Research design
Participants reported to the Human Performance Laboratory (HPL) for one familiarization session prior to experimental trials. During the familiarization session, participants were informed of all procedures and familiarized with all performance measures to reduce the possibility of a learning effect. Participants performed two experimental time trials with a minimum of 48 h between each trial. The average time between trials was 5.3 ± 2.7 days. During each trial, participants consumed either regular Turkish coffee (TC), or a decaffeinated Turkish coffee (DC) from the same manufacturer (Strauss Coffee, Lod, Israel).
Participants reported to the HPL 2-h post-prandial and had not exercised for at least 24 h prior to each trial. In addition, participants were asked to not alter their dietary habits, but to replicate their dietary habits for the day before and day of each trial, and were requested to avoid caffeine consumption the day of the trial. Upon arrival to the laboratory participants were asked to lie comfortably in a supine position prior to the insertion of a cannula. Participants were then requested to sit comfortably for 15 min prior to the baseline (BL) blood draw. Following the BL blood draw participants were provided either the TC or DC. Additional blood draws occurred 30-min following coffee ingestion (30+), prior to the 5 km run (PRE), and immediately following the 5 km run (IP). Cardiovascular and subjective measures were also assessed at similar time points. Performance measures included reaction assessment, 5 km time trial and cognitive function. During the time trial oxygen consumption (VO2), minute ventilation (V
E), respiratory exchange ratio (RER), and heart rate (HR) were assessed. The study protocol is depicted in Fig. 1.
Participants
For inclusion in the study, participants were required to have a VO2peak greater than 35 ml∙kg−1∙min−1, and had to be regular caffeine and/or coffee consumers to increase homogeneity of the sample. Regular consumption was defined as ≥200 mg per day or ≥1 cup of coffee per day; the reported intake was 277 ± 183 mg · day−1. After initial testing, 23 recreationally-active individuals (men = 10, women = 13) between the ages of 18 and 35 were recruited to participate in this study. Two female participants were removed due to health reasons not associated with the study, and one for non-compliance. Therefore, data for 10 men (25.5 ± 1.8 y; 1.77 ± 0.05 m; 84.1 ± 7.1 kg; 48.0 ± 4.3 ml · kg · min−1) and 10 women (22.7 ± 2.0 y; 1.64 ± 0.07 m; 61.76 ± 5.28 kg; 42.0 ± 3.1 ml · kg · min−1) were included in the final analysis.
All participants completed a questionnaire to assess their ability to participate in physical activity and to ascertain any prior supplementation. Individuals who reported using substances that could potentially mask the effects of caffeine were excluded from the study. If participants were supplements that did not interfere with the effects of caffeine (e.g. protein), they were asked to continue their supplementation, and diet regiment so that the only difference between trials was the ingestion of TC or DC. Individuals self-reported to be free of musculoskeletal injury as determined by a PAR-Q. Participants were excluded if they had any history of cardiovascular disease, metabolic, renal, hepatic, or musculoskeletal disorders or were taking any other medication as determined by the medical history questionnaire. Following an explanation of all procedures, risks and benefits, each participant provided his/her informed consent to participate in the study. The Institutional Review Board of the University of Central Florida approved the research protocol.
Treatment beverages
During each experimental trial, participants ingested either TC or DC 60-min prior to the 5 km time trial. Participants were provided 0.136 g coffee · kg−1 BW mixed in 200 ml of boiling water, resulting in 3 mg · kg−1 BW of caffeine per serving (e.g., for a participant with an average body weight of 70 kg, this was equivalent to 1.5 teaspoons of coffee powder). The Turkish coffee provided is marketed as Elite Turkish Coffee (Strauss Coffee, Lod, Israel) and the content provided was recommended by the manufacturer. There was no difference in appearance or taste between the TC and DC.
