Preliminary Evaluation of Interocular Agreement and Biometric Associations of Cone Photoreceptor Metrics in Healthy Young Adults Using Confocal Retinal Imaging

This study evaluated the interocular agreement and within-eye repeatability of cone photoreceptor metrics measured using a confocal cone analysis system in healthy young adults, as well as their associations with ocular dominance and axial length. This cross-sectional study included 94 eyes from 47 healthy participants. Cone density and cone diameter were measured from fovea-centered 3 × 3 mm images obtained using the MicroClear CRO system (MicroClear Biomedical Co., Ltd., Suzhou, China). Magnification-corrected and original uncorrected cone measurements were analyzed. Interocular agreement was assessed using paired comparisons, intraclass correlation coefficients (ICC), and Bland–Altman analysis. Within-eye repeatability was evaluated from two repeated measurements per eye using the within-subject standard deviation, coefficient of repeatability (CoR), ICC, and Bland–Altman analysis. Associations with axial length were examined using Spearman correlation and generalized estimating equation (GEE) models. No significant interocular differences were observed for magnification-corrected cone density (mean difference, −763 cells/mm²; 95% confidence interval [CI], −2,799 to 1,273; p = 0.455; ICC = 0.679) or cone diameter (mean difference, 0.07 μm; 95% CI, −0.12 to 0.27; p = 0.455; ICC = 0.359). After exclusion of four measurements with automated segmentation failure, within-eye repeatability was good for both cone density (CoR = 4,937 cells/mm²; ICC = 0.924) and cone diameter (CoR = 0.40 μm; ICC = 0.941). Cone photoreceptor parameters were also similar between dominant and nondominant eyes. At the participant level, magnification-corrected cone density showed a moderate negative correlation with axial length (Spearman ρ = −0.545; p < 0.001), and this association remained significant in the adjusted GEE model (β = −5,914.39 cells/mm²/mm; 95% CI, −8,567.73 to −3,261.06; p < 0.001). In contrast, axial length was not significantly associated with uncorrected cone density in either the participant-level correlation analysis (Spearman ρ = −0.121; p = 0.419) or the adjusted GEE model (β = −1,798.92 cells/mm²/mm; 95% CI, −4,104.46 to 506.62; p = 0.126). These findings indicate good within-eye repeatability after exclusion of segmentation-failure measurements, while the association between axial length and cone density was sensitive to the magnification-correction procedure and should therefore be interpreted cautiously.