Journal Logo

BASIC SCIENCES: Original Investigations

Substrate Metabolism and Exercise Performance with Caffeine and Carbohydrate Intake


Author Information
Medicine & Science in Sports & Exercise: December 2008 - Volume 40 - Issue 12 - p 2096-2104
doi: 10.1249/MSS.0b013e318182a9c7
  • Free


Several studies report enhanced exercise capacity after caffeine ingestion (5,13,14,17,28). It was originally proposed that caffeine mobilized free fatty acids (FFA) from adipose tissue, resulting in higher rates of fat oxidation and sparing of muscle glycogen (5,17). More recently, this so-called "metabolic" theory has been dismissed as a universal explanation for the ergogenic effect of caffeine on endurance exercise performance, and it seems that caffeine exerts its effect via central fatigue mechanisms or by facilitating muscle function (11). Carbohydrate (CHO) ingestion can also delay the onset of fatigue and enhance exercise performance (3,7,9,21). However, the benefits of CHO ingestion have been attributed to the maintenance of plasma glucose concentrations and high rates of CHO oxidation late in exercise when muscle and liver glycogen stores are low(3).

Interestingly, two recent studies suggest that caffeine, in addition to a direct effect on exercise performance, may have another effect when ingested with CHO. These studies suggest that caffeine can increase the absorption (31) and oxidation (38) of CHO ingested during exercise. Firstly, Van Nieuwenhoven et al. (31) reported that caffeine (1.4 mg·kg−1) coingested with glucose (0.5 g·min−1) during 90 min of cycling produced higher rates of intestinal glucose absorption compared with glucose alone. Furthermore, in a previous study from our laboratory, we observed a 26% increase in the rate of exogenous CHO oxidation during 2 h of cycling at 64% V(dot)dot;O2max with caffeine (10 mg·kg−1) and glucose (0.8 g·min−1) ingestion compared with glucose alone (38). In that study, plasma glucose kinetics was not measured, and therefore we could only speculate about the effects of caffeine on glucose metabolism. It is also unknown whether this increase in exogenous CHO oxidation would result in an additional performance benefit.

Despite the large body of evidence for the separate ergogenic effects of caffeine and CHO, very few studies have investigated the effect of combined caffeine and CHO intake on exercise performance. One study (25) reported improved 1 h of time trial (TT) cycling performance when moderate amounts of caffeine (2.1, 3.2, and 4.5 mg·kg−1) were coingested with a 7% CHO solution during exercise. All three doses of caffeine resulted in faster performance times than placebo or CHO. Unfortunately, measures of substrate metabolism were not reported, and the mechanisms underlying improved performance remain unclear. A later study by Cox et al. (6) examined the effect of caffeine coingested with a 6.3% CHO solution on substrate metabolism and exercise performance during 2 h of steady-state (SS) cycling and a subsequent 7 kJ·kg−1 TT. In that study, caffeine ingestion enhanced TT performance by ∼3%; however, caffeine ingestion did not alter plasma FFA concentrations or RER, suggesting that caffeine can improve endurance exercise performance in the absence of altered substrate metabolism.

In a previous study from our laboratory, we used a very high dose of caffeine in combination with a relatively high CHO ingestion rate (38). These high rates of ingestion may not represent what athletes use in the field, and therefore the purpose of the present study was to determine exogenous CHO oxidation and glucose kinetics with a smaller dose of caffeine and lower rate of CHO ingestion during exercise. Compared with our previous study, subjects ingested less caffeine (5.3 vs 10 mg·kg−1) and less CHO (0.71 vs 0.80 g·min−1). We hypothesized that caffeine, when coingested with glucose, would increase exogenous CHO oxidation compared with glucose alone. More specifically, that caffeine would increase Ra glucose without increasing liver glucose output and increase Rd glucose resulting in higher rates of exogenous CHO oxidation. A second purpose was to determine whether combined ingestion of caffeine and CHO enhanced TT cycling performance compared with CHO alone. We hypothesized that, compared with CHO alone, the ingestion of caffeine and CHO would enhance TT cycling performance.



Ten endurance-trained male cyclists (age = 27 ± 7 yr, body mass = 71.4 ± 8.5 kg, V(dot)dot;O2max = 65.7 ± 6.1 mL·kg−1·min−1, and maximal power output = 345 ± 32 W; mean ± SD) volunteered to participate in this study. Habitual caffeine intake was assessed by questionnaire, and all subjects were identified as caffeine users to varying extent (186 ± 101 mg·d−1, range = 70-400 mg·d−1). Subjects were informed of the potential risks involved with the experimental procedures before providing their written consent. The study was approved by the Local Research Ethics Committee (Walsall REC, UK).

General design.

