|
|
||||||||
1 From the University Department of Ophthalmology, Manchester Royal Eye Hospital, Manchester; the 2 Visual Sciences Laboratory, Department of Optometry and Neuroscience, University of Manchester Institute of Science and Technology, Manchester; and the 3 Department of Optometry and Vision Sciences, Cardiff University, United Kingdom.
| Abstract |
|---|
|
|
|---|
METHODS. The optical density of MP was measured psychophysically in 46 subjects ranging in age from 21 to 81 years with healthy maculae and in 9 healthy eyes known to be at high-risk of AMD because of advanced disease in the fellow eye. Each eye in the latter group was matched with a control eye on the basis of variables believed to be associated with the optical density of MP (iris color, gender, smoking habits, age, and lens density).
RESULTS. There was an age-related decline in the optical density of macular pigment among volunteers with no ocular disease (right eye: r2 = 0.29, P = 0.0006; left eye: r2 = 0.29, P < 0.0001). Healthy eyes predisposed to AMD had significantly less MP than healthy eyes at no such risk (Wilcoxons signed rank test: P = 0.015).
CONCLUSIONS. The two most important risk factors for AMD are associated with a relative absence of MP. These findings are consistent with the hypothesis that supplemental lutein and zeaxanthin may delay, avert, or modify the course of this disease.
| Introduction |
|---|
|
|
|---|
Human macular pigment (MP) consists of the two hydroxycarotenoids, lutein (L) and zeaxanthin (Z), with concentrations that peak at the center of the fovea.9 MP is an effective filter of damaging blue light, which causes photo-oxidative retinal injury, because of its absorbance spectrum and its prereceptor location.10 Further, L and Z are powerful antioxidants with the ability to quench the triplet state of photosensitizers and singlet oxygen,11 12 to react with free radicals,13 and to retard the peroxidation of membrane phospholipids.14 Consequently, it has been hypothesized that MP protects against AMD. Because the macular pigment is entirely of dietary origin,15 a protective effect would have important health care implications.
There are many variables that have been investigated as potential risk factors for AMD. These include: exposure to sunlight,4 5 16 17 18 19 20 tobacco use,16 17 21 22 light iris color,16 17 18 23 24 25 26 27 race,28 29 genetic predisposition,30 31 32 female gender,4 17 28 33 cardiorespiratory disease,16 34 35 diet,36 37 and hypermetropia.16 35 However, considerable controversy persists regarding all these putative risk factors. AMD in the fellow eye and increasing age, however, are two risk factors on which there is consensus.21 33 34 38 39 40 41 It was this observation that prompted us to correlate MP optical density measurements with age and to compare MP density measurements in the healthy fellow eye of patients with unilateral neovascular AMD with matched eyes not at high risk of development of AMD.
| Materials and Methods |
|---|
|
|
|---|
Nine white patients with advanced neovascular AMD in one eye, but no macular abnormalities in the fellow (high-risk) eye, were recruited. A diagnosis of neovascular AMD was made according to the criteria published by the International ARM Epidemiologic Study Group and required the demonstration of choroidal neovascularization in the presence of soft drusen and/or pigmentary changes.43 Each high-risk eye was matched with that of a control subject based on iris color, age (±10 years), smoking habits, gender, and lens density. Optical density of MP was then compared in a casecontrol manner. The matching process was masked to MP measurements.
The project was endorsed by the Central Manchester Local Research Ethics Committee (reference number, CEN/98/100). Informed consent was obtained from all subjects, and the tenets of the Declaration of Helsinki were observed.
Measurement of Macular Pigment
The apparatus used to derive the optical density of MP, using
heterochromatic flicker photometry (HFP), is described
elsewhere44
and is schematically represented in Figure 1
. The apparatus consists of a three-channel free-viewing
(nonMaxwellian view) optical system. Channel 1 provides the matching
stimulus with a wavelength that can be varied in increments of 1 nm
(bandwidth: 6 nm). Channel 2 provides the blue reference light, an LED,
the output spectrum of which peaks at 476 nm (calibrated with a
spectrophotometer). The blue LED was set to give a stimulus luminance
of 20 candelas (cd)/m2, thus providing retinal illumination
of approximately 2.6 log troland (td).
|
Channels 1 and 2 each contain a shutter driven by a pen motor (AL20/10; Lectromed, Letchworth Garden City, UK), thus allowing the test stimulus to alternate between the two channels. A computer program is used to control the wavelength and intensity of the matching stimulus (channel 1).
