Elevations in oxygen consumption during exercise increase metabolic processes such as the production of reactive oxygen species (ROS), which can lead to oxidative damage of cellular lipids, proteins and nucleic acids(Reference Evans1). Antioxidants defend the body against such oxidative stress and contribute to the maintenance of a healthy antioxidant:oxidant balance. There are two forms of antioxidants: enzymes such as catalase, glutathione peroxidase and superoxide dismutase; dietary antioxidants such as vitamins C and E, carotenoids, flavonoids, Zn and Se(Reference Chew and Park2). Carotenoids are found in plants, function as accessory pigments in photosynthesis and include lycopene, β-carotene and lutein. The major sites for storage of carotenoids in the body are adipose tissue and the liver, although they are found in other tissues including the kidney, lungs and prostate(Reference Rock3). These carotenoids can be mobilised during states of increased oxidative stress such as exercise(Reference Aguiló, Tauler and Fuentespina4).
Exercise-induced ROS are released from mitochondria in exercising muscles or from inflammatory cells such as phagocytes(Reference Jacob and Burri5). ROS generation can increase the secretion of pro-inflammatory mediators via the effects of NF-κB on the cell nucleus(Reference Chew and Park2). Consequently, exercise may increase the release of pro-inflammatory mediators(Reference Moldoveanu, Shephard and Shek6–Reference Kimura, Suzui and Nagao11).
PUFA, which are important components of cell membranes, are vulnerable to oxidation and can attenuate a pro-inflammatory response(Reference Hill, Worthley and Murphy12–Reference Li, Ruan and Powis14). Carotenoids can protect PUFA from ROS damage(Reference Wang, Shinto and Connor15). A change in dietary intake of carotenoids potentially alters ROS oxidative damage to PUFA and the release of inflammatory mediators.
The aim of the present study was to examine the effect of the dietary restriction of antioxidants on the plasma concentrations of carotenoids and inflammatory mediators at rest and in response to exercise in healthy endurance-trained males.
Experimental methods
Subjects and study design
Seventeen healthy endurance-trained male adults aged 18–35 years were recruited to participate in the study. All participants were non-smokers and did not take any vitamin or mineral supplements or medications that would affect oxidative stress or inflammatory mediators. The institution's Human Research Ethics Committee approved the study and all participants provided informed written consent before participation.
Participants attended the laboratory on three separate occasions. At the first visit, participants completed a treadmill VO2max test, had anthropometric measurements taken and were given dietary instructions. Percentage body fat was calculated from skinfold thickness measurements (triceps, subscapula, biceps, iliac crest, super spinatus, abdominal, front thigh and midcalf) using the Womersley & Dumin equation(Reference Womersley and Dumin16). Four-day weighed food records were completed before the second and third visits. For visits 2 and 3, participants arrived at the laboratory after an overnight fast (consumption of water was allowed) and provided blood samples (20 ml) before and after a treadmill exercise trial. Participants were asked to refrain from physical activity for 24 h before the exercise tests. Visits 2 and 3 were separated by 2 weeks of dietary antioxidant restriction.
VO2max test
The VO2max exercise test was conducted on a motorised treadmill (Powerjog Treadmill Model JM100, Expert Fitness, Mid Glamorgan, Wales). Ventilation (breaths/min), FEO2 and FECOs were measured, and VO2 and VCO2 were calculated from ventilation, FEO2 and FECOs by a computerised online gas analysis system (SensorMedics 2900c; SensorMedics Corporation, Yorba Linda, CA, USA) calibrated to known gases. Heart rate was monitored throughout the test by electrocardiogram (SensorMedics 2900c; SensorMedics Corporation). The initial running speed of each participant was 10 km/h with a gradient of 0 %. Speed increased 2 km/h every 2 min until the participant's maximum voluntary speed was achieved. If VO2max was not achieved with maximum voluntary speed, the gradient was then increased 2 % every 2 min until voluntary exhaustion was achieved. VO2max, heart rate, RER, exercise time, speed and gradient were recorded. VO2max was the highest 30s VO2(Reference Ferrier, Nestel and Taylor17) recorded together with the criteria of a plateau in oxygen uptake, RER >1·15 and a heart rate of 220 − age ± 10 beats/min.
