Meridional anisotropy of visual-evoked potentials and contrast sensitivity in young adults with high astigmatism

Purpose Refractive (or manifest) astigmatism, a common refractive error among the Native American and Asian Chinese population, influences visual development if not appropriately corrected in early life. Unless the circle of least confusion falls precisely at the retina, the orientation-dependent optical feature of astigmatism leads to visual deficits that often degrade one meridian more than the others. This meridional anisotropy has been consistently reported in contrast sensitivity at high spatial frequencies and for grating acuity. However,

previous electrophysiology studies that have primarily determined the

electrophysiological response of the primary visual cortex to a low spatial frequency (3–4 cycles per degree [cpd]) stimulus have yielded conflicting results. This study aimed to investigate the effect of astigmatism on meridional anisotropy using contrast sensitivity (CS) and steady-state visual evoked potentials (ssVEP) over a broad spectrum of spatial frequency.

Methods Thirty-two healthy young adults (18–35 years) with best corrected distance visual acuity (BCDVA) of logMAR 0.0 or better were recruited and divided into two refractive-error groups: (1) Highly astigmatic group (HAS, n = 14): spherical-equivalent error (SE) C - 6.00D, cylindrical error (cyl) C 2.00DC, with negative cylindrical axis 180 ± 20; (2) Non-astigmatic group (NAS, n = 18): SE C - 00D, cyl B 0.50DC. The ssVEP measured the electrophysiological response of the primary visual cortex for sinusoidal gratings oriented horizontally and vertically. Grating spatial frequencies (SF) of 0.6, 1.3, 3, 6, and 12 cpd were used as stimuli. For each grating orientation and spatial frequency, the first harmonic response extracted by a Fourier transform was analyzed. The CS for horizontal and vertical gratings was assessed psychophysically through a spatial four-alternative forced-choice procedure with a 3-down-1-up modified staircase protocol at 0.6, 1.4, 3, 6, and 12 cpd. Both ssVEP and CS were measured with full optical correction.

Results The grating orientation significantly affected the VEP amplitude (ANOVA repeated measures, p = 0.04) and CS (p = 0.002) in the HAS group; both were generally lower for the horizontal than vertical gratings. However, the effect was non-significant in the NAS group (p C 0.17). When comparing the two refractive-error groups, the VEP amplitudes showed significant interactions between refractive error groups and SF for both horizontal and vertical gratings

(mixed model ANOVA, p\0.001), but pairwise comparisons for individual SF were not significant after Bonferroni’s correction. For CS, there was a significant interaction between refractive error groups and SF for horizontal gratings (mixed model ANOVA, p = 0.02) but not for vertical gratings (p = 0.51). Bonferroni’s post hoc test revealed significantly lower CS at 12 cpd in the HAS group than in

the NAS group (p = 0.03). No significant group effect or interaction was found for the vertical grating.

Conclusions Highly astigmatic subjects exhibited significantly lower VEP and CS for horizontal than vertical gratings, even though their BCDVA were clinically normal. In addition, their CS for horizontal gratings at high spatial frequency was significantly lower than that in non-astigmatic subjects. The pattern of meridional anisotropy matches the orientation-dependent optical blur created by their astigmatism when uncorrected.