Peripapillary Choroidal Thickness in Adult Chinese: The Beijing Eye Study

Abstract

Purpose.: To measure peripapillary choroidal thickness (PPCT) and to assess its associations.

Methods.: The population-based cross-sectional Beijing Eye Study 2011 included 3468 participants. Detailed medical and ophthalmic examinations were performed. We measured PPCT by spectral-domain optical coherence tomography (SD-OCT) with a 3.4-mm scan circle centered on the optic nerve head.

Results.: Peripapillary choroidal thickness measurements were available for 3060 (88.2%) study participants with a mean age of 64.4 ± 9.6 years (range, 50–93 years). Mean global PPCT was 134 ± 53 μm (range, 35–348 μm). Peripapillary choroid was thickest in the superior region (155 ± 60 μm), followed by the temporal region (144 ± 75 μm; P < 0.001); nasal region (139 ± 55 μm; P < 0.001); and inferior region (110 ± 45 μm; P < 0.001). In multivariate analysis, thicker PPCT was associated with younger age (P < 0.001; standardized coefficient β: −0.33; correlation coefficient B: −1.95; 95% confidence interval (CI): −2.25, −1.65); shorter axial length (P < 0.001; β: −0.11; B: −5.39; 95% CI: −7.85, −2.93); smaller parapapillary α zone (P = 0.01; β: −0.06; B: −5.46; 95% CI: −9.73, −1.19); and smaller β zone (P < 0.001; β: −0.14; B: −8.29; 95% CI: −11.12, −5.46); better best corrected visual acuity (logMAR; P = 0.002; β: −0.05; B: −14.75; 95% CI: −28.59, −0.91), and higher prevalence of early age-related macular degeneration (P = 0.04; β: 0.05; B: 9.11; 95% CI: 0.42, 17.80) and intermediate age-related macular degeneration (P = 0.001; β: 0.08; B: 10.90; 95% CI: 4.46, 17.33). It was not significantly (all P > 0.05) associated with blood pressure, blood concentration of lipids, intraocular pressure and prevalence of glaucoma, diabetic retinopathy, and retinal vein occlusions. The decrease of PPCT with longer axial length occurred predominantly in the temporal region.

Conclusions.: Peripapillary choroidal thickness is thickest superiorly and thinnest inferiorly. It decreases by 2 μm per year of life and by 5 μm per diopter of myopia. Thinner PPCT is correlated with larger parapapillary α and β zones. The association of thinner PPCT with lower best corrected visual acuity may warrant further study.

Since Spaide and colleagues¹ described the enhanced depth imaging mode of spectral-domain optical coherence tomography (SD-OCT) to visualize the choroid, studies have been focused on the measurement of the subfoveal choroidal thickness in normal individuals and patients with macular diseases.^2–12 The first investigations that evaluated the relationship between subfoveal choroidal thickness and age and which were performed by Margolis and Spaide² and by Ikuno and colleagues⁵ reported that the subfoveal choroidal thickness decreased by 15.6 or 14 μm, respectively, per decade of life. The Beijing Eye Study revealed that in normal persons, the subfoveal choroidal thickness decreased by approximately 4 μm per year of older age and by 15 μm per diopter (D) of increasing myopia.⁸ Subfoveal choroidal thickness was additionally associated with male sex and deeper anterior chamber. These associations explained why under normal conditions, the subfoveal choroidal thickness, with a mean of approximately 250 μm in emmetropic individuals aged 65 years, could range in a population between a minimum of close to 0 and a maximum of approximately 850 μm.⁸ Interestingly, best corrected visual acuity (BCVA) was associated with thicker subfoveal choroid after adjusting for ocular and systemic parameters underlining the clinical importance of subfoveal choroidal thickness measurements.¹³ Some macular disorders were associated with an abnormally thick subfoveal choroid, such as polypoidal choroidal vasculopathy, central serous chorioretinopathy, Vogt-Koyanaga-Harada's disease and idiopathic choroidal neovascularization, other maculopathies, in particular myopic retinopathy, were characterized by an abnormally thin subfoveal choroid, while a third group of disorders was not associated with an abnormal subfoveal choroidal thickness.^{2–7,9–12,14–16} 

