Content uploaded by Venediktos Kapetanakis
Author content
All content in this area was uploaded by Venediktos Kapetanakis on Sep 24, 2015
Content may be subject to copyright.
doi: 10.1136/oem.2009.048785
2009 2010 67: 293-300 originally published online October 9,Occup Environ Med
H Ross Anderson, Ruth Ruggles, Kiran D Pandey, et al.
and Allergies in Childhood (ISAAC) AsthmaPhase One of the International Study of
rhinoconjunctivitis and eczema in children:
world-wide prevalence of asthma,
Ambient particulate pollution and the
http://oem.bmj.com/content/67/5/293.full.html
Updated information and services can be found at:
These include:
References
http://oem.bmj.com/content/67/5/293.full.html#related-urls
Article cited in:
http://oem.bmj.com/content/67/5/293.full.html#ref-list-1
This article cites 27 articles, 13 of which can be accessed free at:
service
Email alerting box at the top right corner of the online article.
Receive free email alerts when new articles cite this article. Sign up in the
Notes
http://oem.bmj.com/cgi/reprintform
To order reprints of this article go to:
http://oem.bmj.com/subscriptions go to: Occupational and Environmental MedicineTo subscribe to
group.bmj.com on May 9, 2010 - Published by oem.bmj.comDownloaded from
Ambient particulate pollution and the world-wide
prevalence of asthma, rhinoconjunctivitis and eczema
in children: Phase One of the International Study of
Asthma and Allergies in Childhood (ISAAC)
H Ross Anderson,
1
Ruth Ruggles,
1
Kiran D Pandey,
2
Venediktos Kapetanakis,
1
Bert Brunekreef,
3
Christopher K W Lai,
4
David P Strachan,
1
Stephan K Weiland,
5
the ISAAC Phase One Study Group
ABSTRACT
Objectives To investigate the effect of ambient
particulate matter on variation in childhood prevalence of
asthma, rhinoconjunctivitis and eczema.
Methods Prevalences of asthma, rhinoconjunctivitis and
eczema obtained in Phase One of the International Study
of Asthma and Allergies in Childhood (ISAAC) were
matched with city-level estimates of residential PM
10
obtained from a World Bank model. Associations were
investigated using binomial regression adjusting for GNP
per capita and for clustering within country. For countries
with more than one centre, a two stage meta-analysis
was carried out. The results were compared with
a meta-analysis of published multi-centre studies.
Results Annual concentrations of PM
10
at city level
were obtained for 105 ISAAC centres in 51 countries.
After controlling for GNP per capita, there was a weak
negative association between PM
10
and various
outcomes. For severe wheeze in 13e14-year-olds, the
OR for a 10
m
g/m
3
increase in PM
10
was 0.92 (95% CI
0.84 to 1.00). In 24 countries with more than one centre,
most summary estimates for within-country associations
were weakly positive. For severe wheeze in 13e14-year-
olds, the summary OR for a 10
m
g/m
3
increase in PM
10
was 1.01 (0.92 to 1.10). This result was close to
a summary OR of 0.99 (0.91 to 1.06) obtained from
published multi-centre studies.
Conclusions Modelled estimates of particulate matter
at city level are imprecise and incomplete estimates of
personal exposure to ambient air pollutants.
Nevertheless, our results together with those of previous
multi-centre studies, suggest that urban background
PM
10
has little or no association with the prevalence of
childhood asthma, rhinoconjunctivitis or eczema either
within or between countries.
INTRODUCTION
Childhood asthma is a major cause of illness and
disability in childhood.
1
It shows large variations in
occurrence world-wide and over time,
23
but the
reasons for these variations and trends are largely
unknown. Asthma is a condition characterised by
inflammation and hyper-responsiveness of the air-
ways, and those with asthma frequently report
that air pollution aggravates or precipitates their
asthma. It is therefore not surprising that ambient
air pollution is widely believed to be one possible
cause of variations in prevalence through effects on
incidence, severity or prognosis. The particulate
component of air pollution is widely measured as
particulate matter with aerodynamic diameter less
than 10
m
m (PM
10
), this being small enough to
penetrate the intrathoracic respiratory tract. PM
10
is a complex mixture arising from different sources
with a range of physicochemical characteristics.
While some components of PM
10
are likely to be
more toxic than others, these differences have been
difficult to quantify.
45
Experimental evidence based
on acute exposures suggests that particulate matter
has the potential to increase airway reactivity,
increase inflammatory responses in the lung and
enhance allergic immune responses,
67
but studies
at environmental concentrations have been equiv-
ocal.
8
Various reviews have concluded that while
there is evidence that short-term increases in PM
10
may aggravate symptoms of asthma, the evidence
for effects of chronic exposure to PM
10
on asthma
incidence and prevalence is weak.
9e11
There is some
evidence from birth cohort studies linking traffic
exposure to atopic sensitisation.
12e14
The evidence for effects on asthma of chronic
exposure to particulate matter may be divided into
those studies which have investigated the effects of
What this paper adds
<The reasons for the wide international variations
in the prevalence of childhood asthma, rhino-
conjunctivitis and eczema are not understood.
<One factor might be exposure to ambient
particulate matter since this has been associ-
ated with exacerbations of asthma and may also
play a role in the increased prevalence of
asthma symptoms and allergy observed in some
traffic-proximity studies.
<In a study of over half a million children from
105 cities in 51 countries, we found little or no
evidence of associations with modelled city-
level residential PM
10
.
<The results suggest that community levels of
ambient particulate matter are unlikely to
explain international variations in prevalence.
<Future investigations of this topic should employ
improved exposure assessment and control for
confounding factors at the individual level.
