&EPA
United States
Environmental Protection
Agency
EPA/600/R-06/164F | August 2008 | www.epa.gov
Review of the Impacts
of Climate Variability
and Change on Aeroallergens
and Their Associated Effects
National Center for Environmental Assessment
Office of Research and Development, Washington, DC 20460
-------
EPA/600/R-06/164F
August 2008
A Review of the Impacts of Climate Variability and
Change on Aeroallergens and
Their Associated Effects
Global Change Research Program
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
-------
DISCLAIMER
This document is a final report. It has been subjected to within-agency peer review and
to review by a panel of independent scientists. The report does not constitute U.S.
Environmental Protection Agency policy. Mention of trade names or commercial products does
not constitute endorsement or recommendation for use.
Photo Credits
Photographs on the report cover are provided by permission from the following:
Front Cover:
Young boy using inhaler
By permission from the MetroHealth System, Cleveland, OH.
Photographer: Karen Kutoloski, DO.
See http://www.metrohealth.org/
The image appears at http://www. metrohealth. org/body. cfm?id= 15 52
Common ragweed
By permission from Auburn University Agriculture Department at http://www.ag.auburn.edu/
The image appears at http://www.ag.auburn.edu/.. ./weedid/commonragweed.html
Microscopic pollen grains
By permission from Asthma Helpline at http://www.asthmahelpline.com/
The image appears at http://www.asthmahelpline.com/photos%20site/pollen%20grains.jpg
Back Cover:
Pine cones
By permission from Affinity Research at http ://www. affinity-research.com
The image appears at http://www.affinity-research.com/ourresearchinterests
Mugwort
By permission from Artistic Gardens at http://www.artisticgardens.com
The image appears at
https://www.artisticgardens.com/catalog/product info.php?productsid=162&osCsid=3281b490
53f34dldd2d976705900d305
Microscopic pollen from prairie hollyhock
In the public domain at the U.S. National Aeronautic and Space Administration (NASA)
The image appears at
http://science.nasa.gov/headlines/v2008/images/pollen/pollenprairiehollyhock.jpg
ii
-------
CONTENTS
LIST OF TABLES v
LIST OF FIGURES vi
LIST OF ABBREVIATIONS vii
PREFACE viii
AUTHORS AND REVIEWERS ix
EXECUTIVE SUMMARY xi
1. INTRODUCTION 1-1
2. AEROALLERGENS AND ASSOCIATED ALLERGIC DISEASES IN THE
UNITED STATES 2-1
2.1. AEROALLERGENS 2-1
2.1.1. Pollen 2-1
2.1.2. Mold 2-10
2.1.3. Indoor Allergens 2-12
2.2. ASSOCIATED ALLERGIC DISEASES 2-13
2.2.1. Allergic Rhinitis 2-16
2.2.2. Asthma 2-20
2.2.3. Atopic Dermatitis 2-22
2.2.4. Cross-Reactivity 2-23
3. HISTORICAL TRENDS IN AEROALLERGENS AND ALLERGIC DISEASES
IN THE UNITED STATES 3-1
3.1. AEROALLERGENS 3-1
3.2. HISTORICAL TRENDS OF ALLERGIC DISEASES 3-3
3.2.1. Asthma 3-4
3.2.2. Allergic Rhinitis 3-6
3.2.3. Atopic Dermatitis 3-7
4. IMPACTS OF CLIMATE CHANGE ON AEROALLERGENS 4-1
4.1. PRODUCTION OF AEROALLERGENS 4-1
4.1.1. Pollen 4-2
4.1.2. Mold 4-12
4.2. DISTRIBUTION OF AEROALLERGENS 4-14
4.3. DISPERSAL 4-17
4.4. ALLERGEN CONCENTRATION 4-18
in
-------
CONTENTS (continued)
4.5. CLIMATE VARIABILITY AND ALLERGIC DISEASES 4-19
4.5.1. Timing of Aeroallergen Product!on and Subsequent Illness 4-20
4.5.2. Aeroallergen Production, Allergen Content, and Subsequent Illness 4-22
4.5.3. Distribution and Dispersion of Aeroallergens and Subsequent Illness 4-25
4.5.4. Observational Studies of Weather, Aeroallergens, and Illness 4-27
4.5.5. Linkages Among Air Pollution, Aeroallergens, and Allergic Diseases 4-30
5. ECONOMIC AND QUALITY-OF-LIFE IMPACTS OF ALLERGIC DISEASES 5-1
5.1. COST OF ILLNESS—METHODOLOGY 5-9
5.2. COST COMPONENTS 5-10
5.2.1. Direct Medical Costs 5-11
5.2.2. Direct Nonmedical Costs 5-11
5.2.3. Indirect Costs 5-11
5.2.4. Intangible Costs 5-11
5.2.5. Hidden Costs 5-12
5.3. SOURCES OF VARIABILITY IN COST ESTIMATES FROM COI
STUDIES 5-12
5.3.1. Reference Year 5-12
5.3.2. Cost Components 5-12
5.3.3. Discount Rate 5-12
5.3.4. Definition of Disease 5-13
5.3.5. Prevalence vs. Incidence Approach 5-13
5.3.6. Scope and Perspective of Estimation 5-14
5.4. COST OF ILLNESS—ESTIMATES 5-14
5.4.1. Allergic Rhinitis 5-14
5.4.2. Asthma 5-17
5.4.3. Atopic Dermatitis 5-20
6. FUTURE RESEARCH 6-1
7. SUMMARY AND CONCLUSIONS 7-1
REFERENCES R-l
IV
-------
LIST OF TABLES
2-1. Selection of clinically relevant aeroallergens in the United States 2-2
2-2. Geographic distribution of major clinically relevant tree pollen in the United States 2-5
2-3. Pollen seasons of the maj or clinically relevant tree pollen in the United States 2-6
2-4. Geographic distribution of major clinically relevant grass pollen in the United
States 2-7
2-5. Pollen season of the major clinically relevant grass pollen in the United States 2-8
2-6. Geographic distribution of major clinically relevant weed pollen in the United
States 2-9
2-7. Pollen season of the major clinically relevant weed pollen in the United States 2-10
2-8. Geographic distribution of major clinically relevant mold in the United States 2-11
2-9. Geographic distribution of maj or clinically relevant indoor allergens 2-13
2-10. Allergic diseases correlated with the major clinically relevant aeroallergens 2-17
2-11. Cross-reactivity of major clinically relevant aeroallergens 2-24
5-1. National hospital statistics, 2003—principal diagnosis only (all conditions) 5-2
5-2. Allergic rhinitis national hospital statistics, 2003—principal diagnosis only 5-3
5-3. Allergic rhinoconjunctivitis national hospital statistics, 2003—principal
diagnosis only 5-4
5-4. Asthma national statistics, 2003—principal diagnosis only 5-5
5-5. Atopic dermatitis national statistics, 2003—principal diagnosis only 5-8
5-6. Annual cost of allergic rhinitis/rhinoconjunctivitis estimates, in 2005 $ by cost
category 5-15
5-7. Annual cost of asthma estimates, in 2005 $ by cost category 5-18
5-8. Annual cost of atopic dermatitis estimates, in 2005 $ by cost category 5-21
-------
LIST OF FIGURES
3-1. Prevalence of asthma in the United States from 1970 through 1996 3-4
3-2. Prevalence of asthma in the United States by geographic region from 1970
through 1996 3-5
3-3. Prevalence of allergic rhinitis (hay fever) in the United States from 1970 through
1996 3-6
3-4. Prevalence of allergic rhinitis (hay fever) in the United States by geographic
region from 1970 through 1996 3-7
4-1. Pollen production in A. artemisiifolia for three springtime dormancy release
cohorts grown at two CC>2 concentrations (380 ppm and 700 ppm) 4-10
4-2. (A) Schematic depiction of phenotypic frequencies (mean phenotype) for a
population at a location along a climate gradient where fitness maximum is C.
