Skip to main content
NCBI home page
Annals of Clinical and Translational Neurology logoLink to Annals of Clinical and Translational Neurology
. 2019 Sep 26;6(10):1996–2013. doi: 10.1002/acn3.50892

Genetics of diabetic neuropathy: Systematic review, meta‐analysis and trial sequential analysis

Yating Zhao 1, Ruixia Zhu 1, Danni Wang 2, Xu Liu 1,✉
PMCID: PMC6801182  PMID: 31557408

Abstract

Objective

Diabetic neuropathy (DN) is one of the most common complications of diabetes that occurs in more than 67% of individuals with diabetes. Genetic polymorphisms may play an important role in DN development. However, until now, the association between genetic polymorphisms and DN risk has remained unknown. We performed a systematic review, meta‐analysis, and trial sequential analysis (TSA) of the association between all genetic polymorphisms and DN risk.

Methods

Relevant published studies examining the relationship between all genetic polymorphisms and DN were obtained based on a designed search strategy up to 28 February 2019. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated to assess overall pooled effects of genetic models as well as in subgroup analyses. Sensitive analysis and publication bias were applied to evaluate the reliability of the study. Moreover, TSA was conducted to estimate the robustness of the results.

Results

We conducted a systematic review of a total of 1256 articles, and then 106 publications reporting on 136 polymorphisms of 76 genes were extracted. We performed 107 meta‐analyses on 36 studies involving 12,221 subjects to derive pooled effect estimates for eight polymorphisms. We identified that ACE I>D, MTHFR 1298A/C, GPx‐1 rs1050450, and CAT ‐262C/T were associated with DN, while MTHFR C677T, GSTM1, GSTT1, and IL‐10 ‐1082G/A were not. Sensitivity analysis, funnel plot, and Egger’s test displayed robust results. Furthermore, the results of TSA indicated sufficient sample size in studies of ACE, GPx‐1, GSTM1, and IL‐10 polymorphisms.

Interpretation

Our study assessed the association between ACE I>D, MTHFR C677T, MTHFR 1298A/C, GPx‐1 rs1050450, CAT ‐262C/T, GSTM1, GSTT1, and IL‐10 ‐1082G/A polymorphisms and DN risk. We hope that the data in our research study are used to study DN genetics.

Introduction

As a global public threat, diabetes mellitus (DM) is a life‐long disease that involves multiple organs and systems, and the morbidity of diabetes among adults could rise to 552 million by 2030.1, 2 As the most common complication of diabetes, diabetic neuropathy (DN) including diabetic autonomic neuropathy and somatic sensorimotor neuropathy has a prevalence of 8% in newly diagnosed diabetic patients and over 50% in patients with a long course of disease.3, 4 DN may produce a series of clinical manifestations including numbness, tingling, pain, and/or weakness which considerably decrease the quality of life in patients.5 Currently, the risk factors and pathogenesis of DN have drawn increasing attention.

Many factors are known to be associated with DN susceptibility, including smoking, obesity, poor glycemic control, and duration of diabetes, but there are still some potential factors leading to the occurrence of DN, such as genetic variants.6, 7 In 1997, Vague P et al. first found an association between the ATP1 A1 gene polymorphism and DN risk.8 Since then, an increasing number of studies have been carried out to investigate the association between various genetic polymorphisms and DN susceptibility, such as ACE I/D, MTHFR C677T and GSTM1.9, 10 For example, in 2012, Jurado et al.11 reported that the ID genotype of the ACE I/D polymorphism had a protective effect on the development of DN. However, others drew a completely different conclusion in that the ID genotype may lead to an increased DN risk.2, 12 Similarly, a significant association between the MTHFR gene C677T mutation and DN was observed by Yigit in 2013,13 which could not be replicated by Russo in 2016.14

Till now, the findings of individual studies were not always consistent, and no systematic review covered all genetic polymorphisms has been reported. To fill this gap in medical literature worldwide, we performed the first systematic review and meta‐analysis involving all the available evidence in the field of genetic variants and DN susceptibility.

Materials and Methods

Search strategy

A comprehensive literature search was performed in the PubMed and Embase databases up to 28 February 2019, using the following terms: “diabetic neuropathy/diabetic polyneuropathy/diabetic peripheral neuropathy/DPN/cardiovascular autonomic neuropathy/CAN” and “polymorphism/variant/genotype/allele/SNP/mutation”. As a complement, we also checked the reference list of the meta‐analyses and review articles on genetic association for DN, in case the references they used had been missed in original search.

Inclusion criteria

Studies were included if they met the following conditions: (1) case–control studies; (2) for the association between any genetic polymorphism and DN susceptibility; (3) sufficient allele and genotype data to calculate the odds ratios (ORs) with 95% confidence intervals (CIs); (4) studies published in English. If two papers included the same dataset, but one included additional data not found in the other paper, only the later was included. Any genetic polymorphism with three or more published studies was included in our meta‐analysis.

Data extraction

By using a standardized form, two investigators independently extracted the following data: the name of the first author, publication year, region, ethnicity, sample size, allele and genotype frequencies, genotyping methods, age‐ and gender‐matched status, type of diabetes, type of neuropathy, Newcastle‐Ottawa Quality Assessment Scale (NOS) score, and P value for Hardy–Weinberg equilibrium (HWE) in the control group. The quality of studies was evaluated using the NOS and scores >5 were considered to be of high quality, otherwise, they were thought to be with low quality.

Meta‐analysis

We used Stata 12.0 software to conduct the meta‐analysis for each genetic polymorphism to determine the pooled ORs and 95% CIs. We calculated the pooled results under all five genetic models (allelic, recessive, dominant, homozygous, and heterozygous model). Heterogeneity was measured by the I 2 statistic, and I 2 > 50% was considered significant heterogeneity. The random‐effects model was used if significant heterogeneity existed or else the fixed‐effects model was adopted. Subgroup analyses were performed based on ethnicity, genotyping methods, age‐ and gender‐matched status, HWE status of controls, quality of studies, source of control, type of diabetes, and type of neuropathy. The sensitivity analyses were conducted by sequentially omitting each study to detect the stability of pooled results and source of heterogeneity. Publication bias was explored using visual inspection of the funnel plot and Egger’s test. P < 0.05 was considered to be statistically significant.

Trial sequential analysis

Meta‐analysis may lead to a false‐positive or negative conclusion.15 Hence, we used trial sequential analysis (TSA) to reduce these statistical errors.16 TSA is a novel statistical analysis method that uses a combination of techniques that provides required information size (RIS), a threshold of statistically significant effect, for evaluating whether sufficient evidence is included and whether a result is reliable or not, in meta‐analysis. Additionally, a threshold of futility could be tested by TSA to find a conclusion of no effect before reaching the information size by using TSA software (version 0.9.5.10 beta) (Copenhagen Trial Unit, Centre for Clinical Intervention Research, Rigshospitalet, Copenhagen, Denmark). We computed the RIS based on an alpha risk of 5%, a beta risk of 20%, a relative risk reduction of 20% and a two‐sided boundary type. For those analyses that the Z‐curve reached the RIS line or monitoring the boundary line or futility area, it indicates that enough samples are included in the studies, and their results are credible. Otherwise, the amount of information is not large enough, and more evidence is needed.

Results

Study selection

In total, 1256 articles were retrieved according to our search strategy. First, we excluded 1032 articles by duplicate screening as well as title and abstract reviewing. Second, after full‐text reviewing, 118 studies containing 60 letters, reference abstracts and reviews; 38 studies not relevant to DN; 12 studies not focused on DN susceptibility; and eight studies not written in English were excluded. Third, 106 eligible articles were selected in our systematic review, and the relationship between all 136 genetic polymorphisms and DN susceptibility was extracted and listed in Table S1. Finally, for any polymorphism with three or more published studies and sufficient genotype data to extract, we keep it into our meta‐analysis. A total of 36 studies were involved in the meta‐analysis, and the entire process of study selection is shown in Figure 1.

Figure 1.

Figure 1

Open in a new tab

Flow diagram of the study selection process.

Study characteristics

Thirty‐six studies with 4515 cases and 7706 controls were included in the meta‐analysis according to the inclusion and exclusion criteria.13, 14, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39 The general characteristics of the studies are summarized in Table 1. Among the 36 studies, 6 were related to ACE I/D, 8 to MTHFR C677T, 3 to MTHFR 1298A/C, 5 to GPx‐1 rs1050450, 3 to CAT ‐262C/T, 4 to GSTM1 and GSTT1 and 3 to IL‐10 ‐1082G/A. In these studies, 26 studies were performed in the Caucasian population and 10 remaining studies were performed in the Asian population. The genotyping methods included polymerase chain reaction‐restriction fragment length polymorphism, TaqMan, polymerase chain reaction‐sequence specific primers, and amplification refractory mutation system‐polymerase chain reaction. For the quality of studies, all of them except four14, 18, 21, 37 scored more than 5 in NOS. In addition, for the HWE of controls, most of the articles met HWE equilibrium, while 10 studies failed.24, 25, 26, 30, 31, 32, 37, 39

Table 1.

