Hyperglycemia and long duration of diabetes are widely recognized risk factors for diabetic retinopathy, but inherited susceptibility may also play a role because retinopathy aggregates in families. A genome-wide linkage analysis was conducted in 211 sibships in which ≥2 siblings had diabetes and retinal photographs were available from a longitudinal study. These sibships were a subset of 322 sibships who had participated in a previous linkage study of diabetes and related traits; they comprised 607 diabetic individuals in 725 sibpairs. Retinal photographs were graded for presence and severity of diabetic retinopathy according to a modification of the Airlie House classification system. The grade for the worse eye was adjusted for age, sex, and diabetes duration and analyzed as a quantitative trait. Heritability of diabetic retinopathy in this group was 18% (95% CI 2–36). A genome-wide linkage analysis using variance components modeling found evidence of linkage on chromosome 1p. Using single-point analysis, the peak logarithm of odds (LOD) was 3.1 for marker D1S3669 (34.2 cM), whereas with multipoint analysis the peak LOD was 2.58 at 35 cM. No other areas of suggestive linkage were found. We propose that an area on chromosome 1 may harbor a gene or genes conferring susceptibility to diabetic retinopathy.

Diabetic retinopathy, a common and serious complication of diabetes (1), is one of the leading causes of blindness (2). Many risk factors for diabetic retinopathy have been identified, including poor glycemic control, hypertension (3), albuminuria, and diabetes duration (47). Retinopathy is also familial. A study of identical twins found a concordance for diabetic retinopathy in 21 of 31 pairs with type 1 diabetes (68%) and in 35 of 37 pairs with type 2 diabetes (95%) (8). In the Diabetes Control and Complications Trial cohort, the odds ratio (OR) for severe retinopathy when a relative had retinopathy was 3.1 and the pairwise correlation for retinopathy grade was 0.187 for family members (9). In a cohort with type 2 diabetes, the OR for any grade of retinopathy for siblings of subjects with retinopathy was 4.3 (10). More recently, in a study of retinopathy among Mexican-American families, no significant difference was found for the presence of any retinopathy according to the presence of retinopathy in a sibling, but there was a significant OR for the presence of severe retinopathy (11).

To date, however, no specific genetic loci influencing susceptibility to diabetic retinopathy have been identified. Candidate gene studies have, for the most part, reported conflicting results (12), and there have been few genome-wide studies. Our group previously conducted genome-wide linkage analysis for retinopathy in diabetic Pima Indians who had participated in a linkage study designed to identify loci for diabetes and obesity (13,14). The results of this analysis showed only weak evidence for linkage on chromosomes 3 and 9 (13). However, the analysis was conducted by assessment of allele sharing among sibling pairs affected by retinopathy, and quantitative information on retinopathy severity was not used. Furthermore, at the time this preliminary analysis was conducted, retinal photographs were only available on a relatively small subset of participants. In the present study, genome-wide linkage analyses were conducted for participants with updated retinal photographs of quantitative retinopathy severity scores.

Data for the present analysis come from an ongoing longitudinal study among the Pima Indians of the Gila River Indian Community in central Arizona. Community residents are invited to undergo examinations at intervals of 2 years from the age of 5 years onward. These examinations include measures of BMI, blood pressure, plasma glucose, and A1C. Albumin-to-creatinine ratio was measured in a spot urine specimen. Diabetes was diagnosed on the basis of a 75-g oral glucose tolerance test according to World Health Organization criteria (15) or a previously documented clinical diagnosis. All participants ≥15 years old have retinal photographs taken through dilated pupils (two standard 45° fields for each eye) using a Canon CR4–45NM fundus camera, centered at the optic disk and at the macula. When subjects attended multiple examinations over the course of the study, the examination closest to a person having a 10-year duration of diabetes was selected.

A genome-wide linkage scan was previously performed on 1,311 individuals selected from the longitudinal study in 332 nuclear families potentially informative for linkage with diabetes and related traits (14). For the current analyses, siblings from the genome-wide linkage group with a diagnosis of diabetes and graded retinal photographs (n = 607) were selected. The retinal photographs were graded in a standardized manner without knowledge of clinical details according to a modification of the Airlie House classification system and a modification of the Early Treatment of Diabetic Retinopathy Study severity system of diabetic retinopathy, which assigns a numeric score to each eye (Table 1) (16). A mean retinopathy score was calculated for each subject [(score in right eye + score in left eye)/2]; when only one eye was gradable, the score for that eye was taken as the mean score. In addition to the mean score, the score for the worse eye was also analyzed. Although these traits were highly correlated (r = 0.98), we analyzed both traits because the worse eye score may represent an individual's susceptibility to retinopathy, whereas the mean score may represent the precise estimate of the current level of retinopathy.