Blood sampling
During each experimental trial, all blood samples were obtained using a 20-gauge Teflon cannula placed in a superficial forearm vein using a three-way stopcock with a male luer lock adapter. The cannula was maintained patent using an isotonic saline solution (Becton Dickinson, Franklin Lakes, NJ). Blood samples were collected at BL, 30+, PRE, and IP into a Vacutainer® tube, containing sodium heparin, and was kept chilled prior to each blood draw. Following collection samples were subsequently centrifuged at 4,000 × g for 15 min. The resulting plasma was placed into separate 1.8-mL microcentrifuge tubes and frozen at −80 °C for later analysis.
Biochemical analysis
Of the 20 participants to complete the study 4 requested to not have blood drawn, thus a sample size of n = 16 was used for blood measures. Plasma glucose and lactate concentrations were determined using an automated analyzer (Analox GM7 enzymatic metabolite analyzer, Analox instruments USA, Lunenburg, MA). To eliminate inter-assay variance, all samples were analyzed in duplicate by a single technician. Coefficient of variation for each assay was 2.60 % for glucose and 0.79 % for lactate.
Plasma caffeine concentrations were quantified using high performance liquid chromatography (HPLC) with chromatographic conditions based upon previous work done in our lab [19]. Chromatography was performed on an Agilent Infinity 1260 HPLC (Agilent Technologies, Santa Clara, CA) consisting of a degasser, binary pump, auto-sampler, column thermostat, and photodiode array detector. A Zorbax Eclipse Plus C18 (4.6 × 150 mm, 5-μm) column and Zorbax analytical guard column (4.6 × 12.5 mm, 5-μm) were used for separation. Data were collected using OpenLAB chromatography data system, ChemStation edition.
All reagents were of HPLC grade. Caffeine, beta-hydroxyethyl-theophylline, sodium phosphate monobasic and sodium phosphate dibasic were purchased from Sigma-Aldrich (St. Louis, MO) to create the stock solution. Acetonitrile was purchased from Fisher Scientific (Pittsburgh, PA). HPLC grade water was prepared by reverse-osmosis and purified using a Milli-Q Direct 8 water purification system (EMD Millipore, Billerica, MA).
A 40 μg∙mL−1 stock solution of caffeine, theobromine and beta-hydroxyethyl-theophylline was prepared in water and sonicated. Twelve calibration standards were prepared from the stock solution in the range of 0.039 – 40 μg∙mL−1 by serial dilution of 1 mL of the stock solution. Beta-hydroxyethyl-theophylline (internal standard) working solution was prepared in water (10 μg∙mL−1).
An internal plasma sample was collected to serve as control and analyzed every 50 samples. Calibration standards, samples, and controls were prepared in the same fashion for linearity. 60 μL of the calibration standards or 50 μL of sample or quality control sample was added to 1.5 mL microcentrifuge tubes. 10 μL of the internal standard was subsequently added to the samples and controls, followed by 140 μL of chilled acetonitrile for deproteinization. Standards, samples, and controls were then vortexed vigorously for 30 s and placed in a refrigerator (4 °C) for two hours followed by centrifugation at 14,000 g for 15 min in a microcentrifuge to allow the protein to form a pellet. The supernatant (150 μL) was collected and subsequently transferred to a 0.45 μm polytetrafluoroethylene syringeless filter vial (GE Healthcare Mini-Uniprep™, Piscataway, NJ). A concentration of 300 μL of sodium phosphate buffer (mobile phase) was then added to the vial. The solution was filtered and injected into the HPLC using an auto-sampler.
The mobile phase consisted of 25 mM sodium phosphate (pH 7.0 ± 0.05 at 40 °C) and acetonitrile at a volume to volume ratio of 90:10. Buffer pH was achieved by mixing 5.41 g sodium phosphate monobasic anhydrous, and 7.80 g sodium phosphate dibasic anhydrous in 4 L of water at 40 °C. Buffer composition was calculated using Buffer Maker computer software (Marki, Poland) and verified using an Oakton pH 11 portable meter (Oakton Instruments, Vernon Hills, IL). Analysis was carried out under isocratic conditions via binary mixing of aqueous and organic phases at a flow rate of 1.5 mL∙min−1 under a system pressure of approximately 90 bars. Chromatograms were recorded at 275 nm with a run time of 6 min. Duplication of retention times for a known standard was used to verify column equilibrium prior to analysis.
