December 2020
Volume 61, Issue 14
Open Access
Genetics  |   December 2020
The Effect of Genetic Variants Associated With Age-Related Macular Degeneration Varies With Age
Author Affiliations & Notes
  • Tina Schick
    AugenZentrum Siegburg, MVZ ADTC Siegburg GmbH, Siegburg, Germany
    Department of Ophthalmology, University Hospital of Cologne, Cologne, Germany
  • Laura Lorés-Motta
    Department of Ophthalmology, Donders Institute for Brain, Cognition and Behaviour, Radboud University Medical Centre, Nijmegen, the Netherlands
  • Lebriz Altay
    Department of Ophthalmology, University Hospital of Cologne, Cologne, Germany
  • Lars G. Fritsche
    Center for Statistical Genetics, Department of Biostatistics, School of Public Health, University of Michigan, Ann Arbor, Michigan, United States
  • Anneke I. den Hollander
    Department of Ophthalmology, Donders Institute for Brain, Cognition and Behaviour, Radboud University Medical Centre, Nijmegen, the Netherlands
    Department of Human Genetics, Donders Institute for Brain, Cognition and Behaviour, Radboud University Medical Centre, Nijmegen, the Netherlands
  • Sascha Fauser
    Department of Ophthalmology, University Hospital of Cologne, Cologne, Germany
    F. Hoffmann-La Roche, Basel, Switzerland
  • Correspondence: Tina Schick, Department of Ophthalmology, University Hospital of Cologne, Kerpener Straße 62, 50937 Cologne, Germany; ristau.tina@gmail.com
  • Footnotes
    *  TS and LL-M contributed equally to the work presented here and should therefore be regarded as equivalent authors.
Investigative Ophthalmology & Visual Science December 2020, Vol.61, 17. doi:https://doi.org/10.1167/iovs.61.14.17
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Tina Schick, Laura Lorés-Motta, Lebriz Altay, Lars G. Fritsche, Anneke I. den Hollander, Sascha Fauser; The Effect of Genetic Variants Associated With Age-Related Macular Degeneration Varies With Age. Invest. Ophthalmol. Vis. Sci. 2020;61(14):17. doi: https://doi.org/10.1167/iovs.61.14.17.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: The prevalence of age-related macular degeneration (AMD) increases dramatically with age. This large collaborative study investigates the effects of 51 late-AMD–associated genetic variants in different ages, focusing on individuals above the age of 90 years.

Methods: The study included 27,996 individuals of the International AMD Genomics Consortium; 14,539 showed late AMD (51.9%) and 13,457 were controls (48.1%). Four age groups were compiled: 60 to 69 years, n = 6514, AMD = 2210 (33.9%); 70 to 79 years, n = 12228, AMD = 6217 (51.7%); 80 to 89 years, n = 8285, AMD = 5326 (64.3%); and ≥90 years, n = 969, AMD = 686 (70.8%). The effect sizes of 51 AMD-associated genetic variants were calculated for all age groups and were compared among the age groups.

Results: Six variants were associated with late AMD in individuals ≥ 90 years of age (P ≤ 0.0006). For rs10922109 and rs570618 (both in CFH), the minor allele (MA) was protective, and minor allele frequency (MAF) increased with age in cases and controls. For rs116503776 in C2/CFB/SKIV2L, the MA was protective, and MAF increased in cases. For rs3750846 in ARMS2/HTRA1, the MA increased risk, and MAF was lower in cases with increasing age. For rs6565597 in NPLOC4/TSPAN10, the MA increased risk. For rs5754227 in SYN3/TIMP3, the MA was protective, and there was no consistent variation in MAF with age. Variants in CFH and ARMS2 showed lower effect sizes at greater age. Interaction analysis showed strong age-related effects for rs570618 (P = 2.24 × 10–7) and rs3750846 (P = 0.001). Total genetic risk was lower in individuals ≥ 90 years old (area under the curve [AUC], 0.795) than in those 70 to 79 years old (AUC, 0.831; P = 0.03)

Conclusions: Effect sizes and MAF of genetic risk factors for late AMD differed among the age groups. These results could guide future work on AMD risk assessment in older individuals.

