Association of MTHFR Gene with Type 2 Diabetes Mellitus: A Case-Control Study and Meta-Analysis Open Access

Type 2 diabetes mellitus (T2DM) is a chronic metabolic disease that accounts for more than 90 percent of diabetes patients [1]. The prevalence of diabetes has increased significantly in the past 3 decades, and it has become a severe public health problem for human health. According to the latest report, about 537 million adults worldwide had diabetes in 2021 [2]. If these trends continue, 643 million people by 2030 or 783 million people by 2045 will have diabetes [3, 4]. And the resulting complications have become more serious with the significant increase in the prevalence of diabetes [5]. Complications of diabetes seriously damage the physical and mental health of patients, reduce the patient’s quality of life, and also impose a heavy economic burden on their families and society [6].

The occurrence of type 2 diabetes is the result of a combination of several factors. Genetic factors and environmental factors (overweight/obesity, dietary factors, physical activity, social and economic conditions) are the main factors [5, 7]. A genome-wide association study has identified more than 400 loci that are closely related to an increased risk of T2DM [8]. Therefore, genetic polymorphisms may be related to the development of T2DM.

The most common form of genetic variation in humans is single nucleotide polymorphisms (SNPs). SNPs can alter gene expression and function. Polymorphisms are also used to predict the risk of disease [9]. MicroRNAs (miRNAs) contribute to epigenetic regulation. In the human genome, there are approximately 45,000 SNPs of miRNA targeting sites are expected to influence gene expression which makes it possible to predict the risk of disease [10]. The same may be true for MTHFR gene and T2DM. The MTHFR gene produces a functioning MTHFR enzyme. This enzyme converts MTHFR (methylenetetrahydrofolate reductase), a folate-dependent enzyme, catalyzes the reduction of 5,10-methylenetetrahydrofolate to 5-methyltetrahydrofolate and helps regulate homocysteine levels. The polymorphism of the MTHFR gene locus will affect the enzyme activity, cause folate metabolism pathway obstruction, and then affect the body’s folate metabolism, thus leading to elevated homocysteine level, and increasing the risk of coronary heart disease, peripheral vascular disease, and cerebrovascular disease [11]. There are more than one genetic polymorphism in the MTHFR gene that could lead to change of MTHFR enzyme activity. Hence, SNPs in miRNAs may be related to susceptibility and development of T2DM. Several case-control studies have shown inconsistent results regarding the association between MTHFR gene polymorphisms and T2DM risk [12‒15].

Therefore, in this study, we systematically explore the association between MTHFR gene (C677T and A1298C) polymorphisms and T2DM. First, we examined the effect of MTHFR gene polymorphism on susceptibility to T2DM using a large sample case-control study. Second, we conducted a meta-analysis of published studies to explore the association of MTHFR gene polymorphisms with T2DM susceptibility. After preliminary literature search, it was found that the A1298C polymorphism (rs1801131) was not significantly associated with T2DM. Therefore, only the C677T polymorphism (rs1801133), whose results were controversial, was meta-analyzed. This study attempts to explore and clarify the association between MTHFR and T2DM and provide new ideas for the prevention and treatment of T2DM.

We accordingly selected the C677T(rs1801133) and A1298C(rs1801131) SNPs of MTHFR to verify their association with T2DM. The data came from the Anhui Healthy Longevity Survey (AHLS), the details of which have been reported elsewhere [16]. Inclusion criteria for the T2DM group: (1) age ≥55 years old; (2) fasting blood glucose ≥7.0 mmol/L or a 2-h post-glucose level ≥11.1 mmol/L or the hospital clinical diagnosis was T2DM; (3) no family history of DM. Inclusion criteria for the control group: (1) age ≥55 years old, (2) fasting blood glucose <6.1 mmol/L and OGTT 2-h PG <7.8 mmol/L. Exclusion criteria for two groups: (1) lack of compliance and willingness to cooperate with the investigation; (2) patients with severe consumptive diseases(including hyperthyroidism, malignant tumor, and severe trauma); (3) the information required for the study was incomplete. According to the inclusion and exclusion criteria, a total of 1,184 individuals were included in the case-control study. Among them, 538 patients were T2DM group and other 646 individuals were in the control group.