(B) Schematic depiction of fitness optima (red) for a species that ranges across
a climate gradient 4-16
4-3. Schematic diagram of the relationship between global climate change and the
rise in asthma prevalence and severity, via impacts of climate change on plant
and pollen attributes 4-20
4-4. Potential air pollution-related health effects of climate change 4-21
5-1. Age-adjusted death rates for asthma by race and sex, United States (1951-2002) 5-6
5-2. Physician office visits for asthma, United States (1989-2001) 5-7
5-3. Hospitalizations for asthma, United States (1980-2002) 5-7
5-4. Three dimensions of economic evaluation of clinical care 5-9
5-5. Cost inventory diagram 5-10
VI
-------
LIST OF ABBREVIATIONS
A&E Accident and Emergency
AAAAI American Academy of Allergy Asthma and Immunology
BLS CPI Bureau of Labor Statistics Consumer Price Index
CCF Climate Change Futures
CDC Centers for Disease Control and Prevention
COI cost of illness
DALY disability adjusted life years
ED emergency department
ENSO El Nino-Southern Oscillation
GCMs general circulation models
GCRP Global Change Research Program
IPCC Intergovernmental Panel on Climate Change
JCAAI Joint Council of Allergy, Asthma and Immunology
NAST National Assessment Synthesis Team
NCHS National Center for Health Statistics
ORD Office of Research and Development
PEFRs peak expiratory flow rates
U.S. EPA U.S. Environmental Protection Agency
WTP willingness to pay
vn
-------
PREFACE
The Environmental Protection Agency's Global Change Research Program (GCRP) is an
assessment-oriented program within the Office of Research and Development (ORD) that
focuses on assessing how potential changes in climate and other global environmental stressors
may impact air quality, water quality, ecosystems, and human health in the United States. The
Program's focus on human health is consistent with the Strategic Plan of the U.S. Climate
Change Science Program—the federal umbrella organization for climate change science in the
U.S. government. It is responsive to the research agenda set out in the Health Sector Assessment
of the First National Assessment of the Potential Consequences of Climate Variability and
Change in the United States.
Since 1998, the National Center for Environmental Assessment within the ORD has
assessed the consequences of global change on weather-related morbidity, on vector- and water-
borne diseases, and on airborne allergens and ambient pollutants, especially tropospheric ozone
and fine particles. Through its assessments, this Program has provided timely scientific
information to stakeholders and policy makers to support them as they decide whether and how
to respond to the risks and opportunities presented by climate change.
Because health is affected by a variety of social, economic, political, environmental, and
technological factors, assessing the health impacts of global change is a complex challenge. As a
result, health assessments in the GCRP look beyond epidemiological and toxicological research
to develop integrated health assessment frameworks that consider the effects of multiple stresses,
their interactions, potential adaptive responses, and location-specific impacts. This report
assesses the state of the scientific literature and examines the potential effects of climate
variability and change on aeroallergens and their associated health outcomes in the United States.
Vlll
-------
AUTHORS AND REVIEWERS
The Global Change Research Program (GCRP) in the Environmental Protection
Agency's (U.S. EPA's) National Center for Environmental Assessment in the Office of Research
and Development prepared this document in conjunction with scientists at ABT Associates (EPA
Contract No. GS-10F-0146L). Janet Gamble, of the GCRP's assessment staff, provided overall
direction for the project and prepared the report together with Colleen Reid, an ASPH Fellow at
U.S. EPA, and Ellen Post and Jason Sacks of ABT Associates.
Authors
Janet L. Gamble, Ph.D.
Global Change Research Program
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
Colleen E. Reid, MPH
ASPH Fellow
U.S. Environmental Protection Agency
Washington, DC
Ellen Post
ABT Associates, Inc.
Bethesda, MD
Jason Sacks
ABT Associates, Inc.
Bethesda, MD
U.S. EPA Reviewers for the Internal Review Draft
This external review draft incorporates comments on the internal review draft by U.S. EPA
scientists, including:
• Bob Frederick, Office of Research and Development, National Center for Environmental
Assessment
• Julie Damon, Office of Research and Development, National Center for Environmental
Assessment
• Brooke Hemming, Office of Research and Development, National Center for Environmental
Assessment
• Jason Samenow, Office of Air and Radiation, Office of Atmospheric Programs
• Maryjane Selgrade, Office of Research and Development, National Health and Environmental
Effects Research Laboratory
ix
-------
AUTHORS AND REVIEWERS (continued)
U.S. EPA Panel of External Reviewers
The report has been reviewed by a panel of external reviewers, including
• Paul J. Beggs, Macquarie University
• Kristie Ebi, ESS, LLC
• Patrick L. Kinney, Columbia University School of Public Health
• Estelle Levitin, University of Tulsa
The final report incorporates the suggestions set forward by the external panel.
