bs_bs_banner
Clinical and Experimental Ophthalmology 2015; 43: 544–549 doi: 10.1111/ceo.12505
Original Article Association of erythropoietin gene polymorphisms with retinopathy in a Chinese cohort with type 2 diabetes mellitus Qi Song MS RN,1 Yue Zhang MD,2 Yongzhong Wu PhD,3 Fang Zhou MD2 and Yi Qu MD PhD2 1
Operating Room and 2Department of Geriatrics, Qilu Hospital of Shandong University, and 3State Key Lab of Crystal Materials, Shandong University, Jinan, China
ABSTRACT
frequency in rs551238 was significantly lower in DR group (additive OR, 0.42; 95%CI, 0.21–0.38; P = 0.016; recessive OR, 0.40; 95%CI, 0.20–0.79; P = 0.010) or PDR group (additive OR, 0.18; 95%CI, 0.05–0.62; P = 0.002; recessive OR, 0.18; 95%CI, 0.05–0.61; P = 0.002). No significant differences were detected in the distributions of rs1617640 genotype or all polymorphisms’ alleles between groups NDR and DR, PDR or NPDR. Haplotype analyses did not provide any evidence for the correlation between the three polymorphisms and DR.
Background: This study was performed to investigate the correlation of erythropoietin gene polymorphisms with diabetic retinopathy (DR) in a Chinese cohort with type 2 diabetes mellitus (T2DM). Design: Case–control study held in Qilu hospital of Shandong University, China. Participants: 792 T2DM individuals were involved, classified as groups DR (n = 448) and non-DR (NDR; n = 344). The DR group was subdivided into groups proliferative DR (PDR; n = 220) and non-PDR (NPDR; n = 228). Methods: Three single-nucleotide polymorphisms, rs1617640, rs507392 and rs551238, in the erythropoietin gene were genotyped Main Outcome Measures: Odds ratios (ORs) for the effects of erythropoietin gene polymorphisms on DR risk. Results: The genotype CC frequency in rs507392 was significantly lower in DR group (additive: OR, 0.45; 95% confidence interval [CI], 0.23–0.89; P = 0.027; recessive: OR, 0.44; 95%CI, 0.23–0.86; P = 0.012) or PDR group (additive OR, 0.18; 95%CI, 0.05–0.63; P = 0.002; recessive OR, 0.19; 95%CI, 0.06–0.66; P = 0.003) than in NDR group. The genotype CC
Conclusions: Our data suggest that rs507392 and rs551238 in the erythropoietin gene probably act to lessen the risk for DR and PDR in the Chinese T2DM cohort. Key words: Chinese, erythropoietin, retinopathy, singlenucleotide polymorphism, type 2 diabetes mellitus.
INTRODUCTION Diabetic retinopathy (DR) is the leading cause of blindness among working-age adults with diabetes mellitus, contributing to an overall 4.8% blindness globally.1,2 Clinically, DR is characterized by increased vascular permeability, tissue ischaemia and angiogenesis, all of which eventually cause a catastrophic loss of vision. There is evidence that vascular endothelial growth factor (VEGF),
■ Correspondence: Dr Yi Qu, Department of Geriatrics, Qilu Hospital of Shandong University, No. 107, Wenhuaxi Road, Jinan 250012, China. Email: [email protected] Received 14 October 2014; accepted 13 January 2015. Conflict of interest: None declared. Funding sources: This study was partly supported by Natural Science Foundation of Shandong Province (ZR2013HZ003), Fundamental Research Funds of Shandong University (2014JC015) and Scientific & Technologic Project of Shandong Province (2014GSF118128). © 2015 Royal Australian and New Zealand College of Ophthalmologists
Erythropoietin with diabetic retinopathy erythropoietin (EPO), interleukin-6 and advanced glycation end products are involved in the etiopathogenesis of DR.3–6 EPO is considered to have an angiogenic potential equivalent to VEGF.7 EPO, a glycoprotein hormone that stimulates the production of red blood cells, has been reported in the human retina.8 Vitreous levels of EPO in patients with diabetic macular oedema (DME) or proliferative diabetic retinopathy (PDR) were found to be significantly higher than in non-diabetic ones.8,9 It is assumed that the hypoxia resulting from longstanding hyperglycaemia is a major stimulus for production of intraocular EPO.10,11 Recent studies, however, have showed that several forms of an erythropoiesis-stimulating agent is able to prevent lesions associated with the early stages of DR as well as reverse retinal ischaemia in animals (rodent) models of diabetes.12–14 Intravitreal administration of EPO was shown to inhibit breakdown of the blood– retinal barrier. Li et al.15 reported intravitreal injections of EPO to be associated with short-term visual improvements in certain cases of chronic DME. Thus, it is likely that EPO plays a multifaceted role in DR. The human EPO gene is located on chromosome 7q21. Three EPO gene variants rs1617640, rs507392 and rs551238, were shown to be associated with increased risk of DR in Caucasian T2DM subjects.16,17 So far, no study has correlated EPO gene variants with DR in Chinese T2DM subjects. Therefore, this study was performed to investigate whether the single-nucleotide polymorphisms (SNPs) on EPO gene, rs1617640, rs507392 and rs551238, influence the susceptibility to DR in a Chinese T2DM cohort. In the study conducted on a Caucasian T2DM cohort,17 the allele T in SNP rs1617640 was shown to be associated with an increased risk of PDR and end-stage renal disease (ESRD). As a candidate gene involved in influencing susceptibility to DR and ESRD, we only observed the EPO gene polymorphisms in retinopathy in this study.
