International Journal of Sports Physiology and Performance, 2016, 11, 164 -171 http://dx.doi.org/10.1123/ijspp.2015-0133 © 2016 Human Kinetics, Inc.
ORIGINAL INVESTIGATION
Oral Presence of Carbohydrate and Caffeine in Chewing Gum: Independent and Combined Effects on Endurance Cycling Performance Katherine T. Oberlin-Brown, Rodney Siegel, Andrew E. Kilding, and Paul B. Laursen The oral presence of carbohydrate (CHO) and caffeine (CAF) may independently enhance exercise performance, but their influence on performance during prolonged exercise is less known. Aim: To determine the independent and combined effects of CHO and CAF administered in chewing gum during a cycling time trial (TT) after prolonged exercise. Method: Eleven male cyclists (32.2 ± 7.5 y, 74.3 ± 6.8 kg, 60.2 ± 4.0 mL · kg–1 · min–1 V˙O2peak) performed 4 experimental trials consisting of 90-min constant-load cycling at 80% of their second ventilatory threshold (207 ± 30 W), followed immediately by a 20-km TT. Under double-blinded conditions, cyclists received placebo (PLA), CHO, CAF, or a combined CHO+CAF chewing gum at 0-, 5-, 10-, and 15-km points of the TT. Results: Overall TT performance was similar across experimental and PLA trials (%mean difference ± 90%CL 0.2% ± 2.0%, 0.4% ± 2.2%, 0.1% ± 1.8% for CHO, CAF, and CHO+CAF). Compared with PLA, mean power output tended to be higher in the first 2 quarters of the TT with CHO (1.6% ± 3.1% and 0.8% ± 2.0%) and was substantially improved in the last 2 quarters during CAF and CHO+CAF trials (4.2% ± 3.0% and 2.0% ± 1.8%). There were no differences in average heart rate (ES 0.6). Conclusion: After prolonged constant-load cycling, the oral presence of CHO and CAF in chewing gum, independently or in combination, did not improve overall performance but did influence pacing. Keywords: buccal absorption, power output, pacing, ergogenic, sport nutrition Carbohydrate (CHO) ingestion during exercise improves endurance performance over prolonged (>60 min) durations of moderate-intensity exercise (~65–70% maximal oxygen consumption ˙ O2max]).1 More recently, however, the presence of CHO in the oral [V cavity has been shown to improve performance over short-duration ˙ O2max) exercise2–4 irrespective of (75%V endogenous levels of CHO (muscle and liver glycogen). While the exact mechanisms remain to be identified, it is suggested that CHO’s ergogenic nature may be centrally mediated, involving activation of oral receptors and subsequently brain reward centers, allowing for greater motivation, reduced perception of effort, and increased corticomotor excitability.5,6 For example, Chambers et al5 showed, via functional magnetic resonance imaging, that mouth rinsing with both glucose and nonsweet maltodextrin enhanced afferent activity in the insula, anterior cingulate cortex, and striatum, areas of the brain associated with arousal and motor output. In addition, Gant et al6 showed the oral presence of CHO resulted in increased motor-evoked potentials and a subsequent 2% increase in motor output during a maximal voluntary isometric contraction in both fresh and fatigued muscle states (ie, independent of the extent of muscle fatigue and perception of voluntary force) and was not affected by plasma glucose concentration. Collectively, these findings suggest that CHO receptors in the mouth play an important Oberlin-Brown and Laursen are with High Performance Sport New Zealand, Auckland, New Zealand. Siegel and Kilding are with Sports Performance Research Inst New Zealand (SPRINZ), Auckland University of Technology, Auckland, New Zealand. Address author correspondence to Katherine Oberlin-Brown at [email protected]. 164
role in transducing energy density afferent information, which may influence work capacity and exercise performance, even in fatigued or energy depleted states.6 Caffeine (CAF) (1,3,7-trimetilxanthine) is widely used by athletes as an ergogenic aid for enhancing training and competition performance. Extensive evidence shows that CAF exhibits ergogenic effects across a wide range of exercise intensities, durations, and modes (for review see Burke7). Proposed mechanisms underlying its ergogenic effects have traditionally included increased free-fatty-acid oxidation, sparing of endogenous glycogen, and increased mobilization of intracellular calcium.8 However, it is now acknowledged that CAF’s primary ergogenic effect is centrally oriented, involving its antagonist action on adenosine receptors in the brain and the central nervous system.9 Consequently, CAF improves exercise performance by lowering feelings of pain and effort perception10 and improving motor function.11 The pharmokinetics of CAF are well understood. After ingestion, CAF quickly enters the bloodstream via gastrointestinal absorption. Absorption is typically completed approximately 1 hour after ingestion, although factors such as dosage and dose formulation can alter absorption-rate constants (Kamimori et al12). Peak CAF plasma concentrations are usually attained 15 to 120 minutes postingestion,12,13 and as such it has been routine for researchers and athletes to administer oral doses of CAF 1 hour before performance to ensure peak plasma concentrations.14 However, limited research exists to verify the efficacy of such practice on performance. Ryan et al13 showed that timing commencement of a 7-kJ/kg cycling time trial (TT) with peak plasma CAF did not improve performance compared with control (42.6 ± 0.27 and 41.8 ± 0.32 min CAF caffeine –1 h and –2 h before exercise [40.7 ± 0.15 min]). Moreover, CAF
Downloaded by ETSU on 09/19/16, Volume 11, Article Number 2
