Background

Brucellosis is a quite normal zoonotic infection, which is caused by immediate contact with animals infected with Brucella or its products. Although the number of Brucellosis patients is relatively small, it remains a severe problem and it is very popular locally in most areas of Asia and Africa [1, 2]. Some patients show onset fever, some show fatigue or sweating. Multiple clinical manifestations can be shown by Brucellosis. What’s worse, nontypical clinical presentation brings difficulties to diagnosis. The exact pathogenesis of Brucellosis remains unknown. It has been reported that cell-mediated immunity is considered to play a crucial role in immunity response to Brucellosis infection [3]. CD4+ and CD8+ T lymphocytes are considered as playing a key role in cellular immunity as they can release IFN-λ and activate the functions in macrophages [4, 5]. Additionally, Interleukin-10 (IL-10) is a crucial cytokine contributing to resist inflammation, which makes various biological effects on multiple types of cell. After the infection of Brucella pathogen, IL-10 can lead to the production drawdown of IFN-λ and inhibition of macrophages function [6].

IL-6 is another cytokine, not only involving in the process of inflammation response but also regulating multiple biological processes such as metabolism, regeneration and nervous system [7]. It has been demonstrated that polymorphism can influence the expression of cytokine and play a crucial role in infectious diseases [8, 9].

There are quite a few literatures demonstrating that hereditary factors play crucial parts in the development of Brucellosis [10,11,12,13,14]. Apart from IL-6 -174 G/C polymorphism, a large number of documents have focused on IL-10 promoter polymorphisms, including position − 1082(G → A) (rs1800896), − 819(C → T) (rs1800871) and − 592(C → A) (rs1800872). However, inconsistent results were obtained. In the present study, we carry out a meta-analysis to obtain a more precise estimate of the above polymorphisms. In addition, the present study is a multi-ethnic study because ethnic populations may contribute tremendous influence to genetic polymorphisms of IL-10 and IL-6.

Methods

Search strategy

The present meta-analysis was performed on account of predefined protocol named Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group [15]. Two databases consisting of PubMed and Embase were put into use by searching keywords as follows: (“IL” or “Interleukin”) and “Brucellosis” and “polymorphism”. The expiration date was the end of February, 2020. We reviewed titles and abstracts of all citations and retrieved studies. We did not limit anything of literatures such as its language, regional culture, publication year and sample size. After these studies were received, their corresponding references were also investigated for other possible studies.

Inclusion and exclusion criteria

The inclusion criteria consisted of three items: (a) a case-control research or short communication; (b) assessment of relationship between IL polymorphisms and Brucellosis susceptibility; (c) providing enough data to judge the final result or the present data can predict the final result. Accordingly, other types of research were excluded such as case report, review article.

Data extraction

The whole information and data were independently extracted by first author (Yueyuan Wu, Shuzhou Yin and Xiaochun Jin) and the final information was reviewed by final referees (Xu Chen and Youtao Zhang). The present study is a multi-ethnic study. Different ethnic descents were categorized as African, Asian, and Caucasian. To get a accurate result, three authors (Shuzhou Yin, Yueyuan Wu and Xiaochun Jin) were asked to check all data and information. If they could not reach an agreement, they would check the above data and information again and have a meeting trying to come to an agreement. If the controversial results still existed, the final referees (Xu Chen and Youtao Zhang) would be asked to give the final decision.

Methodological quality assessment

Three authors (Shuzhou Yin, Yueyuan Wu and Xiaochun Jin) were asked to perform methodological quality evaluation according to predefined assessment criteria (Table 1), which was based on the items of Jiang et al. [16]. The scores with the range of 0(lowest) to 18(highest) were based on several assessment items consisting of credibility of controls, matching criteria, diagnostic criteria, genotyping examination, sample size and Hardy-Weinberg equilibrium. The above assessment items were performed based on popular epidemiological methods and Brucellosis special characteristic. Literatures whose scores less than 12 were considered as “low-quality” literatures. However, the literatures with scores equal to or more than 12 were regarded as “high-quality” studies.

Table 1 Detailed evaluation criteria focused on included studies
Full size table

Statistical analysis

Both the OR and 95%CI were estimated to make the assessment of association power in four different genetic models. For position − 174 G/C of IL-6, there were allele comparison (G versus C), homozygote comparison (GG versus CC), recessive model (GG versus GC/CC), and dominant model (GG/GC versus CC). For position − 1082 G/A of IL-10, there were allele comparison (A versus G), homozygote comparison (AA versus GG), recessive model (AA versus AG/GG), and dominant model (AA/AG versus GG). For position − 819 C/T of IL-10, there were allele comparison (T versus C), homozygote comparison (TT versus CC), recessive model (TT versus TC/CC), and dominant model (TT/TC versus CC). For position − 592 loci of IL-10, there were allele comparison (A versus C), homozygote comparison (AA versus CC), recessive model (AA versus AC/CC), and dominant model (AA/AC versus CC). The χ2-test based Q-statistic and I2 statistics were used. If there was no evident heterogeneity, the fixed-effects model would be applied [17]. If not, the random-effects model would be used [18]. Sensitivity analysis was used to evaluate the stability of the results. Funnel plots and Egger’s test were applied to detect the potential publication bias [19]. All statistics were conducted using Stata software (version 12.0; StataCorp LP, College Station, TX, USA).

Results

Eligible studies

A flow chart of the search course is displayed in Fig. 1. Seven literatures were enrolled in our meta-analysis [10, 20,21,22,23,24,25]. The detailed information of seven studies was listed in Table 2.