METHODS Subjects A total of 792 T2DM individuals were involved in this case–control study. They were classified into two groups: participants with DR (DR group, n = 448) and without DR (NDR group, n = 344). The DR group was further subdivided into those with proliferative DR (PDR group, n = 220) and nonproliferative DR (NPDR group, n = 228). Participants from Qilu Hospital of Shandong University were enrolled from 2009 through 2012. Diagnosis and classification of T2DM were based on
545 clinical features, according to the guidelines of the Expert Committee Report of the American Diabetes Association.18 The patients with ESRD were excluded. All subjects underwent ophthalmic examinations, including best-corrected visual acuity, slitlamp biomicroscopy and fundus examination, all performed by the same ophthalmologist. Retinopathy status was graded according to the Early Treatment Diabetic Retinopathy Study.19 Blood samples were collected from each participant. The study was approved by the Institutional Review Board of Qilu Hospital of Shandong University and conformed to the tenets of the Declaration of Helsinki. Written informed consent was obtained from each individual prior to participation in the study.
Genotyping Three EPO gene SNPs, rs1617640, rs507392 and rs551238, were selected. These three SNPs exhibited high linkage disequilibrium (LD) with each other.17 The SNP rs1617640 is located in the promoter of EPO gene. The SNP rs551238 is located in the 3'-HRE (hypoxia-responsive element) of the EPO gene close to the hypoxia inducible factor -1 (HIF-1) binding site. These SNPs are deemed to influence the expression of EPO or associated with DR.16,17 Total genomic DNA was extracted from peripheral venous blood of each individual using a DNA extraction and purification kit (TIANamp Blood DNA Kit; Tiangen Biotech, Beijing, China) according to the manufacturer’s instructions. SNP genotyping was conducted using the polymerase chain reaction (PCR)–ligase detection reaction (LDR) sequence method. PCR was carried out with the GenAmp PCR system9600 (Applied Biosystems, Foster City, CA, USA). The following primer pairs were used: rs1617640, 5'-TGAGAGA CCAGCTAGTCTTG-3' and 5'-TATGGCTTCTGGAA ACCCTG-3'; rs507392, 5'-CAAATATCTGGGACTA CCAC-3' and 5'-AGTGAGATCCCCCATCTCTA-3'; rs551238, 5'-CCTCCCTCTCCTTGATGACA-3' and 5'-AGCCTGTCTGACCTCTCGAC-3'. Target DNA sequences were amplified in a total volume of 20 μL, containing 1-μL genomic DNA and 19-μL master mix, which contained Taq DNA polymerase, primer mix, dNTP, Tris-HCL, KCL, MgCl2 and Taq polymerase (2×). Thermal cycling conditions were 15 min at 95°C followed by 35 cycles of 30 s at 94°C, 1 min at 56°C, and 1 min at 72°C and final extension for 7 min at 72°C. LDR was performed in a total volume of 10 μL containing 1-μL PCR product, 0.05-μL Taq DNA ligase, 1-μL probe mix, 1-μL PCR buffer and 6.95-mL H2O. LDR conditions were 2 min at 95°C followed by 35 × 30 s at 94°C and 2 min at 50°C. The LDR fluorescence products were differentiated by ABI sequencer 377 (Applied Biosystems).
© 2015 Royal Australian and New Zealand College of Ophthalmologists
546
Song et al.
