J Appl Physiol 111: 1372–1379, 2011.
First published August 11, 2011; doi:10.1152/japplphysiol.01028.2010.
Caffeine intake improves intense intermittent exercise performance
and reduces muscle interstitial potassium accumulation
Magni Mohr,1,2 Jens Jung Nielsen,1 and Jens Bangsbo1
1
Copenhagen Muscle Research Centre, Department of Exercise and Sport Sciences, University of Copenhagen, Denmark;
and 2Sport and Health Sciences, College of Life and Environmental Sciences, St. Lukes Campus, University of Exeter, Exeter,
United Kingdom
Submitted 31 August 2010; accepted in final form 8 August 2011
Mohr M, Nielsen JJ, Bangsbo J. Caffeine intake improves intense
intermittent exercise performance and reduces muscle interstitial po-
tassium accumulation. J Appl Physiol 111: 1372–1379, 2011. First
published August 11, 2011; doi:10.1152/japplphysiol.01028.2010.—
The effect of oral caffeine ingestion on intense intermittent exercise
performance and muscle interstitial ion concentrations was examined.
The study consists of two studies (S1 and S2). In S1, 12 subjects
completed the Yo-Yo intermittent recovery level 2 (Yo-Yo IR2) test
with prior caffeine (6 mg/kg body wt; CAF) or placebo (PLA) intake.
In S2, 6 subjects performed one low-intensity (20 W) and three
intense (50 W) 3-min (separated by 5 min) one-legged knee-extension
exercise bouts with (CAF) and without (CON) prior caffeine supple-
mentation for determination of muscle interstitial K⫹ and Na⫹ with
microdialysis. In S1 Yo-Yo IR2 performance was 16% better (P ⬍
0.05) in CAF compared with PLA. In CAF, plasma K⫹ at the end of
the Yo-Yo IR2 test was 5.2 ⫾ 0.1 mmol/l with no difference between
the trials. Plasma free fatty acids (FFA) were higher (P ⬍ 0.05) in
CAF than PLA at rest and remained higher (P ⬍ 0.05) during
exercise. Peak blood glucose (8.0 ⫾ 0.6 vs. 6.2 ⫾ 0.4 mmol/l) and
plasma NH3 (137.2 ⫾ 10.8 vs. 113.4 ⫾ 13.3 ␮mol/l) were also higher
(P ⬍ 0.05) in CAF compared with PLA. In S2 interstitial K⫹ was
5.5 ⫾ 0.3, 5.7 ⫾ 0.3, 5.8 ⫾ 0.5, and 5.5 ⫾ 0.3 mmol/l at the end of
the 20-W and three 50-W periods, respectively, in CAF, which were
lower (P ⬍ 0.001) than in CON (7.0 ⫾ 0.6, 7.5 ⫾ 0.7, 7.5 ⫾ 0.4, and
7.0 ⫾ 0.6 mmol/l, respectively). No differences in interstitial Na⫹
were observed between CAF and CON. In conclusion, caffeine intake
enhances fatigue resistance and reduces muscle interstitial K⫹ during
intense intermittent exercise.
fatigue; potassium; free fatty acids; repeated exercise; muscle ion
kinetics
CAFFEINE is classified as part of the methylxanthine family of
drugs and is commonly consumed by athletes as an ergogenic
aid (45), since it was removed from the World Anti-Doping
Agency list of prohibited substances (46). A caffeine ingestion
of 3– 6 mg/kg body mass has been shown to improve fatigue
resistance during various types of exercise such as time trial
cycling (35), sprint running (20, 44), and repeated sprint
running (11). In trained athletes caffeine improved 1 km
cycling performance (47) whereas caffeine had no effect on
either muscle strength and power (4, 47) or 200-m swimming
(40) in trained athletes. In addition, no effect of caffeine intake
was seen in power output during intense cycling exercise (24).
In a study by Stuart et al. (44) caffeine administration 1 h
before a simulated rugby union game improved passing accu-
racy and 30-m sprint speed. Thus the performance-enhancing
Address for reprint requests and other correspondence: J. Bangsbo, The
August Krogh Bldg., IFI, Universitetsparken 13, DK-2100 Copenhagen Ø,
Denmark (e-mail: jbangsbo@ifi.ku.dk).
1372 8750-7587/11 Copyright © 2011 the American Physiological Society
effects of caffeine administration appear to be specific to the
type of exercise performed, but the effect on repeated intense
exercise has not been determined. The Yo-Yo intermittent
recovery test, level 2 (Yo-Yo-IR2), is well-described (7, 31,
37) and has been shown to have a highly anaerobic nature, and
test performance correlates with the amount of intense exercise
performed in team sports (7). Thus, to examine the effect of
Downloaded from http://jap.physiology.org/ by 10.220.33.3 on June 7, 2017
caffeine intake on intense intermittent exercise as observed in
team sport, the Yo-Yo IR2 test can be applied.
Caffeine has a number of physiological effects. It been
shown to elevate the catecholamine levels (22, 24, 32, 44), the
systemic FFA (21), NH3 concentrations, sarcoplasmatic retic-
ulum Ca2⫹ handling (1, 17) as well as lower the respiratory
exchange ratio (RER) values (5). Caffeine may also increase
