Purpose: To evaluate the performance of microperimetry and low-luminance visual acuity for detecting late age-related macular degeneration (AMD) onset.
Methods: Two hundred ninety-two individuals with bilateral large drusen in the Laser Intervention in the Early Stages of AMD study underwent best-corrected visual acuity, low-luminance visual acuity, and microperimetry testing as well as multimodal imaging to detect late (neovascular or atrophic) AMD onset. The performance of the change in the measurement from baseline of each of visual function test for detecting late AMD onset was compared.
Results: The area under the receiver operating characteristic curve for detecting neovascular and atrophic AMD onset was not significantly different for low-luminance visual acuity (area under the receiver operating characteristic curve = 0.71 and 0.56, respectively) and microperimetry (area under the receiver operating characteristic curve = 0.82 and 0.62, respectively) compared with best-corrected visual acuity (area under the receiver operating characteristic curve = 0.57 and 0.56, respectively; P ≥ 0.126 for all). There was also only a fair degree of agreement between the three visual function measures for detecting the onset of neovascular and atrophic AMD (κ ≥ 0.24).
Conclusion: Microperimetry, low-luminance visual acuity, and best-corrected visual acuity demonstrate limited performance for detecting the earliest onset of late AMD. It remains to be established whether they perform better than current methods designed to enable self-detection of neovascular AMD onset, such as Amsler grid testing.
Trial registration: ClinicalTrials.gov NCT01790802.