Peripheral contrast discrimination in high myopia

Purpose: Myopic retinal stretching enlarges the space between retinal cells and increases the critical area for spatial summation, potentially leading to a remodeling of neural contact. Thus, this study aimed to investigate the effect of high myopia on contrast discrimination at the fovea and peripheral retina.

Methods: Thirty-six young participants (aged 19 to 41 years), including 17 high myopes (HM, spherical equivalent: -15.00 to -8.88D) and 19 emmetropes (EM, spherical equivalent: -0.63 to 1.00D), were recruited. Monocular threshold versus pedestal contrast (TvC) curves were generated at the fovea, 10° and 20° temporal retina across eight pedestal contrasts (0.1%, 0.32%, 1%, 3.2%, 5.6%, 10%, 17.8%, 32%). The contrast increment threshold at each pedestal level was determined from three-down-one-up staircases in a two-interval-forced-choice paradigm. Vertical Gabor patches (2 cycles per degree) generated by Psykinematix were used as the visual stimuli. Contrast detection thresholds (i.e., 0% pedestal contrast) were also measured for normalizing the TvC curves. The contrast increment and detection thresholds for the fovea and 10° temporal retina were tested at 2 meters and 20° temporal retina at 1 meter. The slopes of the TvC curve from 3.2% to 32% pedestal levels in log-log coordinate and the normalized contrast increment thresholds were compared between the two groups.

Results: The normalized foveal TvC curves of HM and EM were highly overlapped, and both reached their trough at approximately 1% normalized pedestal contrast, i.e., when the tested pedestal contrast was close to the detection threshold. In contrast, the peripheral TvC curves of HM were slightly elevated, especially at 0.1%, 0.32%, and 1% pedestal contrast tested at the 20° temporal retina. The difference in increment threshold at 1% pedestal contrast reached a significant level (median in HM vs. EM, 1.98% vs. 1.24%, p=0.035). No significant between-group difference in the slopes of the TvC curve were found at the fovea (median in HM vs. EM, 0.82 vs. 0.81, p=0.60), 10° (0.78 vs. 0.71, p=0.14), or 20° retina (HM vs. EM, 0.72 vs. 0.72, p=0.32).

Conclusions: Opposing to our hypothesis, retinal stretching in HM did not affect the foveal contrast discrimination. However, it appeared to diminish the facilitation effect of the TvC function in the peripheral retina, which might be due to the alteration in spatial pooling at low contrast.