JEPonline Journal of  Exercise Physiologyonline ISSN 1097-9751 An International Electronic Journal for Exercise Physiologists Vol 2 No 1 January 1999 Environmental Exercise Physiology Endurance performance effects of hyperoxic vs. normoxic breathing during interval training in female cyclists JEANNE F. NICHOLS, DAVID W. DOUGLASS, MICHAEL J. BUONO, SHAY MCKELVEY, AND SIMON MARSHALL Department of Exercise & Nutritional Sciences, San Diego State University, San Diego, CA JEANNE F. NICHOLS, DAVID W. DOUGLASS, MICHAEL J. BUONO, SHAY MCKELVEY, AND SIMON MARSHALL. Endurance performance effects of hyperoxic vs. normoxic breathing during interval training in female cyclists. JEPonline Vol 2 No 1 1999. The purpose of this study was to determine the differential effects on performance of hyperoxic vs. normoxic breathing during high-intensity interval training in female cyclists. Eighteen cyclists (34.9 ± 1.7 yr, VO2max 56.8± 6.5 ml·kg-1·min-1, mean ± SD) were randomly assigned (single blind) to either hyperoxic (FIO2 = 40-45%) or normoxic training, which consisted of eight, 2-min maximum effort intervals with a 4-min recovery performed once per week for six weeks. Outcome variables measured pre/post training on a Lode cycle ergometer with a Vmax 29 metabolic cart were VO2max, lactate threshold (LT), leg fatigue (sec) at 110% peak power output (PPO110), peak power output (PPO), and a 2.4 km hill-climb time-trial. Group (Normoxic/Hyperoxic) by time (pre/post) repeated measures ANOVA demonstrated no significant interactions for any dependent variable, no significant main effects by group (p>0.05), and a significant main effect for time for PPO110 (p = 0.001) and PPO (p < 0.022). These data indicate that both training groups responded similarly to the training regimen. Under the conditions in this study, interval training under hyperoxic conditions produced no added benefits to performance compared to training in a normoxic state. Future studies should examine training with higher concentrations of inspired oxygen as well as greater training time while in a hyperoxic state. Introduction Competitive athletes are continually in search of training techniques to enhance performance. It has been generally accepted that during exercise at sea level the pulmonary system of normal, healthy individuals is capable of maintaining arterial oxygen saturation at near resting levels. However, recent evidence (1-3) demonstrates that this may not apply to highly trained endurance athletes whose increased capacity of their cardiovascular system allows them to achieve very high levels of metabolic demand. Several studies (1-3) have shown that highly trained endurance athletes undergo a phenomenon known as exercise induced hypoxemia (EIH) during maximal or near maximal exercise. When this occurs, the percent oxygen saturation of hemoglobin (%SaO2) decreases from its normal level of 97% to levels below 92% (2). The resultant decrease in oxygen carrying capacity of the blood could be a limiting factor in endurance exercise performance (4-5). When hypoxic conditions are induced through the manipulation of the fractional concentration of inhaled oxygen (FIO2) during incremental ramp exercise, both VO2max and work rate are significantly reduced compared to normoxic conditions (5). Conversely, when athletes inhale high concentrations of oxygen (FIO2 0.26 - 0.70) during intense exercise, hypoxemia can be reversed and aerobic capacity and/or performance improved (4, 6-8). Furthermore, increasing FIO2 during exercise results in an increase in PaO2 (9, 10). Taken together, these findings support the notion that limitations in pulmonary oxygen diffusion contribute to the performance decrement observed with hypoxemia. While reductions in oxygen supply likely contribute to the decrement in VO2max during hypoxia, very recent evidence suggests that peripheral factors may also play a role (11). Using step-wise multiple regression analysis, Robergs et al. (11) provided evidence that the decrease in VO2max with increasing hypoxia is not uniform in all individuals, as has been suggested previously (6, 12). Furthermore, individuals with the highest values of VO2max do not necessarily demonstrate the greatest decrements. Robergs et al. (11) demonstrated that having a low lactate threshold and high lean body mass also contributed significantly to the decrement in VO2max during hypoxia. These findings indicate that muscle metabolic factors, possibly associated with decreases in muscle oxygen diffusion, also contribute to the performance decrement during hypoxia. The practical applications of the findings mentioned above were of interest in the present study. Specifically, given the effects on performance of elevated FIO2 during acute exercise (7, 8), it was of interest to know whether any training adaptations occur under similar conditions. The present study was