Characteristics of case–control studies included in the meta‐analysis.

First author Year Region Ethnicity Case Control Genotype distribution Genotyping method Age and gender matched Type of diabetes Type of DN NOS P for HWEa
Case Control
ACE I>D           II ID DD II ID DD            
Inanir, A. 2013 Turkey Asian 235 281 43 91 101 63 123 95 NA Matched T1DM & T2DM DN 7 0.058
Mansoor, Q. 2012 Pakistan Asian 276 496 59 161 56 161 230 105 NA NA T2DM DN 5 0.177
Stephens, J. W. 2006 UK Caucasian 173 399 25 87 61 78 199 125 NA Matched T2DM DN (sensorimotor) 6 0.940
Costacou, T. 2006 USA Caucasian 114 256 86 28 200 56 NA Matched T1DM DN (sensorimotor) 7 NA
Degirmenci, I. 2005 Turkey Asian 65 75 6 38 21 19 35 21 NA NA T2DM DN 5 0.568
Ito, H. 2002 Japan Asian 21 63 14 6 1 26 27 10 NA Matched T2DM DN (sensorimotor) 7 0.506
MTHFR C677T           CC CT TT CC CT TT            
Kakavand Hamidi, A. 2018 Iran Asian 141 107 73 62 6 53 42 12 PCR‐RFLP Matched T2DM DN (sensorimotor) 7 0.408
Jimenez‐Ramirez, F. J. 2017 Puerto Rico Caucasian 89 400 72 8 9 184 159 57 PCR‐RFLP Matched T2DM DN 7 0.020
Fekih‐Mrissa, N. 2017 Tunisia Caucasian 16 144 4 12 0 52 90 2 NA Matched T2DM DN 7 0.000
Russo, G. T. 2016 Italy Caucasian 79 184 27 52 51 133 NA Unmatched T2DM DN (sensorimotor) 5 NA
Yigit, S. 2013 Turky Asian 230 282 123 85 22 180 93 9 PCR‐RFLP Matched T1DM & T2DM DN (sensorimotor) 8 0.469
Wang, H. 2012 China Asian 101 149 20 50 31 28 100 21 PCR‐RFLP Matched T2DM DN 7 0.000
Costacou, T. 2006 USA Caucasian 114 256 47 67 88 168 NA Matched T1DM DN (sensorimotor) 7 NA
Ambrosch, A. 2001 German Caucasian 43 22 15 25 2 8 12 2 PCR‐RFLP Matched T2DM DN 7 0.402
MTHFR 1298A/C           AA AC CC AA AC CC            
Kakavand Hamidi, A. 2018 Iran Asian 118 106 68 47 3 67 39 0 PCR‐RFLP Matched T2DM DN (sensorimotor) 7 0.020
Jimenez‐Ramirez, F. J. 2017 Puerto Rico Caucasian 89 400 41 43 1 251 138 11 PCR‐RFLP Matched T2DM DN 7 0.118
Fekih‐Mrissa, N. 2017 Tunisia Caucasian 16 144 10 6 0 82 42 20 NA Matched T2DM DN 7 0.001
GPx‐1 rs1050450           CC CT TT CC CT TT            
Buraczynska, M. 2017 Poland Caucasian 406 838 167 179 60 468 281 89 NA Matched T2DM DN 7 0.000
Tang, T. S.‐a 2012 UK Caucasian 211 558 79 108 24 265 224 69 PCR‐RFLP Matched T1DM & T2DM DN (sensorimotor) 6 0.047
Tang, T. S.‐b 2012 UK Caucasian 63 319 22 38 3 163 137 19 PCR‐RFLP Matched T1DM & T2DM DN (sensorimotor) 6 0.160
Matsuno, S.‐a 2011 Japan Asian 79 94 62 17 0 87 7 0 PCR‐RFLP Matched T2DM DN (sensorimotor) 7 0.708
Matsuno, S.‐b 2011 Japan Asian 25 148 22 3 0 127 21 0 PCR‐RFLP Matched T2DM DN (DAN) 7 0.353
CAT −262C/T           TT CT CC TT CT CC            
Snahnicanova, Z. 2018 Slovak Caucasian 34 80 1 13 20 6 32 42 TaqMan Matched T1DM DN (sensorimotor) 6 0.978
Kasznicki, J. 2016 Poland Caucasian 100 129 4 30 66 7 43 79 PCR‐RFLP Matched T2DM DN (sensorimotor) 6 0.719
Babizhayev, M. A. 2015 Russia Caucasian 216 250 53 80 83 96 74 80 NA Matched T1DM DN 7 0.000
GSTM1 null/present           Null Present Null Present            
Stoian, A. 2015 Romania Caucasian 42 42 18 24 22 20 PCR‐RFLP Matched T2DM DN (sensorimotor) 6 NA
Babizhayev, M. A 2015 Russia Caucasian 216 250 278 154 344 156 NA Matched T1DM DN 7 NA
Zaki, M. A. 2015 Egypt Caucasian 27 27 10 13 3 1 NA Matched T2DM DN 6 NA
Vojtkova, J. 2013 Slovak Caucasian 19 27 10 9 10 17 NA Matched T1DM DN (DAN) 7 NA
GSTT1 null/present           Null Present Null Present            
Stoian, A. 2015 Romania Caucasian 42 42 7 35 8 34 PCR‐RFLP Matched T2DM DN (sensorimotor) 6 NA
Babizhayev, M. A. 2015 Russia Caucasian 216 250 160 272 170 330 NA Matched T1DM DN 7 NA
Zaki, M. A. 2015 Egypt Caucasian 27 27 7 0 4 16 NA Matched T2DM DN 6 NA
Vojtkova, J. 2013 Slovak Caucasian 19 27 3 16 13 14 NA Matched T1DM DN (DAN) 7 NA
IL‐10‐1082G/A           GG GA AA GG GA AA            
Canecki‐Varžić, S. 2018 Croatia Caucasian 204 96 45 90 27 28 38 11 Taqman NA T2DM DN 5 0.742
Rodrigues, K. F. 2015 Brazil Caucasian 42 60 3 20 19 9 27 24 PCR‐SSP Unmatched T2DM DN (sensorimotor) 6 0.757
Kolla, V. K. 2009 India Asian 198 202 32 42 124 13 41 148 ARMS PCR Matched T2DM DN (sensorimotor) 6 0.000
Open in a new tab

DN, diabetic neuropathy; NOS, Newcastle‐Ottawa Quality Assessment Scale; HWE, Hardy–Weinberg Equilibrium; NA, not available; PCR‐RFLP, polymerase chain reaction and restriction fragment length polymorphism; PCR‐SSP, polymerase chain reaction‐sequence specific primers; ARMS PCR, amplification refractory mutation system polymerase chain reaction methods; DAN, diabetic autonomic neuropathy; T1DM, type 1 diabetes mellitus; T2DM, type 2 diabetes mellitus.

a

HWE in control.

Association between genetic polymorphisms and DN risk

ACE I>D

The ACE I>D polymorphism was investigated in six studies along with DN (884 cases, 1570 controls).17, 18, 19, 20, 21, 22 A significant association was uncovered between the ACE I>D genetic polymorphism and DN risk under allelic and homozygous models (D vs. I: OR = 1.23, 95% CI = 1.08–1.39; DD vs. II: OR = 1.50, 95% CI = 1.15–1.95) (Fig. 2). Furthermore, stratified analyses based on ethnicity, quality of studies, matched status, type of diabetes and type of neuropathy were conducted for allele, recessive, and dominant models, with results presented in Table 2. Finally, increased susceptibility was found in the recessive model in the high‐quality study group as well as in the age‐ and gender‐matched group. We subsequently performed sensitivity analyses to explore the influence of an individual study on the pooled results, and our results did not change when omitting each study in the allelic and homozygous models (Figure S1).

Figure 2.

Figure 2

Open in a new tab

Forests for ACE I>D polymorphism and DN risk. (A) allele model (D vs. I); (B) homozygous model (DD vs. II). DN, diabetic neuropathy.

Table 2.

Summary ORs and 95% CIs of ACE I>D polymorphism and DN risk.

