Age-related macular degeneration (AMD) remains a leading cause of modifiable vision loss in older adults. Chronic oxidative injury and compromised antioxidant defenses represent essential drivers in the development of retinal neurodegeneration. Overwhelming free radical species formation results in mitochondrial dysfunction, as well as cellular and metabolic imbalance, which becomes exacerbated with increasing age. Thus, the depletion of systemic antioxidant capacity further proliferates oxidative stress in AMD-affected eyes, resulting in loss of photoreceptors, neuroinflammation, and ultimately atrophy within the retinal tissue.
1. Introduction
Age-related macular degeneration (AMD) is the leading cause of irreversible blindness among older adults in developed countries, affecting roughly one in eight individuals aged 60 years or greater
[1,2,3,4][1][2][3][4]. Comprehensive reports estimate that 200 million people currently live with AMD globally, with this aggregate expected to increase further and approach 300 million by 2040
[3]. Similar rising trends in the U.S. are expected to reach 5.4 million people by the year 2050
[1,4,5][1][4][5]. Although the majority of these are of Caucasian descent, the disease is not limited to individuals of Caucasian origin; Asian, Hispanic, and African-American populations tend to develop intermediate dry AMD and wet polypoidal choroidal vasculopathy with greater incidence
[5,6,7][5][6][7]. The projected growth in prevalence among adults developing noncommunicable eye diseases, such as AMD, can be attributed to the demographic transition consistent with an aging global population
[1,3,8,9,10][1][3][8][9][10]. Due to its chronic nature, wherein this incurable disease requires steady long-term management, AMD has become, and will remain, a public health challenge for both high- and low-income countries, with considerable socio-economic implications and rises in healthcare expenditures
[1,2,3,8,9,10,11,12,13,14][1][2][3][8][9][10][11][12][13][14].
Aging remains one of the primary risk factors in AMD
[1,4,9,15][1][4][9][15]. While cellular senescence is inherent to biological aging, these perturbations in photoreceptor cells and retinal pigment epithelium (RPE) are thought to bring about the neurodegenerative onset characteristic of age-onset maculopathy. Additional non-modifiable risk factors include sex, confirmed family history of AMD, and strong genetic factors that may further predispose individuals to this condition
[1,5,12,16,17,18,19,20,21][1][5][12][16][17][18][19][20][21]. Genetic variants associated with complement factor H (CFH) and AMD susceptibility gene 2 (ARMS2) are well-established risk factors for the development and progression of AMD
[16,17,18,19,20,21,22,23,24,25][16][17][18][19][20][21][22][23][24][25] (single nucleotide polymorphisms that are found on chromosomes 1q31 and 10q26, respectively). Conversely, modifiable risk factors for AMD include dietary behaviors, smoking status, cardiovascular disease, as well as metabolic comorbidities
[5,26,27,28,29,30][5][26][27][28][29][30]. Among these, individuals who currently smoke (and past-smokers) carry significantly greater risk of incident AMD
[26,27,28,31][26][27][28][31].
The etiopathogenesis of AMD is complex and multifactorial. It is postulated that early and intermediate stages of maculopathy are predominated by oxidative stress and low-grade inflammatory activation in aging retinae
[32,33,34,35,36,37,38,39,40,41][32][33][34][35][36][37][38][39][40][41].
Figure 1 provides a summary of the neuroprotective mechanisms provided by macular carotenoids. A comprehensive review of the precise molecular processes, by which carotenoids offer protection against photo-oxidative damage, has been discussed in detail elsewhere
[42]. As a consequence of its extremely high metabolic activity and constant exposure to light, the outer retina is known to be particularly vulnerable to photo-oxidative injury and mitochondrial dysfunction, prompting the overproduction of free radical species
[42,43,44,45,46][42][43][44][45][46]. A growing body of evidence implicates that compromised antioxidant capacity may also serve a crucial role in AMD pathology, a sequela, which occurs predominantly in response to the chronic cycles of sustained oxidative stress, paired with the concomitant depletion of endogenous antioxidants
[42,43,46,47,48][42][43][46][47][48]. Local inhibition of these antioxidant defense mechanisms (to counteract the accumulation of toxic byproducts and cellular debris) plays a significant role in perpetuating subsequent neurodegenerative damage onto the surrounding tissues through immunostimulatory activity
[35,40,41,43,46][35][40][41][43][46]. In fact, outer retinal lesions originating from oxidative insult have been shown to mediate a para-inflammatory state or an adaptive immune response to dysregulated complement activation
[32,33,34,35,36,37,38,40,41,45,49][32][33][34][35][36][37][38][40][41][45][49]. This interdependence between cellular senescence and redox imbalance likely represents essential facets contributing to neurodegenerative onset and disease progression in AMD.
