Profiles of Macular Pigment Optical Density and Their Changes Following Supplemental Lutein and Zeaxanthin: New Results from the LUNA Study

Abstract

Purpose.: Based on latest analyses disclosing an inverse association between ring-like structures in macular pigment (MP) spatial profile and age-related macular degeneration, we performed additional analyses of MP measurements obtained in participants of our earlier lutein nutrition effects measured by autoflourescence (LUNA) study to disclose if oral lutein (L) and zeaxanthin (Z) can attenuate, amplify, or generate a ring structure.

Methods.: A total of 97 subjects attended the last follow-up visit 3 months after discontinuation of a 6-month trial of 12 mg L and 1 mg Z supplementation. Of the subjects, 11 eyes had a secondary peak (ring-like structure) and 8 had an implied pericentral plateau/shoulder on the slope of MP density profile (intermediate distribution).

Results.: L and Z intake led to a general shift toward higher MP values in eyes without ring structure. The difference between mean optical density of retinal MP (Diff MPOD) at last follow-up and baseline was +0.16 density units (D.U.) at 0° eccentricity. Increments at 0.25°, 0.5°, 1°, and 2° (all P < 0.0001) decayed exponentially with higher eccentricity. MPOD showed comparatively slight central changes in eyes with ring and intermediate distribution (diff_MPOD at 0° +0.03 and +0.09), and increased at minimum (+0.06, P = 0.01) and maximum (+0.07, P = 0.02) of the ring, and at inner (+0.07, P = 0.04) and outer (+0.09, P = 0.01) radius of the pericentral “shoulder.”

Conclusions.: Ring structures were neither attenuated nor generated de novo following supplementation. Individuals with second peak/implied plateau in the slope of the profile appear to have the most effective retinal stabilization of L and Z located at a pericentral rather than the central location.

Introduction

The three carotenoids lutein (L), zeaxanthin (Z), which are both of dietary origin, ¹ and meso-zeaxanthin (MZ), which is derived mainly from retinal lutein, ² accumulate in the macula where they are referred to as the macular pigment (MP). ^3–6 As these substances have (blue-) light-screening and antioxidative properties, they are believed to protect against the development of age-related macular degeneration (AMD), the leading cause of blindness in the developed world. ⁷  

L and Z consumption is known to increase serum levels of these carotenoids ⁸ and supplementation also has been shown to raise the optical density of retinal MP (MPOD). ⁹ Recently, an inverse association between levels of serum L and Z, and risk for AMD has been demonstrated. ¹⁰ It was noted that some of the risk factors of AMD are related to low ocular MP, including age, ^3,11 tobacco and drinking habits, ^12–14 obesity, ⁵ and genetic background with family history of AMD. ^15–17 However, investigations of a hypothetical inverse association between MPOD concentration and AMD occurrence have been quite contradictory. Some studies found higher MPOD in persons with healthy maculae than in those afflicted with AMD, ^3,18 while others, however, could not confirm this relationship. ^19,20 The relative spatial distribution of MP might be of greater importance for the putative protective effect against the disease than the absolute amount of MP in single locations. In histologic ²¹ and clinical ^22,23 studies a wide variation of the spatial distribution of MP has been demonstrated, and several investigators described ring-like structures with a secondary peak (“ring”) at a distance of about 0.7° from the fovea in the slope of the MPOD profile. ^24,25 These structures, characterized by an annulus of higher MPOD superimposed on a central exponentially declining MP distribution, recently have been found to be inversely related with AMD and associated positively with a more favorable state of health. ²⁶ Moreover, recent supplementation studies reported on the disappearance of central “dips” (defined by a dip of MPOD at 0.25°, a rise at 0.5°, and then a steady decline to the periphery ²⁷ ) in the MP distribution following augmentation of the carotenoids in the center of the fovea. ^{28, 30}  

