Ophthalmic Genetics, 26:61–67, 2005Copyright c© Taylor and Francis Inc.ISSN: 1381-6810 (print) / 1744-5094 (online)DOI: 10.1080/13816810490967944

RESEARCH REPORT

A Discordant Sib-Pair Linkage Analysis of Age-RelatedMacular Degeneration

Susan L. SantangeloPsychiatric & Neurodevelopmental Genetics Unit, Center for Human Genetic Research, MassachusettsGeneral Hospital, Charlestown, MA, Department of Psychiatry, Harvard Medical School, Boston, MA,and Department of Epidemiology, Harvard School of Public Health, Boston, MA, USA

Chen-Hsing Yen, Stephen Haddad, and Jesen FagernessPsychiatric & Neurodevelopmental Genetics Unit, Center for Human Genetic Research,Massachusetts General Hospital, Charlestown, MA, USA

Christine HuangPsychiatric & Neurodevelopmental Genetics Unit, Center for Human Genetic Research,Massachusetts General Hospital, Charlestown, MA, and Department of Epidemiology,Harvard School of Public Health, Boston, MA, USA

Johanna M. SeddonOphthalmology/Epidemiology Unit, Massachusetts Eye and Ear Infirmary, Boston, MA,Department of Ophthalmology, Harvard Medical School, Boston, MA, and Department of Epidemiology,Harvard School of Public Health, Boston, MA, USA

Background: Age-related macular degeneration (AMD) is the leading cause of visual impairmentand blindness among older adults in the United States and throughout the developed world.Etiological research implicates both genetic and environmental components. Our prior genomescan in 511 affected sib-pairs and other relative pairs identified significant or suggestive linkagesignals on chromosomes 1, 2, 3, 6, 8, 10, 12, 16, and 22. Purpose: To search for genetic locifor AMD using the extremely discordant sib-pair (EDSP) method of linkage analysis, whichuntil now has never been applied to the study of AMD. Methods: The EDSP method is a morepowerful approach than standard methods which rely on relative pairs selected at random orpairs concordant for the phenotype. The EDSP approach has also been characterized as the onlydesign that is uniformly powerful in nearly all genetic situations. Thus, substantial reductionsin sample size can be achieved. Study population: The study sample for analysis included 110EDSPs from 40 families that comprise a subset of the 158 families studied in a prior genome-widescan using affected relative pairs. Results: Evidence for linkage was found on chromosomes 1q,2q, 6q, 19p, and 20q. The regions identified on chromosomes 1q and 2q were the same regionsidentified in our prior analysis, whereas the identified region on 6q was approximately 80 cMdistal to our previous signal. Discussion: Within this study population, we have narrowed thefocus to chromosomes 1q, 2q, 6q, 19p, and 20q in our search for AMD loci. However, given thefact that a gene was recently identified on chromosome 1q, future family- and population-basedanalyses should concentrate on testing for associations with candidate gene variants in the otheridentified chromosomal regions in searches for additional AMD loci.

Keywords AMD; genetics; genome scan

Accepted 4 April 2005.Address correspondence to Susan L. Santangelo, Sc.D., Psychiatric

& Neurodevelopmental Genetics Unit, Center for Human Genetic Re-search, Massachusetts General Hospital, 149 13th Street, Charlestown,MA 02129, USA. E-mail: ssantang@psych.mgh.harvard.edu

INTRODUCTIONAge-related macular degeneration (AMD) is the leading

cause of visual impairment and blindness among older adultsin the United States and throughout the developed world. This

61

Oph

thal

mic

Gen

et D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 10

/27/

14Fo

r pe

rson

al u

se o

nly.

62 S. L. SANTANGELO ET AL.

disease affects the central regions of the retina and choroid,thereby disrupting the central vision of many patients. Roughly30% of adults aged 75 or older have some signs of AMD, andabout 6-8% of the individuals in this age group have the ad-vanced stages of the disease which cause visual loss.1,2 Theprevalence of AMD is expected to rise to an even higher rate asthe population ages. AMD can profoundly impact the quality oflife and independence of the patient, as it may lead to the inabil-ity to read, write, and drive. Unfortunately, treatment optionsare limited and are restricted to a small subset of patients, oftenproviding only short-term relief from the disease. Therefore, in-vestigators must search for the underlying causes of AMD sothat it may one day be preventable.

