Secondary Logo

Journal Logo

Caffeine Effects on Short-Term Performance during Prolonged Exercise in the Heat


Medicine & Science in Sports & Exercise: April 2008 - Volume 40 - Issue 4 - p 744-751
doi: 10.1249/MSS.0b013e3181621336
APPLIED SCIENCES: Physical Fitness and Performance

Purpose: To determine the effect of water, carbohydrate, and caffeine ingestion on fatigue during prolonged exercise in the heat.

Methods: Seven endurance-trained cyclists (V˙O2max = 61 ± 8 mL·kg−1·min−1) pedaled for 120 min at 63% V˙O2max in a hot-dry environment (36°C; 29% humidity), ingesting either no fluid (NF), water (WAT) to replace 97% fluid losses, the same volume of a 6% carbohydrate-electrolyte solution (CES), or each of these treatments along with ingestion of 6 mg of caffeine per kilogram of body weight (NF + CAFF, WAT + CAFF, and CES + CAFF). At regular intervals during exercise, maximal cycling power (PMAX) was measured. Before and after exercise, maximal voluntary contraction (MVC), voluntary activation (VA), and electrically evoked contractile properties of the quadriceps were determined.

Results: Without fluid replacement (NF and NF + CAFF), subjects were dehydrated by 3.8 ± 0.3%, and rectal temperature reached 39.4 ± 0.3°C, while it was maintained at 38.7 ± 0.3°C in trials with rehydration (P < 0.05). Trials with caffeine ingestion increased PMAX by 3% above trials without caffeine (P < 0.05). MVC reductions after exercise were larger with NF (−11 ± 5%) than for the rest of the trials (P < 0.05). MVC was reduced in WAT compared with CES + CAFF (−6 ± 4 vs 2 ± 4%; P < 0.05). However, NF + CAFF maintained MVC at the level of the CES trial. VA showed the same treatment response pattern as MVC. There were no differences in electrically evoked contractile properties among trials.

Conclusion: During prolonged exercise in the heat, caffeine ingestion (6 mg·kg−1 body weight) maintains MVC and increases PMAX despite dehydration and hyperthermia. When combined with water and carbohydrate, caffeine ingestion increases maximal leg force by increasing VA (i.e., reducing central fatigue).

Exercise Physiology Laboratory, Universidad de Castilla-La Mancha, Toledo, SPAIN

Address for correspondence: Ricardo Mora-Rodriguez, Ph.D., Castilla-La Mancha University, Avda. Carlos III, s/n. 45071, Toledo, SPAIN; E-mail:

Submitted for publication August 2007.

Accepted for publication November 2007.

Prolonged exercise in a hot environment negatively affects endurance performance and may also affect short-term, high-intensity performance. The hyperthermia induced by prolonged exercise in the heat is the factor seemingly responsible for the neuromuscular fatigue that manifests in reductions in quadriceps maximal voluntary contraction (MVC) (12,25) and maximal cycling power (PMAX) (11). Supporting the involvement of hyperthermia in fatigue, it has been found that passive heating (26,27) or exercise-induced hyperthermia (23,25) reduce the ability to voluntarily activate leg muscles to contract and reach peak muscle force. Oral rehydration with water counteracts hyperthermia during prolonged exercise in the heat and maintains PMAX above a control dehydration trial (11). The addition of carbohydrates to the rehydration drink better maintains PMAX (11) and quadriceps MVC in subjects that exercise prolonged enough to drive blood glucose to hypoglycemic levels (22).

Apart from rehydrating, another option to mitigate fatigue caused by prolonged exercise in the heat is to maintain the body's ability to voluntarily activate the muscles despite hyperthermia. Caffeine, a well-established ergogenic aid for endurance performance (13), could potentially stimulate muscle force and power when fatigue ensues. Data on the effects of caffeine on short-term performance, however, are uncertain. PMAX has been found to be either increased (2) or not affected (14,29) by caffeine ingestion (3-7 mg·kg−1 body mass) when tested in a thermoneutral environment. Caffeine ingestion (6 mg·kg−1) increased quadriceps MVC and voluntary activation (VA) after a 2-min protocol (24) that induced local muscle fatigue. In contrast, others have not found an ergogenic effect of caffeine on MVC despite using similar caffeine doses and subjects (19,28).