undertaken to determine any possible performance benefits to training under hyperoxic conditions. To date, only one such study has been published (13). That study reported significant increases in work capacity after six weeks of high-intensity hyperoxic cycle training. However, the study design did not include a normoxic control group, therefore, the gains in work capacity could not be attributed to the hyperoxic breathing alone, but possibly to a training effect of the exercise itself. Furthermore, the study was conducted at altitude, where an increase in FIO2 would be expected to improve VO2max and performance. Therefore, the purpose of the present study was to examine, at sea-level, the differential training effects of normoxic vs. hyperoxic interval training in athletes at the peak of their training. Methods Subjects Eighteen highly trained female cyclists at the peak of their personal training programs were recruited from a women's race team and by word-of-mouth in the cycling community. Inclusion criteria included: women, 21-45 yr of age; currently cycle training at least eight hours per week; no known cardiovascular disease, as screened with the PAR-Q questionnaire (14); willingness to be randomized to a normoxic or hyperoxic condition. The subjects were asked whether they experienced any breathing problems during exercise, and if so, whether they used any prescription medication for that problem. One subject had a history of exercise-induced bronchoconstriction (EIB) and used Ventolin before her workouts. Because she indicated that use of the inhaler prevented EIB, she was allowed to remain in the study. This subject was randomized to the control group. The subjects included one former professional triathalete, one world class mountain cyclist, three nationally-ranked masters cyclists and one competitive racer of 15 years. Among the twelve remaining subjects, all but three were competitive cyclists and all were recruited at the peak of their competitive season. During the course of the study two subjects assigned to the treatment group became injured and were dropped from the study. The final sample size included nine control and seven treatment subjects. Procedures This study was approved by the University's Committee on the Protection of Human Subjects. Laboratory testing was conducted under barometric pressures ranging from PB 750 -756 Torr. Measurements included height, weight, aerobic capacity (VO2max), lactate threshold (LT), power output at a blood lactate concentration of 4 mM (LT4mm), and peak power output (PPO) achieved at VO2max. The performance measurements included a test of muscular resistance to fatigue at 110% of each subject's peak power output (PPO110), and a 2.4 km hill-climb time-trial on an outdoor course familiar to all subjects (HC). All testing was repeated within 10 days of the completion of the training program. VO2max and lactate threshold Aerobic capacity was assessed with an incremental cycle test on an electrically-braked cycle ergometer (Lode Excaliber, Groningen, The Netherlands). Before exercise, a baseline blood sample was obtained by finger lancet using 50 ul heparinized capillary tubes, and analyzed immediately, in duplicate, for whole blood lactate with the YSI Sport Lactate Analyzer (YSI Incorporated, Yellowsprings, Ohio). The analyzer was calibrated with standard solutions prior to each test. Subjects then warmed up for approximately 10-15 minutes at approximately 75-100 W at their preferred cadence. Following the warm-up, the resistance was increased to 100 W for the first stage, which was four minutes. The protocol consisted of 25 W increments every three minutes thereafter. During the last 30 seconds of each stage, and three minutes after VO2max was reached, a blood sample was obtained. Once the subject reached a lactate level of 4 mM the protocol changed to one minute stages of 25 W increments until exhaustion, which was defined as the point at which the subject could no longer turn the cranks. During the test, subjects breathed through a Hans Rudolph face mask with a 1-way valve (Hans Rudolph, Kansas City, Missouri) connected by 2" tubing to the metabolic cart. Expired gases were collected with a Vmax 29 metabolic cart (Sensor Medics, Anaheim, Ca.). The oxygen and carbon dioxide analyzers were calibrated with known gases prior to each max test. Heart rate was monitored throughout the test with a Unique CIC heart monitor (Polar, CIC, Incorporated, Port Washington, NY). Percent saturation of hemoglobin (% SaO2) was measured with a finger Oxyshuttle pulse oximeter (Sensor Medics, Anaheim, Ca.) every three minutes during the max tests. During the training program %SaO2 was measured at random 3-5 times per subject for 15 sec each. The pre and post %SaO2 @ VO2max values for each subject were averaged to determine %SaO2 @ VO2max. The decision to use the pulse