To examine the effects of lutein and zeaxanthin supplementation on retinal function using multifocal electroretinograms (mfERG) in patients with early age-related macular degeneration (AMD).
Randomized, double-masked, placebo-controlled trial.
One hundred eight subjects with early AMD were randomly assigned to receive 10 mg/d lutein (n = 27), 20 mg/d lutein (n = 27), 10 mg/d lutein plus 10 mg/d zeaxanthin (n = 27), or placebo (n = 27) for 48 weeks. Thirty-six age-matched controls without AMD were also enrolled to compare baseline data with early AMD patients. MfERG responses and macular pigment optical densities (MPODs) were recorded and analyzed at baseline and at 24 and 48 weeks.
There were significant reductions in N1P1 response densities in ring 1 to ring 3 in early AMD patients compared with the controls ( P < .05), whereas neither N1P1 response densities in ring 4 to ring 6 nor P1 peak latencies significantly changed. After 48-week supplementation, the N1P1 response densities showed significant increases in ring 1 for the 20 mg lutein group and for the lutein and zeaxanthin group, and in ring 2 for the 20 mg lutein group. The increases in MPOD related positively to the increases in N1P1 response density in ring 1 and ring 2 for nearly all active treatment groups. N1P1 response densities in ring 3 to ring 6 or P1 peak latencies in all rings did not change significantly in any group.
Early functional abnormalities of the central retina in the early AMD patients could be improved by lutein and zeaxanthin supplementation. These improvements may be potentially attributed to the elevations in MPOD.
Age-related macular degeneration (AMD) is a progressive degenerative eye disease specifically targeting the macula, the cone-rich region of the retina responsible for highest visual acuity. In developed countries it is the leading cause of irreversible blindness among people over the age of 50. The number of persons with AMD is expected to increase by as much as 50%, from 1.75 million to 2.95 million, between 2000 and 2020 in the United States. With increasing longevity, this disease will place enormous social and economic burden on healthcare resources. Currently, treatment strategies for certain types of exudative AMD have emerged; however, the majority of these patients will still progress to legal blindness, and there is no proven treatment available for most affected persons with early AMD. Therefore, it may be more preferable and of critical importance to intervene in this disease at an earlier stage so as to slow its progression at a time before substantial visual impairment has occurred.
The xanthophyll carotenoids, lutein and zeaxanthin, have specific distribution patterns in human tissue and are concentrated in the macula. The presence of these xanthophylls is thought to provide a unique function in this vital ocular tissue. Several epidemiologic studies, but not all, have indicated that higher levels of lutein and zeaxanthin in diet are associated with a lower risk of AMD. Our previous meta-analysis of the potential protective effects of lutein against AMD reported reductions in risk of early AMD by 4% and late AMD by 26%, suggesting that lutein might help to delay and prevent the progression from early- to late-stage AMD. Clinical studies of lutein in patients with AMD showed a benefit on visual performance; however, little is known about the effects of lutein on the maintenance of function and structural integrity of the macula. The multifocal electroretinogram (mfERG) is a noninvasive technique that allows simultaneous recording of focal electroretinographic responses from multiple retinal locations. It has the advantage of providing an objective assessment of retinal function and can detect and monitor functional changes of macula and the progression of macular disorders, even in the absence of visual loss. Furthermore, although the concentrations of lutein and zeaxanthin peak in the central fovea, zeaxanthin is the dominant carotenoid at this location. This specific distribution of the xanthophyll carotenoids suggests that zeaxanthin may play an essential role in the center of the retina; but until recently, the research specifically concerning the efficacy of zeaxanthin is still limited.
Therefore, we conducted a randomized, double-masked, placebo-controlled trial to investigate the effects of 48-week supplementation with lutein and zeaxanthin on retinal function by mfERG in a group of community-dwelling patients with early AMD.
Recruitment was directed to subjects with probable AMD, aged 50 to 79 years, from the local communities and from congregate living sites in Beijing. All study candidates underwent standard general and ophthalmic examination to screen for study eligibility. Clinical diagnosis of early AMD was established by slit-lamp examination and ophthalmoscopy using a Goldmann noncontact lens, as well as color fundus photograph (Exwave HAD 3CCD; Sony Electronics Inc, Park RIdge, New Jersey, USA) after pupillary dilation using 0.5% tropicamide and 0.5% phenylephrine (Santen Pharmaceutical Co. Ltd, Osaka, Japan), when either of the following lesions in the macular area of at least 1 eye was identified: soft distinct or indistinct drusen; or areas of retinal pigmentary abnormalities, without the presence of signs of late AMD. The AMD was independently classified by 2 masked ophthalmologists in accordance with the Age-Related Eye Disease Study (AREDS) classification and grading system.
