Significant Relationships Concerning Exponential Transmission or Penetration

This paper reviews the statistical evidence linking human mortality with air pollution due to sulfur oxides and particulates. Three types of analyses are discussed: episodic increases in mortality coincident with high pollution, perturbations in mortality in a given city as a time-series in relation to perturbations in air quality, and comparisons of geographic differentials in both pollution and mortality rates. The paper concludes that there are no reliable statistical associations between SO _x and mortality, but that particulates are in some cases associated with excess mortality. Establishing whether the association is in fact causal would require elimination of potential confounding effects such as occupational exposures.