October 2016
Volume 57, Issue 13
Open Access
Genetics  |   October 2016
Impact of Genetic Loci Identified in Genome-Wide Association Studies on Diabetic Retinopathy in Chinese Patients With Type 2 Diabetes
Author Affiliations & Notes
  • Chloe Y. Y. Cheung
    Department of Medicine, The University of Hong Kong, Hong Kong
  • Elaine Y. L. Hui
    Department of Medicine, The University of Hong Kong, Hong Kong
  • Chi-Ho Lee
    Department of Medicine, The University of Hong Kong, Hong Kong
  • Kelvin H. M. Kwok
    Department of Medicine, The University of Hong Kong, Hong Kong
  • Rita A. Gangwani
    Department of Ophthalmology, The University of Hong Kong, Hong Kong
  • Kenneth K. W. Li
    Department of Ophthalmology, United Christian Hospital, Hong Kong
  • Jeffrey C. W. Chan
    Department of Ophthalmology, United Christian Hospital, Hong Kong
  • Yu-Cho Woo
    Department of Medicine, The University of Hong Kong, Hong Kong
  • Wing-Sun Chow
    Department of Medicine, The University of Hong Kong, Hong Kong
  • Michele M. A. Yuen
    Department of Medicine, The University of Hong Kong, Hong Kong
  • Rachel L. C. Wong
    Department of Medicine, The University of Hong Kong, Hong Kong
  • Carol H. Y. Fong
    Department of Medicine, The University of Hong Kong, Hong Kong
  • Aimin Xu
    Department of Medicine, The University of Hong Kong, Hong Kong
    Research Center of Heart, Brain, Hormone and Healthy Aging, The University of Hong Kong, Hong Kong
    State Key Laboratory of Pharmaceutical Biotechnology, The University of Hong Kong, Hong Kong
  • David S. H. Wong
    Department of Ophthalmology, The University of Hong Kong, Hong Kong
  • Pak-Chung Sham
    Department of Psychiatry, The University of Hong Kong, Hong Kong
    Center for Genomic Sciences, The University of Hong Kong, Hong Kong
  • Karen S. L. Lam
    Department of Medicine, The University of Hong Kong, Hong Kong
    Research Center of Heart, Brain, Hormone and Healthy Aging, The University of Hong Kong, Hong Kong
    State Key Laboratory of Pharmaceutical Biotechnology, The University of Hong Kong, Hong Kong
  • Correspondence: Karen S. L. Lam, Department of Medicine, The University of Hong Kong, Queen Mary Hospital, 102 Pokfulam Road, Hong Kong; ksllam@hku.hk
  • Pak-Chung Sham, Center for Genomic Sciences, The University of Hong Kong, 6/F, HKJC Building for Interdisciplinary Research, 5 Sassoon Road, Pokfulam, Hong Kong; pcsham@hku.hk
  • Footnotes
     CYYC and EYLH are joint first authors.
Investigative Ophthalmology & Visual Science October 2016, Vol.57, 5518-5524. doi:https://doi.org/10.1167/iovs.16-20094
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Chloe Y. Y. Cheung, Elaine Y. L. Hui, Chi-Ho Lee, Kelvin H. M. Kwok, Rita A. Gangwani, Kenneth K. W. Li, Jeffrey C. W. Chan, Yu-Cho Woo, Wing-Sun Chow, Michele M. A. Yuen, Rachel L. C. Wong, Carol H. Y. Fong, Aimin Xu, David S. H. Wong, Pak-Chung Sham, Karen S. L. Lam; Impact of Genetic Loci Identified in Genome-Wide Association Studies on Diabetic Retinopathy in Chinese Patients With Type 2 Diabetes. Invest. Ophthalmol. Vis. Sci. 2016;57(13):5518-5524. https://doi.org/10.1167/iovs.16-20094.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose: Diabetic retinopathy (DR) is a common microvascular complication of type 2 diabetes (T2DM). Genome-wide association studies (GWAS) had identified novel DR-susceptibility genetic variants in various populations. We examined the associations of these DR-associated single nucleotide polymorphisms (SNPs) with severe DR in a Chinese T2DM cohort.

Methods: Cross-sectional case-control studies on sight-threatening DR (STDR) and proliferative DR (PDR) were performed. We genotyped 38 SNPs showing top association signals with DR in previous GWAS in 567 STDR cases, including 309 with PDR and 1490 non-DR controls. Multiple logistic regression models with adjustment for conventional risk factors, including age, sex, duration of diabetes, and presence of hypertension, were employed.