Each subject completed four experimental trials consisting of 105-min steady-state (SS) cycling at 55% Wmax (62% V(dot)dot;O2max) followed by a time trial (TT) lasting approximately 45 min. During the first trial, subjects ingested plain water (WAT) that served as a familiarization trial. Thereafter, subjects ingested either of the following: placebo (PLA), glucose (GLU), or glucose plus caffeine (GLU + CAF) solutions. The three experimental trials were performed in random order, using a double-blind crossover design, and separated by at least 7 d. Exogenous CHO oxidation was determined from the 13C enrichment of expired breath, and eight subjects received a primed continuous [6,6-2H2]glucose infusion for determination of plasma glucose kinetics (Ra and Rd glucose).

Preliminary testing.

One week before the start of the experiment, subjects performed an incremental test to exhaustion on an electromagnetically braked cycle ergometer (Lode Excalibur Sport, Groningen, The Netherlands) for the determination of maximal oxygen uptake (V(dot)dot;O2max) and maximal power output (Wmax) as previously described (1). Briefly, subjects began cycling at 95 W, followed by 35-W increments every 3 min thereafter. Breath-by-breath measurements were performed throughout exercise using an automated online gas analysis system (Oxycon Pro; Jaeger, Wuerzburg, Germany). Wmax values were used to determine 55% and 75% Wmax, which were later used in the experimental trials.

Diet and activity before testing.

To reduce the possibility that prior exercise/diet could influence measures of substrate metabolism and exercise performance, subjects recorded food intake and physical activity for 2 d before the first trial. The importance of maintaining the same diet and activities before all future trials was emphasized in considerable detail, and all subjects kept their original records to use as a guide. In addition, subjects followed a specific exercise/diet regimen, which has been shown to reduce the background shift (change in 13C) from endogenous substrate stores (34). Briefly, 5-7 d before each experimental trial, subjects performed an exhaustive bout of exercise to oxidize any 13C-enriched glycogen and refrained from eating foods with high natural 13C abundance thereafter. This included commercially available sports drinks and CHO derived from C4 plants, such as maize and cane sugar. Finally, subjects were asked to refrain from strenuous exercise and avoid alcohol and caffeine intake for 24 h before all trials.

Experimental trials.

On the morning of an experiment trial, subjects reported to the laboratory at 8 a.m. after a 10- to 12-h overnight fast. After sitting quietly for 10-15 min, a Teflon catheter (Venflon; Becton Dickinson, Plymouth, UK) was inserted into an antecubital vein of each arm to allow stable isotope infusion and repeated blood sampling during exercise. Resting blood (10 mL) and expired breath samples were collected immediately before subjects began cycling at 55% Wmax (190 ± 6 W). At the onset of exercise, a primed continuous [6,6-2H2]glucose infusion was administered (priming dose = 26.4 μmol·kg−1, continuous rate = 0.65 μmol·kg−1·min−1) using a calibrated syringe pump (Asena GS Syringe Pump; Alaris Medical Systems, Basingstoke, UK). Further blood (10 mL) and expired breath samples were collected every 15 min until cessation of exercise, along with measures of V(dot)dot;O2 and V(dot)dot;CO2, using an automated online gas analysis system (Oxycon Pro; Jaeger). In addition, RPE were taken every 15 min using the 6- to 20-point scale. HR was recorded continuously (15-s intervals) throughout exercise using a radiotelemetry HR monitor (Polar 625X, Kempele, Finland).

On completion of SS, the ergometer was adjusted to the cadence-dependent (linear) mode, and subjects were required to complete a set amount of work (688 ± 56 kJ) as fast as possible. The total amount of work to be performed was calculated using the formula:

The linear factor was individually adjusted, so that 75% Wmax was obtained when the subject pedaled at their preferred cadence. The only information available to the subjects during TT was elapsed work and percentage of work performed (i.e., 0% at the start and 100% on completion). Furthermore, subjects were not given any feedback on their performance until completion of the entire study. Throughout TT, no blood or respiratory measures were taken, and every effort was made to minimize any possible distractions. On completion of TT, a final blood sample (5 mL) was obtained.

Beverage ingestion.

At the onset of exercise, subjects consumed 5.5 mL·kg−1 followed by 2 mL·kg−1 at 15-min intervals throughout SS. Total fluid intake during SS was 1.25 ± 0.15 L, which provided 75 ± 9 g of CHO (0.71 ± 0.09 g·min−1) in the GLU and GLU + CAF trials. In addition, GLU + CAF provided 375 ± 45 mg of caffeine (5.3 mg·kg−1). To quantify exogenous CHO oxidation, CHO solutions were prepared using corn-derived glucose, which has a high natural abundance of 13C. In addition, ∼0.11 g·L−1 of [U-13C]glucose (99%; Cambridge Isotope Laboratories, Andover, MA) was added to the CHO solutions, resulting in final enrichments of 103.88 and 103.54 δ‰ versus Pee Dee Bellemnitella (PDB) for GLU and GLU + CAF, respectively. Throughout TT, subjects ingested plain water ad libitum with no further ingestion of the experimental beverages.


