Participants

We recruited 300 healthy white subjects with no eye symptoms among the patients and their relatives visiting our clinic for a routine ophthalmologic examination, and also among the hospital staff. All participants were > 18 years old. All subjects signed an informed consent form after the study protocol and its purpose were explained to them in detail. Participants first underwent a complete ophthalmic examination, including, biomicroscopy, intraocular pressure (IOP) measurement, fundus test, and refraction. The medical records of the individuals were also examined. For each patient, one eye was randomly selected for the final analysis.

Every eye included in our study had a vision equal to or better than 20/40, a sphere between +/-5 diopters and a cylinder between +/-3 diopters. Subjects recently undergoing cataract surgery (<1 year) or with a medical history of IOP >21 mmHg, glaucoma, or a macular, neuro-ophthalmological or retinal disease were excluded. Further exclusion criteria were a history of neurological disease or any uncontrolled disease.

Optical coherence tomography

An SD-OCT macular examination was performed without pupil dilation using the Spectralis OCT (Heidelberg Engineering, Heidelberg, Germany) in a dark room in each participant on the same day. The scan was conducted on a 20×20 degree cube with 49 raster lines separated by 120 μm, each containing 1046 pixels. This instrument's automatic eye tracking technology maintains fixation on the retina. Only well-centered, well-segmented images with a signal strength of >20db were used for analysis. All scans were performed and checked by the same experienced ophthalmologist and no manual adjustments to retinal layer segmentation were made.

Macular and inner retinal layer thicknesses were measured on 1, 3 and 6 mm rings as indicated in the ETDRS macular map. The variables recorded for each eye or participant were the means of each set of EDRS subfield thicknesses recorded for each retinal layer and the volumes of each of these layers. The 1 mm rings was named central circle, the 3 mm ring was named inner circle and the 6 mm ring was named outer circle. Volumes are automatically calculated by Spectralis OCT.

Using the new Spectralis segmentation software (6.0c version), the following measurements were acquired in each subject: overall retinal thickness, and thicknesses of the macular retinal nerve fiber layer (mRNFL), ganglion cell layer (GCL), inner plexiform layer (IPL), inner nuclear layer (INL), outer plexiform layer (OPL), outer nuclear layer (ONL—including Henle's fiber layer), retinal pigmentary epithelium (RPE) and photoreceptor layer (PR—being this layer the thickness between the outer limiting membrane and the Burch membrane) (Fig 1). All scans were performed by the same experienced operator and no manual adjustments to retinal layer segmentation were made.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Spectralis segmentation of the retinal nerve fiber layer (mRNFL), ganglion cell layer (GCL), inner plexiform layer (IPL), inner nuclear layer (INL), outer plexiform layer (OPL), outer nuclear layer (ONL), retinal pigmentary epithelium (RPE) and photoreceptor layer (PR).

https://doi.org/10.1371/journal.pone.0194169.g001

Statistical analysis

Data were obtained for one randomly selected eye of each participant. Descriptive statistics are reported as the mean, range and standard deviation. All p values were adjusted using the Bonferroni factor. Correlation between different measurements was defined through Pearson correlation coefficients. The effects of age and sex were analyzed by simple regression analysis. All statistical tests were performed using the SPSS software package version 15 (SPSS Inc., Chicago, IL, USA). Significance was set at p < 0.05.

Retinal layer thickness changes associated with age

The overall macular thinning observed here with age is consistent with the findings of studies including those of Kanai et al.,[2] Manassalorn et al.,[3] and Appukuttan et al.[4] in which negative correlation was detected between every retinal ETDRS zone but the central zone and age. In contrast, Grover et al.[8] found no significant differences with age perhaps because of the small number of subjects examined (n = 50).

As we age, neurons in the inner retinal layers are lost.[9] According to histological findings, between 0.3 and 0.6% of neurons are lost every year.[10] Ooto et al.[11] reported a linear decrease with age of -0.07 and -0.05 μm per year in GCL and IPL thickness, respectively, as measured using the Topcon 3D OCT (OCT- 1000; Topcon, Tokyo, Japan) in 256 healthy volunteers. In the present study, we observed correlation between reduced GCL and IPL thickness and age, indicating a rate of GCL and IPL thinning of -0.051 and -0.054 μm per year respectively. These figures are similar to the rates reported for other OCT studies [4, 12, 13]and are also in line with the results of histological studies [14], but in a lower rate of 0.17 and 0.16% of reduction every year in GCL and IPL thickness.