Results: The strongest association was found at INSR rs2115386, an intronic SNP of INSR: Padjusted = 9.13 × 10−4 (odds ratio [OR],1.28; 95% confidence interval [95%CI], 1.11–1.48) for STDR, and Padjusted= 1.12 × 10−4 (OR [95%CI],1.44 [1.20–1.74]) for PDR. rs599019 located downstream of COLEC12 (Padjusted = 0.019; OR [95%CI],1.19 [1.03–1.38]) and rs4462262 located at an intergenic region between ZWINT and MRPS35P3 (Padjusted = 0.041; OR [95%CI],1.38[1.01–1.89]) also were significantly associated with STDR, but not with PDR alone. On the other hand, MYT1L-LOC729897 rs10199521 (Padjusted = 0.022; OR [95%CI],1.25 [1.03–1.51]) and API5 rs899036 (Padjusted = 0.049; OR [95%CI],1.36 [1.00–1.85]) showed significant independent associations only with PDR. Similar results were obtained when hemoglobin A1c also was included in the adjustment models.

Conclusions: We demonstrated the significant and independent associations of several GWAS-identified SNPs with DR in Chinese T2DM patients with severe DR. The findings on INSR rs2115386 are supportive of the role of insulin resistance, or the compensatory hyperinsulinemia, in the pathogenesis of DR.