Oral Carbohydrate and Caffeine: Effect on Endurance Performance 165
immediately prior (–5 min) resulted in a 4.9% faster performance time than with control (ES –0.99) despite plasma CAF levels being comparable. Paton et al15 showed that providing 300 mg CAF, via a chewing gum, during high-intensity cycle exercise (2 × 5 × 30-s sprint/30 s active recovery) improved mean power output by 5.4% (90%CL ±3.6%; ES 0.25). Thus, from a number of viewpoints it appears that the ergogenic effect of CAF may not be proportional to the plasma CAF concentration and that the faster buccal absorption (~5 min) may deliver CAF to adenosine-receptor sites faster. The ability to perform an “end spurt” during the final stages of endurance competitions—when fatigue is highest—is considered critical and is often the moment that determines an athlete’s overall finishing position. For example, the winning margins at the 2012 London Olympic Games in the men’s triathlon and marathon were 0.17% and 0.34%, respectively. In light of the limited research examining the ergogenic effect of CHO and CAF presence in the oral cavity under these conditions, the primary aim of this study was to examine the independent and combined effects of CHO and CAF chewing gums on self-paced cycling TT performance after a period of prolonged cycling. We hypothesized that both CHO and CAF would improve time-trial performance independently but that in combination they would exhibit a greater ergogenic effect due to increased central stimulation of brain areas that regulate subconscious pacing (ie, CHO acting on motor reward center and CAF on the adenosine receptors in the brain).
Methods Participants Eleven endurance-trained male cyclists or triathletes (mean ± SD age 32.2 ± 7.5 y, mass 74.3 ± 6.8 kg, sum of 8 skinfolds 67.8 ± ˙ O2peak 60.2 ± 4.0 mL · kg–1 · min–1, maximum aerobic 13.3 mm, V power 331 ± 34 W) with at least 1 year of endurance-training history volunteered for this study. Participants were provided with written and verbal information concerning the experimental procedures and potential risks before providing written consent to participate. The study was approved by the Auckland University of Technology ethics committee. Participants were asked to maintain their regular training regimen throughout the duration of the study and to refrain from strenuous exercise on the day before each test. At the time of testing, participants were riding an average of 10 ± 2 h/wk (range 8–14 h/wk). Trials were separated by at least 3 days (range 3–7 d) and commenced at the same time of day to minimize the effects of diurnal variation. Participants were provided with instructions on how to achieve a minimum of 6 g CHO/kg body weight (BW) on the day before each trial and 1 g CHO/kg BW in their pretrial meal and abstain from CAF consumption on the day of their experimental trials. The diet was recorded by the participant for the familiarization trial, checked by the primary investigator, and then replicated for the remaining 4 experimental trials. All tests were performed on a Velotron cycle ergometer (RacerMate, Seattle, WA), which was set up according to the dimensions of each participant’s own bicycle and fitted with each cyclist’s own cycling pedals. All cycling tests were performed in a well-ventilated and temperature-controlled laboratory (18.6°C ± 0.8°C and 60% ± 7% relative humidity).
Incremental Cycling Test An incremental cycling test to exhaustion was used to determine V˙O2peak. The test started at a power output of 100 W and consisted
of a continuous step protocol with 5-minute stages. Power output was progressively increased by 50 W at the end of each 5-minute stage until volitional exhaustion. Participants cycled at their individualized and self-selected optimal cadence, so that an accurate reflection of maximum aerobic power could be determined. The test was terminated voluntarily by the participant or when cadence dropped below 60 rpm. Gas exchange was measured using a ParvoMedics TrueOne 2400 diagnostic system (Salt Lake City, UT). The workload (W and W/kg) corresponding to the second ventilation threshold (VT2) was determined using the methods of Davis et al.16 To familiarize participants with the 20-km TT distance after a fatiguing bout of exercise, participants performed a 20-km TT after completion of the incremental test, along with a full familiarization (described herein) on the second visit.
Experimental Trials Participants completed a total of 4 experimental trials, in a controlled, double-blind, repeated-measures, crossover experimental design. Participants served as their own control under 1 of 4 experimental conditions: chewing gum containing CHO (4 × ~1.8 g sucrose per piece + artificial sweeteners [AS]), CAF (4 × 50 mg CAF per piece + AS), CHO+CAF (4 × ~1.8 g sucrose + 50 mg CAF per piece + AS), or placebo (PLA) (4 × AS). A total of 200 mg (average of 2.7 mg/kg BW) CAF was selected to simulate the actual competition doses used by athletes for competition.7 Each trial began with a 90-minute constant-load cycle consisting of a standardized warm-up of 3 minutes at a power output equivalent to 40% of VT2, followed by 87 minutes at 80% of VT2 (207 ± 30 W, 63% ± 6% of maximal aerobic power, 67% ± 6% of V˙O2peak). The determined exercise intensity of 80% VT2 power output was selected to standardize relative work levels and to mimic realistic race conditions.17 Throughout the constant-load phase, participants self-selected pedaling rate as the Velotron ergometer controls power output irrespective of cadence. Time elapsed, cadence, and power output were visible throughout.