Exercise protocols
Participants performed the following treadmill exercise trials during the second and third visits to the laboratory. The participant's running speed commenced at the workload that elicited 60 % of VO2max established during the first visit and was maintained for 30 min after which the speed was increased by 2 km/h every 2 min until the participant's maximum voluntary speed was achieved, then the gradient was increased 2 % every 2 min until voluntary exhaustion. Heart rate was monitored throughout the test by electrocardiogram. The exercise tests were carried out in a thermally controlled environment (22 ± 2°C, 40–60 % relative humidity). A similar exercise protocol that also involved high-intensity exercise (the incremental test to exhaustion was substituted for a 5 min run at 90 % VO2max) has been shown to affect lipid peroxidation in a similar cohort(Reference Kanter, Nolte and Holloszy18).
Exercise performance measurements
Ratings of perceived effort were determined using a fifteen-point Borg scale(Reference Borg19) and were obtained during the last minute of the 30-min submaximal portion of the exercise trial. Time to exhaustion during the incremental test following the 30-min submaximal exercise was used to measure exercise performance.
Dietary intervention
Between visits 1 and 2, the participants followed their habitual antioxidant diet (H-AO). Between visits 2 and 3, the participants followed a restricted antioxidant diet (R-AO) compared to their H-AO diet for 2 weeks. Participants restricted their intake of fruits and vegetables to 1–2 servings/d as this protocol has been shown to reduce plasma antioxidant concentrations in a 2-week period(Reference Record, Dreosti and McInerney20). Participants were also asked to avoid the consumption of other foods that are high in antioxidants including tea, cod liver oil and wheat germ oil.
Dietary analysis
Participants completed a 4-d weighed food record for two work days and two weekend days. Participants were given a set of electronic scales and instructions on how to complete the food record during visit 1, and collected food records before visit 2 while following the H-AO diet and before visit 3 while following the R-AO diet. Food records were analysed using FoodWorks program (version 2.1, Build 146, Xyris Software, Highgate Hill, QLD, Australia).
Daily energy estimation
The Cunningham equation was used(Reference Thompson and Manore21) to estimate the daily energy requirements of participants. RMR multiplied by an activity factor of 1·8 is the best predictor of mean daily energy requirements for an athletic population(Reference Thompson and Manore21).
Blood sample collection
Blood samples were collected pre-exercise and immediately post-exercise. A needleless cannula was inserted into the superficial vein of the forearm before exercise and was removed after the final blood draw. Blood samples were collected into sterile EDTA tubes and stored on ice. A maximum of 20 ml whole blood was collected per blood draw. Within 2 h of collection, the whole blood was centrifuged (3000 g, 10 min, 4°C) to separate plasma, which was then stored at − 80°C for later analysis.
Analysis of carotenoid composition of plasma
Carotenoids (α- and β-carotenes, lutein/zeaxanthin, lycopene and β-cryptoxanthin) were extracted from plasma and analysed via HPLC(Reference Barua, Kostic and Olson22). Separation was achieved using a 5 μm ODS Hypersil column (100 mm × 2·1 mm inner diameter).
Analysis of inflammatory mediator composition of plasma
All assays were performed on plasma samples according to the kit manufacturer's instructions in a single batch. IL-6 concentrations were analysed via enzyme-amplified sensitivity immunoassay with a detection limit of 2 pg/ml and 100 % specificity (enzyme-amplified sensitivity immunoassay; BioSource, Nivelles, Belgium), TNF-α concentrations were analysed via ELISA with a detection limit of 3 pg/ml and 100 % specificity (BioSource), leukotriene B4 (LTB4) concentrations were analysed via enzyme immunoassay with a detection limit of 13 pg/ml and 100 % specificity (enzyme immunoassay; Cayman Chemical Company, Ann Arbor, MI, USA).
