Abstract

Objectives: To apply a new method for estimating the association between daily ambient particulate matter air pollution (PM) and daily mortality to data from over 100 United States cities contained in the National Morbidity, Mortality, and Air Pollution Study (NMMAPS) database and to see whether the results from the 90 cities NMMAPS analysis are robust to this different modelling approach. This new method has recently been shown to provide improved estimates for the association between PM and daily mortality when every-day PM data are unavailable. It avoids the need for selecting a lag of PM at which the mortality effects of PM are to be investigated.

Methods: With the aid of analytical methods and databases developed for NMMAPS, Poisson log linear models controlling for long term trends and weather effects were used to estimate the association between PM and mortality for cities in the NMMAPS database using the new method. A two stage Bayesian hierarchical model was then used to combine city specific estimates to form a national average PM mortality effect estimate.

Results: A 10 μg/m³ increase in PM was associated with a 0.12% increment in total mortality and a 0.17% increment in cardiovascular and respiratory mortality. These results are consistent with those found in the NMMAPS analysis.

Conclusions: There is a statistically significant association between short term changes in PM and mortality on average for the cities contained in the NMMAPS database. These findings are further evidence that this widespread pollutant adversely affects public health.