-------
EXECUTIVE SUMMARY
This report presents a survey of the current state of knowledge of the potential impacts of
climate change and variability on aeroallergens—pollen, mold, and indoor allergens—in the
United States and the allergic diseases associated with them. Allergies are prevalent in the
United States and impose substantial economic and quality-of-life burdens. A recent nationwide
survey reported that 54.6% of people in the United States test positive for one or more allergens
(American Academy of Allergy Asthma and Immunology [AAAAI], 1996-2005). Among
specific allergens, dust mites, rye, ragweed, and cockroaches caused sensitization in
approximately 25% of the population (Arbes et al., 2005).
Allergies are the sixth most costly chronic disease category in the United States,
collectively costing the health care system approximately $21 billion annually (AAAAI,
1996-2005). The three main allergic diseases that have been associated with exposure to
aeroallergens—allergic rhinitis (hay fever), asthma, and atopic dermatitis (eczema)—
individually and collectively impose both substantial health effects and large economic burdens.
The direct medical costs of asthma and allergic rhinitis (hay fever) are estimated to be
$12.5 billion and $6.2 billion per year, respectively (in 2005 U.S. dollars) (AAAAI, 1996-2005);
the direct medical costs of atopic dermatitis (eczema) are estimated to be $1.2-$5.9 billion per
year (in 2005 dollars) (Ellis et al., 2002).
While data suggest that aeroallergen levels have remained relatively stable, the
prevalence of allergic diseases in the United States has increased over the last 30 years, a trend
that appears to be mirrored in other countries as well. The causes of this upward trend are as yet
unclear. Because the economic impacts of allergic diseases associated with aeroallergens and the
quality-of-life impacts on those individuals who suffer from them are already substantial, any
climate change-induced enhancement or continuation of this trend in the United States would be
of particular concern.
General projections of climate change and potential impacts on aeroallergens
While climate change projections are uncertain, multiple efforts have been made to
derive future climate scenarios based on projected concentrations of greenhouse gases and
models that simulate the climate system. The United Nations Intergovernmental Panel on
xi
-------
Climate Change (IPCC) projects that by the year 2100, the annual surface temperature increases
are projected to range from 1-3°C near the coasts in the conterminous United States to more
than 5°C in northern Alaska (IPCC, 2007). Along with increasing temperatures, other effects of
climate change, such as changes in precipitation and increases in extreme weather events, have
also been anticipated. These changes, including increased CC>2 concentrations, could impact the
production, distribution, and dispersion of aeroallergens along with the allergen content and the
growth and distribution of the weeds, grasses, trees, and mold that produce them. Shifts in
aeroallergen production and, subsequently, human exposure, may result in changes in the
severity and prevalence of symptoms in individuals with allergenic illnesses suffering from
allergic diseases.
The literature does not provide definitive data or conclusions, however, on how climate
change might impact aeroallergens and, subsequently, the severity or prevalence of allergic
diseases in the United States. This is, in part, because studies are of necessity often narrowly
defined, and a single study is unlikely to encompass the broad subject of weather, aeroallergens,
and allergic illness. There is also an inherent uncertainty as to how the climate will change,
especially at a regional level. In addition, the etiology of allergic diseases, especially asthma, is
complex and has a gene-environment interaction that is not yet well understood. Finally, there
are numerous other factors that affect aeroallergen levels and the severity and prevalence of
associated allergic diseases, such as changes in land use, air pollution, adaptive responses, and
modifying factors; many of which are difficult to assess or characterize.
Nevertheless, some tentative conclusions can be drawn about the potential impact of
climate change on aeroallergens and the associated allergic diseases through inferences regarding
the links between (1) climate change and the characteristics of aeroallergens and (2) those
aeroallergen characteristics and the associated allergic diseases.
Other research has shown that preseason temperature and precipitation have been
consistently important factors for pollen and mold production. Overall, experimental and
observational data as well as models indicate the following likely changes in aeroallergen
production, distribution, dispersal, and allergen content as a result of climate change in the
United States:
xn
-------
• Pollen production is likely to increase in many parts of the United States, with the
possible exception of the Southeast;
• Phenologic advance is likely to occur for numerous species of plants, especially trees
(Root et al., 2003);
• There will likely be changes in the distribution of pollen-producing species, including the
possibility of extinction in some cases (Joyce et al., 2001);
• Intercontinental dispersal (e.g., of pollen) is possible, facilitating the introduction of new
aeroallergens into the United States (Husar et al., 2001); and
• Increases in allergen content, and thus, potency, of some aeroallergens are possible
(Beggs, 2004; Beggs and Bambrick, 2005).
Research on the potential effects of climate change on tree and grass pollen production in
the United States is limited, but there are more of these types of studies. In general, the literature
to date suggests that preseason temperature and precipitation are important projectors of both
tree and grass pollen production. To the extent that climate change results in changes in these
two meteorological variables, then we would expect corresponding changes in tree and grass
pollen production, all else being equal. The evidence to date suggests that the nature of the
changes may be region- and species-specific. Although this does not necessarily imply increased
pollen production, a consistent finding from international research is earlier start dates for pollen
seasons, especially in trees (Clot, 2003).
Among weed pollen, common ragweed (Ambrosia artemisiifolia L.) is recognized as a
significant cause of allergic rhinitis (hay fever) in the United States and there is relatively more
research on the response of this weed to climatic variables, especially in the context of climate
change. Several researchers have used controlled environments to examine ragweed response to
carbon dioxide levels and temperature, the two covariates for which models reliably project
increased levels in the future. The experimental results have consistently demonstrated that
doubling carbon dioxide levels from current (350 umol/mol) to projected future levels
(700 umol/mol) would result in a 60 to 90% increase in ragweed pollen production (Ziska and
Caulfield, 2000; Wayne et al., 2002). A field study demonstrated ragweed grew faster, flowered
earlier, and produced significantly greater aboveground biomass and ragweed pollen at urban
locations than at rural locations (Ziska et al., 2003). Because urban locations are warmer and
have higher concentrations of CC>2 than rural locations, this may have implications for the impact
xiii
-------
of climate change on ragweed pollen production overall. In summary, studies of ragweed in
controlled environments and in field studies clearly show that pollen production can be expected
to increase with increased temperature and carbon dioxide levels.
There is limited but inconsistent evidence of increasing trends in mold production.
