OBJECTIVES: The reported results of the relationship between genetic polymorphisms of tumor necrosis factor (TNF)-α -238, -308 locus and ankylosing spondylitis (AS) susceptibility are not always consistent. This article aims to perform a meta-analysis to collect all the relevant studies to date to further clarify the relationship between those genetic polymorphisms and AS. METHODS: A computer search was carried out up to September 2011 for literature pertaining to AS and TNF-α polymorphisms. RESULTS: Twenty-two articles were included, with 2,506 cases of AS and 3,023 normal controls. We searched for genotypes A allele vs. G allele, AA vs. GG + GA, and GA + AA vs. GG in a fixed/random-effects model. The effect summary odds ratios (ORs) and 95 % confidence intervals (CIs) were obtained, which shows there was no association between genetic polymorphisms and AS. As the heterogeneity was observed, in a subgroup analysis by ethnicity, the degree of risk of two genes with AS susceptibility was similar in populations of European and Asian origin. For the human leukocyte antigen (HLA)-B27+ population, results were not statistically significant. CONCLUSION: ORs of various comparisons indicate that there is no association between TNF-α -238, -308 polymorphisms and AS susceptibility in the overall population and in the subgroup of Asian and non-Asian descent.
OBJECTIVES: The reported results of the relationship between genetic polymorphisms of tumor necrosis factor (TNF)-α -238, -308 locus and ankylosing spondylitis (AS) susceptibility are not always consistent. This article aims to perform a meta-analysis to collect all the relevant studies to date to further clarify the relationship between those genetic polymorphisms and AS. METHODS: A computer search was carried out up to September 2011 for literature pertaining to AS and TNF-α polymorphisms. RESULTS: Twenty-two articles were included, with 2,506 cases of AS and 3,023 normal controls. We searched for genotypes A allele vs. G allele, AA vs. GG + GA, and GA + AA vs. GG in a fixed/random-effects model. The effect summary odds ratios (ORs) and 95 % confidence intervals (CIs) were obtained, which shows there was no association between genetic polymorphisms and AS. As the heterogeneity was observed, in a subgroup analysis by ethnicity, the degree of risk of two genes with AS susceptibility was similar in populations of European and Asian origin. For the humanleukocyte antigen (HLA)-B27+ population, results were not statistically significant. CONCLUSION: ORs of various comparisons indicate that there is no association between TNF-α -238, -308 polymorphisms and AS susceptibility in the overall population and in the subgroup of Asian and non-Asian descent.