Efficacy of Diacetate Esters of Macular Carotenoids: Effect of Supplementation on Macular Pigment

The accumulation of the carotenoids lutein, zeaxanthin, and mesozeaxanthin in the center of the human retina, and known as the macula lutea or macular pigment, is believed to protect the retina from age-related macular degeneration. Since the macular pigment is of dietary origin, supplements containing the relevant carotenoids are readily available. In this study, we compared the changes in macular pigment over a 24-week supplementation period for two groups of 24 subjects each assigned to either of two supplement formulations, 20 mg/day of lutein or 20 mg equivalent free carotenoids of a combination of diacetate esters of the macular carotenoids. The latter group responded with a larger increase (0.0666 ± 0.0481) in macular pigment optical density than the former group (0.0398 ± 0.0430), driven largely by the older subjects. The difference was statistically significant (p=0.0287). There was a general trend towards smaller increases in macular pigment for those subjects whose baseline value was high. However, the trend was only significant (p < 0.05) for subjects in the diacetate group. No differences in response could be attributed to the gender of the subjects. We also observed no indication that the use of statin drugs by a few of the older subjects influenced their responses.


Introduction
e macular carotenoids, or xanthophylls, lutein (L), zeaxanthin (Z), and mesozeaxanthin (MZ) are found throughout the human retina but are particularly concentrated in and around the fovea [1]. As both antioxidants and blue light blockers, they are believed to protect against degenerative retinal diseases such as age-related macular degeneration [2] and also, potentially, diabetic retinopathy [3] and retinitis pigmentosa/Usher syndrome [4]. Because the density of macular pigment in the retina responds positively to dietary supplementation with the macular carotenoids [5], it is appropriate to study di erent formulations with regard to their e cacy of absorption into the retinal tissues. e major source of L for the supplement industry is the marigold bloom (Tagetes erecta) where L is found in esteri ed form (56% lutein dipalmitate, 36% lutein dimyristate, and 8% lutein monomyristate [6]). However, the extraction and puri cation process, which typically involves alkaline saponi cation, results in free L. In order to obtain L esters from T. erecta, a food-grade solvent is employed [7]. Both free and esteri ed L are commercially available in the supplement form. A comparative study found that the bioavailability of the esteri ed form was higher than that for the free form as indicated by the increased uptake into the blood serum of free L [6]. (Following ingestion, the L esters are hydrolyzed prior to reaching the circulation.) e suggested reason for the higher bioavailability was the better dispersion, solubilization, and incorporation of L esters into micelles formed in the digestive process, compared with free L. However, this study contradicts a more recent crossover study by Chung et al. [8] in which no signi cant di erences in serum response were observed when the subjects were taking free L or L ester supplements.
L esters can also be synthesized from free L, and the same is true for the other macular carotenoids, Z and MZ. e purpose of the present study was to determine the e ciency of absorption of a mixture of diacetate esters of L, Z, and MZ (Micro Mic ™ ), relative to free L. Free L appears to be the most common commercially available supplement aimed at improving the health of the eye and was therefore considered an appropriate control for this study. Since the ultimate target tissue for the carotenoids is the retina, the e ciency of absorption was determined by directly measuring retinal levels of macular pigment.

Study Duration.
Blood serum levels of carotenoids generally respond rapidly to carotenoid supplementation, reaching a plateau after about 4 weeks [9]. However, because the macular pigment responds more slowly, we adopted a 24-week supplementation period. Past studies have shown that during such a time period, signi cant changes in macular pigment optical density (MPOD) can generally be anticipated [9,10].

Subject Demographics.
A total of 48 subjects were recruited from the students, faculty, and sta at Florida International University. Subjects signed an IRB-approved informed consent form, and the study complied with IRB regulations as well as the Declaration of Helsinki. In order to compare possible age e ects on the response to supplementation, we recruited subjects in two age ranges, 18 to 30, and over 50 years of age. e subjects were then split into two supplement groups: Group 1, consisting of 24 subjects, received 20 mg per day of Micro Mic in a L : MZ : Z ratio of 10 : 10 : 2; Group 2, also consisting of 24 subjects, received 20 mg per day of L (note that commercially available L contains approximately 5% of Z). With 24 subjects in each group and a desired power of 0.80, a signi cant di erence between groups of 0.03 in the change of MPOD, with σ � 0.035, would be realizable. Subject demographics are summarized in Table 1. e study was a single-blinded study with the subjects not being informed which supplement they were receiving. Subjects were not asked to modify their diets, and an assessment of their normal dietary intake of xanthophylls was not included in this study.