The traits (mean retinal score and worse eye score) for each individual were adjusted for age, sex, and diabetes duration using linear regression. To ensure that the assumption of multivariate normality was met, which is required for the genetic analyses, the residuals from the regression analyses were then normalized with an inverse Gaussian transformation before analysis. Variance components methods were used to calculate heritability (i.e., the proportion of variance potentially due to the additive effects of genetic factors) (17,18). CIs for the heritability were calculated by a systematic exploration of the likelihood surface (19). To further assess the extent of familial aggregation according to severity of retinopathy, the prevalence of retinopathy or varying severity was assessed in siblings of an “index” individual according to the presence or absence of retinopathy in the index sibling (the oldest [i.e., first born] diabetic sibling in each sibship was taken as the index). Thus, ORs of retinopathy in siblings were calculated using logistic regression models for varying levels of retinopathy using the retinopathy level (worse eye grade ≥ 21, “any retinopathy”; worse eye grade ≥ 31, “moderate nonproliferative diabetic retinopathy or more”; or worse eye grade ≥ 60, “proliferative diabetic retinopathy”) for the older sibling as the risk factor for a corresponding degree of retinopathy in the younger siblings. These models were adjusted for age, sex, and diabetes duration of the younger sibling and were fit by generalized estimating equations to account for correlation among siblings from the same family.

Genome-wide linkage analysis.

A total of 516 autosomal microsatellite markers were previously typed in a genome-wide linkage scan (14), of which 503 were typed in the laboratory of J. Weber at Marshfield Medical Research Foundation (20,21), and 13 were typed at Glaxo-Wellcome. For duplicate samples, the median rate of agreement was 97%, and no marker had an agreement rate <90%. Markers were checked for Mendelian errors, and genetic distances were determined (22). For each marker, the method of Curtis and Sham (23) was used to estimate identity by descent (IBD) for each sibling pair. For multipoint analyses, the method of Fulker and Cherny (24) was used to calculate multipoint IBD estimates in intervals of 1 cM across each chromosome. Variance components methods were used to assess linkage across the whole chromosome (17,18). These methods involved fitting a linear mixed model to estimate the trait mean (μ) and three components of variance. Variance was partitioned into 1) an additive “monogenic” component linked to the region of interest (σM2), 2) a “polygenic” component that incorporated overall familial effects (σG2), and 3) an “environmental” component that incorporated effects unique to the individual (σE2). Assuming no recombination between the trait and the marker loci, the phenotypic variance-covariance matrix (Ω) for individuals in a pedigree can be represented by the formula: Ω = ΦσM2 + ΠσG2 + IσE2, where Φ is a matrix of the expected proportion of alleles shared IBD, Π is a matrix of the estimated proportion of alleles actually shared IBD for a particular marker, and I is an identity matrix. The likelihood ratio test was used to assess the hypothesis of no linkage at a given location (σM2 = 0) against the alternative of linkage (σM2 > 0). For presentation purposes, this test is shown as a logarithm of odds (LOD) score. All statistical analyses were performed using SAS (SAS Institute, Cary, NC).

From 211 families, 607 subjects were identified, giving a total of 725 sibpairs. The majority of families (n = 110) contributed one sibpair to the analysis, but up to nine siblings from a single family were included (Table 2). Excluding the larger families (≥7 siblings) from the analyses did not appreciably change the results, so the findings for the whole set are presented. Subject characteristics are shown in Table 3, and the distribution of retinopathy grades are shown in Table 4. Retinopathy grade was strongly correlated with diabetes duration (r = 0.49 for the worse eye score, r = 0.50 for the mean eye score) and with albumin-to-creatine ratio (0.40 for the worse eye score, r = 0.41 for the mean eye score), A1C (r = 0.23 for both worse eye score and mean eye score), diastolic blood pressure (r = 0.22 for the worse eye score, r = 0.23 for the mean eye score), systolic blood pressure (r = 0.17 for the worse eye score, r = 0.18 for the mean eye score), and age (r = 0.13 for the worse eye score, r = 0.14 for the mean eye score). BMI was negatively correlated with retinopathy grade (r = −0.30 for the worse eye score, r = −0.31 for the mean eye score). All correlation coefficients shown are Spearman correlations and P < 0.01 for each.

The heritability of retinopathy in this sample for the worse eye score was 18% (95% CI 2–36), and for the mean eye score, the heritability was 16% (1–33). For each family, the retinopathy grade in the first-born sibling (n = 211, 123 with worse eye grade of 10, 15 with worse eye grade of 21, 61 with worse eye grade of 31–41, and 12 with worse eye grade of ≥51) was used as a risk factor for younger siblings developing retinopathy. When the first-born sibling had any retinopathy (worse eye score ≥21), the risk for a younger sibling to have any retinopathy was modestly elevated but was not statistically significant (OR 1.10 [95% CI 0.61–1.97], P = 0.75 adjusted for age, sex, and diabetes duration of the younger sibling and for diabetes duration of the index sibling). The OR was similar for moderate nonproliferative retinopathy or more severe retinopathy (worse eye score ≥31) (1.11 [0.61–2], P = 0.73), but there was a greater risk for proliferative retinopathy (worse eye score ≥60) (2.46 [0.85–7.15], P = 0.10).

