Air Pollution and Mortality in Latin America The Role of Education
Loading...
Date
2008
Journal Title
Journal ISSN
Volume Title
Publisher
LIPPINCOTT WILLIAMS & WILKINS
Abstract
Background: People with less education in Europe, Asia, and the United States are at higher risk of mortality associated with daily and longer-term air pollution exposure. We examined whether educational level modified associations between mortality and ambient particulate pollution (PM10) in Latin America, using several timescales.
Methods: The study population included people who died during 1998-2002 in Mexico City, Mexico; Santiago, Chile; and Sao Paulo, Brazil. We fit city-specific robust Poisson regressions to daily deaths for nonexternal-cause mortality, and then stratified by age, sex, and educational attainment among adults older than age 21 years (none, some primary, some secondary, and high school degree or more). Predictor variables included a natural spline for temporal trend, linear PM10 and apparent temperature at matching lags, and day-of-week indicators. We evaluated PM10 for lags 0 and I day, and fit an unconstrained distributed lag model for cumulative 6-day effects.
Results: The effects of a 10-mu g/m(3) increment in lag 1 PM10 on all nonextemal-cause adult mortality were for Mexico City 0.39% (95% confidence interval = 0.131/-0.65%); Sao Paulo 1.04% (0.71%-1.38%); and for Santiago 0.61% (0.40%-0.83%. We found cumulative 6-day effects for adult mortality in Santiago (0.86% [0.48%-1.23%]) and Sao Paulo (1.38% [0.85%-1.91%]), but no consistent gradients by educational status.
Conclusions: PM10 had important short- and intermediate-term effects on mortality in these Latin American cities, but associations did not differ consistently by educational level.
Methods: The study population included people who died during 1998-2002 in Mexico City, Mexico; Santiago, Chile; and Sao Paulo, Brazil. We fit city-specific robust Poisson regressions to daily deaths for nonexternal-cause mortality, and then stratified by age, sex, and educational attainment among adults older than age 21 years (none, some primary, some secondary, and high school degree or more). Predictor variables included a natural spline for temporal trend, linear PM10 and apparent temperature at matching lags, and day-of-week indicators. We evaluated PM10 for lags 0 and I day, and fit an unconstrained distributed lag model for cumulative 6-day effects.
Results: The effects of a 10-mu g/m(3) increment in lag 1 PM10 on all nonextemal-cause adult mortality were for Mexico City 0.39% (95% confidence interval = 0.131/-0.65%); Sao Paulo 1.04% (0.71%-1.38%); and for Santiago 0.61% (0.40%-0.83%. We found cumulative 6-day effects for adult mortality in Santiago (0.86% [0.48%-1.23%]) and Sao Paulo (1.38% [0.85%-1.91%]), but no consistent gradients by educational status.
Conclusions: PM10 had important short- and intermediate-term effects on mortality in these Latin American cities, but associations did not differ consistently by educational level.
Description
Keywords
PARTICULATE MATTER, MEXICO-CITY, TIME-SERIES, SOCIOECONOMIC-STATUS, TEMPORAL PATTERN, HEALTH, PARTICLES, SANTIAGO, PM10, POPULATION