Locus N * Allele (D vs. I) Recessive (DD vs. ID + II) Dominant (ID + DD vs. II)
OR (95% CI) P I2 (%) OR (95% CI) P I2 (%) OR (95% CI) P I2 (%)
Total 6 1.23 (1.08–1.39) 0.002 49.3 1.17 (0.97–1.41) 0.101 0 1.40 (0.91–2.14) 0.126 65.0
Ethnicity
Asian 4 1.18 (0.89–1.58) 0.252 61.8 1.16 (0.91–1.46) 0.229 37.7 1.36 (0.76–2.44) 0.301 73.6
Caucasian 2 1.21 (0.93–1.56) 0.152 – 1.19 (0.88–1.62) 0.260 0 1.43 (0.87–2.33) 0.157 –
Quality of studies
High‐quality studies 4 1.06 (0.72–1.56) 0.764 71.8 1.27 (1.01–1.59) 0.043 0 1.03 (0.57–1.87) 0.919 66.5
Matched status
Age and gender matched 4 1.06 (0.72–1.56) 0.764 71.8 1.27 (1.01–1.59) 0.043 0 1.03 (0.57–1.87) 0.919 66.5
Type of diabetes
T2DM 4 1.15 (0.87–1.53) 0.333 60.1 1.05 (0.82–1.34) 0.688 0 1.40 (0.77–2.55) 0.268 72.4
T1DM 1 –   – 1.16 (0.69–1.95) 0.569 – –   –
Type of neuropathy
Sensorimotor neuropathy 2 0.75 (0.25–2.20) 0.596 83.5 1.14 (0.84–1.53) 0.404 0 0.77 (0.20–2.99) 0.700 82.5
Open in a new tab

ORs, odds ratios; CIs, confidence intervals; DN, diabetic neuropathy; T2DM, type 2 diabetes mellitus; T1DM, type 1 diabetes mellitus.

*

Numbers of comparisons.

MTHFR C677T and 1298A/C

Totally, there were 8 studies13, 14, 20, 23, 24, 25, 26, 27 (813 cases, 1544 controls) associated with MTHFR C677T and DN involved in the meta‐analysis. Five of eight studies were performed in the Caucasian population, and the other three studies were performed in the Asian population. The pooled results of the five genetic models did not show any significant difference (Fig. 3). Further subgroup analyses were conducted, and no significant result was observed (Table 3).

Figure 3.

Figure 3

Open in a new tab

Forests for MTHFR C677T polymorphism and DN risk. (A) allele model (T vs. C); (B) recessive model (TT vs. TC + CC); (C) dominant model (TC + TT vs. CC); (D) homozygous model (TT vs. CC); (E) heterozygous model (TC vs. CC). DN, diabetic neuropathy.

Table 3.

Summary ORs and 95% CIs of MTHFR C677T polymorphism and DN risk.

Locus N * Allele (T vs. C) Recessive (TT vs. TC + CC) Dominant (TC + TT vs. CC)
OR (95% CI) P I2 (%) OR (95% CI) P I2 (%) OR (95% CI) P I2 (%)
Total 8 0.93 (0.56–1.54) 0.784 87.1 1.16 (0.50–2.71) 0.732 75.2 0.81 (0.50–1.31) 0.396 81.6
Ethnicity
Asian 3 1.22 (0.82–1.81) 0.321 74.5 1.53 (0.46–5.10) 0.489 85.2 1.22 (0.94–1.59) 0.135 43.3
Caucasian 5 0.69 (0.28–1.68) 0.416 82.5 0.68 (0.34–1.35) 0.272 0 0.66 (0.33–1.30) 0.227 79.8
Genotyping method
PCR‐RFLP 5 0.89 (0.50–1.57) 0.686 89.6 1.13 (0.46–2.78) 0.799 80.2 0.77 (0.36–1.64) 0.494 88.9
Others 3 1.24 (0.58–2.64) 0.580 – 1.73 (0.08–37.54) 0.728 – 0.80 (0.57–1.12) 0.200 0
Type of diabetes
T2DM 6 0.82 (0.47–1.44) 0.497 84.4 0.88 (0.34–2.33) 0.803 73.4 0.73 (0.40–1.33) 0.299 78.5
T1DM 1 –   – –   – 0.75 (0.47–1.18) 0.207 –
Type of neuropathy
Sensorimotor neuropathy 4 1.14 (0.58–2.24) 0.701 86.9 1.09 (0.13–9.48) 0.939 91.1 0.97 (0.66–1.42) 0.864 64.0
Controls in HWE 3 1.09 (0.65–1.83) 0.742 75.6 0.90 (0.17–4.71) 0.895 83.6 1.27 (0.96–1.68) 0.092 30.7
Open in a new tab

ORs, odds ratios; CIs, confidence intervals; DN, diabetic neuropathy; PCR‐RFLP, polymerase chain reaction and restriction fragment length polymorphism; T2DM, type 2 diabetes mellitus; T1DM, type 1 diabetes mellitus; HWE, Hardy–Weinberg Equilibrium.

*

Numbers of comparisons.

We included three studies (223 cases, 650 controls) published on the relationship between the MTHFR 1298A/C polymorphism and DN in this meta‐analysis.23, 24, 25 Two of them were performed in the Caucasian population and the other one in the Asian population. Using the AA genotype as the reference, two genetic models revealed a significant association between the MTHFR 1298A/C polymorphism and DN (CC + AC vs. AA: OR = 1.44, 95% CI = 1.03–2.01; AC vs. AA: OR = 1.51, 95% CI = 1.07–2.11; Fig. 4). In addition, the stratified analyses according to ethnicity suggested that MTHFR 1298C/T was correlated with DN in the Caucasian population (CC + AC vs. AA: OR = 1.57, 95% CI = 1.02–2.41).

Figure 4.

Figure 4

Open in a new tab

Forests for MTHFR 1298A/C polymorphism and DN risk. (A) dominant model (CC + AC vs. AA); (B) heterozygous model (AC vs. AA). DN, diabetic neuropathy.

GPx‐1 rs1050450

Five studies31, 32, 33 (784 cases, 1957 controls) were combined to analyze the association between the GPx‐1 rs1050450 polymorphism and DN. Three of five studies were performed in the Caucasian population, and the other two studies were conducted in the Asian population. The pooled OR values of four models revealed a significant association between GPx‐1 rs1050450 and DN risk (T vs. C: OR = 1.43, 95% CI = 1.26–1.64; TT + CT vs. CC: OR = 1.74, 95% CI = 1.46–2.08; TT vs. CC: OR = 1.58, 95% CI = 1.17–2.12; CT vs. CC: OR = 1.78, 95% CI = 1.48–2.14; Fig. 5). Stratification accounting for the type of diabetes revealed increased DN risk in the T2DM group (Table 4). Additionally, a similar relationship was detected under allelic and dominant models in the group with Caucasian ethnicity, sensorimotor neuropathy and controls in HWE (Table 4). In addition, each single study was omitted sequentially, without obvious alteration of overall statistical significance in sensitivity analysis (Figure S1).

Figure 5.

Figure 5

Open in a new tab

Forests for GPx‐1 rs1050450 polymorphism and DN risk. (A) allele model (T vs. C); (B) dominant model (TT + CT vs. CC); (C) homozygous model (TT vs. CC); (D) heterozygous model (CT vs. CC). DN, diabetic neuropathy.

Table 4.

Summary ORs and 95% CIs of GPx‐1 rs1050450 polymorphism and DN risk.

Locus N * Allele (T vs. C) Recessive (TT vs. CT + CC) Dominant (TT + CT vs. CC)
OR (95% CI) P I2 (%) OR (95% CI) P I2 (%) OR (95% CI) P I2(%)
Total 5 1.43 (1.26–1.64) 0.000 33.5 1.21 (0.92–1.59) 0.182 29.2 1.74 (1.46–2.08) 0.000 5.6
Ethnicity
Caucasian 3 1.42 (1.24–1.62) 0.000 18.4 1.21 (0.92–1.59) 0.182 29.2 1.72 (1.44–2.07) 0.000 0
Asian 2 1.74 (0.48–6.26) 0.399 64.2 Excluded   – 1.80 (0.45–7.19) 0.404 67.0
Type of diabetes
T2DM 3 1.57 (1.32–1.87) 0.000 37.6 1.46 (1.03–2.07) 0.035 – 1.84 (1.46–2.30) 0.000 36.6
Type of neuropathy
Sensorimotor neuropathy 3 1.33 (1.09–1.62) 0.005 49.9 0.89 (0.56–1.41) 0.624 0 1.72 (1.32–2.25) 0.000 29.5
Autonomic neuropathy 1 0.84 (0.24–2.91) 0.778 – Excluded   – 0.83 (0.23–3.00) 0.770 –
Both 1 1.55 (1.29–1.85) 0.000 – 1.46 (1.03–2.07) 0.035 – 1.81 (1.42–2.30) 0.000 –
Controls in HWE 3 1.55 (1.10–2.19) 0.013 41.1 0.79 (0.23–2.75) 0.711 – 1.99 (1.29–3.09) 0.002 34.5
Open in a new tab

ORs, odds ratios; CIs, confidence intervals; DN, diabetic neuropathy; PCR‐RFLP, polymerase chain reaction and restriction fragment length polymorphism; T2DM, type 2 diabetes mellitus; HWE, Hardy–Weinberg Equilibrium.

*

Numbers of comparisons.