Age-related macular degeneration (AMD) is a retinal disease and the main cause of severe vision impairment above the age of 50 in Western countries.1 The early forms of AMD show drusen between Bruch's membrane and the retinal pigment epithelium. Late AMD is characterized by geographic atrophy of the outer retina and/or choroidal neovascularization with accumulation of fluid in or under the retina, resulting in vision loss. Multiple environmental and genetic factors are involved in the etiology of AMD.2,3 Several AMD susceptibility loci have been identified.4 Genome-wide association studies (GWASs) and large-scale resequencing initiatives have identified a large number of single nucleotide polymorphisms (SNPs) that increase the risk for AMD.5 The International Age-Related Macular Degeneration Genomics Consortium (IAMDGC) recently reported 34 independent AMD risk loci and 52 variants that were independently associated with late AMD.6 Several variants increase the risk of late AMD, such as rs3750846 in ARMS2/HTRA1. In addition, variants have been found in genes involved in the complement system, such as CFH, C2/CFB, C3, and CFI, that either increase or decrease the risk of AMD. Other risk-inferring and risk-decreasing variants have been found in genes involved in, for example, collagen pathways (e.g., COL15A1, COL8A1, LOXL2, COL4A3/COL4A4, MMP9, PCOLCE), lipid pathways (e.g., APOC2/APOE, CETP, LIPC, ABCA1, PLTP), and extracellular matrix pathways (e.g., TNXB, VAV1, VEGFA, ACTG1, BCAR1, COL4A4, ITGA7, MYL2, VTN). 
Aging is a very important factor in the pathogenesis of AMD. The prevalence of AMD in individuals above the age of 90 years has been reported to be almost 60%, with almost 1/3 of the AMD cases (28.3%) being late AMD.7 Especially in individuals of European ethnicity, the prevalence of the late forms of the disease increases rapidly after the age of 75 years.8,9 Because of the high prevalence of AMD in older individuals, the increased aging of the population adds to the health care burden due to AMD. Risk factors may not be equal in all age groups; therefore, evaluation of the risk factors in the different age groups, especially very old individuals, could lead to improved late-AMD risk assessment and patient care. Previously, we demonstrated that the known genetic risk alleles rs1061170 in CFH and rs10490924 in ARMS2 showed significantly lower effects on AMD development in individuals above the age of 90 years, whereas non-genetic factors maintained comparable effects with advanced age.10 However, so far, most studies have included only small numbers of participants, particularly older individuals, and have evaluated only a limited number of genetic variants. Thus, our understanding of the genetic risk factors for the different age groups, especially those 90 years of age and older, is still very limited. 
In this collaborative case–control study, we used the large IAMDGC genotyped cohort to investigate the effects of 51 known AMD-associated variants in 969 individuals 90 years of age and older compared to younger age groups. It is possible that age-specific genetic risk effects could be utilized to obtain more accurate disease risk assessments in older age groups. 
Methods
Study Population
Our analyses are based on individual participant data from 23 studies of the IAMDGC.6 In our study, we included data for 27,996 individuals in the IAMDGC study who were between 60 and 102 years of age. We divided them into four age groups: 60 to 69 years (n = 6514), 70 to 79 years (n = 12,228), 80 to 89 years (n = 8285), and 90 years and older (n = 969). These individuals passed subject-level quality control as previously described.6 Only unrelated individuals of European ancestry were included in the analysis. 
Our study included healthy controls and late-AMD cases with documentation of the disease status. The AMD status of all included individuals was determined by fundus examination and/or fundus photography. Sixteen of the 23 included study sites additionally used fluorescein angiography and/or optical coherence tomography images for the grading of choroidal neovascularization. The individuals were recruited from studies based on ophthalmology clinics and population-based studies. Four studies recruited their controls from among spouses or friends. The late-AMD cases had geographic atrophy or choroidal neovascularization related to AMD. Controls had fewer than five drusen and no pigmentary abnormalities in either eye, and they had to be defined as “no AMD” based on the grading system of the study site. Due to only a very small number of early or intermediate AMD cases among the individuals above 90 years of age, those individuals with early or intermediate AMD were not included in this study. Further inclusion and exclusion criteria, as well as detailed information on ophthalmological grading, quality control of genetic data, and imputation, are described elsewhere.6 
The study followed the tenets of the Declaration of Helsinki and was approved by the ethics committee of the University Hospital in Cologne (Ethics Commission of Cologne University's Faculty of Medicine) and the local ethics review boards at the other participating sites. Informed written consent was obtained from each participant. 
Genotyping
Genotyping and imputation using the 1000 Genomes reference panel16 were performed as described previously.6 
The genotypes of the initial IAMDGC study are available from the Database of Genotypes and Phenotypes under accession phs001039.v1.p1, and GWAS summary statistics are available at http://amdgenetics.org/. Very rare variants with an allele frequency (AF) below 0.001 were excluded from the analysis. For this reason, the very rare variant rs121913059 (CFH p.Arg1210Cys) was excluded, and we analyzed the genotype dosages of 51 independently associated common and rare variants distributed across 34 loci (Supplementary Table S3; see also Ref. 6). 
Statistical Analysis
Genetic associations of the 51 genetic variants and AMD status were assessed as previously described by using allele dosages and by means of Firth bias-corrected likelihood-ratio tests as implemented in the Efficient and Parallelizable Association Container Toolbox (EPACTS) adjusted for PC1, PC2, and source of DNA.6 Bonferroni correction was applied and P < 0.001 was considered statistically significant. Additionally, a sensitivity analysis was performed that included 885 individuals in each age group matched for sex, AMD status, source of DNA, and study cohort. Selection of the individuals for the sensitivity analysis was based on exact matches and performed using the nearest neighbor matching method of the R package MatchIt (R Foundation for Statistical Computing, Vienna, Austria). 
Interaction analyses were performed with computing models that included two age groups, and the interaction term was estimated for each pair of age groups. Moreover, an overall interaction term of age and SNP was calculated that included all individuals in the analysis. Regression models were calculated for the different age groups and included all 51 genetic variants; the area under the receiver operating characteristic curve (AUC) was also computed. Differences between the AUCs were calculated using the R package pROC. Power calculations were performed along the lines of Cohen17 using the R package pwr based on the odds ratios (ORs) for the overall population reported by Fritsche et al.6 
Results
Demographics
The dataset for individuals 90 years and older included 686 late-AMD cases (70.8%) and 283 controls without AMD (29.2%). The mean age of this study group was 92.08 ± 2.25 years (range, 90–102 years); 351 were male (36.2%), and 618 were female (63.8%). For comparison purposes, 27,027 individuals from other age groups were additionally included in the study (Table 1). All samples belonged to the IAMDGC. Additional information regarding the included IAMGC study cohorts is provided in Supplementary Table S1
Table 1.
 