The investigation included a questionnaire, physical examination, and laboratory test. The participants were investigated by investigators who had been uniformly trained and passed an assessment by face-to-face questionnaire survey. The contents of the questionnaire included demographic characteristics and history of diabetes and other diseases. The measurement of height (m), body weight (kg), waist circumference (cm), and hip circumference (cm) was repeated twice. BMI and WHR were calculated based on that. BMI = weight (kg)/height² (m²); WHR = waist circumference (cm)/hip circumference (cm). BMI was divided into three groups according to the criteria: underweight (<18.5 kg/m²), normal weight (18.5–24 kg/m²), and overweight and obesity (≥24 kg/m²) [17]. Two mL of fasting venous blood was collected to measure FPG after fasting for 8 h.

Five mL of peripheral venous blood from participants was extracted by using EDTA anticoagulation tube after obtaining the informed consent of the subjects. Genomic DNA of blood samples was extracted by using a DNA kit (QIAGEN FlexiGene^® DNA) in the laboratory and placed in a −80°C refrigerator. Horizon^®58 Agarose gel Level Electrophoresis Instrument (Biometra Corporation) was used to access the quality of DNA. And the NanoDrop 2000 Microvolume UV-Vis Spectrophotometer (NanoDrop) was used to standardize DNA concentration. SNPscan™ was used for rs1801133 and rs1801131 SNP genotyping.

The basic principle of this technique is to recognize SNP alleles with the high specificity of ligase binding reaction. Then, nonspecific sequences of different lengths were introduced into the end of the binding probe, and a ligase addition reaction was used to obtain the products of different lengths corresponding to the sites. Polymerase chain reaction amplification was performed with universal primers labeled with fluorescence. The amplified products were separated by fluorescence capillary electrophoresis. Finally, the genotypes of each SNP locus were obtained by electrophoretic analysis. Laboratory operation was carried out in strict accordance with the “Laboratory Operation Procedures,” and special personnel are responsible for quality control. Duplicate samples were added to the test samples to test the consistency rate of typing (the consistency rate of typing >99.9% is required).

SPSS 17.0 software was used for statistical analysis. The quantitative data in line with the normal distribution were expressed as mean and standard deviation. And quantitative data that do not conform to the normal distribution were expressed by the median (P25, P75). Categorical variables such as genotypes were expressed in proportions. The chi-test was used to check the Hardy-Weinberg equilibrium (HWE). The relationship between T2DM and different genotypes was analyzed using logistic regression. Five different models (genotype model, recessive model, dominant model, homozygote model, heterozygote model) of genetic analysis were employed and then stratified by age, gender, and BMI. Odds ratio (OR) and 95% confidence interval (95% CI) were used for expressing the results. The test level was p < 0.05.

According to the Systematic Reviews and Meta-Analyses (PRISMA) guidelines [18], a systematic literature searching of databases including PubMed, Web of Science, and CNKI was conducted to collect articles published before to July 1, 2022. The search terms are as follows: (“type 2 diabetes mellitus” or “T2DM”) and (“MTHFR” or “C677T” or “rs1801133”) (PRISMA Checklist has been provided in online suppl. Fig. 1; for all online suppl. material, see https://doi.org/10.1159/000545731).

Articles were included according to the following criteria: (1) they are observational studies (including cohort and case-control studies); (2) have two groups (T2DM group and control group); (3) the frequencies of genotypes for the rs1801133 polymorphism in the T2DM group and the control group had been provided. Articles were excluded according to the following criteria: (1) repetitive studies, (2) inadequate data (without genotype data), (3) articles with a Newcastle-Ottawa Scale (NOS) score below 5.

The search results were screened sequentially by two independent investigators (HQ and CGM) according to the predefined inclusion/exclusion criteria. The basic data extracted included the year of publication, name of the first author; ethnicity of study; sample sizes of the T2DM and control groups; and the frequency of each variant of the MTHFR C677T gene polymorphism. Substantial differences between the two investigators were resolved by the group members. The quality of the literature was assessed according to Newcastle-Ottawa Scale (NOS) [19]. The NOS ranges from 0 to 9, and a higher score represents a higher quality of the study.

Stata 17.0 software was used for statistical analysis. The chi-test was used to check the HWE. OR and 95% CI were used for expressing the association between the MTHFR C677T gene polymorphism (rs1801133) and T2DM. p ≤ 0.05 indicated statistical significance. In meta-analyses, a fixed-effect model was used if I² ≤ 50%, a random-effects (RE) model was used when I² ≤ 50% [20]. Five different models of genetic analysis (dominant model, recessive model, homozygote model, heterozygote model, allele model) were employed in meta-analyses. The presence of publication bias was detected by Funnel plots and Egger’s test.