Subjects were excluded if they had late AMD (choroidal neovascularization or geographic atrophy), unstable chronic illness, or eye disorders other than macular degeneration, including macular edema, macular holes, central serous chorioretinopathy, or macular epiretinal membrane. We did not enroll subjects who had taken drugs known to affect visual function within 1 month prior to enrollment (eg, chloroquine or oxazepam). Subjects who were vegetarian or had a history of retina-vitreous surgery or photodynamic therapy were also ineligible.
One hundred eight participants with early AMD (mean age, 69.1 ± 7.4 years) met the study criteria and were included in the study; 107 completed 48 weeks of treatment. Thirty-six age-matched controls without AMD (defined as absence of soft drusen or pigmentary abnormalities; mean age, 68.0 ± 7.9 years) were also enrolled according to the same exclusion criteria used for early AMD patients.
After enrollment, eligible participants with early AMD were randomly allocated to 1 of the 4 groups in a 1:1:1:1 ratio according to a list of computer-generated random numbers in sex-stratified blocks of 8. Participants, study and clinical center personnel, and the data analysts were unaware of the treatment allocation through study completion. Only the pharmacist technician, who was not involved in the recruitment or assessment of participants, had access to the randomization list. The different capsules were identical in size, weight, and color.
Participants were randomly assigned to receive placebo (n = 27), 10 mg of lutein (n = 27), 20 mg of lutein (n = 27), or 10 mg lutein plus 10 mg zeaxanthin (n = 27) once a day for 48 weeks. Adherence to treatment was evaluated by monthly interviewing of the patient and by capsule counts. All participants were requested to maintain their usual diet and to abstain from taking supplements containing carotenoids. Dietary intake was assessed at baseline using a validated 120-item food frequency questionnaire.
MfERGs and macular pigment optical density assessments were performed at baseline (week 0) and at 24 and 48 weeks after the initiation of treatment. The controls without AMD were given only a baseline examination.
MfERGs were recorded using a RETIscan system (Version 3.21, Roland Consult, Inc, Brandenburg, Germany), according to the recommended guidelines of the International Society for Clinical Electrophysiology of Vision (ISCEV) for basic mfERG. Pupils of the patients were maximally dilated to a diameter more than 7 mm with 0.5% tropicamide and 0.5% phenylephrine (Santen Pharmaceutical Co, Ltd, Osaka, Japan) and the cornea was anesthetized with 0.4% oxybuprocaine hydrochloride (Santen Pharmaceutical Co, Ltd). Retinal signals were acquired with a bipolar contact lens electrode (Hansen Ophthalmic, Coralville, Iowa, USA) filled with carboxymethylcellulose sodium (Refresh Celluvisc; Allergan Inc, Irvine, California, USA).
The stimulus consisted of 103 hexagonal elements presented on the CRT display with a frame rate of 75 Hz. Each hexagon was alternated pseudorandomly between white (200 cd/m 2 ) and black (less than 1.0 cd/m 2 ) according to a standard m-sequence. The recordings were performed under room light conditions, and a red central cross was presented for fixation. In each video frame, each stimulus element had an equal probability of being white or black, maintaining the overall mean luminance of the stimulus display at a fairly constant value. Retinal signals were band-pass filtered from 10 to 300 Hz, amplified 100 000 times, and sampled at 1200 Hz. Each recording consisted of 16 segments and was approximately 8 minutes long. During the test, the patient fixated on a small circle in the central hexagon. Recording quality was monitored by observation of the real-time signal voltage. Recording segments contaminated by either electrical artifacts or loss of fixation were rejected and repeated.
The mfERG responses for the hexagons across the retina were separated into 6 concentric rings (rings 1 to 6) for data analysis. The response amplitudes in each ring were measured between the first negative trough (N1) and the first positive peak (P1), yielding the N1P1 response densities (amplitudes per unit retinal area in nV/deg 2 ). The P1 peak latencies (ms) of the positive waveform were also measured.