Demographics of the Study Cohort
Table 1.
 
Demographics of the Study Cohort
Association of Genetic Variants with AMD in Individuals 90 Years and Older
The associations with late AMD were determined for 51 AMD-associated variants present in the study cohort (Supplementary Table S2). 
A total of six SNPs were significantly associated with AMD in the individuals 90 years and older after Bonferroni correction for 51 tests (P < 0.001): rs10922109 and rs570618 in the CFH locus, rs116503776 in the C2/CFB/SKIV2L locus, rs3750846 in the ARMS2/HTRA1 locus, rs6565597 in the NPLOC4/TSPAN10 locus, and rs5754227 in the SYN3/TIMP3 locus (Table 2). 
Table 2.
 
Significantly Associated Variants in the 90 Years and Older Group in All Age Groups
Table 2.
 
Significantly Associated Variants in the 90 Years and Older Group in All Age Groups
For rs10922109 and rs570618, minor allele frequency (MAF) increased in both cases and controls with increasing age, but the increase was greater in cases than in controls. For ARMS2 rs3750846, there were no age effects of MAF in the controls, but MAF in cases decreased with greater age. For C2/CFB/SKIV2 rs116503776, there was no consistent relationship between MAF and age in the controls, but in cases an increasing frequency of MAF with age was observed. For NPLOC4/TSPAN10, there was no consistent change of MAF with age, as MAF dropped only in the controls ≥ 90 years of age. In SYN3/TIMP3 rs5754227, no relationship between MAF with age in cases and controls could be detected (for details, see Table 2). Compared to the effects reported in the overall population (see also Supplementary Table S3),6 smaller effects were detected in the nonagenarian and older group for rs10922109 and rs570618 in the CFH locus and for rs3750846 in the ARMS2/HTRA1 locus (Table 2). Moreover, rs6565597 in NPLOC4/TSPAN10 and rs5754227 in SYN3/TIMP3 showed larger effects in the individuals 90 years and older, whereas rs116503776 in C2/CFB/SKIV2L showed similar effects (Table 2). 
Power analysis of all 51 variants analyzed showed that the study design had greater than 80% power to detect the association of a third variant at the CFH locus, rs61818925 (Fig. 1); however, no association was detected for this particular variant in the 90 years and older group (Supplementary Table S2). 
Figure 1.
 
