Macular Choroidal Thickness Profile in a Healthy Population Measured by Swept-Source Optical Coherence Tomography

Abstract

Purpose.: To determine choroidal thickness (CT) profile in a healthy population using swept-source optical coherence tomography (SS-OCT).

Methods.: This was a cross-sectional, noninterventional study. A total of 276 eyes (spherical equivalent ±3 diopters [D]) were scanned with SS-OCT. Horizontal CT profile of the macula was created measuring subfoveal choroidal thickness (SFCT) from the posterior edge of retinal pigment epithelium (RPE) to the choroid–sclera junction. Three determinations were performed at successive points 1000 μm nasal and five more temporal to the fovea. Subjects were divided into five age groups.

Results.: The mean SFCT was 301.89 ± 80.53 μm (95% confidence interval: 292.34–311.43). The mean horizontal macular choroidal thickness (MCT) was 258.69 ± 64.59 μm (95% confidence interval: 251.04–266.35). No difference in CT was found between men and women. Mean SFCT of the different study groups was 325.6 ± 51.1 (0–10 years), 316.7 ± 90.1 (11–20 years), 313.9 ± 80.3 (21–40 years), 264.6 ± 79.3 (41–60 years), and 276.3 ± 88.8 μm in subjects older than 60 years (P < 0.001; ANOVA test). Mean horizontal MCT was 286.0 ± 43.5, 277.7 ± 68.2, 264.0 ± 61.9, 223.4 ± 62.2, and 229.7 ± 66.1 μm, respectively (P < 0.001; ANOVA test). The CT profile was different for each age group.

Conclusions.: To our knowledge, this is the first population study of CT of healthy eyes across a broad range of age groups using SS-OCT. As has been determined using spectral-domain OCT, CT decreases with advancing age, especially after age 40. There were no differences due to sex. The greatest CT variation takes place in temporal sectors.

Introduction

In recent years the choroid and its role in posterior segment disease has become an increasing subject of study. Ultrasonography, ¹ magnetic resonance imaging (MRI), ² and Doppler laser have been employed to study the choroid, but due to insufficient resolution are of limited use. On the other hand, indocyanine green (ICG) angiography reveals useful clinical information but does not provide cross-sectional images of the choroid for in vivo study. ^3,4  

The introduction of optical coherence tomography (OCT) and its continuous development represent a clear breakthrough in choroidal imaging as it provides deeper, higher-resolution imaging of the eye layers with brief acquisition times. ^5,6 Initially, time-domain OCT (TD-OCT) was the technology available to study the posterior segment, but because of poor penetration below the retinal pigment epithelium (RPE) and relatively low resolution, TD-OCT could not be employed for choroidal imaging. In 2006, spectral-domain OCT (SD-OCT) became commercially available. Despite its obvious advantages over TD-OCT, signal roll-off with depth and signal attenuation by pigmented tissues or media opacities still precluded choroidal imaging in most eyes. Spaide et al. ⁷ introduced a technique to allow choroidal imaging using SD-OCT devices: enhanced depth image OCT (EDI-OCT), which provides consistent choroidal visualization in most eyes and allows quantitative and reproducible thickness measurements. The most recent innovative technology available for OCT imaging is high-penetration, swept-source longer-wavelength OCT (SS-OCT). ^8–11 Copete et al. ¹² and Ruiz-Moreno et al. ¹³ affirmed that reliable measurement of CT was possible in 100% of eyes using an SS-OCT device. 

