Vernier thresholds and alignment bias in control, suspect, and glaucomatous eyes.

Detection of vernier offsets requires integration of spatial information from many retinal receptive fields. It was postulated that diminution of the number of ganglion cells in glaucoma may diminish the spatial sampling grain, compromise these integrative processes, and raise vernier thresholds. To test this hypothesis, vernier thresholds were measured in 45 patients with primary open-angle glaucoma (POAG), 45 age-matched glaucoma suspects, and 45 age-matched controls. Alignment bias was measured in 43 subjects from each diagnostic group. Measurements were also obtained in an additional 29 control subjects to evaluate the effect of age on vernier thresholds. All subjects had acuities of 6/7.5 or better. Vernier thresholds were significantly greater in glaucomatous eyes and suspect eyes than in control eyes (p = 0.0002). Mean thresholds were increased by 64% in glaucomatous eyes and 47% in suspect eyes; however, there was significant overlap between the groups. A subgroup of 15 eyes with early glaucomatous damage without localized visual field defects also had vernier thresholds significantly greater than control eyes (p = 0.02). The variances of the alignment biases were significantly greater in the glaucomatous and suspect eyes than in controls eyes (p < 0.05). Our results demonstrate that significant deficits of vernier detection can be demonstrated in some glaucomatous eyes, even in the absence of defects on automated perimetry. Because of the overlap between diagnostic groups, measurement of vernier thresholds cannot be used to distinguish POAG from controls. However, elevation of vernier thresholds may serve as a marker for the early diffuse central dysfunction that occurs in some eyes with glaucoma. The significance of these findings in terms of the mechanisms of glaucomatous damage is discussed.