Article pubs.acs.org/jpr
Metabolomics-Based Analysis of Banana and Pear Ingestion on Exercise Performance and Recovery David C. Nieman,*,† Nicholas D. Gillitt,‡ Wei Sha,§ Mary Pat Meaney,† Casey John,† Kirk L. Pappan,# and Jason M. Kinchen# †
Human Performance Laboratory, Appalachian State University, North Carolina Research Campus, Kannapolis, North Carolina 28081, United States ‡ Dole Nutrition Research Laboratory, North Carolina Research Campus, Kannapolis, North Carolina 28081, United States § Bioinformatics Services Division, University of North Carolina at Charlotte, North Carolina Research Campus, Kannapolis, North Carolina 28081, United States # Metabolon, Inc., Durham, North Carolina 27713, United States S Supporting Information *
ABSTRACT: Bananas and pears vary in sugar and phenolic profiles, and metabolomics was utilized to measure their influence on exercise performance and recovery. Male athletes (N = 20) cycled for 75 km while consuming water (WATER), bananas (BAN), or pears (PEAR) (0.6 g carbohydrate/kg each hour) in randomized order. UPLC-MS/MS and the library of purified standards maintained by Metabolon (Durham, NC) were used to analyze metabolite shifts in pre- and postexercise (0-h, 1.5-h, 21-h) blood samples. Performance times were 5.0% and 3.3% faster during BAN and PEAR versus WATER (P = 0.018 and P = 0.091, respectively), with reductions in cortisol, IL-10, and total leukocytes, and increases in blood glucose, insulin, and FRAP. Partial Least Square Discriminant Analysis (PLS-DA) showed a distinct separation between trials immediately (R2Y = 0.877, Q2Y = 0.457) and 1.5-h postexercise (R2Y = 0.773, Q2Y = 0.441). A total of 107 metabolites (primarily lipid-related) increased more than 2-fold during WATER, with a 48% and 52% reduction in magnitude during BAN and PEAR recovery (P < 0.001). Increases in metabolites unique to BAN and PEAR included fructose and fruit constituents, and sulfated phenolics that were related to elevated FRAP. These data indicate that BAN and PEAR ingestion improves 75-km cycling performance, attenuates fatty acid utilization and oxidation, and contributes unique phenolics that augment antioxidant capacity. KEYWORDS: carbohydrate, cycling, polyphenols, metabolites, inflammation, oxidative stress
■
including 6.4 g starch, 3.1 g dietary fiber, and 14.4 g sugars (5.9 g glucose, 5.7 g fructose, 2.8 g sucrose).8 The glycemic index of bananas is 51 (low-to-medium rating), similar to grapes, mangos, pineapples, raisins, orange juice, and honey.9 Bananas are a good source of potassium (422 mg) and vitamin B6 (0.43 mg), and they contain secondary metabolites such as catecholamines, including dopamine at a concentration of 2.5−10 mg in pulp, and a variety of phenolics.8−11 The antioxidant value of bananas described in oxygen radical absorbance capacity (ORAC) units is 1,037 μmol Trolox equivalents (TE), which is similar to kiwi fruit and orange juice.12 Many of banana’s volatile ester and alcohol compounds play important roles in aromatic properties.13 In a previous study, we compared the acute effect of ingesting bananas with water versus a 6% carbohydrate drink (both providing 0.8 g/kg carbohydrate per hour) on 75-km cycling performance and postexercise inflammation and oxidative stress
INTRODUCTION Endurance athletes participating in 2−3 h of prolonged, intensive exercise experience improvements in performance when ingesting 20 to 80 g per hour of fructose, glucose, and sucrose.1−5 Ingestion of carbohydrate mixtures versus glucose alone during exercise improves intestinal absorption and oxidation rates.3,6 Carbohydrates such as glucose, maltose, and sucrose can be oxidized at rates up to 60 g per hour in comparison to fructose and galactose that are oxidized at much lower rates (up to 40 g per hour). Fructose absorption occurs on the mucosal membrane via facilitated transport by GLUT5 transport proteins, and glucose absorption occurs via the energy-dependent sodium-glucose linked SGLT1 transporter.6 When glucose, fructose, and sucrose are ingested simultaneously, overall intestinal absorption rates for carbohydrate, sodium, and water increase.3 Bananas and plantains are among the world’s leading fruit crops, and are an important source of carbohydrate energy for people around the world.7 One medium Cavendish banana (∼118 g) provides 105 kilocalories and 27 g carbohydrate © XXXX American Chemical Society
Received: September 26, 2015
A
DOI: 10.1021/acs.jproteome.5b00909 J. Proteome Res. XXXX, XXX, XXX−XXX
Article
