Author reply – Corrected Proof

We thank Galvis et al for their letter. We also noted in our study that there are no differences in myopia progression when comparing myopic children with or without atropine treatment in both the intervention (recess outside classroom) and control groups. To assess this, we conducted a separate analysis on myopic children alone. The averages ± standard deviation of myopia progression for atropine and non-atropine users were −0.29±0.70 diopters (D)/year and −0.21±0.79 D/year, with a difference of −0.08 D/year (95% confidence interval (CI), −0.34 to 0.18; t test, P = 0.5509) in the intervention group, and −0.32±0.58 D/year and −0.37±0.77 D/year with a difference of 0.06 D/year (95% CI, −0.28 to 0.40; t test, P = 0.7417; , available at http://aaojournal.org) in the control group. The differences between atropine users and non-atropine users were further estimated by multivariate analysis using the same adjusting variables as in the original paper, and the adjusted differences of atropine users and non-atropine users were −0.05 D/year (95% CI, −0.32 to 0.22; P = 0.7358) in the intervention group, and 0.09 D/year (95% CI, −0.26 to 0.45; P = 0.5987) in the control group. Neither the analysis of variance and multivariate analysis on the 4 combinations of intervention and atropine were significant (P = 0.4583 and 0.5346).