Blood samples were collected into prechilled vacutainers containing K3EDTA (Becton Dickinson) and stored on ice until centrifugation at 2300g for 10 min at a temperature of 4°C. After centrifugation, aliquots of the plasma were immediately frozen in liquid nitrogen and stored at −70°C until further analysis. Plasma samples were analyzed using commercially available spectrophotometric assays for glucose (Glucose HK; ABX Diagnostics, UK), Lactate (Lactic Acid; ABX Diagnostics Northampton, UK) free fatty acids (FFA, NEFA-C Kit; Alpha Laboratories, Eastleigh, UK), glycerol (Enzyplus Glycerol; Diffcamb Ltd., Mansfield, UK), and caffeine (Emit caffeine assay; Dade Behring, Milton Reynes, UK) concentrations using a semiautomatic analyzer (Cobas Mira Plus; ABX Diagnostics).

For the determination of plasma [6,6-2H2]glucose enrichment, plasma samples (150 μL) were deproteinized with 1 mL of acetone and derivatized using 160 μL of a heptafluorobutyric acid-ethylacetate mixture (1:1 ratio). Thereafter, the enrichment of the derivative was measured by gas chromatography/mass spectrometry. Briefly, 1 μL of the derivative was injected into an Agilent 6890N gas chromatograph equipped with a split/splitless injector and 7683 autosampler (Agilent Technologies, Stockport, UK). Mass spectra were obtained using an Agilent 5973N mass-selective detector (Agilent Technologies). Data were acquired by using selective ion monitoring for masses m/z 519 and 521. Breath samples were analyzed for 13C/12C ratio by continuous flow IRMS (GC, Trace GC Ultra; IRMS, Delta Plus XP; both Thermo Finnigan, Herts, UK). From indirect calorimetry (V(dot)dot;O2 and V(dot)dot;CO2) and stable isotope measurements (breath 13C enrichment and plasma [6,6-2H2]glucose enrichment), rates of total fat, total CHO, muscle glycogen, liver-derived glucose, and exogenous glucose oxidation were calculated.


Rates of total CHO and fat oxidation were calculated using stoichiometric equations (24), with the assumption that protein oxidation was negligible:

The isotopic enrichment of expired breath was expressed as δ per milliliter difference between 13C/12C ratio of the sample and a known laboratory reference standard according to the formula of Craig (8):

The δ13C was then related to the international standard PDB.

The rate of exogenous CHO oxidation was calculated using the formula:

where δExp is 13C enrichment of expired breath during exercise, δIng is 13C enrichment of the ingested beverage, δExpbkg is 13C enrichment of expired breath in the WAT trial (background), and k is the volume of CO2 produced by the oxidation of 1 g of glucose (k = 0.7467 L).

A methodological consideration when using 13CO2 in expired air to calculate exogenous CHO oxidation is the trapping of 13CO2 in the bicarbonate pool where a proportion of CO2 arising from the oxidation of glucose is temporarily retained (27). However, during exercise, V(dot)dot;CO2 increases severalfold, so that a physiological SS condition will occur and the 13CO2 in expired air will be equilibrated with the CO2/HCO3 pool. Under exercising conditions, recovery of 13CO2 from the oxidation of 13C-glucose will approach 100% after 60 min when the dilution in the bicarbonate pool becomes negligible (26). Consequently, calculations of substrate oxidation were performed over the final 45 min of exercise (60-105 min).

Rates of appearance (Ra) and disappearance (Rd) of glucose were calculated using the single pool non-SS equations of Steele (29), modified for use with stable isotopes (37).

where F is the infusion rate (μmol·kg−1·min−1), V is the volume of distribution (160 mL·kg−1), C1 and C2 are plasma glucose concentrations (mmol·L) at time points t1 and t2, respectively, and E1 and E2 are plasma glucose enrichment at time points t1 and t2, respectively. Ra represents the sum of hepatic glycogenolysis, gluconeogenesis, and glucose absorbed from the gut. Rd glucose was taken to represent plasma glucose oxidation with the assumption 100% of Rd glucose is oxidized during exercise (22). Because plasma glucose oxidation represents the oxidation of both glucose absorbed from the gut (exogenous glucose) and the contribution from the liver (glycogenolysis and/or gluconeogenesis), liver-derived glucose oxidation was calculated as plasma glucose oxidation minus exogenous CHO oxidation. Muscle glycogen oxidation was calculated as total CHO oxidation minus plasma glucose oxidation.

Statistical analysis.