Caffeine is believed to facilitate central nervous system recruitment of muscle by inhibiting adenosine actions (10). Thus, caffeine may potentially have a larger ergogenic effect during exercise that induces central fatigue rather than after local muscle fatigue (19,28). Supporting this view, Cureton et al. (8) have recently reported that after prolonged exercise in a temperate environment (28°C), caffeine ingestion from a beverage that also contained carbohydrates, taurine, carnitine, and vitamins attenuated the decrease in quadriceps MVC when compared with the ingestion of a carbohydrate-electrolyte solution. However, in the same study, quadriceps VA was not improved, suggesting a direct effect of caffeine on the muscle. Further, no ergogenic effect on MVC was found by adding carbohydrates to the rehydration drink. Beyond the effects on isometric force (i.e., MVC), caffeine ingestion may prevent the declines in cycling peak power (i.e., PMAX) observed during prolonged exercise in the heat (11). However, to our knowledge, the effects of caffeine by itself compared with rehydration with water or with a carbohydrate-electrolyte solution on short-term performance (i.e., MVC and PMAX) during prolonged exercise in the heat have not been investigated.

The purpose of our study was to determine the effect of water, carbohydrate, and caffeine ingestion on maximal quadriceps strength and cycling power during prolonged exercise in the heat. A second purpose was to investigate the mechanisms responsible for the ergogenic effect of these substances by using electrical stimulation. We hypothesized that the coingestion of water (to reduce hyperthermia), carbohydrates (to provide exogenous energy), and caffeine (to increase VA of muscle) would have the greatest ergogenic effect during exercise in the heat.

Back to Top | Article Outline



Seven endurance-trained males who routinely trained for approximately 2 h·d−1, 4-7 d·wk−1, completed this study. Their mean (± SD) age, weight, height, maximal heart rate, and maximal oxygen uptake (V˙O2max) were 27 ± 1 yr, 72 ± 7 kg, 177 ± 5 cm, 191 ± 13 bpm, and 61 ± 8 mL·kg−1·min−1, respectively. The subjects were fully informed of any risks and discomforts associated with the experiment before giving their informed written consent to participate. The study was approved by the local research ethics committee. The subjects were light caffeine consumers (< 60 mg·d−1 from caffeinated soda and lyophilized coffee in milk), and all of them had previously participated in experiments involving endurance performance in hot environments.

Back to Top | Article Outline

Preliminary testing.

All participants completed a continuous incremental cycling test to volitional exhaustion to determine V˙O2max. Then, each subject underwent nine consecutive days of heat acclimatization consisting of 90-min cycling bouts (63% of V˙O2max) in a hot environment (35°C and 27% relative humidity). Trials were held from May to September with outdoor temperatures averaging 26°C, which likely provided some degree of acclimation to the heat. After eight bouts of acclimatization, whole-body sweat rate and heart rate plateaued (i.e., changed less than 5% from previous bout), suggesting full adaptation to the heat of the cardiovascular and thermoregulatory systems. The final acclimation bout was used to familiarize the subjects with the experimental procedures and to measure each individual's sweat rate. The sweat rate was used to calculate fluid-replacement volumes during the experimental trials. After nine acclimatization bouts, subjects underwent four familiarization sessions devoted to obtaining reliable measurements of PMAX and MVC before experimental trials.

Back to Top | Article Outline

Experimental design.

A double-blind, placebo-controlled experimental design was used, with all subjects serving as their own controls. Each subject performed six experimental trials pedaling for 120 min (Monark 818 Varberg) at an intensity that initially elicited 63 ± 5% of their V˙O2max (176 ± 40 W). Experimental trials were conducted randomly and separated by at least 48 h, but no more than 4 d to avoid losing heat acclimation. Exercise was performed in a hot chamber held at 36 ± 3°C and 29 ± 3% relative humidity with a constant 1.9-m·s−1 airflow. The subjects received the following treatments during exercise: 1) no fluid replacement (NF); 2) caffeine ingestion prior to exercise (NF + CAFF); 3) fluid replacement with water (WAT); 4) fluid replacement with water plus caffeine ingestion (WAT + CAFF); 5) fluid replacement with a 6% carbohydrate-electrolyte solution (CES; Gatorade, Quaker Oats Co.); and 6) fluid replacement with a 6% CES plus caffeine ingestion (CES + CAFF; Fig. 1). These experimental trials were chosen to fit a factorial design to evaluate the main effects of carbohydrate and caffeine ingestion, and the potential interactions between them.



