Fletcher, Laura M.*; Engles, Michael†; Hammond, Billy R. Jr.†
The functional role of the macular carotenoids lutein and zeaxanthin (as well as the isomer meso-zeaxanthin) has been considered for well over a century.1 These xanthophyllic molecules, which are found primarily in green, leafy vegetables, are selectively taken up by the retina to the exclusion of other dietary carotenoids. Once there, they embed in the Henle fiber layer2 anterior to the photoreceptor outer segments. Because lutein and zeaxanthin accumulate in greatest density in the central macular region of the retina, they are together termed the macular pigments (MPs). The amount of MP in the retina is determined primarily by a subject’s diet, although this by no means explains all the individual variation. The multiple sources of influence lead to MP optical density that varies widely from person to person (ranging from so little as to be nearly immeasurable up to ∼1.5 log units of optical density).3
Macular pigment is known to absorb visible light between the wavelengths of about 400 to 520 nm, with peak absorption occurring at 460 nm.2,4 The wide variety in density among subjects thus leads to a situation where the amount of short wave light reaching a subject’s retina can vary significantly from person to person (e.g., an optical density of 0.3 equates to ~50% of 460-nm light being prevented from reaching the retina, whereas an optical density of 1.5 absorbs nearly 98%).3 This variable density filter effectively screens the foveal photoreceptors from short wave light, which has important consequences for vision and for overall eye health. The MP’s location anterior to the photoreceptor outer segments, as well as the known absorption of short wave light, has led to several hypotheses as to its function.
It has been previously demonstrated that short wave light is especially damaging to retinal tissue.5 Any reduction in actinic short wave light reaching the retina would therefore have a protective effect by reducing accumulated damage over the lifetime. This could lead to preserved functioning, especially in the foveal region.6 In addition, lutein and zeaxanthin are effective antioxidants and their ability to quench free radicals in the retina may also confer protection.7 Finally, lutein and zeaxanthin work to protect the retina against damage caused by inflammation.8 Some positive evidence for these “protection hypotheses” of MP has come from observational research linking higher carotenoid consumption to a reduced risk of age-related macular degeneration.9,10
Although a role for protecting the retina appears credible, it is unlikely that we evolved to accumulate these pigments for the protection they confer. Optical effects, which would manifest quite early in life, seem a more probable selective pressure for the evolution of mechanisms allowing for the selective accumulation of these pigments. Walls and Judd,11 for instance, suggested several optical theories for the intraocular yellow chromophores that appear to be so ubiquitous across species. They posited that filtration could reduce the effects of chromatic aberration (the so-called acuity hypothesis). They suggested that colored filters could enhance chromatic borders by selectively absorbing one side of a border more than the other, thus making the edge more distinct. Macular pigment could reduce the discomfort caused by bright lights via a reduction of the effects of glare or dazzle. Finally, MP could absorb atmospheric haze, which is predominantly short wave, thereby extending visual range outdoors. This has been termed the visibility hypothesis and has been modeled by Wooten and Hammond.12
The first three of these hypotheses (i.e., acuity, chromatic border enhancement, and glare reduction) have each been tested empirically in multiple studies.13–18 The visibility hypothesis has only been tested in one relatively small within-subjects study.19 In that study, visibility was assessed by measuring contrast sensitivity at 8 cycles per degree (cpd) with varying levels of MP (simulated with a variable path-length filter containing a solution that matched the absorbance spectrum of the MP) under light conditions that simulated atmospheric haze. The greatest improvements in contrast sensitivity (about 30%) occurred for the initial addition of 0.25 units of optical density at 30′ retinal eccentricity, followed by a plateau effect for optical densities above about 0.50 (an effect that was very similar to the predictions modeled by Wooten and Hammond12).
The idea that the use of yellow filters could improve visual function has existed for some time and is consistent with the common existence of intraocular yellow filters in many diurnal species.19 Optical effects of yellow filtering are quite obvious. For example, absolute thresholds for a yellow target (i.e., short wave deficient) are reduced when presented on a blue background and viewed through a yellow filter.20 One could question, however, whether such effects are ecologically relevant. The existing reviews of the visual effects of wearing yellow filters are quite mixed.21,22 However, roughly half of the studies reviewed by Clark21 and Wolffsohn et al.22 did report improvements. Furthermore, even in studies that, overall, produced null results, it is worth noting that some subjects exhibited substantial improvements in visual performance.