Type 2 diabetes mellitus (T2DM) is a chronic disease currently posing huge burdens on individual and public health worldwide. This disease has been considered as a major source of morbidity and mortality attributable to its associated acute or chronic complications. Diabetic retinopathy (DR) is one of the most common chronic microvascular complications of T2DM and is the leading cause of irreversible loss of vision among working populations in developed countries.1 Prevalence of DR in the Caucasian populations has been estimated to range from 19% to 30% among subjects with T2DM.25 A recent large-scale systematic screening study for DR, using the English National Screening Program grading standard,6 reported a high prevalence of DR in 39% of the Hong Kong Chinese diabetic population.7 The prevalence of the more advanced stage of DR, sight-threatening DR (STDR), which is defined as the presence of either preproliferative DR, proliferative DR (PDR), or maculopathy, was found to have reached 9.8%.7 
Over the past few decades, extensive epidemiologic studies have identified important risk factors for DR, such as duration of diabetes, hypertension (HT), and inadequate control of glycemia. However, clinical studies reveal a substantial variation in the onset and severity of retinopathy, which is not fully explained by these known risk factors.8,9 Indeed, some patients with poor control of glycemia or blood pressure (BP) may be found after long diabetes duration to have no DR.10 These data suggest that genetic factors also have a role in the development of DR. The heritability of DR has been estimated to be approximately 25%.11,12 Understanding the genetic basis of DR will help to identify the underlying pathophysiologic mechanisms. This genetic information also may contribute to the risk profiling of DR in diabetic patients, thus facilitating their early diagnosis and disease management. Robust associations of the DR-susceptibility variants would serve as the basis for their use as genetic markers to supplement the prediction by conventional clinical predictors for more accurate risk stratification. Advancing from the traditional candidate gene approach, which focused only on a small number of potential candidates implicated in the pathogenesis of DR, the genome-wide association studies (GWAS) have provided an alternative strategy for detecting novel genetic loci for DR. Several GWAS in various ethnic groups have identified novel susceptibility genetic variants for DR in T1DM and T2DM cohorts.1316 So far, only a few replication studies have been conducted to validate these novel associations.1720 Whether these DR-associated variants identified in previous GWAS all show an impact on the development of DR in the Chinese population has not been systematically investigated. The primary objective of the present study was to evaluate the associations of the GWAS identified DR-associated single nucleotide polymorphisms (SNPs) with severe DR in Hong Kong Chinese patients with T2DM. 
Subjects and Methods
Subjects
Two cross-sectional case-control studies on STDR and PDR were performed in a total of 2057 Southern Chinese patients with T2DM. This study involved 1490 non-DR controls and 567 STDR cases, including 309 subjects who had PDR. Type 2 diabetes mellitus was defined as fulfilling at least one of the following criteria, according to the American Diabetes Association 2008 diagnostic criteria:21 fasting plasma glucose (FPG) ≥ 7 mmol/l, 2-hour glucose during oral glucose tolerance test (OGTT) ≥ 11.1 mmol/l, showing symptoms of hyperglycemia and a causal plasma glucose ≥ 11.1 mmol/l, or on antidiabetic treatment (insulin or other glucose-lowering medications). All non-DR controls (n = 1490) were recruited from the Hong Kong West Diabetes Registry (HKWDR).22 Unrelated T2DM patients, who were on regular follow-up at the diabetes clinics of the Queen Mary Hospital, were invited to participate in the HKWDR, which was commenced in 2008. At 12- to 18-month intervals, the participants underwent extensive medical assessments and laboratory examinations to assess for the presence of diabetic complications, including DR as determined by the 2-field digital fundus photographs assessed by specialist ophthalmologists. The STDR cases in this study were recruited from the HKWDR, ophthalmology clinics at Queen Mary Hospital, Tseung Kwan O Hospital, and United Christian Hospital, Hong Kong. At assessment, detailed medication, and medical and family histories were recorded using a standardized questionnaire; anthropometric (weight, height, waist circumference) and clinical (age, sex, BP, and hemoglobin A1c [HbA1c]) data were collected. Duration of diabetes was defined as the difference between the time at diagnosis and time at ascertainment of STDR. Hypertension was defined as BP ≥ 140/90 mm Hg or taking antihypertensive drugs. After an overnight fast of at least 8 hours, blood samples for biochemical and genetic analysis were drawn with written informed consents were obtained. Ethical approvals were obtained from the Institutional Review Boards (IRB) of the 3 hospitals (IRB references: UW 07-378; UW 13-237; KC/KE-13-0087/FR-1). All study procedures of this research were in accordance with the Declaration of Helsinki. 
