Introduction

Rheumatoid arthritis (RA) is a most common systemic and autoimmune disorder characterized by synovial inflammation, erosive articular degradation, and joint destruction, leading joint deformity and movement limitation [1]. Although the main etiology of RA is yet to be elucidated, studies demonstrated that both environmental and genetic risk factors may play a major role in the onset and progression of the disease [2]. The prevalence of the disease is almost 0.5–1% among the world population and is more frequently occurred in women than in men (sex ratio 3:1) [3, 4]. According to the most recent studies, genetic factors may contribute to approximately 60% of the total risk in susceptibility to RA [5]. The class II Human leukocyte antigen (HLA) has been identified as the most powerful genetic factors in susceptibility to RA [6, 7]. On the other hand, a number of non-HLA genes, such as Cytotoxic T lymphocyte-associated protein 4 (CTLA4), TNF receptor associated factor 1 (TRAF1), Protein tyrosine phosphatase non-receptor type 22 (PTPN22), Peptidyl arginine deiminase 4 (PADI4), TNF alpha induced protein 3 (TNFAIP3), and Methylenetetrahydrofolate reductase (MTHFR) have been consistently associated with RA predisposition [8,9,10,11,12,13].

The folate biological function is to provide methyl groups required for metabolic processes, such as DNA methylation, synthesis, and repair. Therefore, folate deficiency can disrupt these processes [14]. MTHFR, an essential enzyme in folate homeostasis and metabolic pathway, catalyzes the irreversible reduction of 5,10-methylenetetrahydrofolate into 5-methyltetrahydrofolate, which serves a methylation group toward the conversion of homocysteine to methionine with the precursor of S-adenosylmethionine (SAM) [15]. Reduced function of MTHFR results in hypomethylation of DNA and enhanced levels of homocysteine, which leads to increased secretion of proinflammatory cytokines that are implicated in the pathogenesis of RA [16]. Therefore, genetic variations of MTHFR could impress the susceptibility to RA. On the other hand, genetic variants of the MTHFR gene have been associated with hyperhomocysteinemia, which has been considered as risk factor for cardiovascular diseases in RA patients [17].

MTHFR gene is located on the chromosomal region 1p36.3, and several single nucleotide polymorphisms (SNPs) within this gene have been recognized in susceptibility to diverse autoimmune disorders [18]. Two common polymorphisms C677T (rs1801133) and A1298C (rs1801131) have been recognized. These polymorphisms are correlated with a reduction in MTHFR enzyme activity. The C677T mutation in the MTHFR gene causes alanine (A) to valine (V) amino acid substitution. This situation is an important cause of reduced enzyme activity and leads to a higher level of homocysteine levels [19]. Besides, the A to C change at position 1298 leads to glutamine to alanine substitution, resulting in decreased enzyme activity [20].

Several studies evaluated the associations between MTHFR polymorphisms (C677T and A1298C) and susceptibility to RA in different populations with variable frequency and conflicting results. This discrepancy might be owing to different sample sizes, ethnicity, clinical heterogeneity, and publication bias. To compensate these limitations, we conducted this most updated meta-analysis to investigate whether MTHFR gene polymorphisms play a role in RA proneness.

Methods

This meta-analysis follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [21]. The current study is registered in the International Prospective Register of Systematic Reviews (PROSPERO).

Literature search

A comprehensive search was conducted through MEDLINE and Scopus databases to retrieve all potential publications considered the association between MTHFR (C677T and A1298C) gene polymorphism and susceptibility to RA. The following combinations of key words were searched: (“MTHFR” OR “methylenetetrahydrofolate reductase”) AND (“Rheumatoid Arthritis” OR “Arthritis” OR “RA”) AND (“single nucleotide polymorphism” OR “SNP” OR “polymorphisms” OR “mutation” OR “variation”). The reference list of all studies was cross-checked to find other potential studies which might miss during initial search.

Inclusion and exclusion criteria

We included studies in quantitative analysis if they met the following criteria: (i) all observational studies (cohort or case-control design) considered the association between MTHFR (C677T and A1298C) gene polymorphism and susceptibility to RA; (ii) studies reporting sufficient data to extract or calculate risk estimates with 95% CI; (iii) studies with sufficient information regarding numbers or genotype frequencies in cases and healthy controls. Duplicates, book chapters, letters to editor, animal study, case reports, review articles, and studies with repetitive subjects all were excluded. The application of these criteria recognized 16 and 14 eligible studies for C677T and A1298C SNPs, respectively.