All points where the LOD score exceeded 1 (locus-specific P < 0.03) are shown in Table 5 for the single-point linkage analyses. The strongest evidence for linkage was on chromosome 1 with marker D1S3699 for both the worse eye score and the mean eye score. Results of single-point and multipoint analyses for all chromosomes are shown in Fig. 1 for the worse eye phenotype. Findings were similar for the mean eye score phenotype. In the multipoint analysis, the peak LOD score was on chromosome 1 at 35 cM for both the worse eye and mean eye score phenotypes. The peak multipoint LOD was 2.58 for the worse eye and 1.94 for the mean eye score. No other chromosomal regions showed evidence of linkage for retinopathy with LOD >1.

Both twin studies (8) and familial aggregation studies (9,10) show evidence for genetic susceptibility to diabetic retinopathy. The heritability estimate in the present study shows a modest level of familial aggregation in this population that was significantly different from 0, suggesting a genetic role in the development and/or progression of retinopathy.

The majority of studies on the genetics of diabetic retinopathy have focused on a candidate gene approach. The results of such studies have thus far been somewhat inconsistent, which may reflect, in part, the differing definitions for the phenotype used and the different populations included in the studies. The present genome-wide linkage analysis used variance components modeling to identify the areas of linkage for diabetic retinopathy. This method uses quantitative information about the degree of retinopathy from all siblings. The results showed suggestive evidence for linkage to chromosome 1p36. Genes that have shown evidence of association with any degree of retinopathy among candidate gene studies include inducible nitric oxide synthase gene (25), intracellular adhesion molecule-1 gene (26), and aldose reductase gene (27). The transforming growth factor-β gene has shown evidence of association only with proliferative retinopathy (28). A previous candidate gene study in the Pimas reported a modest association between retinopathy (as assessed by fundoscopy) and a polymorphism in the promoter region of the plasminogen activator inhibitor 1 gene on chromosome 7 (29). None of these genes, however, map to the area of linkage identified in the current study.

In linkage studies, it is customary to require particularly stringent levels of statistical significance. A LOD score >3 (locus-specific P < 0.0001) is conventionally considered statistically significant linkage, whereas a LOD score >2 (locus-specific P < 0.0012) is considered suggestive but not definitive evidence for linkage (30). A previous affected sibpair analysis in a subset of the individuals in the present study showed tentative linkage on chromosomes 3 and 9 (13); however, these regions did not show linkage in the current study. The previous study included a fairly small number of affected sibling pairs (n = 103, comprising 156 individuals, all of whom were included in the present study), and the LOD scores did not achieve levels that were suggestive of linkage by conventional criteria. The present analysis is presumably more powerful because it includes data on a larger number (n = 607) of individuals and because the LOD score on chromosome 1p is considered suggestive of linkage. The present analyses also incorporated information about severity of retinopathy based on fundus photographs, the more severe levels of which tend to be more familial than simply the presence of any retinopathy, which was the phenotype analyzed in the previous study. This is in keeping with results from Mexican-American families in Texas, where higher ORs were found for more severe retinopathy within siblings than for milder disease (11). It is possible that the differences between the two analyses are explained by interactions with other factors (such as duration of diabetes) or, given the relatively small sample size, by chance. Ultimately, replication studies in other populations will be required to more firmly establish susceptibility loci for diabetic retinopathy.

For variance components modeling, a normal distribution of the phenotypic measure is assumed, although the method is robust to moderate violations of normality (17,18). Although retinopathy was treated as a continuous variable in this study, the distribution of retinopathy scores was highly skewed, because the majority of subjects had no retinopathy at their examination. Because of the severely skewed distribution of retinopathy scores, data were normalized (after adjustment for age, sex, and duration of diabetes) before linkage analysis. The normalization procedure ensures that the assumptions of normality are met, although it may compromise the power somewhat compared with analysis without normalization. The procedure allows for adjustment for important potential confounders such as duration of diabetes. The use of variance components analysis assumes that the same gene(s) are responsible for retinopathy in general. However, it is quite plausible that there are genes that cause a susceptibility to proliferative retinopathy, for instance, that are not involved in the susceptibility to nonproliferative retinopathy. The OR analysis certainly suggests a stronger familial risk for severe retinopathy compared with any grade of retinopathy. Because retinopathy is also highly dependent on diabetes duration, subjects studied at short durations without retinopathy may still have the genes for retinopathy susceptibility. An affected sibpair analysis might be a preferable method if this is the case rather than adjustment of the retinopathy grade for diabetes duration as is done here. However, we do not present data here for affected sibpair analysis because the number of affected (any retinopathy) sibpairs was low (n = 98), and thus the analysis lacked adequate power.