Statistical analysis Frequency and genotype distributions of the three studied SNPs were tested for Hardy–Weinberg equilibrium (HWE). Differences in categorical clinical data were compared using Student’s t-test and the chi-squared test. Distribution of alleles, genotypes and haplotype frequencies were estimated using the chi-squared test. Genetic analyses for additive, dominant and recessive models were performed on all available genotype data. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated to measure the strength of the association between EPO gene variants and risk of DR. Logistic regression was used to adjust the risk of DR for age, gender, blood glucose, HbA1c level, triglyceride and duration of disease. Statistical analyses were conducted using SPSS version 17.0 (SPSS Inc., Chicago, IL, USA). All tests were two sided. A P < 0.05 was considered statistically significant. Haplotype frequencies and LD estimations were obtained using version 4.2 Haploview software (Broad Institute of MIT and Harvard, Boston, MA, USA). All haplotypes with a frequency of >3% were selected.
RESULTS Clinical characteristics of the involved subjects were shown in Table 1. The groups NDR and DR were well matched for age, gender, HbA1c levels, blood glucose, body mass index or triglyceride level. Significant differences in blood urea nitrogen levels and total cholesterol were found between the two groups. The mean duration of diabetic patients with DR was slightly greater than those with NDR. According to HWE, genotype distribution for the three SNPs (rs1617640, rs507392 and rs551238) was P > 0.05 for all T2DM subjects, slight deviation of HWE was found in the DR group. Genotype and allele frequency distributions of these three SNPs are
depicted in Table 2 and Tables S1–S2. For SNP rs507392, the frequency of genotype CC was significantly lower in the DR group (additive model: OR, 0.45; 95%CI, 0.23–0.89; P = 0.027; recessive model: OR, 0.44; 95%CI, 0.23–0.86; P = 0.012) or PDR group (additive model: OR, 0.18; 95%CI, 0.05–0.63; P = 0.002; recessive model: OR, 0.19; 95%CI, 0.06– 0.66; P = 0.003) compared with the NDR group, respectively. The frequency of genotype CC in SNP rs551238 was significantly lower in the DR group (additive model: OR, 0.42; 95%CI, 0.21–0.38; P = 0.016; recessive model: OR, 0.40; 95%CI, 0.20– 0.79; P = 0.007) or PDR group (additive model OR, 0.18; 95%CI, 0.05–0.62; P = 0.002; recessive model OR, 0.18; 95%CI, 0.05–0.61; P = 0.002) than in the NDR group. No significant differences were detected in genotype distribution of SNP rs1617640 between DR and NDR, PDR and NDR, and NPDR and NDR groups (P > 0.05, respectively). Moreover, no statistically significant differences were detected in the allele frequencies of the three SNPs between DR and NDR, PDR and NDR, and NPDR and NDR groups. No apparent association was found between allele frequencies of the three SNPs and risk of DR. Genotype CC of rs507392 were still significantly associated with decreased risk of DR (additive model, P = 0.019; recessive model, P = 0.014) and PDR (additive model, P = 0.032; recessive model, P = 0.04) after adjustment for age, gender, blood glucose, HbA1c level, triglyceride and duration of disease. The frequency of genotype CC in SNP rs551238 was found significantly lower in the DR (recessive model, P = 0.036) and PDR (recessive model, P = 0.016) groups than in the NDR group after adjustment for these factors. The three EPO gene polymorphisms were found to lie in an LD block (Fig. S1, Table S3). Haplotype analyses of the three SNPs did not provide any evidence for an association with susceptibility of DR, PDR or NPDR (P > 0.05; Table 3) in this Chinese T2DM cohort.