These findings encouraged us to analyze additionally the different patterns of MP distribution and their thinkable dietary modification. Ring-like structures in the MP distribution being inversely related to AMD, ²⁶ it is of great clinical interest to find out whether this favorable state of macular health can be generated “de novo” when a no-ring distribution is preexisting. Is a ring structure in the MP distribution a predetermined pattern of MP arrangement or can it be produced by oral supplementation of the carotenoids L and Z? Alternatively, as other studies suggest, ^29,30 do focal minima in the slope of the MP profile (as present in subjects with ring-like structures) rather disappear following a carotenoid intake so that a profile with a strictly exponential decrease of MP with eccentricity results? Based on the aforementioned new findings, we performed additional analyses of MP measurements obtained in participants of our earlier lutein nutrition effects measured by autoflourescence ⁹ (LUNA) study. ²⁶ We investigated the profiles of MP and their possible changes in response to supplements consisting of the macular carotenoids and co-antioxidants. Special attention was directed to the preexistence, alteration, or even new occurrence of ring-like structures in the MP profile to disclose if an individual pattern of MP distribution can be modified by oral L and Z. 

Methods

Subjects

Eligibility criteria have been described in detail previously. ⁹ Our study was approved by the local medical ethics committee, and adhered to the tenets of the Declaration of Helsinki. Informed consent was given by each subject before his/her involvement. 

For the duration of six months, the 108 subjects who were in the supplementation group (the intervention arm of the study, I) received a daily supplement that consisting of 12 mg esterified L and 1 mg esterified Z, 120 mg vitamin C, 17.6 mg vitamin E, 10 mg zinc, and 40 μg selenium. This supplement is available commercially as Ocuvite Lutein (Bausch & Lomb, Berlin, Germany). The mean age ± SD of subjects was 71.5 ± 7.1 years, ranging from 51 to 87 years, and there were 68 female and 40 male subjects. 

When the baseline measurement was performed, only one eye was selected for investigation for each subject. We defined as study eye the one with the better image quality; if equal, the eye with the better visual acuity was selected, and where there was no discrepancy between fellow eyes in terms of autofluorescence (AF) image quality or in terms of visual acuity, the right eye was selected arbitrarily. 

Of the 108 volunteers in the I group 100 (92.6%) showed features of AMD, including predominantly drusen (60%), noncentral RPE proliferation (33%), and atrophic changes (7%). The remaining eight subjects had healthy maculae. 

A control group (C) consisted of 28 subjects 57 to 83 years of age, mean ± SD 71 ± 8.1 years. Among these were 12 male and 16 female subjects. In the C group 25 subjects exhibited features of AMD (soft drusen, noncentral RPE proliferation and/or atrophic changes) and 3 had healthy maculae. The I and C groups were similar in terms of age and sex. ⁹ The subjects in the C group did not receive any supplement or controlled dietary changes. 

This study was a nonrandomized, open label, controlled study. Participants of the I arm of the study were examined at baseline (visit 1), and 6 (visit 2), 12 (visit 3), 18 (visit 4), and 24 (visit 5) weeks (when the supplementation was discontinued), and once again 3 months following discontinuation of the supplement (visit 6). To answer the question whether preexisting ring structures in the MP distribution can be modified via carotenoid supplementation, we analyzed and presented the MP profiles in MPOD measurements performed at baseline (visit 1) and three months after discontinuation of the supplement (visit 6). At the last follow-up visit, 97 (89.8%) probands of the intervention arm were investigated. The control group was investigated twice only, at baseline and at mean 29.4 ± 9.3 weeks following recruitment into the study. Of the 28 subjects in the control group 26 attended the second follow-up. 

Our earlier analyses already had shown that the increase in MPOD continued until at least 3 months after discontinuation of supplements. ⁹ The highest values of MPOD during follow-up were measured at this point of time. To present data of highest possible significance that can be compared to our earlier reports, ⁹ we again analyzed MP profiles 3 months after stop of intake and compared them to the baseline findings. 

Measurement of Macular Pigment

We used a Heidelberg retina angiograph (Heidelberg Engineering, Heidelberg, Germany), modified for fundus AF, for analysis. The current AF method uses two excitation wavelengths that are absorbed differentially by MP. ³¹ This method has been described elaborately, ^9,22,32 and used in clinical studies previously. ^{9,16,20,26,32–35} It measures MPOD by determining the attenuation of the fluorescence of lipofuscin caused by MP. The lipofuscin that is located in the RPE cells can be excited in vivo at 400 to 580 nm and then emits its fluorescence in the 500 to 800 nm spectral range. MP absorbs blue light with wavelengths shorter than 550 nm with a peak absorbance at 460 nm, ³² The system software creates density maps of MPOD by comparing images taken with 488 nm excitation light that is absorbed well by the MP, and with 514 nm that is absorbed poorly by MP but does excite lipofuscin. Digital subtraction of the log AF images leads to MP density maps in which the mean relative MPOD values are calculated in concentric annula (1 pixel wide) centered on the fovea. To depict the individual profile of MP distribution, we located these circles at eccentricities of 0°, 0.25°, 0.5°, 1.0°, and 2°. The reference P (defining the offset for no MP) was selected at an eccentricity of 6° as the MP profile normally plateaus for eccentricities larger than 4°. ¹  