Figure 1. Overview of the neuroprotective mechanisms of xanthophylls lutein, zeaxanthin, and meso-zeaxanthin in the central retina.
The body’s intrinsic homeostatic mechanisms for maintaining redox control are comprised of both exogenous and endogenous antioxidant activity for neutralizing free radical species
[42,43,46,47,48][42][43][46][47][48]. The mainstay of preventative nutritional therapies is aimed at the augmentation of exogenous antioxidant defenses through oral supplementation containing nutraceuticals and micronutrients. In particular, xanthophyll carotenoids possess unique properties, serving as potent antioxidants and anti-inflammatory mediators in the retina, and have been demonstrated to benefit the prevention of neurodegenerative retinopathies, such as AMD and diabetic eye disease
[42,43,50,51,52,53][42][43][50][51][52][53]. Greater dietary intake of xanthophylls, via carotenoid supplementation, has been well-documented to offer clinically meaningful benefits in visual performance, in both healthy and diseased states
[54,55,56,57,58][54][55][56][57][58]. However, the comprehensive neuroprotective capacity, afforded by macular carotenoids in clinical management of AMD, has not been thoroughly discussed. Hence, the purpose of this systematic review concentrates on summarizing the available evidence from observational studies and randomized controlled trials, reporting on carotenoid lutein,
meso-zeaxanthin, and zeaxanthin (only in patients with AMD).
1.1. Age-Related Macular Degeneration
Traditionally, color fundus photography and slit lamp biomicroscopy have been the mainstay for the ophthalmic examination of fundus lesions associated with AMD
[15,59,60][15][59][60]. Several disease classification systems have been developed over the years, from population studies
[61,62,63][61][62][63] and clinical-based trials, of which the Age-Related Eye Disease Study (AREDS) clinical severity scale (
Table 1) and its simplified severity scale are most notable
[64,65,66,67][64][65][66][67]. However, inconsistency in disease terminology and ambiguous definitions, relating to the severity of maculopathy, highlight an overwhelming unmet clinical need. Thus, it is highly recommended for all those working in this field to adopt the single diagnosis of AMD proposed by the Beckman classification (
Table 2)
[68], which is then further classified according to disease severity, in agreement with AREDS
[64,65,66,67][64][65][66][67].
Table 1. AREDS clinical severity scale of age-related macular degeneration (AMD)
[65].
Category |
Age-Related Eye Disease Study (AREDS) Classification |
Table 2. Beckman Classification System of AMD.
Beckman Clinical Classification [68] |
AMD Classification 1 |
Drusen |
Pigmentary Abnormalities 2 |
Additional Features |
1 |
No drusen, or non-extensive small drusen only in both eyes |
2 |
Extensive small drusen, non-extensive intermediate drusen, or the presence of pigment abnormalities in at least one eye |
3 |
No apparent aging |
None |
Extensive intermediate drusen, large drusen, or non-central geographic atrophy in at least one eye |
4 |
Advanced AMD as defined by at least one of the following: geographic atrophy, retinal pigment epithelial detachment in one eye, choroidal neovascularization, or scars of confluent photocoagulation; or visual acuity less than 20/32 associated with lesions from non-advanced age-related macular degeneration, including large drusen in the fovea, in only one eye |
None |
n/a |
Normal aging changes |
Small (≤63 μm) |
None |
n/a |
Early AMD |
Medium (>63 μm and ≤125 μm) |
None |
n/a |
Intermediate AMD |
Large (>125 μm) |
Abnormalities present 2 |
n/a |
Late AMD |
Large (>125 μm) |
Abnormalities present 2 |
Neovascular AMD and/or any geographic atrophy |
Retinal drusen and basal laminar deposits are sine qua non features of early AMD
[69,70,71,72][69][70][71][72]. Drusen is formed by the exudative accumulation of various cellular waste products and atherogenic debris (>40% of drusen volume)
[69,70,71,72,73][69][70][71][72][73]. Notably, the detection of subretinal drusenoid deposits above the RPE, also known as reticular pseudodrusen, are correlated with a two-fold increase in the risk of developing late-stage geographic atrophy
[72,74][72][74]. Further retinal injury is evident by the presence of pigmentary changes (depigmentation or hyperpigmentation) found in the RPE, indicating abnormalities associated with intermediate AMD
[15,70][15][70]. Often, these more preliminary stages of AMD do not carry obvious changes in visual function or may present elusive symptoms of mild distortion in central vision, with greater difficulty under low light conditions, such as reading or seeing at night
[15,65][15][65]. Some reports suggest that dark adaptation dysfunction, marked by impaired recovery of light sensitivity mediated by rod photoreceptors under mesopic conditions, may be one of the earliest indications of AMD onset
[75,76,77,78,79,80][75][76][77][78][79][80].
