基于北京市第3次群众体育调查数据,采用定序logistic回归模型分析了体育锻炼对北京市户籍居民健康自评的影响。研究表明:(1)参加体育锻炼对健康自评有正向影响,然而是否是体育人口与健康自评无显著关系。(2)拆分体育人口的3个要素后发现,体育锻炼的频度对健康自评具有显著作用,从每周锻炼1~2次起,直至增加到每天锻炼2次,健康自评均显著高于每周不足1次的居民,且每天锻炼2次的居民健康自评最好;每次锻炼的时间为0.5~1 h、1~1.5 h的居民健康自评均显著高于0.5 h以下居民,而1.5 h以上锻炼时间的居民与0.5 h以下居民无显著差异;锻炼的主观自评强度与健康自评结果呈负向关系。结果提示,保持每周不少于1~2次的多次锻炼,且每次控制在0.5~1.5 h的小强度锻炼更有益于健康。 Based on the third mass sports survey data in Beijing, the effect of physical exercise on the health self-assessment of household permanent residents in Beijing was analyzed using a logistic regression model. The research shows that: (1) Participation in physical exercise has a positive effect on self-assessment of health, but whether or not there is no significant relationship between physical education and health self-assessment. (2) After splitting the three elements of the physical population, we found that the frequency of physical exercise has a significant effect on the health self-assessment, from 1 to 2 times per week until 2 times a day, Residents who were less than 1 times a week were better than themselves, and their health self-evaluation was the best among residents who exercised twice a day. The exercise self-assessment each time was 0.5-1 h and 1-5 h Residents, while those who exercised for more than 1.5 hours had no significant difference with those under 0.5 hours. The intensity of subjective self-evaluation of exercise and self-evaluation of health were negatively correlated. The results suggest that maintaining less than 1 or 2 times a week for multiple exercises, and each control in the 0.5 ~ 1.5 h of small-intensity exercise more beneficial to health.