Etiological research to date suggests that AMD is a complexdisease, caused by the actions and interactions of multiple genesand environmental factors. Studies have shown strong associa-tions between several environmental factors and AMD. The mostestablished demographic/environmental risk factors are increas-ing age3−5 and cigarette smoking.6−9 A higher risk of AMD hasalso been demonstrated in white populations than in Hispanicsand blacks.10,11 In addition, diet plays a role. Positive associa-tions have been found with both fat intake12−15 and obesity,16

and analyses have revealed protective effects from antioxidants,nuts, fish, and omega-3 fatty acids.12,14,15,17

Multiple studies have confirmed that AMD has a sub-stantial genetic component. These include familial aggrega-tion studies,18−23 twin studies,24−29 and segregation analyses.30

Linkage and association studies have attempted to locate thedisease-susceptibility loci for AMD. Linkage studies31−35 haveimplicated regions on nearly every chromosome in the humangenome, while most positive association findings36−44 have notbeen consistently replicated.

This report is a follow-up of our study published in the August2003 issue of the American Journal of Human Genetics45 inwhich we described the results of a genome-wide scan using158 multiplex families with 511 affected sib-pairs (ASPs), aswell as other relative pairs. This prior genome scan identifiedsignificant linkage of AMD to regions on chromosomes 2, 3, 6,and 8, as well as regions of potential interest on chromosomes1, 10, 12, 16, and 22. The present analysis was conducted usingthe extremely discordant sib-pair (EDSP) method within thesame study population in an attempt to increase the likelihoodof finding disease-susceptibility loci.

It has been suggested by many investigators that the EDSPmethod may be a more powerful approach than, for example,the affected sib-pair method for mapping quantitative trait loci(QTLs). The EDSP method has been demonstrated to requirefar fewer sib-pairs than methods that rely on pairs selectedat random (i.e. the regression method for quantitative traitsintroduced by Haseman and Elston46), or pairs concordant for thephenotype (the affected sib-pair method). Risch and Zhang47,48

showed that by using EDSPs, 10- to 40-fold reductions insample size could be achieved relative to randomly sampledsib-pairs.

The major advantage of the EDSP approach for screeningQTLs is that it does not depend on the actual values of the quanti-tative trait, but only on the fact that siblings have extreme values.Risch and Zhang47,48 showed that even when it is not possible toprecisely measure the entire range of the phenotype, a consider-able increase in power can be gained simply by differentiatingamong the individuals from the class of ‘unaffecteds.’ Therefore,the EDSP method may be particularly useful for AMD, since itis not known whether the phenotype can be measured as a trulycontinuously distributed trait and may be only quasi-continuous.

The EDSP approach has been characterized47,48 as the onlydesign that is uniformly powerful in all genetic situations (withthe exception of rare recessives) and superior to other sib-pairapproaches, whether or not alleles are rare, whether there aremultiple genes, and irrespective of allele frequencies, degree ofdominance, and heritability of the trait. Therefore, we decided toconduct this follow-up genome scan for AMD using the EDSPmethod.

METHODSThe methods employed in this study conformed to the tenets

of the Declaration of Helsinki and received approval from theappropriate institutional review boards. Informed consent state-ments were signed by the probands and family members par-ticipating in the study. Family recruitment methods and datacollection procedures were as described in our previous genomescan report.45 All of the families included in this report wereincluded in the previous genome scan utilizing affected relativepairs. AMD classification was based on fundus photography,and grades were assigned ranging from 1 (no disease) to 5 (ex-udative disease) according to the Clinical Age-Related Macu-lopathy Grading System (CARMS) as described previously45,49

and modified from the Age-Related Eye Disease Study.50

Because the EDSP method is most powerful when sibs aremaximally discordant, a sibship was designated as an EDSPwhen at least one sib was affected with more advanced AMD(grades 3B, 4, 5) and at least one sib was unaffected (grade 1 andolder than 60 years). Sibs with grades 2 and 3A were designatedas having an unknown phenotype. The sample included 40 fam-ilies with 110 EDSPs from the overall study population; this is asubset of the 158 families studied in our prior genome scan. Thenumber of sampled individuals was 138, with 50 males and 88 fe-males. Genotyping was performed by the National Heart, Lung,and Blood Institute (NHLBI) Mammalian Genotyping Serviceat the Marshfield Center for Medical Genetics (Marshfield, WI,USA). A total of 404 short-tandem repeat markers from We-ber Screening Set 10 were used, with an average density of 9centimorgans (cM).

