Spatial-Frequency-Dependent Contrast Sensitivity in Early and Intermediate Age-Related Macular Degeneration.

<h4>Objective</h4>To evaluate spatial-frequency-dependent differences in contrast sensitivity across early and intermediate age-related macular degeneration (eAMD and iAMD) and explore the discriminative value of visual function metrics beyond age.<h4>Design</h4>Cross-sectional observational study.<h4>Subjects</h4>Ninety-six participants from the Multimodal Assessment of Visual Function in Intermediate AMD study, including 49 controls, 14 participants with eAMD, and 33 with iAMD.<h4>Methods</h4>Contrast sensitivity was assessed using the quick contrast sensitivity function method at six spatial frequencies from 1 to 18 cycles per degree, with derivation of contrast acuity and area under the log contrast sensitivity function (AULCSF). Additional visual function measures included microperimetry mean sensitivity, best-corrected visual acuity, and low-luminance visual acuity. Linear mixed-effects models evaluated associations between AMD stage, spatial frequency, and contrast sensitivity after adjustment for age- and lens-related covariates. Secondary analyses assessed the discriminative performance of visual function metrics using receiver operating characteristic analyses.<h4>Main outcome measures</h4>Contrast sensitivity across spatial frequencies and discriminative performance of visual function metrics.<h4>Results</h4>In age- and lens-adjusted mixed-effects models, contrast sensitivity differed significantly by AMD stage (<i>P</i> < 0.001), with a significant AMD stage × spatial frequency interaction (<i>P</i> = 0.035). Compared with controls, iAMD showed lower adjusted contrast sensitivity across most spatial frequencies, whereas eAMD differences were most pronounced at intermediate spatial frequencies. Discriminative analyses showed stronger separation of controls from iAMD than from eAMD. Microperimetry mean sensitivity provided the highest numerical discrimination for control versus iAMD and eAMD versus iAMD, whereas high-frequency contrast sensitivity and AULCSF performed best for control versus eAMD. Incremental gains beyond age- and lens-related covariates were modest for eAMD but larger for iAMD and for distinguishing eAMD from iAMD.<h4>Conclusions</h4>Contrast sensitivity differences in AMD are spatial-frequency-dependent and vary by disease stage after adjustment for age and lens-related covariates. Quick contrast sensitivity function-derived metrics showed modest incremental discrimination for eAMD, whereas microperimetry mean sensitivity and selected low-frequency contrast sensitivity measures better distinguished iAMD from controls and eAMD. These findings support spatial-frequency-resolved functional testing as an exploratory measure for AMD characterization; larger, age- and lens-balanced longitudinal cohorts are needed to determine prognostic utility.<h4>Financial disclosures</h4>Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.