There are several potential reasons for these mixed results. One may be the optical configuration of the task. Studies that have used stimuli that closely mimic viewing situations in our natural environments have typically found improvements with yellow filters; studies that do not match the lighting conditions (e.g., sky or sunlight) or task requirements (e.g., seeing in the distance, which causes a short wave shift) often do not show improvements. It is also unlikely that yellow filters would improve vision when viewing stimuli that are scotopic and/or short wave deficient.23 Finally, the lack of consistency may be because no filter studies to date have also measured MP optical density in subjects. The MP is an effective filter of short wave light, and adding an additional filter either may have no effect on performance or may reduce the overall luminance to such a degree that performance is actually superfluous or impaired (as originally noted by Reading and Weale14).
Our atmosphere is filled with particles of various sizes, which come from both natural and human-made sources. They range in size from very small (e.g., air molecules, radius of 10−4 μm) to quite large (e.g., rain droplets, radius of about 103 μm). Collectively, these particles are referred to as aerosols.24 Light from the sun travels through our atmosphere and interacts with these particles in several ways; for example, rays may be absorbed or, more importantly for vision, scattered by these atmospheric particles. The way in which elastic scatter (i.e., scatter in which the frequency of the incident light is preserved) occurs depends on several factors: the size of the scattering particle, the distance between particles and their spatial arrangement, and the refractive index of the particles as compared with their surrounding medium.25 Particle size, as compared with the wavelength of light with which it is interacting, is likely the most important of these. In 1871, Lord Rayleigh demonstrated that when a particle in the atmosphere is much smaller than the incident wavelength of light, the light is scattered according to the inverse fourth power of wavelength (as cited in the work by Bohren and Fraser26). Thus, shorter wavelengths (i.e., “blue” light) are scattered with greater efficiency than longer wavelengths; this is termed Rayleigh scatter.
“Blue haze,” reflecting the short wave dominance of haze suspended around the horizon, when other optical factors are equal (e.g., refractive error), is the single most significant limiting factor that determines the visibility of distant objects.12 If one is viewing a distant object through a veil of blue haze, especially if that object is reflecting mid- to long-wavelength light as is typically the case when an object is viewed along the sight line (because short wave light is scattered more efficiently, the longer wavelengths reflected from an object are more likely to reach the eye), MP (or other similar yellow filters) would, hypothetically, make the object more visible by filtering the intervening haze. The theoretical treatment of the visibility hypothesis by Wooten and Hammond12 suggests that for subjects with equal Snellen acuity, a person with high MP would have a 30% increase in visual range (i.e., how far one can see) over a person with no MP (an effect found in the work by Hammond et al.19). It is reasonable, therefore, that for subjects who are required to perform difficult tasks outdoors as part of their daily job requirements (e.g., pilots), this could translate into a meaningful difference in job performance.
The present study aimed to test the visibility hypothesis of the MP in a controlled, but ecologically valid, way. Xenon light, when paired with a specialized glass filter, can very closely approximate true atmospheric haze (see Fig. 1 for spectral information). For the 12 subjects tested with the full contrast sensitivity function (CSF), it is predicted that the function will be depressed (i.e., that the area under the curve will be less) when measured in the presence of simulated haze; the CSF should be “normal” for the same six targets presented in the short wave–deficient background condition. For the sample of 27 subjects, the target contrast of a single spatial frequency (7.5 cpd) was held constant and the intensity of the simulated haze upon which it was superposed was varied. It is predicted that MP optical density will be inversely related thresholds in the simulated haze condition (i.e., positively related to the intensity required to just barely render the grating target invisible). This same grating target was superposed on a broadband xenon background as well as a short wave–deficient background. Macular pigment optical density should be inversely related to thresholds for the broadband xenon background (because of the large quantity of short wave light in the xenon spectrum); however, thresholds should be unrelated to MP optical density for the target superposed on a short wave–deficient background. The effect of the MP can be spatially obviated by measuring thresholds for the grating target on the same three backgrounds viewed parafoveally. The thresholds obtained in these conditions should be unrelated to MP optical density, providing additional support that filtration by the MP is the driving factor for target visibility.
