OBJECTIVE

Diabetic retinopathy is a sight-threatening microvascular complication of diabetes with a complex multifactorial pathogenesis. A systematic meta-analysis was undertaken to collectively assess genetic studies and determine which previously investigated polymorphisms are associated with diabetic retinopathy.

RESEARCH DESIGN AND METHODS

All studies investigating the association of genetic variants with the development of diabetic retinopathy were identified in PubMed and ISI Web of Knowledge. Crude odds ratios (ORs) and 95% CIs were calculated for single nucleotide polymorphisms and microsatellite markers previously investigated in at least two published studies.

RESULTS

Twenty genes and 34 variants have previously been studied in multiple cohorts. The aldose reductase (AKR1B1) gene was found to have the largest number of polymorphisms significantly associated with diabetic retinopathy. The z−2 micro satellite was found to confer risk (OR 2.33 [95% CI 1.49–3.64], P = 2 × 10−4) in type 1 and type 2 diabetes and z+2 to confer protection (0.58 [0.36–0.93], P = 0.02) against diabetic retinopathy in type 2 diabetes regardless of ethnicity. The T allele of the AKR1B1 promoter rs759853 variant is also significantly protective against diabetic retinopathy in type 1 diabetes (0.5 [0.35–0.71], P = 1.00 × 10−4), regardless of ethnicity. These associations were also found in the white population alone (P < 0.05). Polymorphisms in NOS3, VEGF, ITGA2, and ICAM1 are also associated with diabetic retinopathy after meta-analysis.

CONCLUSIONS

Variations within the AKR1B1 gene are highly significantly associated with diabetic retinopathy development irrespective of ethnicity. Identification of genetic risk factors in diabetic retinopathy will assist in further understanding of this complex and debilitating diabetes complication.

Diabetic retinopathy is a sight-threatening microvascular complication of diabetes. Global population–based data indicate that it is the fifth most common cause of blindness in the world, accounting for ∼4.8% of global blindness (1). Diabetic retinopathy is a leading cause of blindness in industrialized countries (1), and its significance is likely to increase with increasing frequency of diabetes (2). Diabetic retinopathy is defined by the presence of retinal microvascular lesions. Early retinal changes include microaneurysms, hemorrhages, hard exudates, cotton wool spots, intraretinal microvascular abnormalities, and venous beading. These clinical features occur in nonproliferative diabetic retinopathy (NPDR). Growth of abnormal new blood vessels that frequently lead to preretinal and vitreous hemorrhage are the principal hallmarks of proliferative diabetic retinopathy (PDR). Visual impairment in diabetic retinopathy occurs secondary to preretinal or vitreous hemorrhage and diabetic maculopathy (either macular edema or macular ischemia) (3).

The prevalence of diabetic retinopathy is higher in type 1 diabetic patients, with sight-threatening retinopathy reported to be up to 2.5 times more common than in those with type 2 diabetes (4). Conversely, the incidence of macular edema has been reported to be up to two times higher among those with type 2 diabetes (5).

The pathogenesis of diabetes is believed to be multifactorial, with genetic risk factors playing a fundamental role. Recent genome-wide association studies have identified several genetic loci involved in the pathogenesis of both type 1 diabetes and type 2 diabetes (6,7). Pathogenesis of diabetic retinopathy is complex, also having a multifactorial biochemical pathogenesis, primarily because of altered glucose metabolism (8). Glycemic control (9,10) and the duration of diabetes (11,12) have been identified through large longitudinal prospective studies as being major risk factors for the development of diabetic retionopathy. However, genetic factors are also likely to account for the susceptibility to this blinding disease as well as the differences in diabetic retinopathy incidence between individuals with type 1 and type 2 diabetes. It has become evident through familial aggregation studies that susceptibility to complications of diabetes such as diabetic retinopathy also have a heritable component independent of glycemic control and duration of diabetes. The Diabetes Control and Complications Trial revealed a familial tendency for severe diabetic retinopathy in type 1 and type 2 diabetes (13). These findings were also replicated in Mexican Americans (14,15) and other familial risk studies (16,17), independent of associated environmental risk factors. There have been numerous studies investigating candidate genes for diabetic retinopathy susceptibility in subjects of various ethnicities.

The aim of this systematic meta-analysis was to analyze all published studies that investigated the association between genetic risk factors and the development of diabetic retinopathy and met specified inclusion criteria. We sought to determine which of the previously investigated genetic variants are significantly associated with the development of diabetic retinopathy in type 1 or type 2 diabetes and to examine the strength of these associations.

Literature search and data collection.

A systematic literature search was performed to identify all studies published between January 1990 and August 2008 that investigated the association of genetic variants with the development of any form of diabetic retinopathy. The PubMed database (National Center for Biotechnology Information; NCBI), ISI Web of Knowledge (version 4.5), and the Cochrane Library were explored independently by two authors using the following keyword strings: “genetic” AND “diabetic retinopathy” and “gene” AND “diabetic retinopathy.” All retrieved publications were reviewed and compared and studies written in English that contained sufficient case (subjects with diabetes and diabetic retinopathy) and control (subjects with diabetes but without the complication of diabetic retinopathy) subjects' genotype information, such as allele or genotype frequency, were included. The reference list of each relevant publication was also examined to identify additional studies appropriate for inclusion in the meta-analysis. Polymorphisms were included in the meta-analysis if a minimum of two studies had assessed association with diabetic retinopathy development.

Data analysis.

Data were entered into a database and statistical analyses performed using RevMan software (version 4.2; Oxford, U.K.) and SPSS (version 16.0; SPSS, Chicago, IL). Data extraction and data entry was performed independently by two of the authors (S.A. and A.W.H.) and compared. Independent review and resolution by a senior investigator (K.P.B. and J.E.C.) was sought if disagreement between the reviewers occurred.

Ensemble (http://www.ensembl.org/index.html) and dbSNP (http://www.ncbi.nlm.nih.gov/SNP/) were used to locate rs# identifiers for the genetic variants. When an rs# identifier could not be located, the most commonly used name for that specific variant or single nucleotide polymorphism (SNP) was selected. All rs# identifiers were confirmed independently by a senior investigator (K.P.B.).

Analyses were performed for all cases with any form of diabetic retinopathy compared with all diabetic subjects without diabetic retinopathy (control subjects). Subanalysis was subsequently undertaken for NPDR and PDR where possible. Major and minor allele frequencies were calculated from the available genetic data of all reported variants in the included studies and odds ratios (ORs) and 95% CIs were calculated. The allelic association of microsatellite markers was also investigated, whereby the risk conferred by each allele was compared against all other alleles as well as against specific alleles. The Der Simonian and Laird random-effects model was used (18,19). This model utilizes weights that incorporate both within-study and between-study variance. Heterogeneity between studies was calculated as the inverse variance estimate. To minimize genetic heterogeneity, subanalysis was performed for all cohorts of white origin and for datasets in which heterogeneity remained; outlying studies were then removed in a stepwise fashion until homogeneity was achieved (20). For the purpose of this study, white ancestry was defined as being of European descent. Studies including subjects of white and nonwhite ancestry have had their cohorts divided and analyzed individually. Funnel plots were constructed in Revman and Egger's test (21) applied in Stata (version 10.1; Stata, College Station, TX) to investigate for publication bias. A P value of <0.05 was considered statistically significant in all analyses, except for publication bias (Egger's test), where a P value of <0.1 (22) was considered as statistically significant. No adjustments were made for multiple testing.

There were 702 publications identified. Of these, 160 were specific to genetic polymorphisms and the development of diabetic retinopathy (Fig. 1). Twelve of these studies were in a language other than English and were excluded, and a final total of 82 studies were suitable for inclusion as a result of having presented sufficient case and control genotype information. There were 196 polymorphisms identified, with 34 of these having adequate genotype data for inclusion in the meta-analysis. Visual inspection of the funnel plots revealed a symmetrical inverted V shape and Egger's test did not detect significant publication bias (P > 0.1) for all polymorphisms examined by more than five studies (supplementary Fig. 1, available in an online appendix at http://diabetes.diabetesjournals.org/cgi/content/full/db09-0059/DC1).

FIG. 1.

Flow chart of study selection process and included studies.

FIG. 1.

Flow chart of study selection process and included studies.

Seven polymorphisms were investigated by more than five studies and are discussed in detail (Table 1). Details of their study design are provided in supplementary Table 1. Forty-eight studies were included in the analyses for the seven polymorphisms, 20 (42%) of which included subjects of white ancestry (Table 2). Heterogeneity was eliminated in the subanalyses of studies with participants of white ancestry without the requirement of removal of outlying studies, indicating that ethnicity is a major source of heterogeneity generally.

TABLE 1

Meta-analysis of genetic variants and risk for diabetic retinopathy in type 1 diabetes, type 2 diabetes, and both types of diabetes regardless of participant's ethnicity

