Age-related macular degeneration (AMD) is the most important cause of blindness in industrialized countries in individuals over 65 years of age. In the United Kingdom, for example, overall prevalence of late AMD is 2.4% of the population aged 50 years or more (Table 1) increasing to 4.8% in those 65 years of age or more and 12.2% in individuals greater than 80 years of age.1–4 As a consequence, there are estimated to be 513,000 individuals currently in the UK with the visual impairment characteristic of AMD suitable for registration as seriously visually impaired, the number of female cases being greater than males. This figure is predicted to rise to 679,000 by 20204 as the proportion of more elderly individuals in the population continues to increase, a phenomenon likely to be repeated in many countries.

There are two forms of AMD, viz., the ‘atrophic’ or ‘dry’ form and the ‘exudative’ or ‘wet’ form. The dry form is characterized by degeneration of the macular pigment epithelium, choriocapillaris, and photoreceptor cells. Larger areas of degeneration result from the merging of small areas of atrophy and this state is often referred to as ‘geographic atrophy’. By contrast, in the wet form, new choroidal vessels are formed (‘neovascularisation’) and ultimately, retinal detachment, edema, and hemorrhage may occur resulting in degeneration of the macula.

In the last decade, collaboration between geneticists, ophthalmologists, and optometrists has revealed significant evidence that genetic factors are involved in AMD5–8 and play a significant role in the disease alongside other risk factors such as smoking,9–12 diet,13–18 and exposure to sunlight.19,20 Several genes have now been implicated as possible risk factors,21–26 the strongest case being made for genes involved in immune modulation and the complement system. Hence, genetic predisposition to AMD may increase susceptibility to other risk factors such as smoking, diet, and exposure to sunlight. The objectives of this review are to discuss the results of these interdisciplinary studies and most specifically to identify those genes that have been associated with an increased risk of AMD. These genes are discussed in five main categories: (1) immune modulation and the complement system, (2) membrane transport, (3) blood vessel development and vascular modulation, (4) lipid metabolism, and (5) miscellaneous genes. In addition, insights that these genes provide as to pathogenesis of AMD together with protective factors, modifiable risk factors, and risk factor assessment scores are also discussed. A list of abbreviation and acronyms used in this review is given in Table 2.

Many of the genes implicated as possible risk factors for AMD involve single nucleotide polymorphisms (SNP) in which one amino acid within the protein is substituted for another. Many substitutions commonly result from ‘missense mutations’, i.e., the alteration of one base within a codon. SNP may have varying consequences. Those mutations which give rise to the substitution of a similar amino acid within a protein, such as valine for alanine, result in little protein modification. By contrast, other substitutions may be of greater consequence, resulting in modified protein structure, through greater numbers of disulphide bridges, modified hydrophobicity and charge, which in turn, compromise protein function and interaction with other molecules. A further two types of substitution are less common but potentially more disruptive than missense mutations. First, a codon is modified by base substitution from one coding for an amino acid such as glutamate, to a stop or termination codon (‘null mutation’ or ‘nonsense mutation’) and frequently results in a truncated protein with little natural function. Second, ‘splice-site’ mutations, i.e., a substitution in a region separating an exon from an intron can result in the incorporation of intron DNA into protein synthesis, resulting in proteins markedly different from their natural counterparts. These genetic changes often occur as natural allelic variations within the population conferring either increased risk or protection against AMD. By contrast to SNP, copy number variation, i.e., variation in the number of copies of a specific DNA sequence within a gene appears to be less important, although a small number of such associations with AMD have been recently reported.27 An alphabetical list of the major genes currently associated with AMD, their functions, and locations within the genome together with their abbreviations is given in Table 3.

A number of genes linked to the development of blood vessels or modulation of vascular processes have been implicated in AMD and these genes are interesting because of the involvement of the vasculature in wet AMD (neovascularization).

In addition to the above groups of genes, a number of genes of varied or unknown function show some statistical association with AMD. Their significance is more problematic and further research will be necessary to establish their significance.

The most important genes implicated in AMD appear to be those involved in immune modulation and the complement system such as CFH,26,28–31CFB,32C3,22,33 and SERPING123 strongly suggesting that inflammatory processed are important in AMD. It has been concluded that SNP in CFH, C3, HTRA1, and SERPING1 together could account for 45% of the risk of developing AMD.23 Genes associated with membrane transport, e.g., ABCR35–38 and CACNG47, the vascular system, e.g., FGF2,22LOXL121,30 and SELP,31 and with lipid metabolism, e.g., APOE52–54 and LIPC55 are also involved. Of these, the vascular genes are of particular interest because of the involvement in neovascularization. In addition, a series of other putative genes of variable or unknown function could be involved in AMD including ARMS2,58,59ATP synthase,60ERCC6,61 and HTRA121,24,35 and it is likely that many further genes will be identified.

Hence, there are likely to be multiple pathways leading to the same ultimate pathology in AMD, i.e., loss of retinal photoreceptors. Nevertheless, it is possible that carriers of high-risk genotypes may show detectable and specific visual changes before overt pathology is present. Feigle et al.64 assessed the critical fusion frequency (CFF) in individuals with normal vision who were carriers of high-risk variants of genes such as CFH, LOC, and HTRA1. CFF mediated by rods and cones was significantly reduced in gene variant-positive rather than gene variant-negative individuals although CFF mediated by cones alone was unaffected. This was the first study to relate early AMD risk genotypes to visual function suggesting that there may be many other gene/vision relationships to be identified in AMD.

