Nutritional Manipulation of Primate Retinas, I: Effects of Lutein or Zeaxanthin Supplements on Serum and Macular Pigment in Xanthophyll-Free Rhesus Monkeys

Abstract

purpose. The xanthophylls lutein (L) and zeaxanthin (Z) are the primary components of macular pigment (MP) and may protect the macula from age-related degeneration (AMD). In this study, L or Z was fed to rhesus monkeys reared on xanthophyll-free diets to follow the accumulation of serum carotenoids and MP over time.

methods. Eighteen rhesus monkeys were fed xanthophyll-free semipurified diets from birth until 7 to 16 years. The diets of six were then supplemented with pure L and six with pure Z at 3.9 μmol/kg per day (2.2 mg/kg per day) for 24 to 56 weeks. At baseline and 4- to 12-week intervals during supplementation, serum carotenoids were measured by HPLC, and MP density was estimated by two-wavelength reflectometry. Serum carotenoids and MP were also measured in monkeys fed a stock diet.

results. Monkeys fed xanthophyll-free diets had no L or Z in serum and no detectable MP. During supplementation, serum L or Z increased rapidly over the first 4 weeks and from 16 weeks onward maintained similar levels, both several times higher than in stock-diet–fed monkeys. The central peak of MP optical density increased to a relatively steady level by 24 to 32 weeks in both L- and Z-fed groups. Rhesus monkeys fed a stock diet had lower blood concentrations of L than those found in humans and other nonhuman primates.

conclusions. Rhesus monkeys respond to either dietary L or Z supplementation with increases in serum xanthophylls and MP, even after life-long xanthophyll deficiency. These animals provide a potential model to study mechanisms of protection from AMD.

Lutein (L) and zeaxanthin (Z) are dihydroxy carotenoids, also known as xanthophylls, that are selectively concentrated in the macular region of human and nonhuman primate retinas and result in the macula’s characteristic yellow color. L and Z are thought to have a specific role in protecting the macula from light-induced oxidative damage because of their reduction of short-wavelength light exposure and their antioxidant activity. ¹ Therefore, it is plausible that L and/or Z may prevent or slow the progression of age-related macular degeneration (AMD), the leading cause of vision loss in older adults. 

In both humans and macaque monkeys, L and Z have been shown to have distinct distributions within the macula. The concentration of Z peaks more sharply in the fovea, whereas L shows a shallower density profile, so that Z is higher than L in the fovea but lower than L in the periphery. ^{2 3 4} The preferential accumulation of Z in the central fovea occurs even when it is present at lower levels than L in the blood. ^{5 6} Therefore, it would be of interest to determine the relative effectiveness of feeding the two compounds in raising blood levels and increasing macular pigment density. Two previous studies examined the effects of feeding xanthophyll supplements to nonhuman primates. Snodderly et al. ⁷ found that pure Z supplementation of squirrel monkeys fed stock diets produced a rapid increase of Z in plasma but did not affect plasma concentrations of L. Leung et al. ⁸ found that a Z-rich fruit extract, fed daily for 6 weeks, raised Z concentrations in serum, liver, and the macula. 

Population studies have shown moderate or weak positive relationships among dietary L and Z, their concentrations in blood, and macular pigment density. ^{9 10 11} One study that measured all three variables in 280 healthy adults found correlation coefficients of r = 0.20–0.26 between dietary L+Z, serum L or serum Z and macular pigment density measured psychophysically by flicker photometry. ¹¹ Many sources of variability may contribute to the lack of stronger correlations in human populations consuming uncontrolled diets. Sex and body weight or body fat are factors known to affect blood and tissue xanthophyll levels, macular pigment density, and the response to L and Z supplementation. ^{9 12 13} For example, a recent large study found that macular pigment density was significantly associated with serum L, serum Z, and adipose L in men but not in women. ¹⁴ In addition, human studies have shown that supplementing the diet with L and/or Z, or foods rich in L and Z, can raise macular pigment optical density in many subjects, but that both the serum and macular pigment responses are quite variable ^{12 15 16 17 18 19 20} (Garnett KM, et al. IOVS 2002;43:ARVO E-Abstract 2820). Thus, many questions remain regarding the uptake and metabolism of these two compounds, their relative efficacy in raising macular pigment levels, and their role in preventing age-related macular degeneration. 

Some, but not all, epidemiologic studies provide evidence that higher dietary intakes and blood levels of L and/or Z are associated with protection from AMD. ²¹ Most notably, the Eye Disease Case–Control Study reported reduced risk of exudative neovascular AMD in subjects with the highest level of plasma L+Z and in those with higher L+Z intakes. ^{22 23} Combined L and Z intake was the nutritional factor most strongly related to reduced risk, independent of the effects of other carotenoids. Furthermore, spinach and/or collard greens, particularly rich sources of L and Z, were the specific foods associated with the greatest risk reduction. ²³ There also is limited evidence from case–control studies for a more direct association between levels of macular pigment in the retina and the risk of AMD, including lower macular pigment optical density in the unaffected eyes of patients with unilateral AMD and lower L and Z measured biochemically in the peripheral region of donor eyes with AMD. ^{24 25}  

Few studies have evaluated the relative roles of L and Z in AMD risk reduction. One exception is a recent study by Gale et al., ²⁶ who found that the plasma concentration of Z but not L was significantly negatively associated with risk for AMD. In an animal model of light-induced retinal damage, quail consuming supplemental dietary Z had increased retinal Z, which correlated with reduced apoptotic photoreceptor death. ^{27 28} Studies in model lipid membranes have shown that both L and Z prevent oxidative damage, but that Z was more effective during prolonged ultraviolet light exposure. ²⁹  

Because nonhuman primates are the only animals with a macular structure closely resembling that in humans, they are especially valuable for studying factors involved in macular disease. We previously reported that rhesus monkeys reared on xanthophyll-free semipurified diets had no macular pigment and showed an increase in macular transmission defects seen in retinal fluorescein angiograms. ³⁰ Histologic analysis indicated that such transmission defects were due to lipoidal degeneration of retinal pigment epithelial (RPE) cells as well as areas of RPE with reduced melanin and increased lipofuscin content. ³¹ The present studies were made possible by the existence of another, similar group of adult rhesus monkeys. Because these animals had been fed xanthophyll-free semipurified diets from birth, they had no detectable L or Z in serum or adipose tissue and no detectable macular pigment at the beginning of this study. Therefore, the effects of dietary supplementation with individual pure xanthophylls could be followed more readily. Xanthophyll-depleted monkeys provide a particularly useful model for examining metabolism of individual carotenoids and the roles of L and Z in the macula. 