Journal of Proteome Research using metabolomics-based profiling.14 Blood glucose levels and performance did not differ between the banana and 6% carbohydrate trials, exercise-induced increases in inflammation biomarkers, including cytokines and granulocyte phagocytosis, were similar and below levels previously measured during water-only studies in our lab, and aside from higher dopamine during the banana trial, metabolite shifts following 75-km cycling were not statistically different, indicating a similar pattern of fuel substrate utilization. In this study, we sought to extend these findings in several ways. Fruits vary widely in sugar and phenolic profiles. With athletes operating as their own controls, we compared exercise performance and recovery influences of water and two distinctly different fruits---bananas and pears. Pears contain high levels of fructose, with a glucose-to-fructose ratio of 0.44 compared to 1.0 for bananas.8 Thus, the sugar profile of pears after ingestion may not support endurance performance to the same degree as bananas.2,3 Pear phenolics include those not found in bananas, such as arbutin (hydroquinone-β-Dglucopyranoside), which has attracted attention for its antibacterial, anti-inflammatory, antioxidative, antitussive, and skin-whitening effects.15,16 Fruit polyphenols possess strong antioxidative, anti-inflammatory, and antiviral effects in vitro, but they are biologically transformed during absorption and metabolism into conjugated (i.e., sulfated, glucuronated, methylated) phenolics with diminished physiologic effects.17,18 The benefits of phenolic ingestion before and during exercise are currently an active area of research endeavor in our lab and others, with no clear consensus of performance and recovery benefits.17,19,20 In this investigation, we employed global, untargeted metabolomics to better define the influence of banana and pear intake with their widely disparate sugar and phenolic profiles on 75-km cycling performance and recovery in overnight fasted endurance athletes. We hypothesized that, compared to water, ingestion of banana or pears during 75-km cycling trials would improve performance (more so for bananas due to a more favorable sugar profile), attenuate fatty acid mobilization and oxidation, reduce inflammation, increase plasma phenolic levels, and improve antioxidant capacity.
■
completing all aspects of the study using a repeated measures analysis of variance (ANOVA), within participants approach. One to 2 weeks prior to the first 75-km time trial, athletes completed orientation and baseline testing. Demographic and training histories were acquired with questionnaires. Maximal power, oxygen consumption, ventilation, and heart rate were measured during a graded exercise test (25 W increase every 2 min, starting at 150 W) with the Cosmed Quark CPET metabolic cart (Rome, Italy) and the Lode cycle ergometer (Lode Excaliber Sport, Lode B.V., Groningen, Netherlands). Body composition was measured with the Bod Pod body composition analyzer (Life Measurement, Concord, CA). During the 3-day period prior to each 75-km cycling trial, participants were asked to reduce the volume of their exercise training as if preparing for a race, and they ingested a moderatecarbohydrate diet using a food list restricting high fat foods. Participants recorded all food and beverage intake during the 3day period, with macro- and micronutrient intake assessed using the Food Processor dietary analysis software system (ESHA Research, Salem, OR). For each of the three 75-km cycling trials, participants reported to the Human Performance Laboratory at 6:45 am in an overnight fasted state (no food or beverages other than water for at least 9 h), and they provided a pre-exercise blood sample. In accordance with the randomized schedule, participants then ingested 5 mL/kg water only, or 0.4 g/kg carbohydrate from ripe Cavendish bananas or bosc pears. After a 20 min rest, participants warmed up and then began the 75km cycling time trial using their own bicycles on CompuTrainer Pro Model 8001 trainers (RacerMate, Seattle, WA). The ComputTrainer MultiRider software system (version 3.0, RacerMate, Seattle, WA) was used to simulate a moderately difficult, mountainous 75-km course with continuous workload monitoring. Heart rate and rating of perceived exertion (RPE) were recorded every 30 min. Oxygen consumption and ventilation were measured using the Cosmed Quark CPET metabolic cart after 16 km and 55 km cycling (level sections of the course). All participants consumed 3 mL/kg water every 15 min; participants randomized to the banana and pear trials also ingested 0.15 g/kg carbohydrate every 15 min. No other beverages or food were allowed during the cycling time trials and 1.5-h recovery. Blood samples were taken via venipuncture immediately after and 1.5-h after completing the 75-km time trial. The next morning, participants returned in an overnight fasted state to provide a 21-h postexercise blood sample. All blood samples were centrifuged, aliquoted, and stored at −80 °C until analysis. Trials were separated by 2 weeks, after which participants crossed over to the next randomized condition, and repeated all procedures. Participants provided responses to a symptom questionnaire within 10 min of completing each of the 75-km cycling trials.15 The symptom questionnaire included questions on digestive health (feeling full, bloating, diarrhea, and nausea/vomiting), energy levels, focus/concentration, muscle cramping, and overall well-being. Subjects indicated responses using a 12point Likert scale, with 1 relating to “none at all”, 6 “moderate”, and 12 “very high”.