Study selection criteria

The results of initial search were exported to Endnote software. Then two reviewers independently assessed titles and abstracts of all studies. Articles which not follow the eligibility criteria were excluded according to a hierarchical approach. The full-text examination was examined if we could not decide to include or exclude studies based on title and abstract. In particular conditions, if an author has published more than one study by the same case series, the most recently published study was included. Any disagreements were discussed and resolved by consensus.

Data extraction

The detailed data of all eligible studies were extracted according to a standardized extraction form including: the author’s name, journal and year of publication, country of origin, ethnicity, mean or range of age, genotyping method, allele frequency of cases and controls, total sample size of cases and controls, and the number of cases and controls for each genotype. Any discrepancy between two reviewers was solved by mutual discussion.

Quality assessment

Methodological quality of eligible studies was evaluated by Newcastle–Ottawa Scale (NOS), a validated scale for non-randomized studies in meta-analysis [22]. Accordingly, studies were categorized to high quality (7–9), intermediate quality (4–6), and low quality (1–3). Furthermore, chi-square tests were calculated to disclose potential deviation from Hardy–Weinberg equilibrium (HWE) for distribution of the allele frequencies in the control group.

Statistical analysis

The data analyses were carried out using STATA (version 14.0; Stata Corporation, College Station, TX) and SPSS (version 23.0; SPSS, Inc. Chicago, IL). In this study, crud OR and its 95% CI was calculated to estimate the association between MTHFR gene polymorphism and the risk of RA. Genotype models were defined as follows: MTHFR C677T, [dominant model (TT+CT vs. CC), recessive model (TT vs. CT+CC), allelic model (T vs. C), homozygote (TT vs. CC), heterozygote (CT vs. CC)] and MTHFR A1298C, [dominant model (CC+AC vs. AA), recessive model (CC vs. AC+AA), allelic model (C vs. A), homozygote (CC vs. AA), heterozygote (AC vs. AA)]. In consideration of the possible heterogeneity (between-study variability) across included studies, chi-square Q-test was used [23]. Additionally, to show heterogeneity quantitatively, the other index (I2) was calculated. There was significant heterogeneity if I2 values exceeded 50% or the Q statistic had a P value less than 0.1. In the presence of significant heterogeneity, the random-effects model (REM) (DerSimonian–Laird approach) was performed. Otherwise, the fixed-effects model (FEM) (Mantel–Haenszel approach) was performed for combination of data [24, 25]. We used sensitivity analysis to assess the stability of our results. Additionally, the Egger’s test and Begg’s test were applied for publication bias [26, 27].

Results

Study characteristics

The four-phase search and screening process based on the PRISMA statement is presented in Fig. 1. After the removal of duplicates (47 studies), 290 studies were remained. Of these, 247 studies were excluded based on title and abstract and 43 studies excluded by full-text evaluation. Ultimately, 16 studies were qualified and included in quantitative analysis (Fig. 1) [28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43]. All included studies were conducted between 2004 and 2019 and had good methodological score. RFLP-PCR and Taq-man genotyping methods were used by most of included studies. Tables 1 and 2 summarized the characteristics and genotype frequency of the included studies.

Fig. 1
figure 1

Flow diagram of study selection process

Full size image
Table 1 Characteristics of studies included in meta-analysis of MTHFR gene C677T polymorphism
Full size table
Table 2 Distribution of genotype and allele among RA patients and controls
Full size table

Quantitative analysis

Meta-analysis of C677T and RA risk

Overall, 15 studies with 2165 patients and 1751 healthy controls were included in quantitative analysis of MTHFR gene C677T polymorphism and RA risk. The analysis of overall population revealed a significant positive association between C677T SNP and RA risk across dominant model (OR = 1.28, 95% CI, 1.01–1.65, P = 0.04), recessive model (OR = 1.47, 95% CI, 1.15–1.89, P < 0.001), allelic model (OR = 1.28, 95% CI, 1.06–1.56, P = 0.01), but not TT vs. CC model (OR = 1.31, 95% CI, 0.98–1.75, P = 0.07), and CT vs. CC model (OR = 1.12, 95% CI, 0.96–1.32, P = 0.15) (Fig. 2). Additionally, subgroup analysis was performed to evaluate ethnicity-specific effect on the association of C677T SNP and RA risk. Since there was only one study for Jewish, African-American and Mexican ethnicity these studies were excluded from subgroup analysis. The results indicated that the risk of RA in African population increases under all genotype models while these results were repeated in Asian population just for recessive model (OR = 2.18, 95% CI, 1.40–3.41, P < 0.001), allelic model (OR = 1.32, 95% CI, 1.06–1.64, P = 0.01), and TT vs. CC model (OR = 2.07, 95% CI, 1.16–3.69, P = 0.01). No significant association was detected in causations (Fig. 3). Furthermore, we stratified studies according to genotyping method and found that utilization of RFLP-PCR method can affect the results of association between C677T SNP and RA risk (Fig. 4).

