Obstructive sleep apnea and glaucoma: What a 12.5-million-patient data set reveals

Does treating obstructive sleep apnea with positive airway pressure (PAP) therapy raise or lower the long-term risk of glaucoma? Nishida and colleagues set out to characterize that relationship in a large retrospective cohort study, drawing on electronic health record (EHR) data from more than 300 million patients. Their findings, published in Ophthalmology Glaucoma, confirm an association between obstructive sleep apnea (OSA) and incident glaucoma, but stop well short of implicating PAP therapy as a cause.¹

A question that has long kept researchers up at night

OSA is a disorder in which the upper airway repeatedly closes off during sleep, driving cycles of oxygen loss and fragmented rest.¹ PAP therapy, including continuous PAP (CPAP), is the standard management approach.² Prior studies have raised concern about CPAP-associated fluctuations in IOP in patients at risk for glaucoma, but most existing work on this question has involved short observation periods, limited patient numbers, and IOP changes as a proxy rather than confirmed glaucoma diagnoses as an end point.³ Glaucoma ranks among the most consequential causes of permanent vision loss globally, driven by a combination of pressure-related, vascular, and inflammatory injury to the optic nerve.⁴ A prior epidemiologic analysis using a large health research network had reported higher glaucoma prevalence among patients with OSA,⁵ but how PAP use relates to glaucoma development over the long term had not been adequately characterized.¹

How the study was built

The retrospective cohort study drew on de-identified EHR data from Epic Cosmos, an aggregated database spanning approximately 1,800 hospitals and health systems. The study was approved by the institutional review board of the University of California, San Diego, and adhered to the tenets of the Declaration of Helsinki. The study window ran from January 1, 2010, through October 31, 2025. Adults aged 40 years or older were included if they had an index encounter for OSA evaluation, no prior glaucoma diagnosis, and at least 180 days of follow-up. Patients were divided into three mutually exclusive groups:

• Sleep-tested controls without OSA or PAP evidence (n = 343,533)
• OSA without a PAP device record within 180 days (n = 12,162,550)
• OSA with a PAP device record within 180 days (n = 10,524).¹

To address immortal time bias, follow-up began at a 180-day landmark after the index OSA evaluation date. Incident glaucoma was defined as at least two ICD-10-CM glaucoma diagnoses within 365 days, with the first qualifying diagnosis occurring after the landmark. Cox proportional hazards models adjusted for demographics, baseline comorbidities, and pre-index healthcare utilization were used to estimate associations. Inverse probability of treatment weighting and sensitivity analyses were also performed.¹

OSA raises glaucoma risk — with or without a PAP device

The final analytic cohort included 12,516,607 patients with a mean age of 60.4 years; 43.7% were women. Over a mean follow-up of 5.2 years, 153,083 incident glaucoma events occurred. Crude glaucoma incidence per 1,000 person-years was 1.74 in controls, 2.38 in OSA without a PAP device record, and 3.39 in OSA with a PAP device record.¹

In adjusted Cox models, both OSA groups carried higher hazard of incident glaucoma relative to sleep-tested controls. The HR was 1.27 (95% CI, 1.22-1.32; P < .001) for OSA without PAP, and 2.10 (95% CI, 1.80-2.45; P < .001) for OSA with a PAP device record. At 10 years, cumulative glaucoma incidence was 1.58% in controls, 2.30% in OSA without PAP, and 3.86% in OSA with PAP. Despite these elevated hazard ratios, the authors note that absolute risk differences were modest at 5 years and remained within a few percentage points at 10 years.¹

Sensitivity analyses were broadly consistent with the primary findings, though effect estimates were attenuated when the cohort was restricted to participants with higher baseline healthcare utilization or those already engaged in ophthalmic care.¹

Why the PAP finding is not what it seems

The authors are explicit that the higher hazard observed in the PAP-record group should not be read as evidence that PAP therapy itself increases glaucoma risk. Because PAP is preferentially initiated in patients with more severe or symptomatic OSA, the elevated hazard in that group likely reflects underlying disease severity, comorbidity burden, and treatment indication patterns rather than any direct harmful effect of the therapy itself. The authors state clearly that these findings do not support withholding PAP therapy on the basis of glaucoma risk.¹

The persistence of elevated risk across both OSA groups—treated and untreated—is consistent with the possibility that OSA contributes to glaucoma through mechanisms that go beyond IOP. The authors note that hypoxia-driven vascular injury to the optic nerve has been proposed as one such pathway, and that RNFL changes have been documented in OSA patients in the absence of chronically elevated IOP.¹ Because the EHR data set lacked granular ocular measurements including IOP values, RNFL thickness, and visual field data, the study could not evaluate which mechanistic pathway—IOP-related, non-IOP, or both—underlies the observed associations.¹

What the data cannot yet tell us

The authors identify several key limitations. Ocular phenotyping was incomplete in the EHR, and glaucoma was defined by diagnosis codes rather than clinical examination, introducing potential misclassification. OSA severity measures, including the apnea-hypopnea index, were unavailable, and residual confounding by severity is plausible. Objective CPAP adherence data were not captured. PAP exposure relied on device and supply records, which may incompletely reflect actual therapy. The PAP-record group was substantially smaller than the untreated OSA group, increasing susceptibility to selection effects.¹

A call for closer ophthalmic attention in OSA patients

Nishida and colleagues conclude that OSA carries an elevated glaucoma risk relative to a sleep-tested population, whether or not a PAP device record is present, and that clinicians managing patients with OSA should be aware of this association. Rather than pointing to PAP as a driver of that risk, the authors frame their findings as a signal that closer eye care follow-up may be appropriate for this patient population. They call for studies that bring together objective measures of OSA severity, real-world adherence data, and detailed ocular assessments to better understand what is driving the risk gradient seen here.¹

References

1. Nishida T, Mittal R, Weinreb RN, et al. Positive airway pressure and long-term glaucoma risk in obstructive sleep apnea: a real-world cohort study. Ophthalmol Glaucoma. doi:10.1016/j.ogla.2026.05.012

2. Patil SP, Ayappa IA, Caples SM, Kimoff RJ, Patel SR, Harrod CG. Treatment of adult obstructive sleep apnea with positive airway pressure: an American Academy of Sleep Medicine Clinical Practice Guideline. J Clin Sleep Med. 2019;15(2):335-343. doi:10.5664/jcsm.7640

3. Cheong AJY, Wang SKX, Woon CY, et al. Obstructive sleep apnoea and glaucoma: a systematic review and meta-analysis. Eye (Lond). 2023;37(15):3065-3083. doi:10.1038/s41433-023-02471-6

4. Weinreb RN, Aung T, Medeiros FA. The pathophysiology and treatment of glaucoma: a review. JAMA. 2014;311(18):1901-1911. doi:10.1001/jama.2014.3192

5. Vasu P, Wagner IV, Lentz PC, et al. Obstructive sleep apnea as a potentiator of primary open-angle glaucoma and necessity for interventional therapy. Ophthalmol Glaucoma. 2025;8(6):553-559. doi:10.1016/j.ogla.2025.05.005