Eight studies of air pollution in U.S. cities have now shown that fine
particles (the invisible soot emitted by incinerators, automobiles,
power plants and heating units) are presently killing about 60,000
Americans each year. More than a dozen studies have, in one way or
another, confirmed this relationship. Furthermore, there appears to be
no threshold, no level below which effects disappear. This means that
people are being killed by air pollution levels well within existing
To summarize bluntly, any increase in fine particles in the atmosphere
kills someone. The victims remain nameless, but they have been deprived
of life all the same. Mere compliance with federal standards does not
protect the public. Any increase in the number of small particles in
the air elevates the death rate. This has obvious implications for
certain technologies: incinerators and fossil-fuel-powered machines
(automobiles and trucks, power plants and heating units). To protect
public health, these technologies must be avoided, or fitted with
expensive control equipment, or replaced by cleaner alternatives.
People have known for a long time that particles in the air can kill.
In 1952, a dense smog killed 4000 people during one week in London, and
since then no one has doubted the cause-and-effect relationship. The
question, therefore, isn't whether airborne particles can harm humans,
but rather, how much pollution causes how much damage, and, secondly,
is there a threshold, an amount below which no effects are seen?
Throughout the '50s and '60s, complacent authorities assumed there was
a threshold --some amount that was safe. However, after 1975, a
revolution took place in scientific understanding of fine particles and
health. In 1979, the National Research Council of the National Academy
of Sciences, and the United Nations, both published book-length
studies of the dangers of small particles to humans. Here is the
current view: humans evolved in an environment where dust was made up
of large particles. Humans therefore evolved means for protecting
themselves against large particles. Large particles are filtered out by
hairs inside the nose, mucous membranes in the throat and airways, and
other mechanisms. However, modern combustion machines produce small
particles which pass right by these natural protections and then enter
the deep lung. In the deep lung, air comes into contact with a person's
blood stream; this is where oxygen passes into the body and carbon
dioxide passes out with each breath we take. Putting tiny particles of
pollution directly in contact with the surface of the deep lung is a
recipe for trouble. Because of their origin in combustion processes,
most fine particles are coated with toxic materials -- metals like lead
and mercury, or toxic organics like polycyclic aromatic hydrocarbons
(PAHs). So fine particles provide a uniquely efficient carrier, giving
dangerous toxins direct entry into the blood stream.
Armed with new knowledge, in 1987, U.S. Environmental Protection Agency
(EPA) established new, stricter standards for particles in the air. The
1987 standard, which governs today, is expressed in terms of small
particles (also called particulate matter) that measure 10 micrometers
or less in diameter. (A meter is 39 inches and a micrometer is a
millionth of a meter.) These are called respirable or inhalable
particles because, as we saw above, they are small enough to get into
the deep lung where they cause various kinds of damage. The shorthand
way to refer to these pollutants is PM10 (meaning Particulate Matter 10
micrometers or less in diameter). Current U.S. standards say that the
ambient air (the general air we all breathe) may contain no more than
50 micrograms of PM10 particles per cubic meter of air as an annual
average, and the one-day average should exceed 150 micrograms per cubic
meter only one day each year. (A gram is 1/28th of an ounce and a
microgram is a millionth of a gram.)
Since 1987, evidence has been accumulating, showing that the 1987
standards do not protect human health. The question about the existence
of a PM10 threshold was addressed first by Joel Schwartz of U.S.
Environmental Protection Agency (EPA). Schwartz reviewed data on air
pollution and deaths from London, 1958-1972, and showed there was no
threshold down to the lowest observed levels of air pollution. A
study published last month in the NEW ENGLAND JOURNAL OF MEDICINE, of
six U.S. cities, including several that are not heavily polluted, such
as Portage, Wisconsin and Topeka, Kansas, shows death rates increasing
with just 15 micrograms per cubic meter of PM10 pollutants. In all,
at least 8 studies have now shown that PM10 at any level kills people.
It seems clear there is no threshold.
A study of people in Steubenville, Ohio, showed that each increase of
100 micrograms per cubic meter of total suspended particles (of which
PM10 represents about half) is associated with a 4% increase in the
death rate, with no threshold. Interestingly, the Steubenville study
showed that the death rate changes as fine particle levels change, but
not as sulphur dioxide levels change.
In Philadelphia, a close relationship between PM10 pollutants and the
death rate was observed. Once again sulfur dioxide levels did not
correlate with the death rate, but particle concentrations did. Here
each increase of 100 micrograms per cubic meter of total suspended
particles (of which PM10 makes up half) was associated with a 7%
increase in the death rate. There was no threshold.
A study of people in Detroit showed that a 6% increase in the death
rate was associated with each increase of 100 micrograms per cubic
meter of total suspended particles (of which PM10 makes up half).
There was no evidence of a threshold. Sulfur dioxide levels were not
significantly associated with increases in the death rate.