As advancements in technology allow extensive studies of the choroid to be performed, variations in CT and morphology have been associated with conditions such as central serous chorioretinopathy, ^4,14–16 age-related macular degeneration, ^4,16–22 polypoidal choroidal vasculopathy, ^4,16–18,21 myopic maculopathy, ^23–27 posterior uveitis, ^{4,15,16,28–31} and choroidal tumors. ^4,32,33  

Age-related thinning of the choroid of healthy patients studied by SD-OCT has been well documented, ^34–39 as well as differences in CT between adult and pediatric eyes. ¹³ A paper by Ikuno et al. ³⁶ described the CT in healthy subjects ranging from 23 to 88 years of age with SS-OCT technology and CT changes according to change in subjects' age. To our knowledge, however, no previous report has determined the age at which the choroid is thickest, or the normal evolution or changes in thickness with age or sex across a wide span of ages. If choroidal variations do play a role in retinal diseases, the normal CT profile must be known so that it is possible to point out variations as they appear. The aim of this study was to determine CT profile in a large population with healthy eyes using SS-OCT. 

Patients and Methods

This was a cross-sectional, noninterventional study, performed at Vissum Alicante, Spain. The study followed the tenets of the Declaration of Helsinki. The institutional review board of Vissum Alicante approved the study. All examinations were obtained in the afternoon to avoid diurnal variations (16:00–20:00). ^40–42  

We manually measured the CT in 276 eyes from 154 patients. Ninety-three patients (60.4%) were male (164 eyes, 78 right and 86 left), and 61 patients (39.6%) were female (112 eyes, 57 right and 55 left). Their macular area was studied with an SS-OCT system (Topcon Corporation, Tokyo, Japan) after they provided informed consent. Inclusion criteria were best-corrected visual acuity (BCVA) between 20/20 and 20/25, spherical equivalent (SE) between +3 and −3 diopters (D), and no systemic or ocular diseases. Eyes with any history of any retinal disease in the fellow eye were not included. Eyes with SE beyond ±3 D were excluded. 

The SS-OCT device used to image the full-thickness choroid and sclera, ⁴³ which uses a tunable laser as a light source, operated at 100,000-Hz A-scan repetition rate in the 1-μm wavelength region. The device can do image averaging of up to 96 B-scans at each location. For this study, the reference mirror was placed at the deeper position of the retina so that the sensitivity was higher at the choroidal area in macular imaging. A line scanning mode, which produces an OCT image containing 1024 axial scans with a scan length of 12 mm, was employed. This sampling space in object space corresponds to 11.7 μm/pixel. Lateral resolution is set at 20 μm with 24-mm axial eye length, while axial resolution is 8 μm in the retina. ⁴⁴ Lateral and axial resolution are independent. 

Acquisition time was 1 second. This allowed us to obtain good-quality images even in 3-year-old children. A horizontal CT profile of the macula was manually created measuring CT (from the posterior edge of RPE to the choroid–sclera junction) under the fovea using the prototype software. The outer aspect of the lamina fusca, rather than the outer limit of the choroidal vessels, was the landmark used to determine the most distal aspect of the choroid. 

Five further determinations were performed every 1000 μm temporal (T1, T2, T3, T4, and T5) and three more nasal (N1, N2, and N3) to the fovea (Fig. 1). 

Figure 1

 

Example of choroidal thickness measures in all nine locations, from the posterior edge of RPE to the choroid–sclera junction; from N3 position (right) to T5 position (left).

An experienced technician determined refractive errors and BCVA using an autorefractometer (Nidek, Gamagohri, Japan) that was later checked by a certified optometrist. 

To study the possible evolution of the CT, the study group was divided into five subgroups according to age distribution: 0 to 10 (eyes, n = 75), 11 to 20 (n = 48), 21 to 40 (n = 50), 41 to 60 (n = 40), and older than 60 years (n = 63). Mean age was 33.5 ± 24.9 years (from 3 to 95). Mean SE was 0.10 ± 1.36 D (from +3 to −3). 

Two observers determined CT independently and in a masked fashion. 