CAT‐262C/T

The analysis of the CAT ‐262C/T polymorphism associated with DN included 3 studies (350 cases, 465 controls), which were all performed in the Caucasian population.28, 29, 30 Using the CC genotype as a reference, we found a protective effect of the CAT ‐262C/T polymorphism against the susceptibility of DN (T vs. C: OR = 0.71, 95% CI = 0.57–0.87; TT vs. CT + CC: OR = 0.53, 95% CI = 0.36–0.77; TT vs. CC: OR = 0.54, 95% CI = 0.35–0.82; Fig. 6). When stratified by the type of diabetes, a decreased risk was identified in the T1DM group (T vs. C: OR = 0.68, 95% CI = 0.53–0.86; TT vs. CT + CC: OR = 0.51, 95% CI = 0.35–0.76), but not T2DM group.

Figure 6.

Figure 6

Open in a new tab

Forests for CAT‐262C/T polymorphism and DN risk. (A) allele model (T vs. C); (B) recessive model (TT vs. CT + CC); (C) homozygous model (TT vs. CC). DN, diabetic neuropathy.

GSTM1 and GSTT1 null/present

The meta‐analysis including four studies30, 34, 35, 36 (516 cases, 573 controls) about GSTM1 null/present polymorphism and DN reflected no significant difference (OR = 1.21, 95% CI = 0.94–1.56, Fig. 7). Concerning GSTT1 null/present polymorphism, four studies30, 34, 35, 36 (500 cases, 589 controls) were enrolled in the meta‐analysis. The pooled results also failed to show any significant difference (OR = 0.96, 95% CI = 0.30–3.04, Fig. 8). The sensitivity analysis showed no significance after excluding any of the studies (Figure S1).

Figure 7.

Figure 7

Open in a new tab

Forest for GSTM1 null/present polymorphism and DN risk. DN, diabetic neuropathy.

Figure 8.

Figure 8

Open in a new tab

Forest for GSTT1 null/present polymorphism and DN risk. DN, diabetic neuropathy.

IL‐10 ‐1082G/A

In the meta‐analysis of the IL‐10 ‐1082G/A polymorphism and DN, three studies37, 38, 39 were involved (444 cases, 358 controls). The pooled results showed no significance between IL‐10 ‐1082G/A and DN (Fig. 9).

Figure 9.

Figure 9

Open in a new tab

Forests for IL‐10 ‐1082G/A polymorphism and DN risk. (A) allele model (G vs. A); (B) recessive model (AA vs. AG + GG); (C) dominant model (AA + AG vs. GG); (D) homozygous model (AA vs. GG); (E) heterozygous model (AG vs. GG). DN, diabetic neuropathy.

Other genetic polymorphisms associated with DN

In addition to the genetic polymorphisms discussed above, we also found that some other polymorphisms had statistical significance on DN risk in 33 individual studies, such as CACNA 1A rs2248069, CYBA rs4673, FTO rs17817449, IL2RA rs706778, SCN10A rs7375036, CTLA‐4 rs5742909, GNB3 C825T, and NOS3 Glu298Asp.8, 20, 28, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68 Due to the small number of relevant studies or insufficient data for genotype frequency, these studies could not be enrolled in our meta‐analysis. Therefore, we performed a systematic review of these polymorphisms and listed them in Table 5, with the purpose of providing clues in future searches for genetic risk factors of DN.

Table 5.

Systematic review for polymorphisms not enrolled into our meta‐analysis.

Author Year Ethnicity Gene/variant Comparison No. of cases/controls OR (95% CI) P‐value NOS References
Sun, L. 2018 Chinese CACNA 1A/rs2248069 A versus G 143/180 8.27 (3.93–17.40) <0.001 6 40
Sun, L. 2018 Chinese CACNA 1A/rs16030 C versus T 143/180 6.25 (2.86–13.67) <0.001 6 40
Sun, L. 2018 Chinese CACNA 1C/rs216008 C versus T 143/180 2.58 (1.45–1.60) 0.001 6 40
Sun, L. 2018 Chinese CACNA 1C/rs2239050 G versus C 143/180 6.01 (2.59, 13.94) <0.001 6 40
Sun, L. 2018 Chinese CACNA 1H/rs3794619 C versus T 143/180 2.52 (1.52, 4.17) <0.001 6 40
Sun, L. 2018 Chinese CACNA 1H/rs7191246 G versus C 143/180 7.38 (3.11, 17.56) <0.001 6 40
Snahnicanova, Z. 2018 Slovak CYBA/rs4673 C versus T 34/80 5.00 (1.40–19.08) 0.016 6 28
Hubacek, J. A. 2018 Czech FTO/rs17817449 G versus T 474/339 1.59 (1.11–2.29) 0.005 6 41
Zaky, E. A. 2018 Egyptian IL2RA/rs706778 A versus G 200/200 13.63 (7.52–24.71) <0.001 7 42
Ciccacci, C. 2018 Italy MIR499A/rs3746444 A versus G 69/80 1.92 (1.00–3.70) 0.005 7 43
Ezhilarasi, K. 2018 Indian VDR/rs1544410 A versus G 72/– 9.86 (4.88–19.91) 0.001 6 44
Marzban, A. 2017 Iranian HLA‐DQB1/DQB1*02 allele DQB1*02 allele 49/57 – – 7 45
Marzban, A. 2017 Iranian HLA‐DRB1/DRB1*10/DRB1*12 alleles DRB1*10/DRB1*12 alleles 49/57 – – 7 45
Gupta, B. 2017 Indian AR/rs759853 C versus T 356/294 1.97 (1.16–3.35) 0.015 7 46
Lv, Y. 2017 Chinese SCN10A/rs7375036 C versus T – – – 7 47
Kiani, J. 2016 Iranian CTLA‐4/rs5742909 C versus T 49/100 2.56 (1.31–4.98) 0.006 6 48
Ji, Z. Y. 2015 Chinese ADP/rs1501299 G versus T 90/90 2.69 (1.54–4.67) <0.001 8 49
Chen, Y. 2015 Chinese ADP/rs3774261 A versus G 80/80 3.18 (1.77–5.72) <0.001 7 50
Chen, Y. 2015 Chinese ADP/rs3821799 T versus C 80/80 2.31 (1.30–4.08) 0.004 7 50
Ren, Z. 2015 Chinese ICAM‐1/rs1799969 A versus G 399/383 3.70 (1.21–11.28) 0.014 7 51
Ren, Z. 2015 Chinese ICAM‐1/rs281432 C versus G 399/383 1.20 (1.01–1.43) 0.041 7 51
Ren, Z. 2015 Chinese ICAM‐1/rs5498 A versus G 399/383 1.72 (1.03–2.87) 0.037 7 51
Jia, Y. 2015 Chinese GRP78/rs391957 C versus T 97/198 2.23 (1.42–3.52) 0.001 7 52
Ciccacci, C. 2014 Italy MIR128a/rs11888095 C versus T 61/64 2.91 (1.32–6.44) 0.007 6 53
Ciccacci, C. 2014 Italy MIR146a/rs2910164 G versus C 27/100 0.46 (0.22–0.94) 0.032 6 53
Ciccacci, C. 2014 Italy MIR27a/rs895819 A versus G 26/97 3.20 (1.30–7.78) 0.009 6 53
Zhang, X. 2014 Chinese VEGF/C936 T C versus T 204/184 – – 7 54
Groener, J. B. 2013 Germany Glo1/rs4746 C versus A 251/273 – – 7 55
Basol, N. 2013 Turkish IL‐4/VNTR(P1) P1 versus P2 227/241 2.28 (1.46–3.58) <0.001 7 56
Ciccacci, C. 2013 Italian TCF7L2/rs7903146 C versus T 13/49 3.88 (1.53–9.81) 0.015 7 57
Korzon‐Burakowska, A. 2012 Poland OPG/rs3102734 C versus T 44/95 – – 6 58
Korzon‐Burakowska, A. 2012 Poland OPG/rs2073617 T versus C 44/95 – – 6 58
Korzon‐Burakowska, A. 2012 Poland OPG/rs3134069 T versus G 44/95 – – 6 58
Mehrab‐Mohseni, M. 2011 Iranian NOS3/intron 4 VNTR a versus b 146/96 1.80 (1.00–3.70) 0.03 6 59
Tavakkoly‐Bazzaz, J. 2010 Iranian VEGF/‐7 C/T C versus T 82/166 1.91 (1.03–3.60) 0.020 7 60
Kolla, V. K. 2009 Indian IFN‐γ/+874A/T A versus T 198/202 1.40 (1.06–1.90) 0.012 7 39
Chistiakov, D. A. 2009 Russian GNB3/C825T C versus T 100/113 2.44 (1.60–3.73) <0.001 6 61
Yang, L. 2008 Chinese MT1B/rs11076161 A versus G – – – 6 62
Yang, L. 2008 Chinese MT2A/rs10636 G versus C – – – 6 62
Nikitin, A. G. 2008 Russian PARP‐1/Leu54Phe Leu/Phe 93/86 1.66 (1.08–2.54) 0.023 6 63
Nikitin, A. G. 2008 Russian PARP‐1/Val762Ala Val/Ala 93/86 2.88 (1.43–5.77) 0.002 6 63
Papanas, N. 2007 Greek Alpha2B‐AR/I/D I versus D 130/60 – 0.001 7 64
Costacou, T. 2006 USA NOS3/Glu 298 Asp G versus T 114/256 4.86 (1.04–22.72) <0.05 7 20
Rudofsky, G., Jr. 2006 Germany UCP2/G‐866A G versus A – 0.44 (0.24–0.79) 0.007 5 65
Rudofsky, G., Jr. 2006 Germany UCP3/C‐55T C versus T – 0.48 (0.25–0.92) 0.031 5 65
Rudofsky, G., Jr. 2004 Germany TLR4/Asp299Gly + Thr399Ile Asp versus Gly Thr versus Ile – – – 7 66
Benjafield, A. V. 2001 Australia TNFRSF1B/CA16 allele I versus D 69/230 2.10 (1.20–3.80) – 6 67
Shi, H. 1998 Chinese ApoA/S2/S3/S4 S2 versus S3 versus S4 26/150 – – 6 68
Vague, P 1997 Caucasian ATP1A1/restricted allele I versus D 31/50 – – 6 8
Open in a new tab

ORs, odds ratios; CIs, confidence intervals; NOS, Newcastle‐Ottawa Quality Assessment Scale.