In trials with caffeine ingestion (i.e., NF + CAFF, WAT + CAFF, CES + CAFF), caffeine (Durvitan, Seid) was provided in capsules filled to deliver a dose of 6 mg·kg−1 body mass. The capsules were ingested 45 min before the onset of exercise protocol because serum caffeine concentration peaks 30-60 min after ingestion (4,6). In trials without caffeine ingestion (i.e., NF, WAT, and CES), the subjects ingested placebo capsules filled with the same amount of dextrose. The amount of additional energy provided by the dextrose (~2 kcal) was deemed negligible. In trials with fluid replacement (i.e., WAT, CES, WAT + CAFF, and CES + CAFF), subjects ingested a volume equal to 97% of their sweat losses (2.4 ± 0.3 L). Subjects ingested one third of the total volume just before exercise and the rest in four aliquots after 9, 30, 60, and 90 min of exercise. To prevent body cooling caused by fluid ingestion, drinks remained for at least 45 min inside the hot chamber, and their average temperature at ingestion was 32°C. In trials with carbohydrate ingestion (CES and CES + CAFF), an average of 146 ± 21 g was ingested during exercise in combination with the rehydration drink.

Back to Top | Article Outline

Experimental protocol.

The day before each experimental trial, the subjects adopted a similar diet, light exercise, and fluid intake regimen. Food and training diaries were collected, and verbal reminders were given to ensure compliance. The subjects withdrew from all dietary sources of caffeine or alcohol for 48 h before testing, and they reported to the laboratory at the same time of day to avoid the effects of circadian variation on the variables measured (18). Postprandial state was standardized by ingesting a carbohydrate meal (~3 g of carbohydrate per kilogram of body mass) 3 h before the exercise, to promote muscle glycogen storage (7). Two hours before the exercise, the subjects were instructed to drink 500 mL of water to ensure euhydration before exercise.

One hour before the onset of exercise, the subjects entered the hot chamber, warmed up (10 min pedaling at a work rate of 1.5 W·kg−1 body mass), and performed the MVC and electrically evoked quadriceps contractions protocol (described later in this work). Immediately afterward, they ingested the capsules containing caffeine (432 ± 42 mg) or placebo, washed down with a volume of 75 mL of water. After 30 min of rest in a thermoneutral room (22°C), subjects' nude body weight was recorded by using a ± 0.05 kg sensitive scale (Wildcat, Mettler Toledo), and a flexible thermistor (401, YSI) was positioned 15 cm past the anal sphincter. The thermistor was connected to a multichannel A/D board (PowerLab 8SP, ADI) and associated software to display and store rectal temperature readings for 60 s every 10 min throughout the trials. A 22-gauge Teflon catheter was inserted into an antecubital vein for blood sampling. The catheter was frequently flushed with 3-4 mL of 0.9% sterile saline to ensure patency. They reentered the climatic chamber and sat quietly on the cycle ergometer while ambient (heat stress monitor, WBGT, Wibget IST) and rectal temperature (TREC) were recorded. Then, a 5-mL blood sample was withdrawn, and subjects started pedaling.

The subjects cycled continuously for 120 min, interspersed with PMAX measurements after 26, 56, 86, and 116 min of exercise. PMAX was measured during a 4-s maximal effort to overcome the inertial load of a heavy (21.5 kg) Monark cycle ergometer flywheel (5). Oxygen consumption (V˙O2) and carbon dioxide production (V˙CO2) were measured for a 4-min period after 15, 50, and 110 min of exercise, using a computerized open-circuit spirometer (Quark b2, Cosmed, Italy). Carbohydrate oxidation was calculated from V˙O2 and V˙CO2 (15). Blood samples (5 mL) were withdrawn at rest and after 8, 53, and 113 min of exercise.

Immediately after exercise (while subjects remained in the hot chamber), the subjects repeated the MVC and electrically evoked contractions protocol. Instrumentation was then removed, subjects were towel dried, and postexercise nude body weight was recorded to calculate dehydration.

Back to Top | Article Outline

MVC and electrically evoked contractions.

MVC and electrically evoked isometric contractions of the right quadriceps were assessed before exercise (prior to caffeine ingestion) and just after the completion of exercise (to avoid body cooling). The subjects sat upright in an adjustable chair. Their arms were crossed over the chest while straps were fastened across chest, waist, and over the thighs to prevent extraneous body movements. With the right hip and knee flexed at 90°, the right ankle was anchored above the malleolus by a strap connected through a metal link (55 cm) to a strain gauge dynamometer (Tedea Huntleigh 1263, Germany), to measure isometric torque production about the knee. The dynamometer was calibrated with weights of known mass, and the voltage readings were converted to torque (N·m) by an A/D board (Powerlab 8SP, ADI) and related software. The sample rate was set at 1 kHz. Electrical stimulation of the muscle was performed using a four-channel high-voltage stimulator (400 V, Megasonic 313, Medicarin) that delivered a constant-intensity current despite changes in skin impedance caused by dehydration. Stimulation amperage (35-101 mA) was adjusted until torque was 50% MVC, as described previously (20). The stimulus was delivered via three pairs of adhesive patch gel electrodes (4.5 × 4.5 cm, Medicarin) placed on the skin overlying vastus lateralis, vastus medialis, and rectus femoris distally and medially. Because electrodes were removed for the cycling protocol, their placing was marked with an antiallergenic permanent marking pen to ensure consistent positioning after cycling.