Power analysis of 51 variants in individuals 90 years of age and older. Significantly associated variants are marked in red; rs61818925 (marked in blue) did not show significant associations despite more than 80% power.
Figure 1.
 
Power analysis of 51 variants in individuals 90 years of age and older. Significantly associated variants are marked in red; rs61818925 (marked in blue) did not show significant associations despite more than 80% power.
Genetic Risk Effects on AMD in the 90 Years and Older Group Were Different Compared to the Other Age Groups
Associations with late AMD for the 51 genetic variants in the other age groups are presented in Supplementary Table S2. For rs10922109, rs570618 in the CFH locus, and rs3750846 in the ARMS2/HTRA1 locus, the genetic effects decreased with increasing age (Figs. 2a–2c). For rs116503776 (C2/CFB/SKIV2L) and rs5754227 (SYN3/TIMP3), there were no significant differences in the effect sizes on late AMD in the different age groups (Figs. 2d, 2e). Interestingly, rs6565597 (NPLOC4/TSPAN10) showed larger effects in the 90 years and older group compared to two younger age groups (70–79 years and 80–89 years) (Fig. 2f). Details of the interactions in the 90 years and older group are displayed in Table 3
Figure 2.
 
Odds ratios in the different age groups of the six variants showing significant associations in individuals 90 years of age and older. (a) rs10922109 (CFH), (b) rs570618 (CFH), (c) rs3750846 (ARMS2/HTRA1), (d) rs116503776 (C2/CFB/SKIV2L), (e) rs5754227 (SYN3/TIMP3), (f) rs6565597 (NPLOC4/TSPAN10). *P-value interaction with the 90 years and older group < 0.05. (For details, see Table 3.)
Figure 2.
 
Odds ratios in the different age groups of the six variants showing significant associations in individuals 90 years of age and older. (a) rs10922109 (CFH), (b) rs570618 (CFH), (c) rs3750846 (ARMS2/HTRA1), (d) rs116503776 (C2/CFB/SKIV2L), (e) rs5754227 (SYN3/TIMP3), (f) rs6565597 (NPLOC4/TSPAN10). *P-value interaction with the 90 years and older group < 0.05. (For details, see Table 3.)
Table 3.
 
Interaction of Genetic Effects Between the 90 Years and Older Group and Other Age Groups and Interaction with Age as a Continuous Variable of Significantly Associated Variants
Table 3.
 
Interaction of Genetic Effects Between the 90 Years and Older Group and Other Age Groups and Interaction with Age as a Continuous Variable of Significantly Associated Variants
Looking at the allele frequencies of cases and controls among age groups, variations could be detected for some of them with increasing age (Supplementary Table S2). 
The distribution among sex, AMD status, source of DNA, and cohort was variable among the age groups, but there were also differences in the size of these age groups. Therefore, an additional sensitivity analysis was performed in which the age groups were matched on these factors (n = 885 for all age groups). These analyses showed comparable results (Supplementary Table S4). 
Receiver operating characteristic (ROC) curves including all 51 variants for all age groups are presented in Figure 3. The AUCs for the regression models (Table 4, Supplementary Table S5) were 0.825 for individuals 60 to 69 years of age, 0.831 for individuals 70 to 79 years of age, and 0.800 for individuals 80 to 89 years of age. 
Figure 3.
 
ROC curves and AUCs of regression models of the 51 variants in the different age groups. Red, 60 to 69 years; orange, 70 to 79 years; green, 80 to 89 years; blue, 90 years and older. (For details, see Table 4.)
Figure 3.
 
