Between April 2014 and March 2016, 34 children visiting the Tokyo High Myopia Clinic met the criteria of a highly myopic refractive error as defined by the Japanese Ministry of Health and Welfare. One patient was excluded because he was diagnosed with congenital stationary night blindness due to characteristic findings in the electroretinogram. Four patients had unilateral high myopia, and their nonhighly myopic eyes were not included in the study. Another patient was excluded due to parapapillary suprachoroidal cavitation. Out of the remaining 60 eyes of 32 children (mean age: 7.7 ± 4.1 years, mean refractive error: −9.3 ± 3.0 D, mean axial length: 26.5 ± 1.7 mm), 41 eyes of 21 patients showed PDCA and were included in the study group of our investigation. The mean age was 9.4 ± 3.7 years (range, 3–15 years); mean refractive error (spherical equivalent) was −11.5 ± 3.1 D (range, −18.5 to −6.75 D); and mean axial length was 27.5 ± 1.4 mm (range, 24.2–30.5 mm;
Table 1).
Out of 1911 primarily eligible children in the GobiDCES, 1565 (81.9%) children participated in the study, among which 1463 (93.5%) children had EDI-OCT images available for measuring choroidal thickness. Their mean age was 11.8 ± 3.5 years and their mean refractive error was −1.20 ± 2.03 D (
Table 1).
In all 41 eyes of the study group with PDCA, OCT images showed an extreme and abrupt thinning of the temporal parapapillary choroid (
Figs. 1,
2). At 2500 μm nasal to the foveola, CT was <60 μm in 31 of the 41 eyes (76%); ≤50 μm in 25 of the 41 eyes (61%); and ≤25 μm in 11 eyes (27%). In contrast, none of the participants of the GobiDCES, except for the child with PDCA, had a CT as measured at 2500 μm nasally to the foveola of <60 μm (
Fig. 3). Choroidal thickness in the macular region was also thinner (
P < 0.001) in the study group than in the GobiDCES after adjusting for age, refractive error, and corneal refractive power (
Table 2). Thicker CT at 2500 μm nasal to the fovea was significantly associated (
r2 = 0.30) with the control group versus study group (
P < 0.001) after adjusting for younger age (
P = 0.002); higher hyperopic refractive error (
P < 0.001); and higher corneal refractive power (
P < 0.001;
Table 3;
Fig. 3). Thicker CT at 2500 μm temporal to the fovea was significantly associated (
r2 = 0.16) with the control group versus study group (
P < 0.001) after adjusting for higher hyperopic refractive error (
P < 0.001) and higher corneal refractive power (
P < 0.001), while it was not significantly associated with age (
P = 0.17;
Table 4).
The mean ratio of CT at 2500 μm nasal to the fovea to subfoveal CT was significantly lower (
r2 = 0.16) in the study group than in the GobiDCES (
P < 0.001;
β: −0.25;
B: −0.22; 95% CI: −0.27, −0.17) after adjusting for higher refractive error (
P < 0.001;
β: 0.15;
B: 0.08; 95% CI: 0.005, 0.012;
Fig. 4). Age was not significantly associated in that model (
P = 0.30). Mean ratio of CT at 1000 μm nasal to the fovea to subfoveal CT was significantly lower (
r2 = 0.16) in the study group than in the control group (
P < 0.001;
β: −0.26;
B: −0.12; 95% CI: −0.15, −0.10) after adjusting for higher refractive error (
P < 0.001;
β: 0.18;
B: 0.005; 95% CI: 0.003, 0.007). Age was not significantly associated (
P = 0.47).
In contrast, the mean ratio of CT at 2500 μm temporal to the fovea to subfoveal CT was higher (r2 = 0.26) in the study group than in the GobiDCES (P < 0.001; β: 0.36; B: 0.37; 95% CI: 0.31, 0.43) after adjusting for lower refractive error (P < 0.001; β: −0.20; B: −0.01; 95% CI: −0.02, −0.01). Also, the mean ratio of CT at 1000 μm temporal to the fovea to subfoveal CT was higher (r2 = 0.18) in the study group than in the GobiDCES (P < 0.001; β: 0.28; B: 0.13; 95% CI: 0.10, 0.15) after adjusting for lower refractive error (P < 0.001; β: −0.20; B: −0.006; 95% CI: −0.007, −0.004).