Association of Chlamydia trachomatis infections with preterm delivery; a systematic review and meta-analysis

Background Premature birth is a primary cause of infant mortality and its etiology varies in different countries. Chlamydia trachomatis (CT) is a common infectious agent transmitted through sexual contact. The purpose of this study is to investigate the connection between CT infections and preterm birth by meta-analysis. Methods All articles published in literature databases including Google Scholar, PubMed, ISI (Web of Science), Biological Abs, IranMedex, SID, and Scopus were investigated. Twenty-four relevant articles, authored betweenm 1998–2014 were analyzed through a random effects model. Heterogeneity of the studies was evaluated by I2 index. The relationship between years of data collection, sample size, and CT infections with preterm delivery prevalence was examined by meta-regression. Data were analyzed with R and STATA [Ver. 12]. Results The overall prevalence of CT infections leading to preterm deliveries was estimated to be 0.13% (CI 95%: 0.11–0.16). The prevalence of CT infections leading to preterm deliveries were calculated based on the study method including PCR [0.06 (CI 95%: 0.04–0.09)], serology [0.23 (CI 95%: 0.10–0.35)] and culture [0.17 (CI 95%: 0.10–0.24)]. Analysis indicates that women with chlamydia infections were 2.28 more likely to deliver pre-term in comparison with those who were not infected. It can be concluded that chlamydia infections increase the risks of preterm delivery, OR = 2.28 (95% CI:1.64–3.16). Conclusions In regard to the results in numerous studies performed on different continents, this meta- analysis showed a clear association between preterm delivery and prior CT colonization. Electronic supplementary material The online version of this article (10.1186/s12884-018-1868-0) contains supplementary material, which is available to authorized users.


Background
Chlamydia trachomatis (CT) is a gram-negative non-motile bacterium, 0.2-1.4 μm in size, that tend to live inside the cylindrical cells of human epithelium [1]. It is the most common infection transmitted through sexual intercourse and those infected may unconsciously transmit it to their sexual partners [2]. In terms of clinical manifestations, colonization with this bacterium by in large is asymptomatic (80% of cases). Patients often do not seek medical treatment up until the symptoms become prominent [3]. Studies have shown that about four million people in the world are infected with CT [4].
Premature birth is a primary cause of infant mortality and the costs of caring for preterm infants are often high. Complications of preterm birth include hyaline membrane disease, necrotizing enterocolitis, air leak syndrome, and intraventricular hemorrhage of the brain. If the new-born survives, they may face convulsions, hearing disorders, or visual disturbances [5]. Rates of premature births vary in different societies, 5-15% in North America and 1.3% in Iran [6,7].
Different factors can lead to the occurrence of preterm births including heart disease, multiple birth, repeated abortions, diabetes, and genitourinary tract infections. Genitourinary infections may be responsible for 25-40% of preterm births. These infections are caused by bacteria such as Mycoplasma urealyticum (MU), Mycoplasma huminis (MH), Neisseria gonorrhea (NG), Trichomonas vaginalis (T. vaginalis), and Chlamydia trachomatis (CT). The role of CT infections in preterm birth has not been well established [8]. Several studies have been conducted on the outcome of CT infections . The aim of this investigation is to evaluate the results of these studies by meta-analysis.

Literature identification
We used PRISMA guidelines for this article [32]. Our data were based on articles that were published in local and international journals. We conducted the literature searches with the following keywords: CT and preterm delivery, CT and preterm labor, CT and preterm birth, CT and premature delivery, CT and prematurity, and CT and premature child birth. All medical literature published in databases including Google Scholar, PubMed, ISI (Web of Science), Biological Abs, IranMedex, SID, and Scopus between, 1986-2014 were chosen. In the first search, 75 manuscripts published between, 1986-2014 were selected. We excluded all duplicate articles, which left 23 for analysis ( Fig. 1) [33,34].

Inclusion and exclusion criteria
All papers with the keywords from above in the title or the abstract were included in our first list and irrelevant articles were removed from the study. The search process was performed by two persons and adjusted one by one. The most important biases in meta-analysis are: publication bias and selection bias. Publication bias was checked using the Begg's test. We reduced the possibility of selection bias by clearly defining the criteria for data collection in each selected study by two researchers independently. These selected studies became the final list.