Detection of publication bias

Funnel plot and Egger’s test were employed to appraise the publication bias among all eight studies. By visual detection of funnel plots, six genetic variants including ACE I>D, MTHFR C677T, GPx‐1 rs1050450, CAT ‐262C/T, GSTM1 null/present and GSTT1 null/present, showed symmetric shapes, which demonstrated that no publication bias existed and was further confirmed by Egger’s test. In contrast with these variants, we detected mild publication bias in MTHFR 1298A/C and IL‐10 polymorphisms. As for MTHFR 1298A/C, marginal bias could be found in the allelic model (P = 0.025). In the recessive genetic model of IL‐10, a statistically significant difference could be found by Egger’s test (P = 0.023). The visual inspection of the funnel plot and P value of Egger’s test of all included studies are summarized in Figure S2 and Table 6, respectively.

Table 6.

Summary of P values of Egger’s test for various contrasts of genetic polymorphisms and diabetic neuropathy susceptibility.

Polymorphism Allelic model Recessive model Dominant model Homozygous model Heterozygous model
ACE I/D 0.293 0.279 0.579 0.581 0.609
MTHFR C677T 0.512 0.383 0.682 0.712 0.514
MTHFR 1298A/C 0.025 – 0.329 – 0.655
GPx‐1 0.880 0.510 0.933 0.577 0.880
CAT‐262C/T 0.460 0.925 0.669 0.913 0.469
GSTM1 null/present 0.957 – – – –
GSTT1 null/present 0.349 – – – –
IL‐10 0.535 0.023 0.866 0.441 0.936
Open in a new tab

Trial sequential analysis

Among the eight studies mentioned above, three studies performed on the ACE I>D polymorphism, GPx‐1 rs1050450 polymorphism, and IL‐10 ‐1082G/A polymorphism concluded that a sufficient number of samples were used in the analyses, and conclusive results could be obtained. Specifically, in the study of the ACE I>D polymorphism, the Z‐curve of the allelic and homozygous model crossed either the TSA monitoring boundary or RIS line, confirming that the ACE I>D polymorphism was associated with increased DN risk. For the GPx‐1 rs1050450 polymorphism, in the allelic, dominant and heterozygous models, we detected that the Z‐curve exceeded the RIS line, which revealed enough evidence for significant results. With regard to the IL‐10 ‐1082G/A polymorphism, as the Z‐curve entered the futility area in the allelic and dominant models, we came to a confirmed conclusion that the IL‐10 polymorphism had no relationship with DN susceptibility. However, the TSA results of the other five genetic variants did not show adequate information involved in the meta‐analysis. More relevant studies are necessary to prove our findings in the future. The TSA results for all the included studies are shown in Figure S3.

Discussion

As we all know, the systematic review and meta‐analysis approach used in this study is the most comprehensive method to detect genetic risk factors in most human diseases.69 To date, there is no complete systematic review and meta‐analysis reporting the potential association between all genetic polymorphisms and DN risk. Using widely accepted genetic models and subgroup analyses based on ethnicity, HWE status, quality of studies and so on, we performed this comprehensive systematic review that provided empirical support for exploring the relationship between relevant genetic polymorphisms, such as ACE I/D, MTHFR C677T, MTHFR 1298 A/C, GPx‐1 rs1050450, CAT ‐262C/T, GSTM1, GSTT1, IL‐10 ‐1082G/A, and DN susceptibility.

ACE is a key component of the renin–angiotensin system that converts angiotensin (Ang) I to Ang II. Ang II impacts endothelial damage and microcirculatory dysfunction.70 Therefore, insufficient blood supply to peripheral nerves due to microcirculatory dysfunction is considered a possible pathological mechanism of DN.71 As the starting factor affecting Ang II level, ACE activity is influenced by the presence of an insertion (I) or deletion (D) of a 287‐base pair fragment in intron 16 of the ACE gene resulting in a common variant, with the D allele being associated with higher ACE activity.72 This allele has been previously observed to probably associate with microvascular complications of diabetes.73, 74, 75 In this study, we statistically confirmed that the ACE I/D polymorphism was significantly associated with DN risk. The D allele had a 1.23‐fold risk for DN compared with the I allele, and a 50% increased risk of DN was identified in DN patients with the DD genotype compared with the II genotype.

MTHFR is a key regulatory enzyme in homocysteine metabolism that converts homocysteine back to methionine via the re‐methylation pathway.76 Therefore, deficiency of MTHFR increases the odds for hyperhomocysteinemia.77 Meantime, it was reported that homocysteine levels and the prevalence of hyperhomocysteinemia were strongly associated with DN.78 Mutations of the MTHFR gene have been defined, and C677T and A1298C variants are the two of the most explored.77 Both are functional polymorphisms that lead to decreased enzymatic activity, resulting in elevated homocysteine levels.77 The association between MTHFR gene polymorphisms and the susceptibility of DN has been investigated in several studies but with inconsistent results. Therefore, we performed this meta‐analysis involving all the available evidence of these two genetic variants and DN risk. In our study, only the MTHFR 1298A/C polymorphism showed a significant association with DN in the pooled results, while no significant difference was found in the analysis of MTHFR C677T. In vitro studies showed that hyperhomocysteinemia affected nervous function either by direct cytotoxicity or by oxidative damage.79, 80 Oxidative stress is associated with the development of apoptosis in neurons and supporting glial cells and could be the unifying mechanism that leads to nervous system damage in diabetes.81, 82

GPx‐1 is a gene that encodes an antioxidant enzyme. Its main role is protecting cells against oxidative damage by reducing hydrogen peroxide and organic peroxidases to H2O2 with reduced glutathione.83 As one of the GPx‐1 polymorphisms, rs1050450, which reduces the activity of this enzyme, may cause an adverse effect on the vascular system and microvascular complications of diabetes.84, 85 The present study aimed to evaluate the association of the rs1050450 polymorphism in the GPx‐1 gene with DN. For our pooled results, we detected that GPx‐1 rs1050450 showed a significant difference in the risk for DN. In the subgroup analysis, we found a similar result in the Caucasian population, as well as in the T2DM and sensorimotor neuropathy groups. The exact mechanism of the observed effect of GPx‐1 gene polymorphism on susceptibility to DN is unknown. We speculate that changing the capacity of the antioxidant enzyme by the rs1050450 polymorphism may lead to increased oxidative damage which was found to be an important pathophysiological mechanism involved in DN.

CAT is a widespread enzyme that can catalyze the decomposition of H2O2 to water and molecular oxygen, which inactivate free oxygen radicals and peroxides in the process of oxidative stress existing in DN.86 Therefore, CAT plays an important role in the pathogenesis of DN. From the current meta‐analysis of CAT ‐262C/T and DN risk, our findings suggested that the T allele showed a protective effect on DN development, with nearly 29% and 47% decreased susceptibility in the allelic and recessive genetic models, respectively. Additionally, all three studies involved in this meta‐analysis are performed in the Caucasian population. Thus, there may be a low risk for DN in T allele carriers of Caucasians. However, no related study was conducted in an Asian population. The role of CAT ‐262C/T in DN requires further studies for non‐Caucasian populations.

Glutathione S‐transferases (GSTs) are a family of antioxidant enzymes that play important antioxidant roles in the elimination of reactive oxygen species.87 GSTM1 and GSTT1 genes are polymorphic in humans, and the null genotypes are accompanied by a lack of enzyme activity.88 The GSTM1 and GSTT1 polymorphisms have been reported as risk factors for DN in the past but without consistent results. According to our pooled data, none of these two genetic polymorphisms showed a significant difference in the risk for DN. However, due to the limited number of further studies and the inadequate number of included samples indicated in TSA, confirming the association between either of the two genetic polymorphisms and DN is difficult. Future studies with larger sample sizes are required.

