Penna M L, Duchiade M P
Instituto de Medicina Social, Universidad del Estado del Rio de Janeiro, Brasil.
Bol Oficina Sanit Panam. 1991 Mar;110(3):199-207.
This study examines the relationship between air pollution, measured as concentration of suspended particulates in the atmosphere, and infant mortality due to pneumonia in the metropolitan area of Rio de Janeiro. Multiple linear regression (progressive or stepwise method) was used to analyze infant mortality due to pneumonia, diarrhea, and all causes in 1980, by geographic area, income level, and degree of contamination. While the variable "proportion of families with income equivalent to more than two minimum wages" was included in the regressions corresponding to the three types of infant mortality, the average contamination index had a statistically significant coefficient (b = 0.2208; t = 2.670; P = 0.0137) only in the case of mortality due to pneumonia. This would suggest a biological association, but, as in any ecological study, such conclusions should be viewed with caution. We believe that air quality indicators are essential to consider in studies of acute respiratory infections in developing countries.
本研究考察了以大气中悬浮颗粒物浓度衡量的空气污染与里约热内卢大都市区因肺炎导致的婴儿死亡率之间的关系。采用多元线性回归（逐步法），按地理区域、收入水平和污染程度分析了1980年因肺炎、腹泻及所有原因导致的婴儿死亡率。虽然在与三种婴儿死亡率类型相对应的回归分析中纳入了“收入相当于两个以上最低工资的家庭比例”这一变量，但平均污染指数仅在因肺炎导致的死亡率案例中具有统计学显著系数（b = 0.2208；t = 2.670；P = 0.0137）。这表明可能存在生物学关联，但与任何生态学研究一样，此类结论应谨慎看待。我们认为，空气质量指标在发展中国家急性呼吸道感染研究中是必须考虑的因素。