For statistical treatment of the data, the program used was version 17.0 of SPSS for Windows (SPSS, Chicago, IL, USA). Interobserver reproducibility was evaluated using intraclass correlation coefficient (ICC) for each variable measured (mean and 95% confidence interval), coefficient of variation between graders, and Bland-Altman plots. The means of the measures obtained by the two observers were the data used for the rest of the calculations. Kolmogorov-Smirnov test was applied for all data samples in order to check normality. Comparison between groups was performed using Student's t-test when samples were normally distributed or Mann-Whitney test when parametric statistics were not possible. The level of significance used was always the same (P < 0.05). Homogeneity of variances was checked using the Levene test. For comparison of several independent samples, analysis of variance (ANOVA) or Kruskal-Wallis test was used depending on whether normality could be assumed. Bivariate correlations were evaluated using Pearson or Spearman correlation coefficients, depending on whether normality could be assumed or not. For the development of predictive models, linear regression was used. 

Results

Swept-source OCT allowed both independent observers clear visualization of both the RPE and scleral–choroidal junction and therefore accurate measurement of CT in all eyes (100%). Mean subfoveal choroidal thickness (SFCT) was 301.89 ± 80.53 μm (from 99.50 to 539.50; 95% confidence interval: 292.34–311.43). Mean macular horizontal CT was 258.69 ± 64.59 μm (from 99.00 to 455.28; 95% confidence interval: 251.04–266.35). The horizontal CT profile can be seen in Figure 2. 

Figure 2

 

Choroidal thickness profile in general population (microns/measurement locations).

No statistically significant difference in CT was found in men compared to women. The two sexes showed a similar choroidal profile (Fig. 3), although women were found to have a trend toward thinner temporal choroids. Mean SFCT was 303.9 ± 70.9 μm in men versus 296.4 ± 94.1 μm in women (P = 0.483; Student's t-test). Mean horizontal MCT was 260.8 ± 60.9 μm in men versus 255.0 ± 70.7 μm in women (P = 0.227; Mann-Whitney U test). Mean age was 30.39 ± 25.37 in men versus 37.48 ± 23.86 in women (P = 0.006; Mann-Whitney test). Mean SE was 0.15 ± 1.38 D in men versus 0.09 ± 1.34 D in women (P = 0.796; Mann-Whitney test) (Table 1). 

Figure 3

 

Choroidal thickness profile comparison, men (red) versus women (blue) (microns/measurement locations).

Table 1

 

Choroidal Thickness Comparison in Men and Women

	Male	Female	P Test
Choroidal thickness	260.8 ± 60.9	255.0 ± 70.7	0.227, Mann-Whitney
95% CI	251.15–269.76	242.91–267.32
Subfoveolar thickness, MSFT	303.9 ± 70.9	296.4 ± 94.1	0.483, Student's t-test
95% CI	293.22–314.86	281.15–316.13
Age, mean ± SD	30.39 ± 25.37	37.48 ± 23.86	0.006, Mann-Whitney
95% CI	26.48–34.24	33.64–42.65
SE, mean ± SD	0.15 ± 1.38	0.09 ± 1.34	0.796, Mann-Whitney
95% CI	−0.07 to 0.36	−0.19 to 0.30

Mean SFCT of the different study groups (Table 2) was 325.6 ± 51.1 μm in subjects 0 to 10 years, 316.7 ± 90.1 μm in those 11 to 20 years, 313.9 ± 80.3 μm in those 21 to 40 years, 264.6 ± 79.3 μm in those 41 to 60 years, and 276.3 ± 88.8 μm in those older than 60 years, differences that were statistically significant (P < 0.001; ANOVA test). Mean horizontal MCT was 286.0 ± 43.5, 277.7 ± 68.2, 264.0 ± 61.9, 223.4 ± 62.2, and 229.7 ± 66.1 μm for subjects 0 to 10, 11 to 20, 21 to 40, 41 to 60, and >60 years, respectively (P < 0.001; ANOVA test). The CT profile was different in each age group (Figs. 4, 5, 6), showing an evident statistically significant difference; CT was thicker as age decreased except in the group older than 60 years, where it was thicker than among those 41 to 60 years (Table 3). However, differences in mean MCT and SFCT were not significant between subjects ages 41 to 60 versus those older than 60 years (P = 0.629 and 0.502; Student's t-test, respectively). 

Figure 4

 

Choroidal thickness profile in different age groups (microns/measurement locations).