Table 1. Baseline characteristics of the involved subjects
Sex (male/female) Age (year) Duration of diabetes (years) HbA1c (%) Blood glucose (mmol/L) BMI (kg/m2) Triglyceride (mmol/L) BUN (mmol/L) CHO (mmol/L)
NDR (n = 344)
DR (n = 448)
P
163/181 60.16 ± 11.67 10.70 ± 7.01 8.57 ± 2.03 9.11 ± 2.99 26.16 ± 4.75 2.21 ± 1.78 5.88 ± 3.11 5.06 ± 1.59
196/252 62.35 ± 11.92 13.61 ± 7.31 9.31 ± 2.88 8.68 ± 2.83 25.58 ± 4.18 2.05 ± 1.21 10.07 ± 7.63 6.45 ± 1.77
0.31 0.13 0.001 0.07 0.34 0.30 0.44
Clinical and Experimental Ophthalmology 2015; 43: 544–549 doi: 10.1111/ceo.12505
Original Article Association of erythropoietin gene polymorphisms with retinopathy in a Chinese cohort with type 2 diabetes mellitus Qi Song MS RN,1 Yue Zhang MD,2 Yongzhong Wu PhD,3 Fang Zhou MD2 and Yi Qu MD PhD2 1
Operating Room and 2Department of Geriatrics, Qilu Hospital of Shandong University, and 3State Key Lab of Crystal Materials, Shandong University, Jinan, China
ABSTRACT
frequency in rs551238 was significantly lower in DR group (additive OR, 0.42; 95%CI, 0.21–0.38; P = 0.016; recessive OR, 0.40; 95%CI, 0.20–0.79; P = 0.010) or PDR group (additive OR, 0.18; 95%CI, 0.05–0.62; P = 0.002; recessive OR, 0.18; 95%CI, 0.05–0.61; P = 0.002). No significant differences were detected in the distributions of rs1617640 genotype or all polymorphisms’ alleles between groups NDR and DR, PDR or NPDR. Haplotype analyses did not provide any evidence for the correlation between the three polymorphisms and DR.
Background: This study was performed to investigate the correlation of erythropoietin gene polymorphisms with diabetic retinopathy (DR) in a Chinese cohort with type 2 diabetes mellitus (T2DM). Design: Case–control study held in Qilu hospital of Shandong University, China. Participants: 792 T2DM individuals were involved, classified as groups DR (n = 448) and non-DR (NDR; n = 344). The DR group was subdivided into groups proliferative DR (PDR; n = 220) and non-PDR (NPDR; n = 228). Methods: Three single-nucleotide polymorphisms, rs1617640, rs507392 and rs551238, in the erythropoietin gene were genotyped Main Outcome Measures: Odds ratios (ORs) for the effects of erythropoietin gene polymorphisms on DR risk. Results: The genotype CC frequency in rs507392 was significantly lower in DR group (additive: OR, 0.45; 95% confidence interval [CI], 0.23–0.89; P = 0.027; recessive: OR, 0.44; 95%CI, 0.23–0.86; P = 0.012) or PDR group (additive OR, 0.18; 95%CI, 0.05–0.63; P = 0.002; recessive OR, 0.19; 95%CI, 0.06–0.66; P = 0.003) than in NDR group. The genotype CC
Conclusions: Our data suggest that rs507392 and rs551238 in the erythropoietin gene probably act to lessen the risk for DR and PDR in the Chinese T2DM cohort. Key words: Chinese, erythropoietin, retinopathy, singlenucleotide polymorphism, type 2 diabetes mellitus.
INTRODUCTION Diabetic retinopathy (DR) is the leading cause of blindness among working-age adults with diabetes mellitus, contributing to an overall 4.8% blindness globally.1,2 Clinically, DR is characterized by increased vascular permeability, tissue ischaemia and angiogenesis, all of which eventually cause a catastrophic loss of vision. There is evidence that vascular endothelial growth factor (VEGF),
■ Correspondence: Dr Yi Qu, Department of Geriatrics, Qilu Hospital of Shandong University, No. 107, Wenhuaxi Road, Jinan 250012, China. Email: [email protected] Received 14 October 2014; accepted 13 January 2015. Conflict of interest: None declared. Funding sources: This study was partly supported by Natural Science Foundation of Shandong Province (ZR2013HZ003), Fundamental Research Funds of Shandong University (2014JC015) and Scientific & Technologic Project of Shandong Province (2014GSF118128). © 2015 Royal Australian and New Zealand College of Ophthalmologists
Erythropoietin with diabetic retinopathy erythropoietin (EPO), interleukin-6 and advanced glycation end products are involved in the etiopathogenesis of DR.3–6 EPO is considered to have an angiogenic potential equivalent to VEGF.7 EPO, a glycoprotein hormone that stimulates the production of red blood cells, has been reported in the human retina.8 Vitreous levels of EPO in patients with diabetic macular oedema (DME) or proliferative diabetic retinopathy (PDR) were found to be significantly higher than in non-diabetic ones.8,9 It is assumed that the hypoxia resulting from longstanding hyperglycaemia is a major stimulus for production of intraocular EPO.10,11 Recent studies, however, have showed that several forms of an erythropoiesis-stimulating agent is able to prevent lesions associated with the early stages of DR as well as reverse retinal ischaemia in animals (rodent) models of diabetes.12–14 Intravitreal administration of EPO was shown to inhibit breakdown of the blood– retinal barrier. Li et al.15 reported intravitreal injections of EPO to be associated with short-term visual improvements in certain cases of chronic DME. Thus, it is likely that EPO plays a multifaceted role in DR. The human EPO gene is located on chromosome 7q21. Three EPO gene variants rs1617640, rs507392 and rs551238, were shown to be associated with increased risk of DR in Caucasian T2DM subjects.16,17 So far, no study has correlated EPO gene variants with DR in Chinese T2DM subjects. Therefore, this study was performed to investigate whether the single-nucleotide polymorphisms (SNPs) on EPO gene, rs1617640, rs507392 and rs551238, influence the susceptibility to DR in a Chinese T2DM cohort. In the study conducted on a Caucasian T2DM cohort,17 the allele T in SNP rs1617640 was shown to be associated with an increased risk of PDR and end-stage renal disease (ESRD). As a candidate gene involved in influencing susceptibility to DR and ESRD, we only observed the EPO gene polymorphisms in retinopathy in this study.
