Treatment-Related Changes in Heart Rate Variability in Children with Sleep Apnea

Objective: Heart rate variability (HRV), a noninvasive indicator of autonomic regulation of cardiac rhythm, may represent the physiologic burden of obstructive sleep apnea (OSA). We hypothesized that the treatment-related effects of OSA on HRV in children are causally attributable to the improvement in OSA severity. Study Design: Secondary analysis of outcomes from the Childhood Adenotonsillectomy Trial (CHAT). Setting: Analysis of database. Subjects and Methods: Time- and frequency-domain HRV parameters along with polysomnographic (PSG) and demographic variables were obtained from the CHAT study, which compared early adenotonsillectomy (eAT) to watchful waiting (WW) in children with OSA. The relative contributions of PSG variables and covariates to each HRV parameter were quantified. The proportion of changes in HRV parameters causally attributable to changes in OSA severity, measured by the apnea-hypopnea index (AHI) and oxygen desaturation index (ODI), was estimated. Results: In total, 404 children aged 5 to 10 years were included. The median (interquartile range) age was 6 (3-9) years. The median body mass index percentile was 82 (53), 195 (48%) children were male, and 147 (36%) were African American. The average heart rate during PSG was the strongest independent predictor of each HRV parameter (P <.001). Although eAT resulted in statistically significant changes in the majority of HRV parameters, these effects were not causally attributable to treatment-related changes in AHI or ODI. Conclusions: The average heart rate strongly modulates HRV in children with OSA. Although eAT results in discernible changes in HRV, it appears to not be causally attributable to specific treatment-related changes in AHI or ODI.