Analysis of airflow indices in paediatric respiratory disorders

Abstract

AimDuchenne Muscular Dystrophy (DMD) and Prader-Willi syndrome (PWS) are complex congenital disorders in which sleep disordered breathing (SDB) typically develop, and this impacts normal growth and neurological development. Due to the slow progression of SDB pathology and the lack of correlation between SDB pathology and daytime clinical symptoms, these patients are often only actively treated after SDB is fully established. This study aims: (1) to determine whether there are changes in breath timing and breath shape (morphology) during sleep in these paediatric respiratory disorders when compared with healthy children; and (2) whether there is an association between physiological markers of disorder progression with these indices of breath timing and shape.MethodsA retrospective cohort of 13patients with DMD and 11 patients with PWS whom had at least three sequential overnight polysomnograms and 27 healthy controls (2-17 years) were identified from a clinical database. The nasal pressure signal in at least two ten-minute segments of stage 2 non-rapid eye movement (NREM) sleep for each patient was extracted from each retrospective polysomnogram. Each breath was automatically segmented into inspiratory and expiratory phase using a customised algorithm, and these were used quantified using three categories of indices: (1) Timing indices including total breath time (Ttot) and fractional inspiratory time (FIT) (mean and standard deviation, SD); (2) Linear shape descriptors (skewness, kurtosis and variance); and (3) Non-linear shape descriptors (line morphology, length – ‘flatness’ – and variability). In addition, standard physiological markers of disease progression were extracted from the polysomnogram report: total apnoea-hypopnoea index (AHI), body-mass index (BMI), mean SpO2 and mean CO2. Indices of breath timing and shape were compared between the disorder cohorts and healthy controls; and the changes in indices over years for cohorts were compared to standardphysiological measures to determine if there was any association with disorder progression.ResultsIn patients with DMD, the mean FIT was statistically different than healthy controls (0.47 vs. 0.41, p<0.05) and there was a trend that SD of FIT was higher in patients with DMD going on to treatment, than those not (0.045 vs. 0.054). Linear shape indices of inspiratory kurtosis and variance linearly correlated with higher mean CO2 over years. So, too, did expiratory variance and expiratory line length, the latter with higher BMI. For patients with PWS, their SD of FIT (0.09 vs 0.06, p<0.05) was higher in than healthy controls. Their mean CO2 is positively correlated with inspiratory kurtosis and inspiratory line (flatness). The higher AHI and BMI were correlated with inspiratory line variance, butnegatively correlated for expiratory line length and variance.DiscussionBreath timing indices during sleep are different in patients with DMD or PWS relative to healthy controls. They both exhibit different timing characteristics; patients with DMD have a higher mean FIT and this appears to be more variable with impending need for treatment, whereas the patients with PWS have variability in FIT – irrespective of longitudinal trend or BMI changes. These linear and non-linear analyses correlate shape and its variability with negative physiological outcome from first to the last study in some patients with DMD or PWS. For those patients with DMD with negative physiological outcome of higher mean CO2, the inspiratory flow shape changed to a more peaked (higher inspiratory kurtosis) and narrowed shape (low variance). The expiratory flow shape in patients with DMD is more rounded as it is wider (expiratory variance high) with short ‘flat’ periods (short expiratory line length) with higher BMI – the end expiratory pause is reduced. In patients with PWS, the shape changes are different to patients with DMD in that the inspiratory flow shape has shorter ‘flat’ periods (short inspiratory line length), shape rounding, with negative physiological outcome. The expiratory shape in patients with PWS have longer line lengths (long expiratory line length) with higher variability of the line length (larger expiratory line width at half maximum). Inspiratory and expiratory flow limitation (IFL and EFL) exhibit differences in shape characteristics associated with physiological markers in these two paediatric respiratory disorders. These differences may be indicative of a deterioration of respiratory muscles in patients with DMD and an increasing upper-airway resistance and breathing control abnormalities and upper-airway resistance in patients with PWS. These indices, derived from conventional respiratory monitoring instrumentation, may be useful in monitoring the physiological status of such patients and assisting in the clinical decision to treat. They show subtle change with negative physiological outcome that is not observed with the current, coarse metric for SDB, the apnoea-hypopnoea index (AHI).