the muscle Na⫹-K⫹-ATPase activity directly or indirectly
through an elevated catecholamine response, which may delay
fatigue if accumulation of potassium in the muscle interstitium
is involved in fatigue (27, 36, 38, 39, 42). Indeed, Lindinger et
al. (32) and Graham and colleagues (23) demonstrated that
high doses (6 and 9 mg/kg body wt) of caffeine administration
decreased plasma K⫹ concentration during prolonged exercise,
but not during repeated 30-s maximal exercise bouts (24).
However, the effect of caffeine on plasma potassium is con-
troversial, and no data exist about the effect of caffeine on
muscle interstitial potassium and sodium.
Thus the aim of the present study was to examine the impact
of caffeine intake on performance and physiological response,
including changes in muscle interstitial ion concentrations,
during repeated intense exercise.
MATERIALS AND METHODS
Subjects
The study consists of two studies (S1 and S2). The participants in
S1 were eight males and four females with an age of 23.1 ⫾ 0.6 yr,
height of 179.0 ⫾ 3.0 cm, and body mass of 72.6 ⫾ 3.7 kg. They were
all active in team sports at a subelite level and were familiar with
intense intermittent exercise. In S2 six habitually active male subjects
(age: 25.9 ⫾ 1.0 yr; height: 183.0 ⫾ 2.8 cm; body mass: 83.8 ⫾ 3.8
kg) participated. The subjects in S2 had a maximal oxygen uptake of
54.0 ⫾ 1.7 ml O2·kg⫺1·min⫺1. All subjects were informed of risks
associated with the experiment before giving their written consent to
participate. The study conforms to the code of Ethics of World
Medical Association (Declaration of Helsinki) and was approved by
the Ethics Committee of Copenhagen and Frederiksberg communities.
Experimental Design
In S1 the subjects performed the Yo-Yo intermittent recovery test,
level 2 (Yo-Yo IR2), until exhaustion (31) on two separate occasions
separated by 6 days. The tests were carried out at the same time of the
http://www.jap.org
 CAFFEINE AFFECTS PERFORMANCE AND INTERSTITIAL POTASSIUM
day. Fluid and food intake as well as activity level was noted the day
before the first experimental day and mimicked before the second day.
Seventy minutes before the test either caffeine (CAF; MERCK,
Darmstadt, Germany) or placebo (PLA; dextrose) was taken orally in
a gelatin capsule (6 mg/kg body wt; 436 ⫾ 22 mg in total). The blood
glucose concentration was unaffected by the placebo intake. The
experiment was randomized, double-blinded, and counterbalanced.
In S2 the subjects performed a control (CON) and a caffeine (CAF)
trial separated by 80 min. For each trial the subjects performed a
one-legged knee-extensor exercise protocol consisting of two 20-W
exercise periods separated by 10 min of passive recovery, then after
10 min of recovery the subjects completed three 3-min exercise bouts
(EX1, EX2, and EX3) at an intensity of 50 W interspersed by 5 min
of passive recovery (Fig. 1). The first 20-W period was carried out to
obtain a higher reliability of the interstitial K⫹ measurements (39) and
only the second 20-W period was included. The CON trial was
completed first, and after 10 min of recovery caffeine was adminis-
trated orally in a gelatin capsule at a concentration of 6 mg/kg body
wt corresponding to 516 ⫾ 21 mg in total. Then the subject rested for
70 min before starting the CAF-trial, which is sufficient for the muscle
to recover from the exercise performed in CON. The CON and the
CAF trial were performed on the same day to ensure that the
microdialysis probes were placed in the same position. The CON trial
was performed before the CAF trial to avoid possible effects of
caffeine in the CON trial. None of the subjects were heavy coffee
drinkers.
Experimental Protocol
S1. All subjects were familiar with the Yo-Yo IR2 test and had
performed at least two familiarization trials. On the experimental day
the subjects reported to the laboratory in the morning after consuming
a light meal. The subjects were instructed to avoid intake of caffeine
containing food and drinks on the day of the experiment as well as
alcohol consumption and heavy physical activity on the day before the
experiment. Thus no caffeine was ingested in the last 12–15 h before
the experiment. A catheter (Venflon Pro, 18G) was inserted into an
Fig. 1. Schematic overview of the experimental protocol.
J Appl Physiol • VOL 111 • NOVEMBER 2011 •
1373
antecubital vein ⬃30 min before the exercise protocol was initiated
for frequent blood sampling before, during, and after the test. The first
