Associations Between Interleukin-1 Gene Polymorphisms and Sepsis Risk: a Meta-Analysis

An-qiang Zhang,Wei Pan,Jun-wei Gao,Cai-li Yue,Ling Zeng,Wei Gu,Jian-xin Jiang
DOI: https://doi.org/10.1186/1471-2350-15-8
2014-01-01
BMC Medical Genetics
Abstract:Background: Previous epidemiological studies have presented conflicting evidence regarding associations between interleukin-1 (IL-1) polymorphisms and sepsis susceptibility. We have performed a meta-analysis to evaluate possible associations between IL-1 polymorphisms and sepsis risk.Methods: Eligible literature was retrieved from PubMed, Embase and Web of Knowledge databases until Jun 15, 2013. The pooled odds ratio (OR) and 95% confidence interval (CI) were calculated using random-effects model in the overall and subgroup analysis based on ethnicity, sepsis severity and quality score.Results: Eighteen studies addressing five IL-1 polymorphisms were included in this meta-analysis. For IL-1A-889 (rs1800587) polymorphism, significant association was observed in overall comparison for allelic effect (OR = 1.47, 95% CI = 1.01-2.13, P = 0.04). There were no significant associations between either IL-1B-511 (rs16944) or IL-1B-31 (rs1143627) and sepsis susceptibility in overall or subgroup analyses. For IL-1B + 3594 (rs143634) polymorphism, genotype TT decreased sepsis risk in overall analysis (OR = 0.59, 95% CI = 0.36-0.97, P = 0.04), as well as in Caucasian (OR = 0.57, 95% CI = 0.34-0.95, P = 0.03) and sepsis (OR = 0.55, 95% CI = 0.31-0.97, P = 0.04) subgroup analysis. For IL-1RN VNTR polymorphism, significant association was observed in overall comparison for allelic effect (OR = 1.40, 95% CI = 1.01-1.95, P = 0.04). Furthermore, the effect sizes of IL-1RN VNTR on sepsis risk increased with disease severity (septic shock OR > severe sepsis OR > sepsis OR).Conclusions: Our meta-analysis indicated that IL-1A-889, IL-1B + 3954 and IL-1RN VNTR might be associated with sepsis susceptibility. However, further studies with larger sample sizes and from homogenous populations would be necessary to validate these findings.
What problem does this paper attempt to address?