To test the validity of and establish the optimum settings for our technique of MP measurement, we performed preliminary investigations in 12 healthy subjects. Sessions began with a 5-minute period of dark adaptation, during which time the procedure was explained. The viewing distance was 30 cm, thus resulting in a stimulus size of 0.95°, and the fellow eye was occluded throughout the procedure.
With the subject fixating centrally and after increasing the luminance of channel 1 (the matching stimulus) to a level of obvious flicker at a frequency of 25 Hz, the subject was instructed to reduce the luminance of the matching stimulus until flicker could no longer be appreciated (the first point of no flicker). Then, the subject continued to reduce the luminance until the flicker reappeared, which he or she then increased until flicker was once again eliminated (the second point of no flicker). The midpoint of this range of no flicker was taken as the matching luminance. When a subject could not eliminate the flicker entirely, a range of minimum flicker was recorded. This procedure was then repeated while the subject fixated a target located at a nasal eccentricity of 6°.
The absorption spectrum for MP was generated by taking optical density measurements, with central and peripheral fixation, between 450 and 560 nm in 12 subjects, in increments of 10 nm between 450 and 480 nm and in increments of 20 nm between 480 and 560 nm. The differing wavelengths were presented in a pseudorandom fashion, and three to six readings were taken for each wavelength.
When we were satisfied that our method of measuring MP optical density was valid and reproducible, subsequent measurements involved recording the matching luminance using test stimuli of 476 and 560 nm only. In each case, five readings were taken on two separate occasions.
The principle behind MP optical density measurements derived from HFP is as follows. A blue reference light close to the spectral optical density peak of MP (476 nm) alternates with a light that is not absorbed by the pigment (say 560 nm), and flicker is eliminated when the perceived luminance of the two lights is equalized. MP reduces the relative sensitivity of the central retina at various wavelengths by a factor equivalent to the fraction of incident light that it absorbs. Therefore, as MP is optically undetectable at an eccentricity of 6.5°, the difference between the matching luminances obtained from central and peripheral viewing can be used as a measure of MP optical density.
If the luminance of the reference L476 is
constant and L
(the matching luminance)
is variable, the absorption spectrum of MP is calculated by:
![]() | (1) |
is the matching
luminance setting for central fixation,
Lp
is the matching luminance
setting for peripheral fixation, and
OD
is the optical density
difference of MP at wavelength
compared with that of the reference
476 nm.
The absorption spectrum is normalized at 560 nm, because absorption by
MP is zero at and above 560 nm. The optical density of MP is therefore
calculated by
![]() | (2) |
Dietary Carotenoid Intake
Nutritional information was obtained by interview, using a food
frequency questionnaire (FFQ)45
which was modified to take
account of published data regarding the carotenoid content of
foodstuffs46
47
so that dietary intake of L and Z could be
evaluated. No measures were taken to control for seasonal variation in
the availability of foods, because it has been shown that MP optical
density remains stable for long periods in subjects who consume a
relatively constant diet,48
reflecting the low biologic
turnover of the carotenoids in the retina.49
Because the
intake of most nutrients correlates positively with total energy
intake,50
the nutrient values were appropriately adjusted
and are expressed in milligrams per 1000 calories (kilocalories). When
nutrient supplements were used, only the dietary component of the
nutrient was energy adjusted and then added to the supplemented amount.
Energy and nutrient values were retrieved from UK food
tables.51
Dietary intake of L and Z was calculated as
nutrient content of food x portion weight x frequency of
consumption.
Statistical Analysis
Regression analysis and analysis of variance (ANOVA) were used
to test age associations. Statistical comparisons between high-risk
eyes and their matched pairs were made using the nonparametric Wilcoxon
signed rank test, or the paired Students t-test when the
data came from a normal distribution. The reproducibility and
testretest variability of MP measurements are expressed as the
coefficient of repeatability (2
2 variance within
subjects).52
This coefficient represents the value below
which the difference between two successive readings, or two sessions,
will lie with 0.95 probability.
| Results |
|---|
|
|
|---|
Macular Pigment
Absorption Spectrum of Macular Pigment.