All data are expressed as means ± SEM unless otherwise stated. One-way ANOVA was performed to study differences in TT performance between trials. Two-way (trial × time) ANOVA for repeated measures was performed to study differences in substrate metabolism, tracer enrichment, plasma glucose kinetics, and plasma metabolite concentrations. Significant effects were followed up by post hoc comparisons (Tukey HSD). Data analysis was performed using SPSS for Windows version 13.0 software (SPSS Inc., Chicago, IL) or by hand. Significance was accepted at P < 0.05.


SS Exercise

Whole-body oxygen consumption, HR, and RPE.

The workload used during SS (190 ± 6 W) elicited an average V(dot)dot;O2 of 2.89 ± 0.07, 2.85 ± 0.07, and 2.92 ± 0.08 L·min−1 for PLA, GLU, and GLU + CAF, respectively. Hence, the relative exercise intensity was not different between trials (62-63% V(dot)dot;O2max). Average HR during SS was also similar between PLA, GLU, and GLU + CAF (138 ± 3, 136 ± 2, and 136 ± 2 beats·min−1, respectively). However, RPE during SS was significantly lower in GLU + CAF (10.9 ± 0.2, P < 0.05) compared with PLA (12.0 ± 0.3) or GLU (11.8 ± 0.2).

Exogenous CHO oxidation.

Peak exogenous CHO oxidation rates were reached at the end of SS exercise (105 min) and were not significantly different between GLU and GLU + CAF trials (52.6 ± 2.7 and 49.1 ± 2.1 μmol·kg−1·min−1, respectively). Similarly, average exogenous CHO oxidation rates, during the final 45 min of exercise, were not significantly different between GLU and GLU + CAF trials (50.4 ± 2.8 and 46.8 ± 2.3 μmol·kg−1·min−1, respectively).

RER, CHO, and fat oxidation.

As shown in Table 1, RER was significantly lower in PLA compared with the two CHO trials (P < 0.05) but was not significantly different between GLU and GLU + CAF. Accordingly, total CHO oxidation was significantly higher with CHO ingestion compared with PLA (P < 0.05). The relative contribution from CHO oxidation increased from 47 ± 6% in PLA to 59 ± 3% in GLU and 58 ± 4% in GLU + CAF (P < 0.05; Fig. 1).

Whole-body oxygen consumption, RER, and total CHO and fat oxidation during SS exercise (60- to 105-min period).
Relative contribution of substrates to total energy expenditure calculated during the 60- to 105-min exercise period with PLA, GLU, or GLU + CAF ingestion. Values are means (n = 8). aSignificant difference between PLA and GLU, P < 0.05. bSignificant difference between PLA and GLU + CAF, P < 0.05.

Plasma glucose kinetics.

Isotopic SS was achieved during the final 45 min of exercise, with plasma [2H]glucose enrichment ranging between 0.017 and 0.020 TTR in PLA and between 0.010 and 0.011 TTR with CHO ingestion. Rates of appearance (Ra) and disappearance (Rd) of glucose were significantly higher with CHO ingestion than PLA but were not significantly different between GLU and GLU + CAF trials (P < 0.01; Table 2).

Rates of appearance (R a) and disappearance (R d) of glucose, exogenous CHO oxidation, muscle glycogen, and liver-derived glucose oxidation during SS exercise (60- to 105-min period).

Oxidation of CHO sources and their relative contribution to energy expenditure.

The relative contributions of various substrates to energy expenditure are shown in Figure 1. Exogenous CHO oxidation provided 19 ± 1% and 17 ± 1% of total energy in GLU and GLU + CAF, respectively, resulting in significantly lower estimated rates of liver-derived glucose oxidation compared with PLA (P < 0.05). The contribution from liver-derived glucose oxidation decreased from 13 ± 1% in PLA to 4 ± 1% in GLU and 5 ± 1% of in GLU + CAF. Overall rates of endogenous CHO oxidation were not significantly different between the three trials. Estimated muscle glycogen oxidation was not significantly different between the three trials, contributing 34 ± 5%, 36 ± 3%, and 36 ± 3% of total energy during PLA, GLU, and GLU + CAF, respectively.

Plasma metabolites.

Resting plasma glucose, lactate, FFA, and glycerol concentrations were not significantly different between trials (Figs. 2A-D). Plasma glucose concentrations remained stable (4.5-4.8 mmol·L−1) throughout SS with PLA ingestion (Fig. 2A). In the CHO trials, plasma glucose concentrations increased to peak values (∼5.9 mmol·L−1) during the first 30 min and were higher than PLA at several time points throughout SS (P < 0.05). Plasma glucose concentrations were not significantly different between GLU and GLU + CAF (5.29 ± 0.13 and 5.45 ± 0.08 mmol·L, respectively).

Plasma glucose (A), lactate (B), free fatty acid (FFA) (C), and glycerol (D) concentrations at rest and during SS exercise with PLA, GLU, or GLU + CAF ingestion. Values are means ± SEM (n = 10). aSignificant difference between PLA and GLU, P < 0.05. bSignificant difference between PLA and GLU + CAF, P < 0.05. cSignificant difference between GLU and GLU + CAF, P < 0.05.