Oxidative damage, in part induced by bright light exposure, is suspected of playing an important role in the etiology of AMD. The retina is especially vulnerable to oxidative damage because of its high oxygen levels and light exposure in combination with high levels of polyunsaturated fatty acids. Photoreceptor outer segment membranes contain the body’s highest levels of docosahexaenoic acid (DHA, 22:6n–3), a long-chain n–3 fatty acid with six double bonds that confer a particularly high susceptibility to lipid peroxidation. This high DHA content appears to optimize photoreceptor function, ³² but it also may increase the vulnerability to light-induced damage. Retinal DHA levels can be reduced by long-term intake of diets low in n–3 fatty acids. Monkeys, rats, and guinea pigs reared on such diets have abnormal retinal function (reviewed in Jeffrey et al. ³² ), but rats with low retinal DHA levels show protection from acute light damage. ^{33 34} In contrast, human studies suggest a protective effect of dietary n–3 fatty acids. Five recent observational studies found significantly lower risk of development or progression of advanced AMD ^{35 36 37} (SanGiovanni JP, et al. IOVS 2003;44:ARVO E-Abstract 2112) or all forms of AMD ³⁸ in those with higher intakes of long-chain n–3 fatty acids, DHA and/or fish (a rich source of DHA), particularly when combined with low intakes of linoleic acid, the principal n–6 fatty acid. ^{36 37}  

In this study, we were able to examine the possible interaction of retinal DHA with macular pigment, because some of the animals were fed diets low in n–3 fatty acids from birth, whereas others received adequate levels of n–3 fatty acids. Our previous studies showed that monkeys raised on the same low n–3 fatty acid diet as in the present study had 80% less retinal DHA than an adequate n–3 fatty acid group or monkeys fed a standard stock diet. ^{39 40} Animals in the low n–3 fatty acid group had slower visual acuity development ⁴¹ and abnormalities in the electroretinogram, including delayed implicit times and prolonged rod recovery functions. ^{39 42}  

This study was designed to examine the effects of dietary L or Z supplementation in this unique group of xanthophyll-depleted animals, using purified preparations of the individual xanthophylls, in particular pure L made available specifically for this project. Its goal was to determine the time course of the increase of L and Z and their metabolites in the serum and to track increases in macular pigment density. The use of xanthophyll-free animals allowed us to test for differences between the effectiveness of the two different pure xanthophylls, in combination with high or low retinal DHA levels. Other papers in this series will describe related outcomes in the same animals, including quantitative morphology of the retina and RPE, ⁴³ retinal and adipose tissue carotenoid concentrations, and the effects of acute short-wavelength light exposure. 

Methods

Animals and Diets

All procedures were approved by the Institutional Animal Care and Use Committee of the Oregon National Primate Research Center (ONPRC), Oregon Health and Science University, and conformed to NIH guidelines and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

The experimental subjects were 18 rhesus monkeys (Macaca mulatta) reared on one of two semipurified diets, both of which provided adequate levels of all known nutrients including vitamin A (as vitamin A acetate) and α-tocopherol, but contained no detectable carotenoids. The two diets differed only in their fat sources and therefore in the content of n–3 and n–6 fatty acids. In particular, one had very low levels (0.3% of total fatty acids), and the other had adequate levels (8% of total fatty acids) of α-linolenic acid (18:3n–3), an n–3 fatty acid and precursor of DHA. As described in detail previously, ^{39 42} these diets were fed to the mothers of the subjects throughout pregnancy and to the subjects from the day of birth and continuing throughout the study. All animals also received limited amounts of low-carotenoid or carotenoid-free foods such as cereals, fruits (e.g., pineapple and banana), and sweetened gelatin. The subjects were fed three times per day and had fresh water available at all times. They were maintained on a 12:12-hour light–dark cycle with an ambient light level of 50 to 90 lux produced by full-spectrum fluorescent lamps (F32-T8-TL850; Philips, Eindhoven, The Netherlands). 

For xanthophyll supplementation, L was purified and Z was synthesized by DSM Nutritional Products Ltd. (Basel, Switzerland; formerly Roche Vitamins Ltd.) and formulated into gelatin beadlets. Depending on the xanthophyll and batch, the beadlets contained 4% to 9% of the purified xanthophylls, as confirmed by reversed-phase HPLC analysis in our laboratory (described later). This analysis also confirmed that the L, purified in a noncommercial process specifically for this study, contained only all-trans-L and no detectable Z. In the Z beadlets, approximately 90% was in the all-trans form and 10% was present as cis-Z, with no detectable L. The cis-Z isomer was tentatively identified as 13-cis-Z by comparing absorption spectra and HPLC retention time with a known standard. The noncarotenoid portion of the beadlets was identical for the L and Z supplements. 

Beginning at 8 to 16 years of age, while all of the monkeys continued on their semipurified diets, the diets of six were supplemented with pure L and six with pure Z. The remaining six animals received no xanthophyll supplements. The L- and Z-fed groups were balanced to the extent possible based on sex, age, n–3 fatty acid diet group, and body weight. They began supplementation in three cohorts, at different times of the year, with each cohort including two L-fed and two Z-fed animals. Cohort 1 consisted of four females, 7 to 9 years of age, all in the low n–3 fatty acid group. Cohort 2 included three females and one male, 13 to 16 years of age, with one adequate n–3 and one low n–3 animal in each of the two supplement subgroups. Cohort 3 included four males, 9 to 14 years of age, again with one adequate n–3 and one low n–3 animal assigned to each of the xanthophyll supplements. The six animals with unsupplemented diets included four females and two males, 7 to 17 years of age. The characteristics of each subject are listed in Table 1 . 

L and Z supplements were both given at 3.9 μmol/kg per day (2.2 mg/kg per day). In both squirrel monkeys and rhesus monkeys this dose of Z was shown to approximately double plasma Z levels relative to monkeys fed a stock laboratory diet, ^{7 8} and it represents 7.7 times the estimated average daily intake of rhesus monkeys fed a stock diet (described later). The daily doses of beadlets were measured for each animal based on body weight, which was measured at least monthly. The beadlets were inserted into food treats, such as marshmallows or small pieces of fruit, and fed just before the animals’ midday meal of semipurified diet. Beadlets and individual supplement doses were stored in the dark at 4°C. Supplements were initially given 7 d/wk; the frequency was reduced to 4 d/wk after week 52 for cohort 1, week 44 for cohort 2, and week 15 for cohort 3. Cohorts 1 and 2 continued to receive the supplements 4 d/wk for a total duration of at least 56 weeks, and cohort 3 for 24 to 34 weeks. The present paper reports longitudinal in vivo measurements through 56 weeks of supplementation. Therefore, Table 1 indicates for each subject the total duration of supplementation through 56 weeks, as well as the durations of the component periods when supplements were provided 7 and 4 d/wk. For consistency and comparison to the other papers in this series, the table also includes the same information for the length of the entire study, up to the time the animals were killed for the morphologic and biochemical studies reported in the companion papers. 