ROC curves and AUCs of regression models of the 51 variants in the different age groups. Red, 60 to 69 years; orange, 70 to 79 years; green, 80 to 89 years; blue, 90 years and older. (For details, see Table 4.)
Table 4.
 
AUCs of Regression Models for the Different Age Groups Including All Genetic Variants
Table 4.
 
AUCs of Regression Models for the Different Age Groups Including All Genetic Variants
In the group of individuals 90 years and older, the AUC was the lowest: 0.795 (Table 4). The AUC for the individuals 70 to 79 years old differed significantly from the 90+ group (P = 0.026) (Fig. 3Table 4). 
Discussion
In this study, we analyzed 51 genetic variants associated with late AMD in persons 90 to 102 years of age and compared the effect sizes to other age groups. In the oldest group, comprised of individuals 90 years and older, we detected six SNPs significantly associated with late AMD. The strongest associations were detected for rs10922109 and rs570618 located in the CFH locus and rs3750846 in the ARMS2/HTRA1 locus. Previous reports have shown the importance of variants in CFH and ARMS2 in the progression of drusen and progression into advanced AMD.1820 These very highly AMD-associated variants showed smaller effect sizes on late AMD in individuals over 90 years of age compared to the overall population and compared to younger age groups. For rs570618 in CFH and rs3750846 in ARMS2/HTRA1, we observed a strong interaction with age that supports this relationship. These results are in line with previous reports describing lower effects sizes of other AMD-associated variants in CFH in older individuals.1012 
For some variants, allele frequencies in cases and/or controls increased or decreased with increasing age. For rs10922109 and rs570618 in CFH, the minor allele increased protection, especially in cases but also in controls. For rs3750846 in ARMS2, the minor allele was risk conferring. No age-related effects of MAF could be found in controls, but MAF decreased with age in cases. These findings may be explained by a survivorship effect, which has also been discussed by Adams et al.11 
Minor alleles of rs5754227 in SYN3/TIMP3 and rs116503776 in C2/CFB/SKIV2L are infrequent, which could have resulted in no clear or consistent change of MAF with age being detected, and there was not a clear association with age in the interaction models. 
In our analysis, rs61818925 in CFH did not show significant associations with late AMD despite having sufficient power. Therefore, the protective effect of this variant on developing AMD might also be smaller in individuals above the age of 90 years. These findings point toward a smaller role for CFH and ARMS2 in late AMD development at very advanced ages. 
Previous studies have investigated the mortality risk resulting from AMD.2123 These reports describe associations of late AMD with mortality and therefore support the theory of the enrichment of non-risk, AMD-associated allele frequencies in a long-lived population. A large meta-analysis by McGuinness et al.24 also concluded that late AMD is associated with increased rates of mortality of all causes, especially cardiovascular mortality. The authors therefore suggested the existence of shared pathways between late AMD and systemic diseases, which was also shown by Grassmann et al.25 
Longevity studies describe the phenomenon of genetic differences between younger and older populations, explained by an enrichment of “longevity genes” in very old individuals.1315 Interestingly, a recent genome-wide association study for parent lifespan did not detect genome-wide significant associations with longevity or highly AMD-associated variants in CFH or ARMS2.26 The study did detect variant rs429358 in APOE as a factor for survival with an increasing lifespan for carriers of the T-allele, a finding that was also reported by Pilling et al.27 in 2017. In our study, APOE variant rs429358 showed significant associations in individuals 60 to 89 years of age with the C-allele as a protective factor with similar ORs. In contrast, in individuals older than 90 years of age, the effect was smaller and did not reach significance (Supplementary Tables S2). Although these findings may be explained by the smaller size of the oldest group, in our cohort the MAF was lower in very old individuals without AMD compared to other age groups (Supplementary Tables S2). Therefore, APOE may represent an example of a variant that is associated with both longevity and AMD. At higher ages, there are fewer carriers of the C-allele, possibly due to increased mortality. 
