In this population-based study, randomly selected non–highly myopic eyes with an axial length of less than 25.0 mm had a mean horizontal and vertical diameter of the BMO of 1.54 ± 0.17 mm and 1.67 ± 0.17 mm, respectively. Beyond an axial length of 26.0 mm, both diameters increased (
P > 0.001) linearly with axial length (
Figs. 2,
3). The axial elongation–associated increase was slightly more marked for the horizontal diameter than for the vertical diameter, as indicated by an increase in the ratio of horizontal-to-vertical BMO diameter with longer axial length (
Fig. 4). The horizontal BMO diameter minus the horizontal gamma zone width represented the effective opening of BM for the passage of the retinal nerve fibers and retinal vessels. Its mean horizontal diameter decreased with longer axial length in the group of eyes with an axial length of more than 24.0 mm (
Fig. 5). An overhanging of BM into the intrapapillary compartment was observed mainly in the region from the superior to the nasal optic disc pole. The mean length of the overhanging BM of 263 ± 202 μm was significantly correlated with the mean width of parapapillary gamma zone in the location opposite the overhanging BM. As a corollary, the location of the widest intrapapillary BM overhanging was spatially correlated with the location of the widest gamma zone. The largest width of gamma zone was more strongly associated with a longer horizontal BMO diameter (
P < 0.001; β: 0.46) than it was with a larger BM overhang on the side opposite the largest gamma zone location (
P = 0.006; β: 0.14), after adjusting for longer axial length. A higher prevalence of patchy macular atrophies or macular BM defects as part of a myopic maculopathy was associated with a smaller horizontal BMO diameter within the group of eyes with an axial length of ≥28.0 mm.
The size of the BMO has been determined in only few studies so far. Araie and colleagues
31 examined 258 normal eyes of Japanese subjects and measured a mean BMO area of 2.06 ± 0.45 mm
2. That area of 2.06 mm
2 is quite similar to a BMO area of 2.02 mm
2 for eyes with an axial length of <25.0 mm as examined in the present study, if the BMO diameters are taken for the area calculation of an ellipse. The BMO area for the whole population of the present study was 2.21 mm
2. The difference of the value of 2.02 mm
2 as found for the nonmyopic eyes in our study population was due to the association between larger BMO diameters and longer axial length (
Figs. 2,
3). A similar observation was reported by Nakanishi and colleagues.
32
The BMO diameters increased in the present study with longer axial length in a nonlinear manner. There was only a slight enlargement of the BMO with longer axial length in eyes with an axial length of ≤26.0 mm and a relatively steep and linear increase in eyes with an axial length of >26.0 mm (
Figs. 2,
3). Beyond an axial length of 26.0 mm, the horizontal BMO diameter increased by 0.21 mm (95% CI: 0.16, 0.27) and the vertical BMO diameter by 0.12 mm (95% CI: 0.06, 0.18) for each millimeter of axial elongation. Similar observations were made for the relationship between axial length and optic disc size as measured two-dimensionally upon ophthalmoscopy using optic disc photography.
13,33 In these clinical studies, the optic disc size as measured ophthalmoscopically and the size of parapapillary gamma zone started to increase with longer axial length at a cutoff point of approximately 26.0 to 26.5 mm of axial length or a refractive error of −8 diopters. Such values may be helpful in defining the cutoff point between medium myopia and high myopia.
33,34 The axial elongation–associated increase in the BMO diameters was more marked for the horizontal BMO diameter than for the vertical BMO diameter. Consequently, the ratio of the horizontal-to-vertical BMO diameter increased with axial length in the highly myopic subgroup (
Fig. 5). It may suggest that the myopic elongation of the eye takes place more markedly in the horizontal meridian than in the vertical meridian. Consequently, the effect of the BMO expansion caused by the myopic elongation appears to be more profound in the horizontal meridian than in the vertical meridian. If that notion is valid, future studies may examine whether lacquer cracks as linear defects in BM in eyes with myopic maculopathy may run more in a vertical direction than in a horizontal direction.
34
It was unexpected that the horizontal diameter of the effective opening of the BM for the passage of the retinal nerve fibers and retinal vessels, calculated as the difference of the horizontal BMO diameter minus the horizontal gamma zone width, significantly decreased with longer axial length in the group of eyes with an axial length of >24.0 mm (
Fig. 5). Future studies may address whether this finding is of importance for the discussion about the increased susceptibility of highly myopic eyes for glaucomatous optic neuropathy.