MVC of the quadriceps was measured in triplicate. Each MVC lasted 4 s with a 2-min resting period between repetitions. The subjects were strongly encouraged to produce the highest possible force in each MVC and invited to repeat the attempts they considered "not maximal." The attempt with the highest MVC value of torque was used for analysis. VA of quadriceps was assessed by using a supramaximal train of electrical stimuli (1 s at 80 Hz) superimposed onto MVC. Subjects were strongly encouraged to maintain MVC despite the superimposed electrical stimulus. VA was calculated using the central activation ratio formula (17),

where MVC + ES is the isometric force achieved when superimposing electrical stimulation onto MVC. This technique permits investigators to distinguish whether the strength loss after cycling in the heat is attributable to a decreased activation of muscle (i.e., central nervous system fatigue) or to a direct effect on muscle contractility.

After VA determination, the quadriceps muscle was electrically stimulated to evoke a contraction. Two trains of 500-ms electrical stimulus at either 20 or 80 Hz were delivered to the muscle, interspersed by 1 min of rest. The 500-ms stimulation train has been reported to be long enough to induce a plateau in the mechanical response, and it accurately assesses muscle contractile function (21). The highest value of torque production (PEVO), response time (i.e., time from stimulus delivery to onset of the mechanical response, TEVO), and maximal rate of torque development (MTDEVO) were measured at 20 and 80 Hz.

Back to Top | Article Outline

Maximal cycling power.

PMAX measurements were interspersed during prolonged exercise. Within each period of measurement, subjects performed three all-out 4-s sprints followed by 2-min recovery pedaling at a low work rate, which, when combined with the sprints, averaged 63% V˙O2max. For each sprint, subjects positioned their left pedal at 45° (above the horizontal), firmly grabbed the handlebar, and accelerated the Monark's flywheel maximally. The increase in kinetic energy resulting from the acceleration of the flywheel was used to calculate PMAX (5).

Back to Top | Article Outline

Blood analysis.

Blood samples were mixed with ethylenediaminetetraacetic acid (EDTA) in plastic tubes. The plasma was immediately separated by centrifugation (MPW-350R, MedInstruments, Poland) and stored at −80°C for future analysis. At a later date, plasma was analyzed for lactate ([Lac]plasma) and glucose ([Glu]plasma) concentrations (1500 Sports, YSI). A portion (2.5 mL) of each blood sample was allowed to clot, and serum was separated by centrifugation. The serum was analyzed for osmolality (Osmserum) using a freezing point osmometer (mod 300, Advanced Instruments).

Back to Top | Article Outline

Statistical analysis.

Preexercise TREC and serum osmolality (Osmserum), and percent change from pre- to postexercise in MVC and VA, were analyzed with a one-way ANOVA with repeated measures (i.e., treatments). The data collected repeatedly over time (i.e., PMAX) were analyzed with a two-way (treatment by time) ANOVA with repeated measures. After a significant F test (Geisser-Greenhouse correction for the assumption of sphericity), differences between means were identified by using Tukey's HSD post hoc procedure. To test the main effect of caffeine, the data from trials with caffeine ingestion (NF + CAFF, WAT + CAFF, and CES + CAFF) were pooled and compared with those from trials without caffeine ingestion (NF, WAT, and CES) when a caffeine effect but no significant carbohydrate or interaction effects were observed. Similarly, when a carbohydrate effect but no significant caffeine or interaction effects were observed, data from trials with carbohydrate ingestion (CES and CES + CAFF) were pooled and compared with those from trials without carbohydrate ingestion (WAT and WAT + CAFF). The significance level was set at P < 0.05. The data presented are means ± SD.

Back to Top | Article Outline


Body temperature and fluid balance.

Before exercise, TREC and Osmserum were similar among all trials (average of 37.6 ± 0.3°C and 287 ± 5 mOsm·kg−1 H2O, respectively). After 120 min of exercise in the heat without fluid replacement (NF and NF + CAFF trials), the subjects lost 3.8 ± 0.3% of their initial body weight while TREC rose to 39.4 ± 0.5°C. Rehydration with WAT or CES (2.4 ± 0.3 L) prevented body fluid deficit (dehydration 0.8 ± 0.3%) and TREC from exceeding 38.7 ± 0.5°C (lower than NF and NF + CAFF trials; P < 0.05). The addition of caffeine to water (WAT + CAFF) or to CES (CES + CAFF) did not significantly affect TREC (38.8 ± 0.8°C) compared with WAT or CES. From rest, Osmserum increased to 300 ± 8 mOsm·kg−1 H2O at the end of exercise in the no-fluid trials (NF and NF + CAFF), while it was maintained to approximately 290 ± 3 mOsm·kg−1 H2O during the fluid replacement trials (WAT, CES, WAT + CAFF, and CES + CAFF) without an effect of caffeine ingestion.