METHODS
Participants
Participants included 20 male cyclists (ages 25−51 years) who regularly competed in road races (category 1 to 5) and were capable of cycling 75-km at race pace. During the 6-week period when data were being collected, participants maintained their typical training regimen (216 ± 21.4 km cycling per week), and avoided the use of vitamin and mineral supplements, herbs, and medications. Participants signed informed consent, and study procedures were approved by the Institutional Review Board at Appalachian State University. Data were collected at the Human Performance Laboratory at the North Carolina Research Campus in Kannapolis, NC. Research Design
This study utilized a randomized (1:1 allocation, random number generator) crossover approach, and participants engaged in three 75-km cycling time trials while ingesting water only, bananas and water, and pears and water, separated by 2 weeks each. Participants completed the three arms of the study, and data were analyzed with participants operating as their own controls. Data were analyzed from participants
Complete Blood Count, Lactate, Glucose, Insulin, Cortisol
Complete blood counts (CBC) with a white blood cell (WBC) differential were performed using a Coulter Ac.T 5Diff Hematology Analyzer (Beckman Coulter, Inc., Miami, FL). Exercise-induced shifts in plasma volume were calculated using B
DOI: 10.1021/acs.jproteome.5b00909 J. Proteome Res. XXXX, XXX, XXX−XXX
Article
Journal of Proteome Research
at fixed concentrations to ensure injection and chromatographic consistency. One aliquot was analyzed using acidic, positive ion-optimized conditions and the other using basic, negative ion-optimized conditions in two independent injections using separate dedicated columns (Waters UPLC BEH C18−2.1 × 100 mm, 1.7 μm). Extracts reconstituted in acidic conditions were gradient eluted using water and methanol containing 0.1% formic acid, while the basic extracts, which also used water/ methanol, contained 6.5 mM ammonium bicarbonate. A third aliquot was analyzed via negative ionization following elution from a HILIC column (Waters UPLC BEH Amide 2.1 × 150 mm, 1.7 μm) using a gradient consisting of water and acetonitrile with 10 mM Ammonium Formate. The MS analysis alternated between MS and data-dependent MS2 scans using dynamic exclusion, and the scan range was from 80 to 1000 m/z. Metabolites were identified by automated comparison of the ion features in the experimental samples to a reference library of chemical standard entries that included retention time, molecular weight (m/z), preferred adducts, and in-source fragments as well as associated MS spectra and curated by visual inspection for quality control using software developed at Metabolon.25 Identification of known chemical entities was based on comparison to metabolomic library entries of purified standards.