Supplements.
e supplements were provided in the form of identical looking gel caps that subjects were instructed to take with a meal, one per day, throughout the supplementation period. Each gel cap contained 20 mg of carotenoids (Table 1) in vegetable oil. In the case of the diacetate esters, 20 mg refers to the amount of free carotenoids, that is, not including the masses of the diacetate groups. For the lutein gel caps, 20 mg refers to the total amount of carotenoid, ∼95% of which was L and ∼5% was Z. Subjects were also given a 7-day pill organizer to aid in compliance and a schedule for future visits including dates for receiving re lls. Compliance was determined by counting remaining gel caps at the end of the study and using this information to determine the number taken as a percentage of the number that should have been taken.

MPOD Measurements.
MPOD was determined at baseline (week 0), at weeks 6, 12, and 18 and at the conclusion (week 24) of the supplementation period. e instrument employed was the mapcat SF ™ [11], a heterochromatic icker photometer that was used in a customized mode (cHFP). ("Customized" refers to using optimum icker frequencies for the individual subject.) MPOD was obtained in the right eye of each subject except for those whose vision was markedly better in the left eye. e subject viewed a small, circular stimulus, 1.5°in diameter and provided with crosshairs that alternated between blue and green lights provided by LEDs. e blue light is strongly absorbed by the macular pigment while the green light is only weakly absorbed. e result, generally, is a ickering appearance of the stimulus due to mismatched luminances. In order to determine the customized icker frequency, the subject rst viewed the stimulus with the green light switched o . Starting with a high blue light frequency (∼45 Hz), the subject gradually reduced the frequency until icker was just perceived (critical fusion frequency, CFF). For the next stage of the test, the frequency was set to 2/3 of the CFF. With both the blue and green LEDs turned on, the subject altered the intensity of the blue light until, at equiluminance, icker stopped or was minimized. Small adjustments to the frequency were made if the subject reported either a range of no icker (frequency too high) or inability to eliminate icker (frequency to low). e subject's blue light intensity setting at equiluminance re ected the amount of attenuation of the blue light, principally by the macular pigment and also, and increasingly with age, by the lens. e whole procedure was repeated with a larger, 15°stimulus with crosshairs for central xation, and a default frequency set at 5/6 of the CFF. For this phase of the test, the subjects were asked to adjust the blue luminance to eliminate icker around the periphery of the stimulus while ignoring the residual icker at the center. Subjects made ve repeat measurements for each part of the test, and the test was deemed acceptable if the standard error in the mean MPOD was less than 0.015. e MPOD and standard error were calculated automatically by a programmed microprocessor in the mapcat SF. e calculations are described in Bone and Mukherjee [11]. After a brief resting period, the entire test was repeated up to two more times. A weighted mean of the MPODs was calculated together with a standard error according to an algorithm published by Olive et al. [12].

Statistical Analysis.
Results are expressed as means ± SD. Increases in MPOD resulting from supplementation were tested for signi cance using an independent-samples t-test (α � 2). Values of p < 0.05 were considered signi cant. Adjustments for potentially confounding factors such as body mass index were not included in the analyses.

Retention and Compliance.
All but two of the subjects completed the study satisfactorily. Of these, one of the younger subjects from Group 1 had to return to her homeland for personal reasons, and one of the older subjects from Group 2 developed severe di culties with performing the cHFP test. us, reliable data were obtained from 23 subjects in Group 1 (11 young and 12 old), and 23 subjects in Group 2 (12 young and 11 old). e average compliance, based on the percentage of pills taken, and associated standard deviation for each group are shown in Table 2.