Assessment of mold production in response to climate change is derived mainly from
observational analyses of long-term data sets. An analysis in Denver, Colorado showed
Cladosporium increasing, but not co-occurring mold such as Alternaria or Epicoccum (Katial et
al., 1997). An observational study in Derby, UK showed Alternaria increasing (Corden and
Millington, 2001). Another United States study observed increases in mold counts after an El
Nino event (Freye and Litwin, 2001). While the extent of the impact is not altogether
understood, it is clear that climatic factors have an impact on mold production.
Long-term responses to climate change (over 50 to 100 years) are likely to include
species' range or distribution shifts, and in some cases, possible extinction. Some ecological
models suggest that the potential habitats, and thus distribution, for many tree species in the
United States are likely to change, in some cases dramatically, by the end of the 21st century.
Trees favoring cool environments, such as maple and birch, are likely to shift northward,
possibly out of the United States entirely, thus altering the pollen distribution associated with
them (Joyce et al., 2001). The habitats of alpine, subalpine spruce/fir, and aspen communities
are likely to contract significantly in the United States and largely shift into Canada (Joyce et al.,
2001). Potential habitats for oak/hickory, oak/pine, ponderosa pine, and arid woodland
communities are likely to increase in the United States (Joyce et al., 2001). Under certain model
scenarios, the Southeast will experience significant warming trends leading to an expansion of
savannas and grasslands at the expense of forest, again altering the presence of major
aeroallergens in large regions of the country. Note however, conclusions about projected shifts
in the distribution of major vegetation types, as plant species move in response to climate
change, depend on an implicit assumption in the biogeography models that contend that
vegetation will be able to move freely from location to location. This assumption "may be at
least in part unwarranted because of the barriers to plant migration that have been put in place on
landscapes through agricultural expansion and urbanization" (Melillo et al., 2001, p. 82).
There has been only limited research on how climate change could affect the dispersal of
pollen and mold, but there are cases of both pollen and dust being dispersed long distances from
xiv
-------
their release sites. The frequency of such cases may be increased by climate change. One study,
for example, suggests that in Europe, increased strength of westerly winds due to climate change
will enhance the long-range transport of birch pollen already observed to take place from north
and central Europe to Scandinavia (Emberlin, 1994). Transcontinental transport of dust
particulates has also been observed. To the extent that climate change increases the frequency of
weather events that facilitate such transcontinental transport, it could increase the likelihood of
additional aeroallergens being introduced into the United States. Whether long-range transport
of pollen may instead decrease aeroallergen concentration and distribution so as to decrease
human exposures has not been reported.
The links to allergic diseases
Shifts in phenology (periodic phenomena that are related to climate change) are one of
the most consistent findings in studies of plant pollen production. Alterations in the timing of
aeroallergen production in response to weather variables have been clearly demonstrated for
certain tree species, but less so for grass and weed pollen and mold. Analyses of trends in
allergic diseases are based on annual prevalence and generally do not document the seasonal
timing of these diseases within the year. In sensitized individuals, however, exposure clearly
leads to allergic response; thus, it is reasonable to expect that changes in the timing of production
of seasonal aeroallergens would result in corresponding changes in the timing of the associated
seasonal allergic diseases. Thus, the National Assessment Synthesis Team (NAST) (Bernard et
al., 2001) concluded that climate change may affect the timing or duration of seasonal allergic
diseases such as hay fever. However, shifts in the timing of asthma and atopic dermatitis in
response to changes in phenology are not well understood.
Increases in aeroallergen production and/or allergen concentration could impact the
severity and possibly prevalence of allergic illness via sensitivity and response pathways. On the
basis of model projections by the NAST based on climate change scenarios (Melillo et al., 2001),
pollen production, and possibly allergen content, in many areas of the country may increase at
least through the mid-21st century. Exposures to higher concentrations of allergens may lead to
more severe allergic responses (Nielsen et al., 2002). In addition, exposure to elevated pollen
and mold concentrations during sensitization may lead to a greater likelihood of development of
allergies such as rhinitis. Finally, as noted above, additional aeroallergens might be introduced if
xv
-------
long-range transcontinental transport of pollen and/or mold is facilitated by climate
change-induced factors.
Although there is substantial evidence suggesting a causal relationship between
aeroallergens and allergic diseases, it remains unclear which aeroallergens are more highly
associated with causing sensitization and subsequent disease development. We observe this
primarily because of the cross-reactivity of aeroallergens—the ability of two or more
aeroallergens with biochemical similarities to elicit an allergic response in an individual who
may be sensitized to only one of them. Multiple studies have found cross-reactivity among
aeroallergens to be implicated in allergic disease causation.
Not only the type, but also the amount of aeroallergen to which an individual is exposed
is influential in the development of an allergic illness. Similar to what is observed in most
disease causation scenarios, a dose-response relationship between aeroallergen exposure and
sensitization and exacerbation of disease has been observed—i.e., sensitized patients are more
likely to have more severe disease if exposure to allergens is high.
There are, thus, at least three causal pathways for climate change-induced impacts on
aeroallergens to alter the severity and possibly the prevalence of allergic diseases. First, a longer
exposure during sensitization may lead to greater likelihood of the development of allergy.
Second, a higher dose during sensitization may lead to a greater likelihood of development of an
allergy. Third, a higher dose during subsequent exposures (postsensitization) may lead to a more
severe allergic response. For individual patients, we can learn exactly which allergens are
responsible for sensitization.
However, as noted earlier, the etiology of allergic diseases, especially asthma, is complex
and has a gene-environment interaction that is poorly understood. There are numerous other
factors that affect aeroallergen levels and the severity and prevalence of associated allergic
diseases, such as changes in land use, air pollution, adaptive responses, and modifying factors;
many of which are difficult to assess.
Future research
Further progress must be made in documenting and understanding aeroallergen response
to climate, the role of aeroallergens in disease development, and the willingness to pay to avoid
the intangible costs of these allergic diseases. A review of the literature indicates there are
xvi
-------
limited data on aeroallergen trends in the United States. Integrated long-term data series on all
aeroallergens are necessary to clearly document future changes in aeroallergen production and
distribution, as well as allergen content. Additional research on the response of mold and indoor
allergens to climate change would be of particular value. In addition, further experimental and
field studies are needed to examine how allergen content and distribution of aeroallergens may
be altered in response to climate change (Beggs, 2004).
There is a need for better understanding of the role of aeroallergens in disease
development, especially with respect to asthma. Specifically, what is the relative contribution of
different aeroallergens to the development of asthma? There is a need to know what levels of
allergen exposure constitute a risk for asthma development. There is also a need for standardized
approaches for measuring exposures and outcomes in epidemiologic studies (Selgrade et al.,
2006). Finally, the potential synergistic effect of aeroallergens and air pollutants on the
development or exacerbation of allergic diseases is an important area for future research.