Eight of the supplemented monkeys (four L-fed and four Z-fed, cohorts 2 and 3) received acute small-spot (150 μm), coherent short-wavelength light exposures within the fovea and in the parafovea to determine the threshold for photochemical light damage. These studies were performed in one hemiretina of the right eye before the beginning of supplementation and in a second hemiretina after 22 to 28 weeks of supplementation. The results of these studies will be described in a separate report. 

The animals fed semipurified diets were compared with age-matched rhesus monkeys from the ONPRC colony that were fed a standard stock laboratory diet (5047 High Protein Monkey Chow; Ralston Purina, Richmond, IN) plus fruit and vegetables (primarily apples and carrots). The animals were housed under the same general conditions as the experimental diet groups. The carotenoid content of the stock diet included 7 to 10 nmol/g (4–6 μg/g) of L, 7 to 9 nmol/g (4–5 μg/g) of all-trans-Z, no detectable cis-Z, and 0.7 to 1.4 nmol/g (0.4–0.8 μg/g) of β-carotene, primarily from corn and alfalfa. For an estimated stock diet intake of 30 g/kg per day, the intakes per kilogram per day of these carotenoids therefore averaged approximately 0.26 μmol of L (150 μg), 0.24 μmol of Z (135 μg), and 0.035 μmol of β-carotene (19 μg). The supplemental fruits and vegetables (typically one half apple or one-half carrot, approximately three times per week) added an estimated maximum average daily intake of ∼3 nmol/kg of L plus Z, or less than 1% of the intake from the stock diet. Seven of these animals provided photographs for measurement of macular pigment density (females, ages 10–13, body weights 4.9–7.4 kg), and 17 had blood samples drawn for serum carotenoid analyses (15 females and 2 males, ages 6–16 years, body weight 3.8–13.7 kg, including 4 of those with macular pigment measurements). 

Carotenoid Analysis of Diets and Supplements

The carotenoid concentrations of the stock diet and semipurified diets were determined using the official method of analysis of the Association of Official Analytical Chemists ⁴⁴ and were analyzed with the same reversed-phase HPLC system used for serum and tissue analysis. ¹² For the L and Z supplements, an exact amount of the beadlets (∼0.5 mg) was dissolved in 1.0 mL of distilled water. This solution was extracted with 2 mL chloroform-methanol (2:1) three times. The chloroform extract was evaporated to dryness under nitrogen. The residue was redissolved in 1 mL of ethanol, vortexed, and sonicated for 30 seconds and then taken up to 100 mL of ethanol. A 50-μL aliquot was used for HPLC analysis. The limit of detection for both xanthophylls was 0.2 pmol. For each batch of xanthophylls the analysis was performed in triplicate and completed before the beadlets were fed. 

Serum Carotenoid Analyses

For all animals in the experimental diet groups, blood samples were obtained before carotenoid supplementation, 2 and 4 weeks after the beginning of supplementation, and every 4 weeks thereafter. For stock diet-fed control subjects, a single sample was obtained. Fasting blood samples (5–10 mL) were drawn from the saphenous or femoral vein into foil-wrapped tubes under dim light and centrifuged at 800g for 15 minutes to obtain serum, which was stored in the dark at −80°C until analysis. 

Serum was prepared for extraction using 150 μL of sample and 0.5 mL 0.9% saline. Echinenone, in ethanol, was added as an internal standard (Hoffman-La Roche, Inc., Nutley, NJ). The mixture was extracted by using 2 mL CHCl₃:CH₃OH (2:1, vol/vol). The mixture was vortexed and then centrifuged at 800g for 15 minutes at 4°C. The CHCl₃ layer was removed and evaporated to dryness under nitrogen. A second extraction was performed on the mixture using 3 mL hexane. The mixture was vortexed and centrifuged as above. The hexane layer was combined with the first extraction and evaporated to dryness under nitrogen. The residue from serum was redissolved in 150 μL of ethanol, vortexed, and sonicated for 30 seconds. A 50-μL aliquot was used for HPLC analysis. ⁴⁵  

Carotenoids were measured using a reversed-phase, gradient HPLC system, as previously described. ¹² The method yields adequate separation of L, all-trans-Z, cis-Z, cryptoxanthin, α-carotene, 13-cis-β-carotene, all-trans-β-carotene, and 9-cis-β-carotene, as well as four geometrical isomers of lycopene (15-cis, 13-cis, 9-cis, and all-trans lycopenes). ¹² Carotenoids were quantified at 455 nm by determining peak areas in the HPLC chromatograms calibrated against known amounts of standards. All-trans-L, all-trans-Z and 3′didehydrolutein (DDL) standards were a gift of DSM Nutritional Products Ltd. (Basel, Switzerland). Concentrations were corrected for extraction and handling losses by monitoring the recovery of the echinenone internal standard. The lower limit of detection with this method is 0.2 pmol for each carotenoid. ¹²  

Fundus Photography

Macular pigment optical density of the supplemented animals was measured by two-wavelength monochromatic fundus reflectometry at baseline and every 4 weeks after the beginning of supplementation until 56 weeks. For unsupplemented animals and those fed stock diets, one or two measurements were made. Standard color photographs of the retinal fundus were obtained at the same time. For retinal photography, monkeys were initially sedated with ketamine (∼10 mg/kg intramuscularly [IM]), and the pupils were dilated with 1% tropicamide plus 2.5% phenylephrine (2 drops of each per eye). The animals were then anesthetized with propofol (∼10 mg/kg intravenously [IV]). The lids were held open by a speculum, and a gas-permeable hard contact lens was placed on the cornea to prevent drying and optimize image quality. 