GeneVariantRisk alleleComparisonType of diabetesTotal case subjects with risk allele (%)Total control subjects with risk allele (%)WeightOR (95% CI)SENumber of studiesχ2PReferences
ACE INS/DEL at intron 16 287 base pair deletion No diabetic retinopathy vs. any diabetic retinopathy Type 1 651 (55) 458 (52) 32.76 1.06 (0.88–1.27) 0.09 3.67 0.5402 (S1–S6) 
    Type 2 986 (44) 931 (43) 67.24 1.01 (0.89–1.15) 0.06 2.67 0.866 (S3) (S7–S12) 
    Total 1,637 (47) 1,389 (45) 100.00 1.03 (0.93–1.14) 0.05 13 6.51 0.6249 
   No diabetic retinopathy vs. NPDR Type 2 276 (35) 438 (36) 100.00 0.91 (0.75–1.12) 0.10 1.51 0.3786 (S9) (S10) (S12) 
   No diabetic retinopathy vs. PDR Type 1 387 (60) 220 (55) 48.63 1.19 (0.77–1.86) 0.23 8.01 0.4355 (S1) (S2) (S4) (S5) 
   Type 2 221 (41) 438 (36) 51.37 1.12 (0.89–1.42) 0.12 0.21 0.3237 (S9) (S10) (S12) 
    Total 608 (51) 658 (40) 100.00 1.17 (0.94–1.44) 0.11 8.55 0.1515  
   NPDR vs. PDR Type 1 120 (64) 72 (45) 17.02 2.21 (1.44–3.39) 0.22 0.0003 (S4) 
    Type 2 353 (45) 481 (41) 82.98 1.12 (0.92–1.35) 0.1 4.75 0.2512 (S9) (S10) (S12–S15) 
    Total 473 (49) 553 (41) 100.00 1.25 (0.95–1.65) 0.14 12.77 0.1064  
NOS3 rs3138808 393 base pair insertion No diabetic retinopathy vs. any diabetic retinopathy Type 1 21 (11) 42 (22) 9.91 0.43 (0.24–0.76) 0.29 0.00 0.0036 (S16) 
    Type 2 498 (21) 705 (26) 90.09 0.93 (0.74–1.17) 0.12 13.63 0.5408 (S17–S23) 
    Total 519 (21) 747 (26) 100.00 0.87 (0.68–1.11) 0.13 19.38 0.2578 
   No diabetic retinopathy vs. NPDR Type 1 21 (11) 42 (22) 32.47 0.43 (0.24–0.76) 0.29 0.00 0.0036 (S16) 
   Type 2 95 (20) 116 (24) 67.53 0.53 (0.12–2.46) 0.78 22.10 0.421 (S17) (S23) 
    Total 116 (17) 158 (24) 100.00 0.5 (0.19–1.31) 0.49 22.99 0.157  
   No diabetic retinopathy vs. PDR Type 2 43 (17) 116 (24) 100.00 0.42 (0.05–3.69) 1.11 28.95 0.4341 (S17) (S23) 
   NPDR vs. PDR Type 2 43 (17) 95 (20) 100.00 0.79 (0.42–1.5) 0.33 2.61 0.4723 (S17) (S23) 
VEGF rs2010963 No diabetic retinopathy vs. any diabetic retinopathy Type 2 1,204 (55) 1,144 (59) 100.00 0.86 (0.7–1.05) 0.10 14.65 0.1407 (S22) (S24–S29) 
   No diabetic retinopathy vs. NPDR Type 2 242 (53) 328 (65) 100.00 0.62 (0.48–0.81) 0.14 1.93 0.0005 (S24) (S28) (S29) 
   No diabetic retinopathy vs. PDR Type 2 212 (56) 643 (59) 100.00 0.8 (0.6–1.06) 0.14 0.92 0.1166 (S24) (S26) (S28) (S29) 
   NPDR vs. PDR Type 2 211 (56) 242 (53) 100.00 1.32 (0.99–1.76) 0.15 0.81 0.0559 (S24) (S28) (S29) 
AKR1B1 rs759853 No diabetic retinopathy vs. any diabetic retinopathy Type 1 114 (29) 149 (45) 33.33 0.49 (0.36–0.68) 0.16 0.02 <0.0001 (S30–S32) 
    Type 2 508 (34) 460 (26) 66.67 1.16 (0.98–1.36) 0.08 3.96 0.0885 (S33–S36) 
    Total 622 (33) 609 (29) 100.00 0.9 (0.66–1.22) 0.16 25.19 0.4857  
   No diabetic retinopathy vs. NPDR Type 1 8 (27) 28 (43) 14.13 0.49 (0.19–1.27) 0.48 0.00 0.143 (S32) 
   Type 2 175 (43) 147 (39) 85.87 1.17 (0.88–1.55) 0.15 0.09 0.2945 (S33) 
    Total 183 (42) 175 (39) 100.00 1.02 (0.7–1.5) 0.20 3.00 0.9099  
   No diabetic retinopathy vs. PDR Type 1 7 (25) 28 (43) 12.07 0.45 (0.17–1.21) 0.50 0.00 0.1144 (S32) 
   Type 2 123 (35) 147 (39) 87.93 0.84 (0.6–1.19) 0.18 1.19 0.3285 (S33) 
    Total 130 (34) 175 (39) 100.00 0.79 (0.55–1.13) 0.18 2.51 0.1949  
   NPDR vs. PDR Type 1 7 (25) 8 (27) 6.65 0.92 (0.28–2.98) 0.60 0.00 0.8848 (S32) 
    Type 2 123 (35) 175 (43) 93.35 0.76 (0.47–1.2) 0.24 1.95 0.2364 (S33) 
    Total 130 (34) 183 (42) 100.00 0.73 (0.54–0.99) 0.16 2.12 0.0428  
AKR1B1 (CA)n dinucleotide repeat No diabetic retinopathy vs. any diabetic retinopathy Type 1 344 (36) 313 (32) 39.83 1.09 (0.89–1.35) 0.11 5.24 0.3945 (S30) (S32) (S37–S40) 
    Type 2 436 (29) 612 (25) 60.17 1.04 (0.70–1.57) 0.21 46.91 0.8330 (S36) (S41–S48) 
    Total 780 (32) 925 (27) 100.00 1.05 (0.81–1.35) 0.13 15 52.88 0.7264  
  No diabetic retinopathy vs. NPDR Type 1 25 (41) 20 (31) 28.59 1.55 (0.75–3.23) 0.37 0.00 0.2366 (S32) 
   Type 2 97 (27) 81 (37) 71.41 0.65 (0.45–0.94) 0.19 0.03 0.0215 (S48) (S42) 
    Total 122 (29) 101 (36) 100.00 0.83 (0.49–1.42) 0.27 4.38 0.4937  
  No diabetic retinopathy vs. PDR Type 1 132 (43) 147 (37) 37.17 1.29 (0.95–1.75) 0.15 0.30 0.0979 (S39) (S32) 
   Type 2 179 (31) 171 (38) 62.83 0.77 (0.59–1.00) 0.13 1.63 0.0482 (S48) (S41–S43) 
    Total 311 (35) 318 (37) 100.00 0.94 (0.72–1.24) 0.14 8.40 0.6738  
  NPDR vs. PDR Type 1 25 (41) 20 (31) 28.59 1.55 (0.75–3.23) 0.37 0.00 0.2366 (S32) 
    Type 2 97 (27) 81 (37) 71.41 0.65 (0.45–0.94) 0.19 0.03 0.0215 (S48) (S42) 
    Total 122 (29) 101 (36) 100.00 0.83 (0.49–1.42) 0.27 4.38 0.4937  
  Z−2 No diabetic retinopathy vs. any diabetic retinopathy Type 1 362 (37) 261 (27) 41.98 1.95 (1.04–3.66) 0.32 40.66 0.0367 (S30) (S32) (S37–S40) 
    Type 2 385 (26) 171 (7) 58.02 2.64 (1.39–5.01) 0.33 69.93 0.0029 (S36) (S41–S48) 
    Total 747 (30) 432 (13) 100.00 2.33 (1.49–3.64) 0.23 15 120.27 0.0002  
  Z−2 No diabetic retinopathy vs. NPDR Type 1 22 (36) 23 (35) 22.71 1.03 (0.50–2.13) 0.37 0.00 0.9400 (S32) 
   Type 2 137 (38) 55 (25) 77.29 1.87 (1.29–2.71) 0.19 0.20 0.0010 (S48) (S42) 
    Total 159 (38) 78 (28) 100.00 1.64 (1.14–2.35) 0.18 2.26 0.0075  
  Z−2 No diabetic retinopathy vs. PDR Type 1 41 (24) 35 (23) 24.74 1.18 (0.69–2.01) 0.27 0.31 0.5363 (S32) (S40) 
   Type 2 173 (30) 85 (19) 75.26 1.64 (1.21–2.22) 0.16 2.50 0.0016 (S48) (S41–S43) 
    Total 214 (29) 120 (20) 100.00 1.51 (1.16–1.97) 0.14 3.89 0.0023  
  Z−2 NPDR vs. PDR Type 1 19 (18) 12 (14) 18.52 1.39 (0.64–3.04) 0.40 0.00 0.4076 (S40) 
    Type 2 137 (38) 55 (25) 81.48 1.87 (1.29–2.71) 0.19 0.20 0.0010 (S48) (S42) 
    Total 156 (34) 67 (22) 100.00 1.77 (1.26–2.48) 0.17 0.64 0.0009  
  z+2 No diabetic retinopathy vs. any diabetic retinopathy Type 1 153 (16) 224 (23) 40.68 0.58 (0.36–0.93) 0.24 18.16 0.0243 (S30) (S32) (S37–S40) 
    Type 2 331 (22) 521 (21) 59.32 0.97 (0.57–1.65) 0.27 64.22 0.9038 (S36) (S41–S48) 
    Total 484 (20) 745 (22) 100.00 0.79 (0.53–1.17) 0.20 15 104.32 0.2388  
  z+2 No diabetic retinopathy vs. NPDR Type 1 7 (12) 13 (20) 22.07 0.52 (0.19–1.40) 0.51 0.00 0.1955 (S32) 
   Type 2 34 (10) 29 (14) 77.93 0.71 (0.42–1.20) 0.27 0.07 0.2001 (S48) (S42) 
    Total 41 (10) 42 (15) 100.00 0.66 (0.41–1.05) 0.24 0.36 0.0820  
  z+2 No diabetic retinopathy vs. PDR Type 1 48 (16) 70 (18) 38.34 0.84 (0.50–1.40) 0.26 1.26 0.4971 (S39) (S32) 
   Type 2 82 (14) 98 (22) 61.66 0.62 (0.39–0.99) 0.24 4.71 0.0464 (S41–S43) (S48) 
    Total 130 (15) 168 (20) 100.00 0.71 (0.51–0.98) 0.17 6.92 0.0383  
  z+2 NPDR vs. PDR Type 1 7 (12) 13 (20) 22.07 0.52 (0.19–1.40) 0.51 0.00 0.1955 (S32) 
    Type 2 34 (10) 29 (14) 77.93 0.71 (0.42–1.20) 0.27 0.07 0.2001 (S48) (S42) 
    Total 41 (10) 42 (15) 100.00 0.66 (0.41–1.05) 0.24 0.36 0.0820  
GeneVariantRisk alleleComparisonType of diabetesTotal case subjects with risk allele (%)Total control subjects with risk allele (%)WeightOR (95% CI)SENumber of studiesχ2PReferences
ACE INS/DEL at intron 16 287 base pair deletion No diabetic retinopathy vs. any diabetic retinopathy Type 1 651 (55) 458 (52) 32.76 1.06 (0.88–1.27) 0.09 3.67 0.5402 (S1–S6) 
    Type 2 986 (44) 931 (43) 67.24 1.01 (0.89–1.15) 0.06 2.67 0.866 (S3) (S7–S12) 
    Total 1,637 (47) 1,389 (45) 100.00 1.03 (0.93–1.14) 0.05 13 6.51 0.6249 
   No diabetic retinopathy vs. NPDR Type 2 276 (35) 438 (36) 100.00 0.91 (0.75–1.12) 0.10 1.51 0.3786 (S9) (S10) (S12) 
   No diabetic retinopathy vs. PDR Type 1 387 (60) 220 (55) 48.63 1.19 (0.77–1.86) 0.23 8.01 0.4355 (S1) (S2) (S4) (S5) 
   Type 2 221 (41) 438 (36) 51.37 1.12 (0.89–1.42) 0.12 0.21 0.3237 (S9) (S10) (S12) 
    Total 608 (51) 658 (40) 100.00 1.17 (0.94–1.44) 0.11 8.55 0.1515  
   NPDR vs. PDR Type 1 120 (64) 72 (45) 17.02 2.21 (1.44–3.39) 0.22 0.0003 (S4) 
    Type 2 353 (45) 481 (41) 82.98 1.12 (0.92–1.35) 0.1 4.75 0.2512 (S9) (S10) (S12–S15) 
    Total 473 (49) 553 (41) 100.00 1.25 (0.95–1.65) 0.14 12.77 0.1064  
NOS3 rs3138808 393 base pair insertion No diabetic retinopathy vs. any diabetic retinopathy Type 1 21 (11) 42 (22) 9.91 0.43 (0.24–0.76) 0.29 0.00 0.0036 (S16) 
    Type 2 498 (21) 705 (26) 90.09 0.93 (0.74–1.17) 0.12 13.63 0.5408 (S17–S23) 
    Total 519 (21) 747 (26) 100.00 0.87 (0.68–1.11) 0.13 19.38 0.2578 
   No diabetic retinopathy vs. NPDR Type 1 21 (11) 42 (22) 32.47 0.43 (0.24–0.76) 0.29 0.00 0.0036 (S16) 
   Type 2 95 (20) 116 (24) 67.53 0.53 (0.12–2.46) 0.78 22.10 0.421 (S17) (S23) 
    Total 116 (17) 158 (24) 100.00 0.5 (0.19–1.31) 0.49 22.99 0.157  
   No diabetic retinopathy vs. PDR Type 2 43 (17) 116 (24) 100.00 0.42 (0.05–3.69) 1.11 28.95 0.4341 (S17) (S23) 
   NPDR vs. PDR Type 2 43 (17) 95 (20) 100.00 0.79 (0.42–1.5) 0.33 2.61 0.4723 (S17) (S23) 
VEGF rs2010963 No diabetic retinopathy vs. any diabetic retinopathy Type 2 1,204 (55) 1,144 (59) 100.00 0.86 (0.7–1.05) 0.10 14.65 0.1407 (S22) (S24–S29) 
   No diabetic retinopathy vs. NPDR Type 2 242 (53) 328 (65) 100.00 0.62 (0.48–0.81) 0.14 1.93 0.0005 (S24) (S28) (S29) 
   No diabetic retinopathy vs. PDR Type 2 212 (56) 643 (59) 100.00 0.8 (0.6–1.06) 0.14 0.92 0.1166 (S24) (S26) (S28) (S29) 
   NPDR vs. PDR Type 2 211 (56) 242 (53) 100.00 1.32 (0.99–1.76) 0.15 0.81 0.0559 (S24) (S28) (S29) 
AKR1B1 rs759853 No diabetic retinopathy vs. any diabetic retinopathy Type 1 114 (29) 149 (45) 33.33 0.49 (0.36–0.68) 0.16 0.02 <0.0001 (S30–S32) 
    Type 2 508 (34) 460 (26) 66.67 1.16 (0.98–1.36) 0.08 3.96 0.0885 (S33–S36) 
    Total 622 (33) 609 (29) 100.00 0.9 (0.66–1.22) 0.16 25.19 0.4857  
   No diabetic retinopathy vs. NPDR Type 1 8 (27) 28 (43) 14.13 0.49 (0.19–1.27) 0.48 0.00 0.143 (S32) 
   Type 2 175 (43) 147 (39) 85.87 1.17 (0.88–1.55) 0.15 0.09 0.2945 (S33) 
    Total 183 (42) 175 (39) 100.00 1.02 (0.7–1.5) 0.20 3.00 0.9099  
   No diabetic retinopathy vs. PDR Type 1 7 (25) 28 (43) 12.07 0.45 (0.17–1.21) 0.50 0.00 0.1144 (S32) 
   Type 2 123 (35) 147 (39) 87.93 0.84 (0.6–1.19) 0.18 1.19 0.3285 (S33) 
    Total 130 (34) 175 (39) 100.00 0.79 (0.55–1.13) 0.18 2.51 0.1949  
   NPDR vs. PDR Type 1 7 (25) 8 (27) 6.65 0.92 (0.28–2.98) 0.60 0.00 0.8848 (S32) 
    Type 2 123 (35) 175 (43) 93.35 0.76 (0.47–1.2) 0.24 1.95 0.2364 (S33) 
    Total 130 (34) 183 (42) 100.00 0.73 (0.54–0.99) 0.16 2.12 0.0428  
AKR1B1 (CA)n dinucleotide repeat No diabetic retinopathy vs. any diabetic retinopathy Type 1 344 (36) 313 (32) 39.83 1.09 (0.89–1.35) 0.11 5.24 0.3945 (S30) (S32) (S37–S40) 
    Type 2 436 (29) 612 (25) 60.17 1.04 (0.70–1.57) 0.21 46.91 0.8330 (S36) (S41–S48) 
    Total 780 (32) 925 (27) 100.00 1.05 (0.81–1.35) 0.13 15 52.88 0.7264  
  No diabetic retinopathy vs. NPDR Type 1 25 (41) 20 (31) 28.59 1.55 (0.75–3.23) 0.37 0.00 0.2366 (S32) 
   Type 2 97 (27) 81 (37) 71.41 0.65 (0.45–0.94) 0.19 0.03 0.0215 (S48) (S42) 
    Total 122 (29) 101 (36) 100.00 0.83 (0.49–1.42) 0.27 4.38 0.4937  
  No diabetic retinopathy vs. PDR Type 1 132 (43) 147 (37) 37.17 1.29 (0.95–1.75) 0.15 0.30 0.0979 (S39) (S32) 
   Type 2 179 (31) 171 (38) 62.83 0.77 (0.59–1.00) 0.13 1.63 0.0482 (S48) (S41–S43) 
    Total 311 (35) 318 (37) 100.00 0.94 (0.72–1.24) 0.14 8.40 0.6738  
  NPDR vs. PDR Type 1 25 (41) 20 (31) 28.59 1.55 (0.75–3.23) 0.37 0.00 0.2366 (S32) 
    Type 2 97 (27) 81 (37) 71.41 0.65 (0.45–0.94) 0.19 0.03 0.0215 (S48) (S42) 
    Total 122 (29) 101 (36) 100.00 0.83 (0.49–1.42) 0.27 4.38 0.4937  
  Z−2 No diabetic retinopathy vs. any diabetic retinopathy Type 1 362 (37) 261 (27) 41.98 1.95 (1.04–3.66) 0.32 40.66 0.0367 (S30) (S32) (S37–S40) 
    Type 2 385 (26) 171 (7) 58.02 2.64 (1.39–5.01) 0.33 69.93 0.0029 (S36) (S41–S48) 
    Total 747 (30) 432 (13) 100.00 2.33 (1.49–3.64) 0.23 15 120.27 0.0002  
  Z−2 No diabetic retinopathy vs. NPDR Type 1 22 (36) 23 (35) 22.71 1.03 (0.50–2.13) 0.37 0.00 0.9400 (S32) 
   Type 2 137 (38) 55 (25) 77.29 1.87 (1.29–2.71) 0.19 0.20 0.0010 (S48) (S42) 
    Total 159 (38) 78 (28) 100.00 1.64 (1.14–2.35) 0.18 2.26 0.0075  
  Z−2 No diabetic retinopathy vs. PDR Type 1 41 (24) 35 (23) 24.74 1.18 (0.69–2.01) 0.27 0.31 0.5363 (S32) (S40) 
   Type 2 173 (30) 85 (19) 75.26 1.64 (1.21–2.22) 0.16 2.50 0.0016 (S48) (S41–S43) 
    Total 214 (29) 120 (20) 100.00 1.51 (1.16–1.97) 0.14 3.89 0.0023  
  Z−2 NPDR vs. PDR Type 1 19 (18) 12 (14) 18.52 1.39 (0.64–3.04) 0.40 0.00 0.4076 (S40) 
    Type 2 137 (38) 55 (25) 81.48 1.87 (1.29–2.71) 0.19 0.20 0.0010 (S48) (S42) 
    Total 156 (34) 67 (22) 100.00 1.77 (1.26–2.48) 0.17 0.64 0.0009  
  z+2 No diabetic retinopathy vs. any diabetic retinopathy Type 1 153 (16) 224 (23) 40.68 0.58 (0.36–0.93) 0.24 18.16 0.0243 (S30) (S32) (S37–S40) 
    Type 2 331 (22) 521 (21) 59.32 0.97 (0.57–1.65) 0.27 64.22 0.9038 (S36) (S41–S48) 
    Total 484 (20) 745 (22) 100.00 0.79 (0.53–1.17) 0.20 15 104.32 0.2388  
  z+2 No diabetic retinopathy vs. NPDR Type 1 7 (12) 13 (20) 22.07 0.52 (0.19–1.40) 0.51 0.00 0.1955 (S32) 
   Type 2 34 (10) 29 (14) 77.93 0.71 (0.42–1.20) 0.27 0.07 0.2001 (S48) (S42) 
    Total 41 (10) 42 (15) 100.00 0.66 (0.41–1.05) 0.24 0.36 0.0820  
  z+2 No diabetic retinopathy vs. PDR Type 1 48 (16) 70 (18) 38.34 0.84 (0.50–1.40) 0.26 1.26 0.4971 (S39) (S32) 
   Type 2 82 (14) 98 (22) 61.66 0.62 (0.39–0.99) 0.24 4.71 0.0464 (S41–S43) (S48) 
    Total 130 (15) 168 (20) 100.00 0.71 (0.51–0.98) 0.17 6.92 0.0383  
  z+2 NPDR vs. PDR Type 1 7 (12) 13 (20) 22.07 0.52 (0.19–1.40) 0.51 0.00 0.1955 (S32) 
    Type 2 34 (10) 29 (14) 77.93 0.71 (0.42–1.20) 0.27 0.07 0.2001 (S48) (S42) 
    Total 41 (10) 42 (15) 100.00 0.66 (0.41–1.05) 0.24 0.36 0.0820  

