Re: Deaths from fossil fuel burning air pollution

[Jim Muckerheide <jmuckerheide@delphi.com>]

Steve,

As you note, in a prospective study, as a self-selected group, you can't apply
the results outside the population represented. But the results do compare
this large group to itself, presumably against levels of air pollution. It
doesn't reflect responses in children, and is probably not very affected by,
nor representative of, the aged, or people in inner cities, etc. 

But if a biased group of somewhat healthier than average middle-aged
population, that generally has some moderate level of medical care, are
affected by air pollution, that is significant. However, people who are more
at risk (the aged, inner-city or rural groups that may be health-compromised,
and with minimal levels of medical care) could therefore be expected to
express GREATER adverse effects. As you note, just don't take the study as
representing the population as a whole. It does not. 

But it not uninformative (presuming the assessment of pollution-caused adverse
effects with the correlation is credible). 
 
Regards, Jim
============

Steven Dapra wrote:
> 
> Nov. 30
> 
> There has been a lot of electronic ink spilled here (mostly by Drs. Cohen
> and Raabe) over air pollution-induced mortality.  Dr. Cohen initiated the
> debate (on Nov. 17) by discussing a paper by Pope, et al. (Am J Respir Crit
> Care Med 151:669-674;1995).  I would like to present a short quote from the
> Pope et al. paper, and then discuss two uses of this paper elsewhere.
> 
> On page 670, Pope et al. describe their study population, writing:
> 
> "This analysis relied on data for 552,138 men and women drawn from the
> American Cancer Society (ACS) Cancer Prevention Study II (CPS-II), an
> ongoing prospective mortality study of approximately 1.2 million adults
> [citation].  Participants were enrolled by ACS volunteers in the fall of
> 1982.  The resided in all 50 states, the District of Columbia, and Puerto
> Rico, and were usually friends, neighbors, or acquaintances of the ACS
> volunteers.  Enrollment was restricted to persons who were at least 30 yr.
> Of age and who were members of households with at least one individual 45
> yr of age or more."
> 
> This appears to be a self-selected study group.  It certainly is not a
> statistical cross-section of the population of the United States (and
> Puerto Rico), if only because it excludes everyone under the age of 30.
> 
> According to Table 1 (p. 670), the mean age at enrollment was slightly over
> 56 y.  Participants were 56 % female, 94 % white, and four percent black.
> 
> According to the 1998 Statistical Abstract of the United States, the mean
> age of U.S. residents is 35.2 y.  The population is 51% female.  Whites
> constitute 82 % of the population, and blacks 15 %.
> 
> I am not suggesting the study is no good, I am merely saying it seems to be
> self-selected and that the demographics don't reflect those of the U.S. as
> a whole.  (I have omitted Puerto Rico here because its population is so
> small.)
> 
> On Nov. 30, Prof. Cohen refers RADSAFERs to the book "Particles in Our Air:
>  Concentrations and Health Effects," edited by Richard Wilson and John D.
> Spengler (Harvard University Press, 1996).  In 1996 Wilson was in Harvard's
> Department of Physics and its Center for Risk Analysis, and Spengler was in
> Harvard's School of Public Health.
> 
> On October 18, 2000 Susan Gawarecki made a posting to RADSAFE ("Power Plant
> Pollution Linked to 30,000 Premature Deaths Each Year") referring readers
> to a report ("Death, Disease & Dirty Air") prepared by the Clean Air Task
> Force.  This report (p. 4) claims there are 60,000 premature deaths per
> year in the U.S. because of exposure to fine particles.  The Clean Air
> authors refer readers to "Particles in Our Air" (Wilson and Spengler), page
> 212; however the 60,000 deaths figure is first mentioned on page 210, where
> Wilson and Spengler end a general discussion of air pollution studies by
> writing,  "The consistent picture that a public health authority should
> consider is that 60,000 persons might be dying prematurely of air pollution
> related problems in the United States each year [citation]."  The citation is:
> 
> "Shprentz, D.S. Bryner, G.C., and Shprentz J.S., (1996) "Breath Taking:
> Premature Mortality due to Particle Air Pollution in 239 American Cities"
> National [sic] Resources Defense Council (NRDC), New York, NY."  (This
> should be "Natural" Resources, etc.)
> 
> I obtained a copy of "Breath Taking" (dated May, 1996) and began looking
> for the 60,000 figure.  On page 29, "Breath Taking" begins talking about a
> "landmark study . . . [that] linked ambient air pollution data with
> information from an American Cancer Society [ACS] cohort of 1.2 million
> adults from all 50 states."  The citation is to the Pope, et al. study that
> Profs. Cohen and Raabe have been debating on RADSAFE.
> 
> On page 30, Shprentz, et al. write of the Pope, et al. study, "The
> researchers concluded that modest air pollution exposures are shortening
> the lives of Americans by several years."  Their source for this claim is a
> Harvard School of Public Health press release dated March 9, 1995.
> 
> On page 32, Shprentz, et al. write, "Criticisms of the ACS study are that
> the cohort was not drawn from a random sample of the U.S. population and
> that historical pollution levels were not considered in assessing
> cumulative exposures."  For this, they cite a USEPA "Air Quality Criteria
> for Particulate Matter" draft, dated Nov. 24, 1995.
> 
> On page 44, Shprentz, et al. write that their analysis "is not an
> epidemiological study of the relationship between particle concentrations
> and mortality in U.S. cities." and, that their analysis "merely applies the
> relationship observed in the study of the American Cancer Society cohort to
> current particle concentrations is U.S. cities, to gauge the extent and
> significance of the particle pollution problem."
> 
> So . . . everything - the Clean Air Task Force report, Wilson and Spengler,
> and the American Cancer Society study - all comes back to the self-selected
> cohort in the Pope, et al. study.
> 
> Steven Dapra
> sjd@swcp.com
> 
> ************************************************************************
> The RADSAFE Frequently Asked Questions list, archives and subscription
> information can be accessed at http://www.ehs.uiuc.edu/~rad/radsafe.html
************************************************************************
The RADSAFE Frequently Asked Questions list, archives and subscription
information can be accessed at http://www.ehs.uiuc.edu/~rad/radsafe.html