During SS, plasma lactate concentrations were similar in GLU and PLA (1.26 ± 0.16 and 1.23 ± 0.11 mmol·L−1, respectively). However, plasma lactate concentrations during the final 30-45 min of SS were significantly higher in GLU + CAF compared with GLU and PLA (P < 0.05; Fig. 2B).

Plasma FFA concentrations decreased during the first 15-30 min before gradually increasing to reach peak values at the end of exercise. Plasma FFA concentrations were significantly higher in PLA compared with CHO trials during the final 45-60 min of SS (P < 0.05; Fig. 2C). Plasma FFA concentrations were not different between GLU and GLU + CAF. With the onset of exercise, plasma glycerol concentrations increased in all three trials and were significantly higher in PLA than in CHO trials during the final 30-45 min of SS (P < 0.05; Fig. 2D). Plasma glycerol concentrations were not different between GLU and GLU + CAF.

Caffeine was not detected in resting plasma samples of any trial, confirming that subjects abstained from caffeine intake before each trial as directed. Plasma caffeine concentrations increased throughout SS in GLU + CAF (P < 0.01; Fig. 3). Plasma caffeine concentrations were not significantly different after 105 min of SS and on completion of TT (31 ± 2 and 33 ± 2 μmol·L−1, respectively).

Plasma caffeine concentration at rest, during SS exercise, and on completion of TT with GLU + CAF ingestion. Values are means ± SEM (n = 10).

Time Trial

Performance data.

The results of TT are summarized in Table 3. Overall, CHO ingestion enhanced TT performance compared with PLA. Performance times were 47.40 ± 1.30, 45.45 ± 1.07, and 43.45 ± 0.86 min for PLA, GLU, and GLU + CAF, respectively. GLU enhanced TT performance by 4.4% compared with PLA (P < 0.05; 95% confidence interval (CI) = 1.1-7.5%). GLU + CAF enhanced TT performance by 4.6% compared with GLU (P < 0.05; 95% CI = 1.4-7.9%) and 9.0% compared with PLA (P < 0.05; 95% CI = 5.2-12.9%).

Average HR, power output, and time taken to complete a set amount of work.

Plasma metabolites.

On completion of TT, plasma glucose concentrations were significantly lower in PLA than in both CHO trials (Table 4; P < 0.05). Plasma lactate concentrations were not different between PLA and GLU; however, plasma lactate concentrations were significantly higher in GLU + CAF than in PLA (Table 4; P < 0.05). Plasma FFA concentrations tended to be higher in PLA compared with both CHO trials, but this did not reach significance (Table 4; P = 0.121). Plasma glycerol concentrations were significantly higher in PLA compared with both CHO trials (Table 4; P < 0.05).

Plasma glucose, lactate, FFA, and glycerol concentrations on completion of TT.


It is well established that caffeine (5,13,14,17,28) and CHO (3,7,9,21) ingestion can delay the onset of fatigue and enhance exercise performance. It is also accepted that CHO ingestion is beneficial to performance because it reduces the reliance on endogenous CHO stores (20). As such, there is considerable interest in developing ways to increase the delivery and oxidation of CHO ingested during exercise. Exogenous CHO oxidation is potentially limited by gastric emptying, intestinal absorption, hepatic glucose extraction and storage, muscle glucose uptake, or a combination of these factors (20). However, it seems that exogenous CHO oxidation is primarily limited by the rate of intestinal CHO absorption (19,20). One study recently reported increased rates of intestinal glucose absorption with caffeine ingestion (31). Therefore, we hypothesized that caffeine, coingested with glucose, would increase exogenous CHO oxidation when compared with glucose alone. The main finding of the present study was that coingesting caffeine (5.3 mg·kg−1) and glucose (0.71 g·min−1) did not increase exogenous CHO oxidation compared with an isoenergetic amount of glucose alone.

