Study objectives: To determine the effect of varying approaches to the measurement of the respiratory disturbance index (RDI) on identification of sleep disordered breathing (SDB) in children.
Design: Cross-sectional study of SDB in a well-characterized birth cohort (stratified for term and preterm birth) participating in longitudinal studies of cognition and behavior.
Setting: Community-based; overnight studies conducted in participant's homes.
Participants: 433 children, ages 8-11 years, un-referred for clinical assessment of SDB.
Interventions: N/A.
Measurements and results: Participants underwent unattended overnight in-home recording of respiratory inductance plethysmography, pulse oximetry, body position, and heart rate. The relationships among RDIs derived from various definitions of apnea and hypopneas and thresholds for frequency of events were assessed. Median RDI varied by more than 20-fold for definitions using the more liberal criteria for event definition (all respiratory events; i.e., central or obstructive events and hypopneas with no requirement for associated desaturation) to the most conservative definition (using obstructive apneas only or obstructive apnea and hypopneas requiring a 5% associated desaturation). Prevalence estimates for SDB based on RDIs that included central apneas were 40% to 140% higher than those that excluded central apneas.
Conclusions: Different approaches for quantifying RDI contribute to substantial variability in identification and classification of SDB in children and will lead to discrepant estimates of its presence and severity.