the equation of Dill and Costill.21 Blood glucose and lactate were measured using microfuge tubes lined with ethylenediaminetetraacetic acid (EDTA) dipotassium salt (RAM Scientific Inc., Germany), and the YSI 2300 STAT Plus Glucose and Lactate analyzer (Yellow Springs, OH). Serum cortisol and insulin were measured with electrochemiluminescence immunoassays (ECLIA) through a commercial lab (LabCorp, Burlington, NC, USA). Plasma Cytokines
Total plasma concentrations of five inflammatory cytokines [monocyte chemoattractant protein-1 (MCP-1), tumor necrosis factor alpha (TNFα), IL-6, IL-8, and IL-10] were determined using an electrochemiluminescence based solidphase sandwich immunoassay (Meso Scale Discovery, Gaithersburg,MD, USA). All samples and provided standards were analyzed in duplicate, and the intra-assay coefficient of variation (CV) ranged from 1.7% to 7.5%, and the interassay CV ranged from 2.4 to 9.6% for all cytokines measured. Preand postexercise samples were analyzed on the same assay plate to decrease interkit assay variability. FRAP
Total plasma antioxidant power was determined by the ferric reducing ability of plasma (FRAP) assay, a single electron transfer reaction, as previously described by Benzie et al.22 This assay utilizes water-soluble antioxidants in the plasma to reduce ferric iron to the ferrous form identifiable at 593 nm. Samples and standards are expressed as ascorbate equivalents based on an ascorbate standard curve. Intra-assay and interassay coefficients of variation were less than 5% and 7%, respectively.
Statistical Analysis
Data are expressed as mean ± SE. Food record and performance data were compared between conditions using paired t tests. Biomarker data were analyzed using a 3 (condition) x 4 (time) repeated-measures ANOVA, withinparticipants design, with changes over time within conditions contrasted between conditions using paired t tests. For the metabolomics data, raw area counts for each metabolite in each sample were normalized to correct for variation resulting from instrument interday tuning differences. Median values for each run-day were set to 1.0. Missing values were imputed with the observed minimum after normalization. Following log transformation of the normalized data, analysis by two-way ANOVA with repeated measures with post-test contrasts was performed and statistical significance was set at p < 0.05. A false discovery rate estimate (“q-value”) was calculated to adjust for multiple comparisons.26 Fold changes across time points were calculated using group averages of the median scaled intensity values. Partial Least Square Discriminant Analysis (PLS-DA) in SIMCA (Version 14, Umetrics, Umeå, Sweden) was used to detect metabolites that best distinguished the three trials. The default 7-round cross-validation in SIMCA was used to compute the diagnostic Q2Y value which is a measure of model prediction ability. Permutation based validation was used to prevent overfitting. During permutation based validation, the prediction ability of the model was compared with the prediction ability of each model built using each of the 500 permutated data sets. A model was considered valid if its prediction ability was found to be better than 95% of the models built using the permutated data sets. Variable Influence on Projection (VIP) score was calculated for each metabolite based on PLS weight and variability supported in PLS-DA. Metabolites with VIP > 1.2 were considered important metabolites in explaining trial differences. Three PLS-DA analyses were performed. The first one compared banana and water trials; the second compared pear and water trials; and the third compared all three trials at the same time. For each
Global Metabolomic Platform Materials and Methods
Sample preparation was carried out at Metabolon Inc., as previously described.23,24 Briefly, recovery standards were added prior to the first step in the extraction process for quality control purposes. To remove protein, to dissociate small molecules bound to protein or trapped in the precipitated protein matrix, and to recover chemically diverse metabolites, proteins were precipitated with methanol under vigorous shaking for 2 min (Glen Mills Genogrinder 2000) followed by centrifugation. Three types of controls were analyzed in concert with the experimental samples: technical replicates from a pool of extensively characterized human plasma; extracted water process blanks; and a cocktail of standards spiked into every analyzed sample for instrument performance monitoring. Instrument variability was determined by calculating the median relative standard deviation (RSD) for the standards that were added to each sample prior to injection into the mass spectrometers. Overall process variability was determined by calculating the median RSD for all endogenous metabolites present in 100% of the plasma technical replicates. Experimental samples and controls were randomized across the platform run. The ultraperformance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) platform utilized a Waters Acquity UPLC with Waters UPLC BEH C18−2.1 × 100 mm, 1.7 μm columns and a Thermo Scientific Q-Exactive high resolution/accurate mass spectrometer interfaced with a heated electrospray ionization (HESI-II) source and Orbitrap mass analyzer operated at 35,000 mass resolution. The sample extract was dried then reconstituted in acidic or basic LC-compatible solvents, each of which contained 8 or more injection standards C
