Positive Airway Pressure and Long-Term Glaucoma Risk in Obstructive Sleep Apnea: A Cohort Study

To evaluate the association between obstructive sleep apnea (OSA), recorded positive airway pressure (PAP) device use, and incident glaucoma in a large multicenter electronic health record (EHR) cohort.

A retrospective cohort study.

Adults in Epic Cosmos were included if they had an index encounter for OSA evaluation between January 1, 2010, and October 31, 2025, no prior glaucoma diagnosis, and at least 180 days of follow-up. Patients were categorized as sleep-tested controls without OSA, OSA without a PAP device record, or OSA with a PAP device record within 180 days.

Follow-up began at a 180-day landmark after the index OSA evaluation and continued until incident glaucoma, death, or last encounter. Positive airway pressure exposure was defined using a code-based EHR proxy for PAP therapy within 180 days. Cox proportional hazards models adjusted for baseline covariates were used to estimate associations with incident glaucoma. Incidence rates were calculated per 1000 person-years from the landmark to censoring.

Incident glaucoma defined as at least 2 glaucoma diagnosis codes within 365 days, with the first qualifying diagnosis occurring after the 180-day landmark.

The cohort included 12 516 607 patients with a mean (standard deviation) age of 60.4 (11.6) years; 43.7% were women. Over a mean (standard deviation) follow-up of 5.2 (3.7) years, 153 083 incident glaucoma events occurred. Crude incidence rates per 1000 person-years were 1.74 in controls, 2.38 in OSA without a PAP device record, and 3.39 in OSA with a PAP device record. Compared with sleep-tested controls, adjusted hazards of incident glaucoma were higher in OSA without a PAP device record (hazard ratio [HR], 1.27; 95% confidence interval [CI], 1.22-1.32) and OSA with a PAP device record (HR, 2.10; 95% CI, 1.80-2.45). At 10 years, cumulative incidence was 1.58%, 2.30%, and 3.86%, respectively.

Obstructive sleep apnea was associated with a higher risk of incident glaucoma compared with sleep-tested controls. The observed risk gradient does not establish whether PAP itself increases or decreases glaucoma risk relative to untreated OSA because residual confounding by OSA severity and treatment indication is likely. Ophthalmic monitoring may be warranted in patients undergoing clinical evaluation or treatment for OSA.

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.