5 min of the Yo-Yo intermittent recovery test, level 1 (Yo-Yo IR1),
was used as warm-up before the Yo-Yo IR2 test. This procedure has
been shown to markedly elevate the muscle temperature (30, 31). The
Yo-Yo IR2 test was initiated 2 min after the warm-up. A blood sample
was taken at rest, after the warm-up period, after 160, 280, 440, 600,
and 760 s, and at exhaustion in the Yo-Yo IR2 test, as well as 1, 2 and
5 min into recovery.
S2. One-legged knee-extensor exercise was carried out on a mod-
ified Krogh ergometer permitting the exercise to be confined to the
quadriceps muscle (3). The exercise is performed as knee extensions
from an angle of 90° to ⬃170° while flywheel momentum repositions
the relaxed leg during the flexion. EMG recordings have demonstrated
that the hamstring muscles are passive, while the quadriceps muscles
are active in the extension phase (3). To familiarize the subjects with
the exercise ergometer the subjects completed 3– 4 preexperimental
sessions. The subjects were trained to keep a constant kicking fre-
quency of 60 kicks/min (by visual feedback from a display showing
the frequency), and to confine the exercise to the quadriceps femoris
muscle (by verbal and visual feedback based on online force record-
Downloaded from http://jap.physiology.org/ by 10.220.33.3 on June 7, 2017
ings).
On the experimental day the subjects reported to the laboratory in
the morning and followed the same guidelines as described in S1. Six
microdialysis probes were inserted parallel to the muscle fibers of the
vastus lateralis muscle in the experimental leg under local anesthesia
(lidocaine, 1 ml of 20 mg/ml) as previously described (27, 36, 39).
The probes were custom design (length of microdialysis membrane
30 – 40 mm, outer diameter 0.22 mm) as described by Hellsten et al.
(25). The perfusate was Ringer acetate containing (in mM) 130 Na⫹,
2 Ca2⫹, 4 K⫹, 1 Mg2⫹, and 30 Ac⫺. The exact perfusate K⫹
concentration was determined in each experiment. In addition, the
perfusate contained 3 mM of glucose and 201Tl (activity ⬍ 7,000
Bq/ml). The probes were flushed and connected to a pump (CMA
102). Then they were perfused at a rate of 2 ␮l/min for 1 h and 5
␮l/min for the remaining part of the experiment. In addition, a catheter
www.jap.org
1374 CAFFEINE AFFECTS PERFORMANCE AND INTERSTITIAL POTASSIUM
(Venflon Pro, 18G) was placed in an antecubital vein for blood
sampling.
After insertion of the microdialysis probes the subjects rested for at
least 60 min followed by the control trial (CON). Before, during, and
10 min after the second 20-W period dialysate was sampled. During
exercise samples were collected at 0.5–2.5, 2.5– 4.5, 4.5– 6.5, and
6.5–10 min taking into account the delay in collecting the sample from
muscle interstitium (⬃30 s) due to the space in the outlet tubing. In
recovery dialysate was collected after 0.5–2.5, 2.5– 4.5, 4.5– 6.5, and
6.5–10 min. During EX1, EX2, and EX3 dialysate was collected in
1.5-min intervals starting at 0.5 min of exercise. In recovery dialysate
was collected after 0.5–2, 2–3.5, and 3.5–5 min. Blood samples were
drawn at rest, after the second 20-W period, and after all three 50-W
periods (Fig. 1). After CON caffeine was administrated and the
procedures in CAF were the same as in CON. No placebo was given
before CON.
Analysis of Dialysate
The perfusion rate of the individual microdialysis probes was
determined throughout the experiment by weighing the sample tube
before and after collection. Samples were excluded if the perfusion
rate deviated by ⬎30% from the target perfusion rate, or if there was
any sign of hemoglobin in the dialysate. Interstitial K⫹ ([K⫹]int)
(interstitial K⫹) was measured by flame photometry (FLM3, Radiom-
eter, Denmark) using lithium as internal standard. Thallium loss was
determined as previously described (27).
Calculations
The relative loss (RL) of 201Tl was calculated as RLTl ⫽ (perfusate
activity - dialysate activity)/perfusate activity. [K⫹]int was calculated
from the dialysate potassium concentrations, assuming that fractional
201
TI loss from the perfusate was equal to the fractional gain of K⫹ to
the perfusate : [K⫹]int ⫽ K⫹perfusate ⫹ [(K⫹dialysate ⫺ K⫹perfusate)/
RLTl] (27).
Blood Analyses
Blood samples were obtained in 2-ml syringes with heparin. Im-
mediately after sampling, 100 ␮l of heparinized blood was hemolyzed
in an ice-cold 100 ␮l Triton X-100 buffer solution and stored on ice
for ⬃30 min until analyzed for lactate and glucose (YSI 2300, Yellow
Spring Instruments, OH). Plasma was collected from centrifuged
