Acute Caffeine Ingestion's Increase of Voluntarily Chosen Resistance-Training Load After Limited Sleep

Abstract

This study aimed to determine whether caffeine ingestion would increase the workload voluntarily chosen by athletes in a limited-sleep state.
In a double-blind, crossover study, 16 professional rugby players ingested either a placebo or 4 mg/kg caffeine 1 hr before exercise. Athletes classified themselves into nondeprived (8 hr+) or sleep-deprived states (6 hr or less). Exercise comprised 4 sets of bench press, squats, and bent rows at 85% 1-repetition maximum. Athletes were asked to perform as many repetitions on each set as possible without failure. Saliva was collected before administration of placebo or caffeine and again before and immediately after exercise and assayed for testosterone and cortisol.
Sleep deprivation produced a very large decrease in total load (p = 1.98 × 10(-7)). Caffeine ingestion in the nondeprived state resulted in a moderate increase in total load, with a larger effect in the sleep-deprived state, resulting in total load similar to those observed in the nondeprived placebo condition. Eight of the 16 athletes were identified as caffeine responders. Baseline testosterone was higher (p < .05) and cortisol trended lower in non-sleep-deprived athletes. Changes in hormones from predose to preexercise correlated to individual workload responses to caffeine. Testosterone response to exercise increased with caffeine compared with placebo, as did cortisol response.
Caffeine increased voluntary workload in professional athletes, even more so under conditions of self-reported limited sleep. Caffeine may prove worthwhile when athletes are tired, especially in those identified as responders.

Full text

157
International Journal of Sport Nutrition and Exercise Metabolism, 2012, 22, 157 -164
© 2012 Human Kinetics, Inc.
Cook and Drawer are with the United Kingdom Sports Coun-
cil, London, UK. Beaven is with the National Winter Sports
Research Center,Östersund, Sweden. Kilduff is with the
College of Engineering, Swansea University, Swansea, UK.
Acute Caffeine Ingestion’s Increase of Voluntarily
Chosen Resistance-Training Load After Limited Sleep
Christian Cook, C. Martyn Beaven, Liam P. Kilduff, and Scott Drawer
Introduction: This study aimed to determine whether caffeine ingestion would increase the workload voluntarily
chosen by athletes in a limited-sleep state. Methods: In a double-blind, crossover study, 16 professional rugby
players ingested either a placebo or 4 mg/kg caffeine 1 hr before exercise. Athletes classied themselves into
nondeprived (8 hr+) or sleep-deprived states (6 hr or less). Exercise comprised 4 sets of bench press, squats,
and bent rows at 85% 1-repetition maximum. Athletes were asked to perform as many repetitions on each set
as possible without failure. Saliva was collected before administration of placebo or caffeine and again before
and immediately after exercise and assayed for testosterone and cortisol. Results: Sleep deprivation produced
a very large decrease in total load (p = 1.98 × 10–7). Caffeine ingestion in the nondeprived state resulted in a
moderate increase in total load, with a larger effect in the sleep-deprived state, resulting in total load similar
to those observed in the nondeprived placebo condition. Eight of the 16 athletes were identied as caffeine
responders. Baseline testosterone was higher (p < .05) and cortisol trended lower in non-sleep-deprived
athletes. Changes in hormones from predose to preexercise correlated to individual workload responses to
caffeine. Testosterone response to exercise increased with caffeine compared with placebo, as did cortisol
response. Conclusions: Caffeine increased voluntary workload in professional athletes, even more so under
conditions of self-reported limited sleep. Caffeine may prove worthwhile when athletes are tired, especially
in those identied as responders.
Keywords: sleep deprivation, sleep poverty, testosterone, cortisol, individual responses
Original research
Caffeine has long been touted as an ergogenic aid.
With the lifting of the partial ban on its use by the Interna-
tional Olympic Committee in 2004, research on its effects
in relation to athletic performance has intensied. Indeed,
caffeine at appropriate doses appears able to improve
time to exhaustion and other indices of endurance in a
number of different physical activities including rowing,
swimming, cycling, and running and simulated rugby
and soccer performances (Carr, Gore, & Dawson, 2011;
Davis & Green, 2009; Paton, Lowe, & Irvine, 2010;
Stuart, Hopkins, Cook, & Cairns, 2005). Furthermore,
a position stand by the International Society of Sports
Nutrition has declared that the positive effects of caffeine
supplementation are not associated with negative effects
on uid balance that could negatively affect performance
(Goldstein et al., 2010).
In terms of the effects of caffeine on strength per-
formance, the existing literature is equivocal. It has been
suggested that caffeine can enhance contractile force at
submaximal loads (Tarnopolsky, 2008), and Jacobson,
Weber, Claypool, and Hunt (1992) observed that a dose
of 7 mg/kg body weight of caffeine had a benecial effect
on muscle strength. However, the same benets were not
seen in a similar study design by Bond, Gresham, McRae,
and Tearney (1986) using a 5-mg/kg body weight dose.