DOI: 10.1021/acs.jproteome.5b00909 J. Proteome Res. XXXX, XXX, XXX−XXX
Article
Journal of Proteome Research
compared to water trials, but also enhanced energy and ability to focus, and for the banana trial improved overall well-being. No trial differences were found for symptoms of diarrhea, nausea/vomiting, or muscle cramping (data not shown). Exercise induced significant increases in serum cortisol, blood total leukocytes, the ratio of neutrophils to lymphocytes, and plasma IL-6, IL-8, IL-10, TNFα, and MCP-1 (time effects, all P < 0.001). Serum cortisol was significantly lower 1.5-h postexercise in both the banana and pear trials compared to the water-only trial, with lower postexercise concentrations for total blood leukocytes and the ratio of neutrophils to lymphocytes (Table 3). Interaction effects were nonsignificant for plasma IL-6, IL-8, TNF-α, and MCP-1 (Table 3), but substantial postexercise reductions were measured for plasma IL-10 for both the banana and pear trials compared to the water-only trial (interaction effect, P < 0.001). The pattern of increase in FRAP immediately postexercise differed between trials (interaction effect, P = 0.018), with a significant increase in the banana (P < 0.01) and pear (P = 0.05) trials compared to the water trial (Figure 2) Metabolomics analysis revealed 590 biochemicals of known identity. Figure 3 depicts the PLS-DA score scatter plot for the three trials (yellow represents the banana trial, green for pear, blue for water) using ratios between the immediate postexercise and pre-exercise values for each participant (labeled on the plot by identification numbers). The PLS-DA model showed a distinct separation between trials (R2Y = 0.877, Q2Y = 0.457), and metabolites with VIP scores of greater than 1.2 are summarized in the Supporting Information Table S1. A similar PLS-DA score scatter plot using ratios between the 1.5-h postexercise and pre-exercise values also showed a distinct separation between trials (R2Y = 0.773, Q2Y = 0.441) (plot not shown, with the VIP scores listed in Supporting Information Table S1). Both PLS-DA models passed permutation based validation. No valid PLS-DA model could be generated to separate the three trials using ratios between the 21-h postexercise and pre-exercise values. Figure 4 depicts the PLS-DA loading scatter plot for the analysis in Figure 3. This plot displays relationships between metabolites (dots) and each trial (star) with the same color code as Figure 3 and additional colors added to represent metabolites whose post-to-pre-exercise ratios are higher in both the banana and the pear trials (red), both banana and water trials (brown), and both pear and water trials (pink). Metabolites with VIP score less than or equal to 1.2 are represented by black dots. The lists of metabolites using this color scheme can be found in Supporting Information Table S1. A total of 107 metabolites significantly increased more than 2-fold after the 75-km cycling time trial during the water-only trial, as shown in Supporting Information Table S2. Metabolites (N = 35) increasing more than 5-fold immediately postexercise during the water-only trial are listed in Table 4. All 35 of these metabolites were from the lipid superpathway, and all were still significantly elevated 1.5-h postexercise, with a return to near pre-exercise levels for all but 8 metabolites by 21-h postexercise. When banana and water trials were compared, none of the metabolites were found to be significantly different pre-exercise. Immediately postexercise, 13 metabolites were found to be different between the two trials by both repeated measures ANOVA (with q < 0.05) and PLS-DA (with VIP > 1.2), and they had a fold increase of 2 or higher (Table 5). Of these 13 metabolites, seven were from the amino acid super pathway,
subject, ratios between postexercise (immediate, 1.5-h, and 21h) and pre-exercise time points were calculated and used as input data in the PLS-DA analysis. For the exercise recovery analysis, metabolites that significantly increased by more than 2 fold postexercise compared to pre-exercise in the water trial were considered exercise-induced metabolites. The total fold increase (sum of fold increase across all exercise-induced metabolites) for each subject at each time point in each trial was calculated to reflect the total magnitude of increase, and then the average and standard error of the total magnitude was calculated across subjects. Only a fold increase greater than 2 was considered for the calculation of total fold increase in this analysis. Outliers with Z score >3.5 or ←3.5 were removed. The total magnitude of increase of exercise induced metabolites were compared between trials to reflect the difference in the perturbation of these metabolites in the three trials.
