GFRAL, and that mice lacking GDF15 manifest diet-induced obesity suggest that GDF15 may play a physiological role in energy balance. We performed experiments in humans, mice, and cells to determine if and how nutritional perturbations modify GDF15 expression. Circulating GDF15 levels manifest very modest changes in response to moderate caloric surpluses or deficits in mice or humans, differentiating it from classical intestinally derived satiety hormones and leptin. However, GDF15 levels do increase following sustained high-fat feeding or dietary amino acid imbalance in mice. We demonstrate that GDF15 expression is regulated by the integrated stress response and is induced in selected tissues in mice in these settings. Finally, we show that pharmacological GDF15 administration to mice can trigger conditioned taste aversion, suggesting that GDF15 may induce an aversive response to nutritional stress.
Article
Full-text available
Background Intensified training is important for inducing adaptations to improve athletic performance, but detrimental performance effects can occur if prescribed inappropriately. Monitoring biomarker responses to training may inform changes in training load to optimize performance. Objective This systematic review and meta-analysis aimed to identify biomarkers associated with altered exercise performance following intensified training. Methods Embase, MEDLINE, CINAHL, Scopus and SPORTDiscus were searched up until September 2017. Included articles were peer reviewed and reported on biomarkers collected at rest in well-trained male athletes before and after periods of intensified training. Results The full text of 161 articles was reviewed, with 59 included (708 participants) and 42 (550 participants) meta-analysed. In total, 118 biomarkers were evaluated, with most being cellular communication and immunity markers (n = 54). Studies most frequently measured cortisol (n = 34), creatine kinase (n = 25) and testosterone (n = 20). Many studies reported decreased immune cell counts following intensified training, irrespective of performance. Moreover, reduced performance was associated with a decrease in neutrophils (d = − 0.57; 95% confidence interval (CI) − 1.07 to − 0.07) and glutamine (d = − 0.37; 95% CI − 0.43 to − 0.31) and an increase in urea concentration (d = 0.80; 95% CI 0.30 to 1.30). In contrast, increased performance was associated with an increased testosterone:cortisol ratio (d = 0.89; 95% CI 0.54 to 1.24). All remaining biomarkers showed no consistent patterns of change with performance. Conclusions Many biomarkers were altered with intensified training but not in a manner related to changes in exercise performance. Neutrophils, glutamine, urea and the testosterone:cortisol ratio exhibited some evidence of directional changes that corresponded with performance changes therefore indicating potential to track performance. Additional investigations of the potential for these markers to track altered performance are warranted.