Cardiovascular measures
Heart rates and blood pressure (BP) were measured at each assessment time point using a wireless heart rate monitor (Polar® RS800CX, Kempele, Finland) and a mobile mercury sphygmomanometer (American Diagnostic Corporation Diagnostix™ 972, Hauppauge, New York).
Subjective measures
Participants were instructed to assess their subjective feelings of energy, alertness, and focus using a 15-cm visual analog scale (VAS). The scale was anchored by the words “Lowest” and “Highest” to represent extreme ratings where the greater measured value represents the greater feeling. Questions were structured as “My level of energy is”, “My level of alertness is”, and “My level of focus is”. The validity and reliability of VAS in assessing subjective feelings have been previously established [20].
Performance measures
Upper body reaction measurements
Measurement of upper body reaction time was performed on the Dynavision D2 Visuomotor Training Device (D2; Dynavision International LLC, West Chester, OH). The D2 is a light-training reaction device developed to train sensory motor integration through the visual system [21]. It consists of a board (4 ft x 4 ft) that can be raised or lowered relative to the height of the participant. It contains 64 target buttons (lights) arranged into five concentric circles surrounding a center screen that can be illuminated to serve as a stimulus for the participant. Participants were required to assume a comfortable athletic stance and stand at a distance from the board where they can easily reach all of the lights. The board height was kept consistent for all testing trials and was adjusted per participant so the center screen was located just below eye level. A total of three different reaction tests were conducted.
The first assessment measured the participant’s ability to react to a stimulus (light) as it changed position on the board. An initial stimulus (light) was present on the D2 in a random location. The stimulus (light) remained lit until it was touched by the participant. A stimulus (light) then appeared at another random location. The participant was instructed to successfully identify and touch as many stimuli (lights) as possible within 60 s. The number of successful “hits” was recorded for each trial. The ICC of this assessment has been reported to be 0.75 in our laboratory [21].
The second assessment was similar to the previous measure in that participants were also required to react to a visual stimulus (light) as it changed position on the board. However, during this trial the stimulus (light) remained lit for 1 s before it changed to another random location and the participant had to verbally recite a five digit number that was presented on the center screen of the D2 every 5 s. The appearance of the digits placed a cognitive demand on the information processing resources of the participant. The participant was instructed to successfully identify and touch each stimulus before it changed position and score as many touches as possible within 60 s. The number of successful “hits” was recorded for each trial. The ICC of this test has been reported to be 0.73 in our laboratory [21].
The third assessment measured the participant’s visual, motor, and physical reaction times to a visual stimulus with the dominant hand. The test was initiated when the participant placed and held his hand on an illuminated “home” button. At this point, a stimulus (light) was presented randomly in one of five locations, parallel to the home button. Visual reaction time was measured as the amount of time it takes to identify the stimulus (light) and initiate a reaction by taking their hand off the home button. Motor response time was measured as the amount of time it takes to physically touch the stimulus (light) with their hand following the initial visual reaction and was measured as the amount of time between the hand leaving the home button and touching the stimulus (light). Physical reaction time was measured as the total elapsed time from the introduction of the target stimulus (light) to the physical completion of the task (returning to the home button after touching the stimulus). All measures were recorded to the 1/100’s of a second. Participants performed this assessment ten times. The average time for all ten assessments was recorded. In our laboratory, the ICC of this test has been shown to be 0.84 (visual) and 0.63 (motor) [21].
Lower body reaction measurements
Lower body reaction time was measured using a 20-s reaction test on the Quick Board™ (The Quick Board, LLC, Memphis, TN) reaction timer. Participants stood on a board of five circles in a 2 x 1 x 2 pattern. Participants straddled the middle circle and reacted to a visual stimulus located on a display box that depicts one of five potential lights that corresponded with the circles on the board. Upon illumination of a light, the participant attempted to move the dominant foot to the circle that corresponds to the visual stimulus. Upon a successful “hit” with the foot, the next stimulus appeared. The total number of successful attempts during the 20-s test and the average time between the activation of the light and the response to the corresponding circle was recorded. The ICC of the Quick Board™ has consistently shown r > 0.90 in our laboratory [19, 22].