Based on a review of the cost of illness (COI) literature on allergic rhinitis, asthma, and
atopic dermatitis, it is clear that an important research gap is the current lack of assessment of
and, in particular, estimation of willingness to pay to avoid the intangible costs of these diseases.
In addition, better methodologies are needed to address productivity losses, aeroallergen
avoidance, and over-the-counter medication use. Finally, a disease or condition may also
contribute to increased costs as a secondary diagnosis, or as a risk factor for other diseases and
conditions. These hidden costs of comorbidity need to be properly addressed and, if possible,
included in future COI studies.
xvn
-------
XV111
-------
1. INTRODUCTION
Aeroallergens are classified into three groups: pollens (tree, weed, and grass), molds, and
indoor allergens. There is evidence to support a causal relationship between each aeroallergen
within these groups and one or more allergic diseases, including allergic rhinitis (hay fever),
asthma, and atopic dermatitis (eczema). Over the last 30 years, there has been a substantial
increase in the prevalence of allergic diseases within the United States. The prevalence of
asthma and allergenic rhinitis has increased from approximately 8-55 per 1,000 persons and
approximately 55-90 per 1,000 persons, respectively. The underlying reasons behind the
increased prevalence of each illness remain unclear. It has been hypothesized that global climate
change could alter the concentrations, distributions, dispersion patterns, and allergenicities of
aeroallergens in the environment in ways that could further increase the prevalence of allergic
diseases in the United States.
Although climate change projection is still an uncertain science, there have been attempts
to derive future climate scenarios, based on projected concentrations of greenhouse gases and
models that simulate atmospheric circulation. In 2007, the United Nations Intergovernmental
Panel on Climate Change (IPCC) conducted a comprehensive assessment of the science behind
projected climate changes. The IPCC report projected changes in average, minimum, and
maximum temperature; precipitation patterns; and impacts on cyclical climate patterns, such as
the El Nino-Southern Oscillation (ENSO; IPCC, 2007).
The IPCC projects that the average global surface air temperature in the years 2090-2099
is likely to be between 1.8 to 4.0°C warmer than such temperatures in 1980-1999, depending on
which climate scenarios are input into the models. Models projecting the global distribution of
temperature change project that North America will experience more warming than the average
global warming, with the United States projected to have an increase of 2 to 3°C in the western,
southern, and eastern areas, with up to 5°C in annual mean temperature by 2100 projected for
Alaska (IPCC, 2007). Rising minimum temperatures are expected to result in fewer cold days,
frost days, and cold waves globally (IPCC, 2007). Change has already been detected, with
global average temperature increasing by more than 0.7°C over the past 100 years, with
corresponding increases in the frequency of hot days and decreases in the frequency of cold
nights (IPCC, 2007).
1-1
-------
Along with increasing temperatures, other effects of climate change, such as changes in
precipitation, have also been projected. Globally, precipitation is projected to increase in all
areas, with greater certainty for higher latitudes. Projections of precipitation show increases for
the northeastern United States, with decreases for the southwest (IPCC, 2007). Present
observations indicate that annual land precipitation has increased in the eastern United States
(IPCC, 2007).
Overall, heavy precipitation has increased over most land areas, and hurricanes have
increased in intensity in the North Atlantic since records began in 1970. The intensity of heavy
precipitation events, including hurricanes, is projected to increase, although there is less
confidence about projections for their frequency under climate change (IPCC, 2007).
Conversely, episodes of drought have occurred more frequently and intensely in recent decades
in some regions, particularly the tropics and subtropics (IPCC, 2007).
Overall, the frequency of floods, extreme precipitation events, heat waves, and other
extreme weather events have been projected to increase. It is believed that some of the expected
changes will be caused by the projected increase in temperature because higher temperatures can
affect the hydrological cycle by increasing evaporation and allowing more water vapor to be held
in the atmosphere. It remains unclear if the frequency of small-scale weather events such as
thunderstorms and tornadoes will change, because there is insufficient information to include
these in global models (IPCC, 2007).
ENSO events are cyclical changes in sea surface temperatures and air pressure, and they
result in short-term episodes of increased climate variability. El Nino events generally occur in
cycles of 3 to 6 years, and there are periods of higher and lower strength of ENSO on larger time
scales of decades, causing very different impacts globally (IPCC, 2007). ENSO events are
associated with changes in precipitation globally, with some regions becoming wetter during El
Nino events (e.g., eastern North America) and other areas getting drier (e.g., western North
America). The strong influence of ENSO on tropical monsoons means that any changes in
ENSO due to climate change can affect rainfall patterns in many parts of the Pacific region
(IPCC, 2007). El Nino events have occurred with greater frequency over the last few decades,
with the El Nino event of 1997-1998 being the strongest recorded (Sutherst, 2004). All climate
models project that ENSO interannual variability will continue; however, there are many
1-2
-------
differences between models about changes to ENSO due to climate change. Additionally, there
are other measures of global climate variability (IPCC, 2007).
The potential impacts of these climatic changes on aeroallergens and allergic diseases in
the United States are unclear. Current research has focused on examining how specific elements
of climate change (e.g., increased CC>2 levels, changing temperatures, and increased and
decreased regional precipitation) can alter the production, distribution, and allergen content of
aeroallergens. A change in any of these characteristics of aeroallergens could lead to a
substantial increase in the overall prevalence of allergic diseases in the United States, above and
beyond the increase already observed.
This report provides an overview of the literature detailing the potential impacts of
climate change on aeroallergens and their associated allergic diseases in the United States.
Section 2 provides background information on the major aeroallergens in the United States and
the allergic diseases associated with them. Section 3 outlines historical trends in levels of
aeroallergens and the prevalence of their associated allergenic illnesses in the United States.
Section 4 discusses the evidence of links between climate variability and aeroallergens in the
United States, especially the potential indirect impacts on the allergenic illnesses associated with
them and the evidence of links between climate variability and allergic diseases. Section 5
discusses the economic and quality-of-life implications of allergenic diseases on populations in
the United States. Finally, Section 6 addresses the current gaps in knowledge about the impacts
of climate change on aeroallergens and allergenic illnesses.