All MPOD measurements had been performed by the primary investigator (MZ) using the same testing device and protocol throughout following pupil dilation with 5% tropicamide and 2.5% phenylephrine to a diameter of at least 6 mm. 

MP Density Profile and Ring-Like Structure

The profile of MP distribution was displayed graphically by plotting the mean MPOD values for each radius against the distance to the fovea (Fig. 1). In some cases the MP distribution does not decrease monotonously from the center of the fovea to the periphery but shows a bimodal pattern, ^24,25 with a secondary peak of increased MP density on the slope of the profile. In these graphs, MP density ring patterns are visible with a central accumulation of MP surrounded by a small trough of lower MP, and then followed more eccentrically by a “hump” of higher MP concentration. Eyes exhibiting this minimum-maximum pair of MPOD values were graded visually by the examiner as having “ring like” structures (Fig. 2).We classified eyes as having “intermediate distributions” when the distribution of MP was not decaying monotonously as a function of eccentricity, but showed an implied pericentral “shoulder” or even plateau on the slope of the profile without an explicit minimum-maximum pair. ²⁶ Whenever an eye demonstrated an MP distribution characterized by a strictly monotonic decline from the center of the fovea to the periphery without any plateauing or bimodal pattern, it was defined as a “no ring” distribution. Presence of ring-like structures was classified and assigned based on the baseline MPOD measurements. 

Figure 1. 

 

“No ring”: radial density profiles of MPOD and associated MP density maps without ring-like structures at baseline (a, c), and 3 months after stop of supplementation (b, d). MPOD shows strictly monotonic decline from the center of the fovea to the periphery without any plateauing or bimodal pattern.

Figure 2. 

 

“Ring” and “intermediate” type: radial density profiles of MPOD and associated MP density maps. Intermediate distribution in density profile (a) and density map (c) with an implied plateau on the slope of the density profile. Ring-like structure (visible in density map, d) with secondary peak of MPOD and an explicit minimum-maximum pair in density profile (b).

In addition to the measurements at eccentricities of 0°, 0.25°, 0.5°, 1.0°, and 2°, in eyes with ring-like structures MPOD was measured at the second peak, the “maximum,” and at its “minimum” between central and second peak. The respective eccentricities at which the “minimum” and “maximum” of the ring occurred also were documented. In eyes classified as having “intermediate distribution” we measured MPOD at the inner and outer radius of the implied pericentral plateau, and documented the respective eccentricities. 

Statistical Analysis

The MEDCALC for Windows package (Version 9.4.2.0, SAS Institute Inc., Cary, NC) was used for statistical analysis. We calculated the mean values ± SD for MPOD and its changes as differences between baseline and follow-up in the specific categories. We calculated the difference between mean MPOD at last follow-up (in the I group three months after discontinuation of the intake, and in the C group at 29.4 ± SD 9.3 weeks) and baseline (Diff MPOD). Changes of MPOD were assessed using paired t-tests, while comparisons between groups used unpaired statistical methods. A P value of <0.05 was considered statistically significant. 

Results

At the last follow-up visit 3 months after discontinuation of supplementation, 97 eyes of 97 (89.8%) participants were examined, with 11 (11.3%) of them showing a ring-like structure and 8 (8.2%) an “intermediate distribution” at baseline. Of the 28 subjects in the control group 26 attended the second (and last) follow-up at mean 29.4 ± 9.3 weeks following recruitment into the study. Among these, at baseline 4 eyes were classified as having a ring structure and 4 as “intermediate distribution,” and the remaining 18 showed a no ring distribution with a monotonous decrease of MP concentration from the center of the fovea to the periphery. Of interest, all but one of the “C” volunteers exhibiting a ring or intermediate distribution were women. 

