Clinical and Polysomnographic Characteristics of Nonobese and Obese Chinese Patients With Obstructive Sleep Apnea.

Abstract

PURPOSE\nObesity is a risk factor associated with the onset and exacerbation of obstructive sleep apnea (OSA). However, the majority of OSA patients in Asian populations are nonobese. To date, there have been insufficient large-scale studies of the differences in the clinical and polysomnographic features of obese and nonobese OSA patients in this population, and few studies have sought to identify predictors of OSA severity in affected obese and nonobese patients.\n\n\nMETHODS\nWe conducted a case-matched retrospective study, including 652 consecutive Chinese OSA patients (326 nonobese and 326 obese) to assess differences in demographic, clinical, and polysomnographic data between these two groups. Independent predictors of OSA severity were identified through multivariate linear regression analysis.\n\n\nRESULTS\nThe age and gender distributions of our obese and nonobese OSA patient cohorts did not differ significantly (P > 0.05), and rates of comorbidities were comparable in these two patient groups (P > 0.05). Nonobese patients were more likely to report atypical symptoms of OSA, including insomnia (P < 0.001), irritability (P < 0.05), and depressive symptoms (P < 0.01), whereas obese patients were more likely to report typical symptoms of OSA, such as habitual snoring (P < 0.001), witnessed apnea (P < 0.05), and daytime sleepiness (P < 0.001). Relative to nonobese patients, those who were obese exhibited significantly higher apnea-hypopnea index during total sleep time (P < 0.001), apnea-hypopnea index during nonrapid eye movement sleep (P < 0.001), and apnea-hypopnea index during rapid eye movement sleep (P < 0.001), in addition to lower average oxygen saturation (P < 0.001), minimal oxygen saturation (P < 0.001), and a higher oxygen desaturation index (P < 0.001) and arousal index (P < 0.001). Total sleep time was, on average, shorter for nonobese patients (P < 0.05), who also exhibited decreased sleep efficiency and more frequent awakening relative to obese patients (P < 0.05). A multivariate linear regression analysis revealed that neck circumference and waist circumference were independent predictors of OSA severity in obese patients (P < 0.05).\n\n\nCONCLUSIONS\nOn average, OSA was typically less severe in nonobese patients, who were also more likely to experience atypical OSA symptoms relative to obese patients. These results also suggest that the differential contributions of body fat accumulation and distribution to OSA severity may offer insights into the pathogenesis, clinical manifestations, and optimal management of this condition in nonobese and obese patient populations. It is vital that clinicians consider these differences to properly diagnose and treat this debilitating condition.