Abstract

Background: Previous studies have yielded inconsistent findings regarding the association between OSA and incident venous thromboembolism (VTE). More specifically, the impact of OSA on unprovoked VTE was barely evaluated.

Aim: We aimed to evaluate whether apnea hypopnea index (AHI) and nocturnal hypoxemia markers were associated with unprovoked VTE incidence.

Methods: Data from the Pays de la Loire cohort were linked to health administrative data to identify incident unprovoked VTE in patients suspected for OSA and no previous VTE disease. Cox proportional hazards models were used to evaluate the association of unprovoked VTE incidence with AHI and nocturnal hypoxemia markers including the time spend under 90% of saturation (T90), oxygen desaturation index and hypoxic burden.

Results: After a median [inter-quartile range] follow-up of 6.3 [4.3-9.0] years, 104 of 7,355 patients developed unprovoked VTE. The comparison of patients with and without incident VTE showed significant differences for AHI and all nocturnal hypoxemia markers. However, in a fully adjusted model, T90 was the only independent predictor of VTE (HR 1.06; 95%CI [1.01-1.02]; p=0.02). The highest tercile of T90 (>6%) was associated with the highest risk of VTE (HR 1.98; 95%CI [1.05-3.74]; p=0.02 versus T90<1%). Sensitivity analysis showed that the association was stronger towards patients non treated by positive airway pressure.

Conclusion: Patients with more severe OSA as measured by the AHI or nocturnal severity markers are more likely to have incident unprovoked VTE. However, solely nocturnal hypoxemia represents an independent risk factor in adjusted analysis.