One study (Beck et al., 2006) suggested that upper body
strength is enhanced by caffeine consumption with no
effect on lower body strength, while others have seen
the opposite (Astorino, Martin, Schachtsiek, Wong, &
Ng, 2011) or have failed to see any evidence for strength
increases (Williams, Cribb, Cooke, & Hayes, 2008). Asto-
rino et al. (2011) concluded that caffeine had a limited
practical effect before workouts but conceded that this
was highly individual. Many fundamental resistance-
training programs progress by adding total load via
increased repetitions to set weights before increasing
the actual weight load itself (Kraemer & Ratamess,
2005), so motivation to perform additional repetitions,
as observed on the leg press by Astorino et al. (2011)
after caffeine ingestion, may in fact be an important
component in progressive training gain that has received
little focus.
Mild sleep deprivation is common across all aspects
of modern society (Bixler, 2009) and can affect physi-
cal performance and hormonal responses (Mougin et
al., 2001; Remes, Kuoppasalmi, & Adlercreutz, 1985;
Vgontzas et al., 2004). Indeed, increasing nightly sleep
hours by approximately 2 hr from 8 to 10 hr of sleep has

158 Cook et al.
been shown to improve physical-performance attributes
in college athletes (Mah, Mah, Kezirian, & Dement,
2011). Caffeine is commonly used to ameliorate the
effects of limited sleep, and caffeine consumption has
been associated with enhancements in mood, psycho-
motor performance, and vigilance (Dawkins, Shahzad,
Ahmed, & Edmonds, 2011; McLellan et al., 2005). In
terms of improving sporting performance, caffeine has
produced clear benets in cognitive tasks during exer-
cise (Hogervorst et al., 2008) and sport-specic skills in
team-sport players (Foskett, Ali, & Gant, 2009; Stuart et
al., 2005). Furthermore, our research team has reported
that caffeine improved passing skills in rugby during a
competitive test scenario while the athletes were sleep
deprived (Cook, Crewther, Kilduff, Drawer, & Gaviglio,
2011).
Caffeine ingestion also appears to be associated with
increases in free testosterone and cortisol concentrations
that may be important in mediating the adaptive benets
of resistance exercise. In professional athletes, caffeine
elicited small but signicant acute increases in salivary
testosterone across a workout accompanied by larger
increases in the cortisol response in a dose-responsive
manner (Beaven et al., 2008). Similar changes in these
hormones were seen across the combined physical and
mental stress of a simulated rugby skill (Cook et al.,
2011). In contrast, caffeinated chewing gum led to a
substantial increase in free testosterone not accompanied
by any change in cortisol in competitive cyclists (Paton
et al., 2010).
The main aims of the current study, then, were to
observe whether caffeine ingestion inuenced the vol-
untary choice of workload and, if so, whether this effect
was exaggerated after self-reported poor sleep duration.
A secondary aim was to observe whether free concen-
trations of the hormones testosterone and cortisol varied
across these states.
Methods
Participants
Sixteen professional rugby players (age 20.9 ± 0.9 years,
height 1.85 ± 0.06 m, and body mass 97 ± 8 kg; M ± SD)
were recruited to participate in the study. The athletes
chosen were similar in strength level, with maximum lifts
of 170- to 210-kg squat, 130- to 170-kg bench press, and
110- to 130-kg barbell row. All participants were fully
informed of the nature and possible risks of the study
before giving written consent, and the protocol was
approved by the university ethics committee.
Experimental Protocol
All athletes completed a one-repetition maximum (1-RM)
test in the week preceding the start of the experiment for
back squat, bench press, and bent barbell row, and these
data were used to extrapolate each individual’s 85%
1-RM loading used in the four test sessions. Briey, the
1-RM test session consisted of 5 × 50%, 3 × 60%, 2 ×
70%, 1 × 80%, 1 × 90%, 1 × 95%, and then 1 × 100% of
each individual’s previous best effort. If athletes failed,
they had one more attempt, and if they still failed, the
weight was reduced by 2.5 kg until their best lift was
obtained. If they were successful at their 100% lift, the
weight was increased by 2.5 kg until two failed attempts
indicated they had reached their maximum weight. Three
minutes were allowed between efforts.
The study used a randomized, double-blind, placebo-
controlled, balanced trial. Each athlete was asked to
log his nightly sleep time and quality and to turn up for
testing on two occasions when he had obtained 8 hr or
more of good sleep and on two occasions when he had
slept for less than 6 hr (sleep-deprived state), under the
proviso that any two testing sessions be a minimum of
3 days apart. On arrival, the athletes ingested gelatin
capsules containing either lactose (placebo) or 4 mg/kg
body weight of caffeine 1 hr before exercise. Thus, all
participants completed four test sessions over a 4-week
period with caffeine or placebo ingested before exercise
in both a non-deprived and a sleep-deprived state. They
also completed a short questionnaire at the conclusion
of training to ascertain their perception of what they had
ingested and whether they thought it had had an effect
on their training.
All participants were instructed to refrain from
ingesting dietary caffeine before exercise on the test days.