<See Commentary, p 290
1
Division of Community Health
Sciences, MRC-HPA Centre for
Environment and Health, St
George’s, University of London,
London, UK
2
Environment Department,
World Bank, Washington, DC,
USA
3
Institute for Risk Assessment
Sciences, Utrecht University,
Utrecht, the Netherlands and
the Julius Center for Health
Sciences and Primary Care,
University Medical Center,
Utrecht, the Netherlands
4
Department of Medicine and
Therapeutics, The Chinese
University of Hong Kong, Hong
Kong, SAR China
5
Department of Epidemiology,
University of Ulm, Ulm, Germany
Correspondence to
Professor H R Anderson,
Division of Community Health
Sciences, MRC-HPA Centre for
Environment and Health, St
George’s, University of London,
Cranmer Terrace, London SW17
0RE, UK;
r.anderson@sgul.ac.uk
Stephan K Weiland died on 19
March 2007.
Members of the ISAAC Phase
One Study Group are listed in
appendix 1
Accepted 16 August 2009
Occup Environ Med 2010;67:293e300. doi:10.1136/oem.2009.048785 293
Original article
group.bmj.com on May 9, 2010 - Published by oem.bmj.comDownloaded from
proximity to sources of traffic pollution and those which have
investigated the effects of community-level concentrations of
specific pollutants measured by urban background monitoring
stations representing the residential environment. The contri-
bution of traffic pollution to urban background concentrations
varies. Traffic-proximity studies are usually conducted within
a single urban area and tend to find that children living in close
proximity to traffic have an increased risk of respiratory symp-
toms, including those of asthma, relative to those living further
away.
15 16
The effects of community-level exposure on preva-
lence are usually investigated by comparing urban areas using
urban background concentrations of pollutants; these studies
tend to find associations with bronchitic type symptoms rather
than with asthma or allergic symptoms, but interpretation is
frequently hampered by the small number of areas compared,
often as low as two, and by the ecological nature of such studies.
We identified nine studies of children comparing five or more
centres and none found evidence of a statistically significant
positive association between ambient particulate matter and
asthma symptoms.
17e25
This is unlikely to be entirely explained
by exposure measurement error because the same studies often
report other adverse effects, including bronchitic symptoms in
those with and without asthma, reductions in lung function and
increased mortality rates in adults.
Current evidence therefore suggests that despite associations
between asthma prevalence and proximity to traffic, and
between asthma exacerbations and short-term variations in air
pollution measured at community monitors, particulate matter
does not play a role in determining asthma prevalence in
a population. Current evidence is, however, based on compari-
sons of a limited number of cities or communities within indi-
vidual developed countries. The exception is a study of nitrogen
dioxide and asthma prevalence in 62 centres in five largely
European countries.
26
There are no published international
studies of city-level average concentrations of background
particulate matter and asthma prevalence. The role of particu-
late matter in determining variations in asthma prevalence
world-wide has not been investigated.
Phase One of the International Study of Asthma and Allergies
in Childhood (ISAAC) study published prevalence estimates for
symptoms and diagnoses of asthma, rhinoconjunctivitis and
eczema from 6e7-year-old children in 91 centres in 38 countries
and from 13e14-year-old children in 156 centres in 56 coun-
tries.
27
Measures of ambient air pollution in ISAAC centres are
sparse. However, model estimates of annual concentrations of
ambient particle concentrations (PM
10
) in residential areas of all
cities with a population of more than 100 000 are available from
a model developed at the World Bank.
28
These had been used by
WHO to estimate the global impact of particulate matter on
mortality. This provided an opportunity to investigate the
association between city-level estimates of residential exposure
to urban background PM
10
and the prevalence of asthma,
rhinoconjunctivitis and eczema in children in a large number of
cities world-wide.
METHODS
The ISAAC protocol and results for the prevalence of symptoms
of asthma, rhinoconjunctivitis and eczema have been
published.
27 29
Briefly, each centre obtained data on 13e14-year-
old children from a self-completed questionnaire at school.
Optionally, data on 6e7-year-old children were obtained using
parent-completed questionnaires. The schools were selected to
represent a defined geographical area with a target sample of
3000 children per centre per age group.
The International Data Centre in Auckland supplied the raw
survey data. The definitions of current (past 12 months)
symptoms of wheeze, rhinoconjunctivitis and eczema and of
diagnoses (ever) of asthma, hay fever and eczema were the same
as reported previously.
30
Additionally, the 12-month prevalence
of moderate to severe wheezing (‘severe wheeze’) was based on
one or more of: (1) four or more attacks of wheeze, (2) woken by
wheeze on one or two nights per week or (3) wheezing severe
enough to limit speech to only one or two words at a time,
between breaths. The prevalence of severe wheeze among those
with current wheeze was estimated by the proportion of the
corresponding numbers of children in each of these symptom
groups. ‘Atopy ’was defined as rhinoconjunctivitis and/or eczema.
Various sources of data on community exposure to particulate
matter were explored, including the WHO, national electronic
databases and information from the ISAAC centre investigators,
but these were insufficient for our purpose because they were
available for only a minority of centres and when available were
not presented consistently. We therefore used the World Bank
Global Model of Ambient Particulates (GMAPS) which had esti-
mated for 1999 the annual exposure to particulate matter in
residential areas of cities with populations greater than 100 000,
details of which are published elsewhere
28
and may also be found
in online data supplement 1. This is a reduced form of a fixed effect
model developed using available particulate matter measurements
from population based monitoring stations world-wide for
1985e1999. There were 572 locations in 304 cities in 55 countries,
but these were heavily biased towards developed countries. The
determinants in the model included factors such as fuel use and
mix, scale and composition of economic activity, strength of local
pollution regulation, and geographical and atmospheric condi-
tions that affect pollutant transport. The model included
a country-specificfixed effect to control for economic, social and
natural factors not captured by the other explanatory variables.
Model estimates for countries with no monitoring stations are
based on an estimate of country-specificfixed effect from
a secondary model. The model explained 88% of the observed
variation in monitored particulate matter data in the 55 countries
during the 1985e1999 period. The model used in our analysis was
GMAPS 46, which estimated residential PM
10
levels for 1999.
Previouswork with ISAAC Phase One has shown a weak positive
association between gross national product (GNP) per capita and
asthma symptoms.