Twelve subjects (age range, 21 to 32 years; mean age, 26.8 years; six were male) had their full CSF measured under two background conditions (“blue haze” and short wave deficient). Inclusion criteria included ability to perform study tasks, no history of relevant ocular disease, and Snellen acuity better than 20/40 (corrected; as assessed by a standard wall chart). One subject was determined to be deuteranomalous (as assessed by the Ishihara color plates), but as this was unlikely to affect the results, he remained in the study. A separate sample of 27 subjects (age range, 18 to 29 years; mean age, 21.3 years; 11 were male) had their contrast sensitivity thresholds for one spatial frequency (7.5 cpd) measured under the same background conditions, with the addition of a broadband xenon condition to simulate sunlight. The same inclusion criteria were used for these subjects. Two subjects were excluded from analysis for corrected Snellen acuity worse than 20/40. Informed consent was obtained from all subjects before any experimental procedures, and the experiments followed the guidelines of the Declaration of Helsinki as well as the University of Georgia Institutional Review Board.
The apparatus used to measure contrast sensitivity consisted of a three-channel optical system. See the study by Wooten et al.27 for a diagram and detailed explanation of the system. A 1000-W xenon arc bulb served as the light source. Channel 1 provided a uniform circular field of light; channel 2 directly illuminated a 1-degree test grating. The light from these two channels was integrated at beam-splitter 1 and could be independently regulated by a series of neutral density filters and wedges. Consequently, the contrast of the grating, as seen by the subject, could be manipulated by varying the amount of light that back-illuminated the grating to counter the amount of light that came from the diffusing channel. The two channels were counterbalanced such that a range of contrasts from 0.0 to the maximum inherent contrast of the grating could be produced while maintaining a constant luminance. Contrast sensitivity in both conditions was measured at six different spatial frequencies: 2, 4, 8, 15, 22, and 43 cpd, thus allowing a nearly complete CSF to be measured.
Light from channel 3 was passed through a chromatic filter (Schott glass Filter #BG34, UQG Optics Ltd, Barrington, NJ), which in conjunction with the spectral output of the xenon arc lamp very nearly reproduced the blue haze spectrum of skylight (Fig. 1). To simulate the “yellowing” of distant targets, two additional chromatic filters (long filter #2, Edmond Optics, Barrington, NJ) were placed in channels 1 and 2. Light from channel 3 was integrated at beam-splitter 2 with the light from channels 1 and 2. A timed electronic shutter (Uniblitz AOX5; Vincent Associates, Rochester, NY) provided exposure durations for the gratings of 3 seconds. Opening of the shutter was controlled by the experimenter. Care was taken to maintain a constant interstimulus interval of 3 seconds.
Subjects viewed the target with the right eye and had their heads stabilized via a chin-and-forehead rest assembly. Subjects were evaluated using a two-alternative forced-choice staircase method. The initial trial sequence (10 consecutive presentations of the target at a given spatial frequency) began at a clearly visible level to allow the subject to adapt to and learn the task. After the subject had become comfortable with the task, the contrast of the grating was dropped to 0.0 and incrementally increased (a typical step size of about 1.0 to 10% was used to identify the threshold region; a step size as small as 0.1% was used after the threshold region was identified) until at least 90% of the presentations were correctly identified for two consecutive experimental sets (set size, 10 consecutive trials). By starting at a low contrast level and progressively increasing the contrast, adaptation was avoided. This task was repeated until all spatial frequencies were assessed. On each trial, the target grating was tilted clockwise or counterclockwise 11.5 degrees, and the orientation was chosen randomly. The task required the subject to identify the orientation of the grating (e.g., “left” or “right”). Feedback was not provided. After a total of 40 to 60 trials, a cumulative percent-correct value for each contrast level was calculated.
Psychometric functions were generated using the graphing software Origin 7.0. The cumulative percent-correct value at each presented contrast setting for a given subject on a given grating was entered into a table and a best-fit psychometric function was rendered (upper and lower bounds were 100 and 50%, respectively). Thresholds were taken as the value that corresponded to the 75% correct point on the curve. In cases where multiple measures were made at the same spatial frequency in the same condition and in the same subject, averages were used for the purpose of analysis.
