Incidence of VTE in Patients With OSA: A Cohort Study

Abstract

Previous studies suggesting that OSA may be an independent risk factor for VTE have been limited by reliance on administrative data and lack of adjustment for clinical variables, including obesity. Does OSA confer an independent risk of incident VTE among a large clinical cohort referred for sleep-disordered breathing evaluation? We analyzed the clinical outcomes of 31,309 patients undergoing overnight polysomnography within a large hospital system. We evaluated the association of OSA severity with incident VTE, using Cox proportional hazards modeling accounting for age, sex, BMI, and common comorbid conditions. Patients were of mean age 50.4 years, and 50.1%were female. There were 1,791 VTE events identified over a mean follow-up of 5.3 years. In age- and sex-adjusted analyses, each 10-event/h increase in the apnea-hypopnea index was associated with a 4%increase in incident VTE risk (hazard ratio [HR], 1.04; 95%CI, 1.02-1.06). After adjusting for BMI, this association disappeared (HR, 1.01; 95%CI, 0.99-1.03). In contrast, nocturnal hypoxemia had an independent association with incident VTE. Patients with > 50%sleep time spent with oxyhemoglobin saturation< 90%are at 48%increased VTE risk compared with those without nocturnal hypoxemia (HR, 1.48; 95%CI, 1.16-1.69). In this large cohort, we found that patients with more severe OSA as measured by the apnea-hypopnea index are more likely to have incident VTE. Adjusted analyses suggest that this association is explained on the basis of confounding by obesity. However, severe nocturnal hypoxemia may be a mechanism by which OSA heightens VTE risk.