Phenotype Characterization
The current study included Chinese T2DM patients either without retinopathy or with STDR, determined on the basis of digital, color fundal photographs taken with fundus cameras (TRC50-DX type 1A; Topcon, Tokyo, Japan) with 2 photographic fields (45°) for each eye (one centered at the macula and the other centered at the optic disc). Visual acuity was assessed with the Early Treatment of Diabetic Retinopathy Study (ETDRS) chart using the auto-chart projector (Auto-Chart Projector ACP-7EM; Topcon). All STDR cases have been assessed systematically by specialist ophthalmologists who determined the presence and graded the severity of DR according to the English National Screening Program guidelines.6 Cases with STDR were defined as patients with either pre-PDR (graded R2), PDR (graded R3), or showing features of maculopathy (graded M1).7 Non-DR controls were T2DM patients without retinopathy (graded R0). Subjects with background DR (graded R1) or ungradable fundus photographs were excluded from the current study. 
Genetic Analysis
We selected 38 previously reported DR-associated SNPs to examine for their associations with severe DR (STDR and PDR) in a Southern Chinese population. These SNPs have shown top association signals with P < 5 × 10−4 as reported in at least one of the available GWAS1316 published before the commencement of genotyping in June 2014. For the reported SNPs which showed strong linkage disequilibrium (LD; r2 > 0.9) on the 1000 Genome Project for Southern Chinese, only one representative SNP with stronger association was selected for genotyping. Single nucleotide polymorphisms that were monomorphic or with minor allele frequency (MAF) < 1% in the Southern Chinese population were not included in the current study. Genomic DNA for genetic analysis was extracted using the ReliaPrep Blood gDNA Miniprep System (Promega, Madison, WI, USA) extraction kits according to the manufacturer's instructions. All SNPs were genotyped using the Sequenom iPLEX Gold genotyping platform at the Center for Genomic Sciences of the University of Hong Kong. Four SNPs that were incompatible with the Sequenom multiplexing design were replaced by a corresponding proxy SNP (with r2 > 0.8) for genotyping (rs2038823 replaced by rs16953072, rs1970671 replaced by rs1125313, rs10910200 replaced by rs6662352, and rs11867934 replaced by rs117421492). Genotyping of rs9565164 and rs1399634 was unsuccessful and was not included in the analyses. Hardy-Weinberg Equilibrium (HWE) for each SNP was examined by the exact test using PLINK version 1.09.23 To ensure the accuracy of genotype data, we excluded 3 SNPs (rs4470583, rs13163610, and rs6909083) that showed deviation from HWE (P < 0.05/38 SNPs = 1.3 × 10−3), and 4 SNPs (rs117421492, rs1445754, rs1125313, and rs2380261) with a call rate of less than 90%, from further analyses. After quality control, all 3 variants (rs96565164, rs1399634, and rs2380261) identified in a GWAS16 conducted by Sheu et al.16 were excluded for the reasons mentioned above. Thus, for the final analysis, a total of 29 SNPs that were identified from the remaining GWAS1315 and passed quality measures were examined. The average successful call rate of SNPs that passed quality control was 99.0% and the concordant rate was 99.98% based on 20 duplicated samples. 
Statistical Analysis
All statistical analyses were conducted using PLINK version 1.0923 and IBM SPSS Statistics 21. All continuous variables were summarized as mean ± SD or median with interquartile range as appropriate. Variables that did not follow a normal distribution as reflected by a significant P value in the Kolmogorov-Smirnov test were natural-logarithmically transformed before analyses. Continuous and categorical clinical parameters were compared between the case and control groups by 1-way ANOVA and the χ2 test, respectively. The associations between the SNPs and severe DR were examined by the multiple logistic regression analyses with adjustment for the conventional risk factors, including age, sex, duration of diabetes and the presence of HT (Model 1), under an additive model. Hemoglobin A1c also was included as an additional covariate in Model 2. A 2-tailed P < 0.05 was considered as statistically significant. Bonferroni correction for multiple testing was used to correct for multiple comparison. The power of study was calculated using the Genetic Power Calculator (available in the public domain at http://pngu.mgh.harvard.edu/∼purcell/gpc/).24 
Results
Association With STDR
A total of 29 DR-associated SNPs identified in previous GWAS1315 was successfully genotyped in 1490 non-DR controls and 567 STDR cases, among whom 309 had PDR. Table 1 shows the clinical characteristics of the study participants. As expected, the subjects with STDR or PDR had a longer duration of diabetes, more of them also were affected with HT, and they had greater HbA1c when compared to the non-DR subjects. 
Table 1
 
