Concordance of macular pigment measurements obtained using customized heterochromatic flicker photometry, dual-wavelength autofluorescence, and single-wavelength reflectance

This study compares invivo measurements of macular pigment (MP) obtained using customized heterochromatic flicker photometry (cHFP; Macular Metrics Densitometer^™), dual-wavelength fundus autofluorescence (Heidelberg Spectralis^® HRA+OCT MultiColor) and single-wavelength fundus reflectance (Zeiss Visucam^® 200). MP was measured in one eye of 62 subjects on each device. Data from 49 subjects (79%) was suitable for analysis. Agreement between the Densitometer and Spectralis was investigated at various eccentricities using a variety of quantitative and graphical methods, including: Pearson correlation coefficient to measure degree of scatter (precision), accuracy coefficient, concordance correlation coefficient (ccc), paired t-test, scatter and Bland-Altman plots. The relationship between max MP from the Visucam and central MP from the Spectralis and Densitometer was investigated using regression methods. Agreement was strong between the Densitometer and Spectralis at all central eccentricities (e.g. at 0.25° eccentricity: accuracy=0.97, precision=0.90, ccc=0.87). Regression analysis showed a very weak relationship between the Visucam and Densitometer (e.g. Visucam max on Densitometer central MP: R²=0.008, p=0.843). Regression analysis also demonstrated a weak relationship between MP measured by the Spectralis and Visucam (e.g. Visucam max on Spectralis central MP: R²=0.047, p=0.348). MP values obtained using the Heidelberg Spectralis are comparable to MP values obtained using the Densitometer. In contrast, MP values obtained using the Zeiss Visucam are not comparable with either the Densitometer or the Spectralis MP measuring devices. Taking cHFP as the current standard to which other MP measuring devices should be compared, the Spectralis is suitable for use in a clinical and research setting, whereas the Visucam is not.