Limitation also existed in our study. First, several genes have just been investigated in small cohorts and in only Caucasian populations such as GSTT1, GSTM1, and CAT ‐262C/T. Second, we confined the enrolled studies to publications in English. Third, obvious heterogeneity could be detected among some meta‐analyses, such as MTHFR C677T and GSTM1 null/present which influences the credibility of our results. Therefore, we performed subgroup and sensitivity analyses to explore the source of heterogeneity, which was often from different study designs, measurement errors and ethnic diversity. Unfortunately, heterogeneity was not eliminated by these methods, which indicated that all factors mentioned before should be considered together. Fourth, mild publication bias was detected in MTHFR 1298A/C and IL‐10 polymorphisms, and TSA showed inadequate information involved in the analyses for MTHFR, CAT and GST genes. Thus, the comprehensive analyses should be interpreted with caution. Finally, we did not analyze the gene‐gene and gene‐environment interactions in our current meta‐analysis due to insufficient information.

In conclusion, we demonstrated that ACE I/D, MTHFR 1298A/C, GPx‐1 rs1050450, and CAT‐262C/T were associated with DN susceptibility but MTHFR C677T, GSTM1, GSTT1, and IL‐10 −1082G/A were not. More studies performed in different ethnicities with larger sample sizes are required to confirm our findings in the near future.

Conflict of Interest

The authors declare no financial or other conflicts of interests.

Supporting information

Figure S1. Sensitivity analyses for the polymorphisms and DN risk.

Click here for additional data file. (202.5KB, pdf)

Figure S2. Funnel plots for the polymorphisms and DN risk.

Click here for additional data file. (112.7KB, pdf)

Figure S3. Trial sequential analyses for the polymorphisms and DN risk.

Click here for additional data file. (2MB, pdf)

Table S1. Full genetic polymorphism list for systematic review.

Click here for additional data file. (33KB, docx)

Acknowledgments

This work was supported by the National Natural Science Foundation of China (grant no. 81400950, 81501006).

Funding Information

This work was supported by the National Natural Science Foundation of China (grant no. 81400950, 81501006).

Funding Statement

This work was funded by National Natural Science Foundation of China grants 81400950 and 81501006.

