Low-Dose Atropine Eye Drops Pass a Key Corneal Safety Test in Children With Myopia

Low-concentration atropine eye drops have become one of the most widely prescribed treatments for slowing the progression of childhood myopia, but a persistent question has shadowed their use: what do these drops actually do to the cornea, the transparent dome at the front of the eye that provides most of the eye’s focusing power? A new randomised controlled trial from Western Australia, published in Scientific Reports, offers the most detailed answer yet, and the results are largely reassuring. Researchers led by Alvid Soh and Samantha Sze-Yee Lee of the Centre for Ophthalmology and Visual Science at the University of Western Australia found that two years of nightly 0.01% atropine produced only subtle changes in corneal shape, with no evidence that the treatment increases the risk of corneal ectasia, the family of sight-threatening thinning disorders that includes keratoconus.

The study emerged from the Western Australia Atropine for the Treatment of Myopia trial, known as WA-ATOM, in which 153 children aged between 6 and 16 years with progressive myopia were randomly assigned in a 2:1 ratio to receive either 0.01% atropine eye drops or a placebo. The children used their assigned drops for two years, after which the treatment was stopped and the participants were followed for an additional one-year washout period. To track what was happening at the front of the eye, the team used Scheimpflug imaging, a scanning technique that builds a three-dimensional map of corneal curvature, thickness and elevation with far greater precision than the Placido-based topography systems used in routine clinics. Measurements were taken at baseline, at 24 months, and again at 36 months, allowing the researchers to compare how the corneas of treated and untreated children diverged over time.

The motivation for the study came from earlier reports that low-dose atropine might alter corneal curvature, raising the possibility that some of the refractive changes seen during atropine therapy could be attributable to the cornea rather than to a genuine slowing of axial elongation, the growth of the eyeball that drives myopia. If atropine were steepening the cornea, for example, that would flatten the eye’s optical system in a way that mimics a myopia-control effect without actually changing the eye’s length, a distinction with major implications for how clinicians interpret treatment success. The researchers therefore tested the null hypothesis that 0.01% atropine produces no change in corneal parameters or ectasia risk compared with placebo, using linear mixed models to account for the repeated measurements taken from each child over the three-year study window.

The headline finding concerns a parameter called BAD-D, a sophisticated index derived from Scheimpflug-based tomography that combines multiple measures of posterior elevation, corneal thickness and pachymetric progression into a single score used to flag keratoconus and subclinical ectatic disease. Counterintuitively for anyone worried about atropine, the atropine group showed significantly lower BAD-D values than the placebo group at both 24 and 36 months. Relative to baseline, the reductions were 0.075 at two years (95% confidence interval 0.126 to 0.023, p = 0.027) and 0.115 at three years (95% confidence interval 0.168 to 0.063, p = 0.005), with the between-group differences reaching statistical significance at both time points. Lower BAD-D values point away from ectatic change, meaning the treated corneas looked, if anything, slightly more regular than those of untreated children.

There was one measurable shift in corneal curvature worth noting. At 36 months, the atropine group showed a significant steepening of K-max, the steepest keratometric reading on the corneal surface, of +0.301 dioptres (95% confidence interval 0.214 to 0.389, p = 0.039) compared with placebo. A third of a dioptre is a small optical change, roughly a tenth of the annual myopic progression that atropine is prescribed to prevent, and it appeared only after the full treatment period rather than accumulating progressively. Crucially, every other corneal parameter measured, along with cylindrical power, the component of refractive error attributable to corneal and lenticular astigmatism, remained statistically indistinguishable between the treated and placebo groups throughout the study.

The trial also delivered an unexpected and clinically instructive case report. One child in the treatment group, one out of the 153 enrolled, developed keratoconus during the treatment phase, an incidental diagnosis that surfaced through the study’s systematic corneal imaging rather than through any suspicion that the drops were responsible. Keratoconus is a progressive thinning disorder in which the cornea bulges into a cone shape, distorting vision and, in advanced cases, requiring corneal transplantation. Its prevalence in the general population is low, and the researchers note that finding a single case among 153 children is compatible with expected population rates. The case therefore reads not as a warning about atropine but as a demonstration of why careful baseline corneal screening matters when children are enrolled in long-term eye treatments of any kind.

Putting the numbers in context, the authors conclude that no signal suggesting a treatment-related increase in corneal ectatic changes was identified within the limitations of this cohort. That caveat matters. A trial of 153 children, while large enough to detect meaningful between-group differences in tomographic indices, cannot rule out extremely rare adverse effects, and the single keratoconus case cannot be definitively attributed to chance alone. The three-year follow-up, including a full year after treatment stopped, does strengthen the conclusion considerably, because ectatic progression driven by a biomechanical weakening of the cornea would be expected to continue or accelerate rather than stabilise once a causative agent was withdrawn.

The mechanistic backdrop makes these findings particularly interesting. Atropine blocks muscarinic receptors, and although its myopia-control action remains incompletely understood, the drug is known to affect the ciliary muscle, the retina and, in principle, the extracellular matrix of ocular tissues. Because the cornea’s stroma is richly innervated and biomechanically active, some researchers had hypothesised that chronic muscarinic blockade could soften the corneal structure, mimicking the biomechanical changes seen in early ectasia. The WA-ATOM tomography data argue against that concern at the 0.01% concentration, showing that the biomechanical proxy indices moved in the opposite direction and that curvature changes were confined to a fraction of a dioptre at the steepest point.

For clinicians, the practical takeaways are straightforward. The findings provide reassurance from a safety perspective for the growing number of children worldwide who receive nightly low-dose atropine, a treatment whose popularity has surged as evidence accumulates that it can meaningfully slow myopic progression when started early. At the same time, the authors urge clinicians to be aware of the rare possibility of keratoconus appearing during treatment and to refer any patient with suspicious corneal findings for specialist assessment. Because myopia itself is a known association of keratoconus, and because keratoconus typically emerges during adolescence, children undergoing myopia control are precisely the group in whom vigilant corneal monitoring is most defensible, whether or not they are using atropine.

The study also highlights how modern imaging is quietly transforming paediatric eye care. Scheimpflug tomography, once reserved for refractive surgery screening, can now detect subclinical corneal change years before it affects vision, and its deployment in a randomised trial of myopia therapy shows how safety questions about widely adopted treatments can be answered prospectively rather than through anecdote. As low-dose atropine moves toward broader adoption across Asia, Australia, Europe and North America, trials like WA-ATOM provide the kind of controlled, longitudinally imaged evidence base that regulators and clinicians need. For now, the message to parents is measured but positive: the drops their children take each night appear to leave the cornea essentially untouched, and the one serious corneal condition encountered in the trial was an incidental finding that careful screening caught in time.

Subject of Research: Corneal biometry and ectasia safety of low-dose atropine eye drops for childhood myopia control

Article Title: Corneal biometry changes with 0.01% atropine for myopia control: a randomised controlled trial and case of incidental keratoconus

Article References: Soh, A., Lingham, G., Chan, E., Clark, A., Mackey, D. A., & Lee, S. S.-Y. (2026). Corneal biometry changes with 0.01% atropine for myopia control: a randomised controlled trial and case of incidental keratoconus. Scientific Reports. https://doi.org/10.1038/s41598-026-74892-8

Image Credits: AI Generated

DOI: 10.1038/s41598-026-74892-8

Keywords: atropine, myopia control, cornea, keratoconus, corneal ectasia, Scheimpflug imaging, BAD-D, K-max, randomised controlled trial, WA-ATOM, paediatric ophthalmology, corneal biometry