Sleep Apnea Linked to Higher Incident Glaucoma in Real-World Study

Key Takeaways

• Adults evaluated for OSA showed higher later incident glaucoma than sleep-tested controls without OSA.
• Glaucoma incidence increased stepwise from controls to OSA without a PAP record and was highest in the PAP-record subgroup.
• The PAP-record difference did not establish that PAP therapy itself altered glaucoma risk.

In a real-world cohort of more than 12.5 million adults evaluated for obstructive sleep apnea, later glaucoma incidence was higher in patients with OSA than in sleep-tested controls without OSA. The highest adjusted hazard ratio was 2.10 in patients with OSA who had a PAP device record. A smaller adjusted increase in incident glaucoma was also seen in patients with OSA without a PAP device record.

Investigators identified adults evaluated for obstructive sleep apnea between Jan. 1, 2010, and Oct. 31, 2025, and excluded patients with prior glaucoma. The final cohort exceeded 12.5 million patients, with a mean age of 60.4 years; 43.7% were women. Comparison groups included sleep-tested controls without OSA, OSA without a PAP device record, and OSA with a PAP device record within 180 days of evaluation.

During the 5.2-year average follow-up, investigators identified 153,083 incident glaucoma cases. Crude incidence rates per 1,000 person-years were 1.74 in controls, 2.38 in patients with OSA without a PAP record, and 3.39 in those with a PAP record. Adjusted hazard ratios versus controls were 1.27 for OSA without a PAP record and 2.10 for OSA with a PAP record. The corresponding 95% confidence intervals were 1.22 to 1.32 and 1.80 to 2.45, respectively.

At 10 years, cumulative glaucoma incidence was 1.58% in controls, 2.30% in OSA without a PAP record, and 3.86% in the PAP-record group. The investigators stated that these findings do not establish whether PAP therapy itself influences glaucoma risk. They urged cautious interpretation because patients receiving PAP therapy often have more severe sleep apnea, and residual confounding related to treatment indication likely influenced the subgroup differences.