A trial of positive airway pressure for the treatment of sleep apnea in pregnancy

BACKGROUND: The pathophysiology of obstructive sleep apnea in pregnancy remains poorly understood and studies examining the effect of treatment with positive airway pressure on pregnancy have been limited.OBJECTIVE: This study aimed to perform a randomized controlled trial of positive airway pressure treatment for obstructive sleep apnea in pregnancy.STUDY DESIGN: Participants with a body mass index >= 30 kg/m2 underwent polysomnography at 14 to 20 weeks' gestation (visit 1) and those with obstructive sleep apnea (apnea-hypopnea index >= 5 but <50) were enrolled. In phase 1, participants were randomized to autotitrating positive airway pressure vs sham positive airway pressure; in phase 2, the sham arm was replaced with a sleep hygiene control. Participants returned at 28 to 31 weeks' gestation (visit 2). The mean arterial blood pressure, uterine artery Doppler pulsatility index, endoglin, soluble FMS-like tyrosine kinase 1 levels, and placental growth factor levels were measured, as well as fasting glucose and insulin to calculate insulin resistance (homeostatic model assessment for insulin resistance). The primary outcome was a composite of the uterine artery Doppler pulsatility index, soluble FMS-like tyrosine kinase 1 to placental growth factor ratio, and the homeostatic model assessment for insulin resistance. For secondary analyses, each outcome variable was ana-lyzed independently. Adherence to treatment was examined.RESULTS: A total of 241 participants completed visit 1, and 89 (37%) had an apnea-hypopnea index between 5 and 50. Of the those, 51 participants were randomized in phase 1 and 38 in phase 2. There was no significant dif-ference in our primary outcome by treatment group. In secondary analyses, the uterine artery Doppler pulsatility index was lower in participants on autoti-trating positive airway pressure when compared with sleep hygiene controls. Otherwise, there were no differences in the mean arterial blood pressure, angiogenic markers, or metabolic markers in phase 1, phase 2, or across the entire study. The overall adherence to autotitrating positive airway pressure therapy was low, but the mean use was greater in phase 2 (0.3+0.6 hours/-night vs 1.3+2.3 hours/night; P=.10). For those on active therapy, fasting glucose values decreased as adherence increased.CONCLUSION: This randomized controlled trial of autotitrating positive airway pressure in pregnancy did not find any differences in a composite pri-mary cardiometabolic risk profile between the treatment groups. Higher autotitrating positive airway pressure adherence was associated with lower fasting glucose levels. The use of a sham positive airway pressure control arm in phase1 may have negatively impacted adherence to active treatment.