Th NEW ENGLAND IOURNAL FMEDICINE statistically significant.After adjusu sex.race status,body-mass index,and the presence or ab- even hyper on,t ith stroke or death (hazard 02 ciation of age with the out P-0.003 1 2 4 5 other c ariates.The inclusion of antiplatelet ther Year apy in the model also did notinfluence the adjusted No.at Risk azard ratio for the syndro me(data not shown) between the syndrome and death as a single end point (hazard ratio,2.0;95 percent confidence in- and hypertension resulted in an attenuation of the hazard ratio for apnea (1.7 5 percent con- sis (Table 3 ice in Patients with syndrome crease in the risk of stroke or death from any cause 06 as a function of increased severity of sleep apnea .005).Therisko times that in the controls. 02- P-0.0 DISCUSSION ational cohort study svndrome in the development of a first stroke or lemonstrat onts with and tha the association is independent of other cardiovas- cular and cerebrovascular risk factors,including hyperten with recent dats association was found between the obstructive sleep showing that severe obstructive sle apnea syndrome and strokeordeath fromany caus drome increases the risk of fatal and nonfatal car. (hazard o,2.24;95 percent contdence interv iovascular events.25 The broad range of severityof ons orage an statistically significant.The unadjusted hazard ra- verity of the syndrome and increased risk of stroke tios for smok ng,hypertension,and atrial fibrilla- or death from any cause.This correlation confirms tion were in the expected direction,although not the results of previous cross-sectional data show. 2038 N ENGLJ MED 353:19 WWW.NEJM.ORG NOVEMBER 10,2005 Downloaded from www.neim.org on January 23.2010.Copyright 2005 Massachusetts Medical Society.All rights reserved.n engl j med 353;19 www.nejm.org november 10, 2005 The new england journal of medicine 2038 association was found between the obstructive sleep apnea syndrome and stroke or death from any cause (hazard ratio, 2.24; 95 percent confidence interval, 1.30 to 3.86; P=0.004). The associations of age and diabetes with the composite outcome were also statistically significant. The unadjusted hazard ra￾tios for smoking, hypertension, and atrial fibrilla￾tion were in the expected direction, although not statistically significant. After adjustment for age, sex, race, smoking status, alcohol-consumption status, body-mass index, and the presence or ab￾sence of diabetes mellitus, hyperlipidemia, atrial fibrillation, and even hypertension, the obstructive sleep apnea syndrome retained a statistically sig￾nificant association with stroke or death (hazard ratio, 1.97; 95 percent confidence interval, 1.12 to 3.48; P=0.01). The association of age with the out￾comes also persisted after adjustment, but no sta￾tistically significant association was detected for other covariates. The inclusion of antiplatelet ther￾apy in the model also did not influence the adjusted hazard ratio for the syndrome (data not shown). A significant unadjusted association was found between the syndrome and death as a single end point (hazard ratio, 2.0; 95 percent confidence in￾terval, 1.11 to 3.60; P=0.02). Adjusting for age, sex, race, smoking status, alcohol-consumption status, body-mass index, and the presence or absence of diabetes mellitus, hyperlipidemia, atrial fibrillation, and hypertension resulted in an attenuation of the hazard ratio for sleep apnea (1.70; 95 percent con￾fidence interval, 0.92 to 3.16; P=0.09). A trend analysis (Table 3) revealed a stepwise in￾crease in the risk of stroke or death from any cause as a function of increased severity of sleep apnea (P=0.005). The risk of stroke or death in patients in the most severe quartile of sleep apnea was three times that in the controls. We conducted a large observational cohort study examining the role of the obstructive sleep apnea syndrome in the development of a first stroke or death from any cause. Our results demonstrate that the syndrome is associated with an increased inci￾dence of stroke or death from any cause and that the association is independent of other cardiovas￾cular and cerebrovascular risk factors, including hypertension. Our findings are consistent with recent data showing that severe obstructive sleep apnea syn￾drome increases the risk of fatal and nonfatal car￾diovascular events.25 The broad range of severity of the syndrome in our study population allowed us to demonstrate an association between increased se￾verity of the syndrome and increased risk of stroke or death from any cause. This correlation confirms the results of previous cross-sectional data show￾discussion Figure 1. Kaplan–Meier Estimates of the Probability of Event-free Survival among Patients with the Obstructive Sleep Apnea Syndrome and Controls. 1.0 Probability of Event-free Survival 0.8 0.6 0.4 0.2 0.0 0 1 2 3 4 5 6 Controls Patients with syndrome Year No. at Risk Controls Patients with syndrome 1 3 23 33 88 173 227 452 260 543 266 559 325 697 P=0.003 Figure 2. Kaplan–Meier Estimates of the Probability of Overall Survival among Patients with the Obstructive Sleep Apnea Syndrome and Controls. 1.0 Probability of Survival 0.8 0.6 0.4 0.2 0.0 0 1 2 3 4 5 6 Controls Patients with syndrome Year No. at Risk Controls Patients with syndrome 1 3 23 33 89 179 228 465 262 552 266 562 325 697 P=0.02 Downloaded from www.nejm.org on January 23, 2010 . Copyright © 2005 Massachusetts Medical Society. All rights reserved