A dietary log for the preceding 24 hr was collected to
assess caffeine, uid, and food intake, and reminders were
given to ensure dietary compliance. The athletes were by
their own admission light, occasional caffeine consum-
ers, and this was veried by food diaries that were kept
for 2 weeks before the study. The diaries indicated that
consumption of caffeine was equal to or less than 120 mg
caffeine per day, with most participants not consuming
caffeine on a daily basis.
Test Sessions
All exercises were performed at the gymnasium where the
participants were accustomed to training. All participants
were completely familiar with the training protocol and
had undertaken similar sessions over the last 12 months.
They all had a minimum of 2 years of fully recorded and
instructed resistance training. Resistance-training ses-
sions were performed at 11 a.m., with athletes reporting
to the gym at 9:30 a.m. after consuming breakfast and
at least 750 ml of uid. Water was available ad libitum
across the sessions.
The athletes performed a warm-up as follows: 5-min
warm-up on a stationary cycle, unloaded, at 60 rpm and
a nonstop circuit of one set of squats using a 20-kg bar
for 10 repetitions (reps), one set of bench press using a
20-kg for 10 reps, one set of barbell rows using a 20-kg
bar for 10 reps, and one set of squats, bench press, and
barbell row with a 60-kg weight for 10 reps (each with
30 s rest). The subjects then rested for 1 min and began
training.

Caffeine’s Increase of Resistance-Training Load 159
Training consisted of four sets of back squat using
85% of their individually dened 1-RM, four sets of
bench press at 85% 1-RM, and four sets of bent row at
85% 1-RM. Athletes rested for 90 s between sets and
3 min between exercises. They were accompanied by
a trainer during the workout who verbally encouraged
them to perform as many repetitions on each set as they
felt they could without failure. No assistance was given,
and any failed repetitions were not counted. All sets were
attempted. The workload for each exercise was calculated
as the product of total repetitions and load lifted. The sum
of these workloads was dened as the total workload.
Saliva Samples
Whole saliva samples were obtained from each athlete
immediately before caffeine or placebo ingestion, then
again before starting and at the end of the resistance-
exercise session. For each sample, participants were
asked to expectorate 2 ml of saliva into a sterile container.
Saliva samples were stored at –20 °C until assay. Salivary
steroid samples were taken in this study because they are
minimally invasive and have the advantage of reecting
free-steroid concentrations, which are reported to be more
physiologically relevant than total blood levels (Obminski
& Stupnicki, 1997; Vining, McGinley, & Symons, 1983).
To prevent blood contamination of saliva, resulting in an
overestimation of hormone concentrations, subjects were
advised to avoid brushing their teeth, drinking hot uids,
or eating hard foods such as apples in the 2 hr before
providing their sample. Saliva samples were analyzed
in duplicate for testosterone and cortisol using ELISA
kits per manufacturer’s instructions (Salimetrics Ltd.).
Detection limits for the assays were 0.1 pg/ml and 0.01
ng/ml for testosterone and cortisol, respectively. The
intra- and interassay coefcients of variation were <9%
for cortisol and <8% for testosterone.
Statistical Analysis
Changes in the mean of each measure with and with-
out caffeine treatment were used to assess magnitudes
of effects by dividing the changes by the appropriate
between-participants SDs. Pairwise t-statistic compari-
sons were made between conditions, and differences were
interpreted in relation to the likelihood of exceeding the
smallest worthwhile effect with individual change thresh-
olds for each variable. Hormonal and load data were log-
transformed to reduce nonuniformity of error, with effects
derived by back-transformation as percentage changes.
Magnitudes of the standardized effects were interpreted
using thresholds of 0.2, 0.6, and 1.2 for small, moderate,
and large, respectively (Hopkins, Marshall, Batterham,
& Hanin, 2009). Standardized effects of –0.19 to 0.19
were termed trivial. To make inferences about the large-
sample value of an effect, the uncertainty in the effect
was expressed as 90% condence limits. An effect was
deemed unclear if the condence interval overlapped the
thresholds for both small positive and negative effects.
The signicance level was set at p ≤ .05.
Results
None of the athletes were able to perceive beyond chance
whether they had actually received the placebo (which
they had been told was another supplement) or the 4-mg/
kg body weight caffeine dose. Caffeine administration in
the non-sleep-deprived state was associated with substan-
tial and signicant increases in bench press (effect size
[ES] 0.75, p = .039), back squat (ES 0.93, p = .0122),
bent row (ES 0.81, p = .0249), and total workload (ES
1.13, p = .003) compared with the non-sleep-deprived
placebo condition (Figure 1). Caffeine administration
in the sleep-deprived state also produced signicant
increases in bench press (ES 1.16, p = .0023), back squat
(ES 1.39, p = .0004), bent row (ES 1.07, p = .0043), and
total workload (ES 1.47, p = .0002) compared with pla-
cebo sleep-deprived state. With caffeine administration,
the total workload completed in the sleep-deprived state
was not signicantly different (p = .6402) than that seen
in the non-sleep-deprived state of the placebo condition
(Figure 1). Very large increases in the voluntarily chosen
workload were observed between the non-sleep-deprived
caffeine state and the sleep-deprived placebo state: bench
press 2,568 vs. 1,780 kg (p = 2.31 × 10–5), squat 3,316
vs. 2,327 kg (p = 2.20 × 10–5), bent row 1,997 vs. 1,524
kg (p = .0009), and total workload 7,881 vs. 5,631 kg (p
= 1.05 × 10–8). A lack of sleep was associated with sig-
nicant decreases in total workload in both the placebo
(ES 2.33, p = 1.98 × 10–7) and caffeine (ES 1.03, p =
.0058) condition.