1-3
-------
1-4
-------
2. AEROALLERGENS AND ASSOCIATED ALLERGIC DISEASES IN THE
UNITED STATES
2.1. AEROALLERGENS
Aeroallergens are classified into three primary categories: pollen, mold, and indoor
allergens. Clinically relevant aeroallergens in North America, identified in Practice Parameters
for Allergen Immunotherapy, the 2003 publication by the Joint Council of Allergy, Asthma and
Immunology (JCAAI), are shown in Table 2-1.
The JCAAI identified the most common aeroallergens through consensus opinion of
experts, rather than through evidence derived from clinical studies identifying a causal
relationship between aeroallergen exposure and an allergic illness (White et al., 2005). In
contrast, White et al. (2005) defined major tree pollen allergens as those aeroallergens in which
percutaneous reactivity to a given tree pollen extract resulted in more than 50% of all patients
having a positive skin prick test. Using this definition, White et al. (2005) identified American
sycamore, American elm, box-elder, red maple, white ash, cottonwood, and black walnut—all
included in Table 2-1 above—as "major allergens," but they did not recognize mulberry, also
included in Table 2-1, as a major aeroallergen. As this example shows, the inconsistencies in the
definitions of a "major aeroallergen" can result in the identification of clinically relevant
aeroallergens not listed in Table 2-1. Other examples include the grass pollen orchard (Dactylis
glomerata), Kentucky blue grass (Poapratensis), Red top (Agrostis alba), and Sweet vernal
(Anthoxanthum odoratum) (American Academy of Allergy, Asthma & Immunology [AAAAI],
2002). Throughout this report, Table 2-1 will be taken to represent the most common
aeroallergens; however, regional differences, future changes in plant populations, and differences
in the definition of what constitutes a major aeroallergen could result in the addition of
aeroallergens to this list in the future (White and Bernstein, 2003).
2.1.1. Pollen
The major pollen allergens are divided into three subcategories: tree pollen, grass pollen,
and weed pollen. The pollen size for all of the subcategories varies from 5 um to greater than
200 um (Wood, 1986). The pollen of each species has a distinct distribution, season of
pollination, and level of dispersal, as discussed in detail by Kosisky and Carpenter (1997);
2-1
-------
Table 2-1. Selection of clinically relevant aeroallergens in the United States
Latin name
Common name
Tree pollen
Acer negundo
Acer rubra
Alnus rubra
Betula papyri/era
Carya illinoensis
Fraxinus Americana
Juglans nigra
Juniperus ashei
Moms alba
Olea europaea
Plantanus occidentalis
Populus deltoids
Quercus alba
Quercus rubra
Ulmus Americana
Ulmus parvifolia
Ulmus pumila
Box-elder
Red maple
Alder
Paper birch
Pecan
White ash
Black walnut
Mountain cedar
Mulberry
Olive
American sycamore
Eastern cottonwood
White oak
Red oak
American elm
Chinese elm
Siberian elm
Grass pollen
Cynodon dactylon
Festuca elatior
Holcus halepensis
Lolium perenne
Paspalum notatum
Phleum pretense
Bermuda
Meadow fescue
Johnson
Rye
Bahia
Timothy
Weed pollen
Amaranthus retroflexus
Ambrosia artemisiifolia
Artemisia vulgaris
Kochia scoparia
Plantago lanceolata
Rumex acetosella
Salsola kali
Red root pigweed
Short ragweed
Mugwort
Burning bush
English plantain
Sheep sorrel
Russian thistle
2-2
-------
Table 2-1. Selection of clinically relevant aeroallergens in the United States
(continued)
Latin name
Common name
Mold
Alternaria alter nata
Aspergillus fumigatus
Cladosporium (C.
cladosporioides; C. herbarum)
Drechslera or Bipolaris type
(e.g., Helminthosporium solani)
Epicoccum nigrum
Penicillium (P. chrysogenum;
P. expamum)
N/A
N/A
N/A
N/A
N/A
N/A
Indoor allergens
Felis domesticus
Canis familiaris
Dermatophagoides farinae ;
Dermatophagoides pteronyssinus
Blattella germanica
Cat (epithelium)
Dog (epithelium)
House dust mites
German cockroach
Source: Joint Task Force on Practice Parameters (2003).
however, within a pollen type (e.g., tree pollen), there are many similarities across species. In a
study observing pollen levels during a 5-year period in Washington, D.C., Kosisky and
Carpenter (1997) found tree pollen accounts for approximately 90% of the total annual pollen
produced, with weeds and grasses accounting for 6 and 3%, respectively. The results reported by
Kosisky and Carpenter (1997) are consistent with the results of similar studies observing yearly
pollen levels. For example, in a study conducted in Philadelphia and Southern New Jersey,
Dvorin et al. (2001) found that tree pollen accounts for the largest percent (approximately 75%)
of the total annual pollen produced. Although there are clear differences in the amounts of
different types of pollen produced, other factors, including prevailing winds and the pattern of
land use, may also affect the level of airborne allergens in an area (Wood, 1986). The following
sections discuss the defining characteristics of each pollen type, including the distributions of the
relevant plant species within the United States, the pollen seasons, and the levels of pollen
dispersal.
2-3
-------
2.1.1.1. Tree Pollen
Tree pollen accounts for the largest percent of pollen produced during the pollen
season—approximately 75 to 90% (Dvorin et al., 2001; Kosisky and Carpenter, 1997). Of the
total amount produced, however, only a small percentage is pollen generated from clinically
relevant tree species. During a study of 5-year mean tree pollen counts, for example, White et al.
(2005) found that "major allergens" accounted for only 5% or less of the total 5-year mean tree
pollen count.l In a similar study, Kosisky and Carpenter (1997) found oak pollen represented
approximately 57% of the pollen produced during a 6-year period in Washington, D.C.;
however, there are 20 species of oak in the D.C. area, only two of which (white oak and red oak)
produce pollen counted among the major aeroallergens. This suggests that, in the case of
Washington, D.C., allergenic tree pollen may represent only a small percentage of the total tree
pollen produced on a yearly basis. Nevertheless, a small percentage still represents a large
number of persons affected.
Studies conducted by Weber (2003a, b) and White and Bernstein (2003) identified
regions of the major tree pollen allergens in the United States by hardiness zones (i.e., climatic
zones) defined by the United States Department of Agriculture (Weber, 2003a, b). White and
Bernstein (2003) went a step farther and defined geographic locations for each pollen type
through the designation of east or west, with the dividing line between east and west, running
from the middle of Montana diagonally to just east of the southern tip of Texas. One or more
U.S. Census Bureau geographic regions (e.g., Midwest [MW], Northeast [NE], South [S], and
West [W]) was assigned to each tree pollen type based on the tree growth region data provided in
Weber (2003a, b), and White and Bernstein (2003) and National Center for Health Statistics,
(2004) (Table 2-2). Overall, the distribution of tree pollen spans the entire United States, but the
abundance of pollen produced by certain tree species can vary within their defined geographic
region(s).
