Abstract
Purpose.:
The controversial protective effect of macular pigment (MP), consisting of lutein (L) and zeaxantin (Z), in age-related maculopathy (ARM) and its late-stage, age-related macular degeneration (AMD) is discussed. Determinants of MP optical density (MPOD) and its relation to ARM were investigated.
Methods.:
MPOD was accessed at eccentricities of 0.5° and 2.0° from the fovea in 369 participants in the 2.6-year follow-up examination of the prospective Muenster Aging and Retina Study using dual-wavelength analysis of autofluorescence images. ARM was graded from standardized fundus photographs according to the International Classification System.
Results.:
MPOD at 0.5° and 2.0° between pairs and within single eyes was strongly correlated (P < 0.001). Smoking and body mass index showed moderately inverse associations with MPOD at 2.0°, and age was positively related to MPOD at both eccentricities. Serum L, measured at the baseline examination, was significantly associated with MPOD measured at follow-up. Likewise, use of L/Z-containing supplements raised MPOD. Crude mean MPOD increased with ascending stage of ARM. However, adjustment for influential factors and exclusion of L supplement users removed differences of mean MPOD between ARM stages. Considering further the accompanying eye, study eyes with ARM had significantly higher MPOD when the contralateral eye had AMD.
Conclusions.:
MPOD levels showed a high degree of intraindividual concordance and interindividual variability. Long-standing serum L levels, and in particular L supplementation, were the strongest determinants of MPOD. The hypothetical inverse association between MPOD and ARM stage was not confirmed.
Age related macular degeneration (AMD) is the advanced form of age-related maculopathy (ARM) and is the leading cause of blindness in the elderly.
1 Oxidative stress, which refers to tissue damage caused by reactive oxygen intermediates,
2 and retinal damage by short-wavelength (blue) light
3 have been implicated in the etiology and pathogenesis of ARM.
A pigment composed of three carotenoids, lutein (L), zeaxanthin (Z), and
meso-zeaxanthin (
meso-Z), accumulates at the macula, where it is known as macular pigment (MP).
3,4 In humans, L and Z cannot be synthesized de novo and are derived entirely from diet, whereas
meso-Z is largely derived from retinal L.
3,4 Due to its short-wavelength light screening and antioxidant properties, it is believed that MP may afford protection against the development of ARM.
3
The first follow-up examination of the prospective Muenster Aging and Retina Study (MARS) enabled the investigation of the determinants of MP optical density (MPOD) in a large group of patients with different stages of ARM and in eye-healthy controls.
The MARS Study is a longitudinal study designed to identify medical, environmental, and genetic factors with implications for the progression of ARM. From June 2001 to October 2003, we assembled a cohort of 1060 residents of the Muenster (Germany) region.
5 Eligibility criteria for the baseline examination were described in detail previously.
5 In brief, patients with ARM (drusen and/or retinal pigment epithelial [RPE] changes) in at least one eye, no or minimal lens opacity, thereby allowing good visualization of the retina, and age between 60 and 80 years were included into the study. Control subjects were volunteers, spouses, and companions of ARM patients who had no signs of ARM. Between November 2003 and August 2006, we reexamined 828 participants (85.5% of the initial cohort members eligible for reexamination). The median follow-up time was 2.6 years. We replicated the baseline examination protocol without drawing blood, including in addition the measurement of MPOD.
The recruitment and research protocols were reviewed and approved by the Institutional Review Board of the University of Muenster, and written informed consent was obtained from all study participants, in compliance with the Declaration of Helsinki.
All subjects were interviewed by a trained interviewer using a standardized risk factor questionnaire. Detailed information was obtained about demographic characteristics, smoking history, lifestyle, medical history, and the current and past use of medications and vitamin supplements, in particular those containing L and/or Z. Physical examinations included measurements of height, weight, pulse rate, and blood pressure. Blood was drawn only at the baseline examination for biochemical and genetic analyses. Serum concentrations of L and Z were measured using standard methods as described previously
5
After pupil dilatation with tropicamide 0.5% and phenylephrine 2.5%, 30° stereoscopic digital color fundus photographs were taken from both eyes at the baseline as well as at the follow-up examination as described in detail previously.
5 The presence and severity of retinal lesions were graded according to the International Classification and Grading System for ARM.
6 In accordance with the Rotterdam Study classification,
7 the range of ARM signs was stratified into five severity stages. As in previous studies,
8,9 eyes were classified as having no ARM (stage 0 to 1) or ARM (stages 2 to 3) or as AMD (stage 4).
The autofluorescence (AF) method for measuring MPOD has been described previously.