31
For this reason, we decided to control our
analyses at a country level for GNP per capita in 1993.
32
Centres were selected if there was a GMAPS estimate for
annual average PM
10
for their city. Where there was more than
one centre per city (five cities), we selected one centre at
random, to avoid overweighting that city estimate. The number
of centres and some of the sample sizes differ in detail from the
original Phase One reports
2 27
because we included some centres
that were too late for inclusion in the first report. The associa-
tions between PM
10
and the various outcomes were estimated
using a binomial logistic regression model which adjusted for
GNP per capita and allowed for clustering by country according
to the HubereWhite estimates of variance.
For 24 countries with ISAAC centres in two or more cities, the
within-country association with PM
10
was estimated by fitting
a separate binomial logistic regression for each country (without
adjustment for GNP per capita). The country-level results for
PM
10
were subsequently combined using random effects meta-
analysis to obtain a combined estimate.
The proportion of variability between studies attributed to
heterogeneity rather than chance was estimated from I
2
values.
Statistical analyses were performed using STATA.
33
294 Occup Environ Med 2010;67:293e300. doi:10.1136/oem.2009.048785
Original article
group.bmj.com on May 9, 2010 - Published by oem.bmj.comDownloaded from
RESULTS
The prevalence, PM
10
and GNP data for each centre in a city
with an estimate for PM
10
are summarised as quartiles in table 1.
The individual data for each centre are given in online data
supplement 2.
For the 6e7-year age group, PM
10
estimates were obtained for
65 centres in 32 countries. The median annual PM
10
concen-
tration was 37
m
g/m
3
(range 15e136). The median number of
participants per centre was 3007 (range 1104e6533) and the
total number of participants was 190 624. For the 13e14-year
age group, annual PM
10
estimates were obtained from 105
centres in 51 countries. The median annual PM
10
concentration
was 34
m
g/m
3
(range 15e158). The median number of partici-
pants per centre was 3086 (range 1056e5521) and the total
number of participants was 322 529.
For the 6e7- and 13e14-year age groups, the respective
median per capita GNP was US$8030 (range 340e39 640) and
US$3480 (range 100e39 640). The low GNP countries showed
a wide range of PM
10
levels and contained all of the high PM
10
concentrations, while high GNP countries tended to have levels
only in the lower range (below 50
m
g/m
3
)(figure 1). A rank
correlation test found a strong inverse relationship between
GNP and PM
10
(Spearman rho¼0.62).
Figure 2 shows the relationship between severe wheeze and
PM
10
for the 6e7- and 13e14-year age groups. There was
a moderate to weak negative association (Spearman rho¼0.4)
in both groups. The results of the logistic regression analysis
adjusting for GNP per capita and clustering within country are
shown in table 2.
In both age groups, for all outcomes studied, the ORs for
a10
m
g/m
3
increase in PM
10
concentrations remained below
unity, with the exception of severe wheeze among those with
current wheeze in the 13e14-year age group. The upper confi-
dence limit for the OR was also below unity for severe wheeze
and atopy in the 6e7-year age group and for current wheeze in
the 13e14-year age group.
Table 3 shows the results of the meta-analysis of within-
country relationships where there was more than one centre per
country.
Most of the outcomes, including those measuring wheeze and
severe wheeze, showed positive summary estimates, but most
had lower 95% confidence limits below unity. Three of the four
estimates for which the confidence limit did not include unity
were positive (hay fever in the 6e7-year age group, and rhino-
conjunctivitis and atopy in the 13e14-year age group) and one
was negative (asthma diagnosis in the 13e14-year age group). A
more detailed description of one of these meta-analyses, that for
severe wheeze in the 13e14-year age group, is shown in figure 3
which shows the ORs for the individual countries. The indi-
vidual country estimates varied in size and direction, but the
Table 1 Quartile distribution of outcomes, PM
10
and GNP per capita in the 6e7- and 13e14-year age groups
6e7 Years (65 centres, 32 countries) Min Q1 Median Q3 Max
Participants per centre (n) 1104 2418 3007 3414 6533
Current symptoms (% prevalence)
Current wheeze 0.8 6.4 9.3 17.3 27.2
Severe wheeze 0.5 2.1 3.3 7.0 15.3
Severe as % of current wheeze 21.8 32.4 37.9 43.2 62.5
Rhinoconjunctivitis 0.8 3.9 6.4 9.8 14.9
Eczema 0.0 3.3 6.7 10.2 18.4
Atopy 1.1 7.0 12.3 17.4 23.3
Diagnoses ever (% prevalence)
Asthma 1.0 4.1 7.7 14.4 30.8
Hay fever 0.0 4.8 6.7 12.7 29.0
Eczema 0.3 4.7 13.0 18.9 57.2
Air pollution and GNP per capita
PM
10
(
m
g/m
3
) 15 23 37 52 136
GNP per capita* (US$) 340 2750 8030 19380 39640
13e14 Years (105 centres, 51
countries) Min Q1 Median Q3 Max
Participants per centre (n) 1056 2904 3086 3373 5521
Current symptoms (% prevalence)
Current wheeze 1.6 6.8 10.7 16.1 33.5
Severe wheeze 0.7 2.7 4.4 7.2 16.6
Severe as % of current wheeze 21.8 36.7 42.3 48.1 76.6
Rhinoconjunctivitis 1.4 8.4 12.7 16.6 39.7
Eczemay0.3 4.0 6.0 9.3 20.5
Atopyy1.5 12.1 18.4 22.5 46.5
Diagnoses ever (% prevalence)
Asthma 1.4 5.5 9.6 14.1 30.4
Hay fever 0.0 6.1 14.2 24.2 54.4
Eczemay0.2 5.5 10.1 15.7 49.3
Air pollution and GNP per capita
PM
10
(
m
g/m
3
) 15 22 34 49 158
GNP per capitaz(US$) 100 1480 3480 18720 39640
*GNP per capita estimates were based on available data for 31 countries.
yData were available for 104 centres.
zGNP per capita estimates are based on available data for 50 countries.
