Abstract
Purpose.:
To study the influence of macular pigment on retinal straylight in healthy eyes.
Methods.:
This prospective study included 150 eyes of 75 healthy subjects between 12 and 81 years of age (mean, 46.1 years) without a history of ocular surgery or ocular disease known to influence straylight (e.g., cataract). Retinal straylight was measured with the compensation comparison technique, and the known influence of age and axial length was compensated by calculating the base, age, and axial length–corrected (BALC) straylight. Macular pigment optical density (MPOD) was measured using light-emitting diode (LED) heterochromatic flicker photometry. Axial length was determined with a partial coherence biometer and iris color by visual inspection. Data analysis consisted of studying the predictive values between these parameters, after correction for the symmetry between left and right eyes, using linear mixed models.
Results.:
Mean retinal straylight was 1.05 ± 0.18 log units, and the mean MPOD was 0.37 ± 0.19. Age and axial length were found to be important predictors of retinal straylight (P < 0.001 and P = 0.010, respectively) but not of MPOD (P > 0.05). The mean BALC straylight was −0.07 ± 0.13 log units. No significant relationship was found between retinal straylight and MPOD (P > 0.05), even after correction for age and axial length. Also, no significant correlation was found between iris color and BALC straylight or MPOD.
Conclusions.:
In healthy human eyes, retinal straylight values measured with the compensation comparison technique are not significantly correlated with macular pigment optical density.
Retinal straylight is incident light scattered by ocular media that causes a veil of light over the retinal image. This reduces the contrast of the image projected on the retina
1 and causes disability glare, a physiological phenomenon.
2,3 Disability glare may lead to a decreased quality of vision through the blinding effect of a bright light source somewhere in the visual field. Typical examples are a low sun or approaching headlights at night.
1 The amount of retinal straylight is different for each individual and depends on age,
3,4 pigmentation,
5 and axial length,
6 but it is also influenced by pathologic conditions such as cataract
7 and corneal dystrophies.
8 Based on recent studies that suggest that individuals with higher levels of macular pigment optical density (MPOD) are less affected by disability glare,
9,10 macular pigment (MP) has been proposed as another factor that might influence straylight. However, this influence remains to be confirmed.
Macular pigment consists of diet-derived carotenoids that were first described as xanthophyllic carotenoids by Wald
11 and later more specifically identified as lutein, zeaxanthin,
12 and meso-zeaxanthine
13 by Bone et al. In the fovea, MP is predominantly located in the Henle fiber layer, and at the parafoveal site it is located in the plexiform and inner nuclear layers.
14,15 There are several hypotheses regarding the function of the MP. The first, known as the protection hypothesis, is that the carotenoids filter potentially actinic short-wave (blue) light and are antioxidants that protect the retina from oxidative damage. Oxidative stress plays an important role in the etiopathogenesis of age-related macular degeneration and is exacerbated in part by cumulative short-wave light exposure.
9,16–18
Another hypothesis of MP function is the optical hypothesis,
19,20 which states that how far one can see and how well details can be resolved are determined by the light's wavelength and the amount it is scattered. Short wavelengths are known to scatter more, thus dominantly influencing the visual acuity and straylight-related visual discomfort,
16,21 while MP is known to absorb this short-wave energy. Thus, higher levels of MPOD would improve the resolution,
16 which in turn has a beneficial effect on a patient's visual discomfort caused by glare.
9
The aim of this study was to see whether a relationship exists between macular pigment optical density and retinal straylight.
One hundred and fifty eyes of 75 subjects, aged 12.8 to 80.2 years (mean, 46.1 years; SD = 18.7), were included in this study, most of them recruited from the nearby town of Edegem. Any subjects with a history of amblyopia, ocular surgery, early cataract (diagnosed by Scheimpflug images, Oculus Pentacam, Wetzler, Germany), corneal or retinal pathologies, epilepsy, or systemic diseases (e.g., diabetes, systemic macula diseases) were excluded, as well as pregnant women and hard contact-lens wearers. All subjects recruited in the study could be considered as naïve observers for the tests carried out. The data were collected in the framework of Project Gullstrand, a European multicenter study conducted to determine the correlation between ocular biometry and several psychophysical tests in the general population, as well as to determine what levels of visual quality are tolerable before they affect a patient's quality of life. One of the parameters included in Project Gullstrand is retinal straylight. For a subgroup, the MPOD was also measured. This study adhered to the guidelines of the Declaration of Helsinki for research in human subjects and was approved by the ethical committee of the Antwerp University Hospital (ref 10/36/241). Written informed consent was obtained from each volunteer prior to measurements.