The average absorption spectrum, and the individual absorption spectra,
of MP generated by HFP closely matched the composite curve of Wyszecki
and Stiles,53
thus confirming the validity of our
technique (Fig. 2)
. Reproducibility and intersession variability are given by the
coefficients of repeatability of 0.08 and 0.09, respectively.
|
The mean optical density of MP was 0.289 ± 0.156 (range, 0.0240.646), for the right eye. The corresponding values for the left eye were 0.299 ± 0.159 (range, 0.0310.596). There was good interocular agreement of MP optical density (simple regression: r = 0.866; P < 0.0001), with a maximum rightleft difference of 0.135, and the measurements were statistically similar for fellow eyes (Wilcoxon signed rank test: P = 0.68; Fig. 3 ). A statistically significant age-related decline in the optical density of MP was observed (right eye: r2 = 0.24, P = 0.0006; left eye: r2 = 0.29, P < 0.0001; Fig. 4 ).
|
|
=
1.56 ± 0.16; standard-risk eyes:
Lp
= 1.65 ± 0.19; Wilcoxon
signed rank test: P = 0.26), indicating that peripheral
medium- and long-wavelength cone spectral sensitivity was similar for
the matched pairs. Eight of the nine high-risk eyes had less MP than
their matched control eyes, and the difference was statistically
significant (Wilcoxon signed rank test: P = 0.015;
Table 1
).
|
|
|
|
|
| Discussion |
|---|
|
|
|---|
The mean optical density of MP in this study was 0.289 ± 0.156, higher than measurements taken in 217 subjects from Arizona (0.22 ± 0.13) in a recent study,54 but comparable with most previous reports.55 56 There are two possible explanations for the discrepancy between our results and those of the Arizona study. First, Hammond and CarusoAvery used a reference for MP measurement at only 4° retinal eccentricity, where MP is still optically detectable, and slightly lower values are therefore unsurprising. And second, all subjects reported in the current study have been living in the northwest of England since childhood, where ambient light levels are much lower than in Arizona, and it is possible that MP is depleted in response to the oxidant load arising from greater cumulative light exposure. The interindividual variability of MP measurement among our subjects, represented by a range of 0.0124 to 0.646 ± 0.156, is entirely consistent with previous reports.54 57
The current results showing an age-related decline in MP optical density are contrary to some of the early studies57 58 but consistent with the most recent.54 It is important to note, however, that the early reports took no account of recently identified variables that are believed to be related to the optical density of MP, such as iris color,59 tobacco use,60 gender,55 and lens density.61 In 2000, Hammond and CarusoAvery54 reported a statistically significant inverse relationship between the optical density of MP and age (r = -0.14; P < 0.02) among subjects living in Arizona. However, it was unclear whether the ages of the study group were related to the variables that are associated with MP density. This is of particular importance, because those associations were confirmed in that study.54 In the present study, there was no relationship between any of these variables and age. It is of interest that age and years spent in Arizona were positively correlated in Hammond and CarusoAvery, allowing for the possibility that the age-related decline in MP optical density was attributable to cumulative exposure to very high ambient levels of light. That we have reproduced this finding in subjects from a northern European population indicates that other factors play a role.
The age-related decline in MP optical density must be attributable to either inadequate uptake or excessive depletion of the retinal carotenoids. The decline of MP optical density with increasing age may simply reflect the age-related loss of photoreceptors and their axons in which L and Z are found,62 especially in view of the demonstration by Elsner et al.63 of the close spatial relationship between cone photopigment and MP distribution. However, because cones are relatively spared in age-related loss64 and because our technique of HFP excluded rod contributions, this is unlikely. Alternatively, depletion of MP may result from utilization of L and Z in response to the age-related increase in oxidant load.65 66 67 68
Given that micronutrient deficiencies are seen in 18% to 40% of the
elderly population,69
70
71
the age-related decline in MP
may be nutritional in origin. However, because dietary carotenoid
intake is difficult to measure, reflected in the wide variability of
reported values for L and Z consumption (0.84
mg/day),72
73
this is difficult to investigate. Of note, a
recent study has confirmed a significant and positive correlation
between dietary intake of carotenoids and serum levels in the elderly,
with the exception of L and Z, indicating that the macular carotenoids
may be inadequately absorbed in this age group.74
Further,
because carotenoids act synergistically with
-tocopherol and
ascorbate, deficiency of either of these vitamins results in excessive
depletion of its carotenoid coantioxidant.75
76
Beyond
dietary and absorptive factors, it is also possible that age-related
changes in carotenoid transport in blood and accumulation of L and Z in
the retina may be important.