Calibrated monochromatic photographs were taken with a fundus camera (Carl Zeiss Meditech, Jena, Germany) outfitted with custom interference filters (Omega Optical, Burlington, VT). An index of macular pigment density was obtained from the monochromatic photographs by two-wavelength reflectometry. The physical principles underlying this method have been described by Delori et al. ⁴⁶ One wavelength band (peak 486 nm, full width at half maximum [FWHM] 25 nm) is strongly absorbed by the macular pigment, and the comparison wavelength band (peak 538 nm, FWHM 29 nm) is only weakly absorbed by the pigment. For brevity, we refer to these as “blue” and “green,” respectively. The blue-wavelength band was not at the peak of the macular pigment absorbance (approximately 460 nm) but was slightly shifted toward longer wavelengths to avoid interference from other yellow pigments ^{4 47} and to minimize scattering in the ocular media. The difference between the amounts of light reflected in these two wavelength bands provides a measure of pigment density when referenced to the known spectrum of the macular pigment ⁴⁷ (tabulated in Delori et al. ⁴⁶ ). Delori et al. ⁴⁶ found that human macular pigment density determined with a similar reflectometric method correlated highly with psychophysical measurements by heterochromatic flicker photometry, as well as with the results of a new autofluorescence spectrometry method, but that estimates from the reflectometric method were consistently lower by 40% to 50% than those from the other two methods. 

This method assumes that light reflected from the eye has passed through the retina twice, once on the way in and once after being reflected behind the retina. A potential artifact that must be minimized is the effect of specular reflections from the vitreous surface of the retina. These reflections are particularly obvious in the foveal region, where the curved retinal surface creates image distortion and bright highlights. ⁴⁸ To avoid these reflections, we modified the fundus camera by placing apertures in the pupil plane of the illuminating beam so that light entered the eye only through either the nasal or the temporal half of the pupil. With the illumination restricted to these sectors, the camera could be positioned so that the reflections were moved to the side of the fovea, leaving the vertical meridian relatively reflection-free (Fig. 1) . 

In each photographic session, both retinas were photographed with the blue and green filters, first with nasal illumination, then with temporal illumination. This alternation was repeated to produce a set of four blue-green image pairs (two nasal and two temporal) for each eye. The images were recorded on 35-mm black-and-white film (TMAX 400; Eastman Kodak, Rochester, NY). 

A calibration strip of film from the same film lot was separately exposed by contact with a calibrated neutral density wedge (30 steps; A-64013; Dupont, Wilmington, DE) that was uniformly illuminated by light from a flash gun mounted in an integrating sphere. This calibration film strip also had an unexposed region to represent the fog level of the film. Each film roll with retinal images was loaded into a developing tank along with a calibration strip so that the calibration strip and the film roll were developed under identical conditions. Films were developed (T-Max developer; Eastman Kodak) at 20°C for 10 minutes. 

Image Analysis

Each data point represents results obtained from a single blue-green pair of photographs from one eye. The best pair of negatives was selected on the basis of the following criteria: (1) no specular reflections present in the foveal region; (2) fovea centered in the frame without image distortion related to viewing angle; (3) film density in the linear input–output range of the film; and (4) image in good focus. To avoid bias in selecting film frames for analysis, we examined each frame according to a predetermined order. First, the eight frames from the right eye were examined. If a satisfactory pair was found, no further films were examined. If no right eye pairs were adequate, frames from the left eye were examined. In 27% of the sessions, right eye images were unsatisfactory, and left eye images were used. The right eyes of eight of the supplement-fed animals (four L-fed and four Z-fed, groups 2 and 3) had been exposed to small spot (150 μm) short-wavelength light at 22 to 28 weeks of supplementation, but these were not near the vertical meridian where all the macular pigment measurements were made (discussed later) and did not interfere with the optical density measurements. 

Each 35-mm negative of the selected blue-green pair was individually mounted and placed on a uniformly illuminated mechanical stage. The negatives were imaged by a video camera (C2400; Hamamatsu, Hamamatsu City, Japan) containing a chalnicon tube with a wide dynamic range. The video camera was connected to an image-capture board (model IVP-SA; MuTech, Billerica, MA, no longer available) in a computer. In preliminary experiments, we had determined the magnification of the photographs by measuring the distances between retinal vascular landmarks that could be visualized in fundus photographs and in wholemounts of the same retina after the animal was killed. With this information, the system was set up to capture images with a known scaling of 20 μm/pixel at the retina. 

For each blue-green pair of negatives, the image of the blue negative was averaged for 17 frames to minimize the contributions of electrical noise and then stored on disc with an image-analysis program (Image Pro v1.2; Media Cybernetics, Silver Spring, MD). Without moving anything, the same image was captured using custom software that stored the image in one color plane of the image-capture board. The green negative was then placed on the mechanical stage. Its image was displayed as a live overlay on the monitor in another color and mechanically translated and rotated to bring it into register with the blue image. This registration procedure compensated for the inevitable small motions that occurred between taking the blue and the green photographs. Finally, the image analysis program averaged the green image for 17 frames and stored it on disc. 

As a separate step, the density of each roll of film as a function of relative exposure was calibrated by reading the digitized output of the camera for seven steps of the calibration strip that was developed with that roll. These values were hand-entered into a lookup table in the image-analysis program that converted the blue and the green image files into maps of the relative optical density of the retina in the two wavelength bands. 

Once the blue and green image files were converted into density maps, they were subtracted to form a difference image. The density difference profile along a vertical band passing near the center of the fovea was displayed, and the horizontal location of the vertical band was varied to locate the peak difference. At the location of the peak, the average density of a 100-μm-wide row of pixels was calculated at each vertical location and stored as a spreadsheet file. This constituted our best estimate of the density difference along the vertical meridian (Fig. 1) . 

To derive an estimate of the amount of macular pigment, the area under the density difference profile was calculated. The usual form of the profile was a single peak near the center of the fovea that declined to a fairly constant baseline at approximately 0.5 mm eccentricity. However, in some images, a plateau beyond 0.5 mm was not obvious because of noisy profiles or valleys intruding into the plateau region that were probably caused by unwanted reflections from the foveal crest. In these cases, the baseline was set to the density value at 0.5 mm eccentricity. The difference between the density profile and the baseline was multiplied by a constant calculated from the filter transmission curve to estimate the optical density of macular pigment at 460 nm for a single transit of light through the retina. The area between the density profile and the baseline integrated from −0.5 to +0.5 mm vertical (the integrated optical density) was taken as our index of the amount of macular pigment for each retina. This index has the units of optical density at 460 nm × mm. 