Retinal straylight was measured psychophysically using the C-Quant (Oculus Optikgeräte, Wetzler, Germany), a commercially available version of the “compensation comparison method” first proposed by Franssen et al.
22 The subject fixates on two half circles in the center of the test field, which is surrounded by an annular glare source with a radius of 5° to 10°, resulting in an effective average angular value of 7°, around the fovea.
23 During the test, one of the half circles flickers in counterphase with the glare source, and the subject must identify this flickering half circle in a forced-choice comparison. This procedure is repeated 25 times with different intensities of the glare source. This method has proven to give reliable and objective measurements of intraocular straylight.
24,25 All measurements were performed monocularly in a dark room, and when refractive correction was needed, thoroughly cleaned trial lenses were used to correct for the spherical equivalent. Only measurements with an estimated standard deviation (Esd) below 0.08 and a measurement quality parameter of Q > 0.5 were used for analysis.
Retinal straylight(s) is known to be influenced by age and axial length (L), which can be modeled by the following equation
6 :
By subtracting the first two terms in
Equation 1 from the measured straylight log(s), it is possible to define the base and age-corrected (BAC) straylight, a straylight parameter that does not depend on age. Similarly, by subtracting
Equation 1 in its entirety, one can define the base, age, and axial length–corrected (BALC) straylight in which the influence of axial length has been compensated as well. Pupil size was not taken into account as Franssen et al.
26 and Garcia-Lazaro et al.
27 demonstrated that no correction for pupil size is required for a C-Quant straylight meter when estimating average amounts of retinal straylight from samples of eyes.
MPOD was measured psychophysically by measuring foveal and parafoveal sensitivities to lights with wavelengths of 470 to 540 nm, using the method of light-emitting diode (LED) heterochromatic flicker photometry (QuantifEye; Tinsley, Croydon, UK). During the test, the patient fixates first on a 1° blue-green LED central target that flickers with gradually increasing amplitude. As soon as the flicker is detected, the patient must press a button, and the test is repeated at a different wavelength. Then, the patient fixates at a 1.75° red spot located at 8° horizontal eccentricity, and a second set of data are recorded for peripheral viewing. The MPOD is then determined by the difference between the minima obtained from the central measurement and the peripheral measurement.
28,29 The device (QuantifEye; Tinsley) has a good test–retest reliability and a proven validity,
29 but some subjects find it difficult to carry out the peripheral task of the test (which was the case in six eyes of five individuals). For those subjects, the age-estimate MPOD was used.
30 The test is also unsuitable for individuals with a limited visual field, insufficient visual acuity, or learning difficulties,
31 but no such subjects were included in the test population.
Furthermore, we performed axial-length measurements with a partial coherence biometer (Lenstar LS900; Haag-Streit, Koeniz, Switzerland) and measured the refraction with an autorefractometer (AR-700; Nidek, Gamagori, Japan). Iris color was classified by the examiner into three categories, after determination by visual inspection: blue-gray (41 subjects), green (12 subjects), or brown-black (22 subjects).
The data were analyzed using SPSS version 21.0 for Windows (IBM Corp., Armonk, NY). Descriptive statistics were calculated. Since both eyes of each subject were used in the analysis, linear mixed models were used to correct for between-eye correlations.
32 This method gives a
P value to indicate the predictive value that one particular parameter (e.g., MPOD) has on another (e.g., straylight) but cannot provide a Pearson correlation value. Instead, uncorrected Pearson correlation values are given in the tables—one must be aware that these will slightly overestimate the actual Pearson correlation value. Pearson correlation values were also calculated for left and right eyes separately.
A power analysis using simulations demonstrated that using 150 eyes of 75 subjects, and taking the correlation between eyes into account, leads to a statistical power of 86% to detect a correlation of 0.25 at a significance level of 0.05. This significance level of 0.05 was also used throughout the rest of our analysis.
The study population consisted of 150 eyes for which the population descriptives are shown in
Table 1. The mean axial length was 23.64 ± 1.14 mm and ranged from 20.91 to 26.53 mm. The mean values of the measured parameters, together with the uncorrected Pearson correlation coefficients and the corrected predictive values, are listed in
Table 2. Mean retinal straylight was 1.05 ± 0.18 log units, while mean MPOD was 0.37 ± 0.19.
MPOD was not found to be a potentially important predictor for retinal straylight (P = 0.635; Fig. a). As expected from the literature, age was a good predictor for retinal straylight (P < 0.001), and axial length was a good predictor for BAC straylight (P = 0.010). Neither age nor axial length was a good predictor for MPOD (P = 0.994 and P = 0.444, respectively). After adjustment for age and axial length, the average BALC straylight was −0.07 ± 0.13 log units, which was not well predicted by the MPOD (P = 0.784; Fig. b). When looking at the left and right eye separately, for each straylight value, no significant correlation was seen.