Interestingly, in our cohort, rs6565597 in NPLOC4/TSPAN10 showed a significant association with AMD in the 90 years and older age group, despite limited statistical power. In our study, the effect size for NPLOC4/TSPAN10 was higher in the 90 years and older group than in the overall population and was specifically higher than in the individuals 70 to 79 and 80 to 89 years of age, indicating that this variant may be a specific risk factor for late AMD in the elderly. This was also seen in the sensitivity analysis, in which an association of this variant was detected only in the oldest age group. On the other hand, the frequency of the risk-increasing minor allele dropped only in the controls who were 90 years and older; there were no consistent changes in the late-AMD cases and no interaction with age when analyzed as a continuous variable, thus making it difficult to reach a clear conclusion regarding rs6565597 in NPLOC4/TSPAN10
Regression models of all 51 SNPs showed the lowest discriminative power in the group of individuals 90 years and older and a similar AUC for the individuals 80 to 89 years of age. A significantly higher discriminative value was detected in the individuals 70 to 79 years of age. Therefore, we propose that the combined genetic influence of these 51 variants on AMD development decreases in individuals above the age of 80 years. As for rs570618 in CFH and rs3750846 in ARMS2/HTRA1, the strongest associations with age were found in the interaction analysis; these two variants likely are strong factors for lower AUCs in the two oldest groups. In older persons, it is plausible that primary age by itself and other non-genetic environmental factors such as smoking, alcohol use, physical activity, or sunlight exposure may become more relevant for disease risk. 
Our results suggest that, in younger AMD patients, a high genetic risk, especially regarding the common variants in CFH and ARMS2, induces and accelerates the disease at earlier ages. Additionally, rare, highly penetrant variants will provoke disease development at younger age. In older individuals, aging processes in the retina and the subretinal structures may influence disease progression despite the presence of fewer genetic risk factors. 
The results presented here are based on a large case–control cohort with a large number of individuals above the age of 90 years. Because it is very difficult to include very old individuals in clinical studies (especially controls without AMD), there are large differences in the sizes of the age groups that have been investigated. However, the results of our sensitivity analysis with matched group sizes show comparable results and suggest that no bias was introduced due to heterogeneity in the number of individuals, sex, AMD status, source of DNA, and cohort among the age groups. The multicenter case–control design with variable methodology does not allow the investigation of longitudinal or epidemiological parameters and limits reaching conclusions on real association trends with age. Furthermore, our analysis was restricted to 51 already known AMD-associated variants, and non-genetic factors were not included, further limiting interpretation. Future prospective studies using extended datasets and including both genetic and environmental factors are needed to identify additional AMD-associated risk factors in the nonagenarian and older population and to better draw conclusions on true survival effects. 
In summary, in our study, risk alleles in ARMS2 and CFH showed significantly smaller risk effects on AMD development in very old individuals. The influence of 51 known genetic risk factors on late AMD was less in the oldest individuals compared to persons between 70 and 79 years of age. Moreover, we have described for the first time, to the best of our knowledge, the genetic variant rs6565597 in NPLOC4/TSPAN10, which shows larger effects in individuals 90 years and older; therefore, rs6565597 may be a more specific risk factor for older persons. If these results are confirmed in other studies, they may have a significant impact on AMD risk assessment in an increasingly aging population. 
Acknowledgments
The authors thank the excellent International Age-Related Macular Degeneration Genomics Consortium (http://amdgenetics.org/), which is supported by a grant from the National Eye Institute, National Institutes of Health (R01 EY022310). The list of consortium members reflects the author list of the previous publication by Fritsche et al.6 
Disclosure: T. Schick, None; L. Lorés-Motta, None; L. Altay, None; L.G. Fritsche; None, A.I. den Hollander, None; S. Fauser, F. Hoffmann-La Roche (E) 
References
Congdon N, O'Colmain B, Klaver CC, et al. Causes and prevalence of visual impairment among adults in the United States. Arch Ophthalmol. 2004; 122(4): 477–485. [CrossRef] [PubMed]
Spencer KL, Olson LM, Schnetz-Boutaud N, et al. Using genetic variation and environmental risk factor data to identify individuals at high risk for age-related macular degeneration. PLoS One. 2011; 6(3): e17784. [CrossRef] [PubMed]