The finding that exogenous CHO oxidation was similar in GLU and GLU + CAF trials seems in contrast to our previous work (38) but may be related to the smaller dose of caffeine used in the present study as well as other differences in the experimental protocol. As previously mentioned, the main rate-limiting factor for exogenous CHO oxidation is thought to be intestinal CHO absorption. Glucose is absorbed from the intestinal lumen via the sodium-dependent glucose transporter 1 (SGLT1), which, at high ingestion rates (>1.2 g·min−1), may become saturated (19). The relatively low glucose ingestion rate (0.71 g·min−1) used in the present study is below the saturation of SGLT1. Therefore, possibly intestinal glucose absorption was not limiting the rate of exogenous CHO oxidation and that caffeine only exerts its effect once the glucose transporter is saturated. This is supported by the high oxidation efficiency observed in the present study. During the final 45 min of exercise, exogenous CHO oxidation averaged 0.65 g·min−1 in GLU. In other words, 92% of the ingested CHO was being oxidized. The high rate of exogenous CHO oxidation implies that a large amount of the ingested CHO was absorbed from the gastrointestinal tract. Hence, any effect of caffeine on intestinal glucose absorption may have been masked. Although this does appear to offer an obvious explanation for our findings, we believe that the situation is more complex and may not be due to the rate of CHO ingestion alone. As previously mentioned, ingesting 1.4 mg·kg−1 of caffeine during 90 min of cycling exercise increased intestinal glucose absorption by 23% (31). In that study, glucose ingestion rates (0.5 g·min−1) were lower than that of the present study and below the rate at which intestinal transport is thought to become rate-limiting; however, caffeine still increased intestinal glucose absorption. Individual subjects' differences may also have contributed to our findings. For example, during our previous study (38), we observed peak oxidation rates of 0.57 g·min−1 during the GLU trial, which, despite higher rates of CHO ingestion, is actually lower than that of the present study (0.69 g·min−1). This lower oxidation rate is difficult to explain. Both studies recruited cyclists of similarly trained status and exercised them at identical workloads under standardized laboratory conditions. Nonetheless, it appears that individual differences in the capacity for intestinal glucose absorption and/or exogenous CHO oxidation could be a contributing factor in determining whether caffeine increases the availability of CHO ingested during exercise.

Ingesting CHO increased Ra glucose compared with PLA (Table 2). Ra glucose represents glucose appearing from the gut (i.e., intestinal glucose absorption) and hepatic glucose output (23). In the present study, we did not determine the rate of appearance of glucose from the gut. Nonetheless, the observation that Ra glucose was not different between GLU and GLU + CAF trials (Table 2) suggests that caffeine intake did not influence intestinal glucose absorption.

Caffeine ingestion has been reported to reduce glucose uptake in contracting slow-twitch muscle fibers (33), which could potentially limit exogenous CHO oxidation during exercise. Contrary to this, we observed no difference in tracer determined rates of whole-body glucose disposal (i.e., Rd glucose) between GLU and GLU + CAF trials (Table 2). Similarly, using arterial-venous balance techniques, Graham et al. (12) reported that net glucose uptake was unaffected by caffeine ingestion during 60 min of cycling at 70% V(dot)dot;O2max. Collectively, these findings suggest that any negative effects of caffeine on muscle glucose uptake may be limited to resting conditions and possibly the recovery period after exercise.

The elevated plasma lactate concentrations with caffeine ingestion (Fig. 2B) is consistent with the literature (14,15,28,38), but the underlying mechanism is not clearly understood. Some studies attribute higher lactate concentrations with caffeine ingestion to a catecholamine-induced increase in the rate of muscle glycogenolysis. Indeed, epinephrine infusion can increase muscle glycogenolysis and lactate formation during SS exercise (10,35). However, it is important to note that in these studies, epinephrine concentrations were far higher than typically observed with caffeine ingestion. In the present study, estimated rates of muscle glycogen breakdown were similar in all three experimental conditions (Table 2 and Fig. 1), so it seems unlikely that elevated lactate concentrations were due to increased lactate production. In support of this, infusing epinephrine to the concentrations achieved with caffeine ingestion did not increase muscle glycogenolysis or lactate accumulation (2). Furthermore, direct measures of muscle metabolism during cycling exercise demonstrate higher arterial lactate concentrations with caffeine ingestion despite no increase in the rate of lactate release from the exercising leg (12). Taken together, these findings suggest that higher lactate concentrations with caffeine ingestion may be due to reduced clearance by other nonexercising tissues (possibly the liver or resting muscles) rather than a direct effect on the exercising muscles.

A secondary aim of the present study was to determine the effect of combined caffeine and CHO ingestion on TT cycling performance. Overall, CHO ingestion enhanced TT performance compared with placebo (Table 3), which is consistent with several studies reporting an ergogenic effect of CHO during exercise lasting more than 2 h (3,7,9). Ingesting CHO was associated with higher plasma glucose concentrations during SS exercise (Fig. 2A) and on completion of TT (Table 4), which may partially explain the ergogenic effect observed. In addition to the effect of CHO, combined ingestion of caffeine and CHO resulted in a further enhancement in performance (Table 3). GLU + CAF enhanced TT performance by 4.6% compared with GLU and 9.0% compared with PLA. This additive effect of caffeine has been reported by some (6,25) but not all (16,18) recent investigations. The amount of caffeine ingested during the present study (5.3 mg·kg−1 or 375 ± 45 mg) is equivalent to approximately eight caffeinated glucose gels. Interestingly, some studies report performance benefits with far lower and perhaps more practical caffeine doses. For example, Cox et al. (6) reported enhanced TT cycling performance with just 1.5 mg·kg−1 of caffeine. Furthermore, Kovacs et al. (25) reported that 3.2 mg·kg−1 of caffeine was more effective than 2.1 mg·kg−1 but equally as effective as 4.2 mg·kg−1. Clearly, it is difficult to determine the most effective caffeine dose given the range of experimental protocols adopted by the literature, and further research is required. Nonetheless, the findings of the present study and others (6,25) demonstrate that combined ingestion of caffeine and CHO enhances endurance exercise performance compared with CHO alone.