To assess how diabetic retinopathy in the Pima Indians is representative of diabetic retinopathy in the wider population, we can look at previous reports of prevalence, incidence, and progression. Estimates of the prevalence of diabetic retinopathy vary between studies. A recent analysis of pooled studies estimated a crude prevalence of retinopathy of 40.3% in adults aged ≥40 years with diabetes (31). Estimated prevalence tends to be higher among subjects with type 1 diabetes than in subjects with type 2 diabetes. In the Wisconsin Epidemiology of Diabetic Retinopathy (WESDR) cohort, the prevalence of retinopathy in the first 2 years after diagnosis of type 1 diabetes was 7%, rising to almost 100% after 15 years (32). In the WESDR cohort of patients with type 2 diabetes not treated with insulin, prevalence was higher: 23% for 0–2 years' duration and 57.5% at >15 years' duration of diabetes (33). There are also reports of differences in prevalence of retinopathy according to ethnicity, with a higher prevalence reported among Mexican Americans (41.3%) and non-Hispanic blacks (30.4%) than in non-Hispanic whites (15.5%) at a diabetes duration of 5–14 years (34). Prevalence of diabetic retinopathy among Pima Indians is estimated at 37.8% with a mean duration of diabetes of 9 years (35). The prevalence in this study is somewhat lower (25.5%). Reported incidence rates for diabetic retinopathy vary; the UK Prospective Diabetes Study (UKPDS) reported a 6-year cumulative incidence of 22% (3), whereas the WESDR study reported a 34–47% 4-year cumulative incidence (36). Among the Pima Indians, we previously reported a 4-year cumulative incidence of 16.8% (38). Progression rates of existing retinopathy estimates range from 25 to 34% over 4 years in the WESDR cohort (37), and the estimate is 29% in the UKPDS cohort (3). In the Pima Indians, we previously reported a progression rate of 23.5% over a mean period of 3.6 years (4). Studies of risk factors for retinopathy in the Pima Indians identified similar risk factors for incidence and progression of retinopathy as studies in other populations (4). Thus, we believe that genetic factors that influence diabetic retinopathy within the Pima Indians may well also be important in other populations.

The incidence of diabetic retinopathy is strongly related to diabetes duration (4,5,37), and, therefore, the prevalence is higher when subjects are examined after longer durations of diabetes. The longitudinal study design allowed for subjects to be examined at varying durations of diabetes. Confounding by duration was minimized by selecting examinations nearest to 10-year duration, but a wide range of durations were included because many participants only had retinal photographs at other durations of diabetes. Differences in duration were accounted for by adjusting for duration by linear regression. It is possible that genetic factors have a stronger influence on early predisposition to retinopathy or the development of more severe retinopathy, but there were insufficient members with early-onset severe retinopathy to assess this in the present study. Because only ∼4% of the cohort had moderate nonproliferative retinopathy or worse, our observations are primarily limited to relatively early stages of retinopathy and may not describe genetic factors associated with more severe disease. Early retinopathy may be confounded by retinopathy due to other causes (i.e., hypertensive retinopathy) and thus be unrelated to the diabetic process. The phenotype analyzed may also affect the results. The current study presents both worse eye and mean eye scores, and the results are similar, although the peak of linkage is stronger for the worse eye. The mean eye score has the advantage of using information from both eyes, which allows the phenotype to distinguish subtle differences in retinopathy between subjects that may be missed by using the worse eye. However, the worse eye score rather than the average eye score may provide more information on the genetic propensity to the disease if the differences in score between the eyes are primarily due to differences in other factors such as variation in blood pressure in the vessels of each eye. Regardless of which phenotype is used, there are still limitations to the assessment of severity because the current grading system does not differentiate severity of retinopathy as finely as is possible with more retinal views or by more precisely defining a range of severity (i.e., a single microaneurysm is graded the same as eight microaneurysms). Such imprecision in the grading could affect power.

The peak of linkage on the short arm of chromosome 1 has not been studied previously for genes associated with diabetic retinopathy. Although the 1p36 region is rich in genes, many have not yet been fully characterized. Figure 2 shows a diagram of identified genes that fall between the markers D1S228 and D1S522, which defined our linkage peak in the single-point analysis with the peak of linkage found at marker D1S3669 (diagram created using data from the National Center for Biotechnology Information Web site at http://www.ncbi.nlm.nih.gov as of August 2006).

Of the identified genes in the area, those coding for the peptidyl arginine deiminases ([PADI] 1, 2, 3, 4, and 6; Fig. 2, D–H), CASP-9 (Fig. 2, h), CLCN-Ka, and CLCN-Kb (Fig. 2, r and s) are close to our peak of linkage and offer possible mechanisms for diabetic retinopathy. PADI converts arginine residues to citrulline residues. PADI 2 is expressed in the brain, is found in retinal tissue (38), and may play a role in the pathogenesis of multiple sclerosis (39). PADI 4 is expressed primarily in neutrophils and eosinophils (40) and is associated with rheumatoid arthritis (41). Recently, PADI 4 was reported to regulate histone argine methylation and thus could potentially play a broader role in gene expression (42). CASP-9 codes for caspase-9, which is activated by the release of cytochrome c from the mitochondria and itself activates other caspases to induce cell apoptosis (43). In vitro, advanced glycosylation end products induce apoptosis in bovine retinal pericytes, and this effect was inhibited by the use of caspase inhibitors, including one specific to caspase-9 (44). The CLCN-Ka and CLCN-Kb genes code for a kidney-specific chloride channel. A functional mutation in the CLCN-Kb is associated with hypertension (45). However, in this study, the correlation between retinopathy score and blood pressure is modest.

