0533 AROUSAL THRESHOLD, OBESITY, AGE AND RACE PREDICT CONTINUOUS POSITIVE AIRWAY PRESSURE USE AMONG U.S. VETERANS WITH OBSTRUCTIVE SLEEP APNEA

Abstract

Introduction: While the respiratory arousal threshold (ArTH) is a key factor in obstructive sleep apnea (OSA) pathogenesis, limited knowledge exists regarding its distribution, association with physiological and clinical patient characteristics and relationship with continuous positive airway pressure (CPAP) use in large clinical cohorts. Accordingly, we aimed to 1) identify characteristics associated with and 2) assess the impact of low ArTH on CPAP use among U.S. Veterans with OSA. Method(s): Demographics, comorbidities, polysomnographic measures and CPAP use were assessed in U.S. Veterans (n=975) enrolled in a multi-site observational cohort with mean follow-up of 5.5 years. In patients with OSA (apnea hypopnea index, AHI>=5/hour and recommended treatment) presence of low or high ArTH was estimated using previously validated polysomnographic predictors (AHI, nadir nocturnal oxygen saturation and fraction of hypopneas). Associations between low ArTH, demographics, comorbidities and polysomnographic measures were evaluated using univariate, and predictors of regular CPAP use were evaluated by multivariate logistic regression. Result(s): Thirty eight percent of OSA patients exhibited a low ArTH. In bivariate analyses, the odds of exhibiting a low ArTH increased with advancing age and periodic leg movement index (PLMI), while odds of a high ArTH increased with obesity (body mass index, BMI>=30kg/m2), non-white race/ethnicity and hypertension (p-value < 0.05). The odds ratio (OR) for regular CPAP use was markedly reduced in non-obese Veterans with low arousal threshold (OR [95%CI]: 0.38 [0.20,0.72]) as opposed to obese Veterans (1.01 [0.74,1.39]). In multivariate analyses, the interaction term of low ArTH with obesity, increasing age (per 10 years) and non-white race/ethnicity were associated with lower odds of regular CPAP use (ORs of 0.36 [0.19,0.68], 0.86 [0.75,0.97] and 0.68 [0.48,0.96] respectively), after adjustment for gender, sleepiness, AHI, hypertension and PLMI. Conclusion(s): The impact of arousal threshold on CPAP use may not be uniform among all OSA patients. Only in the non-obese patients does a low arousal threshold seem to be associated with reduction of regular CPAP use. Similarly, despite exhibiting a higher arousal threshold, non-white race/ethnicity was associated with lower rates of regular CPAP use, suggesting alternate factors influencing CPAP use disparities in our cohort.