Baseline Findings

Individuals in the I group showed the following baseline characteristics in MP concentration: Individuals with no ring (n = 78, n = 48 females) had a mean baseline MPOD at 0°, 0.25°, and 0.5° eccentricity of 0.8 ± 0.24, 0.61 ± 0.21, and 0.51 ± 0.19 density units (D.U.), respectively (Fig. 3a, dashed line). In individuals with no ring (n = 18, n = 9 females) who were enrolled in the control group the mean baseline MPOD at 0, 0.25°, and 0.5° were slightly lower (0.74 ± 0.25, 0.57 ± 0.20, and 0.49 ± 0.19 D.U., respectively; Fig. 3d, dashed line). 

Figure 3. 

 

MPOD at baseline and following supplementation. Left column: Mean baseline (dashed line) MPOD values (D.U.) for eccentricities of 0°, 0.25°, 0.5°, 1°, and 2° of eyes from study participants in the intervention arm of the study without ring structure (n = 78, n = 48 females; a). Additional values for MPOD at mean eccentricity (°) of the minimum and maximum of the ring (n = 11, n = 8 females; c), and at mean inner and outer radius of the pericentral plateau of eyes with intermediate distribution (n = 8, n = 4 females; b). Continuous lines: the according MPOD values at the final follow-up visit three months after stop of the carotenoid intake. Right column: mean baseline (dashed line) MPOD values (D.U.) for eccentricities of 0°, 0.25°, 0.5°, 1°, and 2° of eyes from volunteers in the control group without ring structure (n = 18, n = 9 females; d). Additional values for MPOD at mean eccentricity (°) of the minimum and maximum of the ring (n = 4, n = 3 females; e), and at mean inner and outer radius of the pericentral plateau of eyes with intermediate distribution (f). The according MPOD values at the second (and last) follow-up at mean 29.4 ± SD 9.3 weeks following recruitment into the study are depicted in continuous lines.

By comparison, in eyes with a ring-like structure of subjects designated I (n = 11, n = 8 females) baseline MPOD values at 0°, 0.25°, and 0.5° eccentricity were lower: 0.68 ± 0.37, 0.47 ± 0.29, and 0.35 ± 0.22 D.U., respectively (Fig. 3b, dashed line). In eyes of control volunteers with a ring-like structure (n = 4, n = 3 females) the mean baseline MPOD valuess at the aforementioned eccentricities were higher compared to those in the I group, but also compared to those of no ring individuals in the C group: 0.89 ± 0.15, 0.72 ± 0.13, and 0.64 ± 0.2 D.U., respectively (Fig. 3e, dashed line). 

In the I arm of the study, the corresponding mean baseline MPOD values of eyes with intermediate distribution (n = 8, n = 4 females) were mostly in between the values found in eyes with and without ring, with 0.77 ± 0.31, 0.54 ± 0.21, and 0.46 ± 0.19 D.U., respectively (Fig. 3c, dashed line). In eyes with intermediate distribution that were part of the control arm of the study (n = 4, n = 4 females; Fig. 3f, dashed line) the mean baseline MPOD values at these eccentricities were 0.89 ± 0.26, 0.65 ± 0.28, and 0.58 ± 0.24 D.U., which also were in between the values in the C eyes with and without ring. In eyes with a ring-like structure among subjects who attended the I arm of the study, the mean eccentricity of the radius of the second peak (“maximum”) was 0.92° ± 0.23°, while the minimum was at 0.48° ± 0.18°, with corresponding mean MPOD values of 0.38 ± 0.19 and 0.31 ± 0.19 D.U., respectively (Fig. 3b, dashed line). In the control eyes the minimum and maximum of the ring were at similar locations (mean 0.49° and 0.91°, respectively) with higher mean MPOD values of 0.57 ± 0.21 and 0.64 ± 0.2 D.U., respectively (Fig. 3e, dashed line). In I eyes with intermediate distribution, mean MPOD values at the inner and outer radius of pericentral “shoulder” were 0.46 ± 0.17 and 0.46 ± 0.16 D.U., respectively, and the respective eccentricities were 0.38° ± 0.18° and 0.72° ± 0.15° (Fig. 3c, dashed line). In control eyes with intermediate distribution mean MPOD values at the inner and outer radius of pericentral “shoulder” were higher than in I eyes (0.56 ± 0.21 and 0.56 ± 0.25 D.U., respectively), and the respective eccentricities also were minimally higher than in I (0.41° and 0.75°, Fig. 3f, dashed line). 