Data extraction
An information checklist for research papers consisted of first author's last name, year of publication, country where the study was carried out, mean age of participants, sample size, study period, method of bacteria identification, the type of study, and existence of a significant to no significant relationship between CT infections with preterm birth. Studies were excluded if they presented insufficient data, if they were mere reviews, or if they were not epidemiologic studies.

Statistical analysis
Variance of prevalence for CT in each study were calculated as binomial distributions. Studies were combined based on their sample sizes and variance of samples. Due to the heterogeneity of the studies, the random effects model was used when combining studies. To assess heterogeneity, the Cochrane Q test and I 2 statistics were used. A p-value less than 5% was considered a significant heterogeneity test. To examine publication bias, the Begg's test and funnel plot was used. A funnel plot is a graphical detection of publication bias that is actually a bivariate scatter plot (x, y) of sample size versus estimates of effect size.
Subgroup analysis was done according to a diagnostic method, type of study, and continents. Meta-regression was used to explore relationship between prevalence of CT associated with preterm delivery including the year of the study and sample size. To find association between Chlamydia infections and the risk of preterm delivery, an odds ratio (OR) was used and statistical analyses were performed in STATA version 11.1 and metan commend.

Results
We used 23 relevant studies (1986-2014) were included in this meta-analysis (Additional file 1.). The overall relatedness of CT infections with preterm delivery was estimated to be 0.13% (CI 95%: 0.11-0.16) (Fig. 2) (Table 1). Analysis showed that pregnant women with chlamydia infections were 2.28 times more likely to deliver prematurely in comparison to those who were not infected.
Our interpretation of the meta-regression data showed that there was no significant relationship between the prevalence of CT and preterm delivery with the year of study (p = 0.33) and sample size (p = 0.21) (Figs 4 and 5).
According to our analysis as publication bias, there were no significant biases in these studies. In fact, most studies were located inside our funnel plot and the results were included in our analysis (Fig. 6).

Discussion
Chlamydia trachomatis is an intracellular pathogen and immunologic response to CT infection elaborate T cell response and cytokine release. Because inflammatory reactions induce damage in tissue, preterm birth may be the consequence of those inflammatory responses [17,18,25,30,35]. In this meta-analysis, it turned out that there is a significant relationship between CT infections and preterm delivery. Our interpretation of the    meta-regression showed that there was no significant relationship between prevalence of CT infections and preterm delivery with the year of study (p = 0.33) and sample size (p = 0.21). The prevalence of CT infections and preterm delivery in Europe and America was relatively high. Among the articles that were studied, we noted some contradictory results. In Mikhova et al. (2006) using PCR methodology found that there was no relationship between CT infections and preterm birth observed [26]. Another PCR study by William et al. (2000) showed that there was a connection between CT infection and preterm birth [18]. In a cross sectional study using culture technique, Ismail et al. (1992) reported a meaningful correlation between Chlamydia trachomatis infections and preterm birth [10]. In another cross sectional study by Silva et al. (1997) using bacterial cultures observed no correlation between CT infection and preterm birth [12]. Bogavac et al. (2001) showed that a significant relationship between the CT infection and preterm birth was observed in a case-control study using serology [19]. Rimbach et al. (1992) using the same method, showed no connection between preterm birth and Chlamydia trachomatis infections [9]. In regard to the overall results based on studies performed on diverse populations, this meta-analysis showed a clear association between preterm delivery and prior colonization with Chlamydia trachomatis.

Conclusions
In regard to the results of numerous studies performed on different continents, this meta-analysis showed a clear association between preterm delivery and prior CT colonization. Funding This study is funded by Kurdistan University of Medical Sciences. The funding body managed the reviewing proposal of this study and they did not have any role in data collection, analysis, and interpretation of data and in writing this manuscript.

Availability of data and materials
All data generated or analyzed during this study are included in this published article.
Authors' contributions AA carried out the study and collected data. RR supervised the study, participated in designing and conducting it, and prepared the original version of the manuscript. KS and FS conducted research and performed the statistical analysis. NA and AM conducted research and revised the manuscript. All authors studied and approved the content of the present manuscript and participated in revising the paper.
Ethics approval and consent to participate Not applicable.

Competing interests
The authors declare that there are no conflicts of interest regarding the publication of this article.

Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.