中文摘要：

目的 分析吸炯对症状性OA患病的影响,探索评价OA患病危险因素的分析方法.方法 调查7 126名山西省农村常住居民,根据是否患有中轴关节及承重大关节的症状性OA相关因素,采用单因素及多因素Logistic回归分析;揭示风险因素之间的交互作用,并深入分析主被动吸烟与其他因素间的相乘交互与相加交互作用;相乘交瓦计算OR值,相加交互计算超相对危险比（RERI）、归因比（AP）、交互作用指数（S）等指标.结果 调查居民7 126名,检出中轴及承重关节OA患者1 480例,患病率20.8％;经全样本多重Logistic回归分析,按α入=0.05水准逐步筛选,结果表明,性别、BMI、吸烟、年龄、土炕取暖、偏食、煤矿井下作业是症状性OA患病的主要影响因素;均衡其他因素影响后,被动吸烟对OA患病有影响;均衡体质量因素后,吸烟未引入分析模型.在各分层样本交互作用分析中,个人吸烟与体质量间存在负向相乘交互作用（OR=0.32,P=0.03）,被动吸烟与体质量及性别间存在相加交互作用[RERI（95％CI）为0.41 （0.01,0.81）＞0,AP（95％CI）为0.25（0.01,0.51）＞0,S=3.04＞1].结论 吸烟对症状性OA患病的影响作用微弱,尚不能认为有统计学意义,推测吸烟对OA的影响作用可能是通过影响体质量而发生的.

英文摘要：

Objective To clarify the effec.t of smoking and related factors in symptomatic osteoarthritis disease,and explore the method to analyze the osteoarthritis （OA） risks.Methods Based on the investigation of 7 126 rural residents in Shanxi province,according to whether patients had related factors of symptomatic osteoarthritis in longitudinal axis arthrosis,the single factor and multiple factors logistic regression analysis were adopted to reveal the interaction between risk factors,and the interaction on multiplicative scale and the interaction on additive scale were both analyzed;interaction on multiplicative scale was estimated by OR,the interaction on additive scale was estimated by the relative excess risk due to interact （RERI）,the attributableproportion due to interaction （AP）,and the synergy index （S）.Results The survey included 7 126 residents,in which 1 480 cases osteoarthritis patients in vertical axis joint with the prevalence of 20.8%.By multiple logistic regression analysis,the whole samples by enter level 0.05 level was stepwisely filtered.The results showed that gender,body mass inlex （BMI）,smoking,age,heatable adobe,picky eater,working under coal mine were main influencing factors for the symptomatic arthritis.Balance other influence factors,passive smoking had an effect on arthritis.Aftcr adjusting for weight,smoking was not included in the final model.In the stratified interaction analysis,there was interaction on multiplicative scale between smoking and weight （OR=0.32,P=0.03）,and the interaction on additive scale between passive smoking and gender or weight exists [RERI （95%CI） 0.41 （0.01,0.81）〉0,AP（95%CI） 0.25（0.01,0.51）〉0,S=3.04〉 1].Conclusion The influence of smoking on osteoarthritis is so weak that it could not be a risk factor and its influence on osteoarthritis is initiated via body.