Among I volunteers, in all three groups of MP distribution mean MPOD eccentricities of 1° and 2° from the center were similar (ring distribution 0.34 ± 0.21 and 0.13 ± 0.08 D.U., respectively; intermediate distribution 0.39 ± 0.18 and 0.13 ± 0.08 D.U., respectively; and no-ring distribution 0.34 ± 0.14 and 0.12 ± 0.06 D.U., respectively (Figs. 3a–1552c, dashed line). In control eyes with no ring distributions, these values were lower than in the two other types of distribution, but similar to those in the I group (ring distribution 0.59 ± 0.16 and 0.23 ± 0.1 D.U., respectively; intermediate distribution 0.5 ± 0.25 and 0.18 ± 0.1 D.U., respectively; and no-ring distribution 0.34 ± 0.13 and 0.13 ± 0.06 D.U., respectively; Figs. 3d–f, dashed line). In I eyes with ring-like structures, the maximum was 55.6% and the minimum 46.3% of the peak MPOD, respectively. In these eyes, the ratio of MP density at the ring's maximum to that at its minimum was 1.2. The MPOD at outer and inner radius of the pericentral “shoulder” in eyes with intermediate distribution was 60% of the peak MPOD, and the ratio of MPOD at outer and inner radius was 1. In control eyes with ring structures maximum and minimum were 71.9% and 64.3% of the peak MPOD, respectively, and maximum-to-minimum ratio was similar to that in I (1.12), while in intermediate eyes the MPOD at outer and inner radius of the pericentral “shoulder” was 63.6% of the peak MPOD, minimally higher than in I eyes with the same MP distribution. 

Response to L and Z Supplementation

A ring structure or at least an intermediate distribution was not generated de novo from a preexisting no ring distribution. Following the supplementation, none of the eyes that had been characterized as having no ring at baseline exhibited a second plateau or shoulder, or even a minimum-maximum pair with a secondary peak of increased MP density on the slope of the profile. In no case did the intake of L and Z lead to the development of a bimodal pattern ^24,25 of the MP profile. After 6 months of supplementation, the changes in the profile of MPOD distribution in individuals without ring structure at baseline indicated a general shift toward higher MPOD values (Figs. 1, 3a, continuous line). The biggest change was observed in the central values (Diff_MPOD = +0.16 D.U. at 0°, P < 0.001). Similarly, the rises at 0.25°, 0.5°, 1°, and 2° were highly significant (all P < 0.0001), but Diff_MPOD diminished with increasing eccentricity (Diff_MPOD at 2° +0.028 D.U., P = 0.002, Fig. 4, black dotted line). 

Figure 4. 

 

Changes of MPOD: Diff_MPOD was calculated as the difference between mean MPOD at last follow-up (in the I group three months after discontinuation of the intake, and in the C group at 29.4 ± SD 9.3 weeks) and baseline. Mean difference between MPOD at last follow-up and baseline for the three groups by eccentricities of 0°, 0.25°, 0.5°, 1°, and 2°, and mean baseline eccentricity (°) of the minimum/maximum of the ring or of the inner/outer radius of the pericentral plateau. Black dotted line: supplemented individuals with “no-ring” distribution, n = 78. Black continuous line: individuals with intermediate distribution, n = 8. Black dashed line: individuals with ring-like structure in the MP distribution, n = 11. Gray dotted line: control individuals with “no-ring” distribution, n = 8. Gray continuous line: individuals with intermediate distribution, n = 4. Gray dashed line: individuals with ring-like structure in the MP distribution, n = 4.

Compared to the no ring eyes with strictly monotonous decline of the density profile graph from the center of the fovea to the periphery, the eyes with plateauing or bimodal pattern in the density profile graph at baseline showed completely different reaction to the L and Z intake with low MP increment in the center and highest increase of MP at the maximum of the ring (or outer radius of the pericentral plateau in intermediate eyes). 