Back to Top | Article Outline

MVC and electrically evoked contractions.

The MVC isometric torque did not differ among treatments before exercise, averaging 301 ± 25 N·m. The reductions in MVC from preexercise values were larger in the NF trial (−11 ± 5%) than in the rest of the trials (Fig. 2A; P < 0.05), except for the WAT trial. Water ingestion, despite offsetting dehydration and hyperthermia, did not prevent a 6 ± 4% decline in MVC. The percent decline in MVC with WAT was larger than that observed during the CES + CAFF trial (P < 0.05). Of note, MVC was reduced less during NF + CAFF relative to NF, despite subjects being as hyperthermic and dehydrated.



The same response pattern was observed for quadriceps VA in all treatments (P < 0.05; Fig. 3A). When pooling data, caffeine ingestion (NF + CAFF, WAT + CAFF, CES + CAFF) maintained MVC isometric torque and VA better than trials without caffeine ingestion (NF, WAT, CES) after 120 min of exercise in the heat (P < 0.05; Figs. 2B and 3B). A similar effect on MVC and VA was observed when comparing carbohydrate ingestion (CES, CES + CAFF) against trials without carbohydrate ingestion (WAT, WAT + CAFF; P < 0.05; Figs. 2B and 3B).



Preexercise PEVO was similar among trials (i.e., 109 ± 10 N·m and 152 ± 13 N·m for 20 Hz and 80 Hz, respectively). Postexercise PEVO remained similar to preexercise levels (i.e., 104 ± 10 N·m and 156 ± 13 N·m for 20 Hz and 80 Hz, respectively). Before exercise, there were no differences in response time (i.e., TEVO = 225 ± 24 ms and 195 ± 18 ms for 20 Hz and 80 Hz, respectively) or maximal rate of torque development (MTDEVO = 201 ± 24 N·m·s−1 and 178 ± 18 N·m·s−1 for 20 Hz and 80 Hz, respectively) among trials, and there were no effects of exercise on these variables in any trial. Furthermore, when pooling data, there were no effects of caffeine or carbohydrate ingestion on electrically evoked contractile properties.

Back to Top | Article Outline

Maximal cycling power.

During the NF trial, PMAX remained close to the 26-min value (1297 ± 190 W) throughout the trial (1302 ± 222 W at 116 min). Rehydration with water (WAT) or CES did not affect the PMAX response. Trials with caffeine ingestion (NF + CAFF, WAT + CAFF, and CES + CAFF) increased PMAX by 3.0 ± 1.0% from the first assessment (min 26) when compared with trials without caffeine ingestion (Fig. 4; P < 0.05). Nevertheless, an ergogenic effect of carbohydrates ingestion on PMAX was not present.



Back to Top | Article Outline

Plasma lactate and plasma glucose.

[Glu]plasma was significantly reduced during the NF trial from a resting value of 4.4 ± 1.6 to 3.8 ± 0.8 mM (at the end of exercise; P < 0.05). In NF + CAFF, and in the trials with water ingestion (WAT and WAT + CAFF), [Glu]plasma responded similarly to NF. Trials with carbohydrate ingestion (CES and CES + CAFF) increased [Glu]plasma after 8 min of exercise (P < 0.05), but there was no difference among trials at the end of exercise. [Lac]plasma did not increase above resting values (from 1.2 ± 0.5 at rest to 1.6 ± 1 mM during exercise) in the NF trial. Trials with carbohydrate ingestion (CES and CES + CAFF) increased [Lac]plasma at 8 min of exercise (~1.7 ± 0.2 mM, P < 0.05), but there were no differences between trials at the end of exercise.

Back to Top | Article Outline

V˙O2 and carbohydrate oxidation.

V˙O2 was similar among trials during the first stages of exercise, and it drifted similarly from the 15-min to 110-min period (5.5 ± 3%; P < 0.05). At the end of exercise (110 min), carbohydrate oxidation was higher during the trials with carbohydrate ingestion (CES and CES + CAFF; 2.5 g·min−1) than in the other trials (1.8 g·min−1; P < 0.05). However, when pooling data, there were no effects of caffeine on V˙O2 or carbohydrate oxidation.