All variants have been examined by a minimum of five studies. Subanalyses have been performed for NPDR and PDR if data were available. References are preceded with S as they correlate with the online supplementary appendix.

TABLE 2

Meta-analysis of genetic variants and risk for diabetic retinopathy in patients with type 1 diabetes, type 2 diabetes, and both types of diabetes, for studies of white populations

GeneVariantRisk alleleComparisonType of diabetesTotal case subjects with risk allele (%)Total control subjects with risk allele (%)WeightOR (95% CI)SENumber of studiesχ2PReferences
ACE INS/DEL at intron 16 287 base pair deletion No diabetic retinopathy vs. any diabetic retinopathy Type 1 651 (55) 458 (52) 50.735 1.06 (0.88–1.27) 0.09 3.67 0.5402 (S1–S6) 
    Type 2 504 (53) 501 (54) 49.265 0.99 (0.83–1.19) 0.09 1.43 0.9182 (S7) (S8) (S10) (S3) 
    Total 1,155 (54) 959 (53) 100 1.02 (0.9–1.16) 0.07 10 5.35 0.7157 
   No diabetic retinopathy vs. PDR Type 1 387 (60) 220 (55) 76.8488 1.19 (0.77–1.86) 0.23 8.01 0.4355 (S1) (S2) (S4) (S5) 
   Type 2 75 (53) 78 (49) 23.1512 1.14 (0.73–1.79) 0.23 0.00 0.5617 (S10) 
    Total 462 (59) 298 (53) 100 1.19 (0.86–1.66) 0.17 8.11 0.2995  
   NPDR vs. PDR Type 1 120 (64) 72 (45) 33.9165 2.21 (1.44–3.39) 0.22 0.00 0.0003 (S4) 
    Type 2 156 (54) 224 (50) 66.0835 1.23 (0.9–1.67) 0.16 0.83 0.1905 (S10) (S14) 
    Total 276 (58) 296 (49) 100 1.52 (1–2.32) 0.22 5.55 0.0509  
NOS3 rs3138808 393 base pair insertion No diabetic retinopathy vs. any diabetic retinopathy Type 1 21 (11) 42 (22) 47.0284 0.43 (0.24–0.76) 0.29 0.00 0.0036 (S16) 
    Type 2 123 (22) 58 (21) 52.9716 1.09 (0.77–1.55) 0.18 0.00 0.6248 (S21) 
    Total 144 (19) 100 (21) 100 0.71 (0.28–1.75) 0.46 7.49 0.4512  
VEGF rs2010963 No diabetic retinopathy vs. any diabetic retinopathy Type 2 387 (55) 247 (62) 100 0.83 (0.65–1.07) 0.13 0.43 0.1601 (S27) (S28) 
AKR1B1 rs759853 No diabetic retinopathy vs. any diabetic retinopathy Type 1 99 (29) 121 (46) 45.1355 0.5 (0.35–0.71) 0.18 0.01 0.0001 (S30) (S31) 
    Type 2 324 (42) 202 (41) 54.8645 1.05 (0.83–1.33) 0.12 0.91 0.6756 (S33) 
    Total 423 (38) 323 (43) 100 0.76 (0.5–1.16) 0.22 12.83 0.2061  
AKR1B1 (CA)n dinucleotide repeat No diabetic retinopathy vs. any diabetic retinopathy Type 1 277 (42) 247 (39) 75.2466 1.11 (0.88–1.39) 0.117 0.88 0.3785 (S30) (S37) (S39) 
    Type 2 79 (32) 67 (42) 24.7534 0.66 (0.44–1.00) 0.21 0.00 0.05 (S42) 
    Total 356 (39) 314 (40) 100 0.97 (0.73–1.27) 0.14 5.47 0.8076  
  No diabetic retinopathy vs. PDR Type 1 107 (43) 127 (38) 51.9056 1.24 (0.89–1.74) 0.17 0.00 0.2008 (S39) 
   Type 2 79 (32) 67 (42) 48.0944 0.66 (0.44–1.00) 0.21 0.00 0.05 (S42) 
    Total 186 (38) 194 (39) 100 0.92 (0.5–1.70) 0.314 5.42 0.7872  
  z−2 No diabetic retinopathy vs. any diabetic retinopathy Type 1 226 (34) 167 (27) 74.1019 1.83 (0.85–3.95) 0.393 15.42 0.1238 (S30) (S37) (S39) 
    Type 2 97 (40) 43 (27) 25.8981 1.78 (1.15–2.74) 0.22 0.00 0.0091 (S42) 
    Total 323 (36) 210 (27) 100 1.80 (1.06–3.06) 0.271 15.89 0.0301  
  z−2 No diabetic retinopathy vs. PDR Type 1 87 (35) 121 (36) 52.3594 0.96 (0.68–1.35) 0.175 0.00 0.8023 (S39) 
   Type 2 97 (40) 43 (27) 47.6406 1.78 (1.15–2.74) 0.22 0.00 0.0091 (S42) 
    Total 184 (37) 164 (33) 100 1.29 (0.7–2.36) 0.309 4.85 0.4166  
  z+2 No diabetic retinopathy vs. any diabetic retinopathy Type 1 103 (16) 159 (25) 76.3483 0.51 (0.25–1.06) 0.372 12.85 0.0722 (S39) 
    Type 2 28 (12) 24 (15) 23.6517 0.73 (0.41–1.31) 0.299 0.00 0.2958 (S42) 
   Total 131 (15) 183 (23) 100 0.56 (0.32–0.97) 0.282 13.71 0.039  
  z+2 No diabetic retinopathy vs. PDR Type 1 41 (17) 57 (17) 63.9459 0.97 (0.62–1.50) 0.224 0.00 0.8814 (S30) (S37) (S39) 
   Type 2 28 (12) 24 (15) 36.0541 0.73 (0.41–1.31) 0.299 0.00 0.2958 (S42) 
    Total 69 (14) 81 (17) 100 0.87 (0.62–1.24) 0.179 0.56 0.455  
GeneVariantRisk alleleComparisonType of diabetesTotal case subjects with risk allele (%)Total control subjects with risk allele (%)WeightOR (95% CI)SENumber of studiesχ2PReferences
ACE INS/DEL at intron 16 287 base pair deletion No diabetic retinopathy vs. any diabetic retinopathy Type 1 651 (55) 458 (52) 50.735 1.06 (0.88–1.27) 0.09 3.67 0.5402 (S1–S6) 
    Type 2 504 (53) 501 (54) 49.265 0.99 (0.83–1.19) 0.09 1.43 0.9182 (S7) (S8) (S10) (S3) 
    Total 1,155 (54) 959 (53) 100 1.02 (0.9–1.16) 0.07 10 5.35 0.7157 
   No diabetic retinopathy vs. PDR Type 1 387 (60) 220 (55) 76.8488 1.19 (0.77–1.86) 0.23 8.01 0.4355 (S1) (S2) (S4) (S5) 
   Type 2 75 (53) 78 (49) 23.1512 1.14 (0.73–1.79) 0.23 0.00 0.5617 (S10) 
    Total 462 (59) 298 (53) 100 1.19 (0.86–1.66) 0.17 8.11 0.2995  
   NPDR vs. PDR Type 1 120 (64) 72 (45) 33.9165 2.21 (1.44–3.39) 0.22 0.00 0.0003 (S4) 
    Type 2 156 (54) 224 (50) 66.0835 1.23 (0.9–1.67) 0.16 0.83 0.1905 (S10) (S14) 
    Total 276 (58) 296 (49) 100 1.52 (1–2.32) 0.22 5.55 0.0509  
NOS3 rs3138808 393 base pair insertion No diabetic retinopathy vs. any diabetic retinopathy Type 1 21 (11) 42 (22) 47.0284 0.43 (0.24–0.76) 0.29 0.00 0.0036 (S16) 
    Type 2 123 (22) 58 (21) 52.9716 1.09 (0.77–1.55) 0.18 0.00 0.6248 (S21) 
    Total 144 (19) 100 (21) 100 0.71 (0.28–1.75) 0.46 7.49 0.4512  
VEGF rs2010963 No diabetic retinopathy vs. any diabetic retinopathy Type 2 387 (55) 247 (62) 100 0.83 (0.65–1.07) 0.13 0.43 0.1601 (S27) (S28) 
AKR1B1 rs759853 No diabetic retinopathy vs. any diabetic retinopathy Type 1 99 (29) 121 (46) 45.1355 0.5 (0.35–0.71) 0.18 0.01 0.0001 (S30) (S31) 
    Type 2 324 (42) 202 (41) 54.8645 1.05 (0.83–1.33) 0.12 0.91 0.6756 (S33) 
    Total 423 (38) 323 (43) 100 0.76 (0.5–1.16) 0.22 12.83 0.2061  
AKR1B1 (CA)n dinucleotide repeat No diabetic retinopathy vs. any diabetic retinopathy Type 1 277 (42) 247 (39) 75.2466 1.11 (0.88–1.39) 0.117 0.88 0.3785 (S30) (S37) (S39) 
    Type 2 79 (32) 67 (42) 24.7534 0.66 (0.44–1.00) 0.21 0.00 0.05 (S42) 
    Total 356 (39) 314 (40) 100 0.97 (0.73–1.27) 0.14 5.47 0.8076  
  No diabetic retinopathy vs. PDR Type 1 107 (43) 127 (38) 51.9056 1.24 (0.89–1.74) 0.17 0.00 0.2008 (S39) 
   Type 2 79 (32) 67 (42) 48.0944 0.66 (0.44–1.00) 0.21 0.00 0.05 (S42) 
    Total 186 (38) 194 (39) 100 0.92 (0.5–1.70) 0.314 5.42 0.7872  
  z−2 No diabetic retinopathy vs. any diabetic retinopathy Type 1 226 (34) 167 (27) 74.1019 1.83 (0.85–3.95) 0.393 15.42 0.1238 (S30) (S37) (S39) 
    Type 2 97 (40) 43 (27) 25.8981 1.78 (1.15–2.74) 0.22 0.00 0.0091 (S42) 
    Total 323 (36) 210 (27) 100 1.80 (1.06–3.06) 0.271 15.89 0.0301  
  z−2 No diabetic retinopathy vs. PDR Type 1 87 (35) 121 (36) 52.3594 0.96 (0.68–1.35) 0.175 0.00 0.8023 (S39) 
   Type 2 97 (40) 43 (27) 47.6406 1.78 (1.15–2.74) 0.22 0.00 0.0091 (S42) 
    Total 184 (37) 164 (33) 100 1.29 (0.7–2.36) 0.309 4.85 0.4166  
  z+2 No diabetic retinopathy vs. any diabetic retinopathy Type 1 103 (16) 159 (25) 76.3483 0.51 (0.25–1.06) 0.372 12.85 0.0722 (S39) 
    Type 2 28 (12) 24 (15) 23.6517 0.73 (0.41–1.31) 0.299 0.00 0.2958 (S42) 
   Total 131 (15) 183 (23) 100 0.56 (0.32–0.97) 0.282 13.71 0.039  
  z+2 No diabetic retinopathy vs. PDR Type 1 41 (17) 57 (17) 63.9459 0.97 (0.62–1.50) 0.224 0.00 0.8814 (S30) (S37) (S39) 
   Type 2 28 (12) 24 (15) 36.0541 0.73 (0.41–1.31) 0.299 0.00 0.2958 (S42) 
    Total 69 (14) 81 (17) 100 0.87 (0.62–1.24) 0.179 0.56 0.455  

Subanalyses were performed for NPDR and PDR if data were available. References are preceded with S as they correlate with the online supplementary appendix.

Data on the remaining 27 SNPs analyzed in the current study, which have been examined by a minimum of two and maximum of five cohorts, are presented in Table 3. The rs2910964 SNP of the α2β1 integrin (ITGA2) gene (OR 1.65 [95% CI 1.26–2.15], P = 2 × 10−4) and rs13306430 of the intercellular cell adhesion molecule 1 (ICAM1) gene (0.56 [0.39–0.81], P = 1.70 × 10−3) were significantly associated with diabetic retinopathy, both being examined in type 2 diabetes only and by two studies in each case.

TABLE 3

Pooled ORs for 27 SNPs and risk for diabetic retinopathy in type 1 diabetes, type 2 diabetes, and both types of diabetes