1 White et al. (2005) identified American sycamore, American elm, box-elder, red maple, red oak, white ash,
cottonwood, and black walnut as "major allergens." They did not classify mulberry as a major allergen; therefore, it
is possible that the total percent contribution of "major allergens" to the 5-year mean tree pollen count could be
slightly larger than 5%.
2-4
-------
Table 2-2. Geographic distribution of major clinically relevant tree pollen in
the United States
Latin name
Acer negundo
Acer rubra
Alnus rubra
Betula papyri/era
Carya illinoensis
Fraxinus americana
Juglans nigra
Juniperus ashei
Moms alba
Olea europaea
Plantanus occidentalis
Populus deltoids
Quercus alba
Quercus rubra
Ulmus Americana
Ulmus parvifolia
Ulmus pumila
Common name
Box-elder
Red maple
Alder
Paper birch
Pecan
White ash
Black walnut
Mountain cedar
Mulberry
Olive
American sycamore
Eastern cottonwood
White oak
Red oak
American elm
Chinese elm
Siberian elm
Geographic region(s)a
MW, ME, S, W
MW, ME, S
W
MW, NE, W
MW, S
MW, NE, S
MW, NE, S
AR, MO, OK,TXb
MW, NE, S
Wc
MW, NE, S
NE, S
MW, NE, S
MW, NE, S
MW, NE, S
MW, NE, S, W
MW, NE, S, W
a MW = Midwest; NE = Northeast; S = South; W = West.
bMountain cedar is located throughout the United States but is highly prevalent in central
Texas and other areas of the southern Great Plains (Levetin and Van de Water, 2003).
°Olive is most prevalent in the Southwest United States (White and Bernstein, 2003).
Sources: Weber (2003a, b), White and Bernstein (2003).
The pollen seasons of the clinically relevant tree species are shown in Table 2-3. The
overall pollen season for tree pollen tends to last from early March to mid-May; although in
some cases, it can run from February to June (Kosisky and Carpenter, 1997; White et al., 2005).
The one exception to this is the unique pollen season of Mountain Cedar (Juniperus ashei),
which ranges from December to January (Levetin and Van de Water, 2003).
The period from late April to early May is of particular importance because this is the
period with the highest pollen prevalence due to considerable overlap of the pollen seasons of
multiple tree species (Dvorin et al., 2001). April, in particular, has been found to have the
highest weekly average pollen concentrations (Kosisky and Carpenter, 1997). During the pollen
2-5
-------
season, multiple tree species will release pollen at the same time, resulting in a significant
amount of pollen being dispersed. The release of pollen from these tree species and,
subsequently, all tree species during the pollen season can result in the weekly average pollen
concentration per tree exceeding 100 grains/m3, with the cumulative pollen abundance over the
pollen season for each tree ranging upwards of 1,800 grains/m3 (Dvorin et al., 2001).
Table 2-3. Pollen seasons of the major clinically relevant tree pollen in the
United States
Latin name
Acer negundo
Acer rubra
Alnus rubra
Betula papyri/era
Carya illinoensis
Fraxinus americana
Juglans nigra
Juniperus ashei
Moms alba
Olea europaea
Plantanus
occidentalis
Populus deltoids
Quercus alba
Quercus rubra
Ulmus Americana
Ulmus parvifolia
Ulmus pumila
Common name
Box-elder
Red maple
Alder
Paper birch
Pecan
White ash
Black walnut
Mountain cedar
Mulberry
Olive
American
sycamore
Eastern
cottonwood
White oak
Red oak
American elm
Chinese elm
Siberian elm
Pollen season
Early spring
Mid-April to Mid-May
February to April
Late April to Late May
April to June
April to May
Late spring (May) to
Early summer
December to January
Spring; April to May
Spring
March to April
March to April
March to May
March to April
February to Marcha
Fall
February to Marcha
Reference
Phadia, 2002
Dvorin etal., 2001
Weber, 2003 a, b
Dvorin etal., 2001
Phadia, 2002
Phadia, 2002
Levetin, 2006; Phadia,
2002
Levetin and Van de
Water, 2003
Levetin, 2006; Phadia,
2002
Phadia, 2002
Levetin, 2006
Levetin, 2006
Dvorin etal., 2001;
Levetin, 2006
Levetin, 2006
Levetin, 2006
Tidwell, 2006
Tidwell, 2006
aPollen season can possibly extend to April (Saint Louis County, 2006).
2-6
-------
2.1.1.2. Grass Pollen
Grass pollen accounts for the smallest percent of pollen produced during the pollen
season—approximately 3 to 10% (Dvorin et al., 2001; Kosisky and Carpenter, 1997). The
literature does not address what percentage of the total grass pollen count is composed of the
clinically relevant grass pollen, however, as it does for tree pollen; therefore, the total amount of
clinically-relevant grass pollen produced on a yearly basis is not well-defined.
As for tree pollen, the distributions of grass pollen within the United States were
determined using data detailed in Weber (2003a, b) and White and Bernstein (2003), and then
extrapolated to the United States regions defined by the U.S. Census Bureau. Table 2-4 shows
the geographic regions of the most common clinically relevant grass pollen in the United States.
Consistent with what has been reported for tree pollen, the distribution of grass pollen can vary
considerably within defined geographic region(s). Grass pollen is usually deposited within 50
miles of its release, and although the exact distance can vary, it will mostly be confined to the
relative vicinity in which it grows (Wood, 1986).
Table 2-4. Geographic distribution of major clinically relevant grass pollen
in the United States
Latin name
Cynodon dactylon
Festuca elatior
Holcus halepensis
Lolium perenne
Paspalum notatum
Phleum pretense
Common name
Bermuda
Meadow fescue
Johnson
Rye
Bahia
Timothy
Geographic region(s)a
MW, S, Wb
MW, NE, S, Wc
MW, S, Wb'd
MW, NE, S, Wc
MW, S, Wb
MW, NE, S, Wc
a MW = Midwest; NE = Northeast; S = South; W = West.
bBermuda, Johnson, and bahia are all located in the southern part of each region from the East to the
West coast of the United States and are becoming increasingly more important in the south
(Phipatanakul, 2005; White and Bernstein, 2003).