20-km TTs After the 90-minute constant-load cycle, participants were allowed 3.5 minutes of rest before commencing the 20-km TT. A verbal warning was given 1 minute before the start and then a 5-second countdown. Each TT began from a standing, stationary start with the starting gear ratio standardized for each participant from the 2 familiarization trials. Cyclists were instructed to complete all TTs in the fastest time possible and to adopt the same self-selected, individualized pacing strategy used previously. Instantaneous feedback for distance was provided, but participants were blinded to power output, cadence, speed, and heart rate. One piece (3 g) of chewing gum (EPSAR Co, Australia) was provided at the start of the 20-km TT and for every 25% of the TT completed (ie, 0, 5, 10, 15 km). The delivery procedure of the gum and application time interval (ie, approximately every 7.5–8 min) was chosen to replicate previous CHO mouth-rinse studies.2,3 Participants were instructed to chew their gum for 3 km (~5 min) before expelling the contents. On completion of the trial, they were asked to guess which gum they believed they had received and whether they experienced gastrointestinal discomfort during the trial. Time, power output, and cadence were automatically recorded at a frequency of 1 Hz by the Velotron software (Velotron Coaching Software, RacerMate, USA) and averaged for each 1 km. To minimize any potential effects of hydration and heat strain on
IJSPP Vol. 11, No. 2, 2016
166 Oberlin-Brown et al
exercise performance, a small fan (wind speed 2.8 ± 0.3 m/s) was positioned in front of the participant and water consumption was ad libitum during both the 90-minute constant-load phase and the 20-km TT.
Blood Analysis A capillary blood sample was drawn preexercise, immediately after the 90-minute constant-load cycle, and 3 minutes after completion of the 20-km TT. Samples were collected from the right ear lobe using standard techniques. During each blood draw, 2 separate samples were collected and analyzed for blood lactate (Lactate Pro analyzer, Arkray, Tokyo, Japan) and blood glucose concentration (CareSens II glucose analyzer, i-SENS Inc, Korea).
Downloaded by ETSU on 09/19/16, Volume 11, Article Number 2
Statistical Analysis Data are presented as mean ± SD and were log transformed before statistical analysis to reduce bias arising from nonuniformity of error. To make inferences about the true (population) effects of CHO and CAF on performance, the uncertainty of the effect was expressed as 90% confidence limits, with probabilities of the true value representing a substantial change in performance.18 An effect was deemed unclear if its confidence interval overlapped the thresholds for substantiveness, meaning that the effect could be substantially positive or negative. An estimate of the smallest substantial change in power output was required to make these inferences. Paton and Hopkins19 estimate the smallest effects of 0.5% to 1.5% in mean power output, based on variability of competitive cyclists’ performance in cycling TT events free of drafting. A smallest substantial change in endurance-TT performance was assumed to be a reduction or increase in performance power
output of 1% or more. For all other measures, 0.2 times the baseline between-subjects SD was used.18
Results Performance Response Performance times and power output during the 20-km TT were not significantly different between experimental trials (Table 1). Pairwise comparisons revealed positive insubstantial improvements for all experimental trials relative to PLA (Table 2). Individual results for each trial are presented in Figure 1. Three participants substantially improved mean power output with CHO versus PLA (6.9%, 2.4%, and 4.6%), 6 had substantial reductions (–1.2, –1.5, –2.5, –3.7, and –2.6), and 2 had trivial differences (0.0% and –0.4%). With CAF, 4 participants substantially improved (1.3, 3.3, 3.6, and 7.9) compared with PLA, 3 performed substantially worse (–3.4, –4.7, and –2.6), and the remaining 3 showed trivial differences (–0.8, 0.0, and –0.3). In the combined CHO+CAF condition, 4 participants showed substantial improvement (1.3, 5.1, 2.3, and 3.4), 3 substantially worsened (–3.5, –2.2, and –2.3), and the remaining 3 presented with trivial differences (0.3, –0.8, and –0.9). Analysis of the pacing response for each 5-km quarter is presented in Figure 2, and the associated pairwise comparisons are presented in Table 3. Mean power output with CHO was higher than PLA in the first 2 quarters of the TT (CHO 288 ± 37 and 266 ± 35 vs PLA 284 ± 40 and 264 ± 35 W for 0–5 km and 5–10 km respectively). In contrast, with CAF and CHO+CAF, mean power output was higher in the final 2 quarters (CAF 263 ± 39 and 284 ± 42 and CHO+CAF 260 ± 35 and 279 ± 43 vs PLA 259 ± 35 and 273 ± 41 W, for 10–15 km and 15–20 km, respectively).
