不同纬度下维生素D受体基因多态性与佝偻病易感性关联性Meta分析  

The association between vitamin D receptor gene polymorphisms and rickets susceptibility:a Meta-analysis

作　　者：张鹏鹏[1] 陈莉莉[1] 李颢[2] 王束玫[1] 

机构地区：[1]山东大学公共卫生学院流行病与卫生统计学研究所,济南250012 [2]山东大学齐鲁医院血液病科,济南250012

出　　处：《中国实用儿科杂志》2012年第3期191-197,共7页Chinese Journal of Practical Pediatrics

摘　　要：目的探讨4个常见维生素D受体(VDR)基因位点(ForkI、BsmI、ApaI、TaqI)单核苷酸多态性(SNPs)与佝偻病易感性关联性并探讨日照量对其的影响。方法制定原始文献的纳入和排除标准及检索策略。检索1980年1月至2011年1月中国期刊全文数据库(CNKI)、中文科技期刊数据库(VIP)、数字化期刊全文数据库(万方)、中国生物医学文献数据库(CBM)及PubMed、CALIS、ScienceDirect数据库中4个位点SNPs与佝偻病易感性关联性关系为研究内容的文献,提取数据。对研究间异质性进行检验并选用适当模型对总人群及高、低纬度两个亚组人群分别进行数据合并,得出合并比值比(OR)及95%置信区间(CI)。探讨研究间异质性来源,最后进行影响性分析及发表偏倚检验。结果共纳入22篇文献。Meta分析结果显示,ForkI位点总人群及亚组人群显性、隐性、共显性遗传模型分析均提示病例f等位基因暴露比低于对照组,差异具有统计学意义。总人群3个遗传模型合并OR(95%CI)分别为0.40(0.25～0.65)、0.41(0.31～0.55)、0.50(0.38～0.67);高纬度人群分别为0.48(0.25～0.93)、0.43(0.23～0.80)、0.55(0.39～0.78);低纬度人群分别为0.38(0.21～0.69)、0.41(0.29～0.57)、0.48(0.34～0.68)。BsmI位点总人群显性、共显性模型及高纬度人群3个遗传模型分析均提示病例B等位基因暴露比高于对照组,差异有统计学意义(P<0.05)。总人群显性、共显性模型合并OR(95%CI)分别为2.15(1.11～4.16)、1.85(1.09～3.16);高纬度人群3个遗传模型合并OR(95%CI)分别为2.86(1.33～6.16)、3.93(1.11～13.89)、2.66(1.33～5.31);低纬度人群差异无统计学意义。ApaI位点仅在总人群隐性模型分析中发现其多态性与佝偻病易感性关联有统计学意义。TaqI位点仅在总人群及高纬度人群隐性模型分析中发现其多态性与佝偻病易感性关联有统计学意义。影响性分析显示Meta分析结果稳定,可信度较高。结论 ForkI位�Objective To assess the relationships between polymorphisms of ForkI,BsmI,ApaI,TaqI in vitamin D receptor gene and the susceptibility to rickets and to investigate whether their associations were dependent upon variants inl atitude by Meta-analysis.Methods A comprehensive search was conducted to identify all studies of the above-mentioned associations.Pooled ORs and their 95%CIs were calculated by either fixed or random effect pooled measure,w hich was selected on the basis of heterogeneity test among studies.Sub-group analysis was performed based on the different latitudes（high/low latitude）.Meta-regression was performed to assess the potentially important covariates exertings ubstantial impact on between-study heterogeneity.An analysis of influence was conducted,which described how robustt he pooled estimator was to removal of each individual study.Publication bias was estimated using modified Eggers linear regression test.Results Compared with thec ontrols,for overall and subgroup populations,theF okI polymorphism had a significantly protective effect on the risk of rickets in dominant,recessive andc odominant models.For overall population thepooled ORs（95%CI）were 0.40（0.25～0.65）,0.41（0.31～0.55）and 0.50（0.38～0.67）respectively.For high latitudea rea population the pooled ORs（95%CI）were 0.48（0.25～0.93）,0.43（0.23～0.80）and 0.55（0.39～0.78）respectively.For low latitude area population the pooled ORs（95%CI）were 0.38（0.21～0.69）,0.41（0.29～0.57）and 0.48（ 0.34～0.68）respectively.The BsmI polymorphism was found contributing significantly to rickets susceptibility for highl atitude area population in dominant,recessive and codominant models and for overall population in dominant and codominant models.For high latitude area population the pooled OR（s95%CI）were 2.86（1.33～6.16）,3.93（1.11～13.89）a nd 2.66（1.33～5.31）respectively.For overall population the pooled ORs（95%CI）were 2.15（1.11～4.16）and 1.85（ 1.09～3.16）respectively

关 键 词：维生素D受体 佝偻病 Meta分析 单核苷酸多态性 

分 类 号：R72[医药卫生—儿科]