Crowding in simulated monovision

Abstract

Purpose: Monovision is the presbyopic vision correction practice of prescribing distance vision in one eye and near vision in the other eye. With blurred image in one eye and clear image in the other, the visual system experiences more binocular masking/suppression/rivalry. About 20%-30% patients with monovision have reported discomfort, indicating certain impairment in visual function. The largely normal visual acuity, which represents the foveal visual function, indicates periphery visual disturbance in those patients. Crowding refers to the reduced object identification ability in the periphery when the target object is flanked by distracters. It is a binocular process and other binocular processes that are ongoing simultaneously could interfere to amplify the impairment to object identification. In this study, we tested whether crowding is much more severe in simulated monovision condition.

Methods: 20 subjects participated in this study. Cycloplegia was induced by instillation of one drop of 1% tropicamide and an artificial pupil with 3mm diameter was applied. Monovision was simulated by full correction in one eye and correction with +2.5 D defocus in the other eye. The crowding effect was quantified by reduction in accuracy when identifying letters (5^° eccentricity from the fixation) with flankers at 5 different spacing (0.44^°, 0.88^°, 1.32^°, 1.76^°, 2.4^°) as compared with no flanker. For each subject, the crowding effect was measured both in normal and simulated monovision condition. Accuracy of letter identification and area of error identification were calculated. Z-test was applied to test if the accuracy of letter identification and area of error identification changed significantly in simulated monovision condition.

Results: The accuracy of letter identification, when tested without flankers, was similar in normal and simulated monovision condition. However, with flankers, the accuracy of letter identification was significantly lower in simulated monovision condition at all spacing (P < 0.05), and the changes of accuracy between them at 5 different spacing were significantly different. (P < 0.05) The area of error identification was significantly higher in simulated monovision condition (3.32 ± 0.47) as compared with normal condition (2.31 ± 0.59). (P < 0.05)

Conclusions: The crowding was much more severe in simulated monovision condition than normal viewing condition, and the difference was dependent of the spacing between the target and flankers.