The same optical system described previously was used to obtain thresholds for the visibility of a single sine-wave grating (7.5 cpd) with varying background condition. This frequency was chosen as it represents peak sensitivity of the full CSF. The contrast within the grating was kept constant for this experiment, and the contrast between the grating and its background was varied. The target was viewed monocularly (right eye only), and head position was made stable with a combination chin-and-forehead rest assembly. Channel 3 was used to create the various backgrounds; channel 2 was used to create a 1-degree grating target; and channel 1 was used to create a fixation light for the parafoveal measures but was blocked for all foveal measures. The target channel containing the grating target was rendered short wave deficient (cutoff, 570 nm) by a chromatic filter (Corning, 51300; Oriel, Stamford, CT). As mentioned previously, a short wave–deficient target most closely resembles what is likely to happen when viewing targets at a distance outdoors.
The grating was kept at a constant radiance for the duration of the experiment (110 nW), and its visibility was tested under three different background conditions, the presentation order of which was randomized for each subject. A blue haze condition was created by pairing the xenon arc light source with the same chromatic filter used in the full CSF experiment (Schott #BG34). A broadband condition intended to mimic sunlight with the absence of haze was created with the xenon arc light source made less intense with neutral density filters. The third background condition was a short wave–deficient background (558 nm; half-power bandwidth, 8 nm; Edmund Optics, Barrington, NJ) that falls outside the absorption spectrum of the MP. This condition was included to spectrally obviate the effect of the MP.
Alternating ascending and descending thresholds were obtained using the method of limits. An average of multiple ascending and three descending trials were obtained for each subject, with a minimum of three trials in each direction or as many as five in each direction (for a total of 6 to 10 trials) depending on subject consistency when making threshold assessments. Each subject’s threshold was defined as the average background luminance at each transition point.
For the parafoveal assessment, the three background conditions were repeated while the subject fixated a red point of light placed 5 degrees nasally. The same ascending and descending method of limits procedure was used. For all subjects, the order of the background conditions was randomized, but all foveal measurements took place before the parafoveal measures.
Macular pigment optical density was measured for all subjects in the single spatial frequency component of the experiment; it was measured only in the right eye. Measurement was done using a Macular Densitometer (Macular Metrics, Rehoboth, MA), which allows for the detailed measurement of retinal levels of lutein and zeaxanthin in a noninvasive manner and in free view. The psychophysical procedure used, customized heterochromatic flicker photometry, has been extensively validated for measuring MP and has been described in detail elsewhere.27 A brief description is as follows: the test stimulus was a circular target 1 degree in size superimposed onto and presented in the center of a 6-degree, 2.75-cd/mm2, 470-nm circular background. Macular pigment is sampled at the edge of a stimulus when using HFP; thus, a target 1 degree in size will measure MP optical density at 30′ retinal eccentricity.3 The test stimulus was composed of a 458-nm measuring field (peak MP absorbance) alternating with a 570-nm, 3.0-cd/mm2 reference field (no MP absorbance). Measuring and reference fields were superposed and presented out of phase at an approximate alternation rate of 12 to 20 Hz in the foveal condition and 8 to 10 Hz in the parafoveal condition. This alternation rate was carefully optimized for each subject to create a narrow (e.g., equivalent to ~0.10 optical density) null zone. Once optimized, subjects adjusted the radiance of the 458-nm measuring field (which was presented in counterphase with the 570-nm reference and yoked to maintain a constant luminance) until the perceived flicker was extinguished. A parafoveal measure at 7 degrees was obtained by having subjects fixate a red light to the left of a 2-degree target while making their judgments. The parafoveal measure falls outside of where MP is optically measurable and thus serves as a reference point for calculating the optical density at the other retinal eccentricities.
Area under the curve for the CSFs measured under both conditions (short wave deficient and simulated blue haze) varied by a factor of about 4.5 to 6. Data were imported to Origin 7.0; each subject’s individual CSF was plotted on linear axes and a line was drawn to connect each point in a series. Area under the curve was taken as the integration of the region under each connecting line and between the lowest and highest spatial frequency measured. Fig. 2 is a plot of the average CSFs for the short wave–deficient and blue haze conditions. Notably, the overall shape of the two functions is the same; however, there is a uniform depression of the CSF measured in the presence of haze, and this depression did not significantly differ across spatial frequencies (F5,42 = 0.54, p = 0.74). Because of this uniform depression, we decided to use a single spatial frequency when assessing the relationship between MP optical density and visibility thresholds. This reduces the testing time considerably and prevents subjects from becoming fatigued or having to return to the laboratory for multiple testing sessions.