Figure 5

 

Mean choroidal thickness (microns) in different age groups.

Figure 6

 

Mean subfoveolar choroidal thickness (microns) in different age groups.

Table 2

 

Choroidal Thickness Comparison in Different Age Groups

	0–10 y	11–20 y	21–40 y	41–60 y	>60 y	P, ANOVA
Mean choroidal thickness	286.0 ± 43.5	277.7 ± 68.2	264.0 ± 61.9	223.4 ± 62.2	229.7 ± 66.1	<0.001
95% CI	276.04–296.08	257.90–297.54	246.41–281.64	203.54–243.34	213.11–246.42
Subfoveal choroidal thickness	325.6 ± 51.1	316.7 ± 90.1	313.9 ± 80.3	264.6 ± 79.3	276.3 ± 88.8	<0.001
95% CI	313.92–337.44	290.54–342.91	291.11–336.78	239.11–290.06	253.94–298.67

Table 3

 

Choroidal Thickness Values According to Age Group in All Measurement Locations

Age	N3	N2	N1	SF	T1	T2	T3	T4	T5	SE, D
0–10:
Mean	161.06	221.63	280.65	325.68	330.51	326.75	320.81	313.57	293.8	0.51
SD	45.02	49.1	54.33	51.13	54.33	56.12	59.72	63.7	60.61	1.49
95% CI	150.80–171.52	210.34–232.93	268.15–293.15	313.92–337.44	318.01–343.01	313.84–339.87	307.07–334.55	298.91–328.22	279.85–307.75	0.13–0.81
11–20:
Mean	159.21	219.79	279.14	316.73	319.23	312.14	300.49	295.02	297.74	−0.21
SD	61.58	80.94	90.6	90.18	80.54	72.16	69.56	80.11	76.06	1.76
95% CI	141.33–177.09	196.29–243.30	252.83–305.44	290.54–342.91	295.84–342.62	291.18–333.09	280.29–320.69	271.76–318.28	275.65–319.82	−0.71 to 0.33
21–40:
Mean	156.37	218.03	277.86	313.94	313.12	298.25	283.55	264.92	250.17	−0.15
SD	54.56	65.16	73.41	80.35	81.79	78.89	77.33	73.84	71.11	1.04
95% CI	140.87–171.88	199.51–236.55	257.00–298.72	291.11–336.77	289.88–336.36	275.83–320.67	261.57–305.53	243.94–285.90	229.96–70.38	−0.50 to 0.10
41–60:
Mean	123.99	177.08	230.56	264.69	258.71	252.91	241.79	234.89	226.33	0.03
SD	58.35	84.35	82.25	79.34	80.8	69.42	65.11	64.75	71.46	0.94
95% CI	105.33–142.65	150.10–204.05	204.26–256.87	239.31–290.06	232.87–284.55	230.71–275.11	220.97–262.61	214.18–255.60	203.47–249.18	−0.27 to 0.33
≥60:
Mean	144.79	201.48	253.04	276.3	265.39	250.77	236.98	228.89	210.25	0.2
SD	63.83	81.51	93.72	88.81	85.47	79.29	69.53	70.76	70.31	1.26
95% CI	128.71–160.86	180.95–222.00	229.44–276.44	253.93–298.67	243.86–286.92	230.80–270.74	219.47–254.49	211.07–246.71	192.54–227.95	−0.13 to 0.51
P test	0.001 Kruskal-Wallis	0.002 ANOVA	0.001 ANOVA	<0.001 ANOVA	<0.001 ANOVA	<0.001 ANOVA	<0.001 ANOVA	<0.001 ANOVA	<0.001 ANOVA	0.016 Kruskal-Wallis

Spherical equivalent was not different between groups except for the group 0 to 10 years (P = 0.016; Kruskal-Wallis test), with 0.51 ± 1.49 D in this group. Mean SE of the other groups was not different (P = 0.429; Kruskal-Wallis test) (Table 3). 