35
The etiology of the parapapillary gamma zone defined as the BM-free parapapillary region has remained elusive so far. The findings obtained in the present study suggest that the development and enlargement of the gamma zone may be due to two mechanisms. If the axial elongation–associated enlargement of the BMO (as reported in previous studies and in this investigation) is more marked than an axial elongation–associated enlargement of the choroidal opening (i.e., the second optic nerve head layer) and an axial elongation–associated enlargement of the peripapillary scleral flange opening (i.e., the third optic nerve head layer or lamina cribrosa), a region outside of the optic disc border (as defined by the peripapillary border tissue of Elschnig) will no longer be covered by BM. Assuming a mostly concentric enlargement of the BMO, it would lead to a circular gamma zone. A second mechanism for the development of gamma zone, in particular at the temporal disc border, may be a shift of the BMO in the direction of the macula. Considering the optic nerve head as a three-layered canal, a temporal shift of the BMO, leaving the choroidal opening and scleral opening behind, would lead to an overhanging of BM into the intrapapillary compartment at the nasal optic disc side and to a lack of BM on the temporal side. Findings in favor of this hypothesis were that the BM overhanging into the intrapapillary compartment was found mostly in the superior to nasal region, that the location of gamma zone was detected mostly in the inferior and temporal region, and that the length of the overhanging BM correlated with the width of gamma zone at the location opposite the overhanging BM (
Fig. 6).
36,37 Recently, an overhanging of BM and the obliqueness of the optic nerve fiber exit in myopic optic nerve heads has also been described and analyzed by Sawada and colleagues.
37 While the hypothesis of a temporal BMO shift has remained unproven as yet, the questions may arise regarding by which mechanism the BMO may be shifted or pushed in the direction of the macula. Recent histomorphometric investigations and experimental studies have suggested that during the process of myopic axial elongation, BM may get elongated or enlarged in the equatorial to retro-equatorial region, and thus the BMO is pushed backward.
38 The underlying choroid and sclera may only passively follow; consequently, the three-layered optic nerve head canal gets an oblique orientation, with a (paradox) course of the retinal nerve fibers running from the macula through the optic nerve head canal in anterior direction before bending backward in the direction of the nasal superior region of the orbit.
It is interesting that a higher prevalence of macular BM defects was associated with a less markedly increased horizontal BMO diameter within the highly myopic group of eyes with an axial length of ≥28.0 mm (
Fig. 7). Macular BM defects have recently been described as a hallmark of myopic maculopathy.
39 Although the present study has only a cross-sectional design,
Figure 7 also suggests that the axial elongation was associated first with an enlargement of the BMO, followed by the development of macular BM defects. One may speculate that the process of myopic axial elongation with an enlargement of the posterior half of the globe leads to an increased strain within BM, so first the physiological hole in the BM, that is, the BMO, enlarges before additional defects in the macular region develop. If the enlargement of BMO is not sufficient to reduce the strain within the BM, macular defects may develop. This notion fits with the observation that within the subgroup of highly myopic eyes, eyes with a macular BM defect as compared to eyes without macular BM defects had a smaller BMO (
Fig. 7). It agrees with the findings made in a cross-sectional hospital-based study on highly myopic eyes in which the number of macular BM defects increased after an increase in the size of parapapillary gamma and delta zone.
40 It also fits with findings that BM at the posterior pole neither elongates nor gets thinner in axially elongated eyes, that it is not elastic, and that it has a relatively high biomechanical strength.
41,42
Limitations of our study should be discussed. First, the participants of our study were not fully population-based recruited because they represented subgroups of the population-based Beijing Eye Study. This opens the possibility of a selection bias, in particular in direction of the highly myopic group. We therefore presented the measurements of the BMO diameters separately for the non–highly myopic subgroup. Second, it can be difficult to delineate the end of BM on the OCT images in eyes with the gamma zone, so this inaccuracy might have increased the noise of the measurements. Despite such a noise, however, the associations were statistically significant; thus, this potential limitation may serve to strengthen the conclusions drawn. Third, in a similar manner, the determination of the length of the overhanging BM into the intrapapillary compartment can be difficult in some eyes since the overlying tissue may prevent a clear visualization. Fourth, the study population included only five eyes with macular BM defects. Although the difference in BMO size between the highly myopic eyes without macular BM defects versus those with macular BM defects was statistically significant (
P = 0.02), the small number of eyes with macular BM defects may not allow a firm conclusion of, but may only hint at, such a correlation of larger BMO size and lower number of macular BM defects (
Fig. 7). Fifth, an inclusion criterion of our study was the detectability of the BMO edges, thus eyes with indiscernible BMO edges were excluded.
43 It may therefore remain unclear whether the results of our study can also fully be transferred to eyes with indiscernible BMO edges.
In conclusion, the horizontal and vertical BMO diameter increased by 0.21 and 0.12 mm, respectively, for each millimeter of axial elongation in eyes with an axial length of >26 mm. This BMO enlargement led predominantly to the development and enlargement of the parapapillary gamma zone, while the horizontal diameter of the effective BMO opening for passage of retinal nerve fibers and retinal vessels got smaller with longer axial length. The parapapillary gamma zone may develop by a temporal BMO shift in medium myopic eyes and by BMO enlargement in highly myopic eyes. A large gamma zone may potentially be protective against macular BM defects in highly myopic eyes.