Seddon JM, Reynolds R, Maller J, Fagerness JA, Daly MJ, Rosner B. Prediction model for prevalence and incidence of advanced age-related macular degeneration based on genetic, demographic, and environmental variables. Invest Ophthalmol Vis Sci. 2009; 50(5): 2044–2053. [CrossRef] [PubMed]
DeAngelis MM, Owen LA, Morrison MA, et al. Genetics of age-related macular degeneration (AMD). Hum Mol Genet. 2017; 26(R2): R246. [CrossRef] [PubMed]
Black JR, Clark SJ. Age-related macular degeneration: genome-wide association studies to translation. Genet Med. 2016; 18(4): 283–289. [CrossRef] [PubMed]
Fritsche LG, Igl W, Bailey JN, et al. A large genome-wide association study of age-related macular degeneration highlights contributions of rare and common variants. Nat Genet. 2016; 48(2): 134–143. [CrossRef] [PubMed]
Hermann M, Caramoy A, Schroder S, Droge K, Kirchhof B, Fauser S. Prevalence of age-related macular degeneration in persons aged 90 years and older in Cologne. Acta Ophthalmol. 2012; 90(6): e500–e501. [CrossRef] [PubMed]
Wong WL, Su X, Li X, et al. Global prevalence of age-related macular degeneration and disease burden projection for 2020 and 2040: a systematic review and meta-analysis. Lancet Glob Health. 2014; 2(2): e106–116. [CrossRef] [PubMed]
Colijn JM, Buitendijk GHS, Prokofyeva E, et al. Prevalence of age-related macular degeneration in Europe: the past and the future. Ophthalmology. 2017; 124(12): 1753–1763. [CrossRef] [PubMed]
Ersoy L, Ristau T, Hahn M, et al. Genetic and environmental risk factors for age-related macular degeneration in persons 90 years and older. Invest Ophthalmol Vis Sci. 2014; 55(3): 1842–1847. [CrossRef] [PubMed]
Adams MK, Simpson JA, Richardson AJ, et al. Can genetic associations change with age? CFH and age-related macular degeneration. Hum Mol Genet. 2012; 21(23): 5229–5236. [CrossRef] [PubMed]
Grassmann F, Fritsche LG, Keilhauer CN, Heid IM, Weber BH. Modelling the genetic risk in age-related macular degeneration. PLoS One. 2012; 7(5): e37979. [CrossRef] [PubMed]
Slagboom PE, Beekman M, Passtoors WM, et al. Genomics of human longevity. Philos Trans R Soc Lond B Biol Sci. 2011; 366(1561): 35–42. [CrossRef] [PubMed]
Kenyon CJ . The genetics of ageing. Nature. 2010; 464(7288): 504–512. [CrossRef] [PubMed]
Beekman M, Blanche H, Perola M, et al. Genome-wide linkage analysis for human longevity: Genetics of Healthy Aging Study. Aging Cell. 2013; 12(2): 184–193. [CrossRef] [PubMed]
1000 Genomes Project Consortium, Abecasis GR, Auton A, et al. An integrated map of genetic variation from 1,092 human genomes. Nature. 2012; 491(7422): 56–65. [CrossRef] [PubMed]
Cohen J . Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates; 1988.
Dietzel M, Pauleikhoff D, Arning A, et al. The contribution of genetic factors to phenotype and progression of drusen in early age-related macular degeneration. Graefes Arch Clin Exp Ophthalmol. 2014; 252(8): 1273–1281. [CrossRef] [PubMed]
Seddon JM, Reynolds R, Yu Y, Daly MJ, Rosner B. Risk models for progression to advanced age-related macular degeneration using demographic, environmental, genetic, and ocular factors. Ophthalmology. 2011; 118(11): 2203–2211. [CrossRef] [PubMed]
Lambert NG, ElShelmani H, Singh MK, et al. Risk factors and biomarkers of age-related macular degeneration. Prog Retin Eye Res. 2016; 54: 64–102. [CrossRef] [PubMed]
Zhu Z, Wang W, Keel S, Zhang J, He M. Association of age-related macular degeneration with risk of all-cause and specific-cause mortality in the National Health and Nutrition Examination Survey, 2005 to 2008. JAMA Ophthalmol. 2019; 137(3): 248–257. [CrossRef] [PubMed]
Fisher DE, Jonasson F, Eiriksdottir G, et al. Age-related macular degeneration and mortality in community-dwelling elders: the age, gene/environment susceptibility Reykjavik study. Ophthalmology. 2015; 122(2): 382–390. [CrossRef] [PubMed]
Age-Related Eye Disease Study 2 Research Group; Papudesu C, Clemons TE, Agron E, Chew EY. Association of mortality with ocular diseases and visual impairment in the Age-Related Eye Disease Study 2: Age-Related Eye Disease Study 2 Report Number 13. Ophthalmology. 2018; 125(4): 512–521. [CrossRef] [PubMed]
McGuinness MB, Karahalios A, Finger RP, Guymer RH, Simpson JA. Age-related macular degeneration and mortality: a systematic review and meta-analysis. Ophthalmic Epidemiol. 2017; 24(3): 141–152. [CrossRef] [PubMed]
Grassmann F, Kiel C, Zimmermann ME, et al. Genetic pleiotropy between age-related macular degeneration and 16 complex diseases and traits. Genome Med. 2017; 9(1): 29. [CrossRef] [PubMed]
Timmers PR, Mounier N, Lall K, et al. Genomics of 1 million parent lifespans implicates novel pathways and common diseases and distinguishes survival chances. Elife. 2019; 8: e39856. [CrossRef] [PubMed]
Pilling LC, Kuo CL, Sicinski K, et al. Human longevity: 25 genetic loci associated in 389,166 UK Biobank participants. Aging (Albany NY). 2017; 9(12): 2504–2520. [CrossRef] [PubMed]
Figure 1.
 
