Establishment of Myopia Occurrence Prediction Model in Children without Myopia Using Cycloplegic Refraction and Prior Axial Length Change.

Abstract

Predicting myopia onset is critical for the identification of premyopia. This study aimed to develop and validate a myopia occurrence prediction model and further investigated potential premyopia criteria. Five-year prospective study. Four thousand four hundred thirty-seven children 6 to 9 years of age from 12 schools. Five valid follow-up visits were completed from 2018 through 2023. At each visit, uncorrected visual acuity was evaluated using the Early Treatment Diabetic Retinopathy Study chart with tumbling E optotypes (Precision Vision). Cycloplegic autorefraction was performed with a desktop autorefractor (KR8800; Topcon Corp). Cycloplegia was induced with 3 drops of 1% cyclopentolate, instilled 5 minutes apart. Refraction measurements were obtained after complete cycloplegia, which was confirmed by the absence of light reflex and a dilated pupil of at least 6 mm in diameter. Axial length (AL) was measured using noncontact partial-coherence laser interferometry (IOL Master 500; Zeiss). Cycloplegic spherical equivalent refraction (SER) and AL. Baseline SER and prior AL change showed strong predictive value. Minimal improvement was found using prior 2-year changes, and hence the final model incorporated prior 1-year AL change, baseline SER, sex, and parents with myopia. The nomogram exhibited excellent calibration, clinical net benefit, and discrimination, with areas under the receiver operating characteristic curve of 0.879 to 0.941 (1- to 4-year prediction) in the primary cohort and 0.877 to 0.928 (1- to 4-year prediction) in the validation cohort. Calibration plots and decision curve analysis showed good consistency (all Brier scores < 0.25) and clinical value. Based on the model, SER thresholds under various conditions were established to map the risk of myopia occurrence in the future. The model was integrated into an open-source online risk calculator. This prospective study combined cycloplegic refraction with prior AL change to develop and validate a myopia prediction model, establishing potential SER thresholds for premyopia definition. Larger-scale studies across diverse populations are warranted to validate these findings. The author(s) have no proprietary or commercial interest in any materials discussed in this article.