The analytical approach outlined by Risch and Zhang47,48

is based on the hypothesis that allele sharing by EDSPs will besignificantly less than the null expectation of 50%. The Rischand Zhang outcome statistic, essentially a one-sided test of themean number of alleles shared identical-by-descent (IBD), isapproximately normally distributed in large samples, such as

Oph

thal

mic

Gen

et D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 10

/27/

14Fo

r pe

rson

al u

se o

nly.

DISCORDANT SIB-PAIR LINKAGE ANALYSIS OF AMD 63

FIG. 1. Chromosome 1 LOD curves. The curves depict the maximum multipoint LOD scores along chromosome 1 from theEDSP analysis. The boxes above the curves show the peaks from the prior ASP linkage scan. These peaks are depicted under thecorresponding markers and labeled GH for the Genehunter result and M/S for the Merlin/SimWalk2 result.

this one. We evaluated this test statistic in our study populationusing the ASPEX package,51 which performs non-parametricsib-pair linkage analysis. Very few parents were available inour sample; as a result, we used the Sib-Phase program ofASPEX, which is tailored for datasets with missing parents.The Sib-Phase program uses additional typed children andallele frequencies to reconstruct and phase the missing parentalgenotypes. This program implements a multipoint approach thatcalculates the probability of the observed marker data at a givenlocus assuming a particular IBD state. This calculation is basedon the known IBD states at the nearest informative flankingmarkers and the recombination distances to the locus position.The Sib-Phase option for analyzing extremely discordantsib-pairs was selected: this option calculates higher LODscores for decreased observed allele sharing. Allele sharingwas assessed using both an additive and a multiplicative model(the additive model does not allow for dominance variance,while the multiplicative model does). Results were quite similarfor the two models, as can be seen in Figures 1 and 2. In thetext, therefore, we highlight only the results from the additivemodel.

RESULTSEvidence for linkage was found on chromosomes 1q, 2q, 6q,

19p, and 20q, with multipoint LOD scores ranging from 1.2 to1.64 (nominal p-values ranged from 0.019 to 0.006). The re-

gions identified on chromosomes 1q and 2q in the 110 EDSPswere the same regions we identified in our prior genome scanof 511 ASPs.45 Our peak LOD score on chromosome 1q wasat 1q25-q31, a region that has received much attention recently,since several other studies have also seen evidence for linkagehere.32,33,41,52,53 In fact, very recently, a gene in this region wasidentified by three separate groups of investigators as being in-volved in AMD.54−56

The peak multipoint LOD score (mlod) in our EDSP anal-ysis of chromosome 1q was 1.64 (p = 0.006) at 190 cM (seeFig. 1). At this location, our prior ASP linkage scan producedan NPL Z-score of 1.95, p = 0.03, using the Genehunter pack-age, and a p-value of 0.03 using a combination of Merlin andSimWalk2.45 The locations of the peaks from our prior scan (butnot the magnitudes) are depicted at the top of Figure 1 under thecorresponding markers and labeled GH for the Genehunter resultand M/S for the Merlin/SimWalk2 result.

Our peak mlod on chromosome 2q was at 2q31.2-q32.3 (seeFig. 2). The mlod was 1.6 (p = 0.006) at 179 cM. At this location,our earlier linkage scan produced a Genehunter NPL Z score of2.03, p = 0.02, and a p-value of 0.01 using Merlin/SimWalk2(depicted in Fig. 2 under the corresponding markers and labeledas in Fig. 1). Iyengar et al.32 and Abecasis et al.31 found signalson chromosome 2q about 75 cM distal to the linkage peaks thatwe found, and Weeks et al.35 found a linkage signal about 40cM proximal to our peaks.

Oph

thal

mic

Gen

et D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 10

/27/

14Fo

r pe

rson

al u

se o

nly.

64 S. L. SANTANGELO ET AL.

FIG. 2. Chromosome 2 LOD curves. The curves depict the maximum multipoint LOD scores along chromosome 2 from theEDSP analysis. The boxes above the curves show the peaks from the prior ASP linkage scan. These peaks are depicted under thecorresponding markers and labeled GH for the Genehunter result and M/S for the Merlin/SimWalk2 result.

In our prior genome scan, we also observed significant link-age on chromosome 2p at 38 cM (p = 0.00001) with Mer-lin/SimWalk2. Weeks et al.,53 Abecasis et al.,31 and Iyengaret al.32 observed signals on chromosome 2p in slightly differentregions (at 95 cM, 10 cM, and 60 cM, respectively).