GNP, gross national product; max, maximum; min, minimum.
Occup Environ Med 2010;67:293e300. doi:10.1136/oem.2009.048785 295
Original article
group.bmj.com on May 9, 2010 - Published by oem.bmj.comDownloaded from
overall random effects summary estimate was very close to zero.
Results were similar when stratified by European versus non-
European and by higher versus lower GNP per capita. Each
subgroup is shown ranked by annual average concentration of
pollution and it is clear that neither the size nor the direction of
the estimates was related to the mean level of PM
10
.
For both age groups, the results for severe wheeze in ISAAC
countries with more than one centre were compared to the
results from a meta-analysis of published multi-city studies
(mostly within one country or region). This is shown in figure 4.
None of the individual published studies found a significant
association between wheeze symptom and PM
10
. There was
little heterogeneity and the overall estimate without the present
study was, for a 10 unit increase in PM
10
, 0.99 (95% CI 0.91 to
1.06).
DISCUSSION
We generally found a weak negative relationship between
concentrations of modelled residential PM
10
at city level and the
centre prevalence of wheeze which persisted after controlling for
GNP per capita and allowing for clustering within country. A
meta-analysis of within-country associations in those countries
with more than one centre found mainly null associations with
respect to the various symptoms with the exception of positive
associations with hay fever in the 6e7-year age group and
rhinoconjunctivitis and atopy in the 13e14-year age group. The
results for severe wheeze were similar for European and non-
European centres and not related to the mean level of PM
10
.
The study design was ecological, with the centre as the unit of
analysis. This is appropriate for a study of ambient air pollution
because the exposure measurement is at an ecological level.
Measurement of chronic exposure at a personal level for a multi-
centre prevalence study is impractical. Although many previous
studies used an ecological design, most had insufficient power
due to the small number of communities compared. Our study is
by far the largest in terms of breadth of exposure and outcome
variables, numbers of units of analysis and precision of centre
prevalence estimates. In ecological analyses, it is possible to have
relationships at a between-country level that are different from
those between areas of the same country. We found that for
severe wheeze, the between-country analysis (allowing for clus-
tering of centres within country) showed a generally convincing
negative association, whereas for the within-country analysis the
summary estimates were more generally null. We do not know
what explains this difference, but neither result supports a posi-
tive association between city-level PM
10
concentrations and the
various outcomes studied.
Our outcome was based on a standardised validated ques-
tionnaire and the asthma symptoms have been shown to
correlate with national hospital admissions and mortality rates
for childhood asthma.
34
Further, prevalence estimates were
obtained independently from the parents of 6e7-year-olds and
from the 13e14-year-olds themselves. The sample size of each
centre was large enough to ensure sufficient precision of the
prevalence estimates.
Figure 2 Association between PM
10
and the prevalence of severe
wheeze in 6e7- and 13e14-year-old children.
Figure 1 Correlation between PM
10
(average for all centres within
a country) and gross national product per capita.
296 Occup Environ Med 2010;67:293e300. doi:10.1136/oem.2009.048785
Original article
group.bmj.com on May 9, 2010 - Published by oem.bmj.comDownloaded from
In using estimates from the GMAPS model, we were able to
include a far wider range of cities than would have been other-
wise possible. In development, the model explained 88% of the
PM
10
concentrations observed in the mainly developed cities for
which monitor data were available, but we do not know how
well it predicted PM
10
in regions that were not well represented
in the development of the model (Middle East, Africa, Asia).
GMAPS model estimates of annual average PM
10
have recently
been available for 1990e2005. PM
10
estimates for 1995 and 1999
are highly correlated (r¼0.98) and use of the estimates for 1999
is not expected to alter any of the findings reported in this paper.
Based on a few actual measurements in cities with ISAAC
centres, we found that there had tended to be a decline in
PM
10
from the mid-1990s to 1999, the year of the modelled
concentrations and that this was steeper in the more polluted
cities. This means that the concentrations in more polluted
cities in 1995 will probably have been underestimated by the
1999 GMAPS model.
It is inevitable that annual average PM
10
concentrations for
a city will be a poor indicator of individual exposure and dose to
the lung. This raises the possibility that measurement error may
be obscuring a real underlying effect. However, exposure esti-
mated at the community level is likely to have Berkson-type
error rather than classical measurement error, so that effect
estimates, although less precise, will not be biased towards the
null.
35
There are also two strands of empirical evidence which
support our use of community-average concentrations of
particulate matter. The first is that similar ecological study
designs have found positive associations with cough symptom
and reduced lung function in children.
11
The second is that city-
level concentrations of particulate matter have been associated
with mortality in multi-city cohort studies and with a wide
range of health outcomes in daily time-series studies. Therefore,
in spite of many issues concerning the measurement of partic-
ulate matter, we think that if an important association with
symptom prevalence had been present, we would have observed
some positive relationships. We did not have the opportunity to
investigate associations with pollutant gases, but the current
evidence from multi-city studies cited earlier does not suggest
that city-average concentrations of ozone, sulphur dioxide or
nitrogen dioxide are related to community asthma prevalence.
Climate factors such as temperature, humidity and rainfall
have been found to have some effect on the prevalence of
asthma, eczema and hay fever in previous analyses of these
ISAAC data.
36
The mechanisms are not understood and it was
unclear to us whether temperature should be included as
a potential confounder. We also noted that climate variables
were included in the GMAPS PM
10
model. We investigated the
association between PM
10
and the same climate variables as used
in the study by Weiland and colleagues and found that there was
no correlation within country between PM
10
and any climate
variable. When we investigated the between-countries correla-
tions, there were few significant correlations and these were
weak. We therefore concluded that the particulate matter results
would be unlikely to be explained by climate factors.
GNP, which was available at country level, was negatively
associated with PM
10
(figure 1) and positively associated with
symptom prevalence in ISAAC.
31
Because the estimation of
PM
10
used GNP per capita as an explanatory variable, we
performed the between-country analysis both with and without
adjustment for GNP. In both cases, the relationship between the
different outcomes and PM
10
was negative, although not
generally significant after controlling for GNP.
Table 3 Meta-analysis of PM
10
and prevalence in countries with more
than one centre*
6e7-Year age group (46 centres, 14 countries)
Outcome PM
10
OR (per 10
m
g/m
3
) 95% CI
Current symptoms (% prevalence)
Current wheeze 1.03 0.95 to 1.13
Severe wheeze 1.00 0.90 to 1.11
Severe as % of current wheeze 0.97 0.92 to 1.01
Rhinoconjunctivitis 1.06 0.96 to 1.17
Eczema 0.98 0.91 to 1.05
Atopy 1.03 0.96 to 1.10
Diagnoses ever (% prevalence)
Asthma 0.96 0.85 to 1.08
Hay fever 1.10 1.01 to 1.19
Eczema 0.97 0.87 to 1.08
13e14-Year age group (77 centres, 24 countries)
Outcome PM
10
OR (per 10
m
g/m
3
) 95% CI
Current symptoms (% prevalence)
Current wheeze 1.05 0.97 to 1.13
Severe wheeze 1.01 0.92 to 1.10
Severe as % of current wheeze 0.97 0.92 to 1.02
Rhinoconjunctivitis 1.15 1.06 to 1.26
Eczemay1.09 0.99 to 1.19
Atopyy1.14 1.05 to 1.24
Diagnoses ever (% prevalence)
Asthma 0.88 0.80 to 0.96
Hay fever 0.98 0.90 to 1.07
Eczemay0.98 0.90 to 1.06
All results are from random effects analysis.
*Estimate for Portugal excludes Funchal (Madeira island).
yEstimates are based on 75 centres, 23 countries.
Table 2 ORs for the association between PM
10
and prevalence at
centre level, adjusted for GNP per capita and allowing for clustering
within country
6e7-Year age group (63 centres, 31 countries)
Outcome PM
10
OR (per 10
m
g/m
3
) 95% CI
Current symptoms (% prevalence)
Current wheeze 0.89 0.80 to 1.00
Severe wheeze 0.88 0.77 to 1.00*
Severe as % of current wheeze 0.97 0.94 to 1.00
Rhinoconjunctivitis 0.93 0.87 to 1.00
Eczema 0.92 0.83 to 1.02
Atopy 0.93 0.86 to 1.00y
Diagnoses ever (% prevalence)
Asthma 0.88 0.76 to 1.02
Hay fever 0.97 0.86 to 1.10
Eczema 0.95 0.85 to 1.06
13e14-Year age group (103 centres, 50 countries)
Outcome PM
10
OR (per 10
m
g/m
3
) 95% CI
Current symptoms (% prevalence)
Current wheeze 0.91 0.84 to 0.99
Severe wheeze 0.92 0.84 to 1.00
Severe as % of current wheeze 1.00 0.97 to 1.03
Rhinoconjunctivitis 0.98 0.92 to 1.04
Eczemaz0.93 0.87 to 1.01
Atopyz0.96 0.90 to 1.02
Diagnoses ever (% prevalence)
Asthma 0.94 0.87 to 1.01
Hay fever 0.92 0.83 to 1.01
Eczemaz0.99 0.92 to 1.06
*0.9988;
y0.9999.
zEstimates are based on 102 centres, 50 countries.
Occup Environ Med 2010;67:293e300. doi:10.1136/oem.2009.048785 297
Original article
group.bmj.com on May 9, 2010 - Published by oem.bmj.comDownloaded from
We had little scope for controlling for confounders at the
individual level. Age and sex were controlled for by design but
the range of individual level confounders customarily controlled
for in prevalence studies (eg, active and passive smoking, other
sources of indoor pollution, dampness, etc) were unavailable.
ISAAC Phase Three, which was conducted during 2001e2002,
included a risk factor questionnaire and future analyses using
this later dataset will have the potential to adjust for some
confounders at an individual level. Notwithstanding these lim-
itations, we think that it is unlikely that any real and substantial
underlying causal association between PM
10
and the various
outcomes has been obscured by unknown confounding factors.
Our results were very similar to those found in published
studies of PM
10
and the prevalence of current asthma symptoms
with five or more study areas
19 20 22 24 25
(figure 4). Four multi-
area studies that could not be adapted for the meta-analysis of
ORs (because the results could not be converted to standardised
ORs) also reported essentially null results for PM
1023
or black
smoke.
17 18 21
The present study, also shown on figure 4, was in
line with these results and the summary estimate for all multi-
centre studies is convincingly null. Unlike the present study
which used modelled data, all of these studies used measured
concentrations of PM
10
or black smoke, and in some cases
employed study-directed monitors. Furthermore, most of these
Figure 3 Meta-analysis of the
association between PM
10
and the
prevalence of severe wheeze in the
13e14-year-age group for countries
with more than one centre. Stratified by
region and ranked by mean
concentration of PM
10
. ORs for 10
m
g/
m
3
PM
10
.
Figure 4 Meta-analysis of published
multi-centre studies of PM
10
and the
prevalence of moderate to severe
wheezing. ORs for 10
m
g/m
3
PM
10
.
298 Occup Environ Med 2010;67:293e300. doi:10.1136/oem.2009.048785
Original article
group.bmj.com on May 9, 2010 - Published by oem.bmj.comDownloaded from
studies controlled for a range of individual-level confounding
factors. The majority observed associations between PM
10
and
bronchitic symptoms such as cough, suggesting that they might
have been capable of detecting associations with wheeze
symptom should these have existed. Unfortunately, the ISAAC
asthma questionnaire did not include questions on cough.
There is some evidence from within-city studies that the
prevalence of asthma symptoms and of allergy may be higher in
populations who are exposed to traffic than in those less
exposed.