Note that if macular pigment has a measurable optical density at eccentricities beyond 0.5 mm, these more peripheral regions will not contribute to the index, but will reduce the density difference and therefore the macular pigment estimate. This fundamental limitation of failing to account for contributions outside the main density peak of the pigment is common to most in vivo methods for measuring macular pigment. ^{46 49}  

Statistical Analyses

The effects of supplement type (L vs. Z) and supplement duration on serum total xanthophylls and on macular pigment optical density were tested with two-way repeated measures analyses of variance (ANOVAs) including the time points with data for all 12 animals (baseline and 4 through 24 weeks) and all time points with data for 8 animals (baseline and 4 through 56 weeks). In addition, a series of analyses was conducted to examine effects of possible confounding or modifying variables. Influences or interactions with discrete variables (gender, n–3 fatty acid status, body weight category and the duration of 7d/wk supplement administration) were explored with 2-way repeated-measures ANOVAs for each of these variables × supplement duration, and by 3-way repeated-measures ANOVAs for each variable in interaction with supplement type (e.g., gender × supplement type × duration). Effects of continuous variables (age and body weight) were explored both by linear regression and by analyses of covariance (ANCOVAs), in which the effect of supplement type was tested while controlling for effects of each of the other variables. The latter analyses used values for single time points (24 or 56 weeks) or average values for each subject for 4 to 24 weeks (n = 12), 4 to 56 weeks (n = 8), or other time periods as appropriate. Analyses for each of the subject variables were performed for both serum xanthophylls and for macular pigment density, but details of the results are reported only when significant. Differences with P < 0.05 were accepted as significant for overall ANOVAs or two-way comparisons. For multiple comparisons, Bonferroni-Dunn post hoc tests were used, in which the combined P-value of all comparisons was controlled at P < 0.05. Post hoc comparisons of analyses involving three groups (L-fed, Z-fed, and control groups) involve three pair-wise comparisons, and therefore in these cases a criterion probability of 0.05/3 = 0.017 was used. 

Results

Serum Carotenoids

The dominant serum carotenoids of normal rhesus monkeys fed the stock diet in this colony were the macular xanthophylls L and Z. The serum L concentration in monkeys fed the stock diet (n = 17) was 0.074 ± 0.009 μmol/L (mean ± SEM), all in the trans form. Their serum Z concentration was 0.081 ± 0.007 μmol/L for Z, of which 0.058 ± 0.009 μmol/L (72%) was in the all-trans form and 0.023 ± 0.004 μmol/L (28%) was cis Z. The serum of the stock diet animals also contained low levels of β-carotene (0.011 ± 0.004 μmol/L) and lycopene (0.034 ± 0.006 μmol/L). 

In contrast, monkeys fed the semipurified diets had no measurable serum L or Z before supplementation. The only carotenoid detected in the serum of the xanthophyll-free animals was lycopene (<0.070 μmol/L). When these animals began receiving L or Z supplements, their serum xanthophyll concentrations rose rapidly in the first 4 weeks, with L reaching a mean of 1.14 μmol/L (range, 0.53–1.85) in the L-fed group and total Z reaching 0.65 μmol/L (range, 0.19–1.43) in the Z-fed group by 4 weeks (Fig. 2) . Serum xanthophyll levels in the supplemented animals exceeded the levels in monkeys fed stock diets by 2 weeks of supplementation and thereafter were approximately 10 times as high for L and 10 to 20 times as high for Z. Although mean levels in the L group were almost twice as high as in the Z group at 4 to 12 weeks, the difference was not statistically significant because of high interindividual variability (repeated measures ANOVA, P = 0.12 for overall effect of diet and P > 0.10 at each time point). In particular, higher means for L-fed animals during this period were due to very high levels in one or two animals per time point. Both the variability and the mean serum L concentrations in the L-fed group decreased after 12 weeks, and, from 16 weeks onward, total xanthophyll concentrations were similar in the two supplement groups (Fig. 2) . Thus, repeated measures ANOVAs for serum total xanthophylls at all time points with data for all 12 animals (baseline and 4 through 24 weeks) and at all time points with data for 8 animals (baseline and 4 through 56 weeks) showed no significant effects of supplement type. There were also no significant effects of supplement duration, except that serum levels at all points from 4 weeks onward were greater than the presupplementation baseline (P < 0.001), and there were no significant interactions between supplement type and duration (P > 0.10). 

ANOVAs, as described in the Methods section, demonstrated no effects on serum total xanthophylls of gender, n–3 fatty acid status, or the duration of 7 d/wk supplements (all P > 0.4) and no interactions of these variables with supplement type. There were also no significant correlations between age or body weight and serum xanthophyll levels (all P > 0.4, r² < 0.1). It should be noted, however, that with these small samples sizes the power to detect such effects was limited. The only effect that approached significance was an effect of body weight on serum xanthophylls early in supplementation. When animals were divided into two weight categories, high and low, relative to mean normative values for each sex, those with higher body weights (four females, four males) had higher serum carotenoid levels during the first 12 weeks than those with lower weights (three females and one male). A three-way repeated-measures ANOVA (supplement × body weight category × time for 4, 8, and 12 weeks) showed trends toward differences for both the type of supplement (L greater than Z, P = 0.058) and for weight category (high greater than low, P = 0.065) but no interaction (P = 0.51). However, an ANCOVA, used to test the difference between the L- and Z-fed groups while controlling for effects of body weight, showed essentially identical results to those in a simple comparison between the two supplement groups (P = 0.124 and 0.120, respectively). 

In the L-fed group, all the serum L was in the all-trans form, and no Z was detected. In the Z group, approximately two thirds of serum Z was all-trans and approximately one third was in the cis form (Fig. 3) , a ratio similar to that found in animals fed the stock diet. Thus, the proportion of the cis form in the serum was considerably higher than the 10% present in the diet, as also found previously in squirrel and cynomolgus monkeys. ^{6 7}  

In addition to the dietary xanthophylls, small amounts of 3′didehydrolutein (DDL) were measured in the serum of some members of the Z-fed group. DDL is a variant of L in which the 3′ hydroxyl has been replaced by a keto group, and its occurrence in the blood of animals fed only Z was unexpected. However, the identification of DDL was confirmed by coelution with a known standard, comparison of absorption spectra with a known standard, and liquid chromatography-mass spectrometry (LC/MS; data not shown). The mean concentration of DDL was greatest at 2 weeks (28 ± 16 nmol/L) and averaged 16 nmol/L over the course of the study. The contribution of DDL to the total serum carotenoid content of these animals was small (<2%) except at 2 weeks (9%) when the levels of serum Z were the lowest. 

