Study objectives: To evaluate the relationship between sleep structure and continuous positive airway pressure (CPAP) delivered by an automatic CPAP (auto-CPAP) machine in patients with obstructive sleep apnea syndrome (OSAS).
Design: Nocturnal polysomnography was performed during CPAP administration by an auto-CPAP machine (Autoset Clinical 1; ResMed; Sydney, Australia).
Setting: Sleep-disorders center in a research institute.
Patients: Fifteen subjects with newly diagnosed OSAS deserving home CPAP treatment.
Measurements and results: During the night, in most cases, the lowest CPAP level was recorded during a prolonged nonrapid eye movement (NREM) sleep period uninterrupted by arousals, whereas the highest level during wake-sleep transitions or NREM sleep fragmented by arousals. In four subjects, rapid eye movement sleep was always associated with increasing CPAP. Sleep efficiency was negatively correlated with CPAP variability, evaluated as the SD of the mean nocturnal CPAP level averaged epoch by epoch (r = 0.63, p < 0.02). Eighty-eight percent of rapid CPAP augmentations (increases by at least 2 cm H(2)O in less-than-or-equal 2 min) were observed during sleep-wake transitions or after arousals/awakenings (Ar/Aw); 63% of such Ar/Aw were not preceded by any detectable respiratory abnormality.
Conclusions: CPAP levels and variations during auto-CPAP application may be mainly related to sleep continuity and efficiency. The recording of a highly variable pressure during auto-CPAP administration in an unattended environment must raise the question whether the patient's sleep quality was acceptable. A poor sleep quality during an autotitration night could lead to an undesirable overestimation of the CPAP level needed for use with fixed-level CPAP machines.