According to the rates of GCL and IPL thinning observed in our study participants, we could predict mean losses of 1.275 μm and 1.35 μm in the thickness of these two retinal layers respectively over the next 25 years. However, this estimate is only a theoretical value obtained from the linear relationship between age and thickness, and fails to consider, for example, if greater thinning is produced from a certain age onwards. Longitudinal studies are needed to confirm these rates of individual retinal nerve fiber layer thinning.

Knowledge of inner retinal layer thickness losses associated with age is important as any thickness reduction found could be the consequence of normal aging or of the progression of a disease such as glaucoma.[15]

Fewer studies have examined the effects of age on outer retinal layer thickness. In this study, we observed discrete positive correlation between age and OPL thickness, while ONL thickness was negatively correlated with age. In the study by Ooto et al.,[11] no significant correlation was found when examining both layers together. Hence, it seems that these layers should not be analyzed together as age may have an opposing effect on each one, possibly explaining the different results obtained.

Although aging has been traditionally associated with inner retinal layer thickness loss, the present results suggest ONL is the layer that undergoes most thinning with age. The outer nuclear layer of the retina is a hypo-reflective band that anatomically corresponds to the photoreceptor bodies. According to our results, we could estimate a mean thickness loss for this layer of 2.45 μm over 25 years, though again this rate needs to be confirmed in longitudinal and/or histological studies. On the other hand, in this study Henle's fiber layer was not measured individually and this may affect our results, since may be Henle's fiber layer the one decreasing with age [16].

The mean decrease in mean PR layer thickness produced in the present study with age is in line with the findings of others such as Won et al.[5] In this last study, significantly lower ETDRS peripheral and central PR thicknesses were noted when comparing subjects aged under 30 years and over 60 years, though no such difference was noted for ETDRS central ring thickness. In contrast, in our results the central circle had the strongest negative correlation with age (R = 0.30, p < 0,001).

In the retinal pigment epithelium, numerous structural changes with age have been described including an increase in the density of residual bodies, lipofuscin accumulation, drusen formation and Bruch's membrane thickening [17]. All these changes can lead to a RPE thickness increase with age when measured by SD-OCT. This increase could be real or the consequence of a reflectivity increase. The RPE thickening with age observed in our study was not significant; however, there have been reports of such an increase in RPE thickness with age.[12]

Retinal layer thickness according to sex

In this study, overall retinal thickness was significantly greater in men than women; the mean thickness difference recorded in the central zone was 7.113 μm. The data collected are consistent with reports by Appukuttan et al.,[4] Mass-in et al.,[18] and Jonas et al.[19] of a significantly higher mean retinal thickness found in male compared to female subjects. In the Jonas et al. study, the mean difference in macular thickness between men and women was 8.04 μm (95% CI 10.6–5.5; p<0.001), similar to the difference observed here.

Won et al.[5] detected significant differences between both sexes in mRNFL, GCL and IPL thicknesses measured in the central, inner and outer rings. Unlike our findings, these authors observed significant differences between IPL inner ring and foveal thickness though their study cohort was considerably smaller than ours (n = 50).

The results of our study also identified a significantly greater mRNFL volume in women than men. Similarly, in the study by Ooto et al.,[11] mean mRNFL thickness was 31.6 +/- 3.07 μm for men and 32.8 +/- 3.7 μm (p = 0.006) for women. Despite being statistically significant, the differences between both sexes in the thickness of individual retinal layers are minimal and bellow the resolution and the variability of Spectralis OCT.

Study limitations

The main limitations of our study were that participants were all European descent such that data collected may not be translatable to other ethnicities. Also its cross-sectional nature means that the significant changes with age detected require longitudinal studies to properly calculate the rate of thinning/thickening of the different retinal layers with age. Since Spectralis OCT resolution nowadays is not able to capture differences in the thickening bellow 4 μm, the thickness changes found in this study will need to be ascertain in future with more precise technology. And finally, the reader should be cautious when translating this results to clinical practice since the R values found in this study suggested that the thickness changes with age are statistically significant but minimal.

In conclusion, our findings indicate thickness changes associated with age in the majority of the retinal layers examined. No significant impacts of age were detected on mRNFL, INL or RPE thicknesses measured with the Spectralis OCT. Our data also confirm the greater overall thickness of the retina in men than women.