°Meadow fescue, Rye, and Timothy are all located in the northern part of each region from the East
to the West coast of the United States (White and Bernstein, 2003).
dThe growing region of Johnson extends slightly further north than that of Bermuda and bahia
(White and Bernstein, 2003).
Sources: Weber (2003a, b), White and Bernstein (2003).
2-7
-------
Unlike tree pollen, some grass pollen, including Bermuda, Johnson, and Bahia, are
produced all year (Weber, 2003a, b). Dvorin et al. (2001) found that for the majority of grasses,
the pollen season tends to last from late April to mid-June, with a secondary peak in early
September. These findings are consistent with what has been defined as the peak grass pollen
season, from May through June (Gonzalez Minero et al., 1998). Table 2-5 shows the pollen
season for each of the clinically relevant grass pollens.
During the peak months of the grass pollen season, the cumulative weekly average
concentration of pollen is typically >100 grains/m3, with the cumulative amount of pollen
produced in a single year not exceeding 2,500 grains/m3 (Gonzalez Minero et al., 1998). During
the pollen season, grass pollen levels can oscillate both during the season and throughout a
region due to anthropogenic factors. The levels can vary depending on the amount of land
covered in grass; the seed mix of sown pastures, and the replacement of haymaking by silage
production, when grasses are cut before they flower (Nielsen et al., 2002).
Table 2-5. Pollen season of the major clinically relevant grass pollen in the
United States
Latin name
Cynodon dactylon
Festuca elatior
Holcus halepensis
Lolium perenne
Paspalum notatum
Phleum pretense
Common name
Bermuda
Meadow fescue
Johnson
Rye
Bahia
Timothy
Pollen season
Late April to mid- June, early September
Late April to mid- June, early September
Late April to mid-June, early September
Late April to mid- June, early September
Late April to mid- June, early September
Late April to mid- June, early September
2.1.1.3. Weed Pollen
Weed pollen accounts for the second greatest percentage of pollen produced during the
pollen season—approximately 6 to 17%. However, the amount produced is significantly less
than the total amount of tree pollen produced in a single year (Dvorin et al., 2001; Kosisky and
Carpenter, 1997). Similar to grass pollen, the literature for weed pollen does not address what
percentage of the total weed pollen count is composed of the clinically relevant weed pollen;
therefore, the total amount of clinically relevant weed pollen produced on a yearly basis is not
clearly defined.
2-8
-------
As for both tree and grass pollen, the distribution of weed pollen within the United States
was determined using data detailed in Weber (2003a, b) and White and Bernstein (2003), and
then extrapolated to the United States regions defined by the U.S. Census Bureau.
Table 2-6 details the geographic regions of the most common weed pollen in the United
States. Although all of the major weeds are located throughout the United States, some are more
highly prevalent in specific regions of the country.
Table 2-6. Geographic distribution of major clinically relevant weed pollen
in the United States
Latin name
Amaranthus retroflexus
Ambrosia artemisiifolia
Artemisia vulgaris
Kochia scoparia
Plantago lanceolata
Rumex acetosella
Salsola kali
Common name
Red root pigweed
Short ragweed0
Mugwort
Burning bush
English plantain
Sheep sorrel
Russian thistle
Geographic region(s)a
MW, NE, S, Wb
MW, NE, S, Wd
MW, NE, S, We
MW, NE, S, W
MW, NE, S, W
MW, NE, S, Wb
MW, NE, S, Wb
a MW = Midwest; NE = Northeast; S = South; W = West.
bFound throughout the United States, but especially in the western half of the United States
(Powell and Smith, 1978; White and Bernstein, 2003).
°Not found in the Pacific Northwest (Phipatanakul, 2005).
dNot found in Utah, Nevada, and California (White and Bernstein, 2003).
eHighly localized to the eastern United States and Pacific Northwest (White and Bernstein,
2003).
Sources: Weber (2003a, b), White and Bernstein (2003).
Unlike the tree and grass pollen seasons, which are relatively consistent across all
species, the pollen season for weeds has been shown in multiple studies to vary across species
(Table 2-7). In some cases, the region of the country in which the weed species is located can
influence the pollen season. For example, in most areas of North America, ragweed pollinates
from August through October, but the pollen season tends to be earlier in northern areas and
progressively later in southern states (Levetin and Van de Water, 2003). Overall, the weed
pollen season is typically defined as mid-August through late-September (Dvorin et al., 2001).
2-9
-------
Table 2-7. Pollen season of the major clinically relevant weed pollen in the
United States
Latin name
Amaranthus retroflexus
Ambrosia artemisiifolia
Artemisia vulgaris
Kochia scoparia
Plantago lanceolata
Rumex acetosella
Salsola kali
Common name
Red root
pigweed
Short ragweed
Mugwort
Burning bush
English Plantain
Sheep sorrel
Russian thistle
Pollen season
Late Summer and
Autumn
March to November21
August to October
Mid-Summer
July to Augustb
April to Mayc; Mid-
August to Late
September
Late Summer and
Autumn
Reference
Phadia, 2002
Weber, 2003 a, b
White and Bernstein,
2003
Phadia, 2002
Weber, 2003 a, b
Dvorinetal., 2001;
Weber, 2003 a, b
Phadia, 2002; Powell and
Smith, 1978
aPollen season is August to October in northern regions of the United States (Dvorin et al., 2001; White and
Bernstein, 2003).
bPollen season may extend slightly longer, May to October, with the peak being May to July (White and Bernstein,
2003).
'Specific to western United States.
During the peak pollen season, from mid-August through late-September, weed pollen
levels may exceed 250 grains/m3 weekly (Dvorin et al., 2001). The total amount of pollen
released during the pollen season can vary from region to region, with the total amount of pollen
released being determined by the prevalence of each weed species in each geographic region of
the United States. For example, although it is found throughout the United States, ragweed has
high pollen counts in the Omaha region of the Midwest, which will highly influence the overall
weed pollen count in that region of the country (Weber, 2003a, b).
2.1.2. Mold
The second major class of clinically relevant aeroallergens is mold. Mold spores are
substantially smaller than pollen grains, ranging in size from 2 to 10 um, and are more abundant
(Burge, 2002). Mold counts are often 1,000-fold greater than pollen counts (Bush and Prