Vortex Vein Morphology and Choroidal Thickness as Predictors of Refractive Status: Model Construction and Mediation Analysis.

PURPOSE: The purpose of this study was to investigate how vortex vein geometry and choroidal structure relate to refractive status and to develop an internally validated prediction framework based on these posterior segment features. METHODS: A cross-sectional study was conducted on 147 participants, including 49 with myopic eyes, 49 with emmetropic eyes, and 49 with hyperopic eyes. Participants were young adults aged approximately 18 years, and axial length (AL) ranged from 23.19 ± 0.09 mm in hyperopic eyes to 24.92 ± 0.30 mm in myopic eyes. All participants underwent multimodal imaging, including ultra-widefield fundus photography, enhanced-depth imaging optical coherence tomography (EDI-OCT), and optical coherence tomography angiography (OCTA). Temporal-superior (TS) vortex vein distances, subfoveal choroidal thickness (SFCT), AL, and spherical equivalent (SE) were analyzed using least absolute shrinkage and selection operator (LASSO) regression, restricted cubic splines (RCS), multinomial logistic modeling, and mediation analysis. RESULTS: Compared with hyperopic eyes, myopic eyes showed significantly longer AL and thinner SFCT (both P < 0.001). LASSO regression retained TS together with established refractive parameters including SFCT and AL in the final predictive model. RCS analysis demonstrated significant nonlinear associations of AL and SFCT with myopia probability (both P < 0.001). The internally validated nomograms achieved AUCs of approximately 0.99 for myopia and 0.97 for hyperopia classification. Mediation analysis showed that SFCT accounted for approximately 64.7% of the association between TS and SE. CONCLUSIONS: Combining vortex vein morphology with choroidal metrics offers a comprehensive structural framework for refractive assessment and suggests potential utility for refractive risk stratification, pending external validation. TRANSLATIONAL RELEVANCE: These findings translate posterior segment imaging biomarkers into proof-of-concept prediction nomograms for refractive risk stratification, with potential future clinical relevance pending external validation.