For the separate set of 25 subjects, average MP optical density was 0.46 at a retinal eccentricity of 30′ (SD = 0.13) but with a wide range (0.15 to 0.91). Recent studies have found similar averages at the same retinal locus.28,29 Mean energy in the background channel at threshold was similar for the haze (mean [±SD] = 6.36 [±1.72] μW) and xenon conditions (mean [±SD] = 6.32 [±1.56] μW) but was somewhat lower for the short wave–deficient background (mean [±SD] = 5.18 [±1.10] μW). This is likely because both the background and target had similar spectral characteristics for this condition. Pearson product moment correlations were performed to determine the association between MP optical density and log energy at threshold for each of the three background conditions; statistical significance was set at p < 0.05. Macular pigment optical density was significantly positively correlated with the log energy required to just barely render the target invisible for both the haze and the xenon background conditions (haze: r = 0.59, p < 0.01; xenon: r = 0.60, p < 0.01) (see Figs. 3 and 4, respectively). As expected, MP optical density was unrelated to log energy at threshold for the short wave–deficient background condition (r = 0.21, p = 0.31) as neither the background nor the target is influenced by the MP (Fig. 5).
Unexpectedly, MP optical density was also significantly related to log energy at threshold for all three parafoveal conditions (haze: r = 0.59, p < 0.01; xenon: r = 0.62, p < 0.01; short wave deficient: r = 0.45, p < 0.05). Semipartial correlations were conducted between MP optical density and the parafoveal thresholds (i.e., controlling for the foveal thresholds in that condition); this renders the correlations nonsignificant. This was conducted to control for individual differences in overall sensitivity as well as differences in task performance. This relationship is thus possibly driven by the very strong correlation between foveal and parafoveal thresholds for each condition (haze: r = 0.89, p < 0.01; xenon: r = 0.73, p < 0.01; short wave deficient: r = 0.65, p < 0.01), although potential additional reasons for the significant correlations are addressed in the discussion.
Scattering of light in the atmosphere places limitations on visibility over great distances by reducing the contrast between the target and the background on which it is viewed. The small particles of the atmosphere preferentially scatter short wave light causing the sky and distant scenes to be tinged blue. Wooten and Hammond12 argued that the natural short wave filters of the human eye (e.g., the MP) will improve visibility by reducing the optical contribution of this blue haze background without significantly affecting the target. The purpose of the present study was to empirically test the visibility hypothesis of MP (i.e., that selective filtration of short wave dominant atmospheric haze can extend visual range outdoors) in an ecologically valid way.
When viewing objects close-up or at intermediate distances, a viewer can rely on both chromatic and luminance edges for object detection/discrimination; however, at greater distances, edges tend to become isoluminant, and chromatic differences may be the only reliable indicator of an edge.30 Hansen and Gegenfurtner31 analyzed about 700 calibrated color images of natural scenes and found that isoluminant edges were not more or less common than pure luminance edges and that, in fact, luminance and chromatic edges occur independently of one another. Thus, chromatic edges seen at a distance are not uncommon in natural scenes. This finding is consistent with the hypothesis that the MP could confer an advantage when viewing objects at a distance. By selectively filtering the intervening haze relative to the target, this chromatic border is enhanced, leading to greater visibility.
It was found that the CSF is uniformly depressed (relative to measurements obtained in short wave–deficient light) by the presence of simulated atmospheric haze. This allowed us to test a single spatial frequency when evaluating the visibility hypothesis of MP.
When testing a single spatial frequency, it was observed that subjects with higher levels of MP required more simulated haze to lose visibility of a short wave–deficient grating target. Log energy at threshold varied by a factor of 2 between subjects with the highest and lowest levels of MP, implying that a subject with high MP optical density would be able to detect a target at a much greater distance (i.e., more atmospheric haze between them and the target) compared with a subject with lower MP optical density. Unlike the previous experiment by Hammond et al.,19 the effect of MP did not appear to plateau for subjects with high MP optical density (a plateau was also suggested by the modeling of Reading and Weale14). A major difference between external filtering (e.g., tinted glasses) and filtering within the eye itself, however, is that sensitivity losses attributed to internal filtering are compensated for by increasing visual gain mechanisms.16 Hence, the well-known loss in visual acuity with decreases in luminance is not sustained when the filter is internal to the eye.