samples and stored at ⫺80°C until further analysis. Plasma FFA,
NH3, potassium, and sodium were analyzed using Hitachi 912 auto-
matic analyzer (Roche Diagnostics, Germany). FFA was determined
fluorometrically after an enzymatic reaction using a kit (WAKO
NEFA C ACS-ACOD Method, Trichem), and NH3 was determined
fluorometrically after an enzymatic reaction using a kit (Roche Diag-
nostics, Germany). Potassium and sodium were determined by the use
of ion selective electrodes (Roche Diagnostics).
Statistics
Values are means ⫾ SE. For each subject in S2 an average [K⫹]int
and [Na⫹]int was calculated from the measurements of each probe.
Mean [K⫹]int and SE were determined from these individual means.
Possible differences in muscle interstitial and blood variables between
CON and CAF as well as between PLA and CAF were tested using a
two-way ANOVA for repeated measurements. In case of significant
main effects a Student-Newman-Keuls post hoc test was used to
identify the points of difference. Significance was set at the 0.05 level.
RESULTS
Yo-Yo IR2 Performance
Yo-Yo IR2 performance was 607 ⫾ 55 m in CAF, which
was 16% better (P ⬍ 0.05) than in PLA (523 ⫾ 55 m).
J Appl Physiol • VOL 111 • NOVEMBER 2011 •
Physiological Response During the Yo-Yo IR2 Test
Plasma K⫹ was 3.6 ⫾ 0.1 and 3.9 ⫾ 0.1 mmol/l at rest in
CAF and PLA, respectively, and rose (P ⬍ 0.05) during the
Yo-Yo IR2 test to 5.2 ⫾ 0.2 and 5.9 ⫾ 0.4 mmol/l at
exhaustion, respectively, with no differences between the two
trials (Fig. 2A). Also peak plasma potassium in CAF and PLA
was not different (5.3 ⫾ 0.2 and 5.9 ⫾ 0.3 mmol/l, respec-
tively). In CAF blood lactate increased (P ⬍ 0.05) from 1.0 ⫾
0.1 to 8.6 ⫾ 0.7 mmol/l during the test with no differences
between the two trials (Fig. 2B). The peak blood lactate
concentration was 11.2 ⫾ 0.6 and 10.8 ⫾ 0.8 mmol/l in CAF
and PLA, respectively.
Blood glucose was 5.8 ⫾ 0.6 mmol/l before the test in CAF,
which tended to be higher (P ⫽ 0.06) than in PLA (4.5 ⫾ 0.5
mmol/l; Fig. 2C). The blood glucose levels at exhaustion were
6.4 ⫾ 0.5 mmol/l in CAF, which was not significantly different
from PLA (4.9 ⫾ 0.3 mmol/l). Peak blood glucose level during
the test was 8.1 ⫾ 0.6 mmol/l in CAF and higher (P ⬍ 0.01)
than in PLA (6.2 ⫾ 0.5 mmol/l). Plasma NH3 was 32.7 ⫾ 5.5
Downloaded from http://jap.physiology.org/ by 10.220.33.3 on June 7, 2017
and 30.0 ⫾ 3.1 ␮mol/l at rest in CAF and PLA, respectively,
and increased (P ⬍ 0.05) during the Yo-Yo IR2 test to 76.7 ⫾
9.8 and 94.0 ⫾ 18.4 ␮mol/l, respectively, with no differences
between trials (Fig. 2D). However, 5 min into recovery the
plasma NH3 was 137.2 ⫾ 10.8 ␮mol/l in CAF, which was
higher (P ⬍ 0.05) than in PLA (105.1 ⫾ 10.7 ␮mol/l). At rest
plasma FFA was 490 ⫾ 77 ␮mol/l in CAF, which was higher
(P ⬍ 0.05) than in PLA (167 ⫾ 31 ␮mol/l; Fig. 2E), and
remained higher (P ⬍ 0.05) in CAF than in PLA during the
entire experiment (Fig. 2E).
Muscle Interstitial Potassium and Sodium
At rest [K⫹]int was 4.3 ⫾ 0.2 and 4.2 ⫾ 0.5 mmol/l in CAF
and CON, respectively. During the 20-W knee-extensor exer-
cise [K⫹]int was lower (P ⬍ 0.05) in CAF than CON and
remained lower (P ⬍ 0.05) during the 10-min recovery period
(Fig. 3A). At rest [Na⫹]int was 134 ⫾ 6 and 134 ⫾ 3 mmol/l in
CAF and CON, respectively. No changes in [Na⫹]int were
observed with exercise during the 20-W period and no statis-
tical differences were found between CAF and CON (Fig. 3B).
Before EX1 [K⫹]int was not significantly different in CAF
and CON, but during EX1 the levels were lower (P ⬍ 0.05) in
CAF than CON with peak values being 5.7 ⫾ 0.3 and 7.5 ⫾ 0.7
mmol/l, respectively (Fig. 4A). In the recovery period after
EX1 [K⫹]int remained lower (P ⬍ 0.05) in CAF than CON .
During and after EX2 [K⫹]int was lower (P ⬍ 0.05) in CAF
than CON (Fig. 4A). Also during EX3 [K⫹]int was lower (P ⬍
0.05) in CAF than CON with peak values being 5.5 ⫾ 0.3 vs.
7.3 ⫾ 0.3 mmol/l, respectively (Fig. 4A).
[Na⫹]int did not change during the 50-W exercise periods
and no differences were found between CAF and CON, al-
though [Na⫹]int in CAF tended to be higher than in CON (P ⫽
0.07; Figs. 3B and 4B).
Plasma Potassium and Metabolites During Knee-Extensor
Exercise
Venous plasma K⫹ was 3.7 ⫾ 0.1 and 3.9 ⫾ 0.0 mmol/l at rest in
CAF and CON, respectively. After the 20-W exercise plasma K⫹
became lower (P ⬍ 0.05) in CAF than in CON (3.8 ⫾ 0.1 and 4.2 ⫾
0.1 mmol/l, respectively). Plasma K⫹ was also lower (P ⬍ 0.05) in
www.jap.org
 CAFFEINE AFFECTS PERFORMANCE AND INTERSTITIAL POTASSIUM 1375