Various nutritional factors may be involved in the pathogenesis of AMD including anti-oxidants and anti-oxidant co-factors, a number of vitamins, zinc, and anti-free radicals such as lutein, zeaxanthin and components of the membrane of photoreceptor cells.65 Consumption of omega-3 fatty acids in combination with anti-oxidants has been suggested as a preventative strategy66 but no fully developed double blind studies have been carried out to date to test this hypothesis. Eating more cold-water fish rather than red meat and adding nuts to the diet have all been suggested to reduce the risk.64 For example, Chua et al.14 suggested that a 40% reduction in early AMD could be achieved by changing to a diet with significantly increased fish consumption.67 In addition, there is evidence that supplementing the intake of DHA and eicosapentaemic acid (EPA) and reducing dietary glycemic index (dGI) may be protective against AMD in the ‘Age Related Eye Disease Study (AREDS) trial.17

A review of all randomized controlled trials comparing anti-oxidant vitamins concluded, however, that there was little benefit of any anti-oxidant combination in AMD.68 They also concluded that there was accumulating evidence that neither vitamin E nor β-carotene in themselves could prevent or delay the onset of AMD. In addition, Delcourt et al.69 studied the relationship between nutrition and AMD and found that individuals taking the antioxidants, lutein, zeaxanthin, and omega-3 poly unsaturated fatty acids had a lower risk of AMD. A similar result was reported by Lecerf and Desmettre.70 By contrast, Chong et al.71 using meta-analysis, studied the effect of dietary anti-oxidants in nine prospective cohorts and three random clinical trials. Pooled results suggested that vitamins A, C, and E, zinc, lutein, zeaxanthin, carotene, P-cryptoxanthin, and lycopene had little or no effect on preventing early AMD. However, there was evidence that antioxidant supplements containing zinc could prevent visual loss in individuals with early AMD.72

Several studies have proposed additional lifestyle changes to reduce the lifetime risk of AMD that should be considered in patients in high genetic risk categories. Hence, high fat intake may be associated with an increased risk of AMD in both women and men. Reducing fat intake to this level means reducing the consumption of red meat and dairy products such as whole milk, cheese, and butter which are high in saturated fats. Higher red meat intake has been positively associated with early AMD whereas consumption of chicken may be negatively associated with AMD.73 There is considerable evidence that the incidence of AMD varies with both ethnicity and geographical location which could be attributed in part to varying diet. Hence, the risk of AMD varies among Americans of different Asian origin74, overall, 5.4% of Asian Americans having dry and 0.49% wet AMD, Chinese and Pakistani Americans having a significantly increased risk of dry AMD compared with non-hispanic whites. By contrast, Japanese Americans have a 29% decreased risk of dry AMD.74 There were no significant differences in the risk of wet AMD in Asian Americans of any ethnicity compared with white Americans. Outside the US, a Japanese study at Funagata75 found that the prevalence of early AMD in individuals greater than 35 years of age was 3.5%, and of late AMD 0.5%, while the age-standardised prevalence in the right eye was 4.1%. In rural China, prevalence of wet and dry AMD is 1.45% and 1.55% respectively, increasing with age.76 An overview of epidemiological studies from around the world72 suggests prevalence of early AMD is similar in the white, black, and hispanic populations, but that whites have a greater prevalence of late AMD. Hence, prevalence studies provide considerable evidence that environmental, and lifestyle differences among populations as well as genetic variation are likely to be involved in determining the risk of AMD.

The identification of specific genetic defects leads to the possibility of more widespread screening for AMD especially in affected families and hence, the possibility of identifying asymptomatic carriers of the disease. A review of research to date suggests that smoking is the single most important modifiable risk factor for AMD.9–12 Studies taken collectively suggest that long-term smoking is associated with approximately a doubling of the risk of late-onset AMD. Hence, it is particularly important that patients with a familial history of the disease should cease smoking. A risk assessment model has been developed for advanced AMD and is available for online use.7 Based on data from 2846 AMD patients followed over 9.3 years, Cox proportional hazard analysis was used to assess the significance of demographic, environmental, phenotypic, and genetic covariates. The final model incorporates age, smoking behavior, family history, and genetic variation (especially of CFH and ARMS2) to predict the risk of AMD.7 Grassman et al.77 have constructed a genetic risk score for AMD based on 13 ‘at risk’ variations of eight individual genes. AMD patients younger than 75 years of age had a significantly higher score compared with those patients older than 75. In addition, Seddon et al.6 have proposed a prediction model for the risk of advanced AMD based on genetic, demographic and environmental variables. This model included several genetic variants including CFH, LOC, C2, CFB, and C3, all of which were related to the prevalence of advanced AMD, and all, with the exception of CFB, significantly related to the progression to advanced AMD. In addition, smoking was independently related to AMD with a multiplicative joint effect with genotype. These advances have also resulted in the development of specific screening tests for AMD, such as the ‘macula risk test’ which involves taking a sample of cells from the cheek of the patient for genetic analysis.78 The results of the test include a risk estimate based on the genetic polymorphisms identified and genetic counselling support. The test has five categories of risk determined by an algorithm that takes into account genetics and also smoking history. Hence, risk category 1 is associated with a lifetime risk of 0.65–4.6% whereas risk category 4 is associated with a 55–94% lifetime risk.

Modifiable risk factors for AMD should be discussed with patients whose lifestyle and/or family history place them in an elevated risk category. Protective modifications include: cessation of smoking, sunglass protection under conditions of elevated light intensity, dietary modifications to reduce intake of saturated fats and increase omega 3 fatty acids. Calculation of AMD risk using current models should be recommended as a tool for patient education. Furthermore, AMD management frequency may be influenced by an assessment of genetic risk, as such provisions are likely to become more widely available. Interactions between different genes and between genes and the environment also play a prominent role in the development of AMD and are likely to be profitable areas of future research.