Pathophysiological Traits in Pediatric Obstructive Sleep Apnea. Associations with Patient Characteristics and Responses to Therapy: A Secondary Analysis of the CHAT Clinical Trial

Abstract Rationale In children with adenotonsillar hypertrophy, there is substantial variation in pediatric obstructive sleep apnea (pOSA) severity, which may be driven by differences in pathophysiological traits (endotypes), including pharyngeal collapsibility, dilator muscle compensation, arousal threshold, and chemoreflex loop gain. Objectives To determine if pathophysiological traits for pOSA vary with participant characteristics, correlate with pOSA severity, and change after adenotonsillectomy. Methods Traits estimation, requiring adequate nasal pressure data quality, was performed during sleep (primary analysis in rapid eye movement REM) from baseline polysomnography (N = 1,117; age 6.9 ± 1.5 yr; body mass index standardized using age- and sex-specific z-score BMI-z 0.89 ± 1.23; total apnea–hypopnea index AHI, 5.1 ± 9.1 events/h; REM AHI, 11.0 ± 20.1 events/h) and postintervention polysomnography (N = 360; adenotonsillectomy or watchful waiting). Associations of each endotype (per standard deviation SD) with AHI and with patient characteristics (race/ethnicity, age, sex, and BMI-z) were characterized using multivariable regression. The effects of adenotonsillectomy on AHI and endotypes were also examined. Results The sample comprised 52% females and children from diverse racial and ethnic backgrounds and geographic sites. Higher REM AHI was observed in Black (7.81 ± 1.01, βadjusted ± standard error of the mean, events/h) and Asian (9.37 ± 3.35) compared with White children; these differences were accompanied by greater collapsibility (0.30 ± 0.09 per SD) in Black children and decreased compensation (−0.99 ± 0.20) in Asian children. Notably, sex and BMI-z were not associated with any endotype. Higher REM AHI was associated with greater collapsibility (13.64 ± 1.73 events/h/SD) and reduced compensation (−4.22 ± 0.98) but not increased loop gain. Increased collapsibility and reduced compensation partially explained higher REM AHI in Black as well as Asian and Indigenous children. Reduced AHI with adenotonsillectomy was accompanied by improved collapsibility in REM (Δ = 16.88% ± 1.94%). Conclusions Increased collapsibility and reduced compensation contribute to higher REM AHI levels in children and may explain an elevated pOSA severity in REM in Black and Asian children. Clinical trial registered with www.clinicaltrials.gov (NCT 00560859).