According to its pathophysiology, late AMD can be divided into neovascular AMD and geographic atrophy (GA), based on distinct clinical manifestations
[15]. Neovascular proliferations are characterized in accordance with the compartment wherein the choroidal neovascularization complex occupies, and the formation of chorioretinal anastomoses is found upon fluorescein angiography. With the advent of optical coherence tomography (OCT) imaging modalities, the anatomical classification of neovascular subtypes is made possible, which includes: type 1 neovascularization (NV) for submacular lesions confined below the RPE, type 2 NV for subretinal lesions found in the space between the photoreceptor layer and RPE, and type 3 NV for intraretinal lesions found within the retinal layers
[15,81,82,83,84,85,86,87,88,89][15][81][82][83][84][85][86][87][88][89]. Subsequently, the aberrant angiogenesis of fragile vasculature is often accompanied by the presence of a retinal hemorrhage, a hard exudate, detachments of the RPE, and chorioretinal fibrotic scarring
[15,82,83,84,87,89,90][15][82][83][84][87][89][90]. On the other hand, non-neovascular complications, seen in GA, are characterized by the demarcated regions of hypopigmentation, in consequence of the cumulative degeneration of the outer neurosensory retina, RPE, and choriocapillaris
[15,64,68,90,91,92][15][64][68][90][91][92]. Perilesional hyperpigmentation delineates areas of ongoing atrophy and the confluent superimposition of RPE cells
[93,94,95][93][94][95]. Both forms of late AMD result in severe visual defects, such as the formation of central scotoma, while peripheral vision remains relatively intact. Neovascular AMD often develops quite rapidly, leading to acute vision loss within a relatively short period of time (days to months); meanwhile, atrophic lesions progress more gradually and may take years, or decades, for symptoms to manifest
[15].
Given the increasing prevalence and substantial implications on quality of life, the detection of these phenotypic lesions in each stage of AMD is of profound importance for disease management and clinical screening. Thus, apposite surrogates of AMD pathology will likely serve a cardinal role in preventing the onset of extensive neurodegeneration within outer retinal layers and irreversible loss of photoreceptor cells.
1.2. Macular Pigment Optical Density in AMD—Background
The xanthophyll carotenoids lutein and zeaxanthin, as well as an isomer of lutein
meso-zeaxanthin, serve an important role in sustaining the integrity of the retina concomitant with optimizing central visual acuity
[42,96,97][42][96][97]. Collectively, these lipid-soluble carotenoids comprise the macular pigment, which forms a yellow spot that is seen during ophthalmoscopy. A recent imaging study determined that the spatial profiles of lutein and zeaxanthin are both localized in the fovea, as previously described
[42,96,97,98,99,100][42][96][97][98][99][100]; however, only zeaxanthin was primarily concentrated in the inner plexiform (IPL), outer plexiform (OPL), and outer nuclear layers
[98]. Meanwhile, lutein distribution was more dispersed throughout the macula, in reduced concentrations, when compared to foveal zeaxanthin levels
[98]. Humans are unable to naturally synthesize lutein and zeaxanthin
[97,101,102][97][101][102]; therefore, they must be acquired through the dietary consumption of foods, such as spinach, kale, and cruciferous green leafy vegetables, as well as corn and egg yolks
[42,97,103,104][42][97][103][104]. On the other hand,
meso-zeaxanthin is a biochemical isomer, also found in the macula, that is configured from lutein metabolism via RPE65 isomerase activity
[105,106][105][106] within the retinal pigment epithelial cells
[42,97,98,101,107][42][97][98][101][107]. A growing body of evidence indicates that the depletion of these carotenoids, marked by low macular pigment optical density (MPOD), may be a clinical biomarker associated with greater risk of incident retinopathy and visual dysfunction
[42,43,108,109,110][42][43][108][109][110].