Table 1 Performance Measures Obtained During the 20-km Time Trial for Each Condition (Mean ± SD) Measure Time (m:s) Mean power output (W)
Placebo
Carbohydrate
Caffeine
Carbohydrate + caffeine
32:27 ± 1:57
32:25 ± 1:45
32:20 ± 1:57
32:26 ± 1:51
270 ± 37
271 ± 35
273 ± 40
270 ± 37
Table 2 Pairwise Comparisons Quantifying the Magnitude of Effect of Different Chewing-Gum Contents on 20-km Time-Trial Mean Power Output Treatment effecta
% effect, mean ± 90%CLb
Effect size, mean ± 90%CLb
Qualitative inferencec
carbohydrate vs placebo
0.2 ± 2.0
0.01 ± 0.12
Possibly trivial
caffeine vs placebo
0.4 ± 2.2
0.03 ± 0.13
Unclear
carbohydrate + caffeine vs placebo
0.1 ± 1.8
0.01 ± 0.11
Possibly trivial
carbohydrate vs placebo
0.3 ± 1.6
0.02 ± 0.10
Possibly trivial
caffeine vs placebo
–0.4 ± 1.5
–0.02 ± 0.09
Unclear
carbohydrate + caffeine vs placebo
–0.1 ± 1.2
–0.01 ± 0.08
Unclear
Experimental treatment effects vs placebo
Within-treatment effects, independent and combined
Units of change are percentages for all measures derived from back-transformed log data. CL: Add and subtract this number to the difference to obtain the 90% confidence limits for the true difference. c Based on smallest beneficial or harmful change in performance of 1%.
a
b
IJSPP Vol. 11, No. 2, 2016
Downloaded by ETSU on 09/19/16, Volume 11, Article Number 2
Oral Carbohydrate and Caffeine: Effect on Endurance Performance 167
Figure 1 — Individual responses to experimental treatments relative to placebo for percentage change in mean 20-km time-trial power output. Black circles = carbohydrate; striped squares = caffeine; white squares = carbohydrate + caffeine; gray bar = smallest worthwhile change of 1%.
Physiological Responses There were no substantial differences in average and maximal heart rates (163 ± 13, 164 ± 11, 162 ± 12, 163 ± 13, and 176 ± 11, 176 ± 10, 178 ± 12, 176 ± 10 for PLA, CHO, CAF, and CHO+CAF, respectively) across conditions, indicative of consistent maximal effort across trials. Cohen effect sizes were 0.2).
in CAF-containing trials—CAF (0.4% ± 2.2%) and CHO+CAF (0.1% ± 1.8%)—compared with PLA for overall performance, although pacing was altered. Specifically, CAF and CHO+CAF exhibited a possibly harmful effect on mean power output for the first 2 quarters of the TT (CAF –1.3 ± 3.6 and –0.4% ± 2.2% and CHO+CAF –1.1% ± 2.4% and –1.2% ± 2.0%). Conversely, in the later stages of the TT, we observed an unclear effect in the third quarter (CAF 1.5% ± 2.2% and CHO+CAF 0.4% ± 2.5%) but a very likely beneficial effect in the last quarter (CAF 4.2% ± 3.0% and CHO+CAF 2.0% ± 1.8%). Given these findings, our results suggest that CAF is not acting via suggested sensory pathways but, rather, via absorptive pathways and eliciting central effects, likely within 5 minutes (Kamimori et al12). Our findings may relate to our CAF-administration methods—4 × 50 mg CAF every 5 km (total = 200 mg; average 2.7 mg/kg BW) during the TT rather than a single bolus before exercise (3.0 mg/kg BW) as used by Ryan et al13 and Paton et al15—resulting in insufficient delivery of CAF to adenosine receptors in the early stages of the TT and, hence, a delayed ergogenic effect. Finally, the fact that performance was substantially improved in the final quarter of the trial, when endogenous CHO stores would have been at their lowest and, conversely, fatigue levels highest, might suggest that providing CAF during endurance exercise may enable cyclists to override fatigue signals associated with exercise-induced fatigue.23,24 Another aspect considered vital in studies of highly trained performers is the individual performance responses. In the current study, results revealed considerable interindividual and intraindividual variation to the independent and combined oral presence of CHO and CAF, suggesting the presence of nonresponders or adverse CAF responders. However, it cannot be ruled out that the inconsistency in performance responses may be due to the small sample size and/or a sampling error variance. In this regard, we acknowledge some limitations with our study. The placebo effect positively influences performance outcomes, as participants are
Although the current work did not yield any definitive performance improvements during a 20-km cycling TT, the data suggest that under conditions of fatigue and reduced glycogen, the oral presence of CHO and/or CAF in chewing gum could temporarily override the anticipatory regulation strategy and subconscious pacing. CHO appears to have a relatively immediate effect—increased power in the first half of the trial—but timing its use toward the latter stages may be critical to prevent a subsequent decline in performance due to overpacing. In contrast, CAF appears to have a delayed ergogenic effect shown through the increased power in the second half of the TT; this suggests that earlier administration of CAF might be best. For example, if athletes had chewed earlier (before the TT) or used a larger single dose at the start, rather than spacing it evenly throughout the TT, the ergogenic effects might have been greater and/or occurred earlier. That both CHO and CAF appeared to acutely alter subconscious motor output during an exercise-fatigued state warrants further investigation to determine whether an ergogenic effect occurs when using an oral CHO gum, in combination with ingestion of CHO early in exercise (
ORIGINAL INVESTIGATION