The group data were biased by large responses
among 8 of the 16 athletes who were identied as high
caffeine responders. Those participants consistently noted
that they believed there was an effect when they were
administered the caffeine dose. Total workload in the
high caffeine responders was signicantly higher than in
nonresponders in the non-sleep-deprived (ES 3.46, p =
8.78 × 10–6) and sleep-deprived (ES 3.37, p = 1.13 × 10–5)
conditions (Figure 2). Indeed, there was no difference in
the voluntarily chosen workload of the nonresponders
between the caffeine and placebo conditions (p = .5329).
There was, however, a clear moderate positive effect of
the 4-mg/kg body weight caffeine dose on the workload
chosen by nonresponders in the fatigued state (ES 1.01,
p = .0533).
Baseline testosterone was clearly elevated in the
non-sleep-deprived groups compared with the groups
that reported a lack of sleep (ES 0.67–1.09; Table 1). The
4-mg/kg body weight caffeine dose produced an increase
in testosterone across the hour before the workout in
the sleep-deprived (5.8% ± 3.3%, p = .0083) and non-
sleep-deprived (3.4% ± 1.8%, p = .005) groups (Figure
3). In contrast, testosterone signicantly decreased over
the same period in both placebo groups (p < .01), with
a small but clear difference between the change in the
sleep-deprived (10.6% ± 3.6%, p = .0001) and non-sleep-
deprived conditions (10.9% ± 2.4%, p = 2.82 × 10–8). The
workout elicited signicant increases in testosterone in
each of the conditions, but the greatest increase was in the

160
Figure 1 — Workload for each exercise performed, M ± SD. NSD = non-sleep-deprived; SD = sleep-deprived; CAF = caffeine;
PLA = placebo; BP = bench press; SQT = back squat; BR = bent row. aSignicantly greater than NSD:PLA. bSignicantly greater
than SD:CAF. cSignicantly greater than SD:PLA. The threshold for signicance was p ≤ .05. Workload is dened as the product
of repetitions performed and load lifted.
Figure 2 — Total workload for responders and nonresponders with placebo and caffeine administration, M ± SD. RSP = caffeine
responders (n = 8); NR = nonresponders (n = 8); CAF = caffeine; PLA = placebo. *Signicantly greater than other conditions; #Clearly
greater than corresponding placebo condition. Caffeine dose was 4 mg/kg body weight. The threshold for signicance was p ≤ .01.

Caffeine’s Increase of Resistance-Training Load 161
non-sleep-deprived caffeine condition (30.1% ± 6.2%, p
= 2.14 × 10–11), which was greater than in any of the other
conditions (p £ .0104). The poorest testosterone response
was observed in the sleep-deprived placebo condition
(Figure 3), which was signicantly worse than in any of
the other conditions (p £ .0035).
Baseline cortisol was clearly elevated in the sleep-
deprived groups compared with the non-sleep-deprived
groups (ES 0.41–0.80; Table 1). Caffeine administration
produced an increase in cortisol across the hour before the
workout in the sleep-deprived (9.0% ± 5.3%, p = .0107)
and non-sleep-deprived (8.1% ± 2.5%, p = .0001) groups.
In contrast, cortisol signicantly decreased over the same
period in both placebo groups (p < .01), with a small but
clear difference between the change in the sleep-deprived
(13.7% ± 5.6%, p = .0007) and non-sleep-deprived
conditions (12.7% ± 2.7%, p = 3.67 × 10–7; Figure 3).
Exercise elicited signicant increases in cortisol in each
of the conditions, with the greatest increase in the non-
sleep-deprived caffeine condition (21.3% ± 8.4%, p =
1.56 × 10–7), which was greater than in any of the other
conditions (p £ .0898).
It was apparent that the increase in testosterone
across the hour before the workout was driven by the
previously identied caffeine responders, with increases
of 6.4% ± 1.6% (p = .0002) and 12.9% ± 2.8% (p = .0001)
in the non-sleep-deprived and sleep-deprived conditions,
respectively. The corresponding changes in testosterone
for the nonresponders were a trivial increase of 0.4%
± 2.1% in the non-sleep-deprived group and a trivial
decrease of 0.8% ± 1.7% in the sleep-deprived group.
These changes were signicantly less than those seen in
the caffeine responders (p = .0012 and 1.19 × 10–5, respec-
tively). Furthermore, among the 8 individuals identied
as caffeine responders, the change in testosterone from
time of caffeine ingestion to immediately before exercise
was strongly correlated with the subsequent increase in
voluntarily chosen workload (R2 = .67, p = .0002).