METHODS Subjects A total of 792 T2DM individuals were involved in this case–control study. They were classified into two groups: participants with DR (DR group, n = 448) and without DR (NDR group, n = 344). The DR group was further subdivided into those with proliferative DR (PDR group, n = 220) and nonproliferative DR (NPDR group, n = 228). Participants from Qilu Hospital of Shandong University were enrolled from 2009 through 2012. Diagnosis and classification of T2DM were based on
545 clinical features, according to the guidelines of the Expert Committee Report of the American Diabetes Association.18 The patients with ESRD were excluded. All subjects underwent ophthalmic examinations, including best-corrected visual acuity, slitlamp biomicroscopy and fundus examination, all performed by the same ophthalmologist. Retinopathy status was graded according to the Early Treatment Diabetic Retinopathy Study.19 Blood samples were collected from each participant. The study was approved by the Institutional Review Board of Qilu Hospital of Shandong University and conformed to the tenets of the Declaration of Helsinki. Written informed consent was obtained from each individual prior to participation in the study.
Genotyping Three EPO gene SNPs, rs1617640, rs507392 and rs551238, were selected. These three SNPs exhibited high linkage disequilibrium (LD) with each other.17 The SNP rs1617640 is located in the promoter of EPO gene. The SNP rs551238 is located in the 3'-HRE (hypoxia-responsive element) of the EPO gene close to the hypoxia inducible factor -1 (HIF-1) binding site. These SNPs are deemed to influence the expression of EPO or associated with DR.16,17 Total genomic DNA was extracted from peripheral venous blood of each individual using a DNA extraction and purification kit (TIANamp Blood DNA Kit; Tiangen Biotech, Beijing, China) according to the manufacturer’s instructions. SNP genotyping was conducted using the polymerase chain reaction (PCR)–ligase detection reaction (LDR) sequence method. PCR was carried out with the GenAmp PCR system9600 (Applied Biosystems, Foster City, CA, USA). The following primer pairs were used: rs1617640, 5'-TGAGAGA CCAGCTAGTCTTG-3' and 5'-TATGGCTTCTGGAA ACCCTG-3'; rs507392, 5'-CAAATATCTGGGACTA CCAC-3' and 5'-AGTGAGATCCCCCATCTCTA-3'; rs551238, 5'-CCTCCCTCTCCTTGATGACA-3' and 5'-AGCCTGTCTGACCTCTCGAC-3'. Target DNA sequences were amplified in a total volume of 20 μL, containing 1-μL genomic DNA and 19-μL master mix, which contained Taq DNA polymerase, primer mix, dNTP, Tris-HCL, KCL, MgCl2 and Taq polymerase (2×). Thermal cycling conditions were 15 min at 95°C followed by 35 cycles of 30 s at 94°C, 1 min at 56°C, and 1 min at 72°C and final extension for 7 min at 72°C. LDR was performed in a total volume of 10 μL containing 1-μL PCR product, 0.05-μL Taq DNA ligase, 1-μL probe mix, 1-μL PCR buffer and 6.95-mL H2O. LDR conditions were 2 min at 95°C followed by 35 × 30 s at 94°C and 2 min at 50°C. The LDR fluorescence products were differentiated by ABI sequencer 377 (Applied Biosystems).
© 2015 Royal Australian and New Zealand College of Ophthalmologists
546
Song et al.