Power analysis of 51 variants in individuals 90 years of age and older. Significantly associated variants are marked in red; rs61818925 (marked in blue) did not show significant associations despite more than 80% power.
Figure 1.
 
Power analysis of 51 variants in individuals 90 years of age and older. Significantly associated variants are marked in red; rs61818925 (marked in blue) did not show significant associations despite more than 80% power.
Figure 2.
 
Odds ratios in the different age groups of the six variants showing significant associations in individuals 90 years of age and older. (a) rs10922109 (CFH), (b) rs570618 (CFH), (c) rs3750846 (ARMS2/HTRA1), (d) rs116503776 (C2/CFB/SKIV2L), (e) rs5754227 (SYN3/TIMP3), (f) rs6565597 (NPLOC4/TSPAN10). *P-value interaction with the 90 years and older group < 0.05. (For details, see Table 3.)
Figure 2.
 
Odds ratios in the different age groups of the six variants showing significant associations in individuals 90 years of age and older. (a) rs10922109 (CFH), (b) rs570618 (CFH), (c) rs3750846 (ARMS2/HTRA1), (d) rs116503776 (C2/CFB/SKIV2L), (e) rs5754227 (SYN3/TIMP3), (f) rs6565597 (NPLOC4/TSPAN10). *P-value interaction with the 90 years and older group < 0.05. (For details, see Table 3.)
Figure 3.
 
ROC curves and AUCs of regression models of the 51 variants in the different age groups. Red, 60 to 69 years; orange, 70 to 79 years; green, 80 to 89 years; blue, 90 years and older. (For details, see Table 4.)
Figure 3.
 
ROC curves and AUCs of regression models of the 51 variants in the different age groups. Red, 60 to 69 years; orange, 70 to 79 years; green, 80 to 89 years; blue, 90 years and older. (For details, see Table 4.)
Table 1.
 
Demographics of the Study Cohort
Table 1.
 
Demographics of the Study Cohort
Table 2.
 
Significantly Associated Variants in the 90 Years and Older Group in All Age Groups
Table 2.
 
Significantly Associated Variants in the 90 Years and Older Group in All Age Groups
Table 3.
 
Interaction of Genetic Effects Between the 90 Years and Older Group and Other Age Groups and Interaction with Age as a Continuous Variable of Significantly Associated Variants
Table 3.
 
Interaction of Genetic Effects Between the 90 Years and Older Group and Other Age Groups and Interaction with Age as a Continuous Variable of Significantly Associated Variants
Table 4.
 
AUCs of Regression Models for the Different Age Groups Including All Genetic Variants
Table 4.
 
AUCs of Regression Models for the Different Age Groups Including All Genetic Variants
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×