The region identified with our EDSPs on chromosome 6q25.3was approximately 80 cM distal to the significant signal weobserved in our 2003 linkage scan. In the present analysis, ourpeak mlod was 1.3 (p = 0.014) at 166.4 cM. In our prior linkagescan, we observed significant signals (p = 0.00001) at 81 cMand 89 cM using Merlin/SimWalk2. Schick et al.34 reportedmultipoint signals in both regions using a quantitative measureof AMD: at approximately 166 cM with p-values of 0.02-0.04and at 84-88 cM with p = 0.04.

Although signals on chromosomes 19 and 20 were observedin our prior linkage scan, they were not as strong as others wefound in that analysis. In the present analysis, we observed apeak LOD score of 1.2 (p = 0.019) at 43 cM on chromosome19p13.1. In our prior analysis, the peak Genehunter NPL scorewas 0.94 (p = 0.17) and was located about 40 cM away on19q. Abecasis et al.31 reported a peak meeting the criterion forsuggestive linkage57 on chromosome 19p at 11 cM.

On chromosome 20q13.1-q13.2, the peak mlod was 1.5 (p =0.009) at 59 cM. The peak Genehunter NPL score we observed inour prior genome scan was 1.20 (p = 0.11). Using a quantitativemeasure of AMD, Iyengar et al.32 reported a signal with anempirical p value of 0.0052, nearly 30 cM distal to our peak, atabout 90 cM.

DISCUSSIONThis linkage analysis for AMD, using the EDSP approach,

identified new regions of interest within our study population onchromosomes 6q, 19p, and 20q. In addition, we re-identified ourprior linkage peaks on 1q25-q31 and 2q31.2-q32.3. The EDSPmethod is more powerful than standard linkage methods whichrely on relative pairs selected at random or pairs concordant forthe phenotype. Moreover, the power of this design is uniformin almost every genetic situation. The EDSP approach does notdepend on the actual values of the trait, but only on the fact thatsiblings have extreme values. Therefore, it is a useful methodfor diseases such as AMD, for which it may not be possible tomeasure the phenotype as a truly continuously distributed trait.

Other research teams have reported linkage signals on chro-mosomes 1q, 2q, 6q, 19p, and 20q as well, and the locationsof our 1q and 6q signals were the same as reported elsewhere.Schick et al.34 found linkage peaks on 6q at 84-88 cM and 166cM, close to the signals we found at 81 cM and 89 cM in ourprior study and at 166.4 cM in the present analysis. Most no-tably, at least five other studies32,33,41,52,53 have also implicatedthe 1q25-q31 chromosomal region.

Several candidate genes for AMD are located in the chromo-somal regions which showed evidence for linkage in this study.The hemicentin gene (FIBL-6) [MIM 608548], first suggestedby Klein et al.52 and Schultz et al.41 as the ARMD1 locus, spansapproximately 600 kilobases (kb) in the 1q24-q31 region andencodes a large extracellular member of the immunoglobulinsuperfamily. Mutations in this gene lead to tissue frailty and cell

Oph

thal

mic

Gen

et D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 10

/27/

14Fo

r pe

rson

al u

se o

nly.

DISCORDANT SIB-PAIR LINKAGE ANALYSIS OF AMD 65

migration defects. Analyses by Iyengar et al.32 found evidencefor linkage to hemicentin variants, although no mutations wereobserved in exon 104 upon examination of the affected individu-als in the linked families. Studies by Abecasis et al.31 and Stoneet al.42 failed to confirm the involvement of the hemicentin genein AMD prevalence.

Several additional candidate genes are located in the 1q24-q31 region. For example, the phosducin gene (PDC) [MIM171490], a candidate for retinitis pigmentosa, spans 18 kb andencodes a phosphoprotein located in retinal rod cells. This genemay participate in the regulation of visual phototransduction orthe integration of photoreceptor metabolism. The myocilin gene(MYOC, also known as TIGR) [MIM 601652], which spans2.3 kb, causes increased intraocular pressure and juvenile open-angle glaucoma.

Recently, three separate research groups identified a com-mon coding variant in the complement factor H gene (CFH)on 1q31 which appears to exert a strong influence on the riskof developing AMD.54−56 These studies suggested that close toone half of all AMD cases in older adults may be attributableto this gene. CFH is a component of the innate immune systemwhich helps to regulate the body’s inflammatory response bypreventing uncontrolled complement activation.58 The causativevariant, Try402His, is located within binding sites for heparin59

and C-reactive protein.60 Binding to either of these factors in-creases the affinity of CFH for complement protein C3b,61,62

which enhances CFH’s ability to inhibit complement’s effect.The inflammatory response has repeatedly been implicated inthe pathophysiology of AMD,63−66 and the discovery of thisrisk allele confirms the role of inflammation in the developmentof AMD.

