Pattern electroretinogram abnormality and glaucoma.

To determine the existence of retinal ganglion cell dysfunction and associated risk factors in glaucoma suspects with increased optic disc cupping and normal visual field. Cross-sectional, observational study. Two hundred glaucoma suspect (GS) patients were identified based on optic disc abnormalities (vertical cup-to-disc ratios [C/D]>0.5; vertical C/D asymmetry >or= 0.2; disc hemorrhages; notching) in association with known glaucoma risk factors (positive family history, African American descent, increased intraocular pressure [IOP]), but normal visual fields. Forty-two patients had early manifest glaucoma (EMG). Sixteen normal black subjects were added to update previous pattern electroretinogram (PERG) normative data and to establish a normal control (NC) group with a racial breakdown comparable with that of the study groups. Pattern electroretinograms were recorded simultaneously from both eyes using skin electrodes and automated analysis; visual fields were monitored with standard white-on-white automated perimetry (SAP) central 24-2 program; vertical C/D was evaluated by an independent reader from stereo disc photographs; and univariate and multivariate statistical analysis between PERG and other outcome measures was evaluated. Pattern electroretinogram amplitude (microV), phase (pi rad), and interocular asymmetry in amplitude and phase (%); and SAP mean deviation (MD; decibels), vertical C/D, age (years), IOP (mmHg), and race (black vs. nonblack). The PERG results were abnormal in at least 1 of the outcome measures in 52% of GS patients and 69% of EMG patients. The PERG amplitude was correlated weakly with both MD (P<0.01) and vertical C/D (P = 0.05). The correlation between PERG amplitude and MD and C/D was stronger (P<0.001) for interocular differences rather than absolute measures. Interocular PERG amplitude asymmetry increased with severity of disease (EMG>GS>NC; P<0.01). The PERG amplitude decline with age was steeper in patients with a more negative MD (P<0.01) and in patients with a more negative MD and a larger vertical C/D (P = 0.06). Black race (but not family history) was associated with lower PERG amplitude (P = 0.005) in GS and EMG patients, but not in normal controls (P = 0.44). The correlation between PERG abnormality and known risk factors for glaucoma indicates that PERG has a predictive potential for the development or progression of the disease, or both.