■
RESULTS The analysis included 20 male cyclists (ages 25 to 51 years) who successfully adhered to all aspects of the study design (see Table 1). Three-day food records collected before each of the Table 1. Participant Characteristics (N = 20) Variable
Mean ± SE
Age (years) Height (m) Weight (kg) Body fat (%) Wattsmax VO2max (mL·kg·−1 min−1) HRmax (beats/min) Training (km/wk)
39.2 ± 1.9 1.78 ± 0.15 77.8 ± 1.7 17.7 ± 1.1 338 ± 8.2 51.0 ± 1.4 177 ± 1.9 216 ± 21.4
three 75-km cycling time trials revealed no significant differences in energy, macronutrient, or micronutrient intake. Energy intake was 2720 ± 169 kcal/day (11.5 ± 0.7 MJ/day), 2715 ± 180 kcal/day (11.5 ± 0.8 MJ/day), and 2663 ± 181 kcal/day (11.2 ± 0.8 MJ/day), with carbohydrate representing 51.7 ± 2.5%, 50.8 ± 1.8%, and 56.0 ± 1.8%, protein 16.7 ± 0.9%, 16.3 ± 0.8%, and 15.3 ± 0.8%, and fat 32.3 ± 1.9%, 32.6 ± 1.3%, and 29.1 ± 1.3% of total energy during the 3-day periods prior to the water, banana, and pear trials, respectively. Performance data for the three 75-km mountainous cycling time trials are summarized in Table 2. Performance times for the 75-km trial were 5.0% and 3.3% faster during the banana and pear trials compared to the water-only trial (P = 0.018 and P = 0.091, respectively). Oxygen consumption and the respiratory exchange ratio (RER) were significantly elevated in both the banana and pear trials relative to the water trial, with mean power output elevated in the banana compared to water trial (P = 0.023). Heart rate, rating of perceived exertion (RPE), plasma volume shift, weight change, and blood lactate change did not differ between the three trials. Blood glucose was significantly elevated postexercise in both the banana and pear trials compared to the water trial (interaction effect, P = 0.004), with a significant increase in serum insulin in the banana compared to water trial (interaction effect, P = 0.014) (Table 2). Symptoms reported by the cyclists following the 75-km cycling trials are summarized in Figure 1. Participants reported more fullness and bloating during the banana and pear D
DOI: 10.1021/acs.jproteome.5b00909 J. Proteome Res. XXXX, XXX, XXX−XXX
Article
Journal of Proteome Research
Table 2. Metabolic and Performance Data during the 75-km Cycling Trials under Water, Banana, and Pear Conditions in Trained Cyclists (N = 20) (mean ± SE) Variable Time (min) VO2 (mL·kg·−1 min−1) VO2 (%VO2max) Watts % Wattsmax HR (beats/min) %HRmax Ventilation (L/min) RPE RER Plasma volume shift (%) Body Weight (kg) Pre-Exercise Post-Exercise Blood lactate (mmol/L) Pre-Exercise Post-Exercise Blood glucose (mmol/L) Pre-Exercise Post-Exercise Serum insulin (pmol/L) Pre-Exercise Post-Exercise
Water
Banana
Pear
162 ± 5.1 34.4 ± 1.2 68.0 ± 2.3 195 ± 9.3 57.5 ± 2.0 143 ± 3.6 80.7 ± 1.8 69.5 ± 3.9 14.0 ± 0.3 0.85 ± 0.01 −10.1 ± 1.4
154 ± 3.3* 36.1 ± 1.2* 71.2 ± 2.4* 208 ± 8.2* 61.5 ± 1.5* 148 ± 3.1 83.7 ± 1.7 74.9 ± 4.5 13.7 ± 0.3 0.88 ± 0.01* −10.7 ± 1.3
156 ± 4.1 36.5 ± 1.1* 72.3 ± 2.4* 204 ± 9.3 60.1 ± 2.0 148 ± 3.9 83.4 ± 1.9 77.9 ± 4.2* 13.7 ± 0.3 0.88 ± 0.01* −12.5 ± 1.2
Interaction 77.7 ± 1.7 77.0 ± 1.7 Interaction 0.64 ± 0.09 2.14 ± 0.21 Interaction 4.39 ± 0.09 4.95 ± 0.20* Interaction 32.5 ± 3.3 41.9 ± 6.5*
P = 0.669 77.8 ± 1.6 77.0 ± 1.6 P = 0.626 0.62 ± 0.06 2.21 ± 0.26 P = 0.004 4.30 ± 0.13 4.86 ± 0.26* P = 0.014 28.3 ± 2.4 37.4 ± 6.5
77.7 ± 1.7 77.0 ± 1.6 0.64 ± 0.07 1.95 ± 0.23 4.26 ± 0.11 3.97 ± 0.24 28.0 ± 3.2 20.0 ± 4.5
*
P < 0.025 compared to the water condition using paired t tests. Interaction effects were analyzed using a 3 (condition) × 2 (time) repeatedmeasures ANOVA, within-participants design, with changes over time within conditions contrasted between conditions using paired t tests.