This study suggests that diabetic retinopathy is modestly heritable in Pima Indians with type 2 diabetes. It also demonstrates that an area of chromosome 1, which has not previously been studied in relation to diabetic retinopathy, has suggestive evidence of linkage. Further linkage studies of diabetic retinopathy may help to define the role of this locus in other populations. Association studies of candidate genes in the region, such as CASP-9, PADI 4, and CLCN-Kb, may also be informative.

FIG. 1.

LOD scores for worse eye score across all chromosomes. Solid line for multipoint analysis and dashed line for single-point analysis.

FIG. 1.

LOD scores for worse eye score across all chromosomes. Solid line for multipoint analysis and dashed line for single-point analysis.

Close modal
FIG. 2.

Diagram of area of peak linkage on chromosome 1 (13,000–19,800 MB) with identified genes indicated by arrows, excluding pseudogenes or predicted genes. a, PRDM2; b, KIAA1026; c, TMEM51; d, EFHD2; e, CTRC; f, ELA2A; g, ELA2B; h, CASP9; i, DNAJC16; j, AGMAT; k, DDI2; l, RSC1A; m, RSC1A1; n, FBLIM1; o, SPEN; p, ZBTB17; q, HSPB7; r, CLCN-Ka; s, CLCN-Kb; t, EPHA2; u, ARHGEF19; v, FBXO42; w, C1orf144; x, SPATA21; y, NECAP2; z, CROCC; A, MFAP2; B, ATP13A2; C, SDHB; D, PADI 2; E, PADI 1; F, PADI 3; G, PADI 4; H, PADI 6; I, RCC2; J, ARHGEF10L; K, ACTL8; L, IGSF21; M, KLHDC7A; N, PAX7; O, TAS1R; P, ALDH4A1.

FIG. 2.

Diagram of area of peak linkage on chromosome 1 (13,000–19,800 MB) with identified genes indicated by arrows, excluding pseudogenes or predicted genes. a, PRDM2; b, KIAA1026; c, TMEM51; d, EFHD2; e, CTRC; f, ELA2A; g, ELA2B; h, CASP9; i, DNAJC16; j, AGMAT; k, DDI2; l, RSC1A; m, RSC1A1; n, FBLIM1; o, SPEN; p, ZBTB17; q, HSPB7; r, CLCN-Ka; s, CLCN-Kb; t, EPHA2; u, ARHGEF19; v, FBXO42; w, C1orf144; x, SPATA21; y, NECAP2; z, CROCC; A, MFAP2; B, ATP13A2; C, SDHB; D, PADI 2; E, PADI 1; F, PADI 3; G, PADI 4; H, PADI 6; I, RCC2; J, ARHGEF10L; K, ACTL8; L, IGSF21; M, KLHDC7A; N, PAX7; O, TAS1R; P, ALDH4A1.

Close modal
TABLE 1

Grading system for retinal photographs

GradeDescription
10 No retinopathy; questionable diabetic retinopathy, usually one questionable MA; retinopathy due to other causes 
21 Definite hard exudates, soft exudates, or intraretinal microvascular abnormalities in the absence of MAs; retinal hemorrhage without MAs; MAs only without other diabetic lesions 
31 MAs and one or more of the following: retinal hemorrhage (but a total of hemorrhages and MAs less than those in standard photograph 2A), hard exudates, soft exudates questionably present, intraretinal microvascular abnormalities questionably present, venous beading questionably present, and venous loops definitely present 
41 MAs and one or more of the following: soft exudates definitely present and intraretinal microvascular definitely present but less than necessary for level 51 
51 MAs and one or more of the following: venous beading definitely present, hemorrhages and microaneurysms equaling or exceeding those in standard photograph 2A, and intraretinal microvascular abnormalities present and equaling or exceeding those in photograph 8A in two fields 
60, 65, 70 Any of several levels of severity of proliferative retinopathy, including neovascularization, fibrous proliferation, vitreous hemorrhage and preretinal hemorrhage, scars of panretinal photocoagulation, and/or retinopathy ungradeable because of vitreous hemorrhage obscuring the retina, phthisis bulbi, or enucleation secondary to a complication of diabetic retinopathy 
GradeDescription
10 No retinopathy; questionable diabetic retinopathy, usually one questionable MA; retinopathy due to other causes 
21 Definite hard exudates, soft exudates, or intraretinal microvascular abnormalities in the absence of MAs; retinal hemorrhage without MAs; MAs only without other diabetic lesions 
31 MAs and one or more of the following: retinal hemorrhage (but a total of hemorrhages and MAs less than those in standard photograph 2A), hard exudates, soft exudates questionably present, intraretinal microvascular abnormalities questionably present, venous beading questionably present, and venous loops definitely present 
41 MAs and one or more of the following: soft exudates definitely present and intraretinal microvascular definitely present but less than necessary for level 51 
51 MAs and one or more of the following: venous beading definitely present, hemorrhages and microaneurysms equaling or exceeding those in standard photograph 2A, and intraretinal microvascular abnormalities present and equaling or exceeding those in photograph 8A in two fields 
60, 65, 70 Any of several levels of severity of proliferative retinopathy, including neovascularization, fibrous proliferation, vitreous hemorrhage and preretinal hemorrhage, scars of panretinal photocoagulation, and/or retinopathy ungradeable because of vitreous hemorrhage obscuring the retina, phthisis bulbi, or enucleation secondary to a complication of diabetic retinopathy 

MA, microaneurysm. From Looker et al. (4).