Besides in the foveal area, the choroid has also been measured in the peripapillary region.^17–26 The peripapillary region is of special interest since optic nerve diseases, in particular glaucomatous optic neuropathy, and axial myopia can be associated with peripapillary atrophic changes such as the peripapillary β, γ, and δ zones.²⁷ In addition, the peripapillary choroid receives its blood from the same source (i.e., the posterior ciliary arteries), as does the peripapillary arterial circle of Zinn-Haller that is located in the sclera at the merging point of the optic nerve dura mater with the posterior sclera and which supplies blood to the lamina cribrosa of the optic nerve head.^28,29 Ischemic optic nerve head diseases may thus show abnormalities in the peripapillary choroid, as has recently been demonstrated in a study on patients with nonarteritic anterior ischemic optic neuropathy.³⁰ Since the available investigations on the peripapillary choroidal thickness usually (1) included relatively few subjects; (2) did not assess all potentially confounding parameters such as axial length, age, sex, anthropometric measures and ocular biometric variables; and (3) had a hospital-based recruitment of study participants with the risk of a bias by a selection artifact, we conducted the present investigation to measure the peripapillary choroidal thickness (PPCT) in a population-based study. 

Methods

The Beijing Eye Study 2011 is a population-based cross-sectional study that was conducted in an urban and a rural region of Beijing in Northern China.^31,32 It was approved by the Medical Ethics Committee of the Beijing Tongren Hospital, adhered to the tenets of the Declaration of Helsinki, and informed written consent was obtained from all participants. The only eligibility criterion was an age of 50 years or older. Out of 4403 eligible individuals, 3468 (78.8%) persons with a mean age of 64.6 ± 9.8 years (range, 50–93 years) took part in the survey. 

The examinations started with a detailed interview, which was conducted by trained research staff and which included standardized questions on the family status, level of education, income, quality of life, psychic depression, physical activity, and known major systemic diseases. Fasting blood samples were taken for measurement of blood lipids, glucose, and glycosylated hemoglobin. We also measured blood pressure, body height and weight, and circumference of the waist and hip. For the ophthalmic examination, we carried out a measurement of best-corrected visual acuity, tonometry, slit-lamp based examination of the anterior and posterior segment of the eyes, optical low-coherence reflectometry (Lenstar 900 Optical Biometer; Haag-Streit, Koeniz, Switzerland) for determination of the ocular dimensions; and after medically inducing a pupillary mydriasis, we performed a digital photography of the cornea, lens, macula and optic disc, and SD-OCT (Spectralis; Heidelberg Engineering, Heidelberg, Germany) of the macula and optic nerve head. Using the photographs of the optic nerve head, we determined the size of parapapillary α zone (defined as an area with irregular pigmentation) and parapapillary β zone (defined as an area with visible sclera and visible large choroidal vessels and being located between the peripapillary ring and α zone). The study design and the techniques have been described in detail previously.^31,32 