Back to Top | Article Outline


The main purpose of this study was to determine the independent and combined effects of water, carbohydrate, and caffeine ingestion in maintaining short-term performance during prolonged exercise in the heat. For this purpose, these substances were ingested separately or in combination, and effectiveness was judged by the ability to maintain PMAX during exercise and quadriceps MVC after 120 min of cycling in the heat. Caffeine ingestion alone or in combination with water and carbohydrates increased PMAX early in exercise (i.e., 26 min) and maintained the effects throughout exercise (Fig. 4). The separate ingestion of caffeine (6 mg·kg−1 body weight) maintained MVC and increased PMAX above the NF control trial, despite similar dehydration and hyperthermia. The combined ingestion of water (replacing 97% of sweat losses), carbohydrate (~70 g·h−1), and caffeine (i.e., CES + CAFF trial) had the biggest effect on preventing the losses in MVC torque (Fig. 2A). The beneficial effects of fluid, carbohydrates, and caffeine on MVC seemed to be mediated by a parallel improvement in VA of leg extensors.

The assessment of muscle VA and electrically evoked contractile properties (PEVO, TEVO, and MTDEVO) permitted investigation into the causes of neuromuscular fatigue and the ergogenic effects of water, carbohydrate, and caffeine. In all trials, VA closely tracked the MVC response (i.e., reductions with NF and increases with CES + CAFF; Figs. 2 and 3), while electrically evoked contractile properties were unaffected by the treatments. These data strongly suggest that the ergogenic effect of water, carbohydrates, and caffeine is related to an improvement in the ability to voluntarily recruit the muscle (improvement on the central nervous system) and is unrelated to improvements in muscle contractile function. In contrast to the results of this study, Cureton et al. (8) have recently reported that the addition of caffeine to a sports drink increased evoked torque (i.e., direct effect on the muscle) without affecting quadriceps VA.

The main differences between the present study and Cureton's (8) are the timing of caffeine administration and the extent of hyperthermia. The caffeine dose in this study was purposively delivered in one bolus, 45 min before exercise. Although we did not measure plasma caffeine concentrations (a limitation of our data), it has been shown in subjects of similar characteristics (i.e., endurance trained cyclists) that ingesting 6 mg·kg−1 of caffeine in one bolus before exercise increases plasma caffeine concentrations above the level observed when the same dose is divided and provided at different times during exercise (4,6). In the study by Cureton et al. (8), however, caffeine was incorporated into a sports drink and ingested in small doses throughout exercise. Accordingly, the subjects achieved peak blood caffeine concentration after exercise, and the larger caffeine effects in MVC were observed 28 min after exercise cessation. In comparing both studies, it could be suggested that separating the ingestion of caffeine and sports drinks is a better strategy than coingestion. Incorporation of caffeine into sports drinks subdues caffeine dose to rehydration volume and timing to thirst or hydration habits. This may attenuate the potential ergogenicity that can be obtained from ingesting caffeine with a schedule independent of fluid ingestion.

Data from the present study support that the main ergogenic effect of caffeine is to counteract the central fatigue induced by hyperthermia during prolonged exercise. Postexercise MVC was tested quickly after finishing the 120 min of pedaling in the heat while subjects remained in the hot chamber to avoid cooling (TREC = 38.7 for rehydration trials; 39.4°C for trials without fluid ingestion). In contrast, Cureton and coworkers (8) tested MVC after 8, 18, and 28 min of rest in a thermoneutral environment (21°C), which likely lowered their core temperature. Thus, it is not surprising that by eliminating the effects of hyperthermia on central fatigue, the cited researchers did not observe effects on VA. In addition, other substances (carnitine, taurine, and vitamins) were coingested with caffeine in the Cureton et al. study, which may have contributed to the differences in the mechanisms for ergogenic effects between studies.

The magnitude of the ergogenic effect of caffeine (i.e., NF + CAFF) on MVC torque is similar to that of rehydrating with CES or WAT + CAFF (Fig. 2A), despite large differences in TREC and dehydration. It seems that caffeine ingestion can counteract the negative effects of dehydration and hyperthermia on maximal muscle performance. Strictly for maximal muscle performance (MVC and PMAX), ingesting a pill with caffeine (6 mg·kg−1 body mass) 45 min before prolonged exercise has similar ergogenic effects to the ingestion of 2440 mL of a sports drink (CES) in mitigating the detrimental effects of dehydration and hyperthermia on MVC. This does not mean that caffeine ingestion is equivalent to fluid ingestion; it is likely that the cardiovascular and thermoregulatory penalties associated with dehydration (i.e., cardiovascular drift and hyperthermia) would offset the ergogenic effects of caffeine if exercise were to be prolonged beyond 120 min.

