Arousal network determines mean of systolic blood pressure and chemoreflex pathway determines dispersion of systolic blood pressure during N2 sleep in patients of moderate to severe obstructive sleep apnea

IntroductionObstructive sleep apnea (OSA) is characterized by recurrent upper airway obstruction during sleep, leading to intermittent hypoxia, sleep fragmentation, and autonomic dysregulation. These disturbances contribute to nocturnal blood pressure (BP) surges and increased cardiovascular risk. While sleep-stage-dependent BP modulation is well established, high-resolution data on sleep-stage- specific systolic BP variability (BPV) in OSA are limited. This study examined beat-to-beat systolic BPV during N2 and rapid eye movement (REM) sleep and its relationship with sleep fragmentation indices in patients with moderate-to-severe OSA. MethodsClinically suspected OSA patients aged 18-65 years underwent overnight level I video polysomnography. Patients with apnea-hypopnea index (AHI) <15 events/h, central sleep apnea >35%, beta-blocker use, or excessive artifacts were excluded. Continuous systolic BP was estimated using a validated pulse transit time (PTT) method calibrated against cuff BP. Artifact-free 20-minute continuous segments were extracted separately from N2 and REM sleep. Sleep fragmentation metrics included number and duration of respiratory events, arousals, and area under the curve of oxygen saturation (AUC SpO2). BPV indices included mean systolic BP, standard deviation (SD), and coefficient of variation (COV). ResultsSixteen patients contributed 16 N2 and 16 REM segments. N2 sleep showed a higher number of respiratory events (p = 0.005) and arousals (p = 0.01) than REM sleep, while event duration and AUC SpO2 were comparable. Mean systolic BP was 126 {+/-} 12.5 mmHg during N2 and 130 {+/-} 14.9 mmHg during REM, with REM significantly higher than N2 (mean difference -3.62 mmHg; p = 0.01). BP variability was highest during REM (SD 7.12 [4.91-9.25] mmHg; COV 5.95 [3.89-6.84%), intermediate during N2 (SD 5.25 [4.02-6.75] mmHg; COV 4.42 [3.17-4.83%), and lowest during wake (p < 0.001). In N2 sleep, arousal duration predicted mean systolic BP (R{superscript 2} = 0.48, p = 0.0025), while AUC SpO2 strongly predicted SD and COV (R{superscript 2} = 0.74-0.79, p < 0.0001). REM-stage correlations were weaker and not predictive. ConclusionSystolic BP variability in OSA is strongly sleep-stage dependent, with REM sleep exhibiting exaggerated BP instability despite fewer respiratory events. Stage specific mechanisms linking arousals and hypoxia to BP regulation may underlie cardiovascular vulnerability in OSA.