The peripapillary choroidal thickness was measured using the spectral domain OCT images obtained by an experienced technician (CXC) after medical mydriasis. The images of OCT were taken in the conventional mode. We used a 360°, 3.4-mm diameter peripapillary circle scan (comprising 100 averaged scans and centered on the optic disc) that has also been used for measurement of the peripapillary retinal nerve fiber layer (RNFL) thickness. Using commercial software (Heidelberg Eye Explorer v. 5.3.3.0; Heidelberg Engineering), an ophthalmologist (RJ) manually moved in a masked manner the segmentation line of the inner limiting membrane to the outer border of the choroid, which was defined as the hyperreflective line of the inner surface of sclera, on the basis of files with segmentation of “Retinal” (with the upper line corresponding to the internal limiting membrane and the lower line corresponding to the retinal pigment epithelium/Bruch's membrane complex). If a clear line could not be detected, the outer choroidal border was defined as a smoothing line connecting the outer limits of the large choroidal vascular spaces. The images were then exported, and using image processing software designed for the current study, the distances between the two segmentation lines were measured at eight locations (temporal, superotemporal, superior, superonasal, nasal, inferonasal, inferior, and inferotemporal) each equidistant (45°) to the next location (Fig. 1). The measurements obtained at these eight locations were averaged as global PPCT, the value of which was used for further statistical analysis. Additionally, we divided the peripapillary region into six sectors and measured the PPCT in each of these sectors, adopting the regional stratification as has also been applied for the measurement of the retinal nerve fiber layer thickness in SD-OCT. It included the temporal sector (90° wide); superotemporal sector (45° wide); superonasal sector (45° wide); nasal sector (90° wide); inferonasal sector (45° wide); and the inferotemporal sector (45° wide). The parameter of the quality of the OCT scans was not included in the data file and into the statistical analysis, since we included only those OCT images whose quality was sufficient for a reliable determination of the anterior and posterior borders of the choroid. Subfoveal choroidal thickness was measured using SD-OCT images that comprised 100 averaged scans and were taken using the enhanced depth imaging mode. Out of seven sections of a matrix scan covering a rectangular area of 5° × 30°, the horizontal section running through the center of the fovea was chosen for the measurement of the subfoveal choroidal thickness. Subfoveal choroidal thickness was defined as the vertical distance from the hyperreflective line of Bruch's membrane to the hyperreflective line of the inner surface of the sclera. It was assessed manually by two ophthalmologists as described in detail previously.⁸ Only one eye of each participant was chosen for statistical analysis. 

Figure 1

 

Measurement illustration of peripapillary choroidal thickness at eight locations. (A) Manual delineation of the outer and inner choroid borders using eye tracking software (Heidelberg Engineering). (B) Measurement of the peripapillary choroidal thickness at eight locations using image-processing software.

We used a statistical software package (SPSS for Windows, version 22.0; SPSS, Inc., Chicago, IL, USA) for the statistical analysis. First, we determined the mean value (presented as mean ± standard deviation) of PPCT in the eight locations, before we used the Wilcoxon signed-rank test to carry out multiple comparisons between PPCT at the various measurement locations. We then performed a univariate linear regression analysis with the global PPCT as a dependent parameter and systemic and ocular parameters as independent parameters. Finally, we carried out a multivariate regression analyses with the global PPCT as the dependent parameter and all of those variables as independent parameters that were significantly associated with mean global PPCT in the univariate analysis. All P values were two-sided and considered statistically significant when the values were less than 0.05. 

Results

Reliable PPCT measurements were available for 3060 (88.2%) eyes (women: 1726, 56.4%) of the 3468 participants of the survey. Mean age was 64.4 ± 9.6 years. It was not normally distributed (Kolmogorov-Smirnov-test; P < 0.001). The group of subjects with available PPCT measurements compared with the group of subjects without PPCT measurements was younger (64.4 ± 9.6 years versus 66.3 ± 10.9 years; P < 0.007) and less myopic (−0.16 ± 1.86 D versus −0.87 ± 3.67 D; P = 0.001), while axial length did not differ between both groups (23.25 ± 1.06 mm versus 23.31 ± 1.67 mm; P = 0.51). The main reason for not identifying the posterior boundary of the choroid was a poor image resolution for some eyes. The study included 154 (5.1%) eyes that had undergone cataract surgery. For the statistical analysis of associations of PPCT with lens thickness, these eyes after cataract surgery were excluded. As defined by the criteria set up by the Age-Related Eye Disease Study, 1250 (41.4 %) eyes had cataract.³³ By the selection criterion, in all of these eyes the anterior border and posterior border of the peripapillary choroid could sufficiently be detected to be measured. 