Correlation Analysis and Linear Regression Models

There exists a high correlation between CT and age, which is statistically significant (Table 4; Fig. 7). This correlation was found to be stronger the farther the CT was measured temporally from the foveal center. Correlations between SE and CT were significant only in subfoveal and nasal measurement points (Table 4). 

Figure 7

 

Correlation between age and mean choroidal thickness (microns, top). Correlation between age and mean subfoveolar choroidal thickness (microns, bottom).

Table 4

 

Correlations of CT With Age and Spherical Equivalent

	N3	N2	N1	SF	T1	T2	T3	T4	T5	MCT
Age, Spearman's Rho	−0.182, P = 0.002	−0.180, P = 0.003	−0.207, P = 0.001	−0.287, P = 10−6	−0.344, P = 4 × 10−9	−0.421, P = 3 × 10−13	−0.454, P = 2 × 10−15	−0.447, P = 6 × 10−15	−0.432, P = 6 × 10−14	−0.385, P = 3 × 10−11
Spherical equivalent, Spearman's Rho	0.178, P = 0.003	0.174, P = 0.004	0.173, P = 0.004	0.141, P = 0.020	0.104, P = 0.086	0.105, P = 0.082	0.082, P = 0.178	0.084, P = 0.166	0.064, P = 0.294	0.147, P = 0.015

Macular CT multiple regression analysis including age and SE was R = 0.407 (r ² = 0.166), showing that subtle changes appear with the variation of this parameter but with a high significance (P < 0.001). The equation predicted is as follows:  

 

 

Following this model, a mean reduction of 10 μm in MCT per decade (1 μm every year) and of 5.3 μm per SE diopter can be predicted. We evaluated the intereye correlation through bivariate correlations. In all measurements the correlation was strong and statistically significant (P < 0.001; Pearson correlation test), with values oscillating from r = 0.680 (T5) to r = 0.830 (N1) and r = 0.864 for MCT. This suggests that similar results should be obtained using one eye (randomly selected) or two eyes in the models developed. In fact, the results predicted are very similar. Evaluating only one eye per patient (taking right or left randomly), the most accurate equation to predict MCT according to age would be the following:  

 

 

The SE is not statistically significant, so the equation to calculate MCT would then be  

 

 

The decrease in MCT is calculated to be 9.4 μm per decade (versus 10 μm when both eyes are included). 

Correlation between mean SFCT and age was R = 0.275, r ² = 0.076. As before, the dependence found is subtle but highly significant (P < 0.001).  

 

 

This model predicts a loss of 8.87 μm of SFCT per decade. 

If only one eye is evaluated per patient (taking right or left randomly), a more accurate calculation yields a SFCT in microns of 333.27 − 0.852*Age (years); R = 0.266, r ² = 0.071. The estimated decrease in SFCT is 8.5 μm per decade (versus 8.87 μm when both eyes are included). 

Intraclass correlation coefficient in CT for the two independent observers varied between 0.973 and 0.987 (Table 5). Bland-Altman plots show this good interobserver correlation (see Fig. 8). There were no statistically significant differences between the variation coefficients obtained by the two observers (Table 6; P > 0.05, Wilcoxon test). 

Figure 8

 

Bland-Altman plots for interobserver correlation in every measurement location, from N3 (upper left) to T5 (bottom).

Table 5

 

ICC in Different Measurement Locations

	ICC	95% Confidence Interval
N3	0.973	0.966, 0.978
N2	0.981	0.976, 0.985
N1	0.981	0.976, 0.985
SF	0.987	0.984, 0.990
T1	0.986	0.982, 0.989
T2	0.981	0.977, 0.985
T3	0.984	0.979, 0.987
T4	0.982	0.978, 0.986
T5	0.981	0.976, 0.985

Table 6

 

Coefficients of Variation of Each Observer at Each Measurement Point and Each Age Group Studied

