2.1. Intermittent Hypoxia

The intermittent hypoxia protocol consisted of five 4-min hypoxic cycles at a targeted arterial oxygen saturation of 90% interspersed with 4-min normoxic cycles. This protocol was based on the intermittent hypoxia protocol used by Burtscher et al. [9], which consisted of progressive changes in the number of hypoxic cycles (3, 4, and 5 cycles), hypoxic duration (3, 4, and 5 min), and hypoxic severity (fraction of inspired oxygen of 0.15, 0,13, and 0.12) over 3 weeks. The 4-min normoxic duration is in accordance with the half-life of 5–8 min of hypoxia inducible factor-1α [12], and was based on previous intermittent hypoxia protocols that used normoxic bouts lasting between 3 to 5 min [9,13]. Due to the high individual variability in hypoxic ventilatory responses, a fixed fraction of inspired oxygen can result in a wide range of arterial oxygen saturation across individuals. Moreover, arterial oxygen saturation was observed to increase over the course of a session of intermittent hypoxia of 5 cycles at a fixed oxygen level [14]. Therefore, intermittent hypoxia was not performed at a fixed oxygen level but at a targeted arterial oxygen saturation in order to induce the same level of hypoxemia for each cycle and in all participants. An arterial oxygen saturation of 90% was chosen based on our previous observations that this arterial oxygen saturation corresponds to a fraction of inspired oxygen of 0.12 ± 0.01 in young healthy individuals. This intermittent hypoxia protocol is therefore considered mild as it represents less than 10 cycles of hypoxia lasting between 15 s to 4 min at a fraction of inspired of oxygen between 0.10 to 0.14 [15].

Participants inhaled hypoxic air through a mask connected to a two-way non-rebreathing valve (Hans Rudolph, Inc, Shawnee, KS, USA), which was itself connected to a five-liter non-diffusing gas bag (Hans Rudolph, Inc, Shawnee, KS, USA). The rebreathing bag was connected to a gas tank of compressed air. Air was made hypoxic by introducing nitrogen in the breathing circuit. The flow of nitrogen was controlled to achieve an arterial oxygen saturation of 90%, as measured by pulse oximetry. Participants in the intermittent normoxia group performed the same protocol, but nitrogen was not introduced in the breathing circuit. Pulmonary gas exchange, hemodynamics, and arterial oxygen saturation were continuously measured during the intermittent hypoxia protocol and intermittent normoxia protocol.