Defining the critical axial length threshold for transition from physiological growth to pathological fundus...

The dynamic evolution of fundus tessellation density (FTD) and peripapillary atrophy (PPA) during childhood myopia progression-and their threshold-associated transitions relative to axial length (AL)-remain incompletely characterized. To address this gap, we prospectively quantified longitudinal FTD and PPA changes in a school-aged cohort with progressive myopia, and modeled their nonlinear dose-response relationships with AL to identify the biomechanical threshold at which fundus remodeling shifts from physiological adaptation to early pathological tissue deformation. A cohort of 371 children aged 8-11 years was prospectively followed for 2 years. FTD and PPA were measured via a validated deep learning system; myopia progression subtypes were stratified via k-means clustering by annual AL elongation rate; longitudinal trajectories and inflection points were estimated via linear mixed-effects models and quadratic regression. Three progression phenotypes emerged: stable (0.147 mm/y), moderate (0.443 mm/y), and rapid (0.827 mm/y). Each 1 mm AL increase correlated with 0.127 mm² (95% CI 0.106-0.148) PPA expansion (p < 0.001) and 0.011 FTD elevation (p < 0.001). PPA showed J-shaped acceleration with inflection at 23.99 mm; FTD showed U-shaped trajectory, declining before 23.66 mm then rising exponentially. This is the first longitudinal study to identify a ∼24 mm (23.66-23.99 mm) biomechanical threshold for early fundus pathology in children-substantially below the adult high-myopia criterion (>26 mm)-supporting a paradigm shift toward earlier AL-driven risk stratification based on tissue biomechanical susceptibility rather than conventional static refractive cutoffs.