9 10
This could be explained by fresh traffic exhaust
being more toxic
37
or by exposure being much higher. There is
therefore a need to reconcile this evidence with that of the
present study. One explanation may be that traffic pollution
becomes less toxic with time and distance from the road,
another is that concentrations of traffic pollution become too
low and diluted with other sources for a health signal to be
detected using prevalence studies such as ours. Yet another is
that the contribution of traffic to the high background levels
found in low GNP countries is likely to be relatively small
because there is far less motorised traffic and far more uncon-
trolled emissions from home heating, industry, energy produc-
tion, etc. An alternative explanation is that the associations
observed in traffic-proximity studies are not due to pollution but
to some other factor. In more recent years, better city-level data
on sources have become available and we plan further studies
using the ISAAC Phase Three data to investigate whether some
sources are more relevant than others.
The strengths of this study lie in its world-wide scope, large
sample size and standardised outcome instrument, but we
recognise its relative weaknesses in exposure assessment and
inability to control for confounding factors at the city and
individual level. It may be possible to address some of these
defects in future studies using the ISAAC data. For the present,
however, we conclude that our results do not support the exis-
tence of an association between city-level concentrations of
residential ambient PM
10
and the prevalences of asthma,
rhinoconjunctivitis or eczema.
Acknowledgements We are grateful to the children and parents who willingly
cooperated and participated in ISAAC Phase One and the coordination and
assistance by the school staff is sincerely appreciated. We thank the Phase One
National Coordinators, Principal Investigators and their colleagues, who helped make
ISAAC Phase Three such a success. We would like to acknowledge and thank the
many funding bodies throughout the world that supported the individual ISAAC
centres and collaborators and their meetings. In particular, we wish to thank the
New Zealand funding bodies, the Health Research Council of New Zealand, the
Asthma and Respiratory Foundation of New Zealand, the Child Health Research
Foundation, the Hawke’s Bay Medical Research Foundation, the Waikato Medical
Research Foundation, Glaxo Wellcome New Zealand, the NZ Lottery Board and Astra
Zeneca New Zealand. Glaxo Wellcome International Medical Affairs supported the
Regional Coordination and the ISAAC International Data Centre. Without help from all
of the above, ISAAC would not have given us all these results from so many countries.
Competing interests None.
Ethics approval Each collaborator (appendix 1) obtained ethical approval for their
respective centre or centres.
Provenance and peer review Not commissioned; externally peer reviewed.
REFERENCES
1. Global Strategy for Asthma Management and Prevention, Global Initiative for Asthma
(GINA) 2007. http://www.ginasthma.org (accessed Mar 2008).
2. ISAAC Steering Committee. Worldwide variations in the prevalence of asthma
symptoms: the International Study of Asthma and Allergies in Childhood (ISAAC). Eur
Respir J 1998;12:315e35.
3. Pearce N, Ait-Khaled N, Beasley R, et al. Worldwide trends in the prevalence of
asthma symptoms: phase III of the International Study of Asthma and Allergies in
Childhood (ISAAC). Thorax 2007;62:758e66.
4. Pope CA III, Dockery DW. Health effects of fine particulate air pollution: lines that
connect. J Air Waste Manag Assoc 2006;56:709e42.
5. World Health Organization Regional Office for Europe 2007, Health relevance of
particulate matter from various sources: report on a WHO workshop, Bonn, Germany,
March 2007 WHO, Copenhagen.
6. Brunekreef B, Holgate ST. Air pollution and health. Lancet 2002;360:1233e42.
7. Riedl M, Diaz-Sanchez D. Biology of diesel exhaust effects on respiratory function.
J Allergy Clin Immunol 2005;115:221e8.
8. Holgate ST, Sandstrom T, Frew AJ, et al. Health effects of acute exposure to air
pollution. Part I: Healthy and asthmatic subjects exposed to diesel exhaust. Res Rep
Health Eff Inst 2003;112:1e30.
9. Department of Health Committee on the Medical Effects of Air Pollutants.
Asthma and outdoor air pollution. London: HMSO, 1995.
10. World Health Organisation. Air quality guidelines: global update 2005, particulate
matter, ozone, nitrogen dioxide and sulphur dioxide. Copenhagen: WHO Regional
Office for Europe, 2006.
11. WHO European Centre for Environment and Health. Effects of air pollution on
children’s health and development - a review of the evidence. Bonn: WHO Regional
Office for Europe, 2005.
12. Brauer M, Hoek G, Smit HA, et al. Air pollution and development of asthma, allergy
and infections in a birth cohort. Eur Respir J 2007;29:879e88.
13. Morgenstern V, Zutavern A, Cyrys J, et al. Atopic diseases, allergic sensitization,
and exposure to traffic-related air pollution in children. Am J Respir Crit Care Med
2008;177:1331e7.
14. Nordling E, Berglind N, Melen E. Traffic-related air pollution and childhood
respiratory symptoms, function and allergies. Epidemiology 2008;19:401e8.
15. Salam MT, Islam T, Gilliland FD. Recent evidence for adverse effects of
residential proximity to traffic sources on asthma. Curr Opin Pulm Med 2008;
14:3e8.
16. Heinrich J, Wichmann HE. Traffic related pollutants in Europe and their effect on
allergic disease. Curr Opin Allergy Clin Immunol 2004;4:341e8.
17. World Health Organisation. Study on chronic respiratory diseases in children in
relation to air pollution. Geneva: WHO, 1980.
18. Florey CD, Swan AV, Van der Lende R, et al.Report on the EC epidemiological
survey on the relationship between air pollution and respiratory health in primary
schoolchildren. Brussels: Commission of the European Communities, 1983.
19. Braun-Fahrlander C, Vuille JC, Sennhauser FH, et al. Respiratory health and
long-term exposure to air pollutants in Swiss schoolchildren. SCARPOL Team.
Swiss Study on Childhood Allergy and Respiratory Symptoms with Respect to
Air Pollution, Climate and Pollen. Am J Respir Crit Care Med 1997;
155:1042e9.