Numerous reports have shown clinical benefits, by raising the levels of xanthophylls in the retina through dietary supplementation, thus, adjunctive carotenoid vitamin therapy may offer enhanced neuroprotection by augmenting MPOD and subsequently preventing further injury
[42,96,97,101,107,111,112,113,114,115,116,117,118,119,120,121,122][42][96][97][101][107][111][112][113][114][115][116][117][118][119][120][121][122]. Higher levels of MPOD are thought to preserve retinal tissue, specifically the layers containing photoreceptors in the fovea, through two primary mechanisms: (1) serving as an innate optical filter against blue light and (2) as protective antioxidants, by neutralizing free radicals and reducing consequent oxidative injury
[97,103,108,112,123,124,125][97][103][108][112][123][124][125]. The peak wavelength of the absorption spectrum of the macular (~460 nm) attenuate proliferation of reactive oxygen species is generated by photosensitizers, such as rod and cone cells, exposed to a range of visible blue light (400–500 nm)
[96,123,126][96][123][126]. This optical filtration is particularly significant, as short-wavelength (blue) light is highly reactive and has the capacity to exacerbate photo-oxidative degeneration in the most sensitive layers of the neurosensory retina
[42,97,123,124,125,126,127][42][97][123][124][125][126][127].
1.3. Measuring MPOD
While several imaging techniques are used to measure MPOD non-invasively within optometry settings, each possess their own set of advantages and disadvantages. The abilities and shortcomings of the MPOD measuring techniques are outlined in more detail elsewhere
[42,95,99,103,127,128,129,130,131,132,133,134,135,136][42][95][99][103][127][128][129][130][131][132][133][134][135][136]. In brief, the standard routine methods of heterochromatic flicker photometry (HFP) and customized flicker photometry (cHFP)
[42,99,103,128,129,130,131][42][99][103][128][129][130][131] utilize a psychophysical approach, wherein the determination of macular pigment levels is reliant upon subjective participation
[137,138,139][137][138][139]. Objective techniques of fundus reflectometry
[114[114][140][141][142][143][144],
140,141,142,143,144], autofluorescence (AFI)
[95,132[95][132][134][135][145],
134,135,145], and resonance Raman spectroscopy
[98,146,147,148,149][98][146][147][148][149] collect MPOD measurements, utilizing physical properties of light within the retina
[42,95,103,128,133,136,150][42][95][103][128][133][136][150].
1.4. MPOD Biomarkers in Clinical AMD
There is an overwhelming need for developing improved biomarkers that underscore the diverse pathology and subtypes found in patients with AMD. While current treatments have shown success for late neovascular AMD, there are a lack of proven therapies involving the mechanisms underlying early/intermediate stages and late atrophic stages of disease; in such cases, AMD develops, in consequence of the compounding cycles of oxidative stress and para-inflammation
[151,152,153][151][152][153]. Therefore, it is critical that therapeutic targets are aimed at ameliorating the perturbations contributing to lesion formation and preventing irreversible retinal neurodegeneration. Biomarkers are important tools, with the capacity to significantly aid the development of novel therapeutics, in addition to investigating the efficacy and overall safety of available treatments
[154,155,156,157][154][155][156][157]. Given that a single biomarker may be appropriate for different clinical utility, it is deemed necessary to clearly define the situation-specific context of how a particular biomarker will be used accordingly
[154,155,156][154][155][156].
While advancements in multimodality imaging have improved the prognosis for diagnosing retinal abnormalities, these modalities have also enabled the measurement of macular pigment status, to serve as a biomarker in multiple settings for AMD. It has been well-documented that MPOD levels are substantially reduced in AMD patients
[107[107][110][158][159][160][161][162][163][164],
110,158,159,160,161,162,163,164], which may be explained, at least in part, to similar risk factors shared between them
[5,21,31,107,110,158,159,160,161,162,163,164,165,166,167,168][5][21][31][107][110][158][159][160][161][162][163][164][165][166][167][168]. Diagnostic assessment, incorporating MPOD measurements, in conjunction with standard fundoscopic imaging, may offer unique clinical insight into the current state of the individual’s retinal health. In fact, macular pigment levels represent the local equilibrium between pro-oxidant stressors and antioxidant defenses in the retina, which can be attributed to its slow biological turnover
[42]. To this accord, MPOD measurement may function as: (1) a prognostic biomarker to appraise the health of neuroretinal layers, (2) a susceptibility/risk biomarker for screening those at risk of incident AMD, and (3) a pharmacodynamic/response biomarker to determine the clinical benefits of carotenoid vitamin therapy in AMD.