Clinical Characteristics of Study Participants
Table 1
 
Clinical Characteristics of Study Participants
Assuming an STDR prevalence of 9.8%,7 the current sample size was sufficient to achieve over 80% power for detecting significant associations of SNPs with MAF ≥ 0.2 and effect sizes greater than 1.25, at a significance level of 0.05. On the other hand, our study was underpowered to detect associations of SNPs with modest effect sizes and lower MAF (See Supplementary Fig. S1). Table 2 shows the association analyses with STDR. Of the 29 SNPs, 13 showed consistent direction of effect with the previous reports, including the strongest association at an intronic SNP rs2115386 of the insulin receptor (INSR) gene. rs2115386 was significantly associated with STDR after adjustment for the conventional confounding factors including age, sex, duration of diabetes, and the presence of HT (Padjusted = 9.13 × 10−4; odds ratio [OR], 1.28; 95% confidence interval [95%CI], 1.11–1.48). This association survived the multiple testing corrections. rs599019 located downstream of COLEC12 (Padjusted = 0.019; OR [95%CI], 1.19 [1.03–1.38]) and an intergenic SNP rs4462262 located at the ZWINT-MRPS35P3 locus (Padjusted = 0.041; OR [95%CI], 1.38 [1.01–1.89]) also showed significant associations with STDR, but were unable to survive the stringent Bonferroni correction for multiple testing. All these associations remained significantly associated with STDR when HbA1c also was included in the adjustment model (INSR rs2115386, Padjusted = 7.18 × 10−4; COLEC12 rs599019, Padjusted = 9.09 × 10−3; ZWINT-MRPS35P3 rs4462262, Padjusted = 0.031). Supplementary Table S1 shows the comparison of association results between the current study and the original reports. 
Table 2
 
Association Analysis of STDR
Table 2
 
Association Analysis of STDR
Association With PDR
We further examined the associations of these SNPs with PDR. Table 3 shows the association analyses with PDR. The INRS rs2115386 again showed a significant association with PDR even after adjustment for the confounding factors (Padjusted = 1.12 × 10−4; OR [95%CI], 1.44 [1.20–1.73]) and this association was able to survive the correction for multiple testing. However, COLEC12 rs599019 and ZWINT-MRPS35P3 rs4462262, which were significantly associated with STDR, failed to show a significant association with the advanced stage PDR alone. On the other hand, an intergenic SNP rs10199521 at the MYT1L-LOC729897 locus (Padjusted = 0.022; OR [95%CI], 1.25 [1.03–1.51]) and an intronic variant of API5 rs899036 (Padjusted = 0.049; OR [95%CI], 1.36 [1.00–1.85]) only showed significant associations with PDR. When HbA1c was included in the adjustment model, INSR rs2115386 (Padjusted = 1.23 × 10−4) and MYT1L-LOC729897 rs10199521 (Padjusted = 0.016) also remained significantly associated with PDR, but API5 rs899036 (Padjusted = 0.058) was unable to reach statistical significance after the adjustment. 
Table 3
 
