Asthma hospitalization in children and air pollution: effect modification by sex.

Epidemiological studies have shown associations between air pollution and asthma hospitalization. However, the effect in susceptible and vulnerable population, as children, and the effect modification by sex are still to be explored. The objective of this study was to evaluate the relationship between exposure to air pollution and pediatrics (0-5 years-old) asthma hospitalizations in the public health system in São Paulo city, from 2008 to 2012, stratified by sex. This is an ecological time series study. Daily records of pollutants (PM10, NO2, O3), minimum temperature and humidity average were obtained from the São Paulo State Environmental Agency. Hospital admissions were obtained from SUS database. Generalized linear Poisson regression model of delay up to 7 days, with third-degree polynomial was used to assess the effect of pollutants on the outcome. An interquartile range increase in PM10 (19 µg/m3) was associated with 2.03% (95%CI: 1.10-2.97) increase in children total asthma admissions and 3.90% (95%CI: 1.39-6.43) for girls, after four days of exposure. An interquartile increase in O3 (49 µg/m3) was associated with a 2.40% (95CI%: 1.36-3.44) increase in children total asthma admissions after six days of exposure, and 3.92% (95%CI: 1.30-6.53) for boys in the same day of the exposition. For NO2, an interquartile range (39 µg/m3) increases in 3.08% (95CI%: 1.95-4.20) children total asthma admissions after five days of exposure; 4.34% (95%CI: 1.75-6.94) for girls after four days of exposure, and a 3.61% (95%CI: 1.08-6.14) for boys in the day of exposure. The accumulated effect for three days NO2 exposure was 4.85% (95%CI: 4.45-5.26) for girls and 5.94% (95%CI: 5.50-6.38) for boy’s asthma admissions. Others pollutants did not show accumulated effect. Boys were more susceptible for NO2.exposure effects This study showed that the worsening of air quality was associated with asthma hospitalization with differences between boys and girls.