Fig. 2
figure 2

Pooled OR and 95% CI of individual studies and pooled data for the association between MTHFR gene polymorphism and RA risk in recessive model (a C677T and b A1298C)

Full size image
Fig. 3
figure 3

Pooled OR and 95% CI of individual studies and pooled data for the association between MTHFR gene polymorphism and RA risk in different ethnicity subgroups and overall populations for dominant model (a C677T and b A1298C)

Full size image
Fig. 4
figure 4

Pooled odds ratio (OR) and 95% confidence interval of individual studies and pooled data in different genotyping method subgroups for dominant model (a C677T and b A1298C)

Full size image

Meta-analysis of A1298C and RA risk

In the quantitative analysis of MTHFR A1298C polymorphism and RA risk, 14 studies containing 2021 patients and 1760 healthy controls were included. The pooled results revealed a significant association in overall population across recessive model (OR = 1.58, 95% CI, 1.21–2.06, P < 0.001) and CC vs. AA model (OR = 1.34, 95% CI, 1–1.78, P = 0.04) (Fig. 2). The results of subgroup analysis by ethnicity rejected potential association between A1298C SNP and RA risk in all three populations (Africans, Asians, Causations) (Fig. 3). Moreover, subgroup analysis based on genotyping method highlighted the influence of RFLP-PCR on the association between A1298C SNP and RA risk (Fig. 4).

Evaluation of heterogeneity and publication bias

During the meta-analysis of MTHFR gene polymorphism, evidence of moderate heterogeneity was detected in some models. However, partial heterogeneity was resolved while the data were stratified by genotyping method and ethnicity. Publication bias was evaluated by funnel plot, Begg’s test, and Egger’s test. Subsequently, there was no obvious evidence of asymmetry according to the funnel plots (Fig. 5), and all P values of Begg’s test and Egger’s test were > 0.05, indicating no evidences of publication biases.

Fig. 5
figure 5

Begg’s funnel plot for publication bias test (a C677T and b A1298C)

Full size image

Sensitivity analysis

The leave-one-out method was used in the sensitivity analysis to explore the effect of individual data on the pooled ORs (Table 3). The significance of ORs was not altered through omitting any single study in the dominant model for C677T and A1298C SNPs, indicating that our results were statistically robust (Fig. 6).

Table 3 Main results of pooled ORs in meta-analysis of MTHFR gene polymorphisms
Full size table
Fig. 6
figure 6

Sensitivity analysis in present meta-analysis investigates the single-nucleotide polymorphisms of MTHFR contribute to risk for RA (a C677T and b A1298C)

Full size image

Discussion

Dysregulated function of MTHFR may lead to hypomethylation of DNA and hyperhomocysteinemia, both have been implicated in RA etiopathogenesis [44], conferring MTHFR as a candidate RA predisposing gene. Several MTHFR gene SNPs have been identified [18], and genome-wide association studies (GWASs) have identified this gene to be associated with genetic susceptibility to RA [45]. In addition, numerous studies have reported that MTHFR gene C677T and A1298C polymorphisms might be contributing genetic factor to RA risk. Although several studies in different ethnic groups have tried to divulge the plausible association between MTHFR gene C677T and A1298C SNPs and RA risk, the observations are still controversial and an apprehensive meta-analysis seems to be indispensable to disclose the bona fide association of these variations with RA disease. Hence, this meta-analysis was performed to reveal the approximation of MTHFR gene C677T and A1298C SNPs and RA risk. According to the pooled analysis, both polymorphisms increased the risk of RA.

MTHFR plays a vital role in the folate metabolism and is involved in the conversion of 5,10-methylenetetrahydrofolate into 5-methyltetrahydrofolate [46]. As a structural analog of folic acid, methotrexate (MTX) and sulphasalazine (SSZ) has been prescribed for the treatment of RA according to its well-known efficacy and less toxicity [47]. These drugs competitively inhibit dihydrofolate reductase (DHFR), an enzyme that participates in the tetrahydrofolate synthesis and leads to increased plasma homocysteine level. However, polymorphisms in the genes of the folate pathway are potentially responsible for interpatient variation in MTX and sulphasalazine efficacy. Studies demonstrated patients homozygous for the mutation in the MTHFR gene had significantly higher baseline homocysteine and heterozygous MTHFR genotype induced significantly higher plasma homocysteine compared with no mutation. Taken together, these findings illustrate the critical role of MTHFR polymorphisms in increasing homocysteine levels [48, 49].