Several mechanisms have been proposed to explain the ergogenic effects of caffeine on performance. It has been found, mostly in vitro, that caffeine could increase force by a direct effect on muscle, but at doses potentially toxic for humans (1). Other mechanisms proposed to explain the ergogenic effect of caffeine focus on its nature as central nervous system (CNS) stimulant. Caffeine is a potent adenosine receptor antagonist at physiologic doses. Davis et al. (9) found that intracerebroventricular infusion of caffeine in rats increased run time to fatigue by blocking the adverse effects of an adenosine receptor agonist. In humans, Kalmar and Cafarelli (16) reported that the ingestion of 6 mg·kg−1 of caffeine increased quadriceps MVC at rest and after a repeated-contractions protocol (24) by increasing quadriceps VA. In our study, testing hyperthermic humans, caffeine also prevented a significant decrease in quadriceps MVC and VA (Figs. 2B and 3B), which enforces the hypothesis that caffeine's effects are directed to the CNS.

Hyperthermia has been proposed as the principal mechanism responsible for force reductions after prolonged exercise in the heat (23). In this study, water ingestion (replacing 97% of sweat losses; WAT trial) reduced hyperthermia (0.7°C), but MVC was still lower than preexercise values (6% reduction). In contrast, the ingestion of a sports drink (CES trial) maintained MVC at preexercise values despite similar body temperature as in the WAT trial. Furthermore, when pooling data, an ergogenic effect of carbohydrate ingestion was observed on MVC maintenance (Fig. 2B). These results suggest that not only body temperature, but also the extra energy provided by the carbohydrates, is important for the maintenance of muscle force during prolonged exercise in the heat.

Hypoglycemia clearly reduces MVC through reductions in VA (i.e., affecting the CNS (22)). The subjects in this study did not reach hypoglycemic levels, but an ergogenic effect of carbohydrate ingestion on MVC through VA was observed (Fig. 3B). Carter and colleagues (3) found that frequent mouth-rinsing with a 6% carbohydrate solution improved endurance performance without affecting plasma glucose concentration. They have hypothesized that carbohydrate feeding stimulates carbohydrate receptors in the oral cavity, affecting motivation. A similar mechanism of CNS facilitation could be present in our experiment. Furthermore, carbohydrate ingestion (CES and CES + CAFF) increased carbohydrate oxidation, which may have provided extra energy to the CNS and allowed the improvements in VA. The mechanism(s) by which carbohydrate ingestion improved MVC after prolonged exercise remains speculative.

In conclusion, caffeine ingestion increases PMAX early in exercise and maintains the effects throughout prolonged exercise in the heat. There is a progressive benefit on leg force preservation (i.e., MVC) when rehydrating with water (97% rehydration; ~40% improvement; NS), adding carbohydrates (70 g·h−1; 80% improvement; P < 0.05), and ingesting caffeine (6 mg·kg−1; 110% improvement; P < 0.05). Of note, the ingestion of caffeine alone has an effect similar to rehydration with a carbohydrate-electrolyte solution on maintaining MVC. Further, it appears that carbohydrates and caffeine maintain quadriceps VA, which suggests that these ingredients are effective at counteracting the central fatigue induced by prolonged exercise in the heat.

The authors wish to thank the subjects for their invaluable contribution to the study. Partial support for this study was provided by the Gatorade Sports Science Institute. Juan Del Coso was supported by a predoctoral fellowship from the Castilla-La Mancha government in Spain.

Back to Top | Article Outline


1. Allen DG, Westerblad H. The effects of caffeine on intracellular calcium, force and the rate of relaxation of mouse skeletal muscle. J Physiol. 1995;487:331-42.
2. Anselme F, Collomp K, Mercier B, Ahmaidi S, Prefaut C. Caffeine increases maximal anaerobic power and blood lactate concentration. Eur J Appl Physiol Occup Physiol. 1992;65:188-91.
3. Carter JM, Jeukendrup AE, Jones DA. The effect of carbohydrate mouth rinse on 1-h cycle time trial performance. Med Sci Sports Exerc. 2004;36(12):2107-11.
4. Conway KJ, Orr R, Stannard SR. Effect of a divided caffeine dose on endurance cycling performance, postexercise urinary caffeine concentration, and plasma paraxanthine. J Appl Physiol. 2003;94:1557-62.
5. Coso JD, Mora-Rodriguez R. Validity of cycling peak power as measured by a short-sprint test versus the Wingate anaerobic test. Appl Physiol Nutr Metab. 2006;31:186-9.
6. Cox GR, Desbrow B, Montgomery PG, et al. Effect of different protocols of caffeine intake on metabolism and endurance performance. J Appl Physiol. 2002;93:990-9.
7. Coyle EF, Coggan AR, Hemmert MK, Lowe RC, Walters TJ. Substrate usage during prolonged exercise following a preexercise meal. J Appl Physiol. 1985;59:429-33.
8. Cureton KJ, Warren GL, Millard-Stafford ML, Wingo JE, Trilk J, Buyckx M. Caffeinated sports drink: ergogenic effects and possible mechanisms. Int J Sport Nutr Exerc Metab. 2007;17:35-55.
9. Davis JM, Zhao Z, Stock HS, Mehl KA, Buggy J, Hand GA. Central nervous system effects of caffeine and adenosine on fatigue. Am J Physiol Regul Integr Comp Physiol. 2003;284:399-404.
10. Fredholm BB, Bättig K, Holmén J, Nehlig A, Zvartau EE. Actions of caffeine in the brain with special reference to factors that contribute to its widespread use. Pharmacol Rev. 1999;51:83-133.
11. Fritzsche RG, Switzer TW, Hodgkinson BJ, Lee S-H, Martin JC, Coyle EF. Water and carbohydrate ingestion during prolonged exercise increase maximal neuromuscular power. J Appl Physiol. 2000;88:730-7.
12. Ftaiti F, Grelot L, Coudreuse JM, Nicol C. Combined effect of heat stress, dehydration and exercise on neuromuscular function in humans. Eur J Appl Physiol. 2001;84:87-94.
13. Graham TE. Caffeine and exercise: metabolism, endurance and performance. Sports Med. 2001;31:785-807.
14. Greer F, Morales J, Coles M. Wingate performance and surface EMG frequency variables are not affected by caffeine ingestion. Appl Physiol Nutr Metab. 2006;31:597-603.
15. Jeukendrup AE, Wallis GA. Measurement of substrate oxidation during exercise by means of gas exchange measurements. Int J Sports Med. 2005;26(Suppl. 1):S28-37.
16. Kalmar JM, Cafarelli E. Effects of caffeine on neuromuscular function. J Appl Physiol. 1999;87:801-8.
17. Kent-Braun JA, Le Blanc R. Quantitation of central activation failure during maximal voluntary contractions in humans. Muscle Nerve. 1996;19:861-9.
18. Krauchi K, Wirz-Justice A. Circadian rhythm of heat production, heart rate, and skin and core temperature under unmasking conditions in men. Am J Physiol Regul Integr Comp Physiol. 1994;267:819-29.
19. Lopes JM, Aubier M, Jardim J, Aranda JV, Macklem PT. Effect of caffeine on skeletal muscle function before and after fatigue. J Appl Physiol. 1983;54:1303-5.
20. Martin V, Millet GY, Martin A, Deley G, Lattier G. Assessment of low frequency fatigue with two methods of electrical stimulation. J Appl Physiol. 2004;97:1923-9.
21. Millet GY, Martin V, Lattier G, Ballay Y. Mechanism contributing to knee extensor strength loss after prolonged running exercise. J Appl Physiol. 2003;94:193-8.
22. Nybo L. CNS fatigue and prolonged exercise: effect of glucose supplementation. Med Sci Sports Exerc. 2003;35(4):589-94.
23. Nybo L, Nielsen B. Hyperthermia and central fatigue during prolonged exercise in humans. J Appl Physiol. 2001;91:1055-60.
24. Plaskett CJ, Cafarelli E. Caffeine increases endurance and attenuates force sensation during submaximal isometric contractions. J Appl Physiol. 2001;91:1535-44.
25. Saboisky J, Marino FE, Kay D, Cannon J. Exercise heat stress does not reduce central activation to non-exercised human skeletal muscle. Exp Physiol. 2003;88:783-90.
26. Thomas MM, Cheung SS, Elder GC, Sleivert GG. Voluntary muscle activation is impaired by core temperature rather than local muscle temperature. J Appl Physiol. 2006;100:1361-9.
27. Todd G, Butler JE, Taylor JL, Gandevia SC. Hyperthermia: a failure of the motor cortex and the muscle. J Physiol. 2005;563:621-31.
28. Williams JH, Barnes WS, Gadberry WL. Influence of caffeine on force and EMG in rested and fatigued muscle. Am J Phys Med. 1987;66:169-83.
29. Williams JH, Signorile JF, Barnes WS, Henrich TW. Caffeine, maximal power output and fatigue. Br J Sports Med. 1988;22:132-4.


©2008The American College of Sports Medicine