Macular Pigment Optical Density

In monkeys fed the unsupplemented carotenoid-free diets, integrated macular pigment optical densities over the central 1 mm were zero or very low, within the noise range of the measurement technique. During L or Z supplementation, integrated macular pigment optical densities rose over the first 24 to 32 weeks but showed no additional consistent increase from 32 to 56 weeks (Fig. 4) . Note that the integrated optical densities during the later phases of supplementation (approximately 2 units of optical density at 460 nm × mm) correspond to a peak difference between the foveal center and the baseline of approximately 0.1 to 0.2 optical density units at 460 nm (see Fig. 1 ). 

A repeated-measures ANOVA for the time points with data for all 12 animals (0, 4, 8, 12, and 24 weeks) found no overall difference between the L- and Z-fed groups (P = 0.29). There was a significant interaction between supplement group and time (P = 0.049), but no significant difference at any individual time point (P > 0.05). The effect of supplement duration was significant overall (P < 0.001), with a significant post hoc difference between the 4- and 24-week time points (P = 0.009); all time points, including 4 weeks, were different from the presupplementation baseline (P < 0.001). A repeated-measures ANOVA for all time points with data for eight animals (every 4 weeks from 0 to 56 weeks except 32 and 40 weeks, when data were missing for one animal per group) also showed no difference between the supplement groups (P = 0.12) but an effect of supplement duration (P < 0.001). Differences were significant between baseline and all other time points (P < 0.001) and between 4 weeks and 28, 44, and 56 weeks (combined P < 0.05). 

ANOVAs showed no effects of sex or n–3 fatty acid group on macular pigment optical density at 4–24 weeks (n = 12) or 4–56 weeks (n = 8) (all P > 0.15). There was also no effect of the duration of 7 d/wk supplementation (15 vs. 24 week duration for the 24-week values, or 44 vs. 52 week duration for the 48- to 56-week values, both P > 0.5). Finally, no correlations were found between either age or body weight and macular pigment density (all P > 0.3, r² < 0.2). However, as noted earlier, the power to detect such effects was limited by the small sample size. 

Macular pigment density was not significantly related to serum total xanthophyll levels. Specifically, levels of macular pigment at 12 or at 24 weeks, or values for each animal averaged over 12–24 weeks, showed no correlation with total serum xanthophyll levels at 4–12 or 12–24 weeks, and macular pigment density at 56 or 48–56 weeks was not related to serum xanthophylls at these same times (all P > 0.2). However, when only trans-xanthophyll serum levels were used (excluding cis-Z in the Z-fed group), the correlation of 24-week serum values with 24-week macular pigment density approached significance (P = 0.081, r ² = 0.20) for the L- and Z-fed groups combined, and was significant for the Z-fed group alone (P = 0.038, r ² = 0.63). 

There also was no significant correlation between the cumulative absolute intake of xanthophylls and macular pigment densities for 12 weeks or 24 weeks (n = 12) or for 56 weeks (n = 8), whether including total xanthophylls (all P > 0.2, r ² < 0.2) or only trans-xanthophylls (all P > 0.4, r² < 0.1). Regression analyses showed that cumulative absolute intakes of total or trans-xanthophylls were not better predictors of macular pigment density than intake per kilogram body weight or the duration of supplementation. 

In addition to the data shown in Figure 4 , a final set of photographs was obtained before the animals were killed at 24 to 103 weeks (6–24 months) for the biochemical and morphometric studies described in the other papers in this series. At that time the macular pigment density (mean ± SEM) for the L-fed group was 2.97 ± 0.39, for the Z-fed group was 2.38 ± 0.27 (not different, P = 0.24), and for the unsupplemented group was 0.18 ± 0.12 (within the noise range of the measurement). 

The mean integrated macular pigment density of the stock diet animals was 4.87 ± 0.51 (mean ± SEM, n = 7). Values for the experimental monkeys after 24 to 56 weeks of supplementation remained low in comparison, despite the fact that serum xanthophyll levels were several times higher in the supplemented groups. After 24 weeks of supplementation of all 12 animals, the mean integrated macular pigment density for the L-fed group was 50% and that for the Z-fed group was just 30% of the mean of animals fed the stock diet. At 56 weeks, the corresponding percentages in the remaining eight animals were 48% and 43% in the L and Z groups, respectively. 

Color fundus photographs did not reveal formation of crystals within the retina at any time during supplementation. 

Discussion

Serum Xanthophylls and Effects of Supplementation

Serum xanthophylls have been measured in many species of primates, including humans, but in most cases Z, L, and their geometrical isomers have not been separated. Despite this limitation, the data in previous studies indicate that rhesus monkeys on normal stock diets have relatively low serum xanthophyll concentrations compared with those in humans, great apes, and most other captive monkey species that have been studied. ^{50 51} Indeed, recent measurements for rhesus monkeys from China show the lowest values for serum L+Z of any Old World primates yet examined. ⁵⁰ In previous studies performed by one of the authors with other collaborators, Z and L, as well as some of their isomers, have been separated in analyses of serum from humans and two species of monkeys. ⁶ Cynomolgus macaques (Macaca fascicularis) and squirrel monkeys (Saimiri sciureus) fed the same stock diet as the rhesus monkeys in the ONPRC colony had serum L concentrations more than twice those of rhesus monkeys, whereas serum Z concentrations were about the same. This finding suggests that rhesus monkeys either absorb L more poorly from normal diets or metabolize it more rapidly than two other common laboratory primates. Even so, the equal, high levels of L or Z supplements fed in the present study resulted in similarly high serum levels of L and Z in the experimental monkeys. 

An even greater contrast occurs when rhesus monkeys are compared with humans, although dietary differences could be a major factor in this comparison. In studies where serum L and Z have been separated, mean human L concentrations ranged from 176 to 328 nmol/L and mean all-trans-Z concentrations from 79 to 88 nM ^{5 16 17} (Garnett KM, et al. IOVS 2002;43:ARVO E-Abstract 2820). These concentrations can be compared to the mean serum L of 54 to 74 nmol/L and mean serum all-trans-Z of 5 to 58 nmol/L in rhesus monkeys found by Leung et al. ⁸ and the present study. Such low serum values, compared with humans and with the other nonhuman primate species, presumably contribute to the relatively low macular pigment densities of rhesus monkeys indicated by our fundus reflectometry, as well as low values found by microdensitometry (Snodderly DM, Sandstrom MM, manuscript in preparation). 