References

  • 1. Meo SA, Memon AN, Sheikh SA, et al. Effect of environmental air pollution on type 2 diabetes mellitus. Eur Rev Med Pharmacol Sci 2015;19:123–128. [PubMed] [Google Scholar]
  • 2. Xu W, Qian Y, Zhao L. Angiotensin‐converting enzyme I/D polymorphism is a genetic biomarker of diabetic peripheral neuropathy: evidence from a meta‐analysis. Int J Clin Exp Med 2015;8:944–948. [PMC free article] [PubMed] [Google Scholar]
  • 3. Singh R, Kishore L, Kaur N. Diabetic peripheral neuropathy: current perspective and future directions. Pharmacol Res 2014;80:21–35. [DOI] [PubMed] [Google Scholar]
  • 4. Deli G, Bosnyak E, Pusch G, et al. Diabetic neuropathies: diagnosis and management. Neuroendocrinology 2013;98:267–280. [DOI] [PubMed] [Google Scholar]
  • 5. Callaghan BC, Cheng HT, Stables CL, et al. Diabetic neuropathy: clinical manifestations and current treatments. Lancet Neurol 2012;11:521–534. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Tesfaye S, Chaturvedi N, Eaton SE, et al. Vascular risk factors and diabetic neuropathy. N Engl J Med 2005;352:341–350. [DOI] [PubMed] [Google Scholar]
  • 7. Adler AI, Boyko EJ, Ahroni JH, et al. Risk factors for diabetic peripheral sensory neuropathy. Results of the Seattle Prospective Diabetic Foot Study. Diabetes Care 1997;20:1162–1167. [DOI] [PubMed] [Google Scholar]
  • 8. Vague P, Dufayet D, Coste T, et al. Association of diabetic neuropathy with Na/K ATPase gene polymorphism. Diabetologia 1997;40:506–511. [DOI] [PubMed] [Google Scholar]
  • 9. Witzel II, Jelinek HF, Khalaf K, et al. Identifying common genetic risk factors of diabetic neuropathies. Front Endocrinol (Lausanne) 2015;6:88. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Politi C, Ciccacci C, D'Amato C, et al. Recent advances in exploring the genetic susceptibility to diabetic neuropathy. Diabetes Res Clin Pract 2016;120:198–208. [DOI] [PubMed] [Google Scholar]
  • 11. Jurado J, Ybarra J, Romeo JH, et al. Angiotensin‐converting enzyme gene single polymorphism as a genetic biomarker of diabetic peripheral neuropathy: longitudinal prospective study. J Diabetes Complications 2012;26:77–82. [DOI] [PubMed] [Google Scholar]
  • 12. Li Y, Tong N. Angiotensin‐converting enzyme I/D polymorphism and diabetic peripheral neuropathy in type 2 diabetes mellitus: a meta‐analysis. J Renin Angiotensin Aldosterone Syst 2015;16:787–792. [DOI] [PubMed] [Google Scholar]
  • 13. Yigit S, Karakus N, Inanir A. Association of MTHFR gene C677T mutation with diabetic peripheral neuropathy and diabetic retinopathy. Mol Vis 2013;19:1626–1630. [PMC free article] [PubMed] [Google Scholar]
  • 14. Russo GT, Giandalia A, Romeo EL, et al. Diabetic neuropathy is not associated with homocysteine, folate, vitamin B12 levels, and MTHFR C677T mutation in type 2 diabetic outpatients taking metformin. J Endocrinol Invest 2016;39:305–314. [DOI] [PubMed] [Google Scholar]
  • 15. Imberger G, Thorlund K, Gluud C, Wetterslev J. False‐positive findings in Cochrane meta‐analyses with and without application of trial sequential analysis: an empirical review. BMJ Open 2016;6:e011890. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Wetterslev J, Jakobsen JC, Gluud C. Trial sequential analysis in systematic reviews with meta‐analysis. BMC Med Res Methodol 2017;17:39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Inanir A, Basol N, Karakus N, Yigit S. The importance of association between angiotensin‐converting enzyme (ACE) Gene I/D polymorphism and diabetic peripheral neuropathy. Gene 2013;530:253–256. [DOI] [PubMed] [Google Scholar]
  • 18. Mansoor Q, Javaid A, Bilal N, Ismail M. Angiotensin‐converting enzyme (ACE) gene II genotype protects against the development of diabetic peripheral neuropathy in type 2 diabetes mellitus. J Diabetes 2012;4:257–261. [DOI] [PubMed] [Google Scholar]
  • 19. Stephens JW, Dhamrait SS, Acharya J, et al. A common variant in the ACE gene is associated with peripheral neuropathy in women with type 2 diabetes mellitus. J Diabetes Complications 2006;20:317–321. [DOI] [PubMed] [Google Scholar]
  • 20. Costacou T, Chang Y, Ferrell RE, Orchard TJ. Identifying genetic susceptibilities to diabetes‐related complications among individuals at low risk of complications: an application of tree‐structured survival analysis. Am J Epidemiol 2006;164:862–872. [DOI] [PubMed] [Google Scholar]
  • 21. Degirmenci I, Kebapci N, Basaran A, et al. Frequency of angiotensin‐converting enzyme gene polymorphism in Turkish type 2 diabetic patients. Int J Clin Pract 2005;59:1137–1142. [DOI] [PubMed] [Google Scholar]
  • 22. Ito H, Tsukui S, Kanda T, et al. Angiotensin‐converting enzyme insertion/deletion polymorphism and polyneuropathy in type 2 diabetes without macroalbuminuria. J Int Med Res 2002;30:476–482. [DOI] [PubMed] [Google Scholar]
  • 23. Kakavand Hamidi A, Radfar M, Amoli MM. Association between MTHFR variant and diabetic neuropathy. Pharmacol Rep 2018;70:1–5. [DOI] [PubMed] [Google Scholar]
  • 24. Jimenez‐Ramirez FJ, Castro LM, Ortiz C, et al. Role of treatment‐modifying MTHFR677C>T and 1298A>C polymorphisms in metformin‐treated Puerto Rican patients with type‐2 diabetes mellitus and peripheral neuropathy. Drug Metab Pers Ther 2017;32:23–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Fekih‐Mrissa N, Mrad M, Ibrahim H, et al. Methylenetetrahydrofolate reductase (MTHFR) (C677T and A1298C) polymorphisms and vascular complications in patients with type 2 diabetes. Can J Diabetes 2017;41:366–371. [DOI] [PubMed] [Google Scholar]
  • 26. Wang H, Fan D, Hong T. Is the C677T polymorphism in methylenetetrahydrofolate reductase gene or plasma homocysteine a risk factor for diabetic peripheral neuropathy in Chinese individuals? Neural Regen Res 2012;7:2384–2391. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Ambrosch A, Dierkes J, Lobmann R, et al. Relation between homocysteinaemia and diabetic neuropathy in patients with Type 2 diabetes mellitus. Diabet Med 2001;18:185–192. [DOI] [PubMed] [Google Scholar]
  • 28. Snahnicanova Z, Mendelova A, Grendar M, et al. Association of polymorphisms in CYBA, SOD1, and CAT genes with type 1 diabetes and diabetic peripheral neuropathy in children and adolescents. Genet Test Mol Biomarkers 2018;22:413–419. [DOI] [PubMed] [Google Scholar]
  • 29. Kasznicki J, Sliwinska A, Kosmalski M, et al. Genetic polymorphisms (Pro197Leu of Gpx1, +35A/C of SOD1, ‐262C/T of CAT), the level of antioxidant proteins (GPx1, SOD1, CAT) and the risk of distal symmetric polyneuropathy in Polish patients with type 2 diabetes mellitus. Adv Med Sci 2016;61:123–129. [DOI] [PubMed] [Google Scholar]
  • 30. Babizhayev MA, Strokov IA, Nosikov VV, et al. The role of oxidative stress in diabetic neuropathy: generation of free radical species in the glycation reaction and gene polymorphisms encoding antioxidant enzymes to genetic susceptibility to diabetic neuropathy in population of type I diabetic patients. Cell Biochem Biophys 2015;71:1425–1443. [DOI] [PubMed] [Google Scholar]
  • 31. Buraczynska M, Buraczynska K, Dragan M, Ksiazek A. Pro198Leu polymorphism in the glutathione peroxidase 1 gene contributes to diabetic peripheral neuropathy in type 2 diabetes patients. Neuromolecular Med 2017;19:147–153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32. Tang TS, Prior SL, Li KW, et al. Association between the rs1050450 glutathione peroxidase‐1 (C > T) gene variant and peripheral neuropathy in two independent samples of subjects with diabetes mellitus. Nutr Metab Cardiovasc Dis 2012;22:417–425. [DOI] [PubMed] [Google Scholar]
  • 33. Matsuno S, Sasaki H, Yamasaki H, et al. Pro198Leu missense polymorphism of the glutathione peroxidase 1 gene might be a common genetic predisposition of distal symmetric polyneuropathy and macrovascular disease in Japanese type 2 diabetic patients. J Diabetes Investig 2011;2:474–482. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. Stoian A, Banescu C, Balasa RI, et al. Influence of GSTM1, GSTT1, and GSTP1 polymorphisms on type 2 diabetes mellitus and diabetic sensorimotor peripheral neuropathy risk. Dis Markers 2015;2015:638693. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35. Zaki MA, Moghazy TF, El‐Deeb MMK, et al. Glutathione S‐transferase M1, T1 and P1 gene polymorphisms and the risk of developing type 2 diabetes mellitus in Egyptian diabetic patients with and without diabetic vascular complications. Alexandria J Med 2015;51:73–82. [Google Scholar]
  • 36. Vojtkova J, Durdik P, Ciljakova M, et al. The association between glutathione S‐transferase T1 and M1 gene polymorphisms and cardiovascular autonomic neuropathy in Slovak adolescents with type 1 diabetes mellitus. J Diabetes Complications 2013;27:44–48. [DOI] [PubMed] [Google Scholar]
  • 37. Canecki‐Varžić S, Prpić‐Križevać I, Mihaljević S, et al. Association between interleukin‐10 gene (‐1082G/A) polymorphism and type 2 diabetes, diabetes‐related traits, and microvascular complications in the Croatian population. Acta Clin Croat 2018;57:71–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. Rodrigues KF, Pietrani NT, Sandrim VC, et al. Association of a large panel of cytokine gene polymorphisms with complications and comorbidities in type 2 diabetes patients. J Diabetes Res 2015;2015:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Kolla VK, Madhavi G, Pulla Reddy B, et al. Association of tumor necrosis factor alpha, interferon gamma and interleukin 10 gene polymorphisms with peripheral neuropathy in South Indian patients with type 2 diabetes. Cytokine 2009;47:173–177. [DOI] [PubMed] [Google Scholar]
  • 40. Sun L, Ma J, Mao Q, et al. Association of single nucleotide polymorphisms in CACNA 1A/CACNA 1C/CACNA 1H calcium channel genes with diabetic peripheral neuropathy in Chinese population. Biosci Rep 2018;38:BSR20171670. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Hubacek JA, Dlouha D, Klementova M, et al. The FTO variant is associated with chronic complications of diabetes mellitus in Czech population. Gene 2018;642:220–224. [DOI] [PubMed] [Google Scholar]
  • 42. Zaky EA, Elgebaly HH, Adly AA, et al. Interleukin 2 receptor alfa gene polymorphism in type 1 Egyptian diabetics. Curr Pediatr Res 2018;22:38–49. [Google Scholar]
  • 43. Ciccacci C, Latini A, Greco C, et al. Association between a MIR499A polymorphism and diabetic neuropathy in type 2 diabetes. J Diabetes Complications 2018;32:11–17. [DOI] [PubMed] [Google Scholar]
  • 44. Ezhilarasi K, Dhamodharan U, Vijay V. BSMI single nucleotide polymorphism in vitamin D receptor gene is associated with decreased circulatory levels of serum 25‐hydroxyvitamin D among micro and macrovascular complications of type 2 diabetes mellitus. Int J Biol Macromol 2018;116:346–353. [DOI] [PubMed] [Google Scholar]