Association Analysis of PDR
Table 3
 
Association Analysis of PDR
Discussion
In the present study, we evaluated the associations of 29 previously reported DR-associated SNPs with STDR and PDR in a total of 2057 Chinese patients with T2DM. We successfully confirmed the associations of several previously reported DR-susceptibility genetic variants in our population, including INSR rs2115386, which remained significantly and independently associated with STDR and PDR even after the stringent Bonferroni correction for multiple testing. 
Among the studied loci, the strongest association was found at the INSR locus. This association could be biologically relevant, since INSR encodes the insulin receptor which has a key role in insulin signaling in insulin-sensitive tissues, including retina.25 In the retina, it has been shown that disrupted insulin receptor signaling leads to cell dysfunction,25 and insulin provides trophic support for retinal neurons through a PI3-kinase/Akt-dependent pathway.26 Animal studies also have demonstrated that diabetes progressively impaired the constitutive retinal insulin receptor/Akt pro-survival signaling.27 On the other hand, insulin receptors on retinal endothelial cells have been shown to regulate the expression of vascular mediators, such as VEGF and have been implicated in the control of retinal endothelial cell growth, neovascularization, and DR.28 Insulin resistance, which often is accompanied by hyperinsulinemia, has been implicated in DR29 and shown to be an independent predictor of PDR in T2DM patients.30 Hyperinsulinemia is a cause of endothelial dysfunction,31 and has been suggested to contribute to the altered retinal microvascular blood flow involved in the development of DR.32 Our findings on the significant association of INSR rs2115386 are supportive of a role of insulin resistance, or the compensatory hyperinsulinemia, in the pathogenesis of DR. 
The association of rs2115386 with severe DR (PDR or diabetic macular edema) was first identified in a GWAS meta-analysis of two T1DM cohorts, including a total of 2829 subjects from the Epidemiology of Diabetes Intervention and Control Trial (EDIC) and the Genetics of Kidney in Diabetes (GoKinD) studies.14 In that study, none of the reported associations had reached genome-wide significance.14 Although not being the most significant association identified in that study, INSR rs2115386 was a variant that showed an association of marginal genome-wide significance (P = 2.86 × 10−6) with severe DR that deserved further investigation.14 The small number of cases with PDR and different inclusion criteria for cases and controls may explain why another study, which examined only 163 cases with non-PDR or PDR and 300 controls with no DR or mild DR, was unable to replicate the association of this SNP.18 Further replication in other independent studies with similar phenotype characterization would be essential to validate this association. rs2115386 is located at intron 2 of INSR and it shows no obvious functional significance (RegulomeDB score, 5).33 It may be linked with the disease-causing variant that is yet to be identified. Fine-mapping and deep-sequencing analyses to further refine the INSR gene region would serve to identify the causative variants. Functional studies to elucidate the role of the insulin receptor in the pathogenesis of DR also are warranted. Our findings that this SNP was significantly associated with STDR and PDR would strongly support INSR as a susceptibility gene for DR. The effect of INSR could possibly be more evident in PDR, as demonstrated in the current study. Our successful replication of this association in a T2DM cohort has provided support for a shared underlying pathogenetic pathway for DR in T1DM and T2DM. 
As with many other complex diseases, the investigations of the genetic basis for DR have benefited from the advanced microarray-based genotyping technologies in recent years. In 2010, the first GWAS for DR involving 103 moderate to severe non-PDR and PDR cases, and 183 normal to early non-PDR controls with T2DM of Mexican-American ancestry was published.13 Several GWAS of DR in different populations subsequently were published.1416 Nonetheless, the results of these GWAS have only yielded limited success with only a few variants being able to reach genome-wide significance (P ≤ 5 × 10−8) and these associations were unable to be replicated in independent GWAS.1316 Notably, these GWAS generally were conducted in relatively small sample sizes, based on different definitions for cases and controls, types of diabetes, and covariates adjustments. These studies also lacked follow-up analyses, except for the study of Sheu et al.,16 which included 2 independent replication cohorts. In the current study, we sought to systematically replicate the top association signals in a T2DM cohort with a reasonable sample size. Assuming a STDR prevalence of 9.8%7 and a MAF of 0.2, our current sample size was sufficient to achieve 68%, 85%, and 94% power to detect a significant association for an effect size of 1.20, 1.25, and 1.30, respectively, at a significance level of 0.05. We were able to detect the significant associations of the INSR variant and several other variants with severe DR with the current sample size. Among these, the INSR rs2115386, COLEC12 rs599019, AIP5 rs899036, and MYT1L rs1019952 were previously identified in the Mexican-American or Caucasian populations. Our successful replication of these SNPs in an Asian population would provide further support of the involvement of these loci in DR. We also replicated the association of ZWINT-MRPS35P3 rs4462262, which was detected in a Taiwan Chinese population. The unsuccessful replications of other variants could have been due to lack of study power to detect those SNPs with relatively lower allele frequency and modest effect size. Furthermore, the difference in definitions for cases and controls and the varying covariates adjustments between studies also might contribute to the unsuccessful replications. Alternatively, some of these signals could be false-positive. Further replication studies should be conducted in independent cohorts of larger sample size to validate or to reject the associations of these GWAS-identified variants. 