Downloaded from http://jap.physiology.org/ by 10.220.33.3 on June 7, 2017

Fig. 2. Plasma K⫹ (A), blood lactate (B), blood glucose (C), plasma NH3 (D), and plasma FFA (E) at rest, during and after the Yo-Yo IR2 test with prior ingestion
of placebo (PLA; Œ) or caffeine (CAF;
). #Significant difference (P ⬍ 0.05) between CAF and PLA (n ⫽ 12 unless otherwise stated).

CAF compared with CON after EX1 (4.0 ⫾ 0.2 and 4.3 ⫾ 0.0 Plasma FFA was 437 ⫾ 103 and 124 ⫾ 41 mmol/l at rest
mmol/l, respectively) but not after EX2 and EX3. At rest plasma Na⫹ in CAF and CON, respectively (P ⫽ 0.07). After the 20-W
was 138 ⫾ 1 and 135 ⫾ 2 mmol/l in CAF and CON, and no changes exercise bout plasma FFA in CAF was higher (P ⬍ 0.05)
were observed during exercise in either CAF or CON. than in CON (726 ⫾ 179 and 200 ⫾ 64 mmol/l, respec-

J Appl Physiol • VOL 111 • NOVEMBER 2011 • www.jap.org

1376 CAFFEINE AFFECTS PERFORMANCE AND INTERSTITIAL POTASSIUM

performing better after caffeine intake. In support, Stuart et al.

(44) substantiated an elevated performance level during re-
peated sprint tests performed during a simulated rugby game in
high-level players after caffeine intake. Moreover, others have
shown improved high-intensity work during a 90-min simu-
lated soccer game (18) and simulated tennis (26). Thus caffeine
ingestion seems to elevate fatigue resistance during team sport-
relevant, high-intensity intermittent exercise. A number of
studies have used short-term all-out exercise protocols to
investigate possible performance-enhancing effects of caffeine
(8, 12, 24, 48). These studies using maximal sprinting exercise
lasting 15–30 s showed no performance-enhancing effects of
caffeine administration, indicating that the ergogenic effects of
caffeine do not appear to have an impact on performance
during these short-term, single-exercise bouts. In contrast, a
large number of observations using both laboratory and field-
based exercise protocols report an elevated fatigue resistance
during exhaustive intense exercise trials lasting from 1 to 15
min (8, 16, 17, 48). Caffeine appears to be especially ergogenic

Downloaded from http://jap.physiology.org/ by 10.220.33.3 on June 7, 2017

Fig. 3. Interstitial K⫹ (A) and Na⫹ (B) concentrations during and after a 10-min
period of one-legged knee-extension exercise at 20 W with (
; CAF) and
without (Œ; CON) prior caffeine intake. #Significant difference (P ⬍ 0.05)
between CON and CAF (n ⫽ 6).

tively), and remained higher during the three 50-W exercise

bouts.
DISCUSSION

The main findings of the present study were that caffeine

ingestion improved fatigue resistance and skeletal muscle
[K⫹]int was lowered during intense intermittent exercise. Thus,
the delayed fatigue development after caffeine intake may be
associated with caffeine-induced acceleration of muscle K⫹
handling.
The present study is the first to report an effect of caffeine
intake on performance during progressive high-intensity inter-
mittent exercise. The Yo-Yo IR2 test has been shown to have
a high rate of anaerobic energy turnover (31), and performance
in the test is correlated to the work rate during the most intense
periods of a soccer game (7). Thus caffeine appears to delay
the type of fatigue that occurs in team sports during short-term
Fig. 4. Interstitial K⫹ (A) and Na⫹ (B) concentrations during three intense
intense intermittent exercise bouts with a large anaerobic 3-min one-legged knee-extension exercise periods at a power output of 50 W
component (21). The caffeine-induced performance enhance- (separated by 5 min recovery) with (
; CAF) and without (Œ; CON) caffeine
ment was clear in the present study with 11 of the 12 subjects intake. #Significant difference (P ⬍ 0.05) between CON and CAF (n ⫽ 6).