E ect of Supplementation on MPOD.
Positive changes in MPOD were obtained for the vast majority of the subjects. Eighteen subjects in Group 1 and fteen subjects in Group 2 were considered as responders based on a change in MPOD greater than twice the standard error in the mean. Negative changes, usually very small, were observed for two subjects in Group 1, and four subjects in Group 2. In all but two of these cases, the presupplementation MPOD was high. e general e ect of a high presupplementation MPOD on the change in MPOD will be discussed later.
A robust response to supplementation is shown for one of the Group 1 subjects in Figure 1. It indicates, via the linear regression line, a remarkably linear relationship that was in fact typical of the majority of subjects. Note that the regression line is weighted using the reciprocal of the variance at each data point as the weighting factor. e average changes in MPOD ± SD from week zero to week 24 are shown for the two groups in Table 3. Also shown are the changes expressed as percentages of the week zero value and the rate of change in MPOD, that is, the slope of the regression line. Using two-tailed Student's t-test, we found a signi cant di erence when comparing the change in MPOD for the two groups (p � 0.0287).

E ect of Presupplementation MPOD on Change in MPOD.
ere was a negative trend between the change in MPOD and the presupplementation MPOD for each group as well as for the combined groups. Figure 2 shows the results for Group 1, the only one exhibiting a signi cant correlation.
e slope of the regression line (−0.154) indicates by extrapolation that, on average, we might expect no change in MPOD with supplementation for a subject whose presupplementation MPOD was ∼0.93. For Group 2, and for Groups 1 and 2 combined, the slope of the regression line, the degree of correlation, and the signi cance were as follows: Group 2 (slope � −0.041, R 2 � 0.033, p � 0.41); Groups 1 and 2 combined (slope � −0.070, R 2 � 0.058, p � 0.106).

E ect of Age on Change in MPOD.
Since the subjects were recruited in either of two age groups (18-30 � "young," >50 � "old"), we were able to determine whether age was a signi cant factor in the MPOD response to supplementation. e MPOD response was quanti ed in three ways as indicated in Table 3: the overall change in MPOD, the percentage change in MPOD, and the rate of change in MPOD. e results ± SD are presented in Table 4 for the combined groups and for each individual group. Also included are the presupplementation MPODs and the p values for two-tailed t-tests to test for signi cant di erences.
Since we found that the change in MPOD was signicantly greater for Group 1 subjects when compared with Group 2 subjects, we made a similar comparison for the two age groups separately. e results are shown in Table 5 and indicate that the change in MPOD was greater for both old and young subjects in Group 1 compared with Group 2, but the di erence in MPOD response was signi cant ( * ) only for the older subjects (p � 0.025).

E ect of Gender on Change in MPOD.
Presupplementation MPOD was lower for the female subjects in the combined groups compared with the males but did not quite reach statistical signi cance. ere were no di erences between males and females for any measure of change in MPOD. e results are summarized in Table 6.

E ect of Statin Use on Change in MPOD.
Statin use was restricted to the >50 year old subjects, six in Group 1 and three in Group 2. e data for the nonusers and users of this drug are summarized in Table 7. Included are the p values for Student's t-test for signi cant di erences.