Only a handful of genes or pathways have been shown to be robustly associated with the development of DR so far. Possibly, the genetic susceptibility to DR could be conferred by rare or low frequency variants that were not covered in the candidate gene analyses and conventional GWAS that mainly focused on the common variants. The recently developed Exome-chip or whole exome/genome sequencing have provided an excellent tool for the discovery of rare, low frequency, or even population-specific disease-susceptibility variants. The gene-based tests, which account for the correlations between variants within the same gene,34 would provide an alternative approach to the single-variant analyses that often lack the power to detect the associations of rare variants. Large-scale meta-analyses with sufficient study power will be essential to detect more genetic variants with modest effect. Due to cost and time considerations in using conventional genotyping platforms, researchers may tend to focus only on the genome-wide significant signals for replications. Potential susceptibility variants that showed marginal associations sometimes may be overlooked for follow-up. Current array-based genotyping technologies allow analyses of much larger number of variants simultaneously at reasonable cost. Application of the array-based genotyping method in future follow-up studies to include more SNPs would facilitate the identification of more potential disease-susceptibility variants. 
The strength of this study was the well-defined phenotype and clear grading of DR severity based on the well-established English National Screening Program guidelines.6 Unlike many others, the current study only included subjects without DR (graded R0) as controls. The associations of SNPs identified in previous GWAS that were based on varying phenotype definitions have been comprehensively studied in this study by examining the more broadly defined STDR (including maculopathy, pre-PDR, and PDR), and the more specifically defined PDR. Potential confounding was taken into consideration by adjustment for the conventional risk factors. To maximize the sample size, we have not specifically imposed an inclusion criterion on the duration of diabetes for controls. Instead, we accounted for this possible confounding by adjustment for duration of diabetes in the analyses. During the preparation of this manuscript, 2 additional GWAS on DR35,36 were published. However, the current study has not examined these newly identified DR-associated variants. Our study would have been strengthened by a more comprehensive examination of all reported DR-susceptibility variants, including the newly identified SNPs as well as those that failed to pass quality check. The current study also was limited by the fact that it was slightly underpowered to detect the associations of variants with lower allele frequencies and modest effect. 
In conclusion, we have successfully confirmed the significant and independent associations of several SNPs identified in previous GWAS with severe DR (STDR and/or PDR) in our Chinese patients with T2DM and these genetic data may contribute to risk profiling in future diabetes management. Our findings indicated that the INSR gene is likely to be a susceptibility candidate for severe DR and supported the involvement of insulin resistance in the pathogenesis of DR. 
Acknowledgments
The authors thank the staffs of the diabetes and ophthalmology clinics of the Queen Mary Hospital, the Tseung Kwan O Hospital, and United Christian Hospital, and participants of this study, for their valuable contributions. 
Supported by the Health and Medical Research Fund of the Food and Health Bureau, HKSAR (Project No. 03144016; CYYC), and Seed Funding Program for Basic Research grant of the University of Hong Kong (EYLH). 
Disclosure: C.Y.Y. Cheung, None; E.Y.L Hui, None; C.-H. Lee, None; K.H.M. Kwok, None; R.A. Gangwani, None; K.K.W. Li, None; J.C.W. Chan, None; Y.-C. Woo, None; W.-S. Chow, None; M.M.A. Yuen, None; R.L.C. Wong, None; C.H.Y. Fong, None; A. Xu, None; D.S.H. Wong, None; P.-C. Sham, None; K.S.L. Lam, None 
References
Centers for Disease Control and Prevention. Blindness caused by diabetes--Massachusetts 1987-1994. JAMA. 1996; 276: 1865–1866.
Younis N, Broadbent DM, Harding SP, Vora JR. Prevalence of diabetic eye disease in patients entering a systematic primary care-based eye screening program. Diabet Med. 2002; 19: 1014–1021. [CrossRef] [PubMed]
Misra A, Bachmann MO, Greenwood RH, et al. Trends in yield and effects of screening intervals during 17 years of a large UK community-based diabetic retinopathy screening program. Diabet Med. 2009; 26: 1040–1047. [CrossRef] [PubMed]
Kostev K, Rathmann W. Diabetic retinopathy at diagnosis of type 2 diabetes in the UK: a database analysis. Diabetologia. 2013; 56: 109–111. [CrossRef] [PubMed]
Thomas RL, Dunstan FD, Luzio SD, et al. Prevalence of diabetic retinopathy within a national diabetic retinopathy screening service. Br J Ophthalmol. 2015; 99: 64–68. [CrossRef] [PubMed]
UK National Screening Committee. Essential Elements in Developing a Diabetic Retinopathy Screening Program. Workbook version 4.4. UK: UK National Screening Committee; 2012. Available at http://diabeticeye.screening.nhs.uk/qualityassurance. Accessed August 8 2014.
Lian JX, Gangwani RA, McGhee SM, Chan CK, Lam CL, Wong DS. Systematic screening for diabetic retinopathy (DR) in Hong Kong: prevalence of DR and visual impairment among diabetic population. Br J Ophthalmol. 2016; 100: 151–155. [CrossRef] [PubMed]
