The Relationship Between Padi4 -94G/A Polymorphism and Rheumatoid Arthritis: A Meta-analysis in Multi Ethnic Groups

Background: A number of studies have reported the association between peptidylarginine deiminase 4 (PADI4) -94G/A polymorphisms and rheumatoid arthritis (RA) risk in different populations, however, the results remained inconclusive. Objecitve: We therefore aim to address this association by performed an updated meta-analysis in multi ethnic groups. Methods: The PubMed and Chinese related databases were searched up to January 2019. The strength of association between PADI -94G/A polymorphism and RA susceptibility was assessed with odds ratios (ORs) and 95% condence intervals (CIs). Results: A total of 22 studies with 14514 RA cases and 21138 controls were nally included in the analysis. Six ethnic groups such as China, Japan, USA, UK, Sweden and Spain were contained. In the overall population, it revealed that PADI -94G/A polymorphism was signicantly associated with an increased risk of RA. In the subgroup analyses by ethnicity, signicant association was found in China as well as in Japan and USA. Conclusions: This meta-analysis demonstrates that the PADI4 -94G/A polymorphisms may represent a signicant risk factor for RA in China, in Japan and USA. Further studies are needed to clarify this nding, since most available studies were conducted among Chinese and Japanese in this study.


Background
Rheumatoid arthritis (RA) is one of the most common chronic systemic in ammatory diseases that cause joint destruction. In the United States, the overall prevalence of de nite RA among adults is approximately 10 per 1,000 (1.0%); rates for women are approximately 2.5 times higher than rates for men [1]. As the world's population ages and the proportion of people over the age of 60 increases, the prevalence of RA has also increased. The etiology and pathogenesis of RA is not fully understood.
However, it has been hypothesized that the greatest impact of increased prevalence of RA is not due to the outcome of aging, but may be associated with genetic and environmental predisposing factors [2][3][4].
Peptidylarginine deiminase 4 (PADI4) -94G/A (rs2240340) has been identi ed as one of the RA susceptibility single nucleotide polymorphism, the results, however, are not frequently reproducible. In retrospect, individual studies with small sample sizes that are known to have low statistical power and yielded poor replication record. Moreover, this lack of reproducibility might also stem from discrepant lifestyle backgrounds. In order to reduce the in uence of the diverse backgrounds, we performed a metaanalysis to assess the relationship between PADI − 94G/A polymorphism and RA risk in multi ethnic groups respectively.

Search strategy and selection criteria
We performed the meta-analysis according to the guidelines of the PRISMA group. The PubMed and Chinese related databases were searched for studies on the relationship between PADI − 94G/A polymorphism and the risk of RA. The search keywords were (PADI4 or peptidylarginine deiminase 4 or -94G/A) and rheumatoid arthritis. The last search was updated on January, 2019. Additional records were identi ed by manual searching. Inclusion criteria: (1) studies on the association between PADI − 94G/A polymorphism and RA, (2) independent cohort or case-control studies in humans, (2) providing su cient genetypes data in cases and controls. Exclusion criteria: (1) duplicate literatures, (2) incomplete data, (3) no controls, (4) review articles.

Data Extraction
Two reviewers extracted the data independently based on the inclusion criteria. Disagreements were resolved by a discussion. The following information was collected from each quali ed study: rst author's name, publication year, ethnicity, sample size, and available genotype information from PADI − 94G/A polymorphism. Titles and abstracts of all potentially relevant articles were screened rstly. Full articles were then scrutinized if the title and abstract were ambiguous.

Statistical analysis
The strength of association between PADI − 94G/A polymorphism and RA susceptibility was assessed with odds ratios (ORs) and 95% con dence intervals (CIs). The distributions of genotypes in controls were tested by Hardy-Weinberg equilibrium (HWE) using the Chi-square test. Four comparisons were performed: (1) allelic contrast, (2) contrast of homozygotes, (3) recessive, and (4) dominant models.
Statistical heterogeneity was measured by Chi-squarebased Q-test. A xed effects model was used when there was no heterogeneity among these studies; otherwise, the random-effects model was used. The signi cance of the pooled ORs were evaluated by a Z-test. Sensitivity analysis was performed by comparing the results of xed-effects model and random-effects model. Begg's funnel plot and Egger's linear regression test were conducted to assess the publication bias. All statistical analyses were conducted using the Stata, version 12 (StataCorp LP, College Station, TX). A P value less than 0.05 was considered to be statistically signi cant.

Results
Description of included studies Figure 1 illustrates the literature search process with a ow chart. Two hundred and twenty-ve articles which examined the association between PADI polymorphisms and RA were identi ed. According to the inclusion and exclusion criteria, eighteen article (22 studies) with PADI − 94G/A and RA [5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22] were nally included in the analysis. The publication year of involved studies ranged from 2003 to 2014. In total, 14514 RA cases and 21138 controls were included in this meta-analysis. Six ethnic groups such as China, Japan, USA, UK, Sweden and Spain were included in our study. The pooled analysis was not performed for Sweden and Spain due to only one study respectively. The characteristics of these included studies are provided in Table 1.