Unfortunately, we cannot provide a clear mechanism for the additional performance effect of caffeine. The original hypothesis proposed by Costill et al. (5,17) almost 30 yr ago suggested that caffeine mobilized FFA from adipose tissue, resulting in higher rates of fat oxidation and sparing of endogenous CHO. In contrast, our data show that caffeine ingestion neither elevated plasma FFA or glycerol concentrations (Figs. 2C and D) nor increased rates of whole-body fat oxidation (Table 1). These findings are in line with other recent reports that plasma FFA concentrations are not always elevated with caffeine ingestion (15), and even if higher plasma FFA concentrations are observed, there is no change in the RER (12,30,32). Possibly the metabolic effects of caffeine are limited to fasting conditions or that CHO feeding might inhibit the beneficial effects of caffeine on fat metabolism (6,25,36). Regardless, it is clear, during the present study at least, that the additional performance benefit of caffeine was not related to changes in either CHO or fat metabolism. An alternative mechanism may be a reduced perception of effort with caffeine ingestion, which, during self-paced tasks such as TT, allows the subject to spontaneously adopt a higher work rate (4). Our data offer some support for this theory with the lowest ratings of perceived exertion during SS exercise being reported in GLU + CAF.

In summary, the coingestion of caffeine (5.3 mg·kg−1) with CHO during exercise enhanced TT performance by 4.6% compared with CHO alone and 9.0% compared with water placebo. However, caffeine did not influence exogenous CHO oxidation or glucose kinetics during SS exercise.

This study was supported by a research grant from GlaxoSmithKline Consumer Healthcare, United Kingdom. The results of the present study do not constitute endorsement by ACSM.


