To test the hypothesis that sleep-related breathing disorder (SRBD) is associated with increasing severity of cardiovascular risk markers.
A cross-sectional study of sleep laboratory patients.
University Hospital Sleep Disorders Centre.
We studied 591 patients referred for a sleep study, all of them without a history of systemic hypertension.
Clinical interview, two unattended sleep studies, and assessment of office blood pressure, cholesterol concentration, alcohol and nicotine consumption and daytime blood gases.
Post-hoc analysis of different cardiovascular risk markers: mean blood pressure, pulse pressure, and the type and grade of systemic hypertension.
Patients were classified as normotensive (blood pressure < 140/90 mmHg, n = 228) or hypertensive (blood pressure ≥ 140/90 mmHg, n = 363) according to office blood pressure measurements. Mixed (systolic and diastolic) hypertension was the most common type of hypertension (n = 182), followed by isolated diastolic hypertension (n = 101), borderline isolated systolic hypertension (n = 70), and isolated systolic hypertension (n = 10). The frequency of mixed hypertension increased with SRBD activity (P < 0.05) and respiratory disturbance index (RDI; the number of breathing disorders per hour of estimated sleep time) was increased in those with mixed hypertension compared with those with normotension (24.8 compared with15.7;t test:P < 0.01). In hypertensive patients classified as having grades 1–3 of hypertension (n = 265, 80 and 18, respectively), there was a progressive increase in RDI (18.9, 27.2 and 30.3, respectively, P < 0.01). Mean blood pressure increased significantly with RDI. Pulse pressure increased significantly with age (P < 0.001), but was unrelated to the degree of SRBD.
We conclude that mean blood pressure and the severity of hypertension, but not pulse pressure, increase with the severity of the SRBD.
Sleep Disorders Centre, Department of Pulmonary Medicine, aSahlgrenska University Hospital, Gothenburg, Sweden, bUniversity of Marburg, Germany.
Received 17 July 2000
Revised 15 October 2000
Accepted 17 November 2000
This study was supported by research grants from the Deutsche Forschungsgemeinschaft (265/3-2), the Bundesministerium für Forschung und Technik (01 KE 8803/8) and the University of Gothenburg.
Correspondence and requests for reprints to Ludger Grote, Department of Clinical Pharmacology, Sahlgrenska University Hospital, 41345 Gothenburg, Sweden. Tel: + 46 31 342 2964; fax +46 31 826723; e-mail: firstname.lastname@example.org