Abstract
In chronic mountain sickness (CMS), increased blood oxygen (O2)-carrying capacity due to excessive erythrocytosis (EE, [Hb] ≥ 21 g/dL) could be offset, especially during exercise by both impaired cardiac output (Q̇t) and O2 diffusion limitation in lungs and muscle. We hypothesized that EE results in reduced peak V̇o2 despite increased blood O2-carrying capacity, and that isovolumic hemodilution (IVHD) improves exercise capacity. In 14 male residents of Cerro de Pasco, Peru (4,340 m), six with and eight without EE, we measured peak cycle-exercise capacity, V̇o2, Q̇t, arterial blood gas parameters, and (resting) blood volume. This was repeated for participants with EE after IVHD, reducing hematocrit by 20% (from 67% to 53%). From these data, we quantified the major O2 transport pathway components (ventilation, pulmonary alveolar-capillary diffusion, Q̇t, and blood-muscle mitochondria diffusion). Participants with EE had similar peak V̇o2, systemic O2 delivery, and O2 extraction as non-EE controls, however, with lower Q̇t and higher arterial [O2]. After IVHD, peak V̇o2 was preserved (but not enhanced), with lower O2 delivery (despite higher Q̇t) balanced by greater O2 extraction. The considerable variance in exercise capacity across the 14 individuals was explained essentially completely by differences in both pulmonary and muscle O2 diffusional conductances and not by any differences in ventilation, [Hb], nor Q̇t. In conclusion, EE does not result in lower peak V̇o2 in Andean males, and IVHD maintains, but does not enhance, exercise capacity.NEW & NOTEWORTHY Male Andean highlanders with and without excessive erythrocytosis (EE) have similar peak V̇o2 at 4,340 m, with higher arterial [O2] in EE and lower cardiac output (Q̇t), thus maintaining similar O2 delivery. Peak V̇o2 in participants with EE was unaffected by isovolumic hemodilution (hematocrit reduced from 67% to 53%), with lower O2 delivery balanced by slightly increased Q̇t and greater O2 extraction. Differences in lung and muscle diffusing capacity, and not hematocrit variation, accounted for essentially all interindividual variance in peak V̇o2.