Discussion
The main nding in this study was that caffeine ingestion
signicantly improved the total voluntary workload lifted
Table 1 Group Salivary Hormone Concentrations Across Treatment Groups
Predose Preworkout Postworkout
Testosterone Cortisol Testosterone Cortisol Testosterone Cortisol
NSD:CAF 136 ± 22b,c 2.07 ± 0.21 140 ± 20a,b,c 2.23 ± 0.21 183 ± 26a,b,c 2.71 ± 0.26a
NSD:PLA 134 ± 31b,c 2.22 ± 0.51 125 ± 30c2.13 ± 0.49 151 ± 32c2.44 ± 0.49
SD:CAF 113 ± 21 2.44 ± 0.54a,d 120 ± 24 2.64 ± 0.46a,c,d 143 ± 29c2.98 ± 0.50a,c
SD:PLA 115 ± 26 2.40 ± 0.53d110 ± 25 2.30 ± 0.51 123 ± 29 2.60 ± 0.49
Note. NSD = non-sleep-deprived; CAF = caffeine; PLA = placebo; SD = sleep-deprived. Testosterone and cortisol concentrations are in pg/ml and
ng/ml, respectively.
aSubstantially greater than NSD:PLA. bSubstantially greater than SD:CAF. cSubstantially greater than SD:PLA. dSubstantially greater than NSD:CAF.
Figure 3 — Salivary hormone responses to caffeine ingestion
and resistance exercise sessions, M ± SD. NSD = non-sleep-
deprived; SD = sleep-deprived; CAF = caffeine; PLA = placebo.
Caffeine dose was 4 mg/kg body weight. All preworkout tes-
tosterone and cortisol values are signicantly different from
the preceding predose value (p ≤ .01). All postworkout
testosterone and cortisol values are signicantly greater
than the preceding preworkout value (p < .01). aSubstantially
greater than corresponding sleep-deprived:caffeine value. bSub-
stantially greater than corresponding sleep-deprived:placebo
value. cSubstantially greater than corresponding non-sleep-
deprived:placebo value.

162 Cook et al.
by professional rugby athletes in a self-reported state of
sleep poverty. There was a similar, although smaller, ben-
ecial effect on total workload in a non-sleep-deprived
state. As loads were kept constant throughout the testing,
the increases in total workload were a direct result of an
increase in the number of voluntarily selected repetitions
in each set. It is important to state, then, that given the
voluntary nature of the repetitions (not forced by coach
or lifting partner), this is likely more about motivation to
perform than an actual physical-performance increase. As
training progression is often based around such increases,
caffeine ingestion appears to be a valuable intervention
when an athlete is tired. The results also demonstrate
that it is important for athletes to avoid states of sleep
poverty due to the negative physical and hormonal effects
observed.
Previously published work has focused on acute
improvements in maximal strength measures in non-
sleep-deprived conditions, where a small effect of caf-
feine was observed (Astorino, Rohmann, & Firth, 2008;
Beck et al., 2006). University of California researchers
have reported that while caffeine did produce signicant
effects on repetitions performed on the leg press, it was
of limited practical value to the training environment
(Astorino et al., 2011). Part of their argument was based
on the importance of progression in extrapolated 1-RM
strength, which clearly is important, rather than main-
tenance of repetitions performed across sequential sets.
However, herein we focus more on a higher maintenance
of repetitions per set at a standard weight, which we
suggest is a valid training gain. Indeed, the ingestion of
caffeine in the non-sleep-deprived state was associated
with an increase in total workload of 2,250 kg in a single
session compared with the sleep-deprived placebo con-
dition. This voluntary increase in workload represents a
superior adaptive stimulus as a result of caffeine ingestion
without an increase in maximal load.
To a large extent our data agree with the ndings of
Astorino et al. (2011), as we found moderate effects of
caffeine ingestion on repetitions performed while non-
sleep-deprived, with the largest effect observable on
squats rather than bench press or bent row. After sleep
deprivation, however, we saw larger effects of caffeine
ingestion, potentially offering more practical signicance
to athletes when programs progress based on added
repetitions, especially given that sleep deprivation is not
uncommon in many athletic groups. Indeed, high train-
ing loads have been associated with sleep disruption in
competitive swimmers (Taylor, Rogers, & Driver, 1997).
It should be noted that ours results demonstrated
large variability, and individual responses in caffeine sus-
ceptibility were apparent. Indeed, some athletes showed
clear changes in voluntary load volume with caffeine
ingestion compared with placebo, while others were not
affected at all. These data were used to identify the high
caffeine responders in the current study along with their
associated distinctive hormonal proles. Astorino et al.
(2011) also clearly reported an individual responsive-
ness, with heavier caffeine consumers tending to respond
positively to caffeine ingestion. A number of studies
now suggest that heavy caffeine consumers are more
responsive to caffeine ingestion and that this may in fact
simply represent amelioration of caffeine-withdrawal
effects induced by the restrictions on caffeine intake that
studies demand before testing (James & Keane, 2007;
Yeomans, Ripley, Davies, Rusted, & Rogers, 2002). We
intentionally chose low consumers of caffeine to avoid
any potential confounding effects of withdrawal rever-
sal. We have previously suggested that in low caffeine
consumers a much lower dose of caffeine may be more
ergogenic than a higher dose (Cook et al., 2011).
