Related Experiment Video
Updated: May 19, 2026
08:03
Midface Hypoplasia and Cranial Base Morphology in Syndromic Craniosynostosis: A Comparative Analysis Study Using a Predictive Regression Model
Published on: November 4, 2025
Influence of Cycloplegia on Axial Length Prediction Models in a Paediatric Sample
Ivo Soares1,2,3, António Baptista4, Oscar Torrado5
1Department of Physics, University of Beira Interior, Covilhã, Portugal. isoares@ubi.pt.
Summary
Accurate axial length (AL) estimation is crucial for monitoring myopia in children. Cycloplegic measurements significantly improve the accuracy and repeatability of AL prediction models, making them a reliable alternative when optical biometers are unavailable.
Area of Science:
- Ophthalmology
- Pediatric Optometry
- Biometry
Background:
- Accurate axial length (AL) estimation is vital for monitoring myopia progression in children.
- Primary care settings often lack access to optical biometers, necessitating alternative methods for AL measurement.
- Prediction models utilizing cycloplegic measurements may offer a reliable solution for AL estimation in pediatric populations.
Purpose of the Study:
- To evaluate the impact of cycloplegia on the accuracy and repeatability of various AL prediction models in children.
- To identify which AL prediction models demonstrate minimal bias under both cycloplegic and non-cycloplegic conditions.
Main Methods:
- Ninety-six children (mean age 12.5 years) had repeated measurements of spherical equivalent refraction (SER), anterior corneal curvature (Kmean), and AL using the Myopia Master, pre- and post-cycloplegia.
- Seven published AL prediction models incorporating SER, Kmean, age, and sex were assessed.
- Performance and repeatability were evaluated using agreement, bias, limits of agreement (LoA), coefficient of repeatability (CoR), and intraclass correlation coefficient (ICC).
Main Results:
- Cycloplegia induced a hyperopic shift (+0.79 D) and improved the repeatability of measured AL (CoR decreased from ~0.14 mm to ~0.09 mm).
- All prediction models showed reduced bias post-cycloplegia, with mean differences ranging from -0.56 to +0.10 mm.
- Models by Morgan et al., Queirós et al., and Lingham et al. exhibited the smallest bias (<0.10 mm) and narrowest LoA (<0.84 mm).
- Spherical equivalent refraction variation accounted for 97-99% of predicted AL change; corneal curvature contributed ≤1.2%.
Conclusions:
- Cycloplegic refraction significantly enhances the accuracy and repeatability of AL prediction models in children.
- The prediction models developed by Morgan et al., Queirós et al., and Lingham et al. demonstrated superior performance.
- These predictive models, particularly with cycloplegic measurements, serve as a valuable alternative in settings lacking optical biometers.