	Coefficient of Variation
	Observer 1	Observer 2
N3	0.379	0.386
N2	0.345	0.351
N1	0.302	0.303
SF	0.266	0.268
T1	0.271	0.270
T2	0.264	0.269
T3	0.269	0.273
T4	0.290	0.289
T5	0.304	0.301
MCT
>60 y	0.287	0.289
41–60 y	0.277	0.280
21–40 y	0.237	0.234
11–20 y	0.248	0.244
0–10 y	0.152	0.153
SFCT
>60 y	0.320	0.324
41–60 y	0.299	0.302
21–40 y	0.258	0.256
11–20 y	0.286	0.285
0–10 y	0.158	0.158

 There were no statistically significant differences between the variation coefficients obtained by each observer (P > 0.05; Wilcoxon test).

Discussion

To our knowledge, there is no previously published large study across all age groups (including pediatric patients) elucidating the normal age-related choroidal thinning that occurs in healthy eyes. Age-dependent thinning is a key factor to establish the role of the choroid in retinal pathology, something that has already been shown in several studies. ^4,14–33 Although Ikuno et al. ³⁶ performed a similar study in which they described the CT in healthy Japanese subjects, their study did not include a wide variation of ages. Furthermore, the loss of SFCT predicted by our study is 0.852 μm per year, far from the 4.32 μm per year predicted by previous papers, leading to different conclusions. 

This is the largest study to date of CT measured with SS-OCT, and it confirms again the accuracy with which SS-OCT can measure the choroid. Swept-source OCT was chosen to perform this study because its characteristics allow the analysis of the choroid with more precision than previous OCT technology from a theoretical point of view. This accuracy has been shown in papers by Copete et al. ¹² and Ruiz-Moreno et al. ¹³ using the same device. Full CT was successfully measured in 100% of the patients in those studies. Using SS-OCT we successfully measured the CT of 100% of the eyes in this study as well. Wei et al. ⁴⁵ successfully measured 93.2% of 3468 patients (largest study reported to date) using Heidelberg Spectralis OCT on the EDI setting. The choice of Manjunath et al. ³⁷ was Cirrus HD-OCT without EDI, which was able to successfully measure the choroid in only 74% of the eyes. Yamashita et al. ⁴⁶ and Shin et al. ⁴⁷ used Topcon SD-OCT and were able to image 63.3% and 90.7%, respectively. Copete et al. ¹² compared SD-OCT and SS-OCT, obtaining good measurements in 70.4% and 100%, respectively. 

Margolis et al. ³⁸ studied a group of 54 eyes with a mean SE of −1.3 D, finding a mean SFCT of 287 μm, similar to the results of Flores-Moreno et al. ⁴⁸ and those of Manjunath et al., ³⁷ who published a mean SFCT of 272 μm in 34 eyes of 51.1-year-old patients. 

Xu et al. ⁴⁹ compared diabetic with nondiabetic patients, finding a mean SFCT of 266 μm in the control group (1795 subjects). 

The choroid has also been studied in pediatric population using SS-OCT. Ruiz-Moreno et al. ¹³ compared both pediatric (83 eyes, 10-year-old patients with mean SE of 0.3 D) and adult (75 eyes, 53-year-old patients with mean SE of 0.16 D) healthy patients, finding a mean SFCT of 312 and 302 μm, respectively. Park and Oh, ⁵⁰ in a recent study with pediatric subjects, found a mean SFCT of 348.4 μm in the 48 eyes studied. 

In the present study, the mean SFCT was found to be 301.89 μm. This figure is approximately 10 μm greater than that found by other authors using SD-OCT. ^{37–39,45–49} It is likely that the introduction of pediatric eyes into the study group resulted in a slight increase in the mean SFCT. Interestingly, these results match those of Ruiz-Moreno et al., ¹³ who with the same device found almost the same CT in their patient group consisting of adults only. Swept-source OCT may result in a CT measurement slightly greater than that seen with SD-OCT. This may be explained by the higher quality of the images and by the fact that the measures were taken from the posterior edge of the RPE to the outer aspect of the lamina fusca, rather than the outer limit of the choroidal vessels. 