In both placebo groups (sleep-deprived and non-
sleep-deprived) salivary testosterone and cortisol tended
(p < .08) to decline from predose to immediately before
exercise. This observation is consistent with other data
showing a circadian fall across the morning hours (Bird &
Tarpenning, 2004; Gachon, Nagoshi, Brown, Ripperger,
& Schibler, 2004). However, the expected circadian
decrease was absent in the caffeine groups, with notable
increases in individuals identied as caffeine responders.
The observation that the caffeine-induced increase in
testosterone was specic to the athletes who responded
positively in terms of increased workload volume sug-
gests that the dose of caffeine sufcient to elicit meaning-
ful training-volume improvements is related to changes
in preworkout hormones. Alternatively, our research
group has demonstrated a role of testosterone in training
gains that may relate to motivation to perform (Cook et
al., 2011; Crewther & Cook, 2010). The current study
provides some support for this hypothesis, which may be
worthy of further exploration in a competitive environ-
ment where caffeine may have more signicant effects.
Caffeine ingestion was also associated with signi-
cant increases in free testosterone and cortisol after the
workout compared with the placebo condition. This phe-
nomenon does not appear to have been well studied. How-
ever, in professional rugby league players we also saw
a small but signicant increase in both the testosterone
and the cortisol response across the workout after caffeine
ingestion (Beaven, Hopkins, et al., 2008). One other study
has reported similar, but larger, effects in testosterone but
did not observe a marked change in cortisol (Paton et
al., 2010). That study in competitive cyclists employed
a chewing-gum delivery method, which may account for
the differences in hormone response observed. It is also
possible that the differential hormonal responses across
the workout are a simple reection of the increased
workload, facilitated through other mechanisms by caf-
feine, as hormone change is known to be inuenced by
intensity of workout (Kraemer et al., 1990; Kraemer &
Ratamess, 2005).
The actual outcome of this caffeine-augmented
increase in hormones across the workout remains equivo-
cal. Indeed, it has been suggested that typical changes in
testosterone concentration are unlikely to have any real
anabolic effect (West et al., 2010). In contrast, others
have suggested that testosterone can both directly and
indirectly inuence muscle and strength gains (Beaven,

Caffeine’s Increase of Resistance-Training Load 163
Cook, & Gill, 2008; Kvorning, Anderson, Brixen, &
Madsen, 2006; Rønnestad, Nygaard, & Raastad, 2011).
Furthermore, the importance of a functional androgen
receptor has been demonstrated (Inoue, Yamasaki,
Fushiki, Okada, & Sugimoto, 1994), suggesting at least
a permissive role for testosterone in muscle adaptation.
In addition, testosterone has been demonstrated to be
dose responsive in priming for a workout, with short-term
effects on the neuromuscular system that have the poten-
tial to mediate adaptation (Crewther, Cook, Cardinale,
Weatherby, & Lowe, 2011).
In conclusion, we found signicant positive effects
of caffeine ingestion on voluntary workload volume in
a resistance-exercise session performed by professional
rugby players. We found large benecial effects of caf-
feine in athletes who reported a poor sleep duration that
could have practical implications in certain responsive
individuals. In these individuals, salivary testosterone
and cortisol were inuenced by caffeine and may in part
be linked to the increase in voluntary workload.
References
Astorino, T.A., Martin, B.J., Schachtsiek, L., Wong, K., &
Ng, K. (2011). Minimal effect of acute caffeine ingestion
on intense resistance training performance. Journal of
Strength and Condioning Research, 25(6), 1752–1758.
PubMed doi:10.1519/JSC.0b013e3181ddf6db
Astorino, T.A., Rohmann, R.L., & Firth, K. (2008). Effect of
caffeine ingestion on one-repetition maximum muscular
strength. European Journal of Applied Physiology, 102(2),
127–132. PubMed doi:10.1007/s00421-007-0557-x
Beaven, C.M., Cook, C.J., & Gill, N.D. (2008). Signicant
strength gains observed in rugby players following specic
RE protocols based on individual salivary testosterone
responses. Journal of Strength and Conditioning
Research, 22(2), 419–425. PubMed doi:10.1519/
JSC.0b013e31816357d4
Beaven, C.M., Hopkins, W.G., Hansen, K., Wood, M.R.,
Cronin, J., & Lowe, T. (2008). Dose effect of caffeine on
testosterone and cortisol responses to resistance exercise.
Internaonal Journal of Sport Nutrion and Exercise
Metabolism, 18(2), 131–141. PubMed
Beck, T.W., Housh, T.J., Schmidt, R.J., Johnson, G.O., Housh,
D.J., Coburn, J.W., & Malek, M.H. (2006). The acute effects
of a caffeine-containing supplement on strength, muscular
endurance, and anaerobic capabilities. Journal of Strength
and Condioning Research, 20(3), 506–510. PubMed
Bird, S.P., & Tarpenning, K.M. (2004). Inuence of circadian
time structure on acute hormonal responses to a single
bout of heavy-resistance exercise in weight-trained men.