For the broadband xenon background, which closely approximates the spectrum of sunlight, the same relationship was found. Xenon light has a significant short wave component, making filtration by the MP a possible factor when detecting the target on a xenon background. In our data, there is about a twofold difference between the subjects with the highest and lowest MP levels in the amount of energy required to lose sight of a central grating target. This is further verified by the overall positive association between MP optical density and the amount of energy in the background at threshold across all subjects. This relationship between MP optical density and log energy at threshold is not observed when the short wave–deficient target is superposed on a 558-nm background because neither the target nor the background is absorbed by the MP.
Macular pigment optical density was positively related to background energy at threshold for all three conditions when viewed parafoveally. There are a few potential nonmutually exclusive reasons for why this unexpected result may have occurred. First, the fixation point was placed only 5 degrees nasally. Although MP optical density has decreased considerably by 5 degrees, some subjects may still have sufficient MP at this spatial location to be optically significant.32,33 Five degrees was chosen because a more eccentric fixation point would have made the task excessively difficult and would have compromised reliability. There may also be a nonoptical effect of MP (e.g., physiological effects as suggested by Renzi and Hammond15). As previously noted, lutein and zeaxanthin embed in the Henle fiber layer anterior to the foveal cones. There are considerable individual differences in the number of foveal cones,34 and these differences extend out to several degrees of eccentricity. If more photoreceptors, which would presumably lead to improved spatial vision, also led to more “space” for accumulating MP, then a relationship between MP and measures of spatial vision may be seen even in tasks for which the optical effects of the MP should not directly play a role.
In sum, the present study provides support for the visibility hypothesis of the MP despite the limitation of a relatively small sample size for each portion of the study. The next obvious step is to conduct a similar study using subjects with a wide range of MP optical density in an outdoor setting for maximum ecological validity. Finding that the same relationship exists outside the carefully controlled laboratory setting would perhaps provide the impetus for conducting measurements of MP (and subsequent supplementation for subjects found to have low levels) in subjects whose visual-motor performance could be enhanced by better vision outdoors. Of additional interest is a closer examination of the potential plateau effect seen in the Hammond et al.19 study. Repeated assessment of visibility thresholds while subjects are being supplemented with lutein and zeaxanthin to increase their MP would be a more naturalistic way to assess this.
Received February 7, 2014; accepted June 27, 2014.
1. Nussbaum JJ, Pruett RC, Delori FC. Historic perspectives. Macular yellow pigment. The first 200 years. Retina 1981; 1: 296–310.
2. Snodderly DM, Brown PK, Delori FC, Auran JD. The macular pigment. I. Absorbance spectra, localization, and discrimination from other yellow pigments in primate retinas. Invest Ophthalmol Vis Sci 1984; 25: 660–73.
3. Hammond BR Jr, Wooten BR, Snodderly DM. Individual variations in the spatial profile of human macular pigment. J Opt Soc Am (A) 1997; 14: 1187–96.
4. Wyszecki G, Stiles WS. Color Science: Concepts and Methods, Quantitative Data and Formulae, 2nd ed. New York, NY: Wiley; 1982.
5. Ham WT Jr, Ruffolo JJ Jr, Mueller HA, Clarke AM, Moon ME. Histologic analysis of photochemical lesions produced in rhesus retina by short-wave-length light. Invest Ophthalmol Vis Sci 1978; 17: 1029–35.
6. Junghans A, Sies H, Stahl W. Macular pigments lutein and zeaxanthin as blue light filters studied in liposomes. Arch Biochem Biophys 2001; 391: 160–4.
7. Khachik F, Bernstein PS, Garland DL. Identification of lutein and zeaxanthin oxidation products in human and monkey retinas. Invest Ophthalmol Vis Sci 1997; 38: 1802–11.
8. Sasaki M, Ozawa Y, Kurihara T, Noda K, Imamura Y, Kobayashi S, Ishida S, Tsubota K. Neuroprotective effect of an antioxidant, lutein, during retinal inflammation. Invest Ophthalmol Vis Sci 2009; 50: 1433–9.
9. Beatty S, Boulton M, Henson D, Koh HH, Murray IJ. Macular pigment and age related macular degeneration. Br J Ophthalmol 1999; 83: 867–77.
10. Age-Related Eye Disease Study 2 Research Group. Lutein + zeaxanthin and omega-3 fatty acids for age-related macular degeneration: the Age-Related Eye Disease Study 2 (AREDS2) randomized clinical trial. JAMA 2013; 309: 2005–15.
