Meta-analysis Of The Association Of Chlamydia Pneumoniae Infection And Stroke

PrefaceWith the development of society, people's living standard has increased year by year. Chronic diseases are important factors, which affect people's health more and more. Chronic diseases not only cause death, but also lead to disability, which would take the patients a deal of burden of diseases. So it is very important to research their risk-factors for preventing and controlling chronic diseases. Chlamydia pneumoniae is one kind of pathogenic microorganisms of respiratory system, and can cause acute bronchitis and pneumonia. Recently, according to the research, other systems may also be affected by Chlamydia pneumoniae infection besides respiratory system. Chlamydia pneumoniae has become a focus of publics. At present, according to a lot of serological and epidemiological studies in domestic and overseas, Chlamydia pneumoniae infection may lead to artherosclerosis, and may be a risk factor of stroke, but still needs further confirmations.In order to assess the association between Chlamydia pneumoniae infection and stroke, this study, based on literature retrieval, analyzed synthetically a lot of the Chinese and English documents which had been published by Meta-analysis. It provided powerful evidences of prevention and treatment and of etiological factor exploration.Material and Method一,Select the Material(一) Inclusion Criteria of ReportThe design proposal of the internalized reports was case-control study, and accorded with instituted the twelve criteria by Horwitz RI, Feinstein AR. The number of case and matching style were not restricted, and the sample size of case group and control group was found precisely in all reports.(二) Inclusion Criteria of CaseStroke included brain hemorrhage and cerebral infarction, finally diagnosed by internationally accepted diagnostic criteria. The patients who had other diseases and infected recently were excluded. Gender and age of subjects were not restricted.(三) The measurement of biochemical indicatorMicroimmunofluorescence test(MIF):1:64≤IgG≤1:256, or IgA≥1:8, if the association of CpnIgM and stroke was mentioned, IgM≤1:16 must be required.Enzyme Linked Immunosorbent Assay(ELISA): The different experiments used different kit specification standards to assess positive and negative values.二,The Search Strategy(一) Search on the webThe search terms: Chlamydia pneumoniae AND stroke OR cerebral apoplexy OR cerebral infarction OR cerebral accident OR ischemic cerebrovascular disease. Search strategy was MeSH term.The searched database included Cochrane, Medline, Elsevier, Proquest, EMBASE, CNKI, VIP, WanFang technological periodical full-text data base. All databases were searched until December, 2006.(二) Search by Hand-madeThe Journal of China epidemiology, The Journal of hypertension, The Journal of China angiocardiopathy, The Journal of China neuromedicine, The Journal of China chronic disease prevention and control, The Journal of China department of neurology, The Journal of China clinical, Chinese Medical Journal, The Journal of China clinical rehabilitation, Chinese Physician Journal, Chinese Immunological Journal. All Journals and periodicals were searched until December 2006. The academic conference in internet, newspapers digest and unpublished documents were searched by contacting with the authors when they're necessary.三,Statistical MethodsThe 24 reports were analyzed by RevMan4.2 software. If the result of heterogeneity test was negative, fixed effect model would be used. On the contrary, random effect model would be used. Furthermore, calculating the pooled OR value and 95% CI, drawing forest picture and accomplishing funnel plot analysis. Eventually, quantitatively and qualitatively observed and calculated the publication bias.Results一,Description of the report and results extracted from the dataWhen two methods were used together to detect CpnIgA, 14 reports were included, and there were 1455 subjects in case group and 694 subjects were positive, meanwhile, there were 1890 subjects in control group and 699 subjects were positive. When detecting CpnIgG, 23 reports were included, and there were 2276 subjects in case group and 1440 subjects were positive, meanwhile, there were 2696 subjects in control group and 1452 subjects were positive. When detecting CpnIgM, 7 reports were included, and there were 495 subjects in case group and 45 subjects are positive, meanwhile, 483 subjects in control group and 33 subjects were positive.When method of MIF was used to detect CpnIgA, 8 reports were included, and there were 820 subjects in case group and 345 subjects were positive, meanwhile, there were 1170 subjects in control group and 385 subjects were positive; When IgG was detected, 11 reports were included, there were 1081 subjects in case group and 711 subjects were positive, meanwhile, there were 1526 subjects in control group and 907 subjects were positive; When IgM was detected, 5 reports were included, there were 330 subjects in case group and 24 subjects were positive, meanwhile, there were 371 subjects in control group and 17 subjects were positive.The detect method of ELISA was used to detect CpnIgA, 6 reports were included, there were 635 subjects in case group and 349 subjects were positive, meanwhile, there were 720 subjects in control group and 314 subjects were positive; When IgG was detected, 12 reports were included, there were 1195 subjects in case group and 729 subjects were positive, meanwhile, there were 1170 subjects in control group and 545 subjects were positive; When IgM was detected, 2 reports were included, and there were 165 subjects in case group and 21 subjects were positive, meanwhile, there were 112 subjects in control group and 16 subjects were positive. All reports were consistent with the standard of case-control study.