Evaluation of Variables to Characterize Respiratory Periodicity during Sleep in Older Adults

Abstract

Reliable markers of early neurological decline might guide interventions to prevent or reverse cognitive decline in older adults. Because cognitive decline is associated with hypoxemia during sleep, the authors examined 3 respiratory periodicity variables in 5 older adults. Subjects were monitored overnight using standard polysomnography. From the inductance band signal, the authors calculated the variability in duration of breathing cycles measured by standard deviation of interbreath intervals (sdIBI), frequency of breathing cycles measured by standard deviation of interbreath frequencies (sdIBF), and amplitude of breathing cycles measured by standard deviation of breathing cycle amplitudes (sdAMP). Logistic regression analysis and kappa coefficients identified variables that reliably detected 5-minute segments having central or obstructive apneas or body movements. An sdIBF 4.5 cpm identified body movements (sensitivity= 0.96, specificity = 0.96, kappa = 0.90). An sdIBI > 1.2 seconds identified central apneas (sensitivity = 0.86, specificity = 0.99, kappa= 0.86), and an sdIBI 1.68 seconds identified segments with 3 central apneas (sensitivity = 0.90, specificity = 0.89, kappa = 0.89). An sdAMP 0.1 V and an sdIBF 1.5 cpm identified obstructive apneas (kappa = 0.91). Data support the potential of these variables to identify central and obstructive apneas and to classify individuals according to different patterns of respiratory periodicity.