Moss SE, Klein R, Klein BE. Ten-year incidence of visual loss in a diabetic population. Ophthalmology. 1994; 101: 1061–1070. [CrossRef] [PubMed]
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. [CrossRef] [PubMed]
Keenan HA, Costacou T, Sun JK, et al. Clinical factors associated with resistance to microvascular complications in diabetic patients of extreme disease duration: the 50-year medalist study. Diabetes Care. 2007; 30: 1995–1997. [CrossRef] [PubMed]
Arar NH, Freedman BI, Adler SG, et al. Heritability of the severity of diabetic retinopathy: the FIND-Eye study. Invest Ophthalmol Vis Sci. 2008; 49: 3839–3845. [CrossRef] [PubMed]
Kuo JZ, Wong TY, Rotter JI. Challenges in elucidating the genetics of diabetic retinopathy. JAMA Ophthalmol. 2014; 132: 96–107. [CrossRef] [PubMed]
Fu YP, Hallman DM, Gonzalez VH, et al. Identification of diabetic retinopathy genes through a genome-wide association study among Mexican-Americans from Starr County Texas. J Ophthalmol. 2010; 2010: 861291. [CrossRef] [PubMed]
Grassi MA, Tikhomirov A, Ramalingam S, Below JE, Cox NJ, Nicolae DL. Genome-wide meta-analysis for severe diabetic retinopathy. Hum Mol Genet. 2011; 20: 2472–2481. [CrossRef] [PubMed]
Huang YC, Lin JM, Lin HJ, et al. Genome-wide association study of diabetic retinopathy in a Taiwanese population. Ophthalmology. 2011; 118: 642–648. [CrossRef] [PubMed]
Sheu WH, Kuo JZ, Lee IT, et al. Genome-wide association study in a Chinese population with diabetic retinopathy. Hum Mol Genet. 2013; 22: 3165–3173. [CrossRef] [PubMed]
Grassi MA, Tikhomirov A, Ramalingam S, et al. Replication analysis for severe diabetic retinopathy. Invest Ophthalmol Vis Sci. 2012; 53: 2377–2381. [CrossRef] [PubMed]
McAuley AK, Wang JJ, Dirani M, Connell PP, Lamoureux E, Hewitt AW. Replication of genetic loci implicated in diabetic retinopathy. Invest Ophthalmol Vis Sci. 2014; 55: 1666–1671. [CrossRef] [PubMed]
Hosseini SM, Boright AP, Sun L, et al. The association of previously reported polymorphisms for microvascular complications in a meta-analysis of diabetic retinopathy. Hum Gene. 2015; 134: 247–257. [CrossRef]
Peng D, Wang J, Zhang R, et al. Common variants in or near ZNRF1, COLEC12, SCYL1BP1 and API5 are associated with diabetic retinopathy in Chinese patients with type 2 diabetes. Diabetologia. 2015; 58: 1231–1238. [CrossRef] [PubMed]
American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care. 2008; 31 (suppl 1): S55–S60. [PubMed]
Hui E, Yeung CY, Lee PC, et al. Elevated circulating pigment epithelium-derived factor predicts the progression of diabetic nephropathy in patients with type 2 diabetes. J Clin Endocrinol Metab. 2014; 99: E2169–E2177. [CrossRef] [PubMed]
Purcell S, Neale B, Todd-Brown K, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007; 81: 559–575. [CrossRef] [PubMed]
Purcell S, Cherny SS, Sham PC. Genetic Power Calculator: design of linkage and association genetic mapping studies of complex traits. Bioinformatics. 2003; 19: 149–150. [CrossRef] [PubMed]
Reiter CE, Gardner TW. Functions of insulin and insulin receptor signaling in retina: possible implications for diabetic retinopathy. Prog Retin Eye Res. 2003; 22: 545–562. [CrossRef] [PubMed]
Barber AJ, Nakamura M, Wolpert EB, et al. Insulin rescues retinal neurons from apoptosis by a phosphatidylinositol 3-kinase/Akt-mediated mechanism that reduces the activation of caspase-3. J Biol Chem. 2001; 276: 32814–32821. [CrossRef] [PubMed]
Reiter CE, Wu X, Sandirasegarane L, et al. Diabetes reduces basal retinal insulin receptor signaling: reversal with systemic and local insulin. Diabetes. 2006; 55: 1148–1156. [CrossRef] [PubMed]
Kondo T, Vicent D, Suzuma K, et al. Knockout of insulin and IGF-1 receptors on vascular endothelial cells protects against retinal neovascularization. J Clin Invest. 2003; 111: 1835–1842. [CrossRef] [PubMed]
Maneschi F, Mashiter K, Kohner EM. Insulin resistance and insulin deficiency in diabetic retinopathy of non-insulin-dependent diabetes. Diabetes. 1983; 32: 82–87. [CrossRef] [PubMed]
Parvanova A, Iliev I, Filipponi M, et al. Insulin resistance and proliferative retinopathy: a cross-sectional case-control study in 115 patients with type 2 diabetes. J Clin Endocrinol Metab. 2004; 89: 4371–4376. [CrossRef] [PubMed]
Arcaro G, Cretti A, Balzano S, et al. Insulin causes endothelial dysfunction in humans: sites and mechanisms. Circulation. 2002; 105: 576–582. [CrossRef] [PubMed]
Jeppesen P, Knudsen ST, Poulsen PL, Hessellund A, Schmitz O, Bek T. Acute hyperinsulinemia increases the contraction of retinal arterioles induced by elevated blood pressure. Am J Physiol Heart Circ Physiol. 2013; 305: H1600–H1604. [CrossRef] [PubMed]
Boyle AP, Hong EL, Hariharan M, et al. Annotation of functional variation in personal genomes using RegulomeDB. Genome Res. 2012; 22: 1790–1797. [CrossRef] [PubMed]
Kang G, Jiang B, Cui Y. Gene-based genomewide association analysis: a comparison study. Curr Genomics. 2013; 14: 250–255. [CrossRef] [PubMed]
Awata T, Yamashita H, Kurihara S, et al. A genome-wide association study for diabetic retinopathy in a Japanese population: potential association with a long intergenic non-coding RNA. PLoS One. 2014; 9: e111715. [CrossRef] [PubMed]
Burdon KP, Fogarty RD, Shen W, et al. Genome-wide association study for sight-threatening diabetic retinopathy reveals association with genetic variation near the GRB2 gene. Diabetologia. 2015; 58: 2288–2297. [CrossRef] [PubMed]
Table 1
 
Clinical Characteristics of Study Participants
Table 1
 
Clinical Characteristics of Study Participants
Table 2
 
Association Analysis of STDR
Table 2
 
Association Analysis of STDR
Table 3
 
Association Analysis of PDR
Table 3
 
Association Analysis of PDR
Supplement 1
×
×

This PDF is available to Subscribers Only

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

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

×