二,The results of Meta-analysisWhen two methods were used together to detect CpnIgA, the pooled OR value was 1.91 and 95%CI was 1.36～2.68, and to detect CpnIgG, the pooled OR value was 1.99 and 95%CI was 1.53～2.60.When CpnIgA and CpnIgG were detected with the method of MIF, there was a statistical difference between stroke patients and health adults, respective OR values were 2.04, 1.78 and 95%CI were 1.19～3.48, 1.18～2.70.When CpnIgA and CpnIgG were detected with the method of ELISA, there was a statistical difference between stroke patients and health adults, respective OR values were 1.80, 2.22 and 95%CI were 1.18～2.76,1.57～3.13.When CpnIgM was detected with two methods together and was detected respectively with the methods of MIF and ELISA, there wasn't statistical difference, and respective OR values were 1.37, 1.79, 0.96 and 95%CI were 0.85～2.20, 0.95～3.39 0.47～1.98.The analysis of funnel plots was carried out in the research of IgM, and because of dissymmetry, the publication bias existed. Calculating the bias quantitatively by equation, 211 negative reports ware included so that the results of Meta-analysis of CpnIgA detected with two methods that were used together can be reversed, and 596 negative reports ware included so that the results of Meta-analysis of CpnIgG detected with two methods that were used together can be reversed. Meanwhile, 90 negative reports were included so that the results of CpnIgG detected with the method of MIF could be reversed, and 212 negative reports were included so that the results of CpnIgG detected with the method of ELISA could be reversed. It demonstrated that the influence of publication bias to the results was less.Discussion一,The Association of Chlamydia pneumoniae infection and strokeAt present, the conclusive explanation of pathogenesy of stroke had been produced by some scholars, and infection had become the focus of studies gradually. More and more scholars considered that stroke was related possibly to vasculitis which chronic infection leads to, and this was the base of stroke attacking and developing. But some other scholars obtained the adverse results, and these were disputing still now.In this Meta-analysis, when two methods were used together to detect CpnIgA and CpnlgG, and the method of MIF and ELISA were used to detect respectively CpnIgA and CpnIgG, infection positive ratio in stroke patients was higher than it in health adults, and the pooled OR value exceeded 1. Furthermore, the sample size was so larger that it had a good representation. To some extent, it indicated that the risk of attacking stroke in patients infected by Chlamydia pneumoniae was higher obviously than that in health adults. When two methods were used together, the method of MIF and the method of ELISA were used to detect CpnIgM respectively, the results didn't have statistical difference. The reason might be that the sample size was too smaller to produce the random error.To sum up, in this Meta-analysis, we concluded that the association of Chlamydia pneumoniae infection and stroke was positive in detecting CpnlgA and CpnIgG. However, because stroke was a multifactorial disease, analyzing one factor only, such as Chlamydia pneumoniae infection, was unilateral. Furthermore, Meta-analysis was an observational study, which analyzed those published previously reports. The bias would be caused in each steps by many conditional restriction, so we must be cautious especially to assess the results.二,The quality of reportsThis Meta-analysis included 24 reports which corresponded into the research criteria and published at domestic and overseas. They had an identical diagnostic criteria, and the chance of diagnostic error bias was smaller. They were designed strictly according to criteria of case-control study. On accounts of searching in Medline and Cochrane database, the chance of allocation bias was smaller. Besides Chinese and English, this article didn't include the reports published by another language, so language bias would be existed. Publication bias also could be found in this study by Funnel plots analysis of the association between increased IgM and stroke. Calculating quantitatively publication bias of the analysis according with criteria by equation, and 221, 596, 90 and 212 negative reports should be included respectively so that the results could be reversed. It indicated that publication bias did not affect the results seriously.To sum up, the association of CpnIgA and CpnIgG tites in serum and attacking stroke was positive probably. But the power of case-control study was not well, furthermore, because of the bias in the analytic process, the results of Meta-analysis which described quantitatively the association between Chlamydia pneumoniae infection and stroke should be adopted cautiously, and the results must be verified by more larger-scale epidemiological research.ConclusionThe discovery of Meta-analysis:Increasing CpnIgA and CpnIgG tites were potential risk factors of stroke in Chlamydia pneumoniae infection patients.