oximeter was based on data from Powers et al. (2) demonstrating pulse oximetry as both a valid and reliable means of estimating %SaO2 during exercise at sea level. Percent saturation was recorded at the end of each stage of the test. EIH was defined as a %SaO2 equal to or less than 91% (2). Test of muscular resistance to fatigue (PPO110) Following a 5-10 minute rest, subjects performed a fatigue test at 110% of their peak power output achieved during the VO2max pre-test. The same absolute power output was subsequently used in the post-test. The resistance was initially set at 100 W while subjects gradually increased their cadence to greater than 110 rpm. The investigator then quickly increased the workload to the designated resistance, and the subject pedaled until cadence fell below 70 rpm. The stopwatch was started the moment the designated resistance was obtained and stopped the moment cadence fell below 70 rpm. The time taken for the investigator to increase the workload from 100W to 110% PPO was tested during 10 separate trials with a subject not involved in the study. The mean and range of scores were: 2.8 ± 0.44 sec, range = 2.31-3.85. Therefore, any change from pre to post test of ± 2 seconds was considered meaningful. Hill climb time-trial The final test, conducted on a separate day between 3 to 7 days from the lab tests, was a 2.4 km hill climb time-trial on an outdoor course selected by the investigator and team coach. The course, which was 7-8% in grade, was familiar to all subjects. Environmental conditions at both pre and post testing were very similar, with the temperature ranging from 25-27° C, relative humidity 70-76%, with a light tail wind (< 10 mph). Individual subjects started the hill climb in one-minute intervals to avoid drafting. Training Following baseline testing, subjects were randomly assigned and blinded to one of two training conditions (normoxic or hyperoxic). Subjects trained in the laboratory under controlled environmental conditions, temperature ranging from 20-23° C, once per week for six weeks. One, rather than two (or more) weekly laboratory training session was selected for this study because it substituted a similar training session in which most of the subjects were engaged once per week. Because 15 of the 18 subjects recruited were currently racing, on the average, one day per week, in addition to participating in another day of intense training, the team coach and investigators believed that more than one laboratory training session could potentially induce excess fatigue and staleness after six weeks. All subjects were fan-cooled during the training. Four to five subjects (2 hyperoxic, 2-3 normoxic) trained simultaneously each training session with their own bikes attached to wind trainers. Training sessions were led by the team's coach. The protocol consisted of eight 2- minute maximum intensity intervals with a four minute recovery interval (1:2 work to recovery ratio). During the interval session each subject breathed through a Hans Rudolph mask connected by tubing to either oxygen tanks (40-45% O2) or room air, depending on condition. The inspired gas from the oxygen tanks was non-humidified. The oxygen was turned on five seconds before each interval and turned off upon completion of that interval. Percent SaO2 was checked by finger oximetry one to two times in each subject during each training session. Statistical analysis Two (group) x two (time) ANOVA with repeated measures was used to test the six dependent variables. An independent t-test was used to detect significance between groups for %SaO2 during training and max testing. The level of significance was set at p < 0.05. Results Table 1 displays the physical characteristics of the subjects. No differences between groups were noted at baseline.   Table 1. Physical characteristics of subjects   Normoxic (N=9) Hyperoxic (N=7) Height (cm) 166.52 ± 5.15 166.19 ± 3.89 Weight (kg) 55.25 ± 3.69 56.67 ± 5.26 Age (yr) 34.89 ± 5.06 34.71 ± 8.30  * Values are expressed as group mean ± SD Table 2 displays the physiological and performance tests of the normoxic and hyperoxic groups pre and post-training.   