In contrast to “no ring” eyes, individuals with intermediate distribution (Fig. 3c, continuous line; Fig. 4, black continuous line) of MP and with ring-like structures (Fig. 3b, continuous line; Fig. 4, black dashed line) showed only minor changes in the center (Diff_MPOD at 0° +0.09 and +0.03 D.U., respectively). These increments were not statistically significant. Therefore, the MPOD increase in the center was significantly higher in eyes without than in eyes with ring-like structures (P = 0.028). 

Figure 4 depicts Diff_MPOD for the three groups of eyes by eccentricities. In individuals with ring-like structures both the radius of maximum and minimum MPOD moved marginally toward higher eccentricities (from 0.92° to 0.94° and from 0.48° to 0.49° degrees, respectively; Fig. 3b, continuous line). 

Of note, when ring-like structures were detectable at baseline they were not removed or attenuated by supplementation. Supplementation did not result in the disappearance of focal minima in the slope of the MP profile (as present in subjects with ring-like structures), but instead led to a very mild amplification of preexisting ring-like structures. The mean MPOD at maximum and minimum increased from 0.38 ± 0.19 to 0.45 ± 0.24, by +0.07 ± 0.05, and from 0.31 ± 0.19 to 0.37 ± 0.25, by +0.06 ± 0.07 D.U. (P = 0.001 and 0.016, respectively; Fig. 3b, continuous line; Fig. 4, black dashed line). Following supplementation of eyes with ring-like structures, the maximum and minimum increased from 55.6% at baseline to 62.8% and from 46.3% to 52.4% of the peak MPOD, respectively. The maximum-to-minimum ratio in these eyes remained stable (baseline 1.2, following supplementation 1.19). In eyes with intermediate distribution of MP, the outer and inner radius of the pericentral “shoulder” shifted apart slightly, which led to a very mild broadening of the plateau (mean eccentricity of outer and inner radius at baseline 0.72° and 0.39°, respectively, and after supplementation 0.75° and 0.35°, respectively). On average, supplementation resulted in the development of a, yet very slight, tendency toward minimum-maximum pair of pericentral MPOD when the indeterminate stage of an intermediate distribution was present. The mean MPOD at outer and inner radius of the pericentral “shoulder” increased from 0.46 ± 0.17 to 0.55 ± 0.15, by +0.09 (+19.5% of baseline MPOD, P = 0.01) and from 0.46 ± 0.17 to 0.53 ± 0.14, by +0.07 (+15.2% of baseline MPOD, P = 0.03), respectively (Fig. 3c, continuous line; Fig. 4, black continuous line). Mean MPOD at outer and inner radius of the pericentral “shoulder” increased from 60% at baseline to 64.3% and from 60% to 62.7% of the peak MPOD, and the ratio of MPOD at outer and inner radius remained stable (from 1–1.02 at the last follow-up visit). Changes in eccentricities of 1° and more were similar in all three types of MP distribution. None of the control eyes that at baseline had been classified as having no ring distribution developed a bimodal pattern of MP distribution during follow-up (Fig. 3d). The MPOD values at all eccentricities showed an insignificant increase of not more than 0.04 D.U. (at 0.25°, Fig. 4, gray dotted line). In control eyes that at baseline had been classified as having a “ring-like” structure, all MPOD values at all eccentricities (including the eccentricities of maximum and minimum of the ring) decreased slightly (Fig. 3e), and none of these changes was significant (Fig. 4, gray dashed line). In intermediate eyes, the changes during follow-up did not exceed 0.02 D.U., which also was not significant (Fig. 4, gray continuous line). 

Discussion

To our knowledge, this is the first study to investigate the effect of L and Z intake on the profiles of MP, with special focus on the preexistence or new occurrence of ring structures. Our new analyses were performed to provide answers to the question whether a ring-like structure in the MP profile can be generated de novo, or whether a preexisting ring can be amplified by oral L and Z, or rather disappears when the MP rearranges toward a more homogeneous distribution. 

