We thank Drs Chen and Zhang for their letter and valuable remarks regarding our recently published article.

The authors pointed out that our description of the study population might contain an inconsistency. In our paper, we included 60 primary open-angle glaucoma (POAG) eyes, which included 34 non–highly myopic eyes and 26 highly myopic eyes, and further included 21 healthy non–highly myopic eyes. Table 2 shows that the number of POAG eyes that were non–highly myopic was 34 in total, which includes 23 superior hemifield visual field (VF) defects and 11 inferior hemifield VF defects. Thus, there is no inconsistency at all.

The authors would like to know the location of the change in retinal microvascular density. Because there were some difficulties in objectively showing the precise pattern of microvascular reduction on optical coherence tomography (OCT) angiography, we decided to use microvascular density in segmented peripapillary areas. Our unpublished data showed that the microvascular reduction in the peripapillary retina highly corresponded to the location of the retinal nerve fiber layer (RNFL) thinning, not beyond it. That is why we speculate that microvascular reduction in the peripapillary retina could occur after RNFL thinning. Because we did not examine the vessel density in the macular area in our study, further studies would be needed to elucidate the association between VF defects and macular microvascular density.

In our paper, the reduction in the circumpapillary RNFL (cpRNFL) thickness was significant not only at the corresponding location of the VF defects but also at the noncorresponding location, while the reduction in the vessel density was limited to the corresponding location. A recent paper showed that peripapillary vessel density was reduced even in glaucoma suspect eyes and OCT angiography vessel density had similar diagnostic accuracy to RNFL thickness measurements for differentiating between healthy and glaucomatous eyes. In any case, there has been no firm evidence that peripapillary vessel density reduces before RNFL thinning so far, and we speculated that peripapillary microvascular reduction identified on OCT angiography represented the secondary loss of capillaries from these findings. However, to answer the cause-and-effect issue between VF and structure, including retinal microcirculation, precisely, longitudinal follow-up studies using OCT angiography, OCT-based structural measurements, and VF examinations would be helpful, as the authors suggested.

Again, we appreciate the valuable comments from Drs Chen and Zhang. Further studies are needed to elucidate some of the outstanding issues.