Candidate genes in the chromosome 2q31.2-q32.3 regioninclude CERKL [MIM 608381] and SLC40A (also knownas SLC11A3) [MIM 604653]. CERKL is a ceramide ki-nase gene that causes autosomal recessive retinitis pigmentosa(RP26); SLC40A is an iron-regulated transporter gene causinghemochromatosis.

Superoxide dismutase 2 (SOD2) [MIM 147460] is a candi-date gene in the chromosome 6q25.3 region. SOD2 encodes anintramitochondrial free-radical scavenging enzyme that is thefirst line of defense against superoxide produced as a byproductof oxidative phosphorylation. This gene product destroys radi-cals which are normally produced within cells and are toxic to bi-ological systems. A hospital-based case-control study40 lookedfor SOD2 polymorphisms in 102 Japanese patients with exuda-tive AMD and 200 controls. They found a significant associationbetween case status and the valine/alanine polymorphism at thetargeting sequence of the SOD2 enzyme.

RAB8 (also known as Oncogene MEL) [MIM 165040] inthe chromosome 19p13.1 region is another AMD candidategene. This member of the RAS superfamily (a family of smallGTP/GDP-binding proteins) may play a role in the transportof proteins from the endoplasmic reticulum to the Golgi andthe plasma membrane. The chromosome 20q13.1-13.2 region is

the location of matrilin 4 (MATN4) [MIM 603897], a candidategene containing 10 exons and spanning approximately 12 kb, forwhich three alternatively spliced variants have been described.This gene encodes a member of the von Willebrand factor Adomain-containing protein family. Proteins of this family arethought to be involved in the formation of filamentous networksin the extracellular matrices of various tissues.

In future studies, family- and population-based analysesshould be conducted to test for AMD associations with geneticvariants located within the candidate genes described above, aswell as other candidates in the regions which showed evidencefor linkage in this study. In addition, a more genomic approachto the investigation of these regions, in which dense SNP typingis carried out across the chromosomal regions without regard tothe identification of candidate genes, may prove fruitful.

Future studies should also test for gene-gene and gene-environment interactions whenever possible. Because AMD isprobably a complex disease, individual gene effects might onlybe detected within certain subgroups of patients with specific en-vironmental exposures. Potential environmental factors to con-sider include cigarette smoking, body mass index (BMI), anddietary variables such as fat, antioxidant, and omega-3 fatty acidconsumption. In addition, a given gene variant might only resultin a detectable phenotype when acting in combination with addi-tional susceptibility alleles either additively or multiplicatively.Investigations of any candidate genes should explore possibleinteractions with the Try402His variant of the CFH gene.

The search for AMD genes remains difficult due to the mul-tiple genetic and environmental effects and interactions thatare likely to be involved. Other characteristics of the diseasealso make this research challenging. AMD onset occurs latein life, and therefore it is often the case that only one genera-tion in the appropriate age range is available for study. Parentsof probands are often deceased and the offspring are often tooyoung to be affected.22 In addition, AMD is a phenotypicallyheterogeneous disease. If it is the case that different sets ofgenes underlie different phenotypic subtypes, then the powerof genetic studies may be diluted if all types of AMD patientsare included together in one analysis. However, if analyses areinstead restricted to one specific subtype of AMD, power maybe compromised by inadequate sample sizes.

Despite all of these difficulties, researchers remain deter-mined to find the disease susceptibility genes, and the recentdiscovery of a causal variant within the CFH gene is an encour-aging development. It is likely that several more AMD suscep-tibility genes remain to be discovered. The results of the presentstudy provide additional information towards this end and callfor further investigations, particularly on chromosomes 2q, 6q,19p, and 20q.

ACKNOWLEDGEMENTSThis work was supported by grants from the National Eye In-

stitute U10-11309, RO1EY011309 (Dr. Seddon), The NationalInstitutes of Health, Bethesda, MD, The Foundation Fighting

Oph

thal

mic

Gen

et D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 10

/27/

14Fo

r pe

rson

al u

se o

nly.

