from the years 2006–2011 on the concentration
of particulate matter in the ambient air (both
PM2.5and PM 10 fractions) were obtained from
the General Inspectorate of Environmental Pro-
tection which is a repository of data collected by
the State Environmental Monitoring agency.
Aggregated annual mortality data for 11 urban
agglomerations in Poland, consisting of cities
with a population above 250,000 inhabitants,
were acquired from the reports of the Department
of Epidemiology of the Center of Oncology (lung
cancer) and from the Central Statistical Office
Local Data Bank (cardiopulmonary and
all-cause mortality).
Exposure to air pollution in each urban
agglomeration (hereafter referred to as ‘city’)
was assessed as the mean 1-h concentration of
PM2.5, subsequently aggregated to 1-year mean
values. The PM2.5 concentration was not
measured at some air monitoring stations in all
relevant years. In those cases, exposure to PM2.
was estimated using a conversion factor deter-
mined from the results of joint PM2.5and PM 10
measurements at the same station (if available) or
different stations located in the same city.
Depending on city, conversion factor ranged
from 0.61 to 0.84.
To calculate the burden of mortality that could
be attributed to PM2.5(population attributable
fraction based on relative risk of mortality arising
from exposure to PM2.5), the exposure-response
functions presented in the report on the European
Perspectives on the Environmental Burden of
Diseases were used (Ha ̈nninen and Knol 2011 ).
The measures of relative risk normalized for unit
exposure to PM2.5in case of lung cancer and
cardiopulmonary diseases were taken from the
results of an American study by Pope
et al. ( 2002 ); those for all-cause mortality were
taken from a report of the World Health Organi-
zation (WHO 2013 ).
3 Results
The annual mean PM2.5 concentration in the
period of 2006–2011 ranged from 14.3 to
52.5μg/m^3 (Fig.1), reaching the highest values
in cities of southern Poland (Cracow, Katowice),
where the emission of air pollutants results pri-
marily from municipal and household sources
(so-called low-stack emission) and from road
transport and heavy industry. The lowest concen-
tration was observed in cities of eastern
(Bialystok, Lublin) and northern Poland
(Gdansk, Szczecin), where the density of emis-
sion sources is lower and there are better climatic
conditions for dispersion of air pollutants. The
mean annual concentration of PM2.5and PM 10
exceeded the limit level according to the CAFE
Directive in 8 and 7 out of the 11 cities, respec-
tively. In 5 of the cities, concentration of PM2.
and PM 10 in 2011 increased in comparison with
that in 2006.
Relative risk of mortality associated with
exposure to PM2.5, calculated for individual
years and urban areas, was used to estimate the
fraction of the population the deaths of who
would have been prevented had the PM2.5con-
centration been reduced to the hypothetical zero
level. The mean values of this population attrib-
utable fraction of all-cause mortality, and mor-
tality from lung cancer and cardiopulmonary
diseases in the years 2006–2011 demonstrate
distribution and variability patterns akin to the
mean concentrations of PM2.5shown in Fig.1,
with the highest value in southern Poland and
lowest in eastern and northern Poland. Data on
the population attributable fraction of mortality
are shown in detail in Table1.
The results of the population attributable frac-
tion and the annual number of deaths from the
health causes above outlined were used to esti-
mate the number of deaths attributable to PM2.
in ambient air. For this purpose the number of
inhabitants living in each city also was used. The
estimate shows that most PM2.5-attributable
deaths occurred in Warsaw, Cracow, and Lodz.
Data on PM2.5-attributable deaths of lung cancer
and cardiopulmonary diseases are shown in
detail in Figs.2 and 3.
Considering the variability of PM2.5level in
different cities, for the sake presentation clarity,
the incidence of deaths attributed to PM2.5was
normalized for 100,000 inhabitants of each city.
Table2 presents the average population of cities
of interest and Fig.4 depicts the incidence of
deaths due to all natural causes, lung cancer,Ambient PM2.5 Exposure and Mortality Due to Lung Cancer and Cardiopulmonary... 11