Mean PPCT was significantly (P < 0.001) the thickest in the superior region (155 ± 60 μm), followed by the temporal region (144 ± 75 μm); the nasal region (139 ± 55 μm; P < 0.001); and finally the inferior region (110 ± 45 μm; P < 0.001; Table 1; Fig. 2). Mean global PPCT (134 ± 53 μm) did not differ significantly (P = 0.40) between men (133.2 ± 55.8 μm) and women (134.8 ± 51.1 μm). In addition, the PPCT measurements obtained at the eight different measurement locations did not differ significantly between men and women in multivariate analysis as described below. 

Table 1

 

Peripapillary Choroidal Thickness Measurements (μm) in The Beijing Eye Study 2011

 

Figure 2

 

Line charts of average peripapillary choroidal thickness in eight locations with different ages/axial lengths. (A) Peripapillary choroidal thickness of different regions 1.7 mm away from optic disc center in each age group. (B) Peripapillary choroidal thickness of different regions 1.7 mm away from optic disc center in each axial length group. I, inferior; IN, inferonasal; IT, inferotemporal; N, nasal; T, temporal; S, superior; SN, superonasal; ST, superotemporal.

In univariate analysis, thicker global PPCT was significantly associated with younger age (P < 0.001; correlation coefficient r = −0.40). Since age was physiologically associated with many other parameters, we analyzed associations between PPCT and these other parameters after adjustment for age (Table 2). It revealed that thicker PPCT was related with the systemic parameters of higher blood concentration of glucose (P = 0.03; standardized correlation coefficient β: 0.04), and with the ocular parameters of shorter axial length (P < 0.001; β: −0.18); thicker lens (P < 0.001; β: 0.07); hyperopic refractive error (P < 0.001; β: 0.15); better best corrected visual acuity (measured in logMAR, P < 0.001; β: −0.07); thicker subfoveal choroidal thickness (P < 0.001; β: 0.69); smaller parapapillary α zone (P = 0.02; β: −0.05); smaller parapapillary β zone (P < 0.001; β: −0.20); thicker retinal nerve fiber layer thickness (P = 0.003; β: 0.05); and higher prevalence of early age-related macular degeneration (P = 0.04; β: 0.04) and intermediate age-related macular degeneration (P < 0.001; β: 0.09). It was not significantly associated with region of habitation (P = 0.87); sex (P = 0.20); body height (P = 0.15); body weight (P = 0.95); waist circumference (P = 0.41); body mass index (P = 0.32); systolic blood pressure (P = 0.33); diastolic blood pressure (P = 0.28); arterial hypertension (P = 0.20); blood concentration of high-density lipoproteins (P = 0.69); low-density lipoproteins (P = 0.18); triglycerides (P = 0.11) and cholesterol (P = 0.62); reported snoring (P = 0.67); intake of aspirin (P = 0.28); number of cigarette package years (P = 0.06); quantity of alcohol consumption (P = 0.10). It was also not significantly associated with the ocular parameters of central corneal thickness (P = 0.58); corneal diameter (P = 0.60); anterior chamber depth (P = 0.46); intraocular pressure (P = 0.70); prevalence of glaucomatous optic neuropathy (P = 0.55); diabetic retinopathy (P = 0.09); late age-related macular degeneration (P = 0.07); and retinal vein occlusions (P = 0.23). 

Table 2

 

Associations Between the Mean Global Peripapillary Choroidal Thickness Measurements (μm) and Ocular and Systemic Parameters in The Beijing Eye Study 2011

 

The multivariate analysis included mean PPCT and dependent variable and all parameters as independent variables that were significantly associated with PPCT in the univariate analysis. Due to collinearity, we dropped refractive error and subfoveal choroidal thickness from the list of independent parameters. Due to missing remaining significance of the association with PPCT, we then dropped step-by-step blood concentration of glucose (P = 0.28); retinal nerve fiber layer thickness (P = 0.28); and lens thickness (P = 0.09). Thicker PPCT was finally significantly associated with younger age (P < 0.001); shorter axial length (P < 0.001); smaller parapapillary α zone (P = 0.01) and smaller β zone (P < 0.001); better best corrected visual acuity (P = 0.002); and higher prevalence of early age-related macular degeneration (P = 0.04) and intermediate age-related macular degeneration (P = 0.001; Table 3). If axially highly myopic eyes with an axial length of more than 26.5 mm were excluded, similar results were obtained. 

