LAUSR

ABSTRACT This study evaluates the accuracy of an objective two‐wavelength fundus autofluorescence technique for the purpose of measuring the macular pigment optical density (MPOD) in Asian pigmented eyes. Potential differences between MPOD values obtained via autofluorescence technique and subjective heterochromatic photometry (HFP) were examined. Inter‐examiner reproducibility between three examiners and test‐retest reliability over five time points were also explored. Subjects were 27 healthy Japanese volunteers aged 24 to 58 (mean±standard deviation, 40.2±9.0) years. An MPOD module of the Spectralis MultiColor instrument configuration (Spectralis‐MP) was used for the autofluorescence technique, and a Macular Metrics Densitometer (MM) was used for HFP. The mean MPOD values at 0.25° and 0.5° eccentricities using the Spectralis‐MP were 0.51±0.12 and 0.48±0.13, respectively. In comparison, the MM based values were 0.72±0.23 and 0.61±0.25, respectively. High correlations between the Spectralis‐MP and MM instrument were found (Pearson's correlation coefficients of 0.73 and 0.87at 0.25° and 0.5° eccentricities, respectively), but there was a systematic bias: the MPOD values by MM method were significantly higher than those by Spectralis‐MP at 0.25° eccentricity. High inter‐examiner reproducibility and test‐retest reliability were found for MM measurements at 0.5° eccentricity, but not at 0.25°. The Spectralis‐MP showed less inter‐examiner and test‐retest variability than the MM instrument at 0.25° and 0.5° eccentricities. We conclude that the Spectralis‐MP, given its high agreement with the HFP method and due to its higher reproducibility and reliability, is well suited for clinical measurements of MPOD levels in Asian pigmented eyes. HIGHLIGHTSMacular pigment optical density (MPOD) was measured using Heiderberg Spectralis.The accuracy of Spectralis to measure MPOD was evaluated in the Japanese population.Spectralis shows good concordance with heterochoromatic flicker photometry.Spectralis shows good inter‐examiner reproducibility and test‐retest reliability.There is a systematic bias between Spectralis and flicker photometry.