GeneVariantRisk alleleComparisonType of diabetesTotal case subjects with risk allele (%)Total control subjects with risk allele (%)WeightOR (95% CI)SENumber of studiesχ2PReferences
AGTR1 rs5186 No diabetic retinopathy vs. any diabetic retinopathy Type 1 98 (29) 200 (33) 73.99 0.82 (0.61–1.10) 0.15 1.87 0.1852 (S6; S12) 
    Type 2 22 (4) 46 (5) 34.26 0.73 (0.43–1.22) 0.26 0.00 0.2251 (S12) 
    Total 142 (14) 246 (15) 100.00 0.8 (0.62–1.03) 0.13 2.05 0.0788  
ITGA2 rs2910964 No diabetic retinopathy vs. any diabetic retinopathy Type 2 258 (46) 137 (34) 100.00 1.65 (1.26–2.15) 0.14 0.05 0.0002 (S49; S50) 
ADRB3 rs4994 No diabetic retinopathy vs. any diabetic retinopathy Type 1 62 (7) 12 (10) 48.01 0.73 (0.38–1.39) 0.33 0.00 0.3320 (S51; S52) 
    Type 2 49 (27) 34 (15) 51.99 2.16 (1.32–3.51) 0.25 0.00 0.0020 (S51; S52) 
    Total 111 (11) 46 (13) 100.00 1.28 (0.44–3.72) 0.54 6.95 0.6516  
   No diabetic retinopathy vs. NPDR Type 1 27 (7) 12 (10) 47.68 0.71 (0.35–1.44) 0.36 0.00 0.3420 (S51; S52) 
   Type 2 21 (22) 34 (15) 52.32 1.71 (0.94–3.14) 0.31 0.00 0.0811 (S51; S52) 
    Total 48 (10) 46 (13) 100.00 1.12 (0.47–2.67) 0.44 3.44 0.7900  
   No diabetic retinopathy vs. PDR Type 1 35 (7) 12 (10) 48.89 0.74 (0.37–1.47) 0.35 0.00 0.3892 (S51; S52) 
    Type 2 28 (31) 34 (15) 51.11 2.68 (1.51–4.75) 0.29 0.00 0.0008 (S51; S52) 
    Total 63 (11) 46 (13) 100.00 1.43 (0.41–5.03) 0.64 7.95 0.5797  
   NPDR vs. PDR Type 1 35 (7) 27 (7) 61.48 1.04 (0.62–1.76) 0.27 0.00 0.8706 (S51; S52) 
    Type 2 28 (31) 21 (22) 38.52 1.56 (0.81–3.01) 0.34 0.00 0.1829 (S51; S52) 
    Total 63 (11) 48 (10) 100.00 1.22 (0.81–1.83) 0.21 0.89 0.3399  
AGT rs4762 No diabetic retinopathy vs. any diabetic retinopathy Type 1 111 (55) 189 (63) 37.28 0.73 (0.51–1.05) 0.18 0.00 0.0875 (S6; S12) 
    Type 2 90 (14) 161 (17) 62.72 0.84 (0.63–1.11) 0.14 0.00 0.2102 (S6; S12) 
    Total 201 (24) 350 (27) 100.00 0.79 (0.64–0.99) 0.11 0.35 0.0418  
APOE ϵ2/ϵ3/ϵ4 ϵ2 No diabetic retinopathy vs. any diabetic retinopathy Type 1 21 (14) 19 (10) 19.55 1.46 (0.75–2.83) 0.34 0.00 0.2620 (S53; S54) 
    Type 2 223 (10) 51 (11) 80.45 1.05 (0.76–1.45) 0.17 0.05 0.7681 (S53; S54) 
    Total 244 (11) 70 (11) 100.00 1.12 (0.84–1.5) 0.15 0.81 0.4470  
FGF2 rs41456044 No diabetic retinopathy vs. PDR Type 2 31 (5) 47 (5) 100.00 1.19 (0.74–1.91) 0.24 0.27 0.4672 (S55; S56) 
FGF2 rs308395 No diabetic retinopathy vs. PDR Type 2 90 (14) 131 (13) 100.00 1.08 (0.81–1.45) 0.15 0.29 0.5882 (S55; S56) 
NOS3 rs1799983 No diabetic retinopathy vs. any diabetic retinopathy Type 2 393 (30) 478 (24) 100.00 1.11 (0.94–1.31) 0.08 1.69 0.2279 (S17-S20) 
NOS3 rs41322052 No diabetic retinopathy vs. any diabetic retinopathy Type 1 105 (42) 106 (42) 25.37 1.01 (0.71–1.44) 0.18 0.00 0.9525 (S57) 
    Type 2 313 (27) 334 (24) 74.63 1.04 (0.75–1.45) 0.17 4.33 0.7956 (S17; S19) 
    Total 418 (29) 440 (27) 100.00 1.06 (0.85–1.33) 0.11 4.77 0.5962  
   No diabetic retinopathy vs. PDR Type 1 105 (42) 106 (42) 77.27 1.01 (0.71–1.44) 0.18 0.00 0.9525 (S57) 
   Type 2 16 (12) 29 (14) 22.73 0.83 (0.43–1.59) 0.33 0.00 0.5686 (S17; S19) 
    Total 121 (31) 135 (29) 100.00 0.97 (0.71–1.32) 0.16 0.28 0.8263  
SLC2A1 rs841853 XbaI- No diabetic retinopathy vs. any diabetic retinopathy Type 1 227 (35) 142 (64) 66.08 0.41 (0.13–1.23) 0.57 8.27 0.1118 (S58; S59) 
HLA DR1–8 DR1 No diabetic retinopathy vs. any diabetic retinopathy Type 1 45 (14) 40 (15) 100.00 0.99 (0.62–1.56) 0.2359 0.2 0.9506 (S60; S61) 
HLA DR1–8 DR7 No diabetic retinopathy vs. any diabetic retinopathy Type 1 39 (12) 38 (14) 100.00 0.59 (0.10–3.60) 0.9247 1.86 0.5658 (S60; S61) 
ICAM1 rs13306430 No diabetic retinopathy vs. any diabetic retinopathy Type 2 150 (36) 89 (50) 100.00 0.56 (0.39–0.81) 0.18 0.01 0.0017 (S62; S63) 
MTHFR rs1801133 No diabetic retinopathy vs. any diabetic retinopathy Type 2 318 (44) 405 (37) 100.00 1.39 (0.99–1.94) 0.17 8.34 0.0563 (S35; S64–S66) 
NPY rs16139 No diabetic retinopathy vs. any diabetic retinopathy Type 2 26 (9) 16 (5) 100.00 2.62 (0.9–7.61) 0.54 1.38 0.0759 (S67; S68) 
PAI-1 rs1799768 5G No diabetic retinopathy vs. any diabetic retinopathy Type 2 355 (44) 252 (45) 100.00 1.06 (0.85–1.32) 0.11 0.00 0.6017 (S10; S35; S69) 
PON1 rs662 No diabetic retinopathy vs. any diabetic retinopathy Type 2 310 (54) 175 (48) 100.00 0.99 (0.44–2.27) 0.42 8.48 0.9896 (S69; S70) 
PON2 rs7493 No diabetic retinopathy vs. any diabetic retinopathy Type 1 192 (57) 191 (49) 54.02 1.37 (1.03–1.84) 0.15 0.00 0.0322 (S70; S71) 
    Type 2 144 (72) 144 (78) 45.98 0.72 (0.46–1.15) 0.23 0.00 0.1685 (S70; S71) 
    Total 336 (62) 335 (58) 100.00 1.02 (0.55–1.91) 0.32 5.33 0.9422  
PPARG rs1801282 No diabetic retinopathy vs. any diabetic retinopathy Type 1 33 (13) 108 (14) 29.35 0.85 (0.56–1.29) 0.21 0.00 0.4413 (S72-S75) 
    Type 2 124 (14) 273 (16) 70.65 0.83 (0.62–1.11) 0.15 3.33 0.2135 (S72-S75) 
    Total 157 (14) 381 (16) 100.00 0.83 (0.67–1.05) 0.12 3.34 0.1173  
AGER rs1800624 No diabetic retinopathy vs. any diabetic retinopathy Type 2 361 (33) 112 (18) 100.00 1.89 (0.53–6.75) 0.65 46.09 0.3271 (S76-S78) 
   No diabetic retinopathy vs. NPDR Type 2 217 (37) 60 (12) 100.00 2.94 (0.46–18.76) 0.95 26.55 0.2540 (S76; S78) 
   No diabetic retinopathy vs. PDR Type 2 67 (26) 93 (20) 100.00 0.93 (0.62–1.38) 0.20 0.88 0.7055  
AGER rs1800625 No diabetic retinopathy vs. any diabetic retinopathy Type 2 125 (13) 94 (15) 100.00 0.94 (0.7–1.27) 0.15 0.32 0.7035 (S76; S78) 
   No diabetic retinopathy vs. PDR Type 2 35 (14) 76 (15) 100.00 0.99 (0.64–1.55) 0.23 0.99 0.9803 (S76; S78) 
VDR rs10735810 No diabetic retinopathy vs. any diabetic retinopathy Type 1 207 (40) 86 (34) 57.59 1.27 (0.93–1.74) 0.16 0.00 0.1282 (S79) 
    Type 2 85 (50) 169 (47) 42.41 1.15 (0.8–1.66) 0.19 0.00 0.4415 (S80) 
    Total 292 (42) 255 (42) 100.00 1.22 (0.96–1.55) 0.12 0.17 0.0978  
VEGF rs25648 No diabetic retinopathy vs. any diabetic retinopathy Type 2 86 (16) 56 (14) 100.00 1.48 (0.42–5.19) 0.64 9.06 0.5363 (S22; S24) 
VEGF rs1570360 No diabetic retinopathy vs. any diabetic retinopathy Type 2 78 (12) 81 (13) 100.00 0.93 (0.67–1.3) 0.17 0.22 0.6740 (S24; S25) 
VEGF rs3095039 No diabetic retinopathy vs. any diabetic retinopathy Type 2 86 (16) 52 (13) 100.00 1.1 (0.47–2.62) 0.44 4.47 0.8230 (S24; S29) 
   No diabetic retinopathy vs. NPDR Type 2 45 (14) 52 (13) 100.00 0.97 (0.33–2.85) 0.55 5.03 0.9560 (S24; S29) 
   No diabetic retinopathy vs. PDR Type 2 41 (18) 52 (13) 100.00 1.44 (0.88–2.36) 0.25 1.13 0.1491 (S24; S29) 
   NPDR vs. PDR Type 2 41 (18) 45 (14) 100.00 1.22 (0.76–1.96) 0.24 1.01 0.4146 (S24; S29) 
VEGF rs35569394 -2549DEL No diabetic retinopathy vs. any diabetic retinopathy Type 1 73 (58) 68 (52) 46.01 1.25 (0.77–2.04) 0.25 0.00 0.3724 (S81) 
    Type 2 252 (65) 81 (45) 53.99 2.28 (1.59–3.26) 0.18 0.00 1.0 × 10.5 (S82) 
    Total 325 (63) 149 (48) 100.00 1.73 (0.96–3.11) 0.30 3.77 0.0677  
VEGF rs699947 No diabetic retinopathy vs. any diabetic retinopathy Type 2 204 (29) 280 (29) 100.00 1.01 (0.56–1.83) 0.30 7.61 0.9741 (S25; S26) 