Statistical analysis Frequency and genotype distributions of the three studied SNPs were tested for Hardy–Weinberg equilibrium (HWE). Differences in categorical clinical data were compared using Student’s t-test and the chi-squared test. Distribution of alleles, genotypes and haplotype frequencies were estimated using the chi-squared test. Genetic analyses for additive, dominant and recessive models were performed on all available genotype data. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated to measure the strength of the association between EPO gene variants and risk of DR. Logistic regression was used to adjust the risk of DR for age, gender, blood glucose, HbA1c level, triglyceride and duration of disease. Statistical analyses were conducted using SPSS version 17.0 (SPSS Inc., Chicago, IL, USA). All tests were two sided. A P < 0.05 was considered statistically significant. Haplotype frequencies and LD estimations were obtained using version 4.2 Haploview software (Broad Institute of MIT and Harvard, Boston, MA, USA). All haplotypes with a frequency of >3% were selected.
RESULTS Clinical characteristics of the involved subjects were shown in Table 1. The groups NDR and DR were well matched for age, gender, HbA1c levels, blood glucose, body mass index or triglyceride level. Significant differences in blood urea nitrogen levels and total cholesterol were found between the two groups. The mean duration of diabetic patients with DR was slightly greater than those with NDR. According to HWE, genotype distribution for the three SNPs (rs1617640, rs507392 and rs551238) was P > 0.05 for all T2DM subjects, slight deviation of HWE was found in the DR group. Genotype and allele frequency distributions of these three SNPs are
depicted in Table 2 and Tables S1–S2. For SNP rs507392, the frequency of genotype CC was significantly lower in the DR group (additive model: OR, 0.45; 95%CI, 0.23–0.89; P = 0.027; recessive model: OR, 0.44; 95%CI, 0.23–0.86; P = 0.012) or PDR group (additive model: OR, 0.18; 95%CI, 0.05–0.63; P = 0.002; recessive model: OR, 0.19; 95%CI, 0.06– 0.66; P = 0.003) compared with the NDR group, respectively. The frequency of genotype CC in SNP rs551238 was significantly lower in the DR group (additive model: OR, 0.42; 95%CI, 0.21–0.38; P = 0.016; recessive model: OR, 0.40; 95%CI, 0.20– 0.79; P = 0.007) or PDR group (additive model OR, 0.18; 95%CI, 0.05–0.62; P = 0.002; recessive model OR, 0.18; 95%CI, 0.05–0.61; P = 0.002) than in the NDR group. No significant differences were detected in genotype distribution of SNP rs1617640 between DR and NDR, PDR and NDR, and NPDR and NDR groups (P > 0.05, respectively). Moreover, no statistically significant differences were detected in the allele frequencies of the three SNPs between DR and NDR, PDR and NDR, and NPDR and NDR groups. No apparent association was found between allele frequencies of the three SNPs and risk of DR. Genotype CC of rs507392 were still significantly associated with decreased risk of DR (additive model, P = 0.019; recessive model, P = 0.014) and PDR (additive model, P = 0.032; recessive model, P = 0.04) after adjustment for age, gender, blood glucose, HbA1c level, triglyceride and duration of disease. The frequency of genotype CC in SNP rs551238 was found significantly lower in the DR (recessive model, P = 0.036) and PDR (recessive model, P = 0.016) groups than in the NDR group after adjustment for these factors. The three EPO gene polymorphisms were found to lie in an LD block (Fig. S1, Table S3). Haplotype analyses of the three SNPs did not provide any evidence for an association with susceptibility of DR, PDR or NPDR (P > 0.05; Table 3) in this Chinese T2DM cohort.
Table 1. Baseline characteristics of the involved subjects
Sex (male/female) Age (year) Duration of diabetes (years) HbA1c (%) Blood glucose (mmol/L) BMI (kg/m2) Triglyceride (mmol/L) BUN (mmol/L) CHO (mmol/L)
NDR (n = 344)
DR (n = 448)
P
163/181 60.16 ± 11.67 10.70 ± 7.01 8.57 ± 2.03 9.11 ± 2.99 26.16 ± 4.75 2.21 ± 1.78 5.88 ± 3.11 5.06 ± 1.59
196/252 62.35 ± 11.92 13.61 ± 7.31 9.31 ± 2.88 8.68 ± 2.83 25.58 ± 4.18 2.05 ± 1.21 10.07 ± 7.63 6.45 ± 1.77
0.31 0.13 0.001 0.07 0.34 0.30 0.44