Table 2. Physiological and performance data  (Values are expressed as group mean ± SD)   Normoxic Normoxic Hyperoxic Hyperoxic Pre Post Pre Post V02max (l·min-1) 3.14 ± 0.14 3.22 ± 0.42 3.22 ± -.39 3.18 ± -.19 PO@LT (Watts) 150 ± 43.30 164 ± 57.71 161 ± 28.35 164 ± 24.40 PO@4nM La (W) 173 ± 51.86 192 ± 53.01 186 ± 27.22 186 ± 23.46 PPO (W) 283 ± 27.95 306 ± 37.03 296 ± 22.49 296 ± 10.21 PPO110% (sec) 80.22 ± 19.83 100.64 ± 26.21 59.95 ± 11.67 75.18 ± 18.49 Hill Climb (min) 6.88 ± 0.65 6.66 ± 0.51 6.50 ± 0.52 6.68 ± 0.62 Results of the 2 x 2 ANOVA for the VO2max test indicated a nonsignificant group effect (F1,13 = 0.03; p > 0.05), a nonsignificant time effect (F1,13 = 1.64; p > 0.05), and a nonsignificant interaction (F1,13 = 0.03; p > 0.05). Power output at LT had a nonsignificant group effect (F1,14 = 0.08; p > 0.05), a nonsignificant time effect (F1,14 = 3.80; p = 0.07), and a nonsignificant interaction (F1,14 = 1.33; p > 0.05). PO @ 4 mM blood lactate had a nonsignificant group effect (F1,14 = 0.03; p > 0.05), a nonsignificant time effect (F1,14 = 2.73; p > 0.05), and a nonsignificant interaction (F1,14 = 2.50; p > 0.05). Peak power output showed a nonsignificant group effect (F1,13 = 0.00; p > 0.05), a significant main effect for time (F1,13 = 6.75; p < 0.05), and a nonsignificant interaction (F1,13 = 3.16; p > 0.05). Both performance tests (PPO110 and HC) had nonsignificant group effects (F1,13 = 4.63, p > 0.05; F1,12 = 0.16, p > 0.05), a significant main effect for time for PPO110 (F1,13 = 19.39, p < 0.05), a nonsignificant time effect for HC (F1,12 = 0.62, p > 0.05), and a nonsignificant interaction (F1,13 = 0.90, p > 0.05; F1,12 = 2.13; p > 0.05) for both. Table 3 displays the means and standard deviations of %SaO2 at VO2max and %SaO2 during training. The percent of subjects who exhibited EIH at VO2max and during training is also presented.   Table 3. Percent SaO2 at VO2maxand during training, and proportion of subjects exhibiting EIH at max and during training   Normoxic Hyperoxic SaO2 (%)@VO2max 90.8±3.46 93.1±2.39 Subjects exhibiting EIH @ VO2max (%) 44.4 28/6 SaO2 (%) during training 89.7±3.9 98.9±0.8 Subjects exhibiting EIH during training (%) 55.6 0.0 * EIH is defined as %SaO2 equal to or less than 91. * VO2max SaO2 (%) is the mean SaO2 from the pre and post VO2max tests for each subject. * Training SaO2 (%) is the mean of 3-5 15-sec pulse oximeter readings for each subject. No group differences in %SaO2 were detected at VO2max, either pre or post training. Percent SaO2 was significantly lower in the normoxic compared to hyperoxic group during training (F1, 15 = 12.32; p < 0.05). Discussion This study provides additional data illustrating that approximately 50% of highly trained endurance athletes, both male and female, experience EIH at sea-level during maximal or near-maximal exercise (1-3). Half of the subjects in the control group desaturated during training, while 38% (6 of 16) of the total sample experienced EIH at VO2max. Given that endurance performance is reduced under hypoxic conditions (5) and that acute exercise performance is enhanced by increasing the fractional concentration of oxygen in inspired air (8) we hypothesized that training under hyperoxic conditions would result in adaptations greater than those achieved through training under normoxic conditions. The results of the present study do not support our hypothesis. Physiological markers of endurance performance were not enhanced by subjects breathing a high concentration of oxygen (40-45%) during a six week interval training program. These findings contradict those of Chick et al. (13) who found significant increases in work capacity after six weeks of hyperoxic training using 70% oxygen. Several factors may be responsible for the contrasting results. First, the Chick et al. study (13) did not employ a normoxic control group. This design flaw makes it difficult to interpret their results. The adaptations reported may not have been due to the hyperoxic condition per se, but to an overall training effect. Second, the Chick et al. study was conducted at moderate altitude, which may have induced EIH at a lower exercise intensity, thus increasing the likelihood of subjects benefiting from training with supplemental oxygen. Another consideration was that the fractional concentration of oxygen used in the present study was considerably lower than that used previously (13); thus, it is unclear as to whether a higher concentration of oxygen would have elicited different results in the present study. Although it has been shown that only a small increase in FIO2 will reverse EIH in exercising athletes, treadmill performance time at 110% of peak work capacity achieved during a VO2max test increases as the fractional concentration of O2 is increased, thus indicating a dose-related response (8). The proposed mechanism for this dose response improvement was based on the finding that pulmonary ventilation decreases as FIO2 is increased, thus decreasing the energy cost of breathing. This would result in improved ventilatory efficiency at a given VO2. The greater efficiency may allow the subject to perform more work as FIO2 is increased, independent of any effect on %SaO2. This suggests that factors unrelated to reducing EIH may play a role in increasing performance by breathing a hyperoxic gas mixture (8). Although most of the research to date related to performance under hypoxic and hyperoxic conditions indicates that pulmonary mechanisms play a major role in determining oxygen delivery to contracting muscle, new research provides evidence that