TABLE 2

Breakdown of sibpairs

Siblings in familyFamiliesSibpairs
110 110 
51 153 
31 186 
11 110 
60 
42 
28 
36 
Total 211 725 
Siblings in familyFamiliesSibpairs
110 110 
51 153 
31 186 
11 110 
60 
42 
28 
36 
Total 211 725 

Data are n.

TABLE 3

Subject characteristics

nMean/frequency
Sex (% male)* 607 36.1 
Age (years) 607 43.8 ± 10.7 
Diabetes duration (years) 607 9.1 (4.4–11.5) 
BMI (kg/m2607 35.8 ± 11.8 
Diastolic blood pressure (mmHg) 601 77 ± 11 
Systolic blood pressure (mmHg) 602 125 ± 20 
ACR (mg/g) 587 26.6 (9.3–94.8) 
A1C (%) 591 8.9 ± 2.5 
Worse eye score 607 10 (10–31) 
Average eye score 607 10 (10–25.5) 
nMean/frequency
Sex (% male)* 607 36.1 
Age (years) 607 43.8 ± 10.7 
Diabetes duration (years) 607 9.1 (4.4–11.5) 
BMI (kg/m2607 35.8 ± 11.8 
Diastolic blood pressure (mmHg) 601 77 ± 11 
Systolic blood pressure (mmHg) 602 125 ± 20 
ACR (mg/g) 587 26.6 (9.3–94.8) 
A1C (%) 591 8.9 ± 2.5 
Worse eye score 607 10 (10–31) 
Average eye score 607 10 (10–25.5) 

Data are means ± SD except for

*

frequency and

median with 25th to 75th percentile.

Before 1990, glycemia was assessed using only the HbAl. Because examinations in this study were performed between 1982 and 2003, the HbA1 measures were converted to A1C using the equation A1C = (0.99 × HbA1) − 1.535, which was derived from samples measured with both assays. ACR, albumin-to-creatine ratio.

TABLE 4

Frequency distribution of retinopathy

Mean retinopathy scoreFrequency
10 386 (63.6) 
11–21 66 (10.9) 
>21–31 41 (6.7) 
>31–41 89 (14.7) 
>41–51 15 (2.5) 
>51–60 2 (0.3) 
>60 8 (1.3) 
Mean retinopathy scoreFrequency
10 386 (63.6) 
11–21 66 (10.9) 
>21–31 41 (6.7) 
>31–41 89 (14.7) 
>41–51 15 (2.5) 
>51–60 2 (0.3) 
>60 8 (1.3) 

Data are n (%).

TABLE 5

Single-point analysis

MarkerChromosome 1 locus (cM)LOD
Worse eye score D1S3669 34.2 3.01 
 D1S1622 53.1 1.01 
Mean eye score D1S3669 34.2 2.40 
MarkerChromosome 1 locus (cM)LOD
Worse eye score D1S3669 34.2 3.01 
 D1S1622 53.1 1.01 
Mean eye score D1S3669 34.2 2.40 

All LOD scores ≥1.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

This research has received support from the Intramural Research Program of the National Institute of Diabetes and Digestive and Kidney Diseases and by the National Eye Institute.

We thank Susan Vitale at the National Eye Institute for her help during the planning of this project and Leslie Baier for her advice in the preparation of this manuscript. We are grateful to the participants of the Gila River Indian Community for their participation in this study, to all at the National Institutes of Health clinic (Sacaton, AZ), and to the staff at the Ocular Epidemiology Reading Center (Madison, WI).