J Appl Physiol • VOL 111 • NOVEMBER 2011 • www.jap.org

 CAFFEINE AFFECTS PERFORMANCE AND INTERSTITIAL POTASSIUM
for speed endurance exercise ranging 1–3 min (15). Thus the
performance-stimulating effect of caffeine seems to occur in
relatively short-term intense exercise with a marked anaerobic
component indicating that caffeine stimulates physiological
mechanisms counteracting this specific type of fatigue.
Fatigue development during high-intensity intermittent ex-
ercise is complex, but an altered resting membrane potential
has been proposed as a central candidate (19, 34, 41). The main
contributor to the depolarized sarcolemmal potential is a
marked efflux of K⫹ from the muscle cell (34) and a concom-
itant accumulation in the muscle interstitium (27, 36, 38, 39,
42). The Yo-Yo IR2 test has been substantiated to exert a
major stimulus on the anaerobic energy systems, and venous
K⫹ concentrations reach the approximate levels (31) observed
by others at the point of fatigue during intense exercise lasting
a few minutes (6, 38). In the present study no differences in
plasma K⫹ between CAF and PLA were observed, but in CAF
it tended to be lower during the entire trial (Fig. 2A). In support
of this Crowe and coworkers (14) found a significant decrease
in plasma K⫹ after caffeine intake, and others have demon-
strated a similar pattern during prolonged exercise (24, 32).
However, any fatiguing effect of elevated extracellular K⫹ is
exerted in the interstitium of the exercising muscles. Therefore,
in S2 the effect of caffeine administration on interstitial K⫹
was studied, and [K⫹]int was markedly lower during and after
both moderate-intensity exercise and repeated intense exercise
(Figs. 3A and 4A). Thus caffeine intake attenuates the accu-
mulation of interstitial K⫹ during repeated intense exercise,
which may delay fatigue as seen in S1. It should be noted that
the average values of a level of 7– 8 mmol/l is not expected to
cause fatigue (27, 36, 38, 39) and are lower than observed in
other studies where exhaustive intense exercise has been ex-
amined. In addition, in S1 plasma potassium was not different
between the caffeine and placebo trial, despite that a marked
difference in [K⫹] int was found in S2, which suggests that
plasma K⫹ during whole body exercise is not a sensitive
measure of [K⫹] int.
There are several possible mechanisms for a caffeine-in-
duced effect on [K⫹]int. First, caffeine may have stimulated the
activity in the Na⫹-K⫹ pump indirectly by an elevated cate-
cholamine response (15, 22, 32). Caffeine could also have had
a direct effect on the muscle, since it has been shown that high
doses of caffeine restore force in K⫹ inhibited muscle in vitro
(10). Furthermore, the elevated glucose concentrations in the
caffeine trial may per se have improved performance through
an effect on the Na⫹-K⫹ pump, as it has been suggested that
high glucose concentrations are preventing the deterioration of
electrical properties of the muscle fiber membrane (29, 33).
The fact that [Na⫹]int tended (P ⫽ 0.07) to be higher in CAF
gives added support to the notion of caffeine-stimulated in-
crease in Na⫹-K⫹ pump activity. Moreover, the rate of [K⫹]int
accumulation was higher in the control trial than in the caffeine
trial in the initial phase of the 20-W exercise period (data not
shown). Thus caffeine may preactivate or accelerate the
Na⫹-K⫹ pump activity and therefore reduce the time lag in
pump activity in the first part of an exercise bout. It has been
shown that the rate of accumulation in [K⫹]int is highest in the
initial part of an intense exercise bout (36, 39), and that the
accumulation rate decreases when intense exercise is repeated
(36), which may associate with an elevated Na⫹-K⫹ pump
activity.
J Appl Physiol • VOL
1377
In the present study plasma FFA was markedly elevated
in CAF before and during the entire test compared with
PLA. Other studies have shown that plasma epinephrine
increases ⬃1 h after caffeine intake in similar doses as in the
present study (13, 24, 44), which most likely caused the
augmented lipolytic activity. Thus it may be proposed that
caffeine may affect performance in an intermittent test by
elevating fat oxidation in the recovery bouts and a concom-
itant glycogen-sparing effect (43). However, it is unlikely
that this has had a significant impact on the fatigue resis-
tance during the exercise due to the relatively short duration
of the test and limited muscle glycogen usage. In support, it
has been shown that the muscle glycogen concentrations are
still high after a Yo-Yo IR2 test of similar duration as in this
study (31).
The caffeine-induced elevation of catecholamine levels
may have had multiple other physiological effects including
a proposed accelerated oxygen uptake kinetics and muscle
glycolysis (21, 45). Greer et al. (24) showed markedly
Downloaded from http://jap.physiology.org/ by 10.220.33.3 on June 7, 2017
higher plasma epinephrine levels before four all-out 30-s
sprints, but the effect disappeared during exercise. More-
over, neither oxygen uptake nor blood lactate was different
during exercise between the caffeine and control trials. In
contrast plasma epinephrine was increased during an entire
⬃80-min simulated rugby game after prior caffeine intake
(44). In the present study blood glucose was higher both
before the Yo-Yo IR2 test as well as the peak levels reached
during the test, indicating increased catecholamine stimula-