Of note, when the LUNA study was designed, the proof of an inverse relation of ring-like structures in the MP distribution and AMD ²⁶ was outstanding and a potential protective effect of these structures MP distribution against the disease if at all was hypothesis. This is the reason why, when the subjects were enrolled, it was not focused especially on the presence or absence of ring structures and, therefore, the number of volunteers with this pattern is quite low. Likewise and unintentionally, the prevalence of ring structures and intermedia distribution is higher in our control group than in the intervention arm. It will be of interest for future studies to enroll larger numbers of individuals with plateauing or bimodal pattern of the MP density graph and use different supplements, including different concentrations of L and/or Z and/or additionally MZ to find out if and how higher concentrations of the carotenoids or the additional MZ might modify the MP profile. ^29,30 Given the large study group with a long period of supplementation, the long follow-up and our method to measure the MP, we felt encouraged to analyze retrospectively our MP density measurements with special focus on recent epidemiologic findings that the individual relative spatial distribution of MP might be of greater importance for the putative protective effect against AMD than the absolute amount of MP in single locations. ²⁶ The strength of the measurement method that was used in the LUNA study to image the MP is that, once the images are gained, a later analysis regarding new aspects is possible. This is, among the possibilities to image MP distribution, an advantage over multipoint heterochromic flicker photometry. 

Baseline Findings

We presented additional analyses on the profiles of MP distribution of 97 LUNA participants, of whom 11 had shown a ring-like structure at baseline, 8 an “intermediate distribution,” and the remaining 78 a strictly monotonous decline with no ring. The presence of ring-like and intermediate MPOD profiles differs from that reported in other studies, probably due to different study sampling procedures. Berendschot et al. describe ring structures in about half of their 53 healthy probands. ²⁴ Thus, in the larger cohort of the epidemiologic Muenster Aging and Retina (MARS) study, almost one in five participants showed a ring-like structure, and almost 14% displayed an “intermediate distribution.” ²⁶ However, about half of the single study eyes were free of ARM in this cohort, while the LUNA population consisted mainly (92.6%) of subjects who exhibited features of AMD. The rare prevalence of ring structures and intermediate distribution of MP in subjects mainly affected by AMD might suggest a potential protective role of the ring structures against development of AMD, but the rather low number of participants in our study and the fact that LUNA was not designed as an epidemiologic investigation to clarify the prognostic impact of MP profiles prohibits further inferences on this relationship. Our baseline finding of a mean eccentricity of 0.92° for the second peak is consistent with results from the MARS cohort using similar measurement procedures and detecting it at 0.85°. ²⁶ Other studies found the maximum of the ring at about 0.7° from the fovea (Delori FC, et al. IOVS 2004; 45:ARVO E-Abstract 1288). ^24,36 In line with our analyses, the mean eccentricity of the MP minimum was 0.48° in the MARS study ²⁶ and in earlier studies by Delori et al. ²⁵ However, the corresponding mean MPOD values of 0.56 and 0.45 D.U. in the MARS study are much higher than our findings of 0.31 and 0.38 D.U., which are better comparable to those of Delori et al. ²⁵ However, in contrast to Delori et al., who reported on MPOD values at the maximum and minimum of 81% and 75% of the peak MPOD, and was supported by the findings in the MARS cohort, the maximum of our ring eyes was 56% and the minimum 46% of the peak MPOD, respectively. The maximum-to-minimum ratio that Delori et al. reported on was 1.09, which is lower than our maximum-to-minimum ratio of 1.2 and that in the MARS study (1.27 in right eyes and 1.23 in left eyes). 

Again, to interpret these values one must keep in mind the different study populations in the MARS 2 and LUNA studies. The MARS population included, while LUNA strictly excluded, individuals with a history of oral L and Z supplementation. Following the intake of 12 mg L and 1 mg Z, the maximum and minimum of the ring in eyes with ring-like structures increased from 56% and 46% to 63% and 52% of the peak MPOD, respectively. It is conceivable that a longer intake of L and Z would have pronounced the differences between minimum and maximum to a higher extent. The differences in the previous carotenoid intake may contribute to the discrepancy of MPOD values between the studies. ²⁶  