GeneVariantRisk alleleComparisonType of diabetesTotal case subjects with risk allele (%)Total control subjects with risk allele (%)WeightOR (95% CI)SENumber of studiesχ2PReferences
AGTR1 rs5186 No diabetic retinopathy vs. any diabetic retinopathy Type 1 98 (29) 200 (33) 73.99 0.82 (0.61–1.10) 0.15 1.87 0.1852 (S6; S12) 
    Type 2 22 (4) 46 (5) 34.26 0.73 (0.43–1.22) 0.26 0.00 0.2251 (S12) 
    Total 142 (14) 246 (15) 100.00 0.8 (0.62–1.03) 0.13 2.05 0.0788  
ITGA2 rs2910964 No diabetic retinopathy vs. any diabetic retinopathy Type 2 258 (46) 137 (34) 100.00 1.65 (1.26–2.15) 0.14 0.05 0.0002 (S49; S50) 
ADRB3 rs4994 No diabetic retinopathy vs. any diabetic retinopathy Type 1 62 (7) 12 (10) 48.01 0.73 (0.38–1.39) 0.33 0.00 0.3320 (S51; S52) 
    Type 2 49 (27) 34 (15) 51.99 2.16 (1.32–3.51) 0.25 0.00 0.0020 (S51; S52) 
    Total 111 (11) 46 (13) 100.00 1.28 (0.44–3.72) 0.54 6.95 0.6516  
   No diabetic retinopathy vs. NPDR Type 1 27 (7) 12 (10) 47.68 0.71 (0.35–1.44) 0.36 0.00 0.3420 (S51; S52) 
   Type 2 21 (22) 34 (15) 52.32 1.71 (0.94–3.14) 0.31 0.00 0.0811 (S51; S52) 
    Total 48 (10) 46 (13) 100.00 1.12 (0.47–2.67) 0.44 3.44 0.7900  
   No diabetic retinopathy vs. PDR Type 1 35 (7) 12 (10) 48.89 0.74 (0.37–1.47) 0.35 0.00 0.3892 (S51; S52) 
    Type 2 28 (31) 34 (15) 51.11 2.68 (1.51–4.75) 0.29 0.00 0.0008 (S51; S52) 
    Total 63 (11) 46 (13) 100.00 1.43 (0.41–5.03) 0.64 7.95 0.5797  
   NPDR vs. PDR Type 1 35 (7) 27 (7) 61.48 1.04 (0.62–1.76) 0.27 0.00 0.8706 (S51; S52) 
    Type 2 28 (31) 21 (22) 38.52 1.56 (0.81–3.01) 0.34 0.00 0.1829 (S51; S52) 
    Total 63 (11) 48 (10) 100.00 1.22 (0.81–1.83) 0.21 0.89 0.3399  
AGT rs4762 No diabetic retinopathy vs. any diabetic retinopathy Type 1 111 (55) 189 (63) 37.28 0.73 (0.51–1.05) 0.18 0.00 0.0875 (S6; S12) 
    Type 2 90 (14) 161 (17) 62.72 0.84 (0.63–1.11) 0.14 0.00 0.2102 (S6; S12) 
    Total 201 (24) 350 (27) 100.00 0.79 (0.64–0.99) 0.11 0.35 0.0418  
APOE ϵ2/ϵ3/ϵ4 ϵ2 No diabetic retinopathy vs. any diabetic retinopathy Type 1 21 (14) 19 (10) 19.55 1.46 (0.75–2.83) 0.34 0.00 0.2620 (S53; S54) 
    Type 2 223 (10) 51 (11) 80.45 1.05 (0.76–1.45) 0.17 0.05 0.7681 (S53; S54) 
    Total 244 (11) 70 (11) 100.00 1.12 (0.84–1.5) 0.15 0.81 0.4470  
FGF2 rs41456044 No diabetic retinopathy vs. PDR Type 2 31 (5) 47 (5) 100.00 1.19 (0.74–1.91) 0.24 0.27 0.4672 (S55; S56) 
FGF2 rs308395 No diabetic retinopathy vs. PDR Type 2 90 (14) 131 (13) 100.00 1.08 (0.81–1.45) 0.15 0.29 0.5882 (S55; S56) 
NOS3 rs1799983 No diabetic retinopathy vs. any diabetic retinopathy Type 2 393 (30) 478 (24) 100.00 1.11 (0.94–1.31) 0.08 1.69 0.2279 (S17-S20) 
NOS3 rs41322052 No diabetic retinopathy vs. any diabetic retinopathy Type 1 105 (42) 106 (42) 25.37 1.01 (0.71–1.44) 0.18 0.00 0.9525 (S57) 
    Type 2 313 (27) 334 (24) 74.63 1.04 (0.75–1.45) 0.17 4.33 0.7956 (S17; S19) 
    Total 418 (29) 440 (27) 100.00 1.06 (0.85–1.33) 0.11 4.77 0.5962  
   No diabetic retinopathy vs. PDR Type 1 105 (42) 106 (42) 77.27 1.01 (0.71–1.44) 0.18 0.00 0.9525 (S57) 
   Type 2 16 (12) 29 (14) 22.73 0.83 (0.43–1.59) 0.33 0.00 0.5686 (S17; S19) 
    Total 121 (31) 135 (29) 100.00 0.97 (0.71–1.32) 0.16 0.28 0.8263  
SLC2A1 rs841853 XbaI- No diabetic retinopathy vs. any diabetic retinopathy Type 1 227 (35) 142 (64) 66.08 0.41 (0.13–1.23) 0.57 8.27 0.1118 (S58; S59) 
HLA DR1–8 DR1 No diabetic retinopathy vs. any diabetic retinopathy Type 1 45 (14) 40 (15) 100.00 0.99 (0.62–1.56) 0.2359 0.2 0.9506 (S60; S61) 
HLA DR1–8 DR7 No diabetic retinopathy vs. any diabetic retinopathy Type 1 39 (12) 38 (14) 100.00 0.59 (0.10–3.60) 0.9247 1.86 0.5658 (S60; S61) 
ICAM1 rs13306430 No diabetic retinopathy vs. any diabetic retinopathy Type 2 150 (36) 89 (50) 100.00 0.56 (0.39–0.81) 0.18 0.01 0.0017 (S62; S63) 
MTHFR rs1801133 No diabetic retinopathy vs. any diabetic retinopathy Type 2 318 (44) 405 (37) 100.00 1.39 (0.99–1.94) 0.17 8.34 0.0563 (S35; S64–S66) 
NPY rs16139 No diabetic retinopathy vs. any diabetic retinopathy Type 2 26 (9) 16 (5) 100.00 2.62 (0.9–7.61) 0.54 1.38 0.0759 (S67; S68) 
PAI-1 rs1799768 5G No diabetic retinopathy vs. any diabetic retinopathy Type 2 355 (44) 252 (45) 100.00 1.06 (0.85–1.32) 0.11 0.00 0.6017 (S10; S35; S69) 
PON1 rs662 No diabetic retinopathy vs. any diabetic retinopathy Type 2 310 (54) 175 (48) 100.00 0.99 (0.44–2.27) 0.42 8.48 0.9896 (S69; S70) 
PON2 rs7493 No diabetic retinopathy vs. any diabetic retinopathy Type 1 192 (57) 191 (49) 54.02 1.37 (1.03–1.84) 0.15 0.00 0.0322 (S70; S71) 
    Type 2 144 (72) 144 (78) 45.98 0.72 (0.46–1.15) 0.23 0.00 0.1685 (S70; S71) 
    Total 336 (62) 335 (58) 100.00 1.02 (0.55–1.91) 0.32 5.33 0.9422  
PPARG rs1801282 No diabetic retinopathy vs. any diabetic retinopathy Type 1 33 (13) 108 (14) 29.35 0.85 (0.56–1.29) 0.21 0.00 0.4413 (S72-S75) 
    Type 2 124 (14) 273 (16) 70.65 0.83 (0.62–1.11) 0.15 3.33 0.2135 (S72-S75) 
    Total 157 (14) 381 (16) 100.00 0.83 (0.67–1.05) 0.12 3.34 0.1173  
AGER rs1800624 No diabetic retinopathy vs. any diabetic retinopathy Type 2 361 (33) 112 (18) 100.00 1.89 (0.53–6.75) 0.65 46.09 0.3271 (S76-S78) 
   No diabetic retinopathy vs. NPDR Type 2 217 (37) 60 (12) 100.00 2.94 (0.46–18.76) 0.95 26.55 0.2540 (S76; S78) 
   No diabetic retinopathy vs. PDR Type 2 67 (26) 93 (20) 100.00 0.93 (0.62–1.38) 0.20 0.88 0.7055  
AGER rs1800625 No diabetic retinopathy vs. any diabetic retinopathy Type 2 125 (13) 94 (15) 100.00 0.94 (0.7–1.27) 0.15 0.32 0.7035 (S76; S78) 
   No diabetic retinopathy vs. PDR Type 2 35 (14) 76 (15) 100.00 0.99 (0.64–1.55) 0.23 0.99 0.9803 (S76; S78) 
VDR rs10735810 No diabetic retinopathy vs. any diabetic retinopathy Type 1 207 (40) 86 (34) 57.59 1.27 (0.93–1.74) 0.16 0.00 0.1282 (S79) 
    Type 2 85 (50) 169 (47) 42.41 1.15 (0.8–1.66) 0.19 0.00 0.4415 (S80) 
    Total 292 (42) 255 (42) 100.00 1.22 (0.96–1.55) 0.12 0.17 0.0978  
VEGF rs25648 No diabetic retinopathy vs. any diabetic retinopathy Type 2 86 (16) 56 (14) 100.00 1.48 (0.42–5.19) 0.64 9.06 0.5363 (S22; S24) 
VEGF rs1570360 No diabetic retinopathy vs. any diabetic retinopathy Type 2 78 (12) 81 (13) 100.00 0.93 (0.67–1.3) 0.17 0.22 0.6740 (S24; S25) 
VEGF rs3095039 No diabetic retinopathy vs. any diabetic retinopathy Type 2 86 (16) 52 (13) 100.00 1.1 (0.47–2.62) 0.44 4.47 0.8230 (S24; S29) 
   No diabetic retinopathy vs. NPDR Type 2 45 (14) 52 (13) 100.00 0.97 (0.33–2.85) 0.55 5.03 0.9560 (S24; S29) 
   No diabetic retinopathy vs. PDR Type 2 41 (18) 52 (13) 100.00 1.44 (0.88–2.36) 0.25 1.13 0.1491 (S24; S29) 
   NPDR vs. PDR Type 2 41 (18) 45 (14) 100.00 1.22 (0.76–1.96) 0.24 1.01 0.4146 (S24; S29) 
VEGF rs35569394 -2549DEL No diabetic retinopathy vs. any diabetic retinopathy Type 1 73 (58) 68 (52) 46.01 1.25 (0.77–2.04) 0.25 0.00 0.3724 (S81) 
    Type 2 252 (65) 81 (45) 53.99 2.28 (1.59–3.26) 0.18 0.00 1.0 × 10.5 (S82) 
    Total 325 (63) 149 (48) 100.00 1.73 (0.96–3.11) 0.30 3.77 0.0677  
VEGF rs699947 No diabetic retinopathy vs. any diabetic retinopathy Type 2 204 (29) 280 (29) 100.00 1.01 (0.56–1.83) 0.30 7.61 0.9741 (S25; S26) 

