Longitudinal Changes in Refractive Development in Highly Hyperopic Children: A 2.6-11.2 Year Follow-up of Preschoolers Diagnosed with High Hyperopia

PURPOSE: This study aims to elucidate the longitudinal refractive and ocular biometric alterations in preschool children with high hyperopia who underwent early interventions. METHODS: We conducted a retrospective analysis of preschool children diagnosed with high hyperopia at Tianjin Medical University Eye Hospital between 2011 and 2023. Inclusion criteria required an initial examination with cycloplegic refraction, bilateral spherical equivalent power (SE) ≥ +5.00D with a difference <1.00D, a minimum two-year follow-up, and at least three ocular biometric measurements. The annual axial growth rate evaluated emmetropization in highly hyperopic children. We applied Restricted Cubic Spline (RCS) models to explore potential nonlinear relationships between age and spherical equivalent, axial length, corneal curvature, and axial length-to-corneal curvature ratio. Additionally, Mixed-effects models were employed to investigate factors associated with changes in refractive error and axial length. RESULTS: The study enrolled 60 eligible subjects, with a median initial diagnosis age of 3.5 years (IQR, 2.8-4.9 years) and a median last visit age of 9.3 years (IQR, 8.1-10.8 years). The average follow-up duration was 5.7 years. RCS analysis revealed notable nonlinear changes in spherical equivalent power, axial length, and axial length-to-corneal curvature ratio, although corneal curvature displayed no statistically significant nonlinear trend. Factors affecting SE changes included the presence of strabismus, the use of cycloplegia, baseline SE, and age. Conversely, changes in axial length solely correlated with baseline axial length and age. CONCLUSION: Highly hyperopic preschool children undergoing early intervention display a marked emmetropization tendency, though most still remain moderately to highly hyperopic, with the progression of refractive changes showing non-uniform patterns with respect to age.