20. Dockery DW, Speizer FE, Stram DO, et al. Effects of inhalable particles on
respiratory health of children. Am J Respir Crit Care Med 1989;139:587e94.
21. Baldi I, Tessier JF, Kauffmann F, et al. Prevalence of asthma and mean levels of air
pollution: results from the French PAARC survey. Pollution Atomospherique et
Affections Respiratoires Chroniques. Eur Respir J 1999;14:132e8.
22. Dockery DW, Cunningham J, Damokosh AI, et al. Health effects of acid aerosols on
North American children: Respiratory symptoms. Environ Health Perspect
1996;104:500e5.
23. Guo YL, Lin YC, Sung FC, et al. Climate, traffic-related air pollutants and asthma
prevalence in middle-school children in Taiwan. Environ Health Perspect
1999;107:1001e6.
24. Peters JM, Avol E, Navidi W, et al. A study of twelve Southern California
communities with differing levels and types of air pollution. I. Prevalence of
respiratory morbidity. Am J Respir Crit Care Med 1999;159:760e7.
25. Shima M, Nitta Y, Ando M, et al. Effects of air pollution on the prevalence and
incidence of asthma in children. Arch Environ Health 2002;57:529e35.
26. Pattenden S, Hoek G, Braun-Fahrlander C, et al. NO2 and children’s respiratory
symptoms in the PATY study. Occup Environ Med 2006;63:828e35.
27. ISAAC Steering Committee. Worldwide variation in prevalence of symptoms of
asthma, allergic rhinoconjunctivitis, and atopic eczema: ISAAC. The International
Study of Asthma and Allergies in Childhood (ISAAC) Steering Committee. Lancet
1998;351:1225e32.
28. Pandey KD, Wheeler D, Ostro B, et al.Ambient particulate matter concentrations in
residential areas of world cities: new estimates based on global model of ambient
pollutants (GMAPS). Washington DC: World Bank, 2003.
29. Asher MI, Keil U, Anderson HR, et al. International Study of Asthma and Allergies in
Childhood (ISAAC): rationale and methods. Eur Respir J 1995;8:483e91.
30. Asher MI, Montefort S, Bjorksten B, et al. Worldwide time trends in the prevalence
of symptoms of asthma, allergic rhinoconjunctivitis, and eczema in childhood: ISAAC
Phases One and Three repeat multi-country cross-sectional surveys. Lancet
2006;368:733e43.
31. Stewart AW, Mitchell EA, Pearce N, et al. The relationship of per capita gross
national product to the prevalence of symptoms of asthma and other atopic diseases
in children (ISAAC). Int J Epidemiol 2001;30:173e9.
32. World Bank. World Development Report. Oxford: Oxford University Press, 1995.
33. Anon. STATA/SE 9.2 Statistics/Data Analysis. College Station, TX: Stata Corporation,
2007.
34. Anderson HR, Gupta R, Kapetanakis V, et al. International correlations between
indicators of prevalence, hospital admissions and mortality for asthma in children. Int
J Epidemiol 2008;37:573e82.
35. Zeger SL, Thomas D, Dominici F, et al. Exposure measurement error in time-series
studies of air pollution: concepts and consequences. Environ Health Perspect
2000;108:419e26.
Occup Environ Med 2010;67:293e300. doi:10.1136/oem.2009.048785 299
Original article
group.bmj.com on May 9, 2010 - Published by oem.bmj.comDownloaded from
36. Weiland SK, Husing A, Strachan DP, et al. Climate and the prevalence of asthma,
allergic rhinitis and atopic eczema in children. Occup Environ Med 2004;61:609e15.
37. McCreanor J, Cullinan P, Nieuwenhuijsen MJ, et al. Respiratory effects of exposure
to diesel traffic in persons with asthma. N Engl J Med 2007;357:2348e58.
APPENDIX 1: ISAAC PHASE ONE STUDY GROUP
ISAAC Steering Committee: N Aı
¨
t-Khaled (Union Internationale Contre la Tuberculose et
les Maladies Respiratoires, Paris, France); G Anabwani (Princess Marina Hospital,
Gaborone, Botswana); HR Anderson (Department of Public Health Sciences,
St George’s Hospital Medical School, London, UK); MI Asher (Department of Paediat-
rics, Faculty of Medical and Health Sciences, The University of Auckland, New Zealand);
R Beasley (Medical Research Institute of New Zealand, Wellington, New Zealand); B
Bjo
¨rkste
´n (Institute of Environmental Medicine, Karolinska Institutet, Stockholm,
Sweden); ML Burr (Department of Epidemiology, Statistics and Public Health, University
of Wales College of Medicine, Cardiff, UK); J Crane (Wellington Asthma Research
Group, Wellington School of Medicine, New Zealand); U Keil (Institut fu
¨r Epidemiologie
und Sozialmedizin, Universita
¨tMu
¨nster, Germany); CKW Lai (Department of Medicine
and Therapeutics, The Chinese University of Hong Kong, SAR China); J Mallol
(Department of Respiratory Medicine, University of Santiago de Chile, Chile); FC
Martinez (Arizona Respiratory Center, University of Arizona, Tucson, Arizona, USA); EA
Mitchell (Department of Paediatrics, Faculty of Medical and Health Sciences, The
University of Auckland, New Zealand); S Montefort (Department of Medicine, University
of Malta, Malta); N Pearce (Centre for Public Health Research, Massey University,
Wellington, New Zealand); CF Robertson (Department of Respiratory Medicine, Royal
Children’s Hospital, Parkville, Australia); JR Shah (Jaslok Hospital & Research Centre,
Mumbai, India); AW Stewart (Population Health, Faculty of Medical and Health
Sciences, The University of Auckland, New Zealand); DP Strachan (Department of Public
Health Sciences, St George’s Hospital Medical School, London, UK); E von Mutius
(Dr von Haunerschen Kinderklinik de Universita
¨tMu
¨nchen, Germany); SK Weiland
(Department of Epidemiology, University of Ulm, Germany); HC Williams (Centre
for Evidence Based Dermatology, Queen’s Medical Centre, University Hospital,
Nottingham, UK).
