Abstract: Objective  To explore the association between CNTNAP2 rs 2710102 polymorphism and autism spectrum disorder(ASD)  by performing a metaanalysis. Methods  Databases,  such as China Biology Medicine disc(CBMdisc),  China National Knowledge Infrastructure,  WanFang data,  CQVIP,  PubMed,  Cochrane Library,  Ovid,  Wiley Online Library, Elsevier Science Direct and Spinger were  searched for the casecontrol studies on the association between the CNTNAP2 gene rs 2710102 polymorphism and ASD.  Stata 12.0 software was used to calculate the odds ratio (OR) and 95% confidence interval(CI),  as well as heterogeneity and publication bias analyses. Results  A total of 7 publications were included in this metaanalysis,  involving a total of 1689 ASD cases and 14227 controls. Pooled analyses showed no association between CNTNAP2 rs2710102 polymorphism and ASD risk using allelic model (C vs T,  OR=1,  95%CI: 0.84, 1.18),  heterozygote model (CT vs CC/TT,  OR=1.01,  95%CI: 0.88,1.17),  dominant model (TT vs CT/TT,  OR=1.09,  95%CI: 0.92,  1.304) and recessive model (CC vs TC/TT,  OR=0.98,  95%CI:0.77，1.07). No heterogeneity was observed in every model suggesting the results were stability,  and there was no evidence of publication bias. Conclusion  CNTNAP2 rs2710102 polymorphisms are not associated with ASD.

Key words: autism spectrum disorder, CNTNAP2 gene, polymorphism, metaanalysis