Author's reply.

Abstract

We thank Drs. Saunders and McCullough for their interest in our article, Axial length targets for myopia control. We regret not citing their important work, although it was published only just before our manuscript's submission. McCullough et al. present axial growth data for young emmetropes along with incident and progressing myopes from the Northern Ireland Childhood Errors of Refraction (NICER) study, which they expand upon in their letter to the editor. We are pleased to see that their data support our recent work. In particular, it is encouraging to see that the 3year progression among 10yearold myopes in NICER is close to that observed in the MiSight clinical trial, and in other recent clinical trials of predominantly nonEast Asian children. What is equally valuable are the corresponding NICER data for emmetropes, which also agree with the predictions used in our paper. Large studies of emmetropic children are rare, and it is sometimes unclear if incident myopes are included in growth data. We could have extracted growth rates form the published figures and used them for comparison. The NICER data plotted by McCullough et al. reflect the growth observed at years 3 and 6 of the study. At the start of this period, the NICER cohort had a mean age of 10.1 years, and 13.1 at the 3year mark. Data were averaged for the 6and 7yearold population at baseline and then eye growth over the 3year intervals are reported. Indeed, it is fortuitous that even though the NICER sample was examined once every three years, one of these visits featured children with a very similar mean age to the MiSight cohort. However, we chose a more nuanced approach, creating virtual cohorts reflecting the actual age distribution of our original trial— between 8 and 12 years at baseline— thereby accounting for both the age distribution of subjects and the nonlinear relationship between axial elongation and age. We would not have been able to use the NICER data for this approach. Drs. Saunders and McCullough go on to recommend criteria for clinical decision making about the efficacy of any treatment, with growth of >0.60 mm over 3 years in 10yearolds representing a lack of success. We caution against this blanket application of mean or median data to individual eyes, although 85% of the MiSighttreated eyes elongated by less than this criterion over 3 years. For example, an eye with slow underlying growth may appear to be experiencing successful treatment but may be nonresponsive. The converse is true for an inherently fastgrowing eye, where a 3year growth of 0.60 mm may represent some relative slowing. However, this is where percentile curves, as proposed by McCullough et al. and others should prove valuable to a clinician. Drs. Saunders, McCullough and colleagues are to be applauded for their ongoing and important work. We look forward to additional contributions, including an expansion of the data, perhaps including annualized axial elongation data in emmetropes and persistent myopes.