oxygen diffusion in skeletal muscle may also be important (11). The interesting finding that the lactate threshold and absolute amount of lean body mass contribute to the decrement in VO2max during hypoxia warrants further research, both during hypoxia and hyperoxia. A factor to consider for the lack of hyperoxic training effects found in the present study was the total volume of training under hyperoxic conditions. Although the training protocol was highly effective in inducing positive changes in several variables, the actual time the subjects were breathing increased oxygen was only 16 minutes (8 intervals x 2 min/interval) per week. Perhaps a more frequent stimulus in the hyperoxic condition would yield different results. This needs to be examined further. An important consideration was that only 29% (2 of 7) of the treatment group actually desaturated during their VO2max test, compared to 44% (4 of 9) of the control subjects. In the control group, those subjects who desaturated during the VO2max test also desaturated (SaO2 <90%) during training. Only one of the remaining five control subjects, who was just above the cut-off for desaturation at VO2max (SaO2 92%), desaturated during training. This finding indicates that those subjects who exhibit EIH at VO2max are more likely to do the same during high-intensity interval training, while those who do not experience EIH at VO2max may not do so during training. Given this, the fact that only two treatment subjects experienced EIH with maximal exercise may have precluded them benefiting from hyperoxic training if they were not hypoxic to begin with. In conclusion, the results of this study indicate that high intensity interval training performed once per week for six weeks can induce significant gains in physiological measures of endurance performance. Although no performance benefits were observed from hyperoxic training, further study on this topic is needed. One recommendation is to examine, under conditions of hyperoxic vs. normoxic training, only those subjects who experience EIH during maximal or near-maximal exercise. Future studies should also manipulate the total volume of training under hyperoxic conditions in order to further examine the potential performance benefits of hyperoxic training.  References 1. Phares, S., J.F. Nichols and M.J. Buono. Oxygen desaturation at VO2max in highly trained female master athletes. Proceedings of the Southwest Regional American College of Sports Medicine Annual Meeting, 1996. 2. Powers SK, Dodd S, Lawler J, Landry G, Kirtley M, McKnight T, et al. Incidence of exercise induced hypoxemia in elite endurance athletes at sea level. Eur J Appl Physiol 1988;58:298-302. 3. Williams JH, Powers SK, Stuart MK. Hemoglobin desaturation in highly trained athletes during heavy exercise. Med Sci Sports Exerc 1986; 19:168-172. 4. Ekblom B, Huot R, Stein EM, Thorstensson AT. Effect of changes in arterial oxygen content on circulation and physical performance. J Appl Physiol 1975; 39(1): 71-75. 5. Hughson RL, Kowalchuk JM. Kinetics of oxygen uptake for submaximal exercise in hyperoxia, normoxia, and hypoxemia. Can J Appl Physiol 1995; 20(2):198-210. 6. Powers SK, Lawler J, Dempsey JA, Dodd S, Landry G. Effects of incomplete pulmonary gas exchange on VO2max. J Appl Physiol 1989; 66(6):2491-2495. 7. Adams RP, Welch HG. Oxygen uptake, acid-base status, and performance with varied inspired oxygen fractions. J Appl Physiol 1980; 49(5): 863-868. 8. Wilson GD, Welch HG. Effects of hyperoxic gas mixtures on exercise tolerance in man. Med Sci Sports Exerc 1975; 7:48-52. 9. Dempsey JA, Hanson PG, Henderson KS. Exercise-induced arterial hypoxemia in healthy human subjects at sea level. J Physiol 1984; 355:161-175. 10. Powers SK, Martin D, Cicale M, Collop N, Huang D, Criswell D. Exercise-induced hypoxemia in athletes; role of inadequate hyperventilation. Eur J Appl Physiol 1992; 65:37-42. 11. Robergs RA, Quintana R, Parker DL, Frankel CC. Multiple variables explain the variability in the decrement in VO2max during acute hypobaric hypoxia. Med Sci Sports Exerc 1998; 30(6):869-879. 12. Buskirk ER, Kollias J, Akers RF, Prolop EK, Reategui EP. Maximal performance at altitude and on return from altitude in conditioned runners. J Appl Physiol 1967; 23:259-266. 13. Chick TW, Stark DM, Murata GH. Hyperoxic training increases work capacity after maximal training at moderated altitude. Chest 1993; 104(6):1759-1762. 14. Thomas S, Reading J, RJ Shephard. Revision of the Physical Activity Readiness Questionnaire (PAR-Q). Can J Sport Sci 1992; 17:338-345.  Address for correspondence:Jeanne F. Nichols, Ph.D., Department of Exercise & Nutritional Sciences, Mail Code 7251, San Diego, CA 92182-0171, [phone: 619-594-1926, fax: 594-6553.  Copyright ©1999 American Society of Exercise Physiologists All Rights Reserved ASEP Table of Contents JEPonline (January 1999)