Role of ventilation on arterial oxygen saturation during exercise

Wood, R.J., 'Role of ventilation on arterial oxygen saturation during exercise', PhD Thesis, The University of Western Australia, 1998.

For some highly-trained endurance athletes during maximal exercise, the pulmonary system is not able to maintain the high arterial oxygen levels as it does for normal healthy untrained individuals. As exercise Ve level has been postulated as playing a significant role in this desaturation among athletes, the purpose of the experiments of this thesis was to examine the relationship between Ve and Sa02 during maximal exercise in trained and untrained individuals under varied experimental conditions.

A validation study during maximal cycling exercise found high correlations of direct measurements of arterial blood, using CO-oximetry, with Ohmeda Biox 3700e (r = 0.91) and Criticare 504 (r =0.94) pulse oximeter measurements. The average of the two oximeters gave a better correlation (r = 0.95) and was therefore used for all SaO2 measurements.

Exercise performance as determined by a maximal 6-minute rowing ergometer test was compared at Canberra (CB, 610 m altitude) and at sea level (SL). Peak V02 was 3.6% lower at CB (SL, 4.12 ± 0.24; CB, 3.97 ± 0.25 1-min1). Peak Ve btps (SL, 152.5 ± 11.1; CB, 149.1 ± 11.7 1-min"1) was similar, though peak Ve stpd (SL, 125.6 ± 9.1; CB, 114.1 ± 9.0 1-min"1) and therefore 02 delivery was lower at altitude. The lower V02 at altitude was concluded to be due to a lack of ventilatory compensation to the lowered PiQ2, possibly affected by a reduction in Sa02. A subsequent study at the same though simulated altitude (ALT), using cycle ergometry with highly-trained athletes, supported these findings. There was a similar peak ve btps (SL, 180.1 ± 18.4; ALT, 175.3 ± 20.3 1-min"1) while peak ve stpd (SL, 145.7 ± 14.9; ALT, 131.6 ± 15.3 l.min-1) was different between conditions. The 6.9% lower V02ma at altitude (SL, 5.48 ± 0.29; ALT, 5.10 ± 0.25 1.min-1) was associated with a lower minimum SaO2 (SL, 89.0 ± 3.3; ALT, 85.1 ± 4.7 %). In untrained subjects, there were no significant differences for SaO2 (SL, 94.7 ± 2.3; ALT, 92.7 ± 2.3 %), V02max (SL, 3.91 ± 0.50; ALT, 3.79 ± 0.49 1-min') and ve (btps and stpd).

To investigate the effect of exercise mode differences in ve and V02 on SaO2, thirteen highly-trained male triathletes performed V02max tests on treadmill (TR), cycle (CY) and arm-crank (AC) ergometers. Ve for TR and CY was higher than for AC (TR, 160.7 ± 22.5; CY, 163.0 ± 22.6; AC, 136.9 ± 23.2 1-min1), and peak V02 were all significantly different from each other (TR. 4.94 ± 0.39; CY, 4.71 ± 0.34; AC, 3.63 ± 0.42 1.min-1), which resulted in Ve/VO2 being higher for AC (TR, 32.5 ± 2.9; CY, 34.6 ± 3.5; AC, 37.7 ± 4.8). SaO2 values at peak V02 were lower only for TR (TR, 91.0 ± 3.3; CY, 93.9 ± 1.9; AC, 95.7 ± 2.0 %). Sa02 values followed the differences in Ve/VO2, though analysis within each exercise mode found no correlation between SaO2 and the other measured variables. However, when all exercise data were pooled (n = 39), significant correlations were found between SaO2 and VE/V02 (r = 0.36), and between S.02 and V02 (r = -0.42).

Seven of these subjects were re-tested on the cycle ergometer after approximately one year. The changes between tests for Ve, V02max, and SaO2 were highly variable; with only V02max significantly different (4.61 ± 0.39 and 4.91 ± 0.56 1.min-1). Analysis of individual cases showed that the two subjects who showed a large decrease in Sa02 between tests also had a large increase in V02max. While the individual differences were highly variable, there appeared to be a trend for changes in SaO2 to follow the VE/V02 in both the individual and mean group data.

The effect of experimentally reducing ve was studied in fourteen male and female athletes of moderate to high fitness level. Each subject performed two V02max tests, one in which Ve was limited by making the subjects breathe solely through the nose (N) the other in which they breathed through the mouth only (0). Despite the large difference in Ve at maximum exercise (0, 133.0 ± 29.1; N, 86.2 ± 23.7 1.min-1), the difference in SaO2 for each condition was small though still significant (0, 95.1 ± 2.0; N, 93.3 ± 3.4 %), and achieved at a much different workload and V02 level (0, 3.90 ±'0.94; N, 3.33 ± 0.85 l.min-1). Therefore, VE/V02 was much lower for the N condition (0, 34.6 ± 4.1; N, 26.0 ± 3.2). Although V02 (02 demand) was lower for N, the reduction in ve and therefore 02 supply was enough to still cause desaturation, indicating the importance of both of these variables in determining SaO2 levels during exercise.

Resting lung function measurements taken from untrained and highly-trained athletes were compared to Sa02 levels achieved during exercise. There was no difference for resting lung function between untrained and highly-trained, although within the highly-trained athletes the variation in Sa02 levels was partially explained by differences in resting lung function.

Finally, to investigate whether voluntary increases in ve could prevent desaturation, seven athletes who had previously demonstrated significant desaturation during maximal exercise, performed three submaximal 4-minute constant-load cycle ergometer tests at 85% of V02max; one a control test with free breathing, the other two using biofeedback to control ventilation, one to mimic the control level and another at a hyperventilation level. Hyperventilation of 15% higher ve than the control level resulted in a small though significantly higherSaO2 (95.5 ± 0.8 versus 94.0 ± 0.9 %). Although the difference was small, this showed the potential for athletes to voluntarily control their ventilation levels to prevent desaturation, and therefore possibly improve performance. These data demonstrate the important role of ve levels in the determination of desaturation in athletes, with a clear trend for a relationship between VE/V02 and Sa02 levels during exercise shown in all studies. However, the role of V02 appears to be just as important.

Related Pages