Demographic and clinical information for the T2DM group and control group is shown in Table 1. There were no significant differences between the T2DM group and the control group in age, gender, and WHR (p = 0.392, 0.639, and 0.521, respectively). Compared with the control group, the BMI and FPG were higher in the T2DM group (t/Z = 6.216, and 15.818, respectively; all p < 0.001).

The genotype frequencies of rs1801133 SNPs GG, GA, and AA were 28.3%, 51.2%, and 20.4% in the control group and 25.1%, 45.7%, and 29.2% in the T2DM group. The genotype distributions rs1801133 in the control group and in the T2DM group conformed to HWE (p = 0.730 and 0.150, respectively). In genotype model, there was no significant difference in genotype between the T2DM and control group in GG vs. GA (p > 0.05); there was a significant difference in genotype between the T2DM and control group in GG vs. AA (OR = 0.620; 95% CI: 0.450–0.855; p = 0.004). After adjustment for age, gender, and BMI there was also a significant difference in genotype between the T2DM and control group in GG vs. AA (OR = 0.639; 95% CI: 0.461–0.885; p = 0.007). There was no significant difference in the recessive model between the T2DM and control group (p > 0.05). There were significant differences in the dominant model, homozygote model, and heterozygote model between the T2DM and control group (OR = 1.605, 95% CI: 1.229–2.095, p = 0.001; OR = 0.620, 95% CI: 0.450–0.855, p = 0.004; OR = 0.625, 95% CI: 0.470–0.830, p = 0.001). After adjustment for age, gender, and BMI there were still significant differences in the dominant model, homozygote model, and heterozygote model between the T2DM and control group (OR = 1.557, 95% CI: 1.202–2.067, p = 0.001; OR = 0.635, 95% CI: 0.458–0.880, p = 0.006; OR = 0.634, 95% CI: 0.475–0.846, p = 0.002), shown in Table 2.

The genotype frequencies of rs1801131 SNPs TT, TG, and GG were 66.9%, 29.9%, and 3.3% in the control group and 68.6%, 28.4%, and 3.0% in the T2DM group. The genotype distributions rs1801131 in the control group and in the T2DM group conformed to HWE (p = 0.995 and 0.996, respectively). There was no significant association between rs1801131 SNPs and T2DM in all genetic models (all p ≥ 0.05). After adjusting for age, gender, and BMI, the difference was also not statistically significant (all p ≥ 0.05), shown in Table 3.

In the subgroup analysis by gender, both male and female, compared with the rs1801133 SNPs AA genotype, the GG + GA genotype was associated with an increased risk of T2DM (OR = 1.679, 95% CI: 1.103–2.556, P_adjusted = 0.016; OR = 1.506, 95% CI: 1.056–2.148, P_adjusted = 0.024). In the female subgroup, compared with the GG genotype, the AA genotype was associated with a decreased risk of T2DM (OR = 0.583, 95% CI: 0.379–0.896, P_adjusted = 0.014). In the male subgroup, compared with the GA genotype, the AA genotype was associated with a decreased risk of T2DM (OR = 0.543, 95% CI: 0.348–0.847, P_adjusted = 0.007). In the subgroup of individuals ≥ 65 years old, compared with the AA genotype, the GG + GA genotype was associated with an increased risk of T2DM (OR = 1.681; 95% CI: 1.263–2.239; P_adjusted <0.001). Similarly, in the subgroup of individuals ≥65 years old, compared with GG or GA genotype, the AA genotype was associated with a decreased risk of T2DM (OR = 0.571, 95% CI: 0.404–0.807, P_adjusted = 0.002; OR = 0.608, 95% CI: 0.448–0.825, P_adjusted = 0.001). In the subgroup of 18.5 kg/m² ≤ BMI <24 kg/m² and BMI ≥24 kg/m², compared with the AA genotype, the GG + GA genotype was associated with an increased risk of T2DM (OR = 1.613, 95% CI: 1.062–2.449, P_adjusted = 0.025; OR = 1.544, 95% CI: 1.074–2.220, P_adjusted = 0.019); compared with the GA genotype, the AA genotype was associated with a decreased risk of T2DM (OR = 0.594, 95% CI: 0.379–0.930, P_adjusted = 0.023; OR = 0.662, 95% CI: 0.450–0.974, P_adjusted = 0.036). In the subgroup of BMI ≥24 kg/m², compared with the GG genotype, the AA genotype was associated with a decreased risk of T2DM (OR = 0.620, 95% CI: 0.399–0.964, P_adjusted = 0.034). No associations between genotype and susceptibility were found in other subgroups, shown in Table 4.