Discussion
Retention of subjects was high with only 2 subjects out of 48 being unable to complete the study. Likewise, Table 2 shows a very high, and almost identical, compliance for the two groups.
e results for Groups 1 and 2 revealed that the Micro Mic diacetate formulation assigned to Group 1 produced an average increase in MPOD that was 67% higher than that produced by the lutein formulation assigned to Group 2. e di erence was statistically signi cant (p < 0.03). is result is     [6] who reported that levels of L in the serum were higher for subjects consuming the esteri ed form of L compared with those consuming free L. We also found that the change in MPOD bore a negative relationship with the presupplementation MPOD (Figure 2), albeit signi cant only for Group 1. However, the di erence in the MPOD increases for Groups 1 and 2 could not be attributed to this nding since the average presupplementation MPODs were 0.499 and 0.477, respectively. We also considered the possible in uence of statins since use of this drug may reduce L and Z levels in the serum [13] and, by extension, in the retina. However, because there were more statin users in Group 1 than in Group 2, we can rule out statin use as a contributor to the larger MPOD increase for Group 1. ere have been a number of other studies that examined the e ect of lutein supplementation on MPOD. Direct comparisons are often di cult because of di erent doses, di erent supplementation periods, or di erent MPOD measurement parameters. For example, Nolan et al. [14] supplemented their subjects with 20 mg per day of a mixture of unesteri ed L, Z, and MZ (10 : 10 : 2). After 6 months, MPOD measured at 0.23°eccentricity increased by ∼0.06. Although this is very similar to the increase observed in the present study for subjects taking the esteri ed 10 : 10 : 2 mixture, MPOD measured with a 1.5°stimulus will always be signi cantly lower than that measured at 0.23°eccentricity [4]. In a study by Schalch et al. [15], one group of subjects received ∼20 mg/day of a mixture of L (∼10 mg) and Z (∼10 mg) for 6 months. MPOD measured by HFP with a 1°s timulus resulted in an average increase in MPOD of ∼15%. Again, this is similar to our own results which yielded ∼16% and ∼10% increases for Groups 1 and 2, respectively. Smaller increases in normal subjects were reported by Aleman et al. [4], whose subjects were supplemented with 20 mg/day of L for 6 months. e average increase was only 0.01 when measured by HFP with a 1°stimulus. However, the average increase was 0.07 for patients with retinitis pigmentosa or Usher syndrome.
As shown in Table 4, age was not a signi cant factor in determining the change in MPOD resulting from supplementation. is was true for Groups 1 and 2 individually or in combination. We also examined the data to see whether the larger change in MPOD for Group 1 subjects compared with Group 2 subjects was age-dependent. As seen in Table 5, both the young and old subjects in Group 1 had a larger change in MPOD than their counterparts in Group 2. However, the di erence was largely due to the older subjects whose increase in MPOD was 144% higher for Group 1 than Group 2 subjects compared with 36% for the younger subjects. e di erence for the older subjects was statistically signi cant (p < 0.05). Table 4 also contradicts a claim, popular among some in the supplement industry, that MPOD declines with age. Although not statistically signi cant, the average presupplementation MPOD for the combined groups was actually higher for the older subjects.
Prior to supplementation, the male subjects had, on average, a 17% higher MPOD than the female subjects, though the di erence was not signi cant. Similar genderbased di erences have been reported previously [16]. A possible reason could be that the diet of the male subjects resulted, on average, in a higher intake of xanthophylls than for the female subjects, but this was not assessed in this study. However, the increase in MPOD was almost identical for males and females.
With the use of statin drugs being commonplace among older persons, statin use was not included as an exclusion criterion. Nor did the study design include a comparison of MPOD responses between statin users and nonusers. Nevertheless, we noted that prior to supplementation, the statin users had, on average, a 22% higher MPOD than the nonusers. On the other hand, the change in MPOD was 29% lower for the statin users, but neither di erence was statistically signi cant, probably owing to the small number (9) of statin users. Because these ndings appear to contradict each other, we are unable to present evidence to either refute or support the earlier nding that statin use was associated with lower serum levels of L and Z [13]. Our results could also simply be a re ection of Figure 2, that is, subjects whose baseline MPOD is high generally have a smaller increase in MPOD. e limitations of this study include the lack of a placebo group. However, our past experience has led us to conclude that MPOD changes over a 6 month period are insigni cant  when subjects are assigned to a placebo group [17,18]. Also, we did not include an assessment of the subjects' dietary intake of xanthophylls which we would expect to in uence their presupplementation MPOD. On the other hand, the supplements provided an approximately 10-fold increase in xanthophyll intake over the average dietary intake, so the subjects' diets would not be expected to in uence changes in MPOD in any signi cant way. Lack of serum analysis of carotenoids in this report might also be considered a limitation; however, the target tissue for this study was the neural retina where the bene cial e ects of the macular carotenoids are believed to occur. Finally, we did not use a mixture of unesteri ed L, Z, and MZ as our control. erefore, it might be argued that the larger increase in MPOD observed for the diacetate group was a result of the inclusion of the zeaxanthin stereoisomers in the supplement rather than the esteri cation. We would argue against this possibility based on past observations, including our own, that neither Z nor MZ appears to produce as large a change in MPOD as L [15,18,19]. Nonetheless, we acknowledge the report by Loughman et al. who found that supplementation with all three macular carotenoids produced signi cant increases in MPOD whereas supplementation with L did not [20].

Conclusions
In conclusion, we have found that a combination of macular carotenoids in diacetate form was more e ective at raising MPOD than unesteri ed L, especially in older subjects. is suggests that these particular esters may be more readily absorbed, that is, more bioavailable, than their unesteri ed counterparts.

Conflicts of Interest
e authors declare that there are no con icts of interest regarding the publication of this paper.