The stereoscopic anisotropy is smaller in elderly population

Abstract

Purpose. The stereoscopic anisotropy is one of the most intriguing phenomena of stereoscopic vision. It shows that the disparity thresholds to detect three-dimensional sinusoidal horizontal corrugations are much lower than for vertical corrugations for spatial frequencies lower than 1 cycles/deg. A recent study has shown that the anisotropy increases during childhood and that visual experience probably plays an important role in its development (Serrano-Pedraza et al., 2016). Here we want to determine the impact that the visual experience has throughout life in the stereoscopic anisotropy. Methods. We performed two experiments testing two age groups of 35 participants each: the young group aged 18 to 45 years and the old group aged 62 to 90 years. We measured disparity thresholds for three-dimensional sinusoidal corrugations of 0.1 cyc/deg, with either vertical or horizontal orientation. Detection thresholds were obtained using Bayesian adaptive staircases. For each participant we computed the anisotropy index by subtracting the thresholds in logarithmic units of the vertical minus the horizontal corrugation. Results. The analyses show that stereo thresholds for vertical corrugations are similar for both groups, however, for horizontal corrugations the thresholds are much lower for the young group. Therefore, the anisotropy was much stronger in the young group (mean, 0.67 ± 0.46) than for the old group (mean, 0.24 ± 0.3). Pearson correlation between the anisotropy index and age shows a negative and significant correlation (r = -0.49; P = 1.83 × 10-5; N = 70), that is, as age advances, the anisotropy decreases. Conclusions. Thus, visual experience plays an important role in the development of stereo vision. Although disparity thresholds for horizontal corrugations in the older group are higher, surprisingly, disparity thresholds for vertical corrugations remain stable and do not change. Therefore, the stereoscopic anisotropy decreases with aging.