tion. However, no difference in blood lactate concentrations
during the Yo-Yo IR2 test was observed between the trials,
indicating similar glycolytic activity. On the other hand the
plasma NH3 rose to significantly higher values in CAF than
PLA, which is in agreement with findings by Greer and
coworkers (24) during repeated sprints. Thus the AMP
breakdown may be accelerated by caffeine intake, but it
does not appear to have affected performance. The ergo-
genic effect of caffeine ingestion has also been linked to
supraspinal or central fatigue, since caffeine apparently
increases motorneuron or motor cortical excitability during
exercise (28). Thus it cannot be excluded that there have
been a central component to the better performance in S1.
In summary, the present study demonstrates that high-inten-
sity intermittent exercise performance is significantly improved
by oral caffeine administration. Moreover, for the first time
caffeine has been shown to lower muscle interstitial potassium
during intense exercise. Thus the performance-enhancing ef-
fect of caffeine intake before intense intermittent exercise may
be associated with an improved interstitial potassium handling
in the exercising muscles.
ACKNOWLEDGMENTS
We thank the subjects for their participation.
GRANTS
The study was supported by The Danish Ministry of Culture (Kulturmin-
isteriets Udvalg for Idrætsforskning).
DISCLOSURES
No conflicts of interest, financial or otherwise, are declared by the author(s).
111 • NOVEMBER 2011 • www.jap.org
1378 CAFFEINE AFFECTS PERFORMANCE AND INTERSTITIAL POTASSIUM
REFERENCES
1. Allen DG, Westerblad H. The effects of caffeine on intracellular calcium,
force and the rate of relaxation of mouse skeletal muscle. J Physiol 487:
331–342, 1995.
2. Allen DG, Lamb GD, Westerblad H. Impaired calcium release during
fatigue. J Appl Physiol 104: 296 –305, 2008.
3. Andersen P, Adams RP, Sjøgaard G, Thorboe A, Saltin B. Dynamic
knee extension as model for study of isolated exercising muscle in
humans. J Appl Physiol 59: 1647–1653, 1985.
4. Astorino TA, Rohmann RL, Firth K. Effect of caffeine ingestion on
one-repetition maximum muscular strength. Eur J Appl Physiol 102:
127–132, 2008.
5. Bangsbo J, Jacobsen K, Nordberg N, Christensen NJ, Graham T.
Acute and habitual caffeine ingestion and metabolic responses to steady-
state exercise. J Appl Physiol 72:1297–303, 1992.
6. Bangsbo J, Madsen K, Kiens B, Richter EA. Effect of muscle acidity on
muscle metabolism and fatigue during intense exercise in man. J Physiol
495: 587–596, 1996.
7. Bangsbo J, Iaia FM, Krustrup P. The Yo-Yo intermittent recovery test:
a useful tool for evaluation of physical performance in intermittent sports.
Sports Med 38: 37–51, 2008.
8. Bell DG, Jacobs I, Ellerington K. Effect of caffeine and ephedrine
ingestion on anaerobic exercise performance. Med Sci Sports Exerc 33:
1399 –1403, 2001.
9. Bruce CR, Anderson ME, Fraser SF, Stepto NK, Klein R, Hopkins
WG, Hawley JA. Enhancement of 2000-m rowing performance after
caffeine ingestion. Med Sci Sports Exerc 32: 1958 –1963, 2000.
10. Cairns SP, Hing WA, Slack JR, Mills RG, Loiselle DS. Different effects
of raised [K⫹]o on membrane potential and contraction in mouse fast- and
slow-twitch muscle. Am J Physiol Cell Physiol 273: C598 –C611, 1997.
11. Carr A, Dawson B, Schneiker K, Goodman C, Lay B. Effect of caffeine
supplementation on repeated sprint running performance. J Sports Med
Phys Fitness 48: 472–478, 2008.
12. Collomp K, Caillaud C, Audran M, Chanal JL, Prefaut C. Effect of
acute or chronic administration of caffeine on performance and on cat-
echolamines during maximal cycle ergometer exercise. C R Seances Soc
Biol Fil 184: 87–92, 1990.
13. Costill DL, Dalsky GP, Fink WJ. Effects of caffeine ingestion on
metabolism and exercise performance. Med Sci Sports Exerc 10: 155–158,
1978.
14. Crowe MJ, Leicht AS, Spinks WL. Physiological and cognitive re-
sponses to caffeine during repeated, high-intensity exercise. Int J Sport
Nutr Exerc Metab 16: 528 –544, 2006.
15. Davis JK, Green JM. Caffeine and anaerobic performance: ergogenic
value and mechanisms of action. Sports Med 39: 813–832, 2009.
16. Doherty M. The effects of caffeine on the maximal accumulated oxygen
deficit and short-term running performance. Int J Sport Nutr 8: 95–104,
1998.
17. Doherty M, Smith PM, Davison RC, Hughes MG. Caffeine is ergogenic
after supplementation of oral creatine monohydrate. Med Sci Sports Exerc
34: 1785–1792, 2002.
18. Foskett A, Ali A, Gant N. Caffeine enhances cognitive function and skill
performance during simulated soccer activity. Int J Sport Nutr Exerc
Metab 19: 410 –423, 2009.
19. Fitts RH. Cellular mechanisms of muscle fatigue. Physiol Rev 74: 49 –94,
1994.
20. Glaister M, Howatson G, Abraham CS, Lockey RA, Goodwin JE,
Foley P, McInnes G. Caffeine supplementation and multiple sprint
running performance. Med Sci Sports Exerc 40: 1835–1840, 2008.
21. Goldstein ER, Ziegenfuss T, Kalman D, Kreider R, Campbell B,
Wilborn C, Taylor L, Willoughby D, Stout J, Graves BS, Wildman R,
Ivy JL, Spano M, Smith AE, Antonio J. International Society of Sports
Nutrition position stand: caffeine and performance. J Int Soc Sports Nutr
7: 5, 2010.
22. Graham TE, Spriet LL. Performance and metabolic responses to a high
caffeine dose during prolonged exercise. J Appl Physiol 71: 2292–2298,
1991.
23. Graham TE, Helge JW, MacLean DA, Kiens B, Richter EA. Caffeine
