OBJECTIVES: Many publications have evaluated the correlation between HLA-B27 polymorphisms and ankylosing spondylitis (AS), with conflicting results. We carried out this new meta-analysis in order to collect all the relevant studies to further clarify the association of HLA-B27 polymorphisms with AS susceptibility. METHODS: Relevant published data were retrieved through Medline, PubMed, Web of Science, CNKI, and the Chinese BioMedical Literature Database on disc. The statistical analysis was conducted using Review Manager Version 5.0 and STATA 11.0. From these data, the odds ratio (OR) with a 95 % confidence interval (95 % CI) was calculated. RESULTS: (1) A total of 38 studies, including 3,410 AS cases and 1,735 healthy controls, were collected in this meta-analysis. (2) Our results showed that B2704 was a risk factor but B2703, B2706, B2707, B2727, B2729, and B2747 may be protective factors for AS worldwide. (3) These subtypes, such as B2701, B2702, B2705, B2708-15, B2717-20, B2723-24, B2733, B2735, B2740, B2746, B2749, and B2767, showed no association with susceptibility to AS. There was a huge difference with previous reports for B2702 and B2705. (4) The B2702, B2704 and B2705 subtypes have existed high heterogeneity but no publication bias. CONCLUSIONS: This meta-analysis in our study suggested that B2704 might be a potential risk factor, however, B2703, B2706, and B2707 might be potential protective factors of AS, especially in Asia.
OBJECTIVES: Many publications have evaluated the correlation between HLA-B27 polymorphisms and ankylosing spondylitis (AS), with conflicting results. We carried out this new meta-analysis in order to collect all the relevant studies to further clarify the association of HLA-B27 polymorphisms with AS susceptibility. METHODS: Relevant published data were retrieved through Medline, PubMed, Web of Science, CNKI, and the Chinese BioMedical Literature Database on disc. The statistical analysis was conducted using Review Manager Version 5.0 and STATA 11.0. From these data, the odds ratio (OR) with a 95 % confidence interval (95 % CI) was calculated. RESULTS: (1) A total of 38 studies, including 3,410 AS cases and 1,735 healthy controls, were collected in this meta-analysis. (2) Our results showed that B2704 was a risk factor but B2703, B2706, B2707, B2727, B2729, and B2747 may be protective factors for AS worldwide. (3) These subtypes, such as B2701, B2702, B2705, B2708-15, B2717-20, B2723-24, B2733, B2735, B2740, B2746, B2749, and B2767, showed no association with susceptibility to AS. There was a huge difference with previous reports for B2702 and B2705. (4) The B2702, B2704 and B2705 subtypes have existed high heterogeneity but no publication bias. CONCLUSIONS: This meta-analysis in our study suggested that B2704 might be a potential risk factor, however, B2703, B2706, and B2707 might be potential protective factors of AS, especially in Asia.