1. Achten J, Gleeson M, Jeukendrup AE. Determination of the exercise intensity that elicits maximal fat oxidation. Med Sci Sports Exerc. 2002;34(1):92-7.
2. Chesley A, Hultman E, Spriet LL. Effects of epinephrine infusion on muscle glycogenolysis during intense aerobic exercise. Am J Physiol. 1995;268:E127-34.
3. Coggan AR, Coyle EF. Metabolism and performance following carbohydrate ingestion late in exercise. Med Sci Sports Exerc. 1989;21(1):59-65.
4. Cole KJ, Costill DL, Starling RD, Goodpaster BH, Trappe SW, Fink WJ. Effect of caffeine ingestion on perception of effort and subsequent work production. Int J Sport Nutr. 1996;6(1):14-23.
5. Costill DL, Dalsky GP, Fink WJ. Effects of caffeine ingestion on metabolism and exercise performance. Med Sci Sports. 1978;10(3):155-8.
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(3):990-9.
7. Coyle EF, Hagberg JM, Hurley BF, Martin WH, Ehsani AA, Holloszy JO. Carbohydrate feeding during prolonged strenuous exercise can delay fatigue. J Appl Physiol. 1983;55:230-5.
8. Craig H. Isotopic standards for carbon and oxygen and correction factors. Geochim Cosmochim Acta. 1957;12:133-49.
9. Febbraio MA, Chiu A, Angus DJ, Arkinstall MJ, Hawley JA. Effects of carbohydrate ingestion before and during exercise on glucose kinetics and performance. J Appl Physiol. 2000;89(6):2220-6.
10. Febbraio MA, Lambert DL, Starkie RL, Proietto J, Hargreaves M. Effect of epinephrine on muscle glycogenolysis during exercise in trained men. J Appl Physiol. 1998;84(2):465-70.
11. Graham TE. Caffeine and exercise: metabolism, endurance and performance. Sports Med. 2001;31(11):785-807.
12. 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. 2000;529:837-47.
13. Graham TE, Hibbert E, Sathasivam P. Metabolic and exercise endurance effects of coffee and caffeine ingestion. J Appl Physiol. 1998;85(3):883-9.
14. Graham TE, Spriet LL. Metabolic, catecholamine, and exercise performance responses to various doses of caffeine. J Appl Physiol. 1995;78(3):867-74.
15. Graham TE, Spriet LL. Performance and metabolic responses to a high caffeine dose during prolonged exercise. J Appl Physiol. 1991;71(6):2292-8.
16. Hunter AM, St Clair Gibson A, Collins M, Lambert M, Noakes TD. Caffeine ingestion does not alter performance during a 100-km cycling time-trial performance. Int J Sport Nutr Exerc Metab. 2002;12(4):438-52.
17. Ivy JL, Costill DL, Fink WJ, Lower RW. Influence of caffeine and carbohydrate feedings on endurance performance. Med Sci Sports. 1979;11(1):6-11.
18. Jacobson TL, Febbraio MA, Arkinstall MJ, Hawley JA. Effect of caffeine co-ingested with carbohydrate or fat on metabolism and performance in endurance-trained men. Exp Physiol. 2001;86(1):137-44.
19. Jentjens RL, Moseley L, Waring RH, Harding LK, Jeukendrup AE. Oxidation of combined ingestion of glucose and fructose during exercise. J Appl Physiol. 2004;96(4):1277-84.
20. Jeukendrup AE. Carbohydrate intake during exercise and performance. Nutrition. 2004;20(7-8):669-77.
21. Jeukendrup A, Brouns F, Wagenmakers AJ, Saris WH. Carbohydrate-electrolytefeedings improve 1 h time trial cycling performance. Int J Sports Med. 1997;18(2):125-9.
22. Jeukendrup AE, Raben A, Gijsen A, et al. Glucose kinetics during prolonged exercise in highly trained human subjects: effect of glucose ingestion. J Physiol. 1999;515:579-89.
23. Jeukendrup AE, Wagenmakers AJ, Stegen JH, Gijsen AP, Brouns F, Saris WH. Carbohydrate ingestion can completely suppress endogenous glucose production during exercise. Am J Physiol. 1999;276:E672-83.
24. Jeukendrup AE, Wallis GA. Measurement of substrate oxidation during exercise by means of gas exchange measurements. Int J Sports Med. 2005;26:S28-37.
25. Kovacs EM, Stegen J, Brouns F. Effect of caffeinated drinks on substrate metabolism, caffeine excretion, and performance. J Appl Physiol. 1998;85(2):709-15.
26. Pallikarakis N, Sphiris N, Lefèbvre P. Influence of the bicarbonate pool and on the occurrence of 13CO2 in exhaled air. Eur J Appl Physiol. 1991;63:179-83.
27. Robert JJ, Koziet J, Chauvet D, Darmaun D, Desjeux JF, Young VR. Use of 13C-labeled glucose for estimating glucose oxidation: some design considerations. J Appl Physiol. 1987;63(5):1725-32.
28. Spriet LL, MacLean DA, Dyck DJ, Hultman E, Cederblad G, Graham TE. Caffeine ingestion and muscle metabolism during prolonged exercise in humans. Am J Physiol. 1992;262(6):E891-8.
29. Steele R. Influences of glucose loading and of injected insulin on hepatic glucose output. Ann N Y Acad Sci. 1959;82:420-30.
30. Tarnopolsky MA, Atkinson SA, MacDougall JD, Sale DG, Sutton JR. Physiological responses to caffeine during endurance running in habitual caffeine users. Med Sci Sports Exerc. 1989;21(4):418-24.
31. Van Nieuwenhoven MA, Brummer RM, Brouns F. Gastrointestinal function during exercise: comparison of water, sports drink, and sports drink with caffeine. J Appl Physiol. 2000;89(3):1079-85.
32. Van Soeren MH, Graham TE. Effect of caffeine on metabolism, exercise endurance, and catecholamine responses after withdrawal. J Appl Physiol. 1998;85(4):1493-501.
33. Vergauwen L, Hespel P, Richter EA. Adenosine receptors mediate synergistic stimulation of glucose uptake and transport by insulin and by contractions in rat skeletal muscle. J Clin Invest. 1994;93(3):974-81.
34. Wagenmakers AJM, Rehrer NJ, Brouns F, Saris WHM, Halliday D. Breath 13CO2 background enrichment at rest and during exercise: diet related differences between Europe and America. J Appl Physiol. 1993;74(5):2353-7.
35. Watt MJ, Howlett KF, Febbraio MA, Spriet LL, Hargreaves M. Adrenaline increases skeletal muscle glycogenolysis, pyruvate dehydrogenase activation and carbohydrate oxidation during moderate exercise in humans. J Physiol. 2001;534:269-78.
36. Weir J, Noakes TD, Myburgh K, Adams B. A high carbohydrate diet negates the metabolic effects of caffeine during exercise. Med Sci Sports Exerc. 1987;19(2):100-5.
37. Wolfe RR, Chinkes DL. Isotope Tracers in Metabolic Research: Principles and Practice of Kinetic Analysis. 2nd ed. Hoboken (NJ): John Wiley and Sons; 2005. pp. 36-41.
38. Yeo SE, Jentjens RL, Wallis GA, Jeukendrup AE. Caffeine increases exogenous carbohydrate oxidation during exercise. J Appl Physiol. 2005;99(3):844-50.


©2008The American College of Sports Medicine