As a prognostic biomarker, MPOD levels may be used to monitor the progression of neurodegenerative changes in the photoreceptors and ganglion cells among patients with early or intermediate AMD. One study found that macular pigment levels were positively correlated with central retinal thickness, along with the neural volume of the ganglion cell layer (GCL), inner plexiform layer, and outer nuclear layer
[169]. Previous reports have demonstrated differential morphology changes on OCT within the outer retinal layers in patients with early AMD, including the thickness and volume of the photoreceptor layer, as well as the RPE-Bruch’s membrane complex
[71,92,170,171,172,173,174,175][71][92][170][171][172][173][174][175]. Similarly, inner retinal alterations are also found in the macular ganglion cell complex, comprised of the IPL, ganglion cell layer, and nerve fiber layer
[175[175][176][177][178][179][180],
176,177,178,179,180], which correspond to the dendrites, cell bodies, and axons of the neurosensory ganglion cells, respectively. Thus, MPOD depletion may serve to help prognosticate visual outcomes before severe impairment develops in early/intermediate AMD patients.
Given its bilateral nature, AMD fellow eyes may be considered to represent the pre-disease condition, in the absence of early retinal lesions, based on the incidence of fellow eye involvement, which increases significantly over time
[158,159,181,182][158][159][181][182]. Recently, Nagai et al. determined the risk of late AMD fellow eyes developing incident maculopathy was significantly associated with the combination of reduced MPOD (<0.65 density units (DU), measured by HFP) and photoreceptor outer segment length (<35 μm on OCT)
[183]. These results suggest MPOD screening may be an important susceptibility/risk biomarker used for the early detection of subclinical neurodegeneration among older adults and eyes with greater risk of developing AMD.
Furthermore, serial measurement of MPOD is used as pharmacodynamic/response biomarkers in randomized clinical trials to evaluate the protective benefits of carotenoid supplementation in patients with AMD, as discussed in more detail below. In summary, MPOD levels could be used to function as susceptibility/risk, monitoring, and pharmacodynamic/response biomarkers, in accordance with FDA-NIH guidelines
[154,155][154][155].
2. Carotenoids and Risk of AMD (Observational Studies)
Currently, dietary modifications remain the mainstay of therapeutic strategies, to potentially delay or prevent both the development and progression of AMD. The Age-Related Eye Disease Study (AREDS) is considered to be among the most influential large-scale clinical trials highlighting the relationship between dietary antioxidants and the risk of AMD progression
[181]. Reports indicate that regular consumption of the AREDS micronutrient formula (containing vitamin C, vitamin E, beta-carotene, and zinc) offered modest benefits, reducing the risk of late AMD progression by up to 25% during a five-year follow-up with at risk patients
[181]. In aging retinae, it is believed that the depletion of endogenous and exogenous antioxidants represents a critical driver in exacerbating neurodegenerative mechanisms. In fact, there is substantial evidence in favor of the neuroprotective association, between greater dietary consumption of carotenoid nutraceuticals, increased lutein and zeaxanthin concentrations in serum, and AMD prevention. A summary of these observational epidemiology studies is outlined in
Table 3.
Table 3. Epidemiology studies on AMD risk associated with dietary intake and/or serum levels of lutein and zeaxanthin.