Oral Presence of Carbohydrate and Caffeine in Chewing Gum: Independent and Combined Effects on Endurance Cycling Performance Katherine T. Oberlin-Brown, Rodney Siegel, Andrew E. Kilding, and Paul B. Laursen The oral presence of carbohydrate (CHO) and caffeine (CAF) may independently enhance exercise performance, but their influence on performance during prolonged exercise is less known. Aim: To determine the independent and combined effects of CHO and CAF administered in chewing gum during a cycling time trial (TT) after prolonged exercise. Method: Eleven male cyclists (32.2 ± 7.5 y, 74.3 ± 6.8 kg, 60.2 ± 4.0 mL · kg–1 · min–1 V˙O2peak) performed 4 experimental trials consisting of 90-min constant-load cycling at 80% of their second ventilatory threshold (207 ± 30 W), followed immediately by a 20-km TT. Under double-blinded conditions, cyclists received placebo (PLA), CHO, CAF, or a combined CHO+CAF chewing gum at 0-, 5-, 10-, and 15-km points of the TT. Results: Overall TT performance was similar across experimental and PLA trials (%mean difference ± 90%CL 0.2% ± 2.0%, 0.4% ± 2.2%, 0.1% ± 1.8% for CHO, CAF, and CHO+CAF). Compared with PLA, mean power output tended to be higher in the first 2 quarters of the TT with CHO (1.6% ± 3.1% and 0.8% ± 2.0%) and was substantially improved in the last 2 quarters during CAF and CHO+CAF trials (4.2% ± 3.0% and 2.0% ± 1.8%). There were no differences in average heart rate (ES 0.6). Conclusion: After prolonged constant-load cycling, the oral presence of CHO and CAF in chewing gum, independently or in combination, did not improve overall performance but did influence pacing. Keywords: buccal absorption, power output, pacing, ergogenic, sport nutrition Carbohydrate (CHO) ingestion during exercise improves endurance performance over prolonged (>60 min) durations of moderate-intensity exercise (~65–70% maximal oxygen consumption ˙ O2max]).1 More recently, however, the presence of CHO in the oral [V cavity has been shown to improve performance over short-duration ˙ O2max) exercise2–4 irrespective of (75%V endogenous levels of CHO (muscle and liver glycogen). While the exact mechanisms remain to be identified, it is suggested that CHO’s ergogenic nature may be centrally mediated, involving activation of oral receptors and subsequently brain reward centers, allowing for greater motivation, reduced perception of effort, and increased corticomotor excitability.5,6 For example, Chambers et al5 showed, via functional magnetic resonance imaging, that mouth rinsing with both glucose and nonsweet maltodextrin enhanced afferent activity in the insula, anterior cingulate cortex, and striatum, areas of the brain associated with arousal and motor output. In addition, Gant et al6 showed the oral presence of CHO resulted in increased motor-evoked potentials and a subsequent 2% increase in motor output during a maximal voluntary isometric contraction in both fresh and fatigued muscle states (ie, independent of the extent of muscle fatigue and perception of voluntary force) and was not affected by plasma glucose concentration. Collectively, these findings suggest that CHO receptors in the mouth play an important Oberlin-Brown and Laursen are with High Performance Sport New Zealand, Auckland, New Zealand. Siegel and Kilding are with Sports Performance Research Inst New Zealand (SPRINZ), Auckland University of Technology, Auckland, New Zealand. Address author correspondence to Katherine Oberlin-Brown at [email protected]. 164
role in transducing energy density afferent information, which may influence work capacity and exercise performance, even in fatigued or energy depleted states.6 Caffeine (CAF) (1,3,7-trimetilxanthine) is widely used by athletes as an ergogenic aid for enhancing training and competition performance. Extensive evidence shows that CAF exhibits ergogenic effects across a wide range of exercise intensities, durations, and modes (for review see Burke7). Proposed mechanisms underlying its ergogenic effects have traditionally included increased free-fatty-acid oxidation, sparing of endogenous glycogen, and increased mobilization of intracellular calcium.8 However, it is now acknowledged that CAF’s primary ergogenic effect is centrally oriented, involving its antagonist action on adenosine receptors in the brain and the central nervous system.9 Consequently, CAF improves exercise performance by lowering feelings of pain and effort perception10 and improving motor function.11 The pharmokinetics of CAF are well understood. After ingestion, CAF quickly enters the bloodstream via gastrointestinal absorption. Absorption is typically completed approximately 1 hour after ingestion, although factors such as dosage and dose formulation can alter absorption-rate constants (Kamimori et al12). Peak CAF plasma concentrations are usually attained 15 to 120 minutes postingestion,12,13 and as such it has been routine for researchers and athletes to administer oral doses of CAF 1 hour before performance to ensure peak plasma concentrations.14 However, limited research exists to verify the efficacy of such practice on performance. Ryan et al13 showed that timing commencement of a 7-kJ/kg cycling time trial (TT) with peak plasma CAF did not improve performance compared with control (42.6 ± 0.27 and 41.8 ± 0.32 min CAF caffeine –1 h and –2 h before exercise [40.7 ± 0.15 min]). Moreover, CAF
Downloaded by ETSU on 09/19/16, Volume 11, Article Number 2
Oral Carbohydrate and Caffeine: Effect on Endurance Performance 165