In Z-fed animals, cis-Z preferentially accumulated in serum, as it accounted for about one third of the serum concentration of total Z, compared with 10% in the Z supplements. The proportion of cis-Z was nearly as high in the stock diet group that received no detectable dietary cis-Z, suggesting in vivo isomerization of trans to cis. The absorption of cis isomers varies among different carotenoids. For example, increased intestinal absorption of cis versus trans isomers and in vivo isomerization is also thought to occur for lycopene, ⁵² whereas trans-β-carotene is better absorbed, as reflected in higher serum levels, than the cis isomer. ⁵³ Our results confirm observations in two other species of monkeys that cis-Z is preferentially accumulated in serum. ⁷ This preferential accumulation could be due to increased intestinal absorption of cis-Z, in vivo isomerization, and/or greater tissue uptake of trans-Z versus cis-Z. 

In addition to the dietary xanthophylls, 3′didehydrolutein (DDL) was found in the serum of Z-fed animals. This finding is interesting in the light of a proposal that DDL might be an intermediate in the retinal transformation of dietary L to 3R,3′S-zeaxanthin (RSZ or meso-zeaxanthin), a major component of macular pigment. ⁵⁴ The retinal biochemistry results from our project establish that RSZ is derived from dietary L (Johnson EJ, unpublished data, 2002), whereas the serum data suggest that the DDL in the retina could originate from Z, and therefore that DDL is unlikely to be an intermediate in the generation of RSZ. 

Although the number of rhesus monkeys that were supplemented is small, the data suggest that L accumulates more rapidly in the serum of some individuals, but then becomes comparable to concentrations of Z as supplementation continues. This pattern suggests the need to strengthen the fragmentary data from supplementation of humans with pure L and pure Z. When the same amounts of L or Z have been ingested by humans for various times, serum Z concentrations have been proportionally higher the longer the duration of supplementation (Z-to-L ratio at 18 to 21 days, 0.09 ¹⁷ ; at 20 weeks, 0.28 ¹⁶ ; and at 6 months, 0.76 (Garnett KM, et al. IOVS 2002;43:ARVO E-Abstract 2820). Given the small number of subjects in these studies (n = 2–8), the results must be considered preliminary, but they suggest that more extensive experimentation may confirm different kinetics for the accumulation of L or Z in the blood of humans as well as monkeys. 

It may be of interest that the variability in serum xanthophyll concentrations was very high in the first 12 weeks but then decreased. This finding may be relevant to the large variation in the response to supplementation in human studies, most of which are of relatively short duration (12–15 weeks); responses to xanthophyll supplementation may become more consistent over a longer period. 

Accumulation of Macular Pigment

In rhesus monkeys with lifelong intake of xanthophyll-free diets and therefore no macular pigment, L or Z supplementation led to high levels of L or Z in the blood and to the accumulation of macular pigment. Measurements of integrated macular pigment density by reflectometry showed similar accumulations after feeding L or Z. These results demonstrate that primate retinas that developed to full maturity in the absence of xanthophylls nevertheless retained the mechanisms to accumulate macular pigment from either L or Z. They therefore offer hope that intervention to increase macular pigment in elderly humans at risk for macular disease could be successful if it is deemed desirable. Furthermore, intervention is likely to be feasible even for people with poor diets who have had low macular pigment densities throughout life. 

The general lack of correlation between serum xanthophylls and macular pigment optical density may be due to a saturation effect caused by the high serum levels of L or Z, so that uptake of xanthophylls into the retina, rather than the circulating blood level, was the limiting factor in the accumulation of macular pigment. The only significant correlation was between the serum concentration of all-trans-Z and macular pigment density within the Z-fed group, which is consistent with the possibility that all-trans-Z is a more effective source for macular pigment than cis-Z. This result is also consistent with our unpublished data indicating that the retina almost exclusively accumulates trans-Z in preference to cis-Z. 

High levels of L or Z supplementation did not lead to deposition of crystals within the retina. This suggests that retinas that have never been exposed to the macular xanthophylls can still incorporate them successfully when exposed to large doses. In contrast, large doses of carotenoids that are not normally present in the retina, such as canthaxanthin, may not be handled as well and can result in formation of crystals. ⁵⁵  

Despite the high serum levels of L or Z in the supplemented animals, the reflectometric estimates of their macular pigment density, even after 1 year of supplementation, were only approximately one half the mean value in control animals fed stock diets. However, the reflectometric method measured macular pigment optical density only in the central 1 mm, and therefore would not detect increased density outside this central density peak, or increases in xanthophyll concentrations occurring throughout a larger retinal area such as that measured biochemically. In fact, our biochemical results indicate that retinal xanthophyll content of the supplemented animals was as high or higher than in control retinas in a 4-mm diameter disc centered on the fovea, and was substantially higher in the periphery (Johnson EJ, unpublished data, 2002). Given these considerations, the reflectometry data may be most useful in demonstrating relative changes in the early uptake of the xanthophylls into the retina but may not provide a measure of total xanthophyll concentrations, especially those outside the central peak of macular pigment density. This limitation may also apply to the use of reflectometry or psychophysical measures of macular pigment optical density in human studies of xanthophyll supplementation, since these in vivo measurements of macular pigment density share the limitation that they are limited to the foveal density peak. 

Another possible contributor to lower macular pigment density in the experimental animals, as measured by reflectometry, is a difference in the molecular orientation of xanthophylls, with less organized deposition in the animals reared on xanthophyll-free diets. Xanthophyll molecules exhibit dichroism, so that their precise orientation in the retinal tissue determines their optical density. ⁵⁶ It is possible that the absence of xanthophylls during development and through most of adult life may have impaired the processes responsible for this ordered orientation, and thereby resulted in lower optical density despite similar xanthophyll concentrations. 

No differences were found in the uptake of xanthophylls into serum or in macular pigment density between animals fed high or low levels of n–3 fatty acids. In comparison to the adequate n–3 fatty acid group, the monkeys in the low n–3 fatty acid group were previously shown to have slower visual acuity development ⁴¹ and electroretinogram abnormalities including delayed recovery of the rod photoresponse. ^{39 42} However, the low n–3 and adequate n–3 groups did not differ in their degree of macular transmission defects assessed by fluorescein angiography (Neuringer, et al. IOVS 1999;40:ARVO Abstract 882). This outcome suggests that the transmission defects, which are related to lipoidal degeneration and loss of melanin in the RPE cells, ³¹ are related to xanthophyll deficiency rather than fatty acid status. Our related study on quantitative morphology of the RPE ⁴³ presents strong evidence that the RPE is sensitive to the absence of macular pigment and responds to xanthophyll accumulation in retinas that previously had none. It also demonstrates differences in the RPE response to xanthophyll supplementation between the low and adequate n–3 fatty acid groups. 