Ruiz-Moreno et al. ¹³ with SS-OCT and Flores-Moreno et al. ⁴⁸ with SD-OCT provide data about mean MCT. Ruiz-Moreno found it to be 275 μm in adults compared to 285 μm in pediatric subjects. Flores-Moreno obtained a result of 257 μm in healthy patients and 115 μm in highly myopic patients. We found that mean MCT was 258.69 μm, similar to these authors' results. Nevertheless, the present study is the first employing a subfoveal 8-mm line of the macula compared to the 6 mm studied by other authors. This was possible due to SS-OCT technology, which provides a more robust imaging of the temporal choroid, as this 2-mm increase was performed on the temporal sector. This may prove to be a clinically important measure, as it is in the temporal sector that greater variations of CT with age take place. 

We found no statistically significant difference when comparing men to women. Subjects of both sexes showed almost identical choroidal profiles with a minimal difference in the temporal sector (Fig. 3), in which the females usually had slightly thinner choroids. Mean SFCT was 303.9 μm in men versus 296.4 μm in women. These slight, nonstatistically significant differences can be explained by the age differences of the two groups, 30.39 years for men and 37.48 years for women. Choroidal thickness was approximately 7 μm greater in men on average, but the men in this study were, on average, 7 years younger. According to the correlation formulas established by this database, CT in a sex- and age-matched group would likely have been almost identical. 

A few studies have been previously performed in adults using high-penetration OCT ^36,43,51 but without comparing different age groups. We analyzed CT comparing several ranges of age with a 20-year gap. The first two age groups spanned from age 0 to 10 and age 10 to 20 years. These ranges were selected because eye growth undergoes two phases, the first up to age 6 years, reaching emmetropization, and the second from 7 to 19 years, when global expansion of the eye takes place. ⁵² Park and Oh ⁵⁰ (mean age of subjects 86.4 months, from 52 to 131 months) and Ruiz-Moreno et al. ¹³ studied CT in pediatric subjects, with only the latter comparing CT simultaneously with adults. The CT profile changed in every group (Figs. 4, 5, 6) and the choroid measured thinner as age increased, showing statistically significant differences. This decrease was found to be progressive until 40 years, at which point the most significant variation took place, from 313.9 μm in the 21- to 40-year-old group to 264.6 μm in the 41- to 60-year-old group (49-μm difference). Interestingly, from 40 years on (41–60 vs. >60 years) the differences in CT were not found to be statistically significant, either mean MCT or SFCT. Variations when only one eye per patient (as in Ray and O'Day ⁵³ ) or both were taken into account were minimal, showing calculated decreases of 9.4 vs. 10 μm per decade in MCT and 8.5 vs. 8.87 μm per patient in SFCT. 

One of the limitations of this study is the fact that the CT was manually determined. There is now an automated software commercially available, Topcon DRI OCT-1 software (version 9.01.003.02), for automated segmentation and thickness measurements in three-dimensional or radial scan mode; however, its accuracy in clinical practice has yet to be widely tested. 

According to our results, the macular CT profile in a healthy population is similar between different age groups, with the choroid in healthy eyes getting thinner with age, particularly when adults older than 40 years are compared to children and younger adults. There were no differences due to sex. The greater CT variation due to age takes place in temporal sectors. 

Acknowledgments

Supported in part by a grant from the Spanish Ministry of Health, Instituto de Salud Carlos III, Red Temática de Investigación Cooperativa en Salud (“Prevención, detección precoz y tratamiento de la patología ocular prevalente, degenerativa y crónica”) (RD12/0034/0011) and by a Research to Prevent Blindness unrestricted grant to the New England Eye Center/Department of Ophthalmology, Tufts University School of Medicine and the Massachusetts Lions Clubs. 

Disclosure: J. Ruiz-Medrano, None; I. Flores-Moreno, None; P. Peña-García, None; J.A. Montero, None; J.S. Duker, Carl Zeiss Meditec, Inc. (F), Optovue, Inc. (F); J.M. Ruiz-Moreno, None