banana, and pear trials are depicted in blue, yellow, and green, respectively. A, B, C, and D represent pre-exercise, postexercise, 1.5-h postexercise, and 21-h postexercise, respectively. The results of the area under curve (AUC) analysis for water, banana, and pear trials were 438,821, 228,803 (P < 0.001 or 48% lower relative to water), and 212,753 (P < 0.001 or 52% relative to water), respectively. The AUCs were not found to be significantly different between the banana and pear trials (P = 0.610).
■
DISCUSSION In agreement with our hypothesis, banana (0.6 g/kg carbohydrate per hour) compared to water ingestion before and during 75-km cycling was associated with a meaningful performance enhancement, participant reports of improved energy and ability to focus, and a modest increase in postexercise antioxidant capacity. Increases in several but not all measures of inflammation were attenuated postexercise in the banana trial, including total leukocytes (34%↓), the neutrophil/lymphocyte ratio (28%↓), plasma IL-10 (59%↓), and serum cortisol (204%↓, 1.5-h postexercise). Additionally, biomarkers associated with carbohydrate availability and oxidation were increased, including blood glucose and fructose, insulin, and the respiratory exchange ratio. Pear versus water ingestion was linked to many of the same advantages, with some biomarker alterations lacking statistical significance due to participant complaints of adverse gastrointestinal symptoms. The concentration of carbohydrate is lower in bosc pears (16.1 g/100 g) than banana (22.8 g/100 g), and pear feedings required a greater volume to achieve target carbohydrate intake levels.
Figure 1. Symptoms reported by the cyclists following the 75-km cycling trials (1 to 12 Likert scale) (mean ± SE). *P ≤ 0.05 compared to the water trial, paired t test analysis.
five from xenobiotics, and one from carbohydrate. All 13 metabolites were significantly elevated 1.5-h but not 21-h postexercise. Figure 5a shows a significant, positive relationship between the pre-to-postexercise change in sulfated phenolics (dopamine, ferulic acid, vanillic acid) and FRAP (r = 0.57, P = 0.008). In a similar fashion, metabolites related to the pear versus water trial are listed in Table 6. Of these 17 metabolites, eight were from the xenobiotics superpathway, six from carbohydrate, two from amino acid, and one from lipid. Fifteen of the 17 metabolites were significantly elevated 1.5-h but none at 21-h postexercise. Figure 5b shows a significant, positive relationship between the pre-to-postexercise change in sulfated phenolics (ferulic acid, vanillic acid) and FRAP (r = 0.53, P = 0.019). The total magnitude of increase of exercise-induced metabolites was compared across trials (Figure 6). The water, E
DOI: 10.1021/acs.jproteome.5b00909 J. Proteome Res. XXXX, XXX, XXX−XXX
Article
Journal of Proteome Research
Table 3. Comparison between Water, Banana, and Pear Trials for Cortisol and Inflammation Biomarkers in Trained Cyclists (N = 20) before and after Cycling 75-km (mean ± SE)a Variable Cortisol (nmol/L) Water Banana Pear Total Leukocytes (109/L) Water Banana Pear Neutrophil/Lymphocyte Water Banana Pear IL-6 (pg/mL) Water Banana Pear IL-8 (pg/mL) Water Banana Pear IL-10 (pg/mL) Water Banana Pear TNFα (pg/mL) Water Banana Pear MCP-1 (pg/mL) Water Banana Pear
Pre-Exercise
Immediate Post-Exercise
1.5-h Post-Exercise
P-values: Time; Interaction
433 ± 22.1 406 ± 13.8 442 ± 19.3
610 ± 49.7 480 ± 35.9 535 ± 38.6
486 ± 46.9 351 ± 27.6* 381 ± 27.6*