Statistical analysis
Results are presented as means and standard deviations. Normality of plasma metabolite data was checked by skewness and kurtosis. All plasma data that were not normally distributed were log-transformed before analysis. Differences between pre- and post-exercise values were compared using 2 × 2 repeated-measures ANOVA. In all analyses, the limit for statistical significance was set at P < 0·05.
The present study was conducted according to the guidelines laid down in the Declaration of Helsinki, and all procedures involving human subjects were approved by the University of Newcastle Human Research Ethics Committee. Written informed consent was obtained from all subjects.
Results
Participant characteristics, dietary intake and exercise performance
Seventeen participants completed the study with one participant withdrawing before visit 2 due to injury. The characteristics of the participants are presented in Table 1 and their dietary intakes are displayed in Table 2. There were NS differences in dietary carbohydrate, protein or fat intake on either diet, whereas dietary intakes of fibre, ascorbic acid and β-carotene were significantly lower on the R-AO diet compared to the H-AO diet. There was NS difference in total energy intake or estimated total energy expenditure between diets. Time to exhaustion on the incremental phase of the exercise did not differ between trials (9·0 (sd 0·3) v. 9·2 (sd 0·3) min), but the rating of perceived effort was significantly higher on the R-AO diet compared to the H-AO diet (11·8 (sd 0·3) v. 12·4 (sd 0·3); P < 0·05).
(Mean values and standard deviations)
DEE, daily energy expenditure.
(Mean values and standard deviations)
* Differences in means between diets (ANOVA).
Plasma carotenoids
The changes in carotenoid concentrations are shown in Fig. 1. Pre-exercise plasma β-carotene and lutein/zeaxanthin concentrations were lower (P < 0·05) on the R-AO diet compared to the H-AO diet; there were NS differences in the pre-exercise concentrations of the other carotenoids. On the H-AO diet, all plasma carotenoid concentrations except lutein/zeaxanthin decreased (P < 0·05) in response to exercise. On the R-AO diet, lycopene and β-cryptoxanthin concentrations decreased (P < 0·05), whereas α-carotene concentrations increased (P < 0·05) in response to exercise. Post-exercise lycopene concentrations were lower (P < 0·05) on the R-AO diet, but no other significant differences in carotenoid concentrations were observed between diets.
The effect of exercise on plasma carotenoid concentrations in endurance-trained male athletes consuming a habitual antioxidant (H-AO) or restricted antioxidant (R-AO) diet (n 17). Mean values were significantly different at * P < 0·05 (H-AO v. R-AO), ** P < 0·05 (pre-exercise v. post-exercise), *** P < 0·01 (pre-exercise v. post-exercise), **** P < 0·001 (pre-exercise v. post-exercise, H-AO v. R-AO). ■, Pre-exercise; □, Post-exercise.
Plasma inflammatory mediators
The changes in inflammatory mediators are shown in Fig. 2. Pre-exercise plasma IL-6 concentrations were similar on both diets. IL-6 concentrations on both the H-AO and the R-AO diets were not affected by exercise. Although not statistically significant, IL-6 concentrations tended to increase (P = 0·1) with exercise on the H-AO diet but did not change on the R-AO diet (P = 0·6). Post-exercise IL-6 concentrations were significantly higher (P < 0·05) on the H-AO diet compared to the R-AO diet. Pre-exercise TNF-α concentrations were 38-fold higher (P < 0·01) and post-exercise TNF-α concentrations were 14-fold higher (P < 0·001) on the R-AO diet compared to the H-AO diet; there was no statistically significant response to exercise on either diet. Pre-exercise plasma LTB4 concentrations were similar on both diets. Plasma LTB4 concentrations decreased (P < 0·001) with exercise on the H-AO diet but did not change significantly on the R-AO diet; post-exercise plasma LTB4 concentrations were lower (P < 0·05) on the H-AO diet compared to the R-AO diet.