Chronobiology Internaonal, 21(1), 131–146. PubMed
doi:10.1081/CBI-120027987
Bixler, E. (2009). Sleep and society: An epidemiological
perspective. Sleep Medicine, 10, S3–S6. PubMed
doi:10.1016/j.sleep.2009.07.005
Bond, V., Gresham, K., McRae, J., & Tearney, R. J. (1986).
Caffeine ingestion and isokinetic strength. Brish Journal
of Sports Medicine, 20(3), 135–137.
Carr, A.J., Gore, C.J., & Dawson, B. (2011). Induced alkalosis
and caffeine supplementation: Effects on 2,000-m rowing
performance. Internaonal Journal of Sport Nutrion and
Exercise Metabolism, 21(5), 357–364. PubMed
Cook, C.J., Crewther, B.T., Kilduff, L.P., Drawer, S., & Gaviglio,
C.M. (2011). Skill execution and sleep deprivation:
Effects of acute caffeine or creatine supplementation—A
randomized placebo-controlled trial. Journal of the
Internaonal Society of Sports Nutrion, 8(1), 2–9.
Crewther, B.T., & Cook, C. (2010). Relationships between sali-
vary testosterone and cortisol concentrations and training
performance in Olympic weightlifters. Journal of Sports
Medicine and Physical Fitness, 50(3), 371–375. PubMed
Crewther, B.T., Cook, C., Cardinale, M., Weatherby, R.P.,
& Lowe, T. (2011). Two emerging concepts for elite
athletes: The short-term effects of testosterone and cortisol
on the neuromuscular system and the dose-response
training role of these endogenous hormones. Sports
Medicine (Auckland, N.Z.), 41(2), 103–123. PubMed
doi:10.2165/11539170-000000000-00000
Davis, J.K., & Green, J.M. (2009). Caffeine and anaerobic
performance: Ergogenic value and mechanisms of action.
Sports Medicine (Auckland, N.Z.), 39(10), 813–832.
PubMed doi:10.2165/11317770-000000000-00000
Dawkins, L., Shahzad, F.Z., Ahmed, S.S., & Edmonds, C.J.
(2011). Expectation of having consumed caffeine can
improve performance and mood. Appete, 57(3), 597–
600. PubMed doi:10.1016/j.appet.2011.07.011
Foskett, A., Ali, A., & Gant, N. (2009). Caffeine enhances
cognitive function and skill performance during simulated
soccer activity. Internaonal Journal of Sport Nutrion
and Exercise Metabolism, 19(4), 410–423. PubMed
Gachon, F., Nagoshi, E., Brown, S.A., Ripperger, J., & Schibler,
U. (2004). The mammalian circadian timing system:
From gene expression to physiology. Chromosoma, 113,
103–112. PubMed doi:10.1007/s00412-004-0296-2
Goldstein, E.R., Ziegenfuss, T., Kalman, D., Kreider, R.,
Campbell, B., Wilborn, C., . . . Antonio, J. (2010).
International Society of Sports Nutrition position stand:
Caffeine and performance. Journal of the Internaonal
Society of Sports Nutrion, 7(1), 5. PubMed
Hogervorst, E., Bandelow, S., Schmitt, J., Jentjens, R.L.P.G.,
Oliviera, M., Allgrove, J., . . . Gleeson, M. (2008). Caffeine
improves physical and cognitive performance during
exhaustive exercise. Medicine and Science in Sports
and Exercise, 40(10), 1841–1851. PubMed doi:10.1249/
MSS.0b013e31817bb8b7
Hopkins, W.G., Marshall, S.W., Batterham, A.M., & Hanin,
J. (2009). Progressive statistics for studies in sports
medicine and exercise science. Medicine and Science in
Sports and Exercise, 41(1), 3–13. PubMed doi:10.1249/
MSS.0b013e31818cb278
Inoue, K., Yamasaki, S., Fushiki, T., Okada, Y., & Sugimoto,
E. (1994). Androgen receptor agonist suppresses exercise-
induced hypertrophy of skeletal muscle. European
Journal of Applied Physiology, 69, 88–91. doi:10.1007/
BF00867933
Jacobson, B.H., Weber, M.D., Claypool, L., & Hunt, L.E.
(1992). Effecto of caffeine on maximal strength and power

164 Cook et al.
in elite male athletes. Brish Journal of Sports Medicine,
26(4), 276–280.
James, J.E., & Keane, M.A. (2007). Caffeine, sleep and
wakefulness: Implications of new understanding about
withdrawal reversal. Human Psychopharmacology, 22(8),
549–558. PubMed doi:10.1002/hup.881
Kraemer, W.J., Marchitelli, L., Gordon, S.E., Harman, E.,
Dziados, J.E., Mello, R., . . . Fleck, S.J. (1990). Hormonal
and growth factor responses to heavy resistance exercise
protocols. Journal of Applied Physiology, 69(4), 1442–
1450. PubMed
Kraemer, W.J., & Ratamess, N.A. (2005). Hormonal responses
and adaptations to resistance exercise and training. Sports
Medicine (Auckland, N.Z.), 35(4), 339–361. PubMed
doi:10.2165/00007256-200535040-00004
Kvorning, T., Anderson, M., Brixen, K., & Madsen, K. (2006).