10 –14 Briefly, it is based on the AF of lipofuscin, which is present in the RPE cells.
10,15 Lipofuscin can be excited in vivo between 400 and 580 nm to emit its fluorescence in the 500–800 nm spectral range, whereas MP absorbs blue-light for wavelengths shorter than 550 nm, with a peak absorbance of 460 nm.
12
In the fovea, excitation light within the absorption range of MP is partially absorbed by the carotenoids, resulting in an area of reduced fluorescence. To measure the MPOD, the dual-wavelength approach of the AF method compares results from two excitation wavelengths that are differentially absorbed by the MP. Therefore, the dual-wavelength technique takes account of the nonuniform distribution of lipofuscin in the RPE
12 but assumes that the shape of the excitation spectrum is constant over the macular area.
10
In our analyses, quantitative imaging was performed using a retina angiograph (Heidelberg Retina Angiogrpah 1; Heidelberg Engineering, Heidelberg, Germany), modified for the measurement of macular pigment. Excitation wavelengths used were 488 nm (well absorbed by MP) and 514 nm (minimally absorbed by MP). This method has been used in clinical studies previously
1,13,14,16 and was described in detail by Trieschmann et al.
13
For this study, we present the mean MP density averaged along an annulus with retinal eccentricity of 0.5° and 2.0° degrees and width of one pixel each. MPOD results are reported in dimensionless density units (D.U.).
MPOD measurements were taken in only 609 out of 828 individuals who participated in the first MARS follow-up examination, as the two-wavelength retina angiograph became available only weeks after the MARS follow-up examinations had started. We excluded subjects with AF images of inadequate quality, most commonly due to insufficient fixation by the study subjects. Furthermore, since central geographic atrophies as well as choroidal neovascularizations affect the measurement of AF, images of eyes with these features, that is, eyes with central AMD, were excluded.
13 We evaluated the process both of image quality appraisal and suitability for analyses considering central pathologies resulting in invalid measurements by repeating it in a masked manner in 100 randomly selected participants. For image quality and suitability for analyses, we obtained a weighted κ value of 0.76 and 0.86, respectively. This indicates a good agreement for quality and a very good agreement for suitability for analyses in evaluation of MPOD measurements.
Finally, study participants with missing information in other relevant study variables were also excluded. We retained measurements in 573 eyes of 369 participants (or 60.6% of those with AF measurements) for analysis in this report. Bilateral MPOD measurements were available for 204 participants. For patient-based analyses, we consistently used one eye only, including the result of the worse eye according to the Rotterdam Classification. If both eyes had equal stages of ARM, measurements of the right eyes were used.
We used only measurements and data obtained at the 2.6-year follow-up examination of the MARS cohort, with the exception of serum L and Z measurements, which had been obtained at baseline. Pearson correlation coefficients (r) were computed within an eye and between pairs of eyes to assess the association of MPOD at 0.5° and 2.0° eccentricity. Likewise, correlations with potential determinants were evaluated. Skewedly distributed factors, such as L and Z serum levels, were logarithmically transformed to achieve more symmetrically distributed values. Descriptive comparisons were made using t-tests for continuous variables and χ2 tests for categorical variables. The impact of influential factors and confounders was evaluated by multivariable linear regression models. P < 0.05 was considered statistically significant. A statistical package (SAS for Windows; version 9.1; SAS Institute, Inc., Cary, NC) was used for analysis.
The 369 participants included in this report were elderly (mean age 71.6 years) and more often female (61.5%). Nearly 50% of the eyes were free of ARM (stage 0 to 1 according to the Rotterdam Study classification) while slightly >50% had signs of ARM (stage 2 to 3 according to Rotterdam Study classification). The proportion of contralateral eyes with AMD (stage 4, according to Rotterdam Study classification) in each stage of the disease is shown in
Figure 1. The average MPOD was 0.57 D.U. at 0.5° and 0.16 D.U. at 2.0° from the center of the fovea. Other characteristics of the study participants are contained in
Table 1.
The intraindividual, between-eye correlation was very strong as MPOD in one eye was closely related to MPOD in the other eye. Based on the 204 subjects where MPOD measurements were available in both eyes, the correlation coefficient in pairs of eyes for MPOD at 0.5° was
r = 0.83 and for MPOD at 2.0° it was
r = 0.86 (
P < 0.001). MPOD showed also substantial intraocular correlations in that MPOD at 0.5° related closely to MPOD at 2.0° (
Fig. 2). The correlation coefficient, based on 573 single eyes, was
r = 0.58 (
P < 0.0001).