Of nine eyes at high-risk of AMD and nine eyes at no such risk, when paired eyes were matched on the basis of variables associated with MP optical density, less MP was seen in the predisposed eye in eight cases. The bilaterality of AMD has recently been investigated in the Blue Mountains Eye Study, in which it was reported that both eyes were affected in 80% of patients with early or late AMD.77 Of 30 fellow eyes in subjects with unilateral neovascular AMD, early AMD was seen in 20 (66%), atrophic AMD in 7 (23%), and a healthy macula in only 3 (10%). The second eye of patients with unilateral neovascular AMD is at high risk of development of the condition because of the significant age-related increase in the bilaterality of neovascular AMD, even after adjusting for variables such as age, smoking, and family history.39 77 The incidence of choroidal neovascularization in the contralateral eye in cases of unilateral neovascular AMD has been estimated to lie between 28% and 35% at 4 years,39 40 78 79 with a 12% risk per annum.41 Of the nine high-risk eyes reported here, three exhibited soft drusen with pigmentary changes within 18 months of testing, and a further three showed these changes with choroidal neovascularization.
The relative absence of MP in eyes predisposed to AMD should be interpreted in the context of the excellent interocular agreement of MP measurements demonstrated in this study and in previous studies.80 81 In other words, low MP optical density in the fellow eye of a patient with neovascular AMD indicates that MP was probably absent in the diseased eye, although the latter cannot be measured because of fibrovascular scarring and loss of central vision. Further, because we matched eyes in terms of putative risk factors for AMD, which are reportedly associated with MP optical density, the observed absence of MP in the predisposed eyes appears to be an independent association with high risk for AMD. However, whether this deficiency of MP resulted in neovascular AMD in the diseased eye and will do the same in the healthy eye or is the result of subclinical disease warrants discussion.
Because MP is located within some part of the photoreceptor cell82 or its membrane,83 and because cone and rod systems appear to be functionally impaired in early AMD,84 85 it is possible that photoceptor loss in preclinical AMD may result in depletion of the pigment. However, this mechanism is unlikely to have played a role in our study, because sensitivity was similar for AMD-predisposed eyes and age-matched control eyes at the parafovea, where AMD typically begins. Further, Curcio et al.62 have shown a differential loss of rods and cones in AMD, with sparing of foveal cones and relative sparing of parafoveal cones in early disease. In brief, therefore, because the healthy predisposed eyes exhibited no clinical signs of disease and rods did not contribute to HFP measurements because we used a frequency of 25 Hz, we do not believe that photoreceptor dropout accounts for the absence of MP we observed in these eyes. This conclusion is consistent with the findings of Bone et al.,86 who found that L and Z concentrations, as measured by high-performance liquid chromatography (HPLC), were significantly reduced in the central and peripheral retina of eyes with AMD, suggesting that the loss of retinal carotenoids is not the result of the disease process.
It is interesting, however, that some predisposed eyes had greater quantities of MP than some of the standard-risk eyes. In other words, our finding does not support the view that there is a critical value below which AMD is likely to develop. Rather, the results suggest depletion of preexisting MP and are therefore consistent with the view that the retinal carotenoids are used in response to an age-related process, possibly oxidative stress.66 Clearly, the main limitation of the present study rests on the small sample size, which reflects the rarity of healthy fellow eyes in patients with unilateral neovascular AMD.
The evidence in support of the view that MP protects against AMD has been reviewed elsewhere.87 The Eye Disease CaseControl Study (EDCC) reported that a high dietary intake and high serum levels of L and Z were associated with a reduced chance of development of AMD.36 88 Parallels between several putative risk factors for AMD and an absence of MP have been observed by Hammond et al.,59 including light iris color, cigarette smoking,60 female gender,55 loss of visual sensitivity,89 and increasing lens density.61 Furthermore, reduced concentrations of L and Z have been demonstrated in the macula and whole retina of human donor eyes with early AMD compared with control subjects.90 Weiter et al.91 noted that the area of central sparing seen in cases of annular macular degeneration, including cases of atrophic AMD, correlated strongly with the lateral extent of MP, which may be due to the absorptive properties of the pigment, in that the lipofuscin fluorophore A2E is known to mediate blue-lightinduced apoptosis of RPE cells.92 93 Indeed, it has even been suggested that the focal reduction in RPE lipofuscin concentration at the fovea is attributable to the protection afforded to the photoreceptor outer segments, the phagocytosed elements of which contribute to lipofuscinogenesis94 by the MP.95 96 Although these findings are consistent with the plausible rationale that MP protects the central retina from blue light damage and oxidative stress, they should be interpreted in the context of our current and incomplete understanding of the disease and with full appreciation of the limitations of the observational nature of the studies involved.
