We would like to thank Mr Tariq and Dr Mitchell for their interest in our article. As noted in their letter, preterm children in our study were more likely to be myopic and to have more myopia than the full-term controls. They invited us to discuss how myopia may have impacted our measurements of retinal nerve fiber layer (RNFL) thickness.
A correlation between refractive error and RNFL thickness has been demonstrated by many authors, particularly the observation that longer axial length is associated with thinner RNFL. Two explanations have been proposed to explain the correlation. First, the diameter of the optical coherence tomography (OCT) scan circle projected onto the retina may be larger because of increased axial length, resulting in the measurement of RNFL at a greater distance from the optic disc margin. Alternatively, stretched, thinner peripapillary RNFL may be present in longer, myopic eyes. Neither of these is applicable to the preterm cohort that we studied.
Unlike school-age myopia, myopia in preterm children is associated with short or normal axial length, thick lens, shallow anterior chamber, and small radius of corneal curvature (Yanni SL, Leffler JN, Birch EE. IOVS 2012;53:ARVO E-Abstract 5854). In our preterm cohort, cycloplegic refractive error was not significantly correlated with global or any sector RNFL thickness (0.02 < r < 0.2, P > .4 for all correlations). We were able to measure axial length using the Lenstar LS900 in 11 of the children in the preterm cohort and 20 of the children in the full-term cohort. Mean (± SD) axial length was 22.5 ± 1.6 mm in the preterm group and 23.3 ± 0.8 mm in the full-term control group—a trend for the preterm infants to have shorter axial length ( P = .07). There was not a significant correlation between axial length and global or any sectoral RNFL thickness (−0.2 < r < 0.3, P > .3 for all correlations). In Figure 2 of our article, we showed the mean RNFL thickness profile for the preterm and full-term groups; we found no significant difference in the angular location of the superotemporal and inferotemporal peaks.
Thus, myopia following preterm birth is associated with abnormal anterior segment development and short or normal axial length, which does not predict thinner RNFL according to either of the predominant hypotheses for the association between RNFL thickness and refractive error. Instead, it is likely that preterm birth is associated with mild optic nerve hypoplasia or mild optic atrophy. We reported that RNFL thickness was 8% lower in preterm compared with full-term children. In addition, optic disc diameter was measured for a subset of children who were sufficiently cooperative to complete an additional spectral-domain OCT test (8 preterm, 36 full-term); optic disc diameter was significantly smaller in the preterm group (mean ± SD: 1369.4 ± 83.8 μm preterm, 1504.7 ± 112.4 μm full-term; P = .003), consistent with hypoplasia.