Table 3

 

Multivariate Analysis of Associations Between Mean Global Peripapillary Choroidal Thickness and Systemic and Ocular Parameters in The Beijing Eye Study 2011

 

If instead of the global PPCT, the PPCT in the various regions was taken for analysis, it revealed that the age-related loss in PPCT took place in a similar amount at all measurement locations (Fig. 2A). The decrease in PPCT with longer axial length occurred predominantly in the temporal region, while the nasal region was almost unaffected (Fig. 2B). 

The mean PPCT as measured in each of the six sectors was 144 ± 69 μm, 155 ± 63 μm, 153 ± 58 μm, 139 ± 53 μm, 117 ± 46 μm, and 113 ± 47 μm as determined in the temporal sector, superotemporal sector, superonasal sector, nasal sector, inferonasal sector, and the inferotemporal sector, respectively. The mean of these six measurements was 138 ± 54 μm. In multivariate analysis, this global PPCT determined as the mean of the measurements obtained in the six sectors showed the same statistical associations with the same parameters as it holds true for the mean global PPCT determined as the mean of the PPCT measured at the eight locations. 

Discussion

In our population-based study population, mean global PPCT was 134 ± 53 μm and ranged between 35 and 348 μm. The peripapillary choroid was thickest superiorly and thinnest inferiorly, while in between, it was thicker in the temporal region than in the nasal region. This pattern of PPCT changed with longer axial length. The decrease in PPCT occurring with axial elongation took place more in the temporal region than in the nasal region. In multivariate analysis, thicker PPCT was associated with younger age, shorter axial length, smaller parapapillary α zone and β zone, better best corrected visual acuity, and higher prevalence of early and intermediate age-related macular degeneration. Peripapillary choroidal thickness was not significantly associated with blood pressure, blood concentration of lipids, intraocular pressure and prevalence of major eye diseases. 

The mean global PPCT of 134 μm agrees with the findings of previous hospital-based studies by Ehrlich et al.¹⁷ and Roberts et al.¹⁸ In a recent study on 76 healthy Chinese subjects with a mean age of 57 ± 13 years and a mean PPCT of 165 ± 40 μm, PPCT was thicker than in our study, also after adjusting for the difference in age.²⁰ Reasons for the discrepancy may be differences in the composition of the study population. In particular, the results of our study agree with the findings from the recent Singapore Malay Eye Study in which the overall mean PPCT was 136 ± 57  μm.³⁴ 

Peripapillary choroidal thickness showed a characteristic pattern, with the thickest PPCT superiorly, followed by the temporal region, the nasal region, and finally the inferior region (Fig. 2). A similar pattern of the PPCT was reported by Chung et al.,³⁵ Vujosevic et al.,²⁶ Ho et al.,²⁵ Tanabe et al.,²¹ and others.²⁰ In the Singapore Malay Eye Study, PPCT was thickest in the superior quadrant (151 ± 60  μm), followed by the nasal (144 ± 58  μm) and temporal quadrants (139 ± 69  μm), and thinnest in the inferior quadrant (111 ± 52  μm).³⁴ Interestingly in the study population examined by Oh et al.,²² the mean PPCT was 191 ± 62 μm, and it was thickest at the temporal quadrant (210 ± 78 μm), followed by the superior region (202 ± 66 μm); nasal region (187 ± 64 μm); and the inferior quadrant (152 ± 59 μm). The reason for the regional distribution of the PPCT has remained elusive so far. It may be discussed that the location of the fissure in the nasal inferior region of the embryonic optic cup may have an influence on the eventual PPCT in the inferior region. 