The hypothesis that MP reduces the risk of development of AMD is particularly enticing because MP is entirely of dietary origin, thus suggesting that the most common cause of blind registration in the western World could be delayed, or even averted, with appropriate dietary modification. Hammond et al.97 have shown that dietary supplements of spinach and corn, representing approximately four times as much L and two to three times as much Z as a typical diet, result in a significant increase in the optical density of MP and the serum concentration of L in most subjects. Of note, after discontinuation of the modified diet, serum levels of the carotenoids returned to normal but MP optical density remained augmented, reflecting the low turnover of these compounds in the retina. However, the subjects involved varied in age from 30 to 65 years and are therefore not representative of the population at risk for AMD.
It remains uncertain whether the age-related decline in MP optical density or the relative absence of MP in predisposed eyes is the result of inadequate dietary intake of L and Z or some other mechanism, and this is an area that requires further investigation. To our knowledge, there are no World Health Organization (WHO) guidelines for optimal nutritional intake of specific carotenoids, and the recommendation of preparations containing these micronutrients cannot be justified on the basis of current evidence. Nevertheless, because MP is entirely of dietary origin, it seems prudent to encourage our patients to eat a balanced diet rich in fruits and vegetables, especially those that are yellow, orange, or dark green.72
In conclusion, we have shown that the two most important risk factors for AMD, age and advanced disease in the fellow eye, are associated with reduced optical density of MP. Ultimately, longitudinal studies involving serial measurements of MP and serum levels of L and Z in a large cohort of subjects are needed to establish whether supplemental L and Z augments MP in those subjects at risk of development of AMD and whether such MP augmentation can delay, avert, or modify the course of the disease.
| Footnotes |
|---|
Submitted for publication June 7, 2000; revised September 8, 2000; accepted September 28, 2000.
Commercial relationships policy: N.
Corresponding author: Stephen Beatty, Clinical Lecturer in Ophthalmology, Academic Department of Ophthalmology, Manchester Royal Eye Hospital, Oxford Road, Manchester, M13 9WH, UK. stephen{at}stiofanbetagh.demon.co.uk
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
E. Cho, S. E Hankinson, B. Rosner, W. C Willett, and G. A Colditz Prospective study of lutein/zeaxanthin intake and risk of age-related macular degeneration Am. J. Clinical Nutrition, June 1, 2008; 87(6): 1837 - 1843. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. J. Kanis, R. P. L. Wisse, T. T. J. M. Berendschot, J. van de Kraats, and D. van Norren Foveal Cone-Photoreceptor Integrity in Aging Macula Disorder Invest. Ophthalmol. Vis. Sci., May 1, 2008; 49(5): 2077 - 2081. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Schmitz-Valckenberg, K. Fan, A. Nugent, G. S. Rubin, T. Peto, A. Tufail, C. Egan, A. C. Bird, and F. W. Fitzke Correlation of Functional Impairment and Morphological Alterations in Patients With Group 2A Idiopathic Juxtafoveal Retinal Telangiectasia Arch Ophthalmol, March 1, 2008; 126(3): 330 - 335. [Abstract] [Full Text] [PDF] |
||||
![]() |
S H M. Liew, C. E Gilbert, T. D Spector, J. Marshall, and C. J Hammond The role of heredity in determining central retinal thickness Br. J. Ophthalmol., September 1, 2007; 91(9): 1143 - 1147. [Abstract] [Full Text] [PDF] |
||||
![]() |
U. E. K. Wolf-Schnurrbusch, N. Roosli, E. Weyermann, M. R. Heldner, K. Hohne, and S. Wolf Ethnic Differences in Macular Pigment Density and Distribution Invest. Ophthalmol. Vis. Sci., August 1, 2007; 48(8): 3783 - 3787. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Iannaccone, M. Mura, K. T. Gallaher, E. J. Johnson, W. A. Todd, E. Kenyon, T. L. Harris, T. Harris, S. Satterfield, K. C. Johnson, et al. Macular Pigment Optical Density in the Elderly: Findings in a Large Biracial Midsouth Population Sample Invest. Ophthalmol. Vis. Sci., April 1, 2007; 48(4): 1458 - 1465. [Abstract] [Full Text] [PDF] |
||||
![]() |
R E Hogg, R S Anderson, M R Stevenson, M B Zlatkova, and U Chakravarthy In vivo macular pigment measurements: a comparison of resonance Raman spectroscopy and heterochromatic flicker photometry Br. J. Ophthalmol., April 1, 2007; 91(4): 485 - 490. [Abstract] [Full Text] [PDF] |
||||