Cumulative association of obstructive sleep apnea severity and short sleep duration with the risk for hypertension

Abstract

Obstructive sleep apnea (OSA) and short sleep duration are individually associated with an increased risk for hypertension (HTN). The aim of this multicenter cross-sectional study was to test the hypothesis of a cumulative association of OSA severity and short sleep duration with the risk for prevalent HTN. Among 1,499 patients undergoing polysomnography for suspected OSA, 410 (27.3%) previously diagnosed as hypertensive and taking antihypertensive medication were considered as having HTN. Patients with total sleep time (TST) <6 h were considered to be short sleepers. Logistic regression procedures were performed to determine the independent association of HTN with OSA and sleep duration. Considering normal sleepers (TST ≥6 h) without OSA as the reference group, the odds ratio (OR) (95% confidence intervals) for having HTN was 2.51 (1.35-4.68) in normal sleepers with OSA and 4.37 (2.18-8.78) in short sleepers with OSA after adjustment for age, gender, obesity, diabetes, depression, current smoking, use of thyroid hormones, daytime sleepiness, poor sleep complaint, time in bed, sleep architecture and fragmentation, and study site. The risk for HTN appeared to present a cumulative association with OSA severity and short sleep duration (p<0.0001 for linear trend). The higher risk for HTN was observed in short sleepers with severe OSA (AHI ≥30) (OR, 4.29 [2.03-9.07]). In patients investigated for suspected OSA, sleep-disordered breathing severity and short sleep duration have a cumulative association with the risk for prevalent HTN. Further studies are required to determine whether interventions to optimize sleep may contribute to lower BP in patients with OSA.

Figure 1. Multivariate adjusted odds ratio (OR) (95% confidence interval [CI]) for hypertension (HTN) associated with various combinations of sleep-disordered breathing and sleep duration.

OSA, obstructive sleep apnea. OR were adjusted for age, gender, obesity, diabetes, depression, current smoking, use of thyroid hormones, excessive daytime sleepiness, poor sleep, time spent in slow wave sleep, overall arousal index, time in bed, and study site. P<0.0001 for linear trend according to the Cochrane-Armitage trend test.