ISAAC International Data Centre: MI Asher, TO Clayton, P Ellwood, EA Mitchell,
Department of Paediatrics, and AW Stewart, School of Population Health, Faculty of
Medical and Health Sciences, The University of Auckland, New Zealand.
ISAAC Phase One principal investigators: Algeria: Dr A Bezzaoucha (Algiers);
Ethiopia: Associate Professor K Melaku (Addis Ababa); Ethiopia: Professor B Seyoum
(Jima); Kenya: Dr FO Esamai (Eldoret), Dr JA Odhiambo (Nairobi); Nigeria: Professor
BO Onadeko (Ibadan); South Africa: Professor R Ehrlich (Cape Town); China: Professor
Y-Z Chen (Beijing), Professor K-H Chen (Chongqing), Professor N-S Zhong
(Guangzhou), Dr M Bao-Shan (Shanghai); Hong Kong: Dr C Lai (Hong Kong 13e14-
year-old children), Professor Y Lung Lau (Hong Kong 6e7-year-old children);
Indonesia: Professor Dr K Baratawidjaja (Bandung); Japan: Professor S Nishima
(Fukuoka); Malaysia: Dr LW Yeong (Ipoh), Associate Professor BS Quah (Kota Bharu);
Philippines: Professor F Cua-Lim (Metro Manilla); Republic of Korea: Dr S-I Lee (Seoul);
Singapore: Professor B-W Lee (Singapore); Taiwan: Professor K-H Hsieh (deceased)
(Taipei); Thailand: Dr P Vichyanond (Bangkok), Associate Professor M Trakultivakorn
(Chiang Mai); Iran: Dr M-R Masjedi (Rasht and Tehran); Kuwait: Dr JA al-Momen
(Kuwait); Lebanon: Dr FM Ramadan (Beirut); Morocco: Professor Z Bouayad (Casa-
blanca and Marrakech), Professor A Bennis (Rabat); Pakistan: Dr ZA Bhutta (Karachi);
Argentina: Dr N Salmun (Buenos Aires and Rosario); Brazil: Professor N Rosa
´rio
(Curitiba), Professor R Stein (Porto Alegre), Dr PGM Bezerra (Recife), Associate
Professor L de Freitas Souza (Salvador), Professor D Sole
´(Sao Paulo); Chile: Dr I
Sanchez (Central Santiago), Dr L Amarales (Punta Arenas), Dr MA Calvo (Valdivia);
Mexico: Professor I Romieu (Cuernavaca); Panama: Dr G Cukier (David-Panama);
Paraguay: Dr JA Guggiari-Chase (Asuncion); Peru: Dr P Chiarella (Lima); Uruguay: Dra
D Holgado (Montevideo); Canada: Professor M Sears (Hamilton), Dr B Taylor
(Saskatoon); USA: Dr V Persky (Chicago (3)), Professor GJ Redding (Seattle); Albania:
Professor A Priftanji (Tirane); Estonia: Dr M-A Riikja
¨rv (Tallinn); Finland: Dr M Kajosaari
(Helsinki), Dr TA Koivikko (Turku and Pori County); Georgia: Dr N Khetsuriani (Kutaisi),
Professor A Gamkrelidze (Tbilisi); Latvia: Dr M Leja (Riga); Poland: Associate Professor
G Lis (Krakow), Dr A Bre
ˆborowicz (Poznan); Roumania: Professor D Dumitrascu (Cluj);
Russian Federation: Professor RM Khaitov (Moscow); Sweden: Dr L Nilsson (Link-
oping), Dr T Foucard (Stockholm/Uppsala); Uzbekistan: Professor T Aripova (Samar-
kand and Tashkent); Australia: Dr D Kennedy (Adelaide), Professor C Robertson
(Melbourne), Professor L Landau (Perth), Dr J Peat (Sydney 6e7-year-old children),
Professor A Bauman (Sydney 13e14-year-old children); New Zealand: Professor MI
Asher (Auckland), Associate Professor P Pattemore (Christchurch), Professor J Crane
(Wellington); India: Dr RM Maheshwari (Akola), Dr MK Joshi (Bombay (16)), Professor
L Kumar (Chandigarh), Dr KC Jain (Jodhpur), Dr TU Sukumaran (Kottayam), Dr N
Somu (Madras (3)), Professor GR Sethi (New Delhi (7)), Dr G Jayaraj (Neyveli), Dr NM
Hanumante (Pune); Austria: Dr J Riedler (Salzburg); Belgium: Professor P Vermeire
(Antwerp); France: Professor D Charpin (Marseilles), Professor P Godard (Montpellier),
Dr C Kopferschmitt-Kubler (Strasbourg); Germany: Professor Dr U Keil (Munster);
Greece: Associate Professor C Gratziou (Athens); Italy: Dr F Forastiere (Cosenza and
Roma), Dr E Chellini (Firenze), Dr L Bisanti (Milano), Dr G Ciccone (Torino), Dr S Piffer
(Trento), Professor A Boner (Verona); Portugal: Dr FD Borges (Funchal), Professor JE
Rosado Pinto (Lisbon), Dr JM Lopes dos Santos (Porto); Spain: Dr RM Busquets
(Barcelona), Dr AD Rubio (Bilbao), Dr AR Asensio (Cadiz), Professor L Garcı
´
a-Marcos
(Cartagena), Dr A Arnedo-Pena (Castellon), Professor F Guille
´n-Grima (Pamplona),
Professor MMM Sua
´rez-Varela (Valencia), Professor A Blanco Quiro
´s (Valladolid);
United Kingdom: Professor HR Anderson (South Thames).
300 Occup Environ Med 2010;67:293e300. doi:10.1136/oem.2009.048785
Original article
group.bmj.com on May 9, 2010 - Published by oem.bmj.comDownloaded from