66 S. L. SANTANGELO ET AL.

Blindness, Owings Mills, MD, Research to Prevent Blindness,New York, NY, and the Massachusetts Lions Research Fund, Inc.We wish to gratefully acknowledge all of the patients, their fami-lies, and the numerous ophthalmologists throughout the countrywho participated in this study.

REFERENCES1. National Advisory Eye Council. Vision research—A national plan:

1999-2003. A Report of the National Advisory Eye Council.2001;1, NIH Publ. No. 98-4120.

2. Seddon JM, Chen C. Epidemiology of age-related macular degen-eration. In: Ryan SJ, editor. Medical retina. 4th ed. St. Louis: C.V.Mosby, 2005; Vol. 2.

3. Eye Diseases Prevalence Research Group. Prevalence of age-related macular degeneration in the United States. Arch Ophthal-mol. 2004;122:564-572.

4. Klein R, Klein BEK, Linton KLP. Prevalence of age-related mac-ulopathy. Ophthalmology. 1992;99:933.

5. Klein R, Klein BEK, Jensen SC, et al. The five-year inci-dence and progression of age-related maculopathy. Ophthalmol-ogy. 1997;104:7-21.

6. Christen WG, Glyim RJ, Manson JE, et al. A prospective study ofcigarette smoking and risk of age-related macular degeneration inmen. J Am Med Assoc. 1996;276:1147-1151.

7. Seddon JM, Hankinson S, Speizer F, et al. A prospective study ofsmoking and age-related macular degeneration. Am J Epidemiol.1995;241:136.

8. Seddon JM, Willett WC, Speizer F, et al. A prospective study ofcigarette smoking and age-related macular degeneration in women.J Am Med Assoc. 1996;276:1141-1146.

9. Tomany SC, Wang JJ, Leeuwen RV, et al. Risk factors for incidentage-related macular degeneration: pooled findings from 3 conti-nents. Ophthalmology. 2004;111:1280-1287.

10. Cruickshanks KJ, Hamman RF, Klein R, et al. The prevalence ofage-related maculopathy by geographic region and ethnicity. ArchOphthalmol. 1997;115:242-250.

11. Sommer A, Tielsch JM, Katz J, et al. Racial differences in thecause-specific prevalence of blindness in East Baltimore. N Engl JMed. 1991;325:1412-1417.

12. Cho E, Hung S, Willett WC, et al. Prospective study of dietary fatand the risk of age-related macular degeneration. Am J Clin Nutr.2001;73:209-218.

13. Mares-Perlman JA, Brady WE, Klein R. Dietary fat and age-relatedmaculopathy. Arch Ophthalmol. 1995;113:743-748.

14. Seddon JM, Rosner B, Sperduto RD, et al. Dietary fat and riskfor advanced age-related macular degeneration. Arch Ophthalmol.2001;119:1191-1199.

15. Seddon JM, Cote J, Rosner B. Progression of age-related maculardegeneration: association with dietary fat, transunsaturated fat, nutsand fish intake. Arch Ophthalmol. 2003;121:1728-1737.

16. Seddon JM, Cote J, Davis N, et al. Progression of age-related mac-ular degeneration: association with body mass index, waist circum-ference, and waste-hip ratio. Arch Ophthalmol. 2003;121:785-792.

17. Seddon JM, Ajani UA, Sperduto RD, et al. Dietary carotenoids,vitamins A, C, and E, and advanced age-related macular degener-ation. J Am Med Assoc. 1994;272:1413-1420.

18. DeJong PT, Klaver CC, Wolfs RC, et al. Familial aggregation ofage-related maculopathy. Am J Ophthalmol. 1997;124:862-863.

19. Klaver CC, Wolfs RC, Assink JJ, et al. Genetic risk of age-relatedmaculopathy: population-based familial aggregation study. ArchOphthalmol. 1998;116:1646-1651.

20. Klein BE, Klein R, Lee KE, et al. Risk of incident age-related eyediseases in people with an affected sibling: the Beaver Dam EyeStudy. Am J Epidemiol. 2001;154:207-211.

21. Piguet B, Wells JA, Palmvang IB, et al. Age-related Bruch’s mem-brane change: a clinical study of the relative role of heredity andenvironment. Br J Ophthalmol. 1993;77:400-403.

22. Seddon JM, Ajani UA, Mitchell BD. Familial aggregation of age-related maculopathy. Am J Ophthalmol. 1997;123:199-206.

23. Silvestri G, Johnston PB, Hughes AE. Is genetic predispositionan important risk factor in age-related macular degeneration? Eye.1994;8:564-568.