Metabolic measures
Peak oxygen consumption (VO2peak) testing
During the familiarization trial an incremental test to volitional exhaustion was performed on a motorized treadmill (Woodway 4Front™, Waukesha, WI) to measure VO
2peak. Open-circuit spirometry (TrueOne 2400® Metabolic Measurement System, Parvo Medics, Inc., Sandy, UT) was calibrated with room air and gases of known concentration, which was used to estimate VO2 (ml∙kg−1∙min−1) by sampling and analyzing breath-by-breath expired gases. VO2peak was determined to be the highest 30-s VO2 value during the test and coincided with at least two of the following three criteria: (a) 90 % of age-predicted maximum heart rate; (b) respiratory exchange ratio > 1.1; and/or (c) a plateau of oxygen uptake (less than 150 mL · min−1 increase in VO2 during the last 60 s of the test). Previous work in our lab has shown the test-retest reliability for VO2peak to be ICC = 0.96 (SEM 1.4 ml.kg.min−1) [23].
5 km time trial measures
Prior to each 5 km time trial, open-circuit spirometry was calibrated with room air and gases of known concentration, which was used to estimate VO2 (ml∙kg−1∙min−1) by sampling and analyzing breath-by-breath expired gases. Expired gases—oxygen (O2), carbon dioxide (CO2), V
E, RER, and heart rate were monitored continuously and expressed as 30-s averages. Ratings of perceived exertion (RPE) were recorded every 5-min during the time trial. The time needed to complete the 5 km was recorded to the nearest tenth of a second.
Multiple object tracking and cognitive assessments
Subsequent to the removal of the intravenous catheter, after the IP blood draw, the participants were assessed for multiple object tracking using a Cave Automatic Virtual Environment (CAVE) system. The CAVE consists of a 7 ft × 7 ft × 7 ft room that includes a canvas projection screen on the front wall which served as the surface for image projection. During each session, the participant wore three dimensional glasses. A three-dimensional image of 8 tennis balls was projected onto the front screen. The participant tracked 4 of the 8 balls that moved in three-dimensions. At the beginning of each trial, the 8 balls appeared frozen on the screen for 2 s while half of them turned grey indicating the balls the participant was to track. After the 2 s, the balls all became the same color again and began to move in three dimensions. At the conclusion of the trial (8 s), the balls froze and a number appeared on each ball. The participant called out the numbers of the four balls they were supposed to be tracking. Velocity of movement began at a slow tracking speed and increased or decreased depending on whether the participant correctly identified the 4 correct balls. Each participant performed 20 trials per session. The velocity of movement that was most successful was recorded. Our laboratory has previously reported the ICC of this test to be 0.77 [24].
Statistical analysis
Prior to statistical procedures, all data was assessed for normal distribution, homogeneity of variance, and sphericity. If assumption of sphericity were violated, a Greenhouse-Geisser correction was applied. Time trial and reaction time performance measures were analyzed with paired student’s t tests. Comparisons between trials were further analyzed using Cohen’s d. Cohen’s d values of 0.20, 0.50, and 0.80 were interpreted as small, medium, and large effect sizes, respectively [25]. All other measures were analyzed using a 2 x 4 [Trial (TC, DC) x Time (BL, +30, PRE, IP)] repeated measures analysis of variance (ANOVA). In the event of a significant interaction, LSD post-hoc tests were used for pairwise comparisons. Due to a significant difference at BL for the VAS energy response, a repeated measures analysis of covariance (ANCOVA) with BL measures serving as the covariate was used for this measure. For effect size, the partial eta squared statistic was calculated, and 0.01, 0.06, and 0.14 were interpreted as small, medium, and large effect sizes, respectively [26]. Additional analysis examined the effect of the trials in those participants that were determined to be responders. A responder was defined as a participant whose time for the 5 km time trial was faster during the TC trial compared to the DC trial. Significance was accepted at an alpha level of p ≤ 0.05. All data are reported as mean ± SD. Statistical analysis was performed using SPSS (version 21.0, SPSS Inc., Chicago, IL).