Suppression of endogenous testosterone production
attenuates the response to strength training: a randomized,
placebo-controlled, and blinded intervention study.
American Journal of Physiology. Endocrinology and
Metabolism, 291, E1325–E1332. PubMed doi:10.1152/
ajpendo.00143.2006
Mah, C.D., Mah, K.E., Kezirian, E.J., & Dement, W.C. (2011).
The effects of sleep extension on the athletic performance
of collegiate basketball players. Sleep, 34(7), 943–950.
PubMed
McLellan, T.M., Kamimori, G.H., Voss, D.M., Bell, D.G., Cole,
K.G., & Johnson, D. (2005). Caffeine maintains vigilance
and improves run times during night operations for special
forces. Aviaon, Space, and Environmental Medicine,
76(7), 647–654. PubMed
Mougin, F., Bourdin, H., Simon-Rigaud, M.L., Nhu, U.N.,
Kantelip, J.P., & Davenne, D. (2001). Hormonal responses
to exercise after partial sleep deprivation and after a
hypnotic drug-induced sleep. Journal of Sports Sciences,
19, 89–97. PubMed doi:10.1080/026404101300036253
Obminski, Z., & Stupnicki, R. (1997). Comparison of the
testosterone-to-cortisol ratio values obtained from
hormonal assays in saliva and serum. Journal of Sports
Medicine and Physical Fitness, 37(1), 50–55. PubMed
Paton, C.D., Lowe, T., & Irvine, A. (2010). Caffeinated chewing
gum increases repeated sprint performance and augments
increases in testosterone in competitive cyclists. European
Journal of Applied Physiology, 110(6), 1243–1250.
Remes, K., Kuoppasalmi, K., & Adlercreutz, H. (1985). Effect
of physical exercise and sleep deprivation on plasma
androgen levels: Modifying effect of physical tness.
Internaonal Journal of Sports Medicine, 6(3), 131–135.
PubMed doi:10.1055/s-2008-1025825
Rønnestad, B.R., Nygaard, H., & Raastad, T. (2011).
Physiological elevation of endogenous hormones results
in superior strength training adaptation. European Journal
of Applied Physiology, 111(9), 2249–2259.
Stuart, G.R., Hopkins, W.G., Cook, C., & Cairns, S.P. (2005).
Multiple effects of caffeine on simulated high-intensity
team sport performance. Medicine and Science in
Sports and Exercise, 37(11), 1998–2005. PubMed
doi:10.1249/01.mss.0000177216.21847.8a
Tarnopolsky, M.A. (2008). Effect of caffeine on the
neuromuscular system—Potential as an ergogenic aid.
Applied Physiology, Nutrition, and Metabolism, 33,
1284–1289. PubMed doi:10.1139/H08-121
Taylor, S.R., Rogers, G.G., & Driver, H.S. (1997). Effects of
training volume on sleep, psychological, and selected
physiological proles of elite female swimmers. Medicine
and Science in Sports and Exercise, 29(5), 688–693.
PubMed doi:10.1097/00005768-199705000-00016
Vgontzas, A.N., Zoumakis, E., Bixler, E.O., Lin, H-M., Follet,
H., Kales, A., & Chrousos, G. (2004). Adverse effects
of modest sleep restriction on sleepiness, performance,
and inflammatory cytokines. The Journal of Clinical
Endocrinology and Metabolism, 89(5), 2119–2126.
PubMed doi:10.1210/jc.2003-031562
Vining, R.F., McGinley, R.A., & Symons, R.G. (1983).
Hormones in saliva: Mode of entry and consequent
implications for clinical interpretation. Clinical Chemistry,
29(10), 1752–1756. PubMed
West, D.W.D., Burd, N.A., Tang, J.E., Moore, D.R., Staples,
A.W., Holwerda, A.M., . . . Phillips, S.M. (2010).
Elevations in ostensibly anabolic hormones with resistance
exercise enhance neither training-induced muscle
hypertrophy nor strength of the elbow exors. Journal of
Applied Physiology, 108(1), 60–67. PubMed doi:10.1152/
japplphysiol.01147.2009
Williams, A.D., Cribb, P.J., Cooke, M.B., & Hayes, A. (2008).
The effect of ephedra and caffeine on maximal strength and
power in resistance-trained athletes. Journal of Strength
and Condioning Research, 22(2), 464–470. PubMed
doi:10.1519/JSC.0b013e3181660320
Yeomans, M.R., Ripley, T., Davies, L.H., Rusted, J., &
Rogers, P.J. (2002). Effects of caffeine on performance
and mood depend on the level of caffeine abstinence.
Psychopharmacology, 164(3), 241–249. PubMed
doi:10.1007/s00213-002-1204-1\

Copyright of International Journal of Sport Nutrition & Exercise Metabolism is the property of Human Kinetics
Publishers, Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email articles for
individual use.