1.
Klein R, Klein BE, Moss SE: Visual impairment in diabetes.
Ophthalmology
91
:
1
–9,
1984
2.
Congdon N, O'Colmain B, Klaver CC, Klein R, Munoz B, Friedman DS, Kempen J, Taylor HR, Mitchell P: Causes and prevalence of visual impairment among adults in the United States.
Arch Ophthalmol
122
:
477
–485,
2004
3.
Stratton IM, Kohner EM, Aldington SJ, Turner RC, Holman RR, Manley SE, Matthews DR: UKPDS 50: risk factors for incidence and progression of retinopathy in type II diabetes over 6 years from diagnosis.
Diabetologia
44
:
156
–163,
2001
4.
Looker HC, Krakoff J, Knowler WC, Bennett PH, Klein R, Hanson RL: Longitudinal studies of incidence and progression of diabetic retinopathy assessed by retinal photography in Pima Indians.
Diabetes Care
26
:
320
–326,
2003
5.
Agardh E, Agardh CD, Koul S, Torffvit O: A four-year follow-up study on the incidence of diabetic retinopathy in older onset diabetes mellitus.
Diabet Med
11
:
273
–278,
1994
6.
Cohen O, Norymberg K, Neumann E, Dekel H: Complication-free duration and the risk of development of retinopathy in elderly diabetic patients.
Arch Intern Med
158
:
641
–644,
1998
7.
Yoshida Y, Hagura R, Hara Y, Sugasawa G, Akanuma Y: Risk factors for the development of diabetic retinopathy in Japanese type 2 diabetic patients.
Diabetes Res Clin Pract
51
:
195
–203,
2001
8.
Leslie RD, Pyke DA: Diabetic retinopathy in identical twins.
Diabetes
31
:
19
–21,
1982
9.
The Diabetes Control and Complications Trial Research Group: Clustering of long-term complications in families with diabetes in the diabetes control and complications trial.
Diabetes
46
:
1829
–1839,
1997
10.
Rema M, Saravanan G, Deepa R, Mohan V: Familial clustering of diabetic retinopathy in South Indian type 2 diabetic patients.
Diabet Med
19
:
910
–916,
2002
11.
Hallman DM, Huber JC Jr, Gonzalez VH, Klein BE, Klein R, Hanis CL: Familial aggregation of severity of diabetic retinopathy in Mexican Americans from Starr County, Texas.
Diabetes Care
28
:
1163
–1168,
2005
12.
Warpeha KM, Chakravarthy U: Molecular genetics of microvascular disease in diabetic retinopathy.
Eye
17
:
305
–311,
2003
13.
Imperatore G, Hanson RL, Pettitt DJ, Kobes S, Bennett PH, Knowler WC, Pima Diabetes Genes Group: Sib-pair linkage analysis for susceptibility genes for microvascular complications among Pima Indians with type 2 diabetes.
Diabetes
47
:
821
–830,
1998
14.
Hanson RL, Ehm MG, Pettitt DJ, Prochazka M, Thompson DB, Timberlake D, Foroud T, Kobes S, Baier L, Burns DK, Almasy L, Blangero J, Garvey WT, Bennett PH, Knowler WC: An autosomal genomic scan for loci linked to type II diabetes mellitus and body-mass index in Pima Indians.
Am J Hum Genet
63
:
1130
–1138,
1998
15.
Diabetes mellitus. Report of a WHO Study Group.
World Health Organ Tech Rep Ser
727
:
1
–113,
1985
16.
Klein R, Klein BE, Magli YL, Brothers RJ, Meuer SM, Moss SE, Davis MD: An alternative method of grading diabetic retinopathy.
Ophthalmology
93
:
1183
–1187,
1986
17.
Amos CI: Robust variance-components approach for assessing genetic linkage in pedigrees.
Am J Hum Genet
54
:
535
–543,
1994
18.
Amos CI, Zhu DK, Boerwinkle E: Assessing genetic linkage and association with robust components of variance approaches.
Ann Hum Genet
60
:
143
–160,
1996
19.
Neale MC, Miller MB: The use of likelihood-based confidence intervals in genetic models.
Behav Genet
27
:
113
–120,
1997
20.
Schwengel DA, Jedlicka AE, Nanthakumar EJ, Weber JL, Levitt RC: Comparison of fluorescence-based semi-automated genotyping of multiple microsatellite loci with autoradiographic techniques.
Genomics
22
:
46
–54,
1994
21.
Dubovsky J, Sheffield VC, Duyk GM, Weber JL: Sets of short tandem repeat polymorphisms for efficient linkage screening of the human genome.
Hum Mol Genet
4
:
449
–452,
1995
22.
Ehm M, Wagner M: A test statistic to detect errors in sib-pair relationships.
Am J Hum Genet
62
:
181
–188,
1998
23.
Curtis D, Sham PC: Using risk calculation to implement an extended relative pair analysis.
Ann Hum Genet
58
:
151
–162,
1994
24.
Fulker DW, Cherny SS: An improved multipoint sib-pair analysis of quantitative traits.
Behav Genet
26
:
527
–532,
1996
25.
Warpeha KM, Xu W, Liu L, Charles IG, Patterson CC, Ah-Fat F, Harding S, Hart PM, Chakravarthy U, Hughes AE: Genotyping and functional analysis of a polymorphic (CCTTT)(n) repeat of NOS2A in diabetic retinopathy.
FASEB J
13
:
1825
–1832,
1999
26.
Kamiuchi K, Hasegawa G, Obayashi H, Kitamura A, Ishii M, Yano M, Kanatsuna T, Yoshikawa T, Nakamura N: Intercellular adhesion molecule-1 (ICAM-1) polymorphism is associated with diabetic retinopathy in type 2 diabetes mellitus.
