Reversible Effect of Falling Ventilatory Drive in Drive-dependent OSA

Inclusion Criteria: * Diagnosed OSA (AHI≥15 events/h reported in a PSG performed within 1 year) or Suspected OSA (snoring, sleepiness, witnessed apneas, other clinical symptoms) * Use of CPAP or other therapies is acceptable; individuals will be asked to withhold treatment for 3 days before each study visit. Individuals who are occupational drivers or operate heavy machinery will not be asked to withhold treatment. Exclusion Criteria: * Any unstable medical conditions * Conditions that could meaningfully raise the cardiovascular risks of brief low-dose hypercapnic-hypoxic inspired gas mixture: heart failure (LVEF\<45% if known), recent cardiovascular event (\<12 mo), recent cerebrovascular event (\<12 mo)\* * Medications known to depress ventilatory drive (e.g. opioids, barbiturates) * Conditions likely to increase arousability from sleep: insomnia * Other sleep disorders that may complicate establishment of sleep: periodic limb movements (periodic limb movement arousal index \> 10/hr), narcolepsy, or parasomnias * For intramuscular electrodes and catheter: allergy to lidocaine * Highly-sensitive gag reflex. Patients with a self-reported 'highly-sensitive gag reflex', including an affirmative response to 'Do you sometimes gag when brushing your teeth?', will not take part in the physiology studies given the placement of an esophageal catheter * For intramuscular electrodes: use of aspirin or other oral anti-platelets / anti-coagulants * For oronasal mask: severe claustrophobia * Pregnancy or nursing * We do not intend to exclude patients with controlled cardiovascular disease (hypertension of any severity, arrhythmias, stents) common in the OSA patient population. The transient gas mixture interventions are mild, short-lived, and act to slow the spontaneous recovery of blood gas levels to prevent cyclic upper airway obstruction as opposed to exacerbating them. Control of breathing studies commonly increase inspired CO2/reduce inspired oxygen (using higher concentrations via rebreathing tests for longer durations) in patients with a range of comorbidities including heart failure. The level of hypercapnic-hypoxia used is equivalent to taking slightly smaller breaths (by about a third, for the standard dose gas mixture 2%CO2/18.5%O2) for several breaths, or skipping a breath (for the highest dose gas mixture 6%CO2/14%O2), physiological changes that typically cause no noticeable oxygen desaturation, and are minimal compared with the effects of the larger ventilation reduction that accompanies OSA.