Neuronal adaptation to simulated and optically-induced astigmatic defocus

It is well established that spatial adaptation can improve visual acuity over time in the presence of spherical defocus. It is less well known how far adaptation to astigmatic defocus can enhance visual acuity. We adapted subjects to "simulated" and optically-induced "real" astigmatic defocus, and studied how much they adapt and how selective adaptation was for the axis of astigmatism.

Ten subjects with a mean age of 26.7±2.4years (range 23-30) were enrolled in the study, three of them myopic (average spherical equivalent (SE)±SD: -3.08±1.42D) and seven emmetropic (average SE±SD: -0.11±0.18D). All had a corrected minimum visual acuity (VA) of logVA 0.0. For adaptation, subjects watched a movie at 4m distance for 10min that was convolved frame-by-frame with an astigmatic point spread function, equivalent to +3D defocus, or they watched an unfiltered movie but with spectacle frames with a 0/+3D astigmatic trial lenses. Subsequently, visual acuity was determined at the same distance, using high contrast letter acuity charts. Four experiments were performed. In experiment (1), simulated astigmatic defocus was presented both for adaptation and testing, in experiment (2) optically-induced astigmatic defocus was presented both for adaptation and testing of visual acuity. In all these cases, the +3D power meridian was at 0°. In experiments (3) and (4), the +3D power meridian was at 0° during adaptation but rotated to 90° during testing. Astigmatic defocus was simulated in experiment (3) but optically-induced in experiment (4).

Experiments 1 and 2: adaptation to either simulated or real astigmatic defocus increased visual acuity in both test paradigms, simulated (change in VA 0.086±0.069 log units; p<0.01) and lens-induced astigmatic defocus (change in VA 0.068±0.031 log units; p<0.001). Experiments 3 and 4: when the axis was rotated, the improvement in visual acuity failed to reach significance, both for simulated (change in VA 0.042±0.079 log units; p=0.13) and lens-induced astigmatic defocus (change in VA 0.038±0.086 log units; p=0.19).

Adaptation to astigmatic defocus occurs for both simulated and real defocus, and the effects of adaptation seem to be selective for the axis of astigmatism. These observations suggest that adaptation involves a re-adjustment of the spatial filters selectively for astigmatic meridians, although the underlying mechanism must be more complicated than just changes in shapes of the receptive fields of retinal or cortical neurons.