Abstract
Purpose: :
To explore the factors influencing the relationship between refractive error and primary open angle glaucoma
Methods: :
Baseline data from patients enrolled in a trial of surgical management of glaucoma were compared with those from a representative population-based sample of residents of the Tanjong Pagar District, Singapore. An optometrist performed a refraction on all subjects. Axial length was assessed using B-mode ultrasound. Anterior chamber depth and central corneal thickness (CCT) were measured using optical pachymetry. Lens opacity was graded after dilation of the pupil using the LOCS III system. Cases and controls were classified using standard criteria. Glaucoma cases had an intraocular pressure >/= 21 mmHg at least once. Gonioscopy was used to confirm an open angle in all cases.
Results: :
There were 125 cases of POAG (mean age: 62.3 years, 94:31 M:F) and 860 controls (58.2 years, 384:476 M:F) with full refractive and biometric data. Glaucoma cases were significantly older (P< 0.001) than controls. POAG patients were also more myopic than controls (Median -1.15D vs. +0.125D, P< 0.001 Mann-Whitney U). Mean axial length was significantly longer in POAG patients (23.87mm) than in controls (23.19 mm, t test P< 0.001). Mean CCT was also greater in POAG than controls (563um vs 541um, P< 0.001). Seventy one cases were using topical (N= 47) or oral (N= 24) carbonic anhydrase inhibitors (CAI’s). No difference in CCT was identified between cases using topical CAI’s and those not (both 552um, P= 0.98), nor between cases using oral CAI’s and not (562um used, 550um not used, P= 0.21). In a multiple logistic regression model including only age, sex and refraction, older age, male sex and myopic refraction were all associated with an increased risk of POAG (P< 0.001 all). However, following an iterative modeling process, and after controlling for lenticular nuclear opacity (LOCS NO, P= 0.04), thicker corneas, longer axial length, older age and male sex remained significant (all P< 0.001), but myopic refraction was no longer significantly associated with increased risk of POAG (P = 0.33).
Conclusions: :
The previously identified relationship between myopia and POAG was confirmed. This relationship is driven by variation in axial length. Ocular biometry may be a superior method of risk profiling in POAG than assessment of refraction.
Keywords: clinical (human) or epidemiologic studies: risk factor assessment • refraction • optic disc