Table 2 contains the results of correlation analyses between influential factors and MPOD. Only factors with statistically significant correlations are presented. We noted that MPOD at 0.5° and 2.0° significantly increased with age whereas it significantly decreased with body mass index (BMI) at 2.0°. Other factors investigated, including blood pressure, showed no statistically significant relation with MPOD. Serum levels of L and Z, measured in blood drawn at baseline, were strongly associated with MPOD measured at follow-up 2.6 years later. The correlations were more pronounced in MPOD at 2.0° than at 0.5° (see
Table 2).
Women and persons who had never smoked had significantly higher averaged MPOD levels than men and former and current smokers in 2.0°, respectively (see
Table 3). Individuals who ever used supplements with L and/or Z had higher MPOD, at both 0.5° and at 2.0° (each
P < 0.0001) (
Table 3). Women significantly more often took supplements containing L (28.6%) than men (15.5%) (
P = 0.0038).
In multivariate analyses, which simultaneously incorporated the influential factors, most of the associations described above remained stable (
Table 4). However, the impact of serum Z on MPOD disappeared almost entirely when serum L was included in the models. The association with baseline serum L remained highly significant even when supplement use was included in the model (
Table 4). On the other hand, exclusion of the 87 individuals with supplement intake even strengthened the association between baseline serum L (logarithmically transformed) and MPOD (regression coefficients 0.124 at 0.5° and 0.049 at 2.0°, each
P < 0.0001).
We evaluated the association of MPOD with ARM in a three-step approach. In crude analyses, disregarding any potential confounding by influential factors, MPOD increased with ascending stage of ARM in the analyzed eye for both 0.5° and 2.0° (
Table 5). Adjustment for influential factors, including supplement use, markedly attenuated this association. Restriction of the adjusted analysis to subjects without known use of supplements containing L (
n = 282) resulted in further attenuations and loss of significance for differences in MPOD at 0.5° or 2.0° between different stages of ARM in analyzed eyes (
Table 5).
We further evaluated the mean MPOD in the study eye, this time taking into account the severity of ARM in the opposite eye. Adjusted mean MPOD at 0.5° in ARM-healthy eyes (combined stages 0 and 1) was not significantly different regardless of whether no ARM, ARM, or AMD was present in the opposite eye (
Fig. 3A). Likewise, mean MPOD at 0.5° in eyes with ARM was also found to be similar when the opposite eye had ARM or no ARM. Only when the contralateral eye had AMD did we observe significantly higher MPOD values in the study eyes with ARM (
Fig. 3B). Of note, mean MPOD at 2.0° in study eyes with ARM or no ARM was not significantly different irrespective of the ARM stage in contralateral eyes (
P = 0.0628–0.9808, respectively).
In our study, crude MP optical density measured at eccentricities of 0.5° and 2.0° slightly increased with ascending stage of ARM. However, this observation was entirely accounted for by influential factors and particularly by use of L-containing supplements. Thus, in an adjusted analysis we detected no differences in MPOD at 0.5° and 2.0° between ARM-healthy eyes and those with different stages of ARM.
The hypothesis that MP protects against ARM is based on the assumption that MP acts as a direct antioxidant as well as a filter of blue, high-energy radiation in the human retina.
3 MP accumulates in high concentrations in the retina and generally peaks at the center of the macula. Typically, the MPOD reaches its half-peak OD at an average of only 1.03° (0.3 mm) retinal eccentricity.
17
Several studies support the hypothesis that MP protects against ARM: Differences in MP levels were observed between donor eyes from subjects with and without AMD
18 as well as between subjects with and without ARM, respectively AMD, measured in vivo.
19,20 On the other hand, several studies found no protective effect of in vivo measured MPOD on different stages of ARM.
8,9,16,21,22 In two of these analyses, ARM stages were classified analogous to our approach, resulting in no differences between ARM-healthy eyes and different stages of ARM.
8,9 One of these studies also found no protective effect of MPOD when considering the incidence of ARM over almost 10 years in a population-based longitudinal study.
8
In concordance with other studies, we found a high degree of agreement between pairs of eyes
23 –25 with interindividual variability in levels
16,21,23 and spatial distribution
16,25,26 of MPOD. Therefore, a relevant genetic regulation of retinal MP levels may be supposed, as confirmed by a study of MPOD measurement in mono- and dizygotic twins,
23 which is modifiable by ingestion of L and Z.
13,19,27
Intake of L is reflected in serum levels of L, which remain stable over a week or two.