Macular Pigment Optical Density
The measurement of macular pigment optical density was performed with a confocal scanning laser ophthalmoscope (Heidelberg Retina Angiograph; Heidelberg Engineering, Heidelberg, Germany) using autofluorescence images centered on the fovea. Subjects were positioned in front of the tabletop and asked to look straight ahead and to remain steady. After using infrared (IR) light, the 488-nm wavelength of light is used to excite autofluorescence. A series of autofluorescence images were obtained for the excitation wavelengths quickly before recovery. An average image was generated from these images using the system software. The center of the fovea was defined as the center of a Gaussian distribution. We quantified macular pigment optical density by calculation of the average image and comparison of foveal and parafoveal autofluorescence. The reproducibility and variability of this measurement technique had been previously reported. The coefficients of variation were less than 5%.
Sample Size and Statistical Analyses
It was estimated that a total sample size of 108 participants provides 90% power to detect a 30% to 40% increase in N1P1 response densities between groups, with a significance level of .05 and a dropout rate of 10%.
The data were analyzed according to the intention-to-treat principle. The Kolmogorov-Smirnov test was used to determine if continuous variables were normally distributed. The significance of differences between the early AMD cases and controls was determined by the independent t test, or Mann-Whitney rank sum test where appropriate, for continuous data and the χ 2 test for categorical data. Correlations between macular pigment optical density and the mfERG values for the AMD cases at baseline were analyzed using the Pearson test. Baseline characteristics among 4 treatment groups were compared using analysis of variance (ANOVA) for continuous variables or the χ 2 test for categorical variables. Absolute changes for each outcome variable at 24 and 48 weeks were normally distributed for macular pigment optical density and the mfERG values. Within-group changes from baseline were assessed using the paired t test, and between-group differences in terms of change from baseline at each time point were tested by analysis of covariance (ANCOVA), with age, sex, smoking, and baseline values included as covariates. The least significant difference procedure was used for multiple pairwise comparisons. Furthermore, overall significance of differences in changes over time was evaluated by repeated-measures ANOVA with time and treatment effects and their interactions, with age, sex, smoking, and baseline values included as covariates. The relationships between change in macular pigment optical density and change in mfERG responses were assessed using Pearson correlation analysis because these 2 variables follow a bivariate normal distribution. All analyses were performed using SPSS statistical software version 11.0 (SPSS Inc, Chicago, Illinois, USA).
The demographic and clinical characteristics of patients with the early AMD cases and of controls are shown in Table 1 . The mean age of the participants was 68.8 years (SD 7.5 years; range 49–79 years). Of the 144 total participants, 84 (58.3%) were female. There was no significant difference in demographic characteristics between early AMD patients and controls. The patients with early AMD had lower N1P1 response densities in ring 1 than controls without AMD ( P < .01). Significant differences in N1P1 response densities in ring 2 and ring 3 between groups were still present, but the magnitude of differences was diminished. The difference in N1P1 response densities in ring 4 to ring 6 was no longer significant between case and control groups. No statistically significant differences were found between these 2 groups in macular pigment optical density or P1 peak latencies. N1P1 response densities in ring 1 were highly associated with macular pigment optical density for early AMD at baseline (Pearson r = 0.31; P < .001), whereas N1P1 response densities in ring 2 to ring 6 were not related to macular pigment optical density ( Figure 1 ). In addition, we found no significant relationships between the P1 peak latencies and macular pigment optical density for early AMD.
