0856 Association of Naturally Occurring Mild Intermittent Hypoxemia and Health Outcomes in Sleep Apnea

Abstract Introduction Hypoxic conditioning, or the intentional exposure of mild intermittent hypoxia (MIH), has been shown to have protective cardiovascular outcomes. We investigated if naturally occurring MIH exposure during sleep has the same effect as controlled exposure. We hypothesized that larger number of respiratory events with MIH is associated with lower risk of adverse outcomes in OSA participants (apnea-hypopnea index ≥ 5 events/hour). Methods Participants from the Sleep and Heart Health Study cohort were analyzed. Respiratory events that were associated with MIH included all those accompanied with a 2-4% desaturation. Those with an oxygen desaturation > 4% were called non-MIH. For each participant, the frequency of MIH (AHI-MIH) and non-MIH (AHI-nonMIH) were calculated. Participants were categorized into 5 groups: those 1) without OSA (AHI 3% or arousal < 5 events/hour; reference group), 2) with OSA and high AHI-MIH (≥ median) but low AHI-nonMIH (< median), 3) with OSA and low AHI-MIH(< median) but high AHI-nonMIH(≥median), 4) with OSA and low AHI-MIH(< median) and low AHI-nonMIH(< median), and 5) with OSA and high AHI-MIH(≥ median) and high AHI-nonMIH(≥ median). Several logistic regression models examined the subgroups’ association with hypertension, diabetes, and Epworth Sleepiness Scale after adjusting for age, gender, race, and body mass index as confounders. In secondary analyses, we adjusted for the “hypoxic burden” as a measure of cumulative event-related intermittent hypoxia. Results 5634 participants were analyzed (means and standard deviations – age: 63.1±11.2 years; BMI: 28.2±5.1 kg/m2; gender: 52.5% females; AHI 17.9±16.1 events/hour). Compared to participants without OSA, individuals with low AHI-MIH but high AHI-nonMIH were at increased risk of hypertension (odds ratio [95% Confidence Interval]: 1.57 [1.25-1.96]). This risk was lower in participants with high AHI-MIH and high AHI-nonMIH(1.19 [0.99-1.43]) followed by the high AHI-MIH/low AHI-nonMIH subgroup (1.06 [0.85-1.31]) and the low AHI-MIH/low AHI-nonMIH subgroup (0.97 [0.82-1.15]). Similar findings were observed for diabetes and sleepiness. After adding hypoxic burden, hazards ratios were reduced, but the same pattern was observed across these groups for all outcomes. Conclusion Presence of MIH events appear to dilute health outcome risks associated with more severe desaturation events. This warrants further investigation to the role of MIH during sleep. Support (if any)