Peripapillary choroidal thickness was not significantly associated with the presence of glaucoma in our study after adjusting for age, axial length, parapapillary α zone and β zone, visual acuity, and prevalence of age-related macular degeneration. It agrees with previous investigations.^17,35–38 Interestingly, Park and colleagues²³ recently compared the PPCT between 48 normal individuals and 108 glaucoma patients and found that PPCT was significantly reduced in eyes with normal-pressure glaucoma while it did not differ between eyes with primary open-angle glaucoma and normal individuals. In a similar manner, the study by Hirooka and colleagues¹⁹ on 50 normal individuals and 52 normal-pressure glaucoma patients revealed that PPCT in the inferonasal, inferior and inferotemporal region was significantly thinner in the normal-pressure glaucoma group as compared to normal subjects. In contrast, a study by Suh et al.³⁹ did not show a significant intereye difference in PPCT in patients with unilateral normal-pressure glaucoma. Roberts and colleagues¹⁸ found that the PPCT in patients with glaucoma who had sclerotic optic disc damage was approximately 25% to 30% thinner compared with that in patients with focal and diffuse optic disc damage and with that in healthy controls. It fits with the results of our study that PPCT was inversely correlated with the size of parapapillary β zone since a large parapapillary β zone is typical for the so-called “sclerotic optic disc damage.” 

With each year increase in age, global PPCT decreased by 2 μm in our study population (Table 2). The age-related decrease in PPCT has also been detected in previous hospital-based investigations. Roberts et al.¹⁸ reported a decline of 11 μm in PPCT per decade of life in a group of healthy subjects. Also Ehrlich et al.,¹⁷ Park et al.,²³ Chung et al.,³⁵ and others reported on an age-associated reduction in PPCT.^19,20 In contrast, Ho et al.²⁵ did not detect a significant correlation between thinner PPCT and younger age. The age-related decline in PPCT is parallel by an age-associated decrease in subfoveal choroidal thickness by approximately 4 μm per year of life.⁸ Since subfoveal choroidal thickness is approximately the double of the mean PPCT, the relative age-related loss in choroidal thickness is similar in the subfoveal region and in the peripapillary region. 

As subfoveal choroidal thickness, thinner PPCT was strongly correlated with longer axial length.⁸ Mean global PPCT decreased by 5 μm for each millimeter increase in axial length (Table 3). In a similar manner, Roberts¹⁸ reported a decrease in PPCT of 10 μm with every axial elongation of 1 mm. In relative terms, the figure of 5 μm decrease in PPCT can roughly be compared with the decrease of subfoveal choroidal thickness by 15 μm.⁸ The association between thinner PPCT and longer axial length was also detected in other investigations.^23,34,40 In contrast, Huang et al.²⁰ did not find a significant association between PPCT and axial length. In the Singapore Malay Eye Study, thicker PPCT was correlated with shorter axial length (P = 0.002); younger age (P = 0.02); lower triglyceride level (P = 0.02); and presence of diabetes (P = 0.04).³⁴ 

In multivariate analysis, thinner PPCT was strongly associated with larger parapapillary β zone (Table 3). It agrees with the results obtained by Takahashi and colleagues⁴¹ who examined 49 highly myopic healthy eyes and reported that the area of the peripapillary myopic crescent was significantly associated with thinner PPCT (P < 0.001) in a multivariate analysis. In a similar manner, Brito⁴² reported an association between thinner PPCT and tilted discs. In contrast to our study, the investigation of Ehrlich et al.¹⁷ showed that parapapillary β zone as a whole and measured in extension of clock hours was not significantly associated with PPCT after adjusting for glaucoma status and age, nor was the spatial distribution of β zone associated with the underlying PPCT. 

While in our study—after adjusting for multiple ocular and general parameters—PPCT was not significantly associated with diabetic retinopathy, Vujosevic and colleagues²⁶ reported that PPCT (and subfoveal choroidal thickness) significantly decreased with increasing level of diabetic retinopathy. In their study, no significant difference in PPCT was found between control individuals and diabetic eyes without detectable diabetic retinopathy. 