All SNPs have been examined by a minimum of two and maximum of five studies. Subanalyses have been performed for NPDR and PDR if data were available. References are preceded with S as they correlate with the online supplementary appendix.

ACE.

The gene encoding ACE is located on chromosome 17q23. Six studies examined the insertion/deletion (INS/DEL) polymorphism in intron 16 of the ACE gene in patients with type 1 diabetes, and seven studies in patients with type 2 diabetes were included in the meta-analysis. The 287 base pair deletion was treated as the risk variant, and there was no statistically significant association with this polymorphism and the development of any form of diabetic retinopathy (Table 1).

Ten studies of subjects with white origin were subanalyzed for all diabetic retinopathy comparisons. The 287 base pair deletion was not found to be significantly associated with any diabetic retinopathy or its subtypes in type 1 or type 2 diabetes (Table 2).

Aldose reductase (AKR1B1).

The aldo-keto reductase family 1 member B1 (AKR1B1) gene (also known as ALR) is located on chromosome 7q35. Associations of two AKR1B1 SNPs with diabetic retinopathy have been reported in the literature: the promoter SNP rs759853 and the (CA)n microsatellite polymorphism located 5′ of the AKR1B1 gene. Six studies have examined the association between the AKR1B1 (CA)n microsatellite with diabetic retinopathy in type 1 and nine studies in type 2 diabetes (Table 1). The three most commonly investigated AKR1B1 variants (z, z+2, and z−2) in the literature were included for analysis.

There was a significant association with the z−2 allele and the development of any diabetic retinopathy (OR 2.33 [95% CI 1.49–3.64], P = 2 × 10−4). Subanalyses revealed a significant association between the z−2 allele and diabetic retinopathy in patients with type 2 diabetes (2.64 [1.39–5.01], P = 2.9 × 10−3), with weaker but statistically significant association also being found for patients with type 1 diabetes (1.95 [1.04–3.66], P = 0.04). A significant association was also found in the NPDR and PDR subgroups.

No statistically significant association was found between the z allele and the development of diabetic retinopathy overall (OR 1.05 [95% CI 0.81–1.35], P = 0.73). However, in the subanalysis for type of diabetic retinopathy, the z allele was significantly protective against NPDR development in type 2 diabetes (0.65 [0.45–0.94], P = 0.02). A significant difference between NPDR and PDR development for the presence of z allele was found also in type 2 diabetes (0.65 [0.45–0.94], P = 0.02). Similarly, the z+2 allele was found to be significantly protective against the development of diabetic retinopathy (0.58 [0.36–0.93], P = 0.02).

Only four studies examining the AKR1B1 (CA)n microsatellite have included subjects of white ancestry (Table 2), with the majority of studies including subjects of only Asian ancestry. In the analyses of studies of white origin, only the z−2 (OR 1.80 [95% CI 1.06–3.06], P = 0.03) and z+2 (0.56 [0.32–0.97], P = 0.04) polymorphisms remained significantly associated with diabetic retinopathy, with the z−2 allele conferring risk and z+2 conferring protection against diabetic retinopathy.

Three studies examined the association of a second AKR1B1 polymorphism (promoter SNP rs759853) with diabetic retinopathy in type 1 diabetes and five studies in type 2 diabetes (Table 1). The T allele was considered the risk variant. Interestingly, analyses revealed protection against diabetic retinopathy with the T allele in type 1 diabetes (OR 0.49 [95% CI 0.36–0.68], P < 1.00 × 10−4). There was no statistically significant association between diabetic retinopathy and rs759853 in type 2 diabetes. However, a borderline association between NPDR and PDR was found (0.73 [0.54–0.99], P = 0.04).

Four studies investigating the AKR1B1 rs759853 included subjects of white ancestry. A significant protection of the T allele against diabetic retinopathy in type 1 diabetes (OR 0.5 [95% CI 0.35–0.71], P = 1.00 × 10−4) remained. Insufficient studies were available for subanalysis of diabetic retinopathy subtypes.

Vascular endothelial growth factor.

The vascular endothelial growth factor (VEGF) gene is located on chromosome 6p12. Six VEGF polymorphisms were included in this meta-analysis, with the rs2010963 polymorphism being the most frequently studied. No studies examining this polymorphism in type 1 diabetes and diabetic retinopathy were located. Seven studies examining this polymorphism in type 2 diabetes and diabetic retinopathy were included in the analyses (Table 1). The G allele has been considered as the risk variant. A significant association between patients with no diabetic retinopathy and those with NPDR (OR 0.62 [95% CI 0.48–0.81], P = 5.0 × 10−4) was identified, yet no significant differences were found between NPDR and PDR development. Meta-analysis revealed no significant association between the VEGF polymorphisms: rs25648, rs1570360, rs3095039, rs35569394, or rs699947 and any type of diabetic retinopathy (Table 3).

Only two studies of participants were available for inclusion in the subanalysis for white ancestry (Table 2). No statistically significant association of the VEGF rs2010963 polymorphism with diabetic retinopathy was found (OR 0.83 [95% CI 0.65–1.07], P = 0.16) and insufficient studies were available for diabetic retinopathy subtype analyses.

Endothelial nitric oxide synthase.

The endothelial nitric oxide synthase (NOS3) gene is located on chromosome 7q35–36. Three NOS3 SNPs (rs1799983, rs41322052, and rs3138808) met the inclusion criteria for meta-analysis. The rs3138808 variant has been the most commonly examined NOS3 diabetic retinopathy polymorphism, and the 393 base pair insertion has been classified as the risk variant. One study examining this polymorphism and diabetic retinopathy in type 1 diabetes and seven studies in type 2 diabetes were included in the analysis (Table 1). There was no statistically significant association between rs3138808 and any form of diabetic retinopathy. Additionally, no significant association between any form of diabetic retinopathy and the rs1799983 (OR 1.11 [95% CI 0.94–1.31], P = 0.23) or rs41322052 (1.06 [0.85–1.33], P = 0.60) were identified (Table 3).

Only two studies were available to be included in the subanalysis for subjects of white ancestry (Table 2). No statistically significant association was found with the NOS3 rs3138808 polymorphism and diabetic retinopathy development (OR 0.71 [95% CI 0.28–1.75], P = 0.45) and insufficient studies were available for diabetic retinopathy subtype analysis.

Diabetic retinopathy remains one of the leading causes of blindness in the developed world (1). There is strong evidence for a genetic component in the development of diabetic retinopathy, independent of other established risk factors (13,,,17). Despite a large number of candidate gene studies on diabetic retinopathy, the genetic evidence thus far for an association with the development of or severity of diabetic retinopathy has been conflicting. This is largely accounted for by varying participant ethnicity, study design, differences in retinopathy grading scales, and in particular suboptimal power.

The current meta-analysis comprehensively assesses the risk of diabetic retinopathy in relation to every published candidate gene meeting inclusion criteria. Eighty-two studies examining 20 genes and 34 SNPs were analyzed. Pathways involved in the pathogenesis of diabetic retinopathy relevant to the genes with the most studied polymorphisms in this meta-analysis are described in detail below.

Renin-angiotensin system.

