Background/aims Conventional Logarithm of the Minimum Angle of Resolution (logMAR) acuity is the current gold standard for assessing visual function in age-related macular degeneration (AMD). However, visual acuity (VA) often remains ‘normal’ when measured with these charts, even with advanced retinal changes. We wished to investigate how VA measurements with the Moorfields Acuity Chart (MAC), which employs high-pass filtered letters, compares to conventional letter charts in subjects with AMD.

Methods Monocular best-corrected VA measurements and test–retest variability (TRV) were compared for conventional and MAC charts in 38 normal observers (mean age 52.1 years) and 80 patients (mean age 80.6 years) with varying degrees of acuity loss owing to AMD. Methods of Bland–Altman and ordinary least-squares regression were employed for data analysis.

Results A proportional bias was confirmed between conventional and MAC measurements (r²=0.133, p=0.001) such that MAC acuity was −0.45 logMAR ‘worse’ at the 0.00 logMAR acuity level, but only −0.26 logMAR ‘worse’ at the 1.00 logMAR level. The mean bias was much smaller in the normal subject group (−0.16 logMAR). Similar TRV (ranging from ±0.09 to ±0.12 logMAR) was found for both charts in both subject groups.

Conclusions VA measurements with the MAC chart appear to be more sensitive to functional loss in AMD compared with conventional letter charts, with similar TRV. Simulations indicate this may be because the high-pass filtered letters are more vulnerable to undersampling as a result of retinal cell loss in the disease process.