Response to Supplementation with L and Z

Ring-like structures have been detected with different imaging modalities, ²⁴ but their anatomic basis is not known fully. Following histologic investigations on retinal layers of Macaque and Cebus monkeys, Snodderly et al. hypothesized that the variability of the profiles between species might be caused by differences in the shape of the foveal depression. ³⁷ As hypothesized before, ^25,36 the differences in the profiles of MP distribution with occurrence of ring structures seemed to be related to anatomic conditions that cannot be influenced. According to our new analyses, supplementation did not lead to a completely new occurrence of a ring structure when the baseline distribution showed a strictly monotonous decline in the density profile graph. In fact, in eyes without a ring structure the intake of L and Z resulted in a general shift of MP profile of toward higher values. The increment of MPOD following the carotenoid intake was most pronounced in the center of the fovea, while it declined as a function of eccentricity. In eyes with a ring-like or intermediate distribution of MP, however, MPOD in the center of the fovea changed to a lesser extent compared to the central changes observed in no-ring eyes. Interestingly, eyes with ring-like and intermediate distribution showed the most pronounced increments of MPOD at the maximum of the ring in “ring eyes,” and at the outer radius of the “shoulder” in “intermediate eyes.” We could demonstrate that, following a 6-month supplementation with L and Z, the mean MPOD rise was slightly stronger at the maximum compared to that at the minimum of the ring in individuals with ring-like structures, and also at the outer radius compared to that at the inner radius of the pericentral “shoulder” in eyes of individuals with intermediate distribution. Indeed, it appears that focal pericentral augmentation of MP can lead to a slight amplification of preexisting ring structures or even disclosure of ring-like structures when the indeterminate stage of an intermediate distribution is present, but we did not detect any case of a new occurrence of a ring structure when the baseline distribution showed a strictly monotonous decline in the density profile graph. 

In eyes with a ring-like distribution, MP accumulation in the minimum of the MP distribution was similar to or even lower than the changes observed at the maximum; hence, the presence of ring structures in the MP distribution also was maintained after supplementation. A recent study described the disappearance of a central “dip” in the spatial profile of MP (where MPOD at 0.25° was lower than at 0.5° eccentricity) in four subjects who consumed 7.3 mg of MZ, 3.7 mg of L, and 0.8 mg of Z daily over an eight-week period. ²⁸ This phenomenon was described as the appearance of a “more typical” spatial profile with highest MPOD at the center. We cannot confirm this kind of “smoothening” of the pericentral MP density profile following our supplementation program. We rather suggest that MPOD profiles represent individually predetermined distribution patterns that do not disappear as a result of supplement intake. 

Conclusion and Summary

The main findings of our new analyses are that the supplementation of L and Z does not lead to a new occurrence or disappearance of ring structures when an individual pattern of MP distribution is given. Plateauing or bimodal patterns in the MP profile seem to represent individually predetermined distribution patterns that may be amplified, but not essentially changed into a pattern characterized by strictly monotonous decline with eccentricity as a result of supplement intake. Our findings are in line with those of Snodderly et al., who assumed the variability of the profiles between species might be caused by differences in the shape of the foveal depression. ³⁷  

The mechanisms of retinal stabilization of carotenoids seem to differ among these different patterns of MP distribution: When no plateauing or bimodal pattern is present, most of the supplemented carotenoids are captured in the center of the fovea, and the accumulation decays as a function of eccentricity. When a bimodal pattern is present, the concentration of MP in the center nearly remains unaffected by oral L and Z, and the main increase of MP values is detected in the region of the maximum of the ring or outer radius of the pericentral plateau. Our findings suggested that the most effective mechanisms of retinal capture and/or stabilization of L and Z in such individuals, who show either a ring-like or an intermediate profile, are located at the pericentral rather than the central location. The arrangement of the present MP toward an annulus of higher MPOD superimposed on a central exponentially declining MP distribution may have protective effects against AMD. ²⁶ Future epidemiologic studies should clarify whether not only the presence of a ring structure in the MP distribution, but also the height of the ring, characterized for example by the maximum-to-peak ratio (which in our subjects could be increased by oral L and Z) or the difference between maximum and minimum are associated inversely with AMD. If an inverse association could be proven, the investigation of the height of ring structures would be an exciting topic for statistical evaluation of future or actual supplementation studies, like the Age-Related Eye Disease Study 2 (AREDS 2; in the public domain at www.AREDS2.org). The enhancement of a ring via long term oral intake of L and Z might be one of the most preferable outcomes of supplementation. 

Footnotes

 Supported by Bausch and Lomb.

Footnotes

 Disclosure: M. Zeimer, Bausch and Lomb (F); M. Dietzel, None; H.W. Hense, None; B. Heimes, Thea Pharma (R); U. Austermann, None; D. Pauleikhoff, Novartis (F, S), Pfizer (F, S), Bayer (F, S), Allergan (F, S)