  • 45. Marzban A, Kiani J, Hajilooi M, et al. HLA class II alleles and risk for peripheral neuropathy in type 2 diabetes patients. Neural Regen Res 2016;11:1839–1844. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46. Gupta B, Singh SK. Association of aldose reductase gene polymorphism (C‐106T) in susceptibility of diabetic peripheral neuropathy among north Indian population. J Diabetes Complications 2017;31:1085–1089. [DOI] [PubMed] [Google Scholar]
  • 47. Lv Y, Zhou L, Tang Z, Dong J. Association and interaction analysis of diabetes mellitus and SCN10A for cardiovascular autonomic neuropathy in a Chinese population. Postgrad Med J 2017;93:344–348. [DOI] [PubMed] [Google Scholar]
  • 48. Kiani J, Khadempar S, Hajilooi M, et al. Cytotoxic T lymphocyte antigen‐4 gene variants in type 2 diabetic patients with or without neuropathy. Iran J Allergy Asthma Immunol 2016;15:220–228. [PubMed] [Google Scholar]
  • 49. Ji ZY, Li HF, Lei Y, et al. Association of adiponectin gene polymorphisms with an elevated risk of diabetic peripheral neuropathy in type 2 diabetes patients. J Diabetes Complications 2015;29:887–892. [DOI] [PubMed] [Google Scholar]
  • 50. Chen Y, Wang J, Wang L, et al. Adiponectin gene polymorphisms are associated with increased susceptibility to diabetic peripheral neuropathy. Biomarkers 2015;20:474–480. [DOI] [PubMed] [Google Scholar]
  • 51. Ren Z, Ji N, Jia K, et al. Association of the intercellular adhesion molecule‐1 gene polymorphisms with type 2 diabetes and diabetic peripheral neuropathy in a Chinese Han population. Genes Genomics 2015;37:69–75. [Google Scholar]
  • 52. Jia Y, Tong Y, Min L. Significance of functional GRP78 polymorphisms in predicting the onset of type 2 diabetic peripheral neuropathy in Chinese population. Neurol Res 2015;37:683–687. [DOI] [PubMed] [Google Scholar]
  • 53. Ciccacci C, Morganti R, Di Fusco D, et al. Common polymorphisms in MIR146a, MIR128a and MIR27a genes contribute to neuropathy susceptibility in type 2 diabetes. Acta Diabetol 2014;51:663–671. [DOI] [PubMed] [Google Scholar]
  • 54. Zhang X, Sun Z, Jiang H, Song X. Relationship between single nucleotide polymorphisms in the 3'‐untranslated region of the vascular endothelial growth factor gene and susceptibility to diabetic peripheral neuropathy in China. Arch Med Sci 2014;10:1028–1034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55. Groener JB, Reismann P, Fleming T, et al. C332C genotype of glyoxalase 1 and its association with late diabetic complications. Exp Clin Endocrinol Diabetes 2013;121:436–439. [DOI] [PubMed] [Google Scholar]
  • 56. Basol N, Inanir A, Yigit S, et al. High association of IL‐4 gene intron 3 VNTR polymorphism with diabetic peripheral neuropathy. J Mol Neurosci 2013;51:437–441. [DOI] [PubMed] [Google Scholar]
  • 57. Ciccacci C, Di Fusco D, Cacciotti L, et al. TCF7L2 gene polymorphisms and type 2 diabetes: association with diabetic retinopathy and cardiovascular autonomic neuropathy. Acta Diabetol 2013;50:789–799. [DOI] [PubMed] [Google Scholar]
  • 58. Korzon‐Burakowska A, Jakobkiewicz‐Banecka J, Fiedosiuk A, et al. Osteoprotegerin gene polymorphism in diabetic Charcot neuroarthropathy. Diabet Med 2012;29:771–775. [DOI] [PubMed] [Google Scholar]
  • 59. Mehrab‐Mohseni M, Tabatabaei‐Malazy O, Hasani‐Ranjbar S, et al. Endothelial nitric oxide synthase VNTR (intron 4 a/b) polymorphism association with type 2 diabetes and its chronic complications. Diabetes Res Clin Pract 2011;91:348–352. [DOI] [PubMed] [Google Scholar]
  • 60. Tavakkoly‐Bazzaz J, Amoli MM, Pravica V, et al. VEGF gene polymorphism association with diabetic neuropathy. Mol Biol Rep 2010;37:3625–3630. [DOI] [PubMed] [Google Scholar]
  • 61. Chistiakov DA, Spitsina EV, Nikitin AG, et al. A splice variant of GNB3 and peripheral polyneuropathy in type 1 diabetes. Dis Markers 2009;26:111–117. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62. Yang L, Li H, Yu T, et al. Polymorphisms in metallothionein‐1 and ‐2 genes associated with the risk of type 2 diabetes mellitus and its complications. Am J Physiol Endocrinol Metab 2008;294:E987–E992. [DOI] [PubMed] [Google Scholar]
  • 63. Nikitin AG, Chudakova DA, Strokov IA, et al. Leu54Phe and Val762Ala polymorphisms in the poly(ADP‐ribose)polymerase‐1 gene are associated with diabetic polyneuropathy in Russian type 1 diabetic patients. Diabetes Res Clin Pract 2008;79:446–452. [DOI] [PubMed] [Google Scholar]
  • 64. Papanas N, Papatheodorou K, Papazoglou D, et al. An insertion/deletion polymorphism in the alpha2B adrenoceptor gene is associated with peripheral neuropathy in patients with type 2 diabetes mellitus. Exp Clin Endocrinol Diabetes 2007;115:327–330. [DOI] [PubMed] [Google Scholar]
  • 65. Rudofsky G Jr, Schroedter A, Schlotterer A, et al. Functional polymorphisms of UCP2 and UCP3 are associated with a reduced prevalence of diabetic neuropathy in patients with type 1 diabetes. Diabetes Care 2006;29:89–94. [DOI] [PubMed] [Google Scholar]
  • 66. Rudofsky G Jr, Reismann P, Witte S, et al. Asp299Gly and Thr399Ile genotypes of the TLR4 gene are associated with a reduced prevalence of diabetic neuropathy in patients with type 2 diabetes. Diabetes Care 2004;27:179–183. [DOI] [PubMed] [Google Scholar]
  • 67. Benjafield AV, Glenn CL, Wang XL, et al. TNFRSFF1B in genetic predisposition to clinical neuropathy and effect on HDL cholesterol and glycosylated hemoglobin in type 2 diabetes. Diabetes Care 2001;24:753–757. [DOI] [PubMed] [Google Scholar]
  • 68. Shi H, Fang J, Yang X, et al. Lipoprotein (a) concentration and apolipoprotein (a) phenotype in subjects with type 2 diabetes mellitus. Chin Med J (Engl) 1998;111:1013–1017. [PubMed] [Google Scholar]
  • 69. Ioannidis JP, Ntzani EE, Trikalinos TA, Contopoulos‐Ioannidis DG. Replication validity of genetic association studies. Nat Genet 2001;29:306–309. [DOI] [PubMed] [Google Scholar]
  • 70. Masi S, Uliana M, Virdis A. Angiotensin II and vascular damage in hypertension: role of oxidative stress and sympathetic activation. Vascul Pharmacol 2019;115:13–17. [DOI] [PubMed] [Google Scholar]
  • 71. Feldman EL, Callaghan BC, Pop‐Busui R, et al. Diabetic neuropathy. Nat Rev Dis Primers 2019;5:41. [DOI] [PubMed] [Google Scholar]
  • 72. Danser AH, Schalekamp MA, Bax WA, et al. Angiotensin‐converting enzyme in the human heart. Effect of the deletion/insertion polymorphism. Circulation 1995;92:1387–1388. [DOI] [PubMed] [Google Scholar]
  • 73. Araz M, Aynacioglu S, Aktaran S, et al. Association between polymorphism of the angiotensin I converting enzyme gene and hypertension in Turkish type II diabetic patients. Acta Medica (Hradec Kralove) 2001;44:29–32. [PubMed] [Google Scholar]
  • 74. Hadjadj S, Gallois Y, Alhenc‐Gelas F, et al. Angiotensin‐I‐converting enzyme insertion/deletion polymorphism and high urinary albumin concentration in French Type 2 diabetes patients. Diabet Med 2003;20:677–682. [DOI] [PubMed] [Google Scholar]
  • 75. Yudkin JS, Andres C, Mohamed‐Ali V, et al. The angiotensin‐converting enzyme gene and the angiotensin II type I receptor gene as candidate genes for microalbuminuria. A study in nondiabetic and non‐insulin‐dependent diabetic subjects. Arterioscler Thromb Vasc Biol 1997;17:2188–2191. [DOI] [PubMed] [Google Scholar]
  • 76. Rozen R. Genetic predisposition to hyperhomocysteinemia: deficiency of methylenetetrahydrofolate reductase (MTHFR). Thromb Haemost 1997;78:523–526. [PubMed] [Google Scholar]
  • 77. Moll S, Varga EA. Homocysteine and MTHFR mutations. Circulation 2015;132:e6–e9. [DOI] [PubMed] [Google Scholar]
  • 78. Wang H, Cui K, Xu K, Xu S. Association between plasma homocysteine and progression of early nephropathy in type 2 diabetic patients. Int J Clin Exp Med 2015;8:11174–11180. [PMC free article] [PubMed] [Google Scholar]
  • 79. Schlussel E, Preibisch G, Putter S, Elstner EF. Homocysteine‐induced oxidative damage: mechanisms and possible roles in neurodegenerative and atherogenic processes. Z Naturforsch C 1995;50:699–707. [DOI] [PubMed] [Google Scholar]
  • 80. Weir DG, Scott JM. The biochemical basis of the neuropathy in cobalamin deficiency. Baillieres Clin Haematol 1995;8:479–497. [DOI] [PubMed] [Google Scholar]
  • 81. Russell JW, Sullivan KA, Windebank AJ, et al. Neurons undergo apoptosis in animal and cell culture models of diabetes. Neurobiol Dis 1999;6:347–363. [DOI] [PubMed] [Google Scholar]
  • 82. Russell JW, Golovoy D, Vincent AM, et al. High glucose‐induced oxidative stress and mitochondrial dysfunction in neurons. FASEB J 2002;16:1738–1748. [DOI] [PubMed] [Google Scholar]
  • 83. Endler M, Saltvedt S, Eweida M, Akerud H. Oxidative stress and inflammation in retained placenta: a pilot study of protein and gene expression of GPX1 and NFkappaB. BMC Pregnancy Childbirth 2016;16:384. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 84. Ravn‐Haren G, Olsen A, Tjonneland A, et al. Associations between GPX1 Pro198Leu polymorphism, erythrocyte GPX activity, alcohol consumption and breast cancer risk in a prospective cohort study. Carcinogenesis 2006;27:820–825. [DOI] [PubMed] [Google Scholar]
  • 85. Wickremasinghe D, Peiris H, Chandrasena LG, et al. Case control feasibility study assessing the association between severity of coronary artery disease with Glutathione Peroxidase‐1 (GPX‐1) and GPX‐1 polymorphism (Pro198Leu). BMC Cardiovasc Disord 2016;16:111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 86. Basic J, Vojinovic J, Jevtovic‐Stoimenov T, et al. The association of CAT‐262C/T polymorphism with catalase activity and treatment response in juvenile idiopathic arthritis. Rheumatol Int 2019;39:551–559. [DOI] [PubMed] [Google Scholar]
  • 87. Ginsberg G, Smolenski S, Hattis D, et al. Genetic polymorphism in glutathione transferases (GST): population distribution of GSTM1, T1, and P1 conjugating activity. J Toxicol Environ Health B Crit Rev 2009;12:389–439. [DOI] [PubMed] [Google Scholar]
  • 88. Seidegard J, Vorachek WR, Pero RW, Pearson WR. Hereditary differences in the expression of the human glutathione transferase active on trans‐stilbene oxide are due to a gene deletion. Proc Natl Acad Sci USA 1988;85:7293–7297. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figure S1. Sensitivity analyses for the polymorphisms and DN risk.

Click here for additional data file. (202.5KB, pdf)

Figure S2. Funnel plots for the polymorphisms and DN risk.

Click here for additional data file. (112.7KB, pdf)

Figure S3. Trial sequential analyses for the polymorphisms and DN risk.

Click here for additional data file. (2MB, pdf)

Table S1. Full genetic polymorphism list for systematic review.

Click here for additional data file. (33KB, docx)

Articles from Annals of Clinical and Translational Neurology are provided here courtesy of Wiley

RESOURCES