Thinner PPCT was significantly associated with worse best corrected visual acuity (Tables 2, 3). It is parallel to the association between thinner subfoveal choroidal thickness and worse visual acuity.¹³ Finally, thicker PPCT was associated with a higher prevalence of early and intermediate age-related macular degeneration. This association may indirectly be caused by the association between shorter axial length or hyperopia and a higher prevalence of age-related macular degeneration.⁴³ 

Several limitations are present in our study. First, The Beijing Eye Study had an overall response rate of 78.8%. Differences between participants and nonparticipants may have led to a selection bias. Among these participants, there were 443 participants with unavailable PPCT including a group of participants with uncertain posterior choroidal boundary owing to failure to detect an extreme thin or thick choroid. The exclusion of these individuals may have led to a bias. Second, subfoveal choroidal thickness shows circadian changes.⁴⁴ The PPCT measurements in our study had not been performed at the same time of the day; however, each participant underwent the OCT examination in a randomized manner with respect to when the readings were obtained. Therefore, it appears unlikely that circadian changes may have influenced the results of our investigation. Third, PPCT in this study was measured only in the right eye of each participant; intereye differences and their associations with intereye differences of other parameters could not be assessed. Fourth, in contrast to the subfoveal choroidal thickness, PPCT was not measured by the enhanced depth imaging mode of the OCT technology, what in the case of subfoveal choroidal thickness determinations might have resulted in problems in detecting the posterior choroidal border. However, since the PPCT was considerably thinner than the subfoveal choroid (134 ± 53 μm versus 254 ± 107 μm), so that the posterior border of the peripapillary choroid compared to the posterior border of the subfoveal choroid was markedly closer to Bruch's membrane, this limitation in the study design may not have markedly affected the measurements and the results and conclusions of our study. Fifth, the outer border of the peripapillary choroidal was defined as a line connecting and smoothing the outer limits of the large choroidal vascular spaces in those eyes in which a clear hyperreflective line of the inner surface of sclera could not be detected. This method is, however, an approximation that does not take into account the suprachoroidal layer. A hyporeflective suprachoroidal layer has been identified at the choroid-scleral junction in nearly half of healthy individuals in previous studies.⁴⁵ Since, however, hyperreflective inner scleral border could be identified in more than 95% of the eyes included into our study, the use of this approximation may not have markedly affected the results. Sixth, PPCT was measured in a circle with a diameter of 3.4 mm and which was centered on the disc center. However, the 3.4-mm scan measure is not 3.4 mm in eyes of different sizes and the OCT device used in our study did not correct for this. Since thinner PPCT was significantly associated with longer axial length, the question arises whether different retinal areas were measured in eyes with a different axial length and whether the difference in the measurement locations could have been the reason for the association between thinner PPCT and longer axial length. If with increasing axial length the measurement location shifts more distant to the optic disc border, it will lead to falsely low PPCT measurements in myopic eyes at locations outside of the papillomacular region. It would lead to falsely high PPCT values at the measurement location temporal to the optic disc, since the choroidal thickness increases in direction towards the fovea. Interestingly, however, the decrease in PPCT with longer axial length was most marked at the temporal location (Fig. 2B), supporting the notion of a general decrease in PPCT with longer axial length. Notwithstanding, the problem of the variability in the exact location of the 3.4-mm measurement ring in dependence of the axial length of the eye examined can have influenced the measurements to a certain degree and may have to be addressed in general in future studies. 

In conclusion, PPCT with a mean of 134 ± 53 μm is thickest superiorly and thinnest inferiorly. It decreases by 2 μm per year of life and by 5 μm per diopter of myopia. The age-related loss in PPCT took place in a similar manner at all measurement locations, while the axial elongation associated decrease in PPCT occurred predominantly in the temporal region. Thinner PPCT was correlated with larger parapapillary α and β zones. The association of thinner PPCT with lower best corrected visual acuity may warrant further studies. 

Acknowledgments

Disclosure: R. Jiang, None; Y.X. Wang, None; W.B. Wei, None; L. Xu, None; J.B. Jonas, Allergan, Inc. (C), Merck Sharp & Dohme Co., Inc. (C), Alimera Co. (C), Boehringer Ingelheim Co. (C), Sanofi Co. (C), P