Choroidal thickness and peripapillary RNFL in relation to axial length and progression in myopic children

Background: Axial length (AL) elongation is a key structural hallmark of myopia progression in children. Identifying early ocular structural characteristics associated with AL growth may help improve risk stratification, although robust predictive models remain limited. Purpose: To explore cross-sectional associations between baseline ocular structural parameters and AL, and to preliminarily evaluate the predictive potential of these parameters for rapid AL progression in myopic children. Methods: A total of 463 myopic children were retrospectively followed for 1 year. Baseline assessments included cycloplegic refraction, ocular biometry, and optical coherence tomography (OCT)-derived measurements of choroidal thickness (ChT) and retinal nerve fiber layer (RNFL). Rapid AL progression was defined as an AL increase of >0.2 mm over 1 year. Cross-sectional correlations between baseline AL, ChT, and RNFL parameters were analyzed. Least absolute shrinkage and selection operator (Lasso) regression with 10-fold cross-validation was used for feature selection, followed by the development and validation of a logistic regression model. Model performance was assessed using the area under the receiver operating characteristic curve (AUC), calibration analysis, and decision curve analysis (DCA). Results: At baseline, AL showed significant cross-sectional associations with ChT and multiple RNFL parameters, with the strongest correlations observed for ChT and nasal-inferior RNFL thickness (RNFL[NI]). Lasso regression retained age, baseline AL, ChT, and RNFL[NI] as candidate predictors of rapid AL progression. In the validation set, the final model demonstrated modest discriminative performance (AUC = 0.703, 95% CI: 0.598-0.801) with acceptable calibration. Decision curve analysis indicated limited but consistent net clinical benefit across a range of threshold probabilities. Conclusion: This study provides exploratory evidence that posterior segment structural parameters are associated with AL and may be useful for predicting subsequent AL progression in myopic children. Although the model's predictive performance was modest, the findings support the hypothesis that OCT-derived biomarkers could contribute to future risk-stratification strategies, warranting further validation and refinement.