immediately prior (–5 min) resulted in a 4.9% faster performance time than with control (ES –0.99) despite plasma CAF levels being comparable. Paton et al15 showed that providing 300 mg CAF, via a chewing gum, during high-intensity cycle exercise (2 × 5 × 30-s sprint/30 s active recovery) improved mean power output by 5.4% (90%CL ±3.6%; ES 0.25). Thus, from a number of viewpoints it appears that the ergogenic effect of CAF may not be proportional to the plasma CAF concentration and that the faster buccal absorption (~5 min) may deliver CAF to adenosine-receptor sites faster. The ability to perform an “end spurt” during the final stages of endurance competitions—when fatigue is highest—is considered critical and is often the moment that determines an athlete’s overall finishing position. For example, the winning margins at the 2012 London Olympic Games in the men’s triathlon and marathon were 0.17% and 0.34%, respectively. In light of the limited research examining the ergogenic effect of CHO and CAF presence in the oral cavity under these conditions, the primary aim of this study was to examine the independent and combined effects of CHO and CAF chewing gums on self-paced cycling TT performance after a period of prolonged cycling. We hypothesized that both CHO and CAF would improve time-trial performance independently but that in combination they would exhibit a greater ergogenic effect due to increased central stimulation of brain areas that regulate subconscious pacing (ie, CHO acting on motor reward center and CAF on the adenosine receptors in the brain).
Methods Participants Eleven endurance-trained male cyclists or triathletes (mean ± SD age 32.2 ± 7.5 y, mass 74.3 ± 6.8 kg, sum of 8 skinfolds 67.8 ± ˙ O2peak 60.2 ± 4.0 mL · kg–1 · min–1, maximum aerobic 13.3 mm, V power 331 ± 34 W) with at least 1 year of endurance-training history volunteered for this study. Participants were provided with written and verbal information concerning the experimental procedures and potential risks before providing written consent to participate. The study was approved by the Auckland University of Technology ethics committee. Participants were asked to maintain their regular training regimen throughout the duration of the study and to refrain from strenuous exercise on the day before each test. At the time of testing, participants were riding an average of 10 ± 2 h/wk (range 8–14 h/wk). Trials were separated by at least 3 days (range 3–7 d) and commenced at the same time of day to minimize the effects of diurnal variation. Participants were provided with instructions on how to achieve a minimum of 6 g CHO/kg body weight (BW) on the day before each trial and 1 g CHO/kg BW in their pretrial meal and abstain from CAF consumption on the day of their experimental trials. The diet was recorded by the participant for the familiarization trial, checked by the primary investigator, and then replicated for the remaining 4 experimental trials. All tests were performed on a Velotron cycle ergometer (RacerMate, Seattle, WA), which was set up according to the dimensions of each participant’s own bicycle and fitted with each cyclist’s own cycling pedals. All cycling tests were performed in a well-ventilated and temperature-controlled laboratory (18.6°C ± 0.8°C and 60% ± 7% relative humidity).
Incremental Cycling Test An incremental cycling test to exhaustion was used to determine V˙O2peak. The test started at a power output of 100 W and consisted
of a continuous step protocol with 5-minute stages. Power output was progressively increased by 50 W at the end of each 5-minute stage until volitional exhaustion. Participants cycled at their individualized and self-selected optimal cadence, so that an accurate reflection of maximum aerobic power could be determined. The test was terminated voluntarily by the participant or when cadence dropped below 60 rpm. Gas exchange was measured using a ParvoMedics TrueOne 2400 diagnostic system (Salt Lake City, UT). The workload (W and W/kg) corresponding to the second ventilation threshold (VT2) was determined using the methods of Davis et al.16 To familiarize participants with the 20-km TT distance after a fatiguing bout of exercise, participants performed a 20-km TT after completion of the incremental test, along with a full familiarization (described herein) on the second visit.
Experimental Trials Participants completed a total of 4 experimental trials, in a controlled, double-blind, repeated-measures, crossover experimental design. Participants served as their own control under 1 of 4 experimental conditions: chewing gum containing CHO (4 × ~1.8 g sucrose per piece + artificial sweeteners [AS]), CAF (4 × 50 mg CAF per piece + AS), CHO+CAF (4 × ~1.8 g sucrose + 50 mg CAF per piece + AS), or placebo (PLA) (4 × AS). A total of 200 mg (average of 2.7 mg/kg BW) CAF was selected to simulate the actual competition doses used by athletes for competition.7 Each trial began with a 90-minute constant-load cycle consisting of a standardized warm-up of 3 minutes at a power output equivalent to 40% of VT2, followed by 87 minutes at 80% of VT2 (207 ± 30 W, 63% ± 6% of maximal aerobic power, 67% ± 6% of V˙O2peak). The determined exercise intensity of 80% VT2 power output was selected to standardize relative work levels and to mimic realistic race conditions.17 Throughout the constant-load phase, participants self-selected pedaling rate as the Velotron ergometer controls power output irrespective of cadence. Time elapsed, cadence, and power output were visible throughout.
