Nocturnal hypoxia is associated with silent cerebrovascular disease in a high-risk Japanese community-dwelling population

Background: Sleep-disordered breathing (SDB) is recognized as a risk factor for cerebrovascular disease. The objective of this study was to investigate the relationship between nocturnal hypoxia and silent cerebral infarct (SCI) in the general population. Methods: In the 2001 annual health check in Nishiarita, Japan, 170 individuals at high risk were screened who met more than three of the following criteria: high blood pressure, hypercholesterolemia, left ventricular hypertrophy by electrocardiography, hemoglobinA(1)c >6.5%, proteinuria, central obesity, heavy smoking habit, heavy drinking, and family history of stroke. Overnight pulse oximetry, brain magnetic resonance imaging, and carotid/ cardiac ultrasonography were performed in 146 (mean age 67.4 +/- 9.0 years) of the 170 individuals in whom pulse oximetry was successfully performed. Results: Subjects were classified into a nocturnal hypoxia group (n = 36) and a nonhypoxia group (n = 110) based on a 3% oxygen desaturation index (ODI) 5.6 times per hour during sleep (highest quartile) by pulse oximetry. The presence of silent cerebral infarct (SCI) (57% v 35%, P =.03) was significantly higher in the hypoxia group than in the nonhypoxia group. The number of SCI was positively correlated with age (r = 0.23, P <.01), systolic blood pressure (r = 0.196, P <.05), and 3% ODI (r = 0.318, P <.001). Even after adjustment for confounding factors using logistic regression analysis, nocturnal hypoxia (odds ratio = 2.2, 95 % confidence interval = 1.10 to 5.30, P =.026) as well as systolic blood pressure and age (10-year increase: odds ratio = 1.22, 95% confidence interval 1.00 to 1.48, P =.048) weie independently associated with SCI in the study subjects. Conclusions: Based on the study results, SDB assessed by overnight pulse oximetry was associated with silent cerebral disease in a high-risk, community-dwelling Japanese population. Am J Hypertens 2005; 18: 1489-1495 (c) 2005 American Journal of Hypertension, Ltd.