Fig. 1
figure 1

Flow diagram for identification of eligible studies for this meta-analysis

Full size image
Table 2 Characteristics of studies included in this meta-analysis
Full size table

Quantitative synthesis of data

Detailed results for the relationship between these polymorphisms and Brucellosis susceptibility are displayed in Table 3. P < 0.05 is considered as a significant association. Overall, significant associations were only found in Asian population of − 819 loci under three genetic models as follows: (Allele model: OR = 0.60, 95%CI = 0.44–0.82, P = 0.001) (Fig. 2), (homozygote comparison: OR = 0.24, 95%CI = 0.09–0.62, P = 0.003) (Fig. 3), (recessive genetic model: OR = 0.22, 95%CI = 0.05–0.91, P = 0.036) (Fig. 4).

Table 3 The general results for the association between IL-10, IL-6 polymorphisms with Brucellosis risk
Full size table
Fig. 2
figure 2

Forest plot of IL-10 -819 C/T polymorphism on Brucellosis risk in allele model (T vs. C)

Full size image
Fig. 3
figure 3

Forest plot of IL-10 -819 C/T polymorphism on Brucellosis risk in homozygote model (TT vs. CC)

Full size image
Fig. 4
figure 4

Forest plot of IL-10 -819 C/T polymorphism on Brucellosis risk in recessive model (TT vs. CC/TC)

Full size image

Sensitivity analysis and publication Bias

Sensitivity Analysis was carried out to indicate single literature’s effect on the final result through continuous removal of individual studies under every genetic model. In the present study, our results could not be affected by any study, suggesting its reliability and robustness. Although slightly asymmetrical funnel plots were discovered (P = 0.806), we could not find any distinct publication bias by Egger’s test under allele model (P = 0.509) or recessive model (P = 0.509).

Discussion

Previous studies have explored the association of IL-10 and IL-6 polymorphisms with Brucellosis risk. The present meta-analysis consists of 7 studies for IL-10 and IL-6 polymorphisms. To the best of our knowledge, the present research was the first to investigate the relationship between IL-10 and IL-6 polymorphisms and Brucellosis risk. Our meta-analysis shows that IL-10 -819 loci polymorphism contributes no risk to Caucasian population but may be associated with decreased risk in Asian population. And other polymorphisms of IL-10 and IL-6 are not related with Brucellosis susceptibility. For IL-6 -174 G/C polymorphism, Budak et al. reported in 2007 that GG genotype was more common in patients of Brucellosis than healthy controls (40 cases and 50 controls) and concluded that GG genotype was a risk factor in the development of Brucellosis [20]. However, the other four literatures reported no association between IL-6 -174 G/C polymorphism and Brucellosis susceptibility. We consider sample size as an important factor contributing to the above discrepancy. Small-study bias is not an emerging event when performing genetic association studies [26]. The current result of IL-10 -819 loci polymorphism may be attributed to race. It is well-known that different races have discrepant genotype number and allele frequency. So that it is necessary to conduct subgroup analysis by race. However, we made a conservative conclusion. Only two literatures with 250 cases and 141 controls were enrolled in our study. Considering the small sample size of eligible studies, a new updated study needs to be urgently conducted. Regarding the significant association found at IL10–1082 AA genotyping in Caucasian, we feel that it is likely a false positive, which is based on two reasons. Firstly, the P value approximates 0.05 and is a critical value. Secondly, the P values of other genetic models are 0.217, 0.484, and 0.916(P > 0.05).

Extensive variations have been established in the frequencies of cytokine polymorphism among different healthy population, including the − 1082 G/A polymorphism, which has been investigated most widely in healthy populations. The allele frequency show wide difference in different countries and regions. For example, the prevalence rate of IL-10 -1082 G allele was reported to be 42.5, 48.9 3.8 and 13.0% in Iranians, Norwegians, Japanese and Koreans, respectively [27,28,29,30].

HWE has been demonstrated to be an important genetic equilibrium law. And if the genotype distribution of control population is not conforming to HWE law, selection bias would occur. It has been reported that several factors would contribute to HWE deviation including fixed mating, or allele do not achieve equilibrium condition and other possible factors, which may lead to inaccurate results of meta-analysis. In our meta-analysis, the problem contributes little influence on our results. Because genotype distributions of all studies are consistent with HWE, which improves reliability of our results greatly.

Genome-wide association study (GWAS) is undoubtedly an effective method to find the sequence of gene mutations, accordingly screening the specific SNP associated with disease. GWAS opens the door to the study of complex diseases by comparing patients’ SNP sites with control groups in whole genome, which identifies all variant allele mutation frequencies. It is regrettable that there is rare GWAS on Brucellosis. Sankarasubramanian etal pulished a study titled “Identification of genetic variants of Brucella spp. through genome-wide association studies” and they found special SNPs, which were closely related with human’s specificity and virulence. They also thought that origin of these special SNPs was early and might derive from B. abortus evolution [31]. However, we think this is just a new start in exploring species-specific SNPs of Brucellosis. In the future, we should investigate further in the genomes and their roles of Brucellosis.

Despite a lot of work we have done, there are still disadvantages existing, which should be remarkable here. Firstly, the present research is still a small-sample study and we should be careful in the results. Secondly, some miscellaneous factors including age, sex and other environment factors are not considered and calculated for final estimates. So that the results of our meta-analysis are on the account of unadjusted estimates. We should also be careful in the results.

Conclusion

In conclusion, this is the first meta-analysis which investigates the association between IL-10 polymorphism and Brucellosis risk. In conclusion, IL-10 -819 loci polymorphism contributes no risk to Caucasian population but may be associated with decreased risk in Asian population. And IL-10 -1082 G/A, 592 loci and IL-6 -174 G/C polymorphism are not associated with Brucellosis risk.