There are two meta-analyses about these hypothesis; Fisher et al. [50] showed that the C677T polymorphism (not A1298C polymorphism) is associated with increased toxicity (OR 1.71, 95% CI 1.32–2.21, P < 0.001). In consistent with them, Song et al. revealed a significant association between the MTHFR C677T and A1298C polymorphisms and MTX toxicity. They suggested that modulation in the activity of MTHFR enzyme as the result of SNPs in the MTHFR gene might be expected reason [51,52,53].

Studies have disclosed the role of MTHFR gene C677T and A1298C polymorphisms in reduced activity of the enzyme [54], which may impress the RA pathogenesis through three major pathways. First, dysfunction of MTHFR enzyme may cause DNA hypomethylation, which has been implicated to be involved in the pathogenesis of RA [55]. Second, MTHFR plays a role in donation of methyl group for methylation of homocysteine to generate methionine; hence, dysregulation of MTHFR might lead to high homocysteine level [16], which has been observed in RA patients compared with healthy subjects [56]. In addition, homocysteine has been shown to trigger the proinflammatory transcription factor of nuclear factor (NF)-κB, which has been shown to be involved in inflammatory settings in RA [44]. Third, GWASs has reported that the chromosomal region harboring the MTHFR gene is associated with genetic susceptibility to RA risk [57, 58]. Hence, it is rational to study the RA pathogenesis with respect to genetic diversity and impaired function of MTHFR.

In this most recent systematic review and meta-analysis, 15 studies with 2165 patients and 1751 healthy controls were included in quantitative analysis of MTHFR gene C677T polymorphism and RA risk as well as 14 studies containing 2021 patients and 1760 healthy controls were included for evaluation of MTHFR gene A1298C polymorphism and RA risk. In addition, we performed a subgroup analysis with respect to the technique applied for genotyping (RFLP-PCR), which had not been performed in the previous meta-analyses. Our analysis indicated a significant positive association between C677T SNP and RA risk in the dominant model (OR = 1.28), recessive model (OR = 1.47), allelic model (OR = 1.28), but not TT and CT genotypes. The results indicated that the risk of RA in African population was increased under all genotype models while these results were repeated in Asian population just for recessive model (OR = 2.18), allelic model (OR = 1.32), and TT genotype (OR = 2.07). However, no significant association was observed in the causations. The analysis of another SNP, namely A1298C, revealed a significant association in overall population according to the recessive model (OR = 1.58) and CC genotype (OR = 1.34). However, the results of subgroup analysis by ethnicity did not show association between A1298C SNP and RA risk in all three populations, including Africans, Asians, and Causations. That notwithstanding, the previous meta-analysis, applying less studies and including less patients and controls, indicated that the recessive model of MTHFR gene C677T polymorphism was not associated with RA risk [59], unlike our meta-analysis that indicated a statistically significant association of this polymorphism with RA risk. Moreover, only the recessive model of MTHFR gene A1298C polymorphism was associated with RA risk in the late meta-analysis that was associated with RA risk in recessive model and CC genotype in the current meta-analysis. This difference originates from the number of studies and subjects included in each meta-analysis.

In this meta-analysis, the subgroup analysis was also conducted based on genotyping technique to further explore the influence of genotyping method on the associations. Subgroup analysis according to the genotyping method indicated that RFLP-PCR method could impress the results of association between MTHFR gene A1298C and C677T SNPs and RA risk. Therefore, the differences in the genotyping method may have significant impressions on the overall observations of the association between MTHFR gene A1298C and C677T SNPs and RA susceptibility.

Despite our attempt to carry out the best possible meta-analysis and conclusion of the available information, a number of limitations can be raised toward this meta-analysis. First, it was not possible to analyze the role of gender, age, lifestyle, drugs, and other genetic variations on the adjusted association of MTHFR gene polymorphisms and RA risk. Therefore, further research in respective of the gene–gene and gene–environment interactions is still required to come up with a more apprehensive estimation of MTHFR gene polymorphisms with RA risk. Second, we only searched the publications with English language. Third, the number of studies available for the subgroup analysis based on the ethnicity was fairly small. This issue did not allow to perform comprehensive subgroup analysis across all populations.

In conclusion, the results of pooled analysis supported significant association between MTHFR gene C677T SNP and RA risk in the dominant, recessive, and allelic model, but not TT and CT genotypes. Nonetheless, the analysis of A1298C SNP revealed a significant association in overall population according to only the recessive model and CC genotype. However, the results of subgroup analysis by ethnicity did not show association between A1298C SNP and RA risk in all three populations, including Africans, Asians, and Causations. The major drawback of the valid estimation of the association of MTHFR gene SNPs with RA risk in this meta-analysis stems from the insufficient amount of original data, which requires further investigations in the future. Furthermore, the role of life style, age, and gender should be taken into consideration in the stratification analyses.