No significant differences were found between the three iris color groups for BALC straylight (blue-gray, −0.05 ± 0.12; green, −0.06 ± 0.11; brown-black, −0.11 ± 0.14; P = 0.104) or MPOD (blue-gray, 0.38 ± 0.17; green, 0.35 ± 0.13; brown-black, 0.38 ± 0.25; P = 0.901).
When light enters the eye, it is scattered by imperfections in the ocular media. This scattering can be subdivided into light scattered toward the retina (forward scatter or retinal straylight) and backwards scattered light (backscatter). Forward scatter is caused by the deviation of light rays away from the normal direction of propagation, thereby reducing retinal contrast and causing glare,
7,33 a well-known visual consequence of intraocular straylight that has recently gained importance when assessing quality of vision. According to the optical hypothesis, MP has a particular role in reducing the effects of light scatter via its light-filtering properties, thus enhancing visual performance and comfort.
18
The main result in this study was the finding that MPOD is not an important predictor of retinal straylight, even when corrected for age and axial length. If the between-eye correlation were not taken into account, the Pearson correlation coefficient between BALC straylight and MPOD would also not have been significant (
r = −0.09;
P = 0.275). Moreover, the value for this correlation coefficient is artificially too high, and its
P value artificially low, owing to the existing between-eye correlation.
32 These results confirm earlier reports in the literature by Loughman et al. and O'Beirne et al. on the absence of a correlation between MPOD and glare sensitivity,
18 or MPOD and retinal straylight (O'Beirne R, et al.
IOVS 2012;53:ARVO E-Abstract 6407). Two studies, Stringham and Hammond
9 and Stringham et al.,
34 on the other hand report a positive and statistically significant correlation between MP and glare (
Table 3), thus confirming the optical hypothesis. This difference in outcome may be due to methodological differences between our work and that of O'Beirne et al. on one side, which studied retinal straylight, and that of Stringham on the other side, which investigated glare sensitivity. As glare sensitivity is a consequence of forward scattered light, it is possible that both parameters relate differently to the MP. Other known consequences of retinal straylight are hazy vision, halos around bright lights, color and contrast loss, and difficulty with against-the-light face recognition.
23
The absence of a correlation between straylight and macular pigment is not surprising as retinal straylight originates from light scatter in the anterior segment, which, observed from the retinal plane, cannot be separated from the retinal image. The MP, with a peak density just above the photoreceptors of the central macula,
14,35 acts as a yellow filter that absorbs the short-wavelength end of the visual spectrum. This absorption occurs equally for both the light of the retinal image and the scattered light, so the proportion between both light sources remains the same. Hence, there is no theoretical reason for MP to influence straylight, regardless of its density. Although retinal straylight is known to occur more in the short-wavelength spectrum,
36 which corresponds well with the absorption spectrum of the MP, the argument above remains valid. However using the C-Quant, we were not able to verify this lack of correlation experimentally for short wavelengths, as it uses an achromatic light source. It is, however, possible that a correlation for short wavelengths exists between MPOD and disability glare, a consequence of straylight, as was already reported by Stringham and Hammond.
9
We also could not find any correlation between MP and age in this population. From the literature, it is unclear whether MP levels vary throughout life since this relationship may be influenced by variables such as dietary intake of carotenoids, body fat, smoking, or other less-known factors.
17 A study of MP levels measured by HPLC in 87 postmortem donor eyes found no significant differences across subjects aged 3 through 95 years.
37 No significant change was also found by Ciulla and Hammond in 390 subjects aged from 18 to 88 years.
38
Although some studies in the literature reported that MP density was lower in blue to gray eyes than in dark eyes,
39,40 no significant differences between the three subgroups were found. A similar result has already been reported by Iannaccone et al.
41
The literature also reports higher straylight values in eyes with light iris colors than in eyes with dark iris colors,
5 especially in light-blue–colored eyes.
42 In this study, no significant differences in BALC straylight were seen between the iris color groups, although the mean BALC straylight value was lowest for the brown-black–colored group, which is in line with previous reports on retinal straylight and iris color. The fact that we combined light-blue and blue irises, as our population had only very few light-blue eyes, is a possible reason of finding no significant differences.
In conclusion, we found that the light-filtering properties of MP do not affect retinal straylight, which is the source of the glare phenomenon. Further investigation is needed on how MP influences the glare phenomenon itself and the resulting visual performance of the human eye.
Supported by the Flemish government agency for Innovation by Science and Technology (Grant IWT 110684). The authors alone are responsible for the content and writing of the paper.
Disclosure: S. Jongenelen, None; J.J. Rozema, None; M.-J. Tassignon, None