11. Walls GL, Judd HD. The intra-ocular colour-filters of vertebrates. Br J Ophthalmol 1933; 17: 641–75.
12. Wooten BR, Hammond BR. Macular pigment: influences on visual acuity and visibility. Prog Retin Eye Res 2002; 21: 225–40.
13. Engles M, Wooten B, Hammond B. Macular pigment: a test of the acuity hypothesis. Invest Ophthalmol Vis Sci 2007; 48: 2922–31.
14. Reading VM, Weale RA. Macular pigment and chromatic aberration. J Opt Soc Am 1974; 64: 231–4.
15. Renzi LM, Hammond BR. The effect of macular pigment on heterochromatic luminance contrast. Exp Eye Res 2010; 91: 896–900.
16. Stringham JM, Hammond BR Jr. Compensation for light loss due to filtering by macular pigment: relation to hue cancellation. Ophthalmic Physiol Opt 2007; 27: 232–7.
17. Stringham JM, Hammond BR Jr. The glare hypothesis of macular pigment function. Optom Vis Sci 2007; 84: 859–64.
18. Stringham JM, Hammond BR. Macular pigment and visual performance under glare conditions. Optom Vis Sci 2008; 85: 82–8.
19. Hammond BR Jr, Wooten BR, Engles M, Wong JC. The influence of filtering by the macular carotenoids on contrast sensitivity measured under simulated blue haze conditions. Vision Res 2012; 63: 58–62.
20. Luria SM. Vision with chromatic filters. Am J Optom Arch Am Acad Optom 1972; 49: 818–29.
21. Clark BA. Color in sunglass lenses. Am J Optom Arch Am Acad Optom 1969; 46: 825–40.
22. Wolffsohn JS, Cochrane AL, Khoo H, Yoshimitsu Y, Wu S. Contrast is enhanced by yellow lenses because of selective reduction of short-wavelength light. Optom Vis Sci 2000; 77: 73–81.
23. Loughman J, Akkali MC, Beatty S, Scanlon G, Davison PA, O’Dwyer V, Cantwell T, Major P, Stack J, Nolan JM. The relationship between macular pigment and visual performance. Vision Res 2010; 50: 1249–56.
24. McCartney EJ. Optics of the Atmosphere: Scattering by Molecules and Particles. New York, NY: Wiley; 1976.
25. Engles M. The Wavelength Dependence of Intraocular Light Scattering: A Review. Saarbrücken, Germany: Lambert Academic Publishing; 2011.
26. Bohren CF, Fraser AB. Colors of the sky. Physics Teach 1985; 23: 267–72.
27. Wooten BR, Hammond BR Jr, Land RI, Snodderly DM. A practical method for measuring macular pigment optical density. Invest Ophthalmol Vis Sci 1999; 40: 2481–9.
28. Hammond BR Jr, Fletcher LM, Elliott JG. Glare disability, photostress recovery, and chromatic contrast: relation to macular pigment and serum lutein and zeaxanthin. Invest Ophthalmol Vis Sci 2013; 54: 476–81.
29. Renzi LM, Hammond BR Jr, Dengler M, Roberts R. The relation between serum lipids and lutein and zeaxanthin in the serum and retina: results from cross-sectional, case-control and case study designs. Lipids Health Dis 2012; 11: 33.
30. Horvath H, Gorraiz J, Raimann G. The influence of the aerosol on the color and luminance differences of distant targets. J Aerosol Sci 1987; 18: 99–112.
31. Hansen T, Gegenfurtner KR. Independence of color and luminance edges in natural scenes. Vis Neurosci 2009; 26: 35–49.
32. Werner JS, Donnelly SK, Kliegl R. Aging and human macular pigment density. Appended with translations from the work of Max Schultze and Ewald Hering. Vision Res 1987; 27: 257–68.
33. van der Veen RL, Berendschot TT, Makridaki M, Hendrikse F, Carden D, Murray IJ. Correspondence between retinal reflectometry and a flicker-based technique in the measurement of macular pigment spatial profiles. J Biomed Opt 2009; 14: 064046.
34. Curcio CA, Sloan KR Jr, Packer O, Hendrickson AE, Kalina RE. Distribution of cones in human and monkey retina: individual variability and radial asymmetry. Science 1987; 236: 579–82.