20-km TTs After the 90-minute constant-load cycle, participants were allowed 3.5 minutes of rest before commencing the 20-km TT. A verbal warning was given 1 minute before the start and then a 5-second countdown. Each TT began from a standing, stationary start with the starting gear ratio standardized for each participant from the 2 familiarization trials. Cyclists were instructed to complete all TTs in the fastest time possible and to adopt the same self-selected, individualized pacing strategy used previously. Instantaneous feedback for distance was provided, but participants were blinded to power output, cadence, speed, and heart rate. One piece (3 g) of chewing gum (EPSAR Co, Australia) was provided at the start of the 20-km TT and for every 25% of the TT completed (ie, 0, 5, 10, 15 km). The delivery procedure of the gum and application time interval (ie, approximately every 7.5–8 min) was chosen to replicate previous CHO mouth-rinse studies.2,3 Participants were instructed to chew their gum for 3 km (~5 min) before expelling the contents. On completion of the trial, they were asked to guess which gum they believed they had received and whether they experienced gastrointestinal discomfort during the trial. Time, power output, and cadence were automatically recorded at a frequency of 1 Hz by the Velotron software (Velotron Coaching Software, RacerMate, USA) and averaged for each 1 km. To minimize any potential effects of hydration and heat strain on
IJSPP Vol. 11, No. 2, 2016
166 Oberlin-Brown et al
exercise performance, a small fan (wind speed 2.8 ± 0.3 m/s) was positioned in front of the participant and water consumption was ad libitum during both the 90-minute constant-load phase and the 20-km TT.
Blood Analysis A capillary blood sample was drawn preexercise, immediately after the 90-minute constant-load cycle, and 3 minutes after completion of the 20-km TT. Samples were collected from the right ear lobe using standard techniques. During each blood draw, 2 separate samples were collected and analyzed for blood lactate (Lactate Pro analyzer, Arkray, Tokyo, Japan) and blood glucose concentration (CareSens II glucose analyzer, i-SENS Inc, Korea).
Downloaded by ETSU on 09/19/16, Volume 11, Article Number 2
Statistical Analysis Data are presented as mean ± SD and were log transformed before statistical analysis to reduce bias arising from nonuniformity of error. To make inferences about the true (population) effects of CHO and CAF on performance, the uncertainty of the effect was expressed as 90% confidence limits, with probabilities of the true value representing a substantial change in performance.18 An effect was deemed unclear if its confidence interval overlapped the thresholds for substantiveness, meaning that the effect could be substantially positive or negative. An estimate of the smallest substantial change in power output was required to make these inferences. Paton and Hopkins19 estimate the smallest effects of 0.5% to 1.5% in mean power output, based on variability of competitive cyclists’ performance in cycling TT events free of drafting. A smallest substantial change in endurance-TT performance was assumed to be a reduction or increase in performance power
output of 1% or more. For all other measures, 0.2 times the baseline between-subjects SD was used.18
Results Performance Response Performance times and power output during the 20-km TT were not significantly different between experimental trials (Table 1). Pairwise comparisons revealed positive insubstantial improvements for all experimental trials relative to PLA (Table 2). Individual results for each trial are presented in Figure 1. Three participants substantially improved mean power output with CHO versus PLA (6.9%, 2.4%, and 4.6%), 6 had substantial reductions (–1.2, –1.5, –2.5, –3.7, and –2.6), and 2 had trivial differences (0.0% and –0.4%). With CAF, 4 participants substantially improved (1.3, 3.3, 3.6, and 7.9) compared with PLA, 3 performed substantially worse (–3.4, –4.7, and –2.6), and the remaining 3 showed trivial differences (–0.8, 0.0, and –0.3). In the combined CHO+CAF condition, 4 participants showed substantial improvement (1.3, 5.1, 2.3, and 3.4), 3 substantially worsened (–3.5, –2.2, and –2.3), and the remaining 3 presented with trivial differences (0.3, –0.8, and –0.9). Analysis of the pacing response for each 5-km quarter is presented in Figure 2, and the associated pairwise comparisons are presented in Table 3. Mean power output with CHO was higher than PLA in the first 2 quarters of the TT (CHO 288 ± 37 and 266 ± 35 vs PLA 284 ± 40 and 264 ± 35 W for 0–5 km and 5–10 km respectively). In contrast, with CAF and CHO+CAF, mean power output was higher in the final 2 quarters (CAF 263 ± 39 and 284 ± 42 and CHO+CAF 260 ± 35 and 279 ± 43 vs PLA 259 ± 35 and 273 ± 41 W, for 10–15 km and 15–20 km, respectively).