Studies of St. Louis, Missouri and Kingston, Tennessee, showed that
the death rate increased 16% (St. Louis) and 17% (Kingston) with each
addition of 100 micrograms per cubic meter of PM10 pollutants to the
air. Associations with gaseous pollutants -- sulfur dioxide, nitrogen
oxides and ozone --did not come close to achieving statistical
In the Utah Valley, a study of the population of Provo revealed that
the daily death rate was closely associated with levels of PM10
pollution. The Utah Valley is unique because PM10 is the only
pollution present there in significant quantities (contributed chiefly
by a steel mill). For every increase of 100 micrograms per cubic meter
of PM10 pollutants, there was a 16% increase in the death rate, and no
threshold was observed.
In all cities, the increase in deaths was most notable among people
older than 65 and in people with chronic obstructive pulmonary disease
(COPD) or cardiovascular disease.
There is a remarkable consistency apparent in all these studies: a 100
micrograms per cubic meter increase in PM10 is always accompanied by an
8% to 17% increase in the death rate. Joel Schwartz, the only EPA
employee ever given a "genius award" by the MacArthur Foundation, re-
analyzed data from London's 1952 killer smog and showed that the death
rate increased 6.4% for each increase of 100 micrograms per cubic meter
total suspended particles, or about 13% for each 100 micrograms per
cubic meter increase in PM10 pollutants--again, remarkably consistent
with the other studies.
No epidemiological study can prove a cause and effect relationship
because it is always possible that some key factor was not considered.
Until last month, skeptics could say smoking might explain why death
rates increase as PM10 concentrations increase. But the study published
last month in the NEW ENGLAND JOURNAL OF MEDICINE looked at 8111 adults
in six American cities and showed that smoking did not explain the
increased death rate observable when PM10 concentrations rise.
Smoking has now been ruled out.
Joel Schwartz recently quoted the British researcher, Bradford Hill,
saying, "All scientific work is incomplete... All scientific work is
liable to be upset or modified by advancing knowledge. That does not
confer upon us a freedom to ignore the knowledge we already have, or to
postpone the action that it appears to demand at a given time." Then
Schwartz added: "At this given time, the knowledge we already have
seems to demand a reduction in population exposure to airborne
 Seven studies are reviewed by Joel Schwartz, "Particulate Air
Pollution and Daily Mortality: A Synthesis," PUBLIC HEALTH REVIEWS
1991/1992 Vol. 19 (1992), pgs. 39-60. For the 8th, see footnote 5. The
60,000 figure is taken from "Air Pollution in Typical U.S. Cities
Increases Death Risk," press release dated May 13, 1991, from the
Harvard School of Public Health, Boston, Mass. describing findings
later reported in Joel Schwartz and Douglas W. Dockery, "Increased
Mortality in Philadelphia Associated With Daily Air Pollution
Concentrations," AMERICAN REVIEW OF RESPIRATORY DISEASE Vol. 145
(1992), pgs. 600-604. Two million deaths occur in the U.S. each year;
according to Schwartz and Dockery, fine particles account for 3%. See
also, Michael Weisskopf, "Particles in the Air Help Kill 60,000 a Year,
Study Says," WASHINGTON POST May 13, 1991, pg. A13.
 National Research Council, AIRBORNE PARTICLES (Baltimore:
University Park Press, 1979).
 United Nations, FINE PARTICULATE POLLUTION (NY: Pergamon Press,
 Joel Schwartz and Allan Marcus, "Mortality and Air Pollution in
London: A Time series Analysis," AMERICAN JOURNAL OF EPIDEMIOLOGY Vol.
131 (1990), pgs. 185-194.
 Douglas Dockery and others, "An Association Between Air Pollution
and Mortality in Six U.S. Cities," NEW ENGLAND JOURNAL OF MEDICINE Vol.
329 (1993), pgs. 1753-1759; see also pgs. 1807-1808.
 Joel Schwartz and Douglas Dockery, "Particulate Air Pollution and
Daily Mortality in Steubenville, Ohio," AMERICAN JOURNAL OF
EPIDEMIOLOGY Vol. 135 (1992), pgs. 12-19; see also pgs. 20 and 23 for
discussion of the Steubenville study.
 Philadelphia study cited in note 1, above.
 Joel Schwartz, "Particulate Pollution and Daily Mortality in
Detroit," ENVIRONMENTAL RESEARCH Vol. 56 (1991), pgs. 204-213.
 Douglas W. Dockery and others, "Air Pollution and Daily Mortality:
Associations with Particulates and Acid Aerosols," ENVIRONMENTAL
RESEARCH Vol. 59 (1992), pgs. 362-373.
 C. Arden Pope III and others, "Daily Mortality and PM10 Pollution
in Utah Valley," ARCHIVES OF ENVIRONMENTAL HEALTH Vol. 47 (1992), pgs.
Descriptor terms: air pollution; morbidity statistics; mortality
statistics; fine particles; particulates; fossil fuels; coal; oil;
natural gas; automobiles; trucks; soot; smoke; exhaust; electric power;
steubenville, oh; oh; philadelphia, pa; pa; detroit, mi; mi; st louis,
mo; mo; kingston, tn; tn; provo, ut; ut; utah valley, ut; pm10; joel
schwartz; douglas dockery; bradford hill; regulations; ambient air