24. Grizzard SW, Arnett D, Haag SL. Twin study of age-related mac-ular degeneration. Ophthalmic Epidemiol. 2003;10:315-322.

25. Hammond CJ, Webster AR, Snieder H, et al. Genetic influenceon early age-related maculopathy: a twin study. Ophthalmology.2002;109:730-736.

26. Klein ML, Mauldin WM, Stoumbos VD. Heredity and age-relatedmacular degeneration: observations in monozygotic twins. ArchOphthalmol. 1994;112:932-937.

27. Meyers SM, Greene T, Gutman FA. A twin study of age-relatedmacular degeneration. Am J Ophthalmol. 1995;120:757-766.

28. Meyers SM, Zachary AA. Monozygotic twins with age-relatedmacular degeneration. Arch Ophthalmol. 1988;106:651-653.

29. Seddon JM, Cote J, Page WF, et al. The U.S. age-related maculardegeneration twin study: relative role of genetic and environmentalinfluences. Arch Ophthalmol. 2005;123:321-327.

30. Heiba IM, Elston RIC, Klein BEK, et al. Sibling correlations andsegregation analysis of age-related maculopathy: The Beaver DamEye Study. Genet Epidemiol. 1994;11:51-67.

31. Abecasis GR, Yashar BM, Zhao Y, et al. Age-related macular de-generation: a high-resolution genome scan for susceptibility lociin a population enriched for late-stage disease. Am J Hum Genet.2004;74:482-494.

32. Iyengar SK, Song D, Klein BE, et al. Dissection of genomewide-scan data in extended families reveals a major locus and oligogenicsusceptibility for age-related macular degeneration. Am J HumGenet. 2004;74:20-39.

33. Majewski J, Schultz DW, Weleber RG, et al. Age-related macu-lar degeneration: a genome scan in extended families. Am J HumGenet. 2003;73:540-550.

34. Schick JH, Iyengar SK, Klein BE, et al. A whole-genome screenof a quantitative trait of age-related maculopathy in sibships fromthe Beaver Dam Eye Study. Am J Hum Genet. 2003;72:1412-1424.

35. Weeks DE, Conley YP, Tsai HJ, et al. Age-related maculopathy:a genomewide scan with continued evidence of susceptibility lociwithin the 1q31, 10q26, and 17q25 regions. Am J Hum Genet.2004;75(2):174-189.

36. Allikmets R. Further evidence for an association of ABCR al-leles with age-related macular degeneration. Am J Hum Genet.2000;67:487-491.

37. Baird PN, Guida E, Chu DT, et al. The epsilon2 and epsilon4 allelesof the apolipoprotein gene are associated with age-related maculardegeneration. Invest Ophthalmol Vis Sci. 2004;45:1311-1315.

38. Hamdi HK, Reznik J, Castellon R, et al. Alu DNA polymorphismin ACE gene is protective for age-related macular degeneration.Biochem Biophys Res Commun. 2002;295:668-672.

Oph

thal

mic

Gen

et D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 10

/27/

14Fo

r pe

rson

al u

se o

nly.

DISCORDANT SIB-PAIR LINKAGE ANALYSIS OF AMD 67

39. Ikeda T, Obayashi H, Hasegawa G, et al. Paraoxonase gene poly-morphisms and plasma oxidized low-density lipoprotein level aspossible risk factors for exudative age-related macular degenera-tion. Am J Ophthalmol. 2001;132:191-195.

40. Kimura K, Isashiki Y, Sonoda S, et al. Genetic association of man-ganese superoxide dismutase with exudative age-related maculardegeneration. Am J Ophthalmol. 2000;130:769-773.

41. Schultz DW, Klein ML, Humpert AJ, et al. Analysis of the ARMD1locus: evidence that a mutation in HEMICENTIN-1 is associatedwith age-related macular degeneration in a large family. Hum MolGenet. 2003;12:3315-3323.

42. Stone EM, Braun TA, Russell SR, et al. Missense variations inthe Fibulin 5 gene and age-related macular degeneration. N Engl JMed. 2004;351:346-353.

43. Zareparsi S, Reddick AC, Branham KE, et al. Association ofapolipoprotein E alleles with susceptibility to age-related maculardegeneration in a large cohort from a single center. Invest Ophthal-mol Vis Sci. 2004;45:1306-1310.

44. Zurdel J, Finckh U, Menzer G, et al. CST3 genotype associatedwith exudative age related macular degeneration. Br J Ophthalmol.2002;86:214-219.