In the subgroup analysis by gender, age, and BMI, there was no significant association between rs1801131 SNPs and T2DM in all genetic models (all p ≥ 0.05), shown in Table 5.

According to the search strategy, a total of 62 articles were retrieved through the above database. After reading the full text, 55 articles were excluded and 7 articles were included (Fig. 1). These articles contained 1,585 subjects in the T2DM group and 1,631 subjects in the control group. Among the seven included studies, four articles involved Asians, and three articles involved Africans. All of the included articles conformed to HWE (p ≥ 0.05). And the quality of included articles was high, with NOS scores being more than 6, shown in Table 6.

In the overall analysis, significant heterogeneity was detected in the recessive model (GA + AA vs. GG: I² = 89.84%, p < 0.01) and allele model (G vs. A: I² = 88.38%, p < 0.01) (Table 5). There was no significant heterogeneity in the overall analysis of the dominant model (GG + GA vs. AA: I² = 35.09%, p = 0.16), homozygote model (A vs. GG: I² = 51.29%, p = 0.09), and heterozygote model (AA vs. GA: I² = 45.70%, p = 0.10) (Table 5). There was no significant association between rs1801133 SNPs and T2DM in all genetic models under RE models (GA + AA vs. GG: OR = 1.47, 95% CI: 0.86–2.54; AA vs. GG: OR = 1.43, 95% CI: 0.89–2.28; G vs. A: OR = 1.32, 95% CI: 0.91–1.91) and FE models (GG + GA vs. AA: OR = 0.83, 95% CI: 0.65–1.05; AA vs. GA: OR = 1.19, 95% CI: 0.93–1.53). We performed an ethnic subgroup analysis to account for heterogeneity. Subgroup analysis showed that there was a significant association in the recessive model (GA+AA vs. GG: I² = 76.52%, p < 0.01; OR = 2.27, 95% CI: 1.16–4.44) under RE models in Asians, shown in Table 7 and Figure 2.

Sensitivity analysis was performed after the one-by-one removal of the articles. And the results were similar to the non-sensitivity analysis. This meta-analysis did not assess publication bias because of the small number (<10) of articles included in this meta-analysis. The statistical power was low and the funnel chart was not sufficient to identify any publication bias.

Defective insulin secretion or impaired biological action and genetic factors are important factors in the pathogenesis of T2DM. Diabetes is a metabolic disease characterized by high blood sugar. Hyperglycemia is caused by defective insulin secretion, impaired biological action, or both [1]. Genome-wide association studies have confirmed dozens of susceptibility genes associated with T2DM [26]. The MTHFR C677T gene produces a functional MTHFR enzyme. This enzyme helps regulate homocysteine levels. If the MTHFR C677T gene polymorphism has a heterozygous or homozygous variant, the enzyme slows down which may lead to increased homocysteine levels [11]. Zhang et al. [27] found that high levels of homocysteine could be spontaneously modified to cysteine sites of insulin receptor precursor proteins via disulfide bonds. This post-translational modification disrupts the maturation of insulin receptors in the endoplasmic reticulum and Golgi apparatus, eventually leading to a significant reduction in the mature form of insulin receptor proteins. The transmission of insulin signals is directly inhibited in the first step, resulting in severe insulin resistance. These results suggest that MTHFR rs1801133 is involved in the occurrence and development of T2DM.

In the case-control study of this article, a significant association was found between rs1801133 and T2DM risk. AA genotype was found to be associated with a reduced risk of T2DM which is consistent with Pathak’s study [12]. After subgroup analysis and adjustment, GG/GA genotypes were found to be associated with an increased risk of T2DM, and AA genotypes were found to be associated with a decreased risk of T2DM in the subgroup of both males and females, individuals ≥65 years old, 18.5 kg/m² ≤ BMI <24 kg/m², and BMI >24 kg/m². Schwammenthal’s study found that vitamins often serve as carriers or coenzymes to help homocysteine complete the metabolic process [28, 29]. In addition, Thomas’s study found that vitamin intake was related to BMI [29, 30]. Whether BMI plays a role in the relationship between rs1801133 and T2DM needs to be further studied. In contrast to our findings, several studies have shown that polymorphism in rs1801133 was not associated with T2DM. Pirozzi’s study of 47 patients with T2DM and 78 controls found that there was no association between MTHFR C677T polymorphism in the development of T2DM [31]. The study by Chang et al. [23] found that MTHFR polymorphisms might play a role in the pathogenesis of T2DM in Caucasians, but it does not apply to Taiwanese. The differences between studies may be due to the small sample size of each study, population heterogeneity, or the small effect size of this genetic polymorphism on T2DM [32]. In the case-control study, there was no significant association between rs1801131 SNPs and T2DM in all genetic models. The difference was also not statistically significant after adjusting for gender, age, and BMI. This is consistent with a lot of previous studies [23, 25].