ACE is involved in the conversion of angiotensin I to angiotensin II (ATII). ATII mediates its hemodynamic effects through signaling via angiotensin type 1 (AT1) and type 2 (AT2) receptors. Vascular remodeling and proliferation occurs mainly via the AT1 receptor (23). All components of the renin-angiotensin system (RAS) have been shown to be expressed in the retina (24). The physiologic effects of ATII in the eye include the regulation of intraocular blood flow and pressure, promotion of capillary growth, enhancing vascular permeability, increasing oxidative stress, and the regulation of cell growth via the expression of various growth factors including VEGF, insulin-like growth factor, and platelet-derived growth factor (23).

Clinical evidence also supports the role for the RAS system in diabetic retinopathy pathogenesis. The EURODIAB Controlled Trial of Lisinopril in Insulin-Dependent Diabetes (EUCLID) trial provided evidence for lisinopril, an ACE inhibitor, decreasing the progression of diabetic retinopathy by 50% (25). In addition, animal studies have shown that ACE inhibitors and AT1 receptor blockers can prevent retinal neovascularization (26).

The ACE INS/DEL polymorphism in intron 16 had the largest number of subjects to be genotyped for any polymorphism and the largest number of studies to include participants of white ancestry. Three-quarters of the studies included participants of white ancestry. There was no statistically significant association of the INS/DEL polymorphism of the ACE gene with diabetic retinopathy or diabetic retinopathy subtypes in type 1, type 2, or combined diabetes. This is consistent with the findings of the majority of the included studies examining this polymorphism in this meta-analysis. Two other meta-analyses have examined the association of the INS/DEL of the ACE gene with diabetic retinopathy development. Fujisawa et al. (27) examined 12 studies, including type 1 and type 2 diabetic subjects. Wiwanitkit (28) examined four studies and included type 2 diabetic subjects only. Both analyses also found no statistically significant association with this polymorphism and the development of diabetic retinopathy.

Polyol pathway.

Aldose reductase (ALR) is a rate-limiting enzyme of the polyol pathway, which catalyzes NADPH-dependent reduction of glucose to sorbitol. This pathway leads to the intracellular accumulation of sorbitol and is primarily active under hyperglycemic conditions (29). Several mechanisms have been proposed to explain the pathogenesis of diabetic microvascular complications, including the induction of osmotic stress and the activation of protein kinase C as well as pathogenic vascular and hemodynamic alterations (8). ALR has been identified in human pericytes, which exhibit an active polyol pathway (30). Animal studies have shown that cultured mural cells from the retinal capillaries of adult rhesus monkeys undergo cellular degeneration after exposure to high glucose levels, with ALR being isolated from these mural cells (31). The sorbitol pathway is also biologically plausible as it is involved in the selective degeneration of human mural cells in NPDR (32). Induced hyperglycemia in dogs with galactosemia has similarly shown retinal vascular changes including microaneurysm formation, degeneration of retinal pericytes, retinal hemorrhages, and nonperfused or acellular vessels (33).

The AKR1B1 gene had the largest number of studies examining the relationship of its polymorphisms to diabetic retinopathy development, regardless of ethnicity. The z−2 microsatellite showed the most significant association with diabetic retinopathy, especially in type 2 diabetes, conferring risk also in PDR and NPDR subtypes. The z+2 and z microsatellite conferred protection against overall diabetic retinopathy and NPDR, respectively, and both were protective against PDR when compared with NPDR in type 2 diabetes. The majority of studies included in this meta-analysis have individually reported a risk for diabetic retinopathy with the z−2 microsatellite; however, the z microsatellite was not found to be statistically significant by most studies, with only a fifth of the studies individually reporting z+2 to be protective against diabetic retinopathy. The vast majority of studies have included participants of non-white ancestry; however, those of white ancestry found the z−2 microsatellite to confer risk and z+2 to confer protection against overall diabetic retinopathy in combined diabetes.

The T allele of the AKR1B1 promoter SNP rs759853 conferred protection against diabetic retinopathy in type 1 diabetes of any ancestry and also of white ancestry alone. This protection against diabetic retinopathy in type 1 diabetes has also been found in the individual studies examining this polymorphism.

VEGF.

VEGF is a multifunctional cytokine that promotes angiogenesis and is a potent mediator of microvascular permeability. Diabetic microvascular changes in the retina lead to hypoxia, a stimulator of VEGF production (34). VEGF has been found to have a significant role in the development of diabetic retinopathy by inducing hyperpermeability of retinal vessels, breakdown of the blood–retinal barrier, and neovascularization in PDR (35,37). Complications such as retinal edema and blinding vitreous hemorrhage arise as a result of the abnormal barrier function of vessels and the growth of new vessels, which are fragile and prone to rupture.

VEGF protein expression has been shown to be influenced by genetic variation in the VEGF gene (38). VEGF levels in the vitreous of patients with PDR are significantly elevated when compared with the vitreous of diabetic eyes without PDR and control subjects without diabetes (39,40). VEGF inhibition has been shown to result in a marked reduction in retinal neovascularization (41) and prevention of the blood–retinal barrier breakdown (42), further supporting its role in diabetic retinopathy development.

The VEGF gene had the largest number of individual SNPs examined in relation to diabetic retinopathy. Of these six polymorphisms, the G variant of the rs2010963 polymorphism was found to significantly protect against the development of NPDR in type 2 diabetes but not with overall diabetic retinopathy development. In keeping with the findings of this meta-analysis, the majority of studies reported no statistically significant association of this polymorphism with any diabetic retinopathy development in type 2 diabetes, regardless of ethnicity.

NO pathway.

Endothelial NO synthase (eNOS) is an enzyme produced by endothelial cells. NO derived from eNOS is a key endogenous vasodilator (43). It is also believed to be important in the promotion of angiogenesis and regulation of VEGF expression (44). A low concentration of eNOS is believed to be necessary to maintain endothelial function (45) with experimental deficiency of eNOS shown to significantly decrease retinal neovascularization in a mouse model (46). Aqueous NO levels have also been found to play an important role in the progression of diabetic retinopathy, with levels significantly higher in active PDR (47). Similarly, NO levels have been found to be significantly elevated in PDR vitreous when compared with nondiabetic subjects (48), making the eNOS gene (NOS3) a biologically plausible candidate for susceptibility to diabetic retinopathy development.

The NOS3 rs3138808 polymorphism was the most studied polymorphism, and the majority of studies included participants of nonwhite ancestry. No statistically significant association was found with diabetic retinopathy development regardless of ethnicity and for whites alone, consistent with the findings of the majority of the included studies. All other NOS3 polymorphisms did not show statistically significant associations with the development of diabetic retinopathy.

Data on the remaining 27 SNPs examined by a minimum of two and maximum of five studies revealed the rs2910964 SNP of the ITGA2 gene and rs13306430 of the ICAM1 gene to be significantly associated with diabetic retinopathy. Both of these variants were examined in type 2 diabetes only and by the minimum allowable two studies each. Unless further replicative studies are undertaken, the importance of these genes in diabetic retinopathy may remain unclear. Zintzaras et al. (49) performed a meta-analysis of five studies examining the link between the C677T polymorphism of the methylenetetrahydrofolate reductase (MTHFR) gene and development of diabetic retinopathy in type 2 diabetes. Unlike our results, a borderline association between C677T transition and the risk of development diabetic retinopathy (OR 1.39 [95% CI 1.05–1.83], P = 0.08) was reported.

Because of the publication deadline for included studies, some polymorphisms, such as those recently reported in the erythropoietin gene (50), were unable to be included in the meta-analysis as further studies have not as yet been reported. The definition of diabetic retinopathy requires the minimum of microaneurysm presence in a diabetic individual (34); however, several definitions and criteria exist for the subclassification of diabetic retinopathy and preferences have been variable among included studies, with many studies defining diabetic retinopathy without the use of standardized scales. Additionally, confounding factors such as glycemic control have not been adjusted for in calculation of the ORs of polymorphisms included in this meta-analysis, as this information provided by included studies have often been incomplete and nonstandardized definitions were used. Therefore, these are accepted as limitations of this meta-analysis, with some being implicit in the meta-analysis concept when dealing with a large number of studies with different design and reporting style.

In conclusion, this meta-analysis has found that sequence variation within the AKR1B1 gene are the most significantly associated with diabetic retinopathy development among those genes qualifying for inclusion. This result supports more detailed research into the role of the AKR1B1 gene in diabetic retinopathy to determine which variants are causally associated and their mechanism of action. Genetic research on diabetic retinopathy susceptibility has the potential for a direct positive impact on patient management by ultimately altering screening regimes in an individualized manner for a disease that is of major and increasing public health concern. Although much work has focused on dissecting the genetics of diabetes itself, considerably less has been conducted into the molecular mechanisms leading to its specific complications, including diabetic retinopathy and its subtypes. Although NPDR and PDR are commonly subanalyzed in genetic studies, very few studies have reported on genetic associations with the common subtype of macular edema, that is a frequent cause of visual disability in diabetic subjects. This is an area that requires further exploration. Future genetic studies should also include genome-wide association studies to identify diabetic retinopathy susceptibility loci that have not been previously considered as candidates to assist with a better understanding of the pathogenesis of this debilitating sight-threatening diabetes complication.

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 was supported by a grant from the Ophthalmic Research Institute of Australia. K.P.B. is a Peter Doherty Fellow of the National Health and Medical Research Council of Australia (NHMRC) and J.E.C. is an NHMRC Practitioner Fellow.

No other potential conflicts of interest relevant to this article were reported.

We acknowledge Richard Woodman for his statistical advice and expertise.

