Prevalence of Obstructive Sleep Apnoea in Glaucoma: the POSAG Study.

Abstract

Purpose : Obstructive sleep apnoea (OSA) has been implicated in the pathophysiology of primary open-angle glaucoma (POAG) but its prevalence in this population is unknown.
The purpose of this study was to establish the prevalence of OSA among patients with POAG and determine whether it is higher than in matched people without glaucoma. We also aimed to examine potential associations between markers of OSA and POAG severity.

Methods : Unselected patients with a diagnosis of POAG and their spouses without glaucoma (controls) were invited. A comprehensive ocular examination (visual field tests, ocular coherence tomography, slit-lamp examination, intraocular pressure, optic disc photography) was performed to confirm or exclude glaucoma. A medical interview focused on sleep history and relevant co-morbidities was conducted.
All patients underwent nocturnal multi-channel respiratory polygraphy. OSA was diagnosed based on apnoea-hypopnoea index (AHI)≥5.
Propensity score matching was used to match the groups for any significant imbalances.

Results : Of 795 potentially eligible subjects, 403 participants: 240 POAG patients and 163 controls were enrolled. The prevalence of OSA was 59.2% (95% CI:53-65%) in POAG patients and 55.2% (95% CI:48-63%) in controls. 22.9% (95% CI:18-28%) POAG patients and 16.6% (95% CI:11-22%) controls were diagnosed with moderate to severe OSA (AHI≥15). Among significant predictors of OSA age and sex differed between the groups (Table1).
163 subjects in each group were successfully matched. There was no significant difference in OSA prevalence between the matched groups (p=0.74 for AHI≥5, p=0.39 for AHI≥15). The median AHI was 5.9 (IQR 2.2-11.8) in the POAG group and 5.5 (IQR 2.6-11.8) in the control group (p=0.65). The level of sleepiness assessed by Epworth score was not significantly different (6.3+/-3.9 vs 6.1+/-3.7, p=0.91).
In a linear regression, AHI was not a significant predictor of Mean Deviation (p=0.93), Visual Field Index (p=0.85) or Retinal Nerve Fibre Layer Thickness (p=0.27) in people with POAG. Nor were other markers of OSA severity (desaturation index, mean saturation, time with saturation<90%) correlated with the glaucoma metrics.

Conclusions : This study confirms high prevalence of OSA in people with POAG which is however not higher than in people without glaucoma who otherwise share the same risk factors for OSA. Our findings do not support the hypothesised association between these two conditions.

This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.