45. Seddon JM, Santangelo SL, Book K, et al. A genomewide scan forage-related macular degeneration provides evidence for linkage toseveral chromosomal regions. Am J Hum Genet. 2003;73:780-790.

46. Haseman JK, Elston RC. The investigation of linkage between aquantitative trait and a marker locus. Behav Genet. 1972;2:3-19.

47. Risch N, Zhang H. Extreme discordant sib pairs for mapping quan-titative trait loci in humans. Science. 1995;268:1584-1589.

48. Risch NJ, Zhang H. Mapping quantitative trait loci with extremediscordant sib pairs: sampling considerations. Am J Hum Genet.1996;58:836-843.

49. Seddon JM, Samuelson L, Page WF, et al. Twin study of maculardegeneration: methodology and application to genetic epidemio-logic studies. Invest Ophthalmol Vis Sci. 1996;38:676.

50. Age-Related Eye Disease Study Research Group. A randomized,placebo-controlled, clinical trial of high-dose supplementationwith vitamins C and E, beta carotene, and zinc for age-related macu-lar degeneration and vision loss. Arch Ophthalmol. 2001;119:1417-1436.

51. Hinds DA, Risch N. The ASPEX package: affected sib-pair exclu-sion mapping. http://aspex.sourceforge.net/

52. Klein ML, Schultz DW, Edwards A, et al. Age-related maculardegeneration: clinical features in a large family and linkage tochromosome 1q. Arch Ophthalmol. 1998;116:1082-1088.

53. Weeks DE, Conley YP, Tsai HJ, et al. Age-related maculopa-thy: an expanded genome-wide scan with evidence of suscepti-

bility loci within the 1q31 and 17q25 regions. Am J Ophthalmol.2001;132:682-692.

54. Edwards AO, Ritter R, Abel KJ, et al. Complement factor Hpolymorphism and age-related macular degeneration. Science.2005;308:421-424.

55. Haines JL, Hauser MA, Schmidt S, et al. Complement factor Hvariant increases the risk of age-related macular degeneration. Sci-ence 2005;308:419-421.

56. Klein RJ, Zeiss C, Chew EY, et al. Complement factor Hpolymorphism in age-related macular degeneration. Science.2005;308:385-389.

57. Lander ES, Kruglyak L. Genetic dissection of complex traits:guidelines for interpreting and reporting linkage results. Nat Genet.1995;11:241-247.

58. Rodriguez DC, Esparza-Gordillo J, Goicoechea DJ, et al. The hu-man complement factor H: functional roles, genetic variations anddisease associations. Mol Immunol. 2004;41:355-367.

59. Blackmore TK, Fischetti VA, Sadlon TA, et al. M protein of thegroup A Streptococcus binds to the seventh short consensus re-peat of human complement factor H. Infect. Immun. 1998;66:1427-1431.

60. Giannakis E, Jokiranta TS, Male DA, et al. A common site withinfactor H SCR 7 responsible for binding heparin, C-reactive proteinand streptococcal M protein. Eur J Immunol. 2003;33:962-969.

61. Fearon DT. Regulation by membrane sialic acid of beta1H-dependent decay-dissociation of amplification C3 convertase ofthe alternative complement pathway. Proc Natl Acad Sci USA.1978;75:1971-1975.

62. Mold C, Kingzette M, Gewurz H. C-reactive protein inhibits pneu-mococcal activation of the alternative pathway by increasing theinteraction between factor H and C3b. J Immunol. 1984;133:882-885.

63. Gurne DH, Tso MO, Edward DP, et al. Antiretinal antibodies inserum of patients with age-related macular degeneration. Ophthal-mology. 1991;98:602-607.

64. Killingsworth MC, Sarks JP, Sarks SH. Macrophages relatedto Bruch’s membrane in age-related macular degeneration. Eye.1990;4(Pt 4):613-621.

65. Mullins RF, Russell SR, Anderson DH, et al. Drusen associatedwith aging and age-related macular degeneration contain proteinscommon to extracellular deposits associated with atherosclerosis,elastosis, amyloidosis, and dense deposit disease. FASEBJ. 2000;14:835-846.

66. Seddon JM, Gensler G, Milton RC, et al. Association between C-reactive protein and age-related macular degeneration. J Am MedAssoc. 2004;291:704-710.

Oph

thal

mic

Gen

et D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 10

/27/

14Fo

r pe

rson

al u

se o

nly.