1.
Resnikoff
S
,
Pascolini
D
,
Etya'ale
D
,
Kocur
I
,
Pararajasegaram
R
,
Pokharel
GP
,
Mariotti
SP
:
Global data on visual impairment in the year 2002
.
Bull World Health Organ
2004
; 
82
:
844
851
2.
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
2004
; 
122
:
552
563
3.
National Health and Medical Research Council.
Management of Diabetic Retinopathy: Clinical Practice Guidelines
.
Canberra
,
NHMRC
2008
4.
Roy
MS
,
Klein
R
,
O'Colmain
BJ
,
Klein
BE
,
Moss
SE
,
Kempen
JH
:
The prevalence of diabetic retinopathy among adult type 1 diabetic persons in the United States
.
Arch Ophthalmol
2004
; 
122
:
546
551
5.
Klein
R
,
Klein
BE
,
Moss
SE
,
Cruickshanks
KJ
:
The Wisconsin Epidemiologic Study of Diabetic Retinopathy. XV. The long-term incidence of macular edema
.
Ophthalmology
1995
; 
102
:
7
16
6.
Todd
JA
,
Walker
NM
,
Cooper
JD
,
Smyth
DJ
,
Downes
K
,
Plagnol
V
,
Bailey
R
,
Nejentsev
S
,
Field
SF
,
Payne
F
,
Lowe
CE
,
Szeszko
JS
,
Hafler
JP
,
Zeitels
L
,
Yang
JH
,
Vella
A
,
Nutland
S
,
Stevens
HE
,
Schuilenburg
H
,
Coleman
G
,
Maisuria
M
,
Meadows
W
,
Smink
LJ
,
Healy
B
,
Burren
OS
,
Lam
AA
,
Ovington
NR
,
Allen
J
,
Adlem
E
,
Leung
HT
,
Wallace
C
,
Howson
JM
,
Guja
C
,
Ionescu-Tirgoviste
C
,
Simmonds
MJ
,
Heward
JM
,
Gough
SC
,
Dunger
DB
,
Wicker
LS
,
Clayton
DG
:
Robust associations of four new chromosome regions from genome-wide analyses of type 1 diabetes
.
Nat Genet
2007
; 
39
:
857
864
7.
Sladek
R
,
Rocheleau
G
,
Rung
J
,
Dina
C
,
Shen
L
,
Serre
D
,
Boutin
P
,
Vincent
D
,
Belisle
A
,
Hadjadj
S
,
Balkau
B
,
Heude
B
,
Charpentier
G
,
Hudson
TJ
,
Montpetit
A
,
Pshezhetsky
AV
,
Prentki
M
,
Posner
BI
,
Balding
DJ
,
Meyre
D
,
Polychronakos
C
,
Froguel
P
:
A genome-wide association study identifies novel risk loci for type 2 diabetes
.
Nature
2007
; 
445
:
881
885
8.
Brownlee
M
:
Biochemistry and molecular cell biology of diabetic complications
.
Nature
2001
; 
414
:
813
820
9.
Diabetes Control and Complications Trial Research Group.
The relationship of glycemic exposure (HbA1c) to the risk of development and progression of retinopathy in the Diabetes Control and Complications Trial
.
Diabetes
1995
; 
44
:
968
983
10.
Stratton
IM
,
Adler
AI
,
Neil
HA
,
Matthews
DR
,
Manley
SE
,
Cull
CA
,
Hadden
D
,
Turner
RC
,
Holman
RR
:
Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35): prospective observational study
.
BMJ
2000
; 
321
:
405
412
11.
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
1984
; 
102
:
520
526
12.
Jerneld
B
,
Algvere
P
:
Relationship of duration and onset of diabetes to prevalence of diabetic retinopathy
.
Am J Ophthalmol
1986
; 
102
:
431
437
13.
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
1997
; 
46
:
1829
1839
14.
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
2005
; 
28
:
1163
1168
15.
Arar
NH
,
Freedman
BI
,
Adler
SG
,
Iyengar
SK
,
Chew
EY
,
Davis
MD
,
Satko
SG
,
Bowden
DW
,
Duggirala
R
,
Elston
RC
,
Guo
X
,
Hanson
RL
,
Igo
RP
 Jr
,
Ipp
E
,
Kimmel
PL
,
Knowler
WC
,
Molineros
J
,
Nelson
RG
,
Pahl
MV
,
Quade
SR
,
Rasooly
RS
,
Rotter
JI
,
Saad
MF
,
Scavini
M
,
Schelling
JR
,
Sedor
JR
,
Shah
VO
,
Zager
PG
,
Abboud
HE
:
Heritability of the severity of diabetic retinopathy: the FIND-Eye study
.
Invest Ophthalmol Vis Sci
2008
; 
49
:
3839
3845
16.
Monti
MC
,
Lonsdale
JT
,
Montomoli
C
,
Montross
R
,
Schlag
E
,
Greenberg
DA
:
Familial risk factors for microvascular complications and differential male-female risk in a large cohort of American families with type 1 diabetes
.
J Clin Endocrinol Metab
2007
; 
92
:
4650
4655
17.
Hietala
K
,
Forsblom
C
,
Summanen
P
,
Groop
PH
:
Heritability of proliferative diabetic retinopathy
.
Diabetes
2008
; 
57
:
2176
2180
18.
DerSimonian
R
,
Laird
N
:
Meta-analysis in clinical trials
.
Control Clin Trials
1986
; 
7
:
177
188
19.
DerSimonian
R
,
Kacker
R
:
Random-effects model for meta-analysis of clinical trials: an update
.
Contemp Clin Trials
2007
; 
28
:
105
114
20.
Bertram
L
,
McQueen
MB
,
Mullin
K
,
Blacker
D
,
Tanzi
RE
:
Systematic meta-analyses of Alzheimer disease genetic association studies: the AlzGene database
.
Nat Genet
2007
; 
39
:
17
23
21.
Egger
M
,
Davey Smith
G
,
Schneider
M
,
Minder
C
:
Bias in meta-analysis detected by a simple, graphical test
.
BMJ
1997
; 
315
:
629
634
22.
Sterne
JA
,
Egger
M
:
Funnel plots for detecting bias in meta-analysis: guidelines on choice of axis
.
J Clin Epidemiol
2001
; 
54
:
1046
1055
23.
Funatsu
H
,
Yamashita
H
:
Pathogenisis of diabetic retinopathy and the renin-angiotensin system
.
Ophthalmic and Physiological Optics
2003
; 
23
:
495
501
24.
Wagner
J
,
Jan Danser
AH
,
Derkx
FH
,
de Jong
TV
,
Paul
M
,
Mullins
JJ
,
Schalekamp
MA
,
Ganten
D
:
Demonstration of renin mRNA, angiotensinogen mRNA, and angiotensin converting enzyme mRNA expression in the human eye: evidence for an intraocular renin-angiotensin system
.
Br J Ophthalmol
1996
; 
80
:
159
163
25.
The EUCLID Study Group.
Randomised placebo-controlled trial of lisinopril in normotensive patients with insulin-dependent diabetes and normoalbuminuria or microalbuminuria
.
Lancet
1997
; 
349
:
1787
1792
26.
Moravski
CJ
,
Kelly
DJ
,
Cooper
ME
,
Gilbert
RE
,
Bertram
JF
,
Shahinfar
S
,
Skinner
SL
,
Wilkinson-Berka
JL
:
Retinal neovascularization is prevented by blockade of the renin-angiotensin system
.
Hypertension
2000
; 
36
:
1099
1104
27.
Fujisawa
T
,
Ikegami
H
,
Kawaguchi
Y
,
Hamada
Y
,
Ueda
H
,
Shintani
M
,
Fukuda
M
,
Ogihara
T
:
Meta-analysis of association of insertion/deletion polymorphism of angiotensin I-converting enzyme gene with diabetic nephropathy and retinopathy
.
Diabetologia
1998
; 
41
:
47
53
28.
Wiwanitkit
V
:
Angiotensin-converting enzyme gene polymorphism is correlated to diabetic retinopathy: a meta-analysis
.
J Diabetes Complications
2008
; 
22
:
144
146
29.
Kador
PF
,
Kinoshita
JH
:
Role of aldose reductase in the development of diabetes-associated complications
.
Am J Med
1985
; 
79
:
8
12
30.
Hohman
TC
,
Nishimura
C
,
Robison
G
:
Aldose reductase and polyol in cultured pericytes of human retinal capillaries
.
Exp Eye Res
1989
; 
48
:
55
60
31.
Buzney
SM
,
Frank
RN
,
Varma
SD
,
Tanishima
T
,
Gabbay
KH
:
Aldose reductase in retinal mural cells
.
Invest Ophthalmol Visual Sci
1977
; 
16
:
392
396
32.
Akagi
Y
,
Kador
PF
,
Kuwabara
T
,
Kinoshita
JH
:
Aldose reductase localisation in human retinal mural cells
.
Invest Ophthalmol Visual Sci
1983
; 
24
:
1516
1519
33.
Engerman
RL
,
Kern
TS
:
Experimental galactosemia produces diabetic-like retinopathy
.
Diabetes
1984
; 
33
:
97
100
34.
Aiello
LP
,
Avery
RL
,
Arrigg
PG
,
Keyt
BA
,
Jampel
HD
,
Shah
ST
,
Pasquale
LR
,
Thieme
H
,
Iwamoto
MA
,
Park
JE
,
Nguyen
HV
,
Aiello
LM
,
Ferrara
N
,
King
GL
:
Vascular endothelial growth factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders
.
N Engl J Med
1994
; 
331
:
1480
1487
35.
Schlingemann
RO
,
van Hinsbergh
VW
:
Role of vascular permeability factor/vascular endothelial growth factor in eye disease
.
Br J Ophthalmol
1997
; 
81
:
501
512
36.
Ferrara
N
:
Role of vascular endothelial growth factor in regulation of physiological angiogenesis
.
Am J Physiol Cell Physiol
2001
; 
280
:
C1358
C1366
37.
Ribatti
D
:
The crucial role of vascular permeability factor/vascular endothelial growth factor in angiogenesis: a historical review
.
Br J Haematol
2005
; 
128
:
303
309
38.
Watson
CJ
,
Webb
NJ
,
Bottomley
MJ
,
Brenchley
PE
:
Identification of polymorphisms within the vascular endothelial growth factor (VEGF) gene: correlation with variation in VEGF protein production
.
Cytokine
2000
; 
12
:
1232
1235
39.
Takagi
H
,
Watanabe
D
,
Suzuma
K
,
Kurimoto
M
,
Suzuma
I
,
Ohashi
H
,
Ojima
T
,
Murakami
T
:
Novel role of erythropoietin in proliferative diabetic retinopathy
.
Diabetes Res Clin Pract
2007
; 
77
(
Suppl. 1
):
S62
S64
40.
Katsura
Y
,
Okano
T
,
Matsuno
K
,
Osako
M
,
Kure
M
,
Watanabe
T
,
Iwaki
Y
,
Noritake
M
,
Kosano
H
,
Nishigori
H
,
Matsuoka
T
:
Erythropoietin is highly elevated in vitreous fluid of patients with proliferative diabetic retinopathy
.
Diabetes Care
2005
; 
28
:
2252
2254
41.
Aiello
LP
,
Pierce
EA
,
Foley
ED
,
Takagi
H
,
Chen
H
,
Riddle
L
,
Ferrara
N
,
King
GL
,
Smith
LE
:
Suppression of retinal neovascularization in vivo by inhibition of vascular endothelial growth factor (VEGF) using soluble VEGF-receptor chimeric proteins
.
Proc Natl Acad Sci U S A
1995
; 
92
:
10457
10461
42.
Qaum
T
,
Xu
Q
,
Joussen
AM
,
Clemens
MW
,
Qin
W
,
Miyamoto
K
,
Hassessian
H
,
Wiegand
SJ
,
Rudge
J
,
Yancopoulos
GD
,
Adamis
AP
:
VEGF-initiated blood-retinal barrier breakdown in early diabetes
.
Invest Ophthalmol Vis Sci
2001
; 
42
:
2408
2413
43.
Li
H
,
Forstermann
U
:
Nitric oxide in the pathogenesis of vascular disease
.
J Pathol
2000
; 
190
:
244
254
44.
Vallance
P
,
Leiper
J
:
Blocking NO synthesis: how, where and why?
Nat Rev Drug Discov
2002
; 
1
:
939
950
45.
Albrecht
EW
,
Stegeman
CA
,
Heeringa
P
,
Henning
RH
,
van Goor
H
:
Protective role of endothelial nitric oxide synthase
.
J Pathol
2003
; 
199
:
8
17
46.
Ando
A
,
Yang
A
,
Mori
K
,
Yamada
H
,
Yamada
E
,
Takahashi
K
,
Saikia
J
,
Kim
M
,
Melia
M
,
Fishman
M
,
Huang
P
,
Campochiaro
PA
:
Nitric oxide is proangiogenic in the retina and choroid
.
J Cell Physiol
2002
; 
191
:
116
124
47.
Tsai
DC
,
Chiou
SH
,
Lee
FL
,
Chou
CK
,
Chen
SJ
,
Peng
CH
,
Kuo
YH
,
Chen
CF
,
Ho
LL
,
Hsu
WM
:
Possible involvement of nitric oxide in the progression of diabetic retinopathy
.
Ophthalmologica
2003
; 
217
:
342
346
48.
Yilmaz
G
,
Esser
P
,
Kociok
N
,
Aydin
P
,
Heimann
K
:
Elevated vitreous nitric oxide levels in patients with proliferative diabetic retinopathy
.
Am J Ophthalmol
2000
; 
130
:
87
90
49.
Zintzaras
E
,
Chatzoulis
DZ
,
Karabatsas
CH
,
Stefanidis
I
:
The relationship between C677T methylenetetrahydrofolate reductase gene polymorphism and retinopathy in type 2 diabetes: a meta-analysis
.
J Hum Genet
2005
; 
50
:
267
275
50.
Tong
Z
,
Yang
Z
,
Patel
S
,
Chen
H
,
Gibbs
D
,
Yang
X
,
Hau
VS
,
Kaminoh
Y
,
Harmon
J
,
Pearson
E
,
Buehler
J
,
Chen
Y
,
Yu
B
,
Tinkham
NH
,
Zabriskie
NA
,
Zeng
J
,
Luo
L
,
Sun
JK
,
Prakash
M
,
Hamam
RN
,
Tonna
S
,
Constantine
R
,
Ronquillo
CC
,
Sadda
S
,
Avery
RL
,
Brand
JM
,
London
N
,
Anduze
AL
,
King
GL
,
Bernstein
PS
,
Watkins
S
,
Jorde
LB
,
Li
DY
,
Aiello
LP
,
Pollak
MR
,
Zhang
K
:
Promoter polymorphism of the erythropoietin gene in severe diabetic eye and kidney complications
.
Proc Natl Acad Sci U S A
2008
; 
105
:
6998
7003

Supplementary data