|Variable||Study Group||P Value|
|Controls (n = 36)||Early AMD Patients (n = 108)|
|Mean age (SD), year||68.0 (7.9)||69.0 (7.4)||.46 a|
|Sex, n (% female)||21 (58.3)||63 (58.3)||>.99 b|
|Race (% Han people)||36 (100.0)||108 (100.0)||>.99 b|
|Current or past smoker, n (%)||3 (8.3)||13 (12.0)||.76 b|
|Dietary nutrient intake|
|Vitamin A (SD), RE||0.75 (0.49)||0.72 (0.45)||.84 a|
|Vitamin C (SD), mg||81.9 (47.2)||83.0 (44.3)||.90 a|
|Vitamin E (SD), mg||7.4 (2.3)||7.5 (2.4)||.78 a|
|Zinc (SD), mg||7.6 (2.9)||8.0 (3.8)||.48 a|
|β-carotene (SD), mg||3.5 (2.4)||3.5 (2.1)||.85 a|
|Lutein and zeaxanthin (SD), mg||2.6 (1.4)||2.6 (1.5)||.89 a|
|Lycopene (SD), mg||0.86 (0.70)||0.73 (0.66)||.15 a|
|N1P1 response densities, nV/deg 2|
|Ring 1 (SD)||85.7 (36.7)||67.4 (28.5)||.002 a|
|Ring 2 (SD)||48.7 (17.5)||40.5 (14.0)||.005 a|
|Ring 3 (SD)||33.3 (9.2)||28.7 (10.1)||.02 a|
|Ring 4 (SD)||22.7 (6.1)||20.9 (6.9)||.16 a|
|Ring 5 (SD)||16.4 (4.9)||15.0 (5.0)||.15 a|
|Ring 6 (SD)||12.6 (4.1)||11.4 (4.0)||.15 a|
|P1 peak latencies, ms|
|Ring 1 (SD)||38.9 (6.4)||40.3 (5.8)||.26 a|
|Ring 2 (SD)||37.7 (3.0)||37.7 (3.4)||.97 a|
|Ring 3 (SD)||36.0 (3.2)||36.2 (2.7)||.57 a|
|Ring 4 (SD)||35.2 (1.9)||35.8 (3.0)||.23 a|
|Ring 5 (SD)||35.5 (2.1)||35.9 (3.3)||.49 a|
|Ring 6 (SD)||35.6 (1.7)||36.5 (3.4)||.12 a|
|MPOD (SD), DU||0.33 (0.16)||0.31 (0.13)||.53 a|
a Based on independent t test.
The baseline demographic characteristics of patients in the 4 treatment groups are summarized in Table 2 . There were no significant between-group differences in any baseline demographic or clinical variables. The changes from baseline in N1P1 response densities over time are shown in Figure 2 . During the first 24 weeks, all active treatment groups increased N1P1 response densities in ring 1, although this did not reach statistical significance in the 10 mg lutein group. An ANCOVA analysis showed participants assigned to the 20 mg lutein group had a much greater increase compared with the placebo group (0.3 vs 20.3; between-group difference, 19.9; 95% confidence interval [CI], 0.03–39.6; P < .05). After 48 weeks, N1P1 response densities in ring 1 showed a mean (standard error [SE]) increase of 18.0 (7.5) in the 10 mg lutein group, an increase of 22.4 (7.4) in the 20 mg lutein group, and an increase of 23.5 (6.9) in the lutein and zeaxanthin group (all P < .05), with maintenance in the placebo group. The changes in N1P1 response densities in ring 1 for the 20 mg lutein group (22.4 vs −0.3; between-group difference, 22.7; 95% CI, 3.1–42.3; P = .02) and the lutein and zeaxanthin group (23.5 vs −0.3; between-group difference, 23.8; 95% CI, 4.2–43.4; P = .02) were significantly greater than those for the placebo group. Results of repeated-measures ANOVA showed that lutein and zeaxanthin supplementation had both the significant treatment effect ( P = .02) and the significant time effect ( P < .001) on improving N1P1 response densities in ring 1. N1P1 response densities in ring 2 also increased progressively over time in all active treatment groups; and significant within-group differences from baseline were only detected in the 20 mg lutein group (10.6; 95% CI, 1.7–19.5; P < .05) at week 48. The changes in N1P1 response densities in ring 2 for the 20 mg lutein group (10.6 vs −0.3; between-group difference, 10.9; 95% CI, 0.2–21.7; P < .05) had a significantly greater increase than those for the placebo group. A significant time effect ( P = .03) with only a tendency for a treatment effect ( P = .19) was observed for N1P1 response densities in ring 2. Significant changes no longer existed in N1P1 response densities in ring 3 to ring 6 in the 4 groups. No significant differences were found between the groups with respect to changes in these N1P1 response densities.
|Variable||Placebo (n = 27)||10 mg Lutein (n = 26)||20 mg Lutein (n = 27)||Lutein and Zeaxanthin (n = 27)||P Value|
|Mean age (SD), y||68.9 (7.6)||69.9 (8.4)||69.0 (6.8)||68.6 (7.0)||.94 a|
|Sex, n (% female)||16 (59.3)||16 (61.5)||15 (55.6)||15 (55.6)||.96 b|
|Race (% Han people)||27 (100.0)||26 (100.0)||27 (100.0)||27 (100.0)||>.99 b|
|Current or past smoker, n (%)||3 (11.1)||3 (11.5)||3 (11.1)||4 (16.7)||.97 b|