The effect of exercise on plasma IL-6, TNF-α and leukotriene B4 (LTB4) concentration in endurance-trained male athletes consuming a habitual antioxidant (H-AO) or restricted antioxidant (R-AO) diet (n 17). Mean values were significantly different at * P < 0·05 (H-AO v. R-AO), ** P < 0·01 (H-AO v. R-AO), *** P < 0·001 (H-AO v. R-AO), **** P < 0·001 (pre-exercise v. post-exercise). ■, H-AO; □, R-AO.
Discussion
The present study has found that restricting the dietary intake of fruit and vegetables to 1–2 serves/d reduces the plasma concentrations of carotenoids, an important group of antioxidants, alters the plasma carotenoid responses to exercise and increases the plasma levels of the inflammatory marker TNF-α. These findings suggest that a diet rich in fruit and vegetables is an important defence against oxidative stress for athletes.
Analysis of the participants' 4-d weighed food records found that their intakes of fibre, ascorbic acid and one marker of dietary carotenoid intake, β-carotene, were reduced on the R-AO diet, suggesting that they adhered to the recommendations to restrict their intake of fruit, vegetables and other major food sources of antioxidants during the study. Plasma concentrations of carotenoids are reflective of the previous several days dietary intake(Reference Yong, Forman and Beecher23), and as the pre-exercise plasma concentrations of carotenoids were lower on the R-AO diet compared to the H-AO diet, this indicates that the participants made the requested dietary changes for the 2 weeks of fruit and vegetable restriction. It also suggests that regular consumption of fruits and vegetables is required to maintain appropriately high levels of carotenoids in the circulation. Furthermore, this suggests that the habitual diet of these trained male endurance athletes has a desirably higher intake of dietary antioxidants sourced from fruit and vegetables.
IL-6 is a major inflammatory mediator released from skeletal muscle during exercise(Reference Fischer24). Dietary restriction of carotenoids did not effect the pre-exercise plasma concentrations of IL-6, but plasma IL-6 concentrations are typically low in endurance-trained adults(Reference Ostrowski, Rohde and Asp8) and in the present study baseline concentrations were comparable to healthy adults(Reference Niebauer, Clark and Webb-Peploe25, Reference Lenn, Uhl and Mattacola26). Exercise can increase the production of IL-6 in healthy trained and untrained adults(Reference Moldoveanu, Shephard and Shek6, Reference Ostrowski, Rohde and Asp8, Reference Castell, Poortmans and Leclercq27, Reference Peake, Suzuki and Wilson28), although the extent of any changes is dependent on the exercise characteristics. There were NS changes in plasma IL-6 concentrations in response to exercise in the present study, although the post-exercise concentrations of IL-6 appear higher on the H-AO diet compared to the R-AO diet. IL-6 was reduced following the R-AO diet post-exercise only. R-AO was expected to increase IL-6; however, it is likely that exercise may have contracted the effect of R-AO diet on IL-6 concentrations. This does merit further examination. Notably, there was no difference in IL-6 concentrations between H-AO and R-AO groups pre-exercise. The implications of this apparent difference, or whether the magnitude of this difference is physiologically significant, are not clear.
