Whereas the prevalence of exercise-induced hypoxemia (EIH) in endurance athletes is commonly reported as approximately 50%, most previous studies have not corrected PaO2 for exercise-induced hyperthermia. Furthermore, although a detrimental effect on aerobic performance has been assumed, no study has measured arterial oxygen content (CaO2) in this context.

Purpose: To determine the effect of temperature-correcting PaO2 values for rectal, arterial blood, esophageal, and exercising muscle temperatures during exercise on CaO2 and the prevalence of EIH.

Methods: Twenty-three trained males (age 26 +/- 5 yr; VO2peak 65.2 +/- 1.6 mL x kg-1 x min-1) performed incremental treadmill exercise to exhaustion with PaO2 corrected for simultaneous temperature measurements at all four sites. EIH was defined as DeltaPaO2 >or= 10 mm Hg.

Results: : With no temperature correction, DeltaPaO2 was -20.8 +/- 5.0 mm Hg and prevalence was 96% (n = 23), but when corrected for rectal temperature, DeltaPaO2 was -14.7 +/- 7.8 mm Hg and prevalence was 73% (n = 20); for arterial blood temperature, DeltaPaO2 was -7.7 +/- 6.5 mm Hg and prevalence was 35% (n = 20); and for esophageal temperature, DeltaPaO2 was -8.1 +/- 7.7 mm Hg and prevalence was 48% (n = 23), although when corrected for active muscle temperature, DeltaPaO2 was +8.2 +/- 7.8 mm Hg and prevalence was 0% (n = 10). There were no significant changes in CaO2 except for uncorrected values, and there was no correlation between DeltaPaO2 and VO2peak.

Conclusions: Although the prevalence of EIH depends on the temperature correction applied to PaO2 values, in no case is there a significant change in CaO2 or any relationship with maximal aerobic power.