Risk Analysis, Vol. 32, No. 5, 2012 DOI: 10.1111/j.1539-6924.2012.01807.x Response Response to Cox Letter: “Miscommunicating Risk, Uncertainty, and Causation: Fine Particulate Air Pollution and Mortality Risk as an Example” Neal Fann, 1 Amy D. Lamson, 1 Susan C. Anenberg, 1 Karen Wesson, 1 David Risley, 2 and Bryan J. Hubbell 1 In his critique of our article, Dr. Cox makes a strong assertion: my co-authors and I have miscommuni- cated the level of uncertainty in our analysis by fail- ing to account explicitly for the possibility that there is no causal relationship between PM 2.5 exposure and premature death—thereby undermining our key claim that PM 2.5 continues to pose a burden to pub- lic health. (1,2) We disagree. We argue here that: (1) there is an expert consensus opinion that there ex- ists a causal relationship between PM 2.5 exposure and premature death; (2) the evidence Dr. Cox offers in support of his view of causality is insufficient; and (3) quantifying the PM 2.5 mortality burden by apply- ing risk coefficients that incorporate varying degrees of certainty regarding causality in fact strengthens our conclusion that “recent levels of PM 2.5 ... pose a nontrivial risk to public health.” (3) In its synthesis of the clinical, toxicological, and epidemiological evidence regarding long-term PM 2.5 exposure and the risk of premature mortality, the U.S. EPA noted in its Integrated Science Assessment (ISA) that “[c]ollectively, the evidence is sufficient to conclude that a causal relationship exists between long-term exposures to PM 2.5 and mortality” (bold in original). (4) This statement reflects both the EPA’s understanding of the current state of the science as well as that of the independent Clean Air Scientific Advisory Committee, comprised of expert health and exposure scientists. (5) In its recent synthesis scientific 1 U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, Research Triangle Park, NC, USA. 2 U.S. Environmental Protection Agency, Office of Atmospheric Programs, Washington, DC, USA. statement, the American Heart Association (AHA) observed that “many credible pathological mecha- nisms ... lend biological plausibility to [the] findings” that PM 2.5 exposure increases the risk of premature death. The AHA concluded that “the overall evi- dence is consistent with a causal relationship between PM 2.5 exposure and cardiovascular morbidity and mortality. (6) Thus, we did not explore the concep- tual and empirical basis supporting a causal relation- ship between PM 2.5 exposure and premature death because we understood this matter to have been re- solved in the literature. Dr. Cox refers to a handful of articles to sup- port his contention that premature death may not be causally related to PM 2.5 exposure. Among these is U.S. EPA (8) which we explore in greater depth later, as well as Ostro et al. (7) and Koop and Tole (2004). (9) As we indicate in the article, we quanti- fied mortality related to long-term exposure to PM 2.5 by applying risk coefficients drawn from an extended analysis of the American Cancer Society (ACS) co- hort (Krewski et al. (10) ), as well as an extended anal- ysis of the Harvard Six Cities (H6C) cohort (Laden et al.). (11) As these are each long-term exposure stud- ies applying the Cox proportional hazards model, the relevance of model specification in the Ostro et al. time series seems less relevant here. We also note that this approach to using ACS and H6C is consis- tent with recent advice from the EPA Science Ad- visory Board in the context of the Agency’s assess- ment of the benefits and costs of the Clean Air Act Amendments of 1990. (12) When referencing the analysis by Koop and Tole (9) , Dr. Cox notes that the study “attempt[s] 768 0272-4332/12/0100-0768$22.00/1 C  2012 Society for Risk Analysis