Due to conflicting results from multiple studies [33‒36], a large meta-analysis was performed in this study on all available case-control studies to obtain a more accurate estimate of association. Our meta-analysis was based on 7 studies involving 1,585 patients with T2DM and 1,631 controls. Results showed that rs1801133 polymorphism was not significantly associated with T2DM risk in all genetic models. Most of these models have heterogeneity, so subgroup analysis was made by ethnicity. Heterogeneity was not detected in most genetic patterns in Asians but was detected in all genetic patterns in Africans. Subgroup analysis showed that the GG genotype may be a risk factor for T2DM susceptibility in Asians but not Africans. Accordingly, we found that rs1801133 was associated with T2DM risk in the Chinese population. These findings advocate for ethnicity-specific screening strategies in diabetes prevention programs, particularly prioritizing Asian populations with elevated genetic risk. The results of this case-control study and meta-analysis suggest that the effect of rs1801133 polymorphism on T2DM may be ethnicity-dependent, which is consistent with the study by Meng et al. [34]. There may be several reasons for the different associations between the rs1801133 gene and T2DM among different ethnicity. First, the differences in genotype distribution among different ethnicity may affect the association between the rs1801133 gene and T2DM [20, 37]. Second, T2DM showed genetic heterogeneity in different populations [38]. Finally, differences in sex, age, BMI/medical history, disease severity, lifestyle habits, etc. may influence the association between genes and disease [39‒41]. Physical activity level influences the MTHFR gene expression in diabetic patients [42]. The study by Ma et al. [15] reported MTHFR rs1801133 seems to synergy between smoking and diabetes susceptibility. This gene-environment interaction highlights the urgency for tobacco control policies targeting genetically susceptible communities. However, most relevant studies do not provide genotype frequencies (such as sex, age, and BMI) for each layer.

This study is the first article to explore the relationship between rs1801133 and T2DM using a case-control study combined with meta-analysis. Strict inclusion and exclusion criteria were established for meta-analysis, and the results were reliable. However, this case-control study and meta-analysis still have some limitations. First, behavioral lifestyle variables associated with T2DM were not taken into account in the case-control study, such as smoking, alcohol consumption, and physical activity. Second, the case-control study did not focus on complications of T2DM, which may limit the clinical value of rs1801133 in assessing prognosis. Third, heterogeneity exists in most genetic models in the whole population. The heterogeneity of African subgroup studies remains high when subgroup analysis by ethnicity. Heterogeneity may be influenced by gene-environment interactions. Most of the included studies did not analyze the correlation between rs1801133 and T2DM based on environmental factors, so it was difficult to explore the source of heterogeneity. Addressing these limitations through multiethnic studies with environmental monitoring will be critical for developing equitable public health interventions. In conclusion, our results suggest that rs1801133 polymorphism have an ethnic-dependent effect on T2DM and are associated with susceptibility to T2DM in Asians. This underscores the necessity of allocating public health resources based on population-specific genetic profiles. However, the association observed in Asians needs to be further validated in a large case-control study.

The authors would like to thank all participants who voluntarily participated in the study.

This study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of Anhui Medical University, Hefei, China (No. 2020H011). Written informed consent was obtained to participate in this study.

The author reports no conflicts of interest in this work.

This study was funded by the Anhui Natural Science Foundation (No. 1808085QH252), the Research Project of Anhui Provincial Department of Education (2024AH051941; gxgnfx2022039), and the Anhui Provincial Health Science and Technology Research Project (2024Bac3003). The funders have no role in the conduct of the research or the decision to submit this report.

Study concept and design: GuiMei Chen and DongMei Zhang; acquisition of data: WeiWei Chang, YuanYuan Wang, HeQiao Zhang, JiaMou Zhou, HuiYan Shen, LinSheng Yang, and DongMei Zhang; analysis and interpretation of data: Qian Huang; drafting of the manuscript: Qian Huang, WeiWei Chang, and YuanYuan Wang; and critical revision of the manuscript for important intellectual content: GuiMei Chen. All authors have reviewed and approved the final manuscript for submission.

Qian Huang, WeiWei Chang, and YuanYuan Wang contributed equally to this work.

The data that support the findings of this study are not publicly available due to their containing information that could compromise the privacy of research participants but are available from the corresponding author on reasonable request.