Summary

Rhesus monkeys with no lifetime intake of xanthophylls retained the ability to absorb and metabolize L or Z supplements, to increase xanthophyll levels rapidly in serum, and to increase macular pigment optical density. Macular pigment densities reached only approximately one half the densities in normal animals, although biochemical analyses of the same animals showed that xanthophyll content of the macular area was as high or higher than in the control subjects. This discrepancy may be accounted for by abnormally high xanthophyll levels outside the central fovea of the supplemented animals. 

⁷  Present affiliation: Department of Ophthalmology, Medical College of Georgia, Augusta, Georgia.

 Supported by DSM Nutritional Products Ltd., Basel, Switzerland (formerly Roche Vitamins Ltd.); Grant DK29930 from the Institute of Diabetes, Digestive and Kidney Diseases; Grant RR00163 from the National Center for Research Resources; Grant 581950-9-001 from the U.S. Department of Agriculture; Grant P30 EY03790 from the National Eye Institute; and a grant from The Foundation Fighting Blindness.

 Submitted for publication December 4, 2002; revised July 22, 2003, and January 27, 2004; accepted April 5, 2004.

 Disclosure: M. Neuringer, DSM Nutritional Products Ltd. (F); M.M. Sandstrom, DSM Nutritional Products Ltd. (F); E.J. Johnson, DSM Nutritional Products Ltd. (F); D.M. Snodderly, DSM Nutritional Products Ltd. (F)

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Composition of Experimental Groups and Subject Characteristics

Animal ID	Sex	Age at Start (y)	Body Weight (kg)*	Supplement Duration to 56 Weeks, †	Supplement Duration to End of Study, †	n–3 Fatty Acid Status
Unsupplemented
776	M	7.0	11.5	0	0	Low
566	M	9.7	7.1	0	0	Low
481	F	11.5	6.5	0	0	Adequate
411	F	12.7	6.1	0	0	Low
300	F	15.0	9.6	0	0	Low
275	F	15.3	7.9	0	0	Adequate
Mean		11.9	8.1
SEM		1.3	0.8
Zeaxanthin-supplemented
642	F	7.9	9.7	56 (52, 4)	62 (52, 10)	Low
567	F	9.7	9.6	56 (52, 4)	62 (52, 10)	Low
224	F	16.4	9.1	56 (44, 12)	103 (44, 59)	Low
217	F	16.5	6.0	56 (44, 12)	103 (44, 59)	Adequate
586	M	11.0	12.1	33 (15, 18)	33 (15, 18)	Low
398	M	14.8	11.3	34 (15, 19)	34 (15, 19)	Adequate
Mean		12.5	9.6
SEM		1.6	0.9
Lutein-supplemented
602	F	8.9	8.5	56 (52, 4)	58 (52, 6)	Low
585	F	9.4	6.1	56 (52, 4)	58 (52, 6)	Low
362	F	13.8	6.1	56 (44, 12)	64 (44, 20)	Low
397	M	13.4	8.3	56 (44, 12)	64 (44, 20)	Adequate
636	M	9.7	10.9	24 (15, 9)	24 (15, 9)	Low
463	M	13.3	11.7	24 (15, 9)	24 (15, 9)	Adequate
Mean		11.4	8.6
SEM		0.9	0.9

*  Body weight at start of study.

†  Values in parentheses indicate number of weeks of daily xanthophyll supplementation, followed by number of weeks of supplementation 4 days/week, which continued until the death of the animals.

Figure 1.

 

Examples of monochromatic photographs (negatives) used to generate macular pigment optical density (MPOD) estimates. White vertical lines: the central 3 mm over which the optical density trace was taken. The reported measurements were derived by integrating under the peak in the central 1 mm of the trace, after establishing a baseline at ±0.5 mm eccentricity, as shown. Dashed lines: the alignment of the white line with the optical density curves in the right column. Top: Normal animal fed a stock diet, with nasal illumination of the right eye; a round camera artifact is visible to the left of the fovea and an arc formed by the reflection from the foveal crest is prominent as well. Middle: xanthophyll-free monkey before supplementation, with nasal illumination of left eye; reflection from foveal crest is to the right of the fovea. Bottom: Same monkey after 52 weeks of supplementation with L, with nasal illumination of right eye; reflection on the left. Left column: photographs with blue light (peak wavelength 486 nm, FWHM 25 nm); the macular pigment appears light, because the blue light is absorbed and less is reflected back to the camera to expose the film. Middle column: Photographs with green light (peak wavelength 538 nm, FWHM 29 nm), which is weakly absorbed by the macular pigment. Right column: macular pigment optical density profiles along the vertical meridian, derived from difference images as described in the Methods section. In the curve on the lower right, there is a small dip at the center that results from a discretely focused reflection at the center of the fovea.

Figure 2.

 

Serum L levels of L-fed monkeys and serum total Z levels of Z-fed monkeys (mean ± SEM) at baseline and at 4 to 56 weeks of xanthophyll supplementation. Differences between the supplement groups were not significant overall or at any time point. n = 6 per supplement group at baseline and at 4 through 24 weeks; n = 4 per group at 2 weeks and at 28 through 56 weeks. See Table 1 for detailed characteristics of monkeys in each group and for each period.

Figure 3.

 

Z isomers in the serum of Z-fed monkeys (mean ± SEM) during the period of supplementation. n = 6 per group at 4 through 24 weeks, and n = 4 per group at 28 through 56 weeks.

Figure 4.

 

Macular pigment integrated optical density in L- and Z-fed monkeys (mean ± SEM) at baseline and at each time point measured during supplementation. n = 6 per group at 0, 4, 8, 12, and 24 weeks, and n = 4 per group at all other time points, except that data were available for only three per group at 32 and 40 weeks. See Table 1 for detailed characteristics of monkeys in each group and for each period.

The authors thank Josephine Gold, Noelle Landauer, Dana Myers, Neal Young, and Audrey Trupp for care of the experimental animals and the preparation and feeding of supplements; Noelle Landauer for assistance with data analysis; John Fanton for veterinary services; Francois Delori and Richard I. Land for assistance in setting up the photographic analyses; and at DSM Nutritional Products Ltd. (formerly Roche Vitamins Ltd.), Wolfgang Schalch and Regina Goralczyk for logistic assistance, Alfred Giger for providing purified Z-free L that would not otherwise have been available, and Oliver Froescheis for technical advice and assistance with the measurement of 3′didehydrolutein.