Table 1 Performance Measures Obtained During the 20-km Time Trial for Each Condition (Mean ± SD) Measure Time (m:s) Mean power output (W)
Placebo
Carbohydrate
Caffeine
Carbohydrate + caffeine
32:27 ± 1:57
32:25 ± 1:45
32:20 ± 1:57
32:26 ± 1:51
270 ± 37
271 ± 35
273 ± 40
270 ± 37
Table 2 Pairwise Comparisons Quantifying the Magnitude of Effect of Different Chewing-Gum Contents on 20-km Time-Trial Mean Power Output Treatment effecta
% effect, mean ± 90%CLb
Effect size, mean ± 90%CLb
Qualitative inferencec
carbohydrate vs placebo
0.2 ± 2.0
0.01 ± 0.12
Possibly trivial
caffeine vs placebo
0.4 ± 2.2
0.03 ± 0.13
Unclear
carbohydrate + caffeine vs placebo
0.1 ± 1.8
0.01 ± 0.11
Possibly trivial
carbohydrate vs placebo
0.3 ± 1.6
0.02 ± 0.10
Possibly trivial
caffeine vs placebo
–0.4 ± 1.5
–0.02 ± 0.09
Unclear
carbohydrate + caffeine vs placebo
–0.1 ± 1.2
–0.01 ± 0.08
Unclear
Experimental treatment effects vs placebo
Within-treatment effects, independent and combined
Units of change are percentages for all measures derived from back-transformed log data. CL: Add and subtract this number to the difference to obtain the 90% confidence limits for the true difference. c Based on smallest beneficial or harmful change in performance of 1%.
a
b
IJSPP Vol. 11, No. 2, 2016
Downloaded by ETSU on 09/19/16, Volume 11, Article Number 2
Oral Carbohydrate and Caffeine: Effect on Endurance Performance 167
Figure 1 — Individual responses to experimental treatments relative to placebo for percentage change in mean 20-km time-trial power output. Black circles = carbohydrate; striped squares = caffeine; white squares = carbohydrate + caffeine; gray bar = smallest worthwhile change of 1%.
Physiological Responses There were no substantial differences in average and maximal heart rates (163 ± 13, 164 ± 11, 162 ± 12, 163 ± 13, and 176 ± 11, 176 ± 10, 178 ± 12, 176 ± 10 for PLA, CHO, CAF, and CHO+CAF, respectively) across conditions, indicative of consistent maximal effort across trials. Cohen effect sizes were 0.2).
in CAF-containing trials—CAF (0.4% ± 2.2%) and CHO+CAF (0.1% ± 1.8%)—compared with PLA for overall performance, although pacing was altered. Specifically, CAF and CHO+CAF exhibited a possibly harmful effect on mean power output for the first 2 quarters of the TT (CAF –1.3 ± 3.6 and –0.4% ± 2.2% and CHO+CAF –1.1% ± 2.4% and –1.2% ± 2.0%). Conversely, in the later stages of the TT, we observed an unclear effect in the third quarter (CAF 1.5% ± 2.2% and CHO+CAF 0.4% ± 2.5%) but a very likely beneficial effect in the last quarter (CAF 4.2% ± 3.0% and CHO+CAF 2.0% ± 1.8%). Given these findings, our results suggest that CAF is not acting via suggested sensory pathways but, rather, via absorptive pathways and eliciting central effects, likely within 5 minutes (Kamimori et al12). Our findings may relate to our CAF-administration methods—4 × 50 mg CAF every 5 km (total = 200 mg; average 2.7 mg/kg BW) during the TT rather than a single bolus before exercise (3.0 mg/kg BW) as used by Ryan et al13 and Paton et al15—resulting in insufficient delivery of CAF to adenosine receptors in the early stages of the TT and, hence, a delayed ergogenic effect. Finally, the fact that performance was substantially improved in the final quarter of the trial, when endogenous CHO stores would have been at their lowest and, conversely, fatigue levels highest, might suggest that providing CAF during endurance exercise may enable cyclists to override fatigue signals associated with exercise-induced fatigue.23,24 Another aspect considered vital in studies of highly trained performers is the individual performance responses. In the current study, results revealed considerable interindividual and intraindividual variation to the independent and combined oral presence of CHO and CAF, suggesting the presence of nonresponders or adverse CAF responders. However, it cannot be ruled out that the inconsistency in performance responses may be due to the small sample size and/or a sampling error variance. In this regard, we acknowledge some limitations with our study. The placebo effect positively influences performance outcomes, as participants are
Although the current work did not yield any definitive performance improvements during a 20-km cycling TT, the data suggest that under conditions of fatigue and reduced glycogen, the oral presence of CHO and/or CAF in chewing gum could temporarily override the anticipatory regulation strategy and subconscious pacing. CHO appears to have a relatively immediate effect—increased power in the first half of the trial—but timing its use toward the latter stages may be critical to prevent a subsequent decline in performance due to overpacing. In contrast, CAF appears to have a delayed ergogenic effect shown through the increased power in the second half of the TT; this suggests that earlier administration of CAF might be best. For example, if athletes had chewed earlier (before the TT) or used a larger single dose at the start, rather than spacing it evenly throughout the TT, the ergogenic effects might have been greater and/or occurred earlier. That both CHO and CAF appeared to acutely alter subconscious motor output during an exercise-fatigued state warrants further investigation to determine whether an ergogenic effect occurs when using an oral CHO gum, in combination with ingestion of CHO early in exercise (