28 Although blood for analyses of serum L in this study was drawn 2.6 years before MPOD measurements were performed, we still found a highly significant, positive association between MPOD and serum levels of L as reported in previous studies.
13,29 We suppose that this is explained by rather stable dietary habits in our elderly study participants.
Use of supplements containing L results in highly elevated levels both of serum L and MPOD, except for so-called nonresponders.
13 Thereby, elevated MPOD could persist even if supplementation of L had been stopped several months before, suggesting a very slow turnover of carotenoids within the retina.
13 We suggest that this may play a role in understanding our finding of elevated levels of MPOD at 0.5° in study eyes with ARM and AMD in the opposite eye. It is well known that patients affected by AMD often consume supplements containing L. Despite our efforts to obtain valid information about current and former supplement intake by standardized interviews, some underreporting may have occurred, and the effects of supplement intake stopped weeks or months before our examination may still have persisted as raised MPOD. This explanation also seems plausible because the observed differences were rather pronounced, and short-term effects of this magnitude could have been caused most likely by supplements containing L.
Our analysis of the impact of severity of ARM and AMD, respectively, in the opposite eye on MPOD in the study eye is comparable with those of Obana et al.
20 Contrary to our results, these authors found a slight decrease of MPOD measured by resonance Raman spectroscopy in eyes with no ARM or ARM when classifying according to ascending stage of ARM and AMD in the opposite eye. Importantly, the data of Obana et al. were not adjusted for influential factors.
The results of the present study show no significant gender differences in MPOD in adjusted models. The initially higher MPOD values found in women could be explained by their more frequent supplement use. These results are in agreement with former studies that found no gender differences in MPOD,
10,16,21 although several studies found higher MPOD in men
30,31 or in women.
32
In our study, MPOD slightly increased with age. The association between age and MPOD is controversial. Some studies find no age dependency of MPOD,
23,33 but others a decline
16,19,31 or an increase of MPOD with age.
10,21,23 Taking into account the limited age range of the participants of the present study (62 to 85 years), our results are comparable with those of Berendschot et al., who found a significant, positive age effect on MPOD in 435 subjects with and without ARM aged 60 to 91 years, as well.
21,33
The observed increase of MPOD with age might be explained by changes of fluorescence at Bruch′s membrane with age, which may alter the fluorescence spectra of the posterior layers of the retina and therefore affect the MP estimates.
10 Even though these changes occur throughout the posterior pole and may therefore not have marked effects on MP estimates,
10 an impact of these age-related changes on MPOD cannot be completely excluded and might play a role in explaining the observed age relationship.
We found slightly lower MPOD in former and current smokers than in persons who never smoked. In adjusted models, these differences were no longer significant. Most studies report lower MPOD levels in smokers,
10,31 while others find no differences depending on smoking.
29,30,33 Thereby some authors assume a dose-dependent effect of smoking on MPOD.
29,31 To differentiate current smokers according to number of cigarettes smoked per day was not sufficiently possible in our elderly study population with only a few remaining current smokers (
n = 13).
BMI was negatively associated with MPOD in our study. This association remained significant for MPOD at 2.0° in adjusted models. Only few studies report on relations of BMI and MPOD. They describe either no association between MPOD and BMI
32 or decreasing levels of MPOD with ascending BMI as reported in our study.
34,35 This could be explained by the fact that up to 80% of the total carotenoids in the body are found in adipose tissue,
36 maybe resulting in lower storage of MP in the retina in obese subjects.
The strengths of the present study are its large number of participants with and without ARM, the detailed information available, and the highly standardized way in which the MPOD measurements and the fundus photographs were evaluated. Limitations are, on the one hand, the cross-sectional study design with prevalent cases of ARM. We attempted to account for this by also considering the stage of ARM of the opposite eye. On the other hand, the spatial distribution of MP might be more relevant than MPOD measurements in defined areas or at the center of the fovea. We present here MPOD measured at two eccentricities from the center of the fovea, which is believed to capture at least some of the spatial distribution of MP.
In conclusion, we found that age, smoking, and BMI exert a weak effect and L serum levels, mostly due to supplementation, a strong effect on foveal MPOD. Our study results are not compatible with a hypothetical protective effect of MP in ARM. However, more detailed measurements of the spatial distribution of MP may help to better understand the role of MP. Likewise, longitudinal studies are needed as they are better suited to analyze the effect of MPOD on ARM occurrence and progression.
Supported in part by grants from Deutsche Forschungsgemeinschaft HE 2293/5-1, 5-2, 5-3; the Intramural IMF fund of the University of Muenster; the Pro Retina Foundation; and the Jackstaedt Foundation.