Diabet Med
19
:
371
–376,
2002
27.
Ichikawa F, Yamada K, Ishiyama-Shigemoto S, Yuan X, Nonaka K: Association of an (A-C)n dinucleotide repeat polymorphic marker at the 5′-region of the aldose reductase gene with retinopathy but not with nephropathy or neuropathy in Japanese patients with type 2 diabetes mellitus.
Diabet Med
16
:
744
–748,
1999
28.
Beranek M, Kankova K, Benes P, Izakovicova-Holla L, Znojil V, Hajek D, Vlkova E, Vacha J: Polymorphism R25P in the gene encoding transforming growth factor-beta (TGF-beta1) is a newly identified risk factor for proliferative diabetic retinopathy.
Am J Med Genet
109
:
278
–283,
2002
29.
Nagi DK, McCormack LJ, Mohamed-Ali V, Yudkin JS, Knowler WC, Grant PJ: Diabetic retinopathy, promoter (4G/5G) polymorphism of PAI-1 gene, and PAI-1 activity in Pima Indians with type 2 diabetes.
Diabetes Care
20
:
1304
–1309,
1997
30.
Morton NE: Significance levels in complex inheritance.
Am J Hum Genet
62
:
690
–697,
1998
31.
Kempen JH, O'Colmain BJ, Leske MC, Haffner SM, Klein R, Moss SE, Taylor HR, Hamman RF: The prevalence of diabetic retinopathy among adults in the United States.
Arch Ophthalmol
122
:
552
–563,
2004
32.
Klein R, Klein BE, Moss SE, Davis MD, DeMets DL: The Wisconsin epidemiologic study of diabetic retinopathy: II. Prevalence and risk of diabetic retinopathy when age at diagnosis is less than 30 years.
Arch Ophthalmol
102
:
520
–526,
1984
33.
Klein R, Klein BE, Moss SE, Davis MD, DeMets DL: The Wisconsin epidemiologic study of diabetic retinopathy: III. Prevalence and risk of diabetic retinopathy when age at diagnosis is more than 30 years.
Arch Ophthalmol
102
:
527
–532,
1984
34.
Wong TY, Klein R, Islam FM, Cotch MF, Folsom AR, Klein BE, Sharrett AR, Shea S: Diabetic retinopathy in a multi-ethnic cohort in the United States.
Am J Ophthalmol
141
:
446
–455,
2006
35.
Nagi DK, Pettit DJ, Bennett PH, Klein R, Knowler WC: Diabetic retinopathy assessed by fundus photography in Pima Indians with impaired glucose tolerance and NIDDM.
Diabet Med
14
:
449
–456,
1997
36.
Knowler WC, Bennett PH, Ballintine EJ: Increased incidence of retinopathy in diabetics with elevated blood pressure: a six-year follow-up study in Pima Indians.
N Engl J Med
302
:
645
–650,
1980
37.
Klein R, Klein BE, Moss SE, Davis MD, DeMets DL: The Wisconsin epidemiologic study of diabetic retinopathy: X. Four-year incidence and progression of diabetic retinopathy when age at diagnosis is 30 years or more.
Arch Ophthalmol
107
:
244
–249,
1989
38.
Vossenaar ER, Zendman AJ, Van Venrooij WJ, Pruijn GJ: PAD, a growing family of citrullinating enzymes: genes, features and involvement in disease.
Bioessays
25
:
1106
–1118,
2003
39.
Mastronardi FG, Moscarello MA: Molecules affecting myelin stability: a novel hypothesis regarding the pathogenesis of multiple sclerosis.
J Neurosci Res
80
:
301
–308,
2005
40.
Asaga H, Nakashima K, Senshu T, Ishigami A, Yamada M: Immunocytochemical localization of peptidylarginine deiminase in human eosinophils and neutrophils.
J Leukoc Biol
70
:
46
–51,
2001
41.
Suzuki A, Yamada R, Chang X, Tokuhiro S, Sawada T, Suzuki M, Nagasaki M, Nakayama-Hamada M, Kawaida R, Ono M, Ohtsuki M, Furukawa H, Yoshino S, Yukioka M, Tohma S, Matsubara T, Wakitani S, Teshima R, Nishioka Y, Sekine A, Iida A, Takahashi A, Tsunoda T, Nakamura Y, Yamamoto K: Functional haplotypes of PADI4, encoding citrullinating enzyme peptidylarginine deiminase 4, are associated with rheumatoid arthritis.
Nat Genet
34
:
395
–402,
2003
42.
Wang Y, Wysocka J, Sayegh J, Lee YH, Perlin JR, Leonelli L, Sonbuchner LS, McDonald CH, Cook RG, Dou Y, Roeder RG, Clarke S, Stallcup MR, Allis CD, Coonrod SA: Human PAD4 regulates histone arginine methylation levels via demethylimination.
Science
306
:
279
–283,
2004
43.
Reed JC: Cytochrome c: can't live with it—can't live without it.
Cell
91
:
559
–562,
1997
44.
Lecomte M, Denis U, Ruggiero D, Lagarde M, Wiernsperger N: Involvement of caspase-10 in advanced glycation end-product-induced apoptosis of bovine retinal pericytes in culture.
Biochim Biophys Acta
1689
:
202
–211,
2004
45.
Jeck N, Waldegger S, Lampert A, Boehmer C, Waldegger P, Lang PA, Wissinger B, Friedrich B, Risler T, Moehle R, Lang UE, Zill P, Bondy B, Schaeffeler E, Asante-Poku S, Seyberth H, Schwab M, Lang F: Activating mutation of the renal epithelial chloride channel ClC-Kb predisposing to hypertension.
Hypertension
43
:
1175
–1181,
2004