Plasma TNF-α concentrations were comparable to healthy adults before exercise(Reference Kimura, Suzui and Nagao11, Reference Fiotti, Giansante and Ponte29) and were not influenced by exercise, the same was observed with IL-6, but unlike IL-6, TNF-α concentrations were substantially higher both before and after exercise on the R-AO diet compared to the H-AO diet. Previous studies have shown varying effects of exercise on plasma TNF-α concentrations(Reference Moldoveanu, Shephard and Shek6, Reference Ostrowski, Rohde and Asp8, Reference Toft, Thorn and Ostrowski10, Reference Kimura, Suzui and Nagao11, Reference Mickleborough, Murray and Ionescu30). Dietary restriction had no effect on pre-exercise plasma LTB4 concentrations, which were comparable to healthy adults(Reference Mayatepek, Okun and Meissner31). LTB4 concentrations increased in response to exercise on the H-AO diet but not the R-AO diet. Previously, varying responses in plasma LTB4 concentrations have been observed with exercise including increases in trained and untrained adults(Reference Mickleborough, Murray and Ionescu30, Reference Hilberg, Deigner and Möller32) and no change in trained athletes(Reference Mickleborough, Murray and Ionescu30, Reference Peake, Suzuki and Hordern33). Post-exercise plasma LTB4 concentrations were lower on the H-AO diet than the R-AO diet. Plasma cytokine changes with exercise are varied, with increases in IL-6 of up to 8000-fold in athletes after a spartathlon(Reference Margeli, Skenderi and Tsironi34), increases in TNF-α of up to 2·6-fold(Reference Ostrowski, Rohde and Asp8) and increases in LTB4 of up to 3-fold(Reference Hilberg, Deigner and Möller32). This indicates that there is great variability in plasma cytokine responses to exercise. Together, the higher TNF-α and LTB4 concentrations on the R-AO diet indicate an elevated inflammatory state when deprived of valuable dietary antioxidants.
Plasma carotenoids have the potential to provide protection from oxidative stress during exercise(Reference Rock3). The plasma concentrations of some but not all carotenoids were lower before exercise on the R-AO diet compared to the H-AO diet, suggesting that carotenoids would be less capable of contributing to antioxidant protection on the R-AO diet. Intense aerobic exercise in fit adults decreases plasma antioxidants(Reference Bergholm, Mäkimattila and Valkonen35). Most plasma carotenoid concentrations decreased in response to exercise on the H-AO diet, but this response was markedly attenuated on the R-AO diet again suggesting that plasma carotenoids contributed less to providing protection from oxidative stress during exercise on the R-AO diet. Although α-carotene levels increased post-exercise following the R-AO diet, exercise was expected to decrease α-carotene on the R-AO diet. There was no difference in α-carotene levels between H-AO and R-AO pre-exercise. This merits further investigation.
Dietary restriction of carotenoids did not alter exercise performance. Participants' times to exhaustion did not differ between diets, but there was an increase in the rating of perceived effort during submaximal exercise on the R-AO diet compared to the H-AO diet. This indicates that dietary carotenoid restriction does not affect short-duration exhaustive exercise performance, but it can increase the participant's perception of effort possibly via an increase in exercise-induced oxidative damage while performing the submaximal exercise typical of endurance activities. The impact of a longer duration of dietary restriction of carotenoids on exercise performance or perceived effort is not known.
In summary, dietary restriction of fruits and vegetables to 1–2 serves/d in healthy endurance-trained men can increase plasma inflammatory mediators and decrease plasma carotenoid concentrations at rest. Post-exercise inflammatory mediators are elevated with dietary carotenoid restriction compared to a diet rich in fruit and vegetables. Exercise performance although unchanged was perceived to be harder when performed after dietary carotenoid restriction. The impact of a longer duration of dietary restriction of carotenoids on oxidative stress and inflammation and whether this has detrimental effects on exercise performance and perceived effort are not known. The present study suggests that a diet rich in carotenoids may be beneficial to combat exercise-induced oxidative stress in athletes performing exercise.
Acknowledgements
The authors would like to acknowledge the study subjects for their participation. There were no conflicts of interests in the present study. The authors have no conflict of interest and the present research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. B. A. P. participated in data collection, performed statistical analysis and drafting of the manuscript. R. C. participated in the design of the study, data collection and drafting of the manuscript. T. A. W. participated in the design of the study, data collection and performed statistical analysis. M. L. G. participated in the design of the study, provided advice on completion of the study and was involved in drafting of the manuscript.