Authors (Year) |
Study Name |
Participants |
Follow-Up |
Assessment of L/Z |
Results |
Seddon (1994) [197][184] |
EDCCS |
356 AMD patients, 520 controls in USA; aged 55–80 years |
- |
Dietary L/Z |
Highest quintile of L/Z intake, such as spinach and collard greens, strongly associated with reduced risk of late AMD |
VandenLangenberg (1998) [201][185] |
Beaver Dam Eye Study |
1709 individuals in USA; aged 43–84 years |
5 years |
Dietary L/Z |
No significant association reported between incident large drusen and dietary intake |
Mares-Perlman (2001) [194][186] |
NHANES III |
8596 individuals in USA; aged ≥40 years |
- |
Dietary L/Z |
Significantly lower risk of pigmentary abnormalities and late AMD in highest L/Z quintiles |
Snellen (2002) [198][187] |
- |
72 AMD patients, 66 controls in Netherlands; aged ≥60 years |
- |
Dietary L/Z |
Low dietary intake significantly associated with higher risk of neovascular AMD |
Cho (2004) [188] |
NHS and HPFS |
77,562 female and 40,866 male health professionals in USA; aged ≥50 years |
18 years;
12 years |
Dietary L/Z |
No significant association between relative risk of age-related maculopathy and vegetable consumption or carotenoid intake |
Van Leeuwen (2005) [200][189] |
The Rotterdam Study |
4170 individuals in Netherlands; aged 55–95 years |
8 years |
Dietary L/Z |
No significant association reported between dietary L/Z intake and incident AMD |
Moeller (2006) [196][190] |
CAREDS |
1787 women in USA; aged 50–79 years |
7 years |
Dietary L/Z |
Protective association among adult women (<75 years) with stable dietary intake and no history of chronic disease |
AREDS Research Group (2007) [51] |
AREDS |
4159 AREDS participants in USA; aged 60–80 years |
- |
Dietary L/Z |
Top quintile of dietary L/Z inversely associated with large drusen, neovascular AMD, and geographic atrophy |
Tan (2008) [199][191] |
Blue Mountains Eye Study |
2454 individuals in Australia; aged 49–93 years |
10.5 years |
Dietary L/Z |
Greater intake of L/Z saw reduced risk developing soft/reticular drusen and neovascular AMD progression |
Cho (2008) [187][192] |
NHS and HPFS |
71,494 female and 41,564 male health professionals in USA; aged 50–79 years |
18 years;
16 years |
Dietary L/Z |
A non-linear, inverse association seen among top quintiles of L/Z intake and neovascular AMD in both cohorts |
Ho (2011) [16] |
The Rotterdam Study |
2167 individuals in Netherlands; aged ≥55 years |
8 years |
Dietary L/Z |
Top tertile of L/Z intake significantly reduced incident early AMD in those with greater genetic risk |
Wu (2015) [202][193] |
NHS and HPFS |
63,443 female and 38,603 male health professionals in USA; aged 50–90 years |
26 years;
24 years |
Dietary L/Z |
Greater consumption of cooked spinach (0.5 cup, >1 serving/wk) inversely associated with intermediate AMD. Late AMD risk significantly lowered by up to 40% with higher L/Z intake |
Arslan (2019) [186][194] |
- |
100 AMD patients, 100 controls in Turkey; aged ≥50 years |
- |
Dietary L/Z |
Non-significant association observed between serum L/Z |
EDCCS Group (1993) [191][195] |
EDCCS |
421 AMD patients, 615 controls in USA; aged 55–80 years |
- |
Serum L/Z |
Protective association with greater serum L/Z levels and risk of neovascular AMD |
Mares-Perlman (1995) [193][196] |
Beaver Dam Eye Study |
167 AMD patients, 167 controls in USA; aged 43–84 years |
- |
Serum L/Z |
No overall association between serum L/Z and risk of late AMD |
Gale (2003) [192][197] |
- |
380 individuals in Sheffield, United Kingdom; aged ≥60 years |
- |
Serum L/Z |
Serum Z strongly associated with risk of incident early and late AMD |
Dasch (2005) [189][198] |
MARS |
586 AMD patients, 182 controls in Germany; aged 59–82 years |
- |
Serum L/Z |
No significant association reported between serum L/Z levels |
Delcourt (2006) [190][199] |
POLA |
640 individuals in Sète, France; aged ≥60 years |
- |
Serum L/Z |
Highest combined serum L/Z has significantly reduced risk |
Michikawa (2009) [195][200] |
- |
722 individuals in Karabuchi Town of Takasaki City, Japan; aged ≥65 years |
- |
Serum L/Z |
No significant association found between serum L/Z |
Zhou (2011) [109] |
- |
174 AMD patients, 89 controls in China; aged 50–88 years |
- |
Serum L/Z |
Significant inverse association between serum Z and neovascular AMD |