Meta-analysis Results
The primary results of this meta-analysis on the association between PADI − 94G/A polymorphism and RA risk are shown in Table 2. In the overall population, it revealed that PADI − 94G/A polymorphism was signi cantly associated with an increased risk of RA in four models ( Table 2, Fig. 2).

Padi − 94g/a Polymorphism With Ra In China And Japan
Ten studies including 2783 cases and 2887 controls identi ed an association between the PADI − 94G/A polymorphism and RA risk in China [5][6][7][8][9][10][11][12][13][14], as well as three studies including 2969 cases and 2093 controls in Japan [15][16][17]. It revealed that PADI − 94G/A polymorphism was signi cantly associated with an increased risk of RA both in China and Japan among all analysis models (

Padi − 94g/a Polymorphism And Ra In Usa
Four studies including 1325 cases and 1174 controls identi ed an association between the PADI − 94G/A polymorphism and RA risk in USA [18][19]. Meta-analysis revealed that PADI − 94G/A polymorphism was signi cantly associated with an increased risk of RA in allelic contrast, in homozygotes and dominant models (

Padi − 94g/a Polymorphism And Ra In Uk
Three studies determined the relationship between PADI − 94G/A polymorphism and RA risk in UK [20][21]. The total sample size for patients with RA and controls was 5691 and 13732, respectively. It revealed that PADI − 94G/A polymorphism was not associated with RA in UK (

Sensitivity Analysis And Publication Bias Diagnosis
To evaluate the sensitivity of this meta-analysis, we compared the consistency between xed-effects model and random-effects model. All the analysis results were not materially altered except the dominant model in USA (Table 2). Hence, the pooled results in this meta-analysis are relatively stable and credible.
The Begg's funnel plot and Egger's test were performed to evaluate the publication bias. As showed in Fig. 3, the shape of the funnel plot did reveal obvious asymmetry. Similarly, the Egger's test showed that there was obvious publication bias in all the included studies (t = 4.29, p = 0.000, Fig. 4).

Discussion
Although the multifactorial nature of RA is well known, genetic factors are considered to be strong determinants of these diseases, thus encouraging researchers to search for the responsible genes. PADI4 has been implicated in the pathogenesis, activity and severity of RA [23][24]. Since the rst positive association between PADI − 94G/A and RA was reported in a Japanese population [15], a number of studies have reported the association between PADI − 94G/A polymorphisms and RA risk in different populations. Till now, there are several published meta-analyses regarding PADI4 -94G/A polymorphisms and RA risk [25][26][27][28][29][30]. Of these, two meta-analyses reported that there was signi cant association between PADI4 -94G/A polymorphism and RA risk both in Asian and European population [26,28], while two metaanalyses reported that there was signi cant association only in Asian individuals [27,29]. However, only one meta-analysis was conducted in a separate ethnic group [30]. Therefore, we performed this metaanalysis to assess the relationship between PADI − 94G/A polymorphism and RA risk in multi ethnic groups respectively.
Our meta-analysis involved 22 studies with 14514 RA cases and 21138 controls. The results showed that a signi cantly elevated risk of RA was associated with all variants of PADI4 -94G/A in the overall analysis. In the subgroup analyses by ethnicity, signi cant association was found in China as well as in Japan and USA. Compared with the previous meta-analyses [25][26][27][28][29][30], the current study involved more research in multi ethnic groups. And the effects of gene-environment interactions with respect to RA risk were also conducted by each separate ethnic group analysis. The sensitivity analysis con rmed the reliability and stability of the meta-analysis. Therefore, the ndings from our meta-analysis provide a strong evidence for the association between PADI4 -94G/A polymorphism and RA in the three contries. It also indicated that the relationship between PADI4 -94G/A polymorphism and RA might be susceptible in different ethnicity.
Nevertheless, there are several limitations to this meta-analysis. First, because the papers searched in our study were limited to those openly published, it is possible that some non-published literature that may meet the inclusion criteria were missed. Second, non-English/Chinese articles and databases were not reviewed in our meta-analysis, thus might introduce some bias. Third, due to the relative small sample size of some studies or lack of necessary information, we did not perform further subgroup analyses. Finally, publication bias existed in our analysis, which indicated that other language studies should be included.
In conclusion, this meta-analysis demonstrates that the PADI4 -94G/A polymorphisms may represent a signi cant risk factor for RA in China, in Japan and USA. Ethnicity seems to play an important role in the genetic association of the disease. Further studies are needed to clarify this nding, since most available studies were conducted among Chinese and Japanese in this meta-analysis. Figure 1 Flow diagram of the literature search.

Figure 2
The forest plots of all selected studies on the association between PADI -94G/A polymorphism and RA susceptibility under allele model. Egger's linear regression.