ingestion does not alter carbohydrate or fat metabolism in human skeletal
muscle during exercise. J Physiol 529: 837–847, 2000.
24. Greer F, McLean C, Graham TE. Caffeine, performance, and metabo-
lism during repeated Wingate exercise tests. J Appl Physiol 85: 1502–
1508, 1998.
J Appl Physiol • VOL 111 • NOVEMBER 2011 •
25. Hellsten Y, Nielsen JJ, Lykkesfeldt J, Bruhn M, Silveira L, Pilegaard
H, Bangsbo J. Antioxidant supplementation enhances the exercise-in-
duced increase in mitochondrial uncoupling protein 3 and endothelial
nitric oxide synthase mRNA content in human skeletal muscle. Free Radic
Biol Med 43: 353–361, 2007.
26. Hornery DJ, Farrow D, Mujika I, Young WB. Caffeine, carbohydrate,
and cooling use during prolonged simulated tennis. Int J Sports Physiol
Perform 2: 423–438, 2007.
27. Juel C, Pilegaard H, Nielsen JJ, Bangsbo J. Interstitial K⫹ in human
skeletal muscle during and after dynamic graded exercise determined by
microdialysis. Am J Physiol Regul Integr Comp Physiol 278: R400 –R406,
2000.
28. Kalmar JM, Cafarelli E. Caffeine: a valuable tool to study central fatigue
in humans? Exerc Sport Sci Rev 32: 143–147, 2004.
29. Karelis AD, Péronnet F, Gardiner PF. Glucose infusion attenuates
muscle fatigue in rat plantaris muscle during prolonged indirect stimula-
tion in situ. Exp Physiol 87: 585–592, 2002.
30. Krustrup P, Mohr M, Amstrup T, Rysgaard T, Johansen J, Steens-
berg A, Pedersen PK, Bangsbo J. The yo-yo intermittent recovery test:
physiological response, reliability, and validity. Med Sci Sports Exerc 35:
697–705, 2003.
31. Krustrup P, Mohr M, Nybo L, Jensen JM, Nielsen JJ, Bangsbo J. The
Yo-Yo IR2 test: physiological response, reliability, and application to elite
Downloaded from http://jap.physiology.org/ by 10.220.33.3 on June 7, 2017
soccer. Med Sci Sports Exerc 38: 1666 –1673, 2006.
32. Lindinger MI, Graham TE, Spriet LL. Caffeine attenuates the exercise-
induced increase in plasma [K⫹] in humans. J Appl Physiol 74: 1149 –
1155, 1993.
33. Marcil M, Karelis AD, Péronnet F, Gardiner PF. Glucose infusion
attenuates fatigue without sparing glycogen in rat soleus muscle during
prolonged electrical stimulation in situ. Eur J Appl Physiol 93: 569 –574,
2005.
34. McKenna MJ, Bangsbo J, Renaud JM. Muscle K⫹, Na⫹, and Cl
disturbances and Na⫹-K⫹ pump inactivation: implications for fatigue. J
Appl Physiol 104: 288 –295, 2008.
35. McNaughton LR, Lovell RJ, Siegler J, Midgley AW, Moore L, Bentley
DJ. The effects of caffeine ingestion on time trial cycling performance. Int
J Sports Physiol Perform 3: 157–163, 2008.
36. Mohr M, Nordsborg N, Nielsen JJ, Pedersen LD, Fischer C, Krustrup
P, Bangsbo J. Potassium kinetics in human muscle interstitium during
repeated intense exercise in relation to fatigue. Pflügers Arch 448: 452–
456, 2004.
37. Mohr M, Krustrup P, Nielsen JJ, Nybo L, Rasmussen MK, Juel C,
Bangsbo J. Effect of two different intense training regimens on skeletal
muscle ion transport proteins and fatigue development. Am J Physiol
Regul Integr Comp Physiol 292: R1594 –R1602, 2007.
38. Nielsen JJ, Mohr M, Klarskov C, Kristensen M, Krustrup P, Juel C,
Bangsbo J. Effects of high-intensity intermittent training on potassium
kinetics and performance in human skeletal muscle. J Physiol 554:
857–870, 2004.
39. Nordsborg N, Mohr M, Pedersen LD, Nielsen JJ, Langberg H,
Bangsbo J. Muscle interstitial potassium kinetics during intense exhaus-
tive exercise: effect of previous arm exercise. Am J Physiol Regul Integr
Comp Physiol 285: R143–R148, 2003.
40. Pruscino CL, Ross ML, Gregory JR, Savage B, Flanagan TR. Effects
of sodium bicarbonate, caffeine, and their combination on repeated
200-m freestyle performance. Int J Sport Nutr Exerc Metab 18:
116 –130, 2008.
41. Sejersted OM, Sjøgaard G. Dynamics and consequences of potassium
shifts in skeletal muscle and heart during exercise. Physiol Rev 80:
1411–1481, 2000.
42. Sjøgaard G. Water and electrolyte fluxes during exercise and their
relation to muscle fatigue. Acta Physiol Scand Suppl 556: 129 –136, 1986.
43. Spriet LL, MacLean DA, Dyck DJ, Hultman E, Cederblad G, Graham
TE. Caffeine ingestion and muscle metabolism during prolonged ex-
ercise in humans. Am J Physiol Endocrinol Metab 262: E891–E898,
1992.
44. Stuart GR, Hopkins WG, Cook C, Cairns SP. Multiple effects of
caffeine on simulated high-intensity team-sport performance. Med Sci
Sports Exerc 37: 1998 –2005, 2005.
45. Tarnopolsky MA. Caffeine and endurance performance. Sports Med 18:
109 –125, 1994.
46. Van Thuyne W, Delbeke FT. Distribution of caffeine levels in urine in
different sports in relation to doping control before and after the removal
www.jap.org
 CAFFEINE AFFECTS PERFORMANCE AND INTERSTITIAL POTASSIUM 1379
of caffeine from the WADA doping list. Int J Sports Med 27: 745–750, 48. Williams JH, Signorile JF, Barnes WS, Henrich TW. Caffeine, maxi-
2006. mal power output and fatigue. Br J Sports Med 22: 132–134, 1988.
47. Wiles JD, Coleman D, Tegerdine M, Swaine IL. The effects of caffeine 49. Williams AD, Cribb PJ, Cooke MB, Hayes A. The effect of ephedra and
ingestion on performance time, speed and power during a laboratory-based caffeine on maximal strength and power in resistance-trained athletes. J
1 km cycling time-trial. J Sports Sci 24: 1165–1171, 2006. Strength Cond Res 22: 464 –470, 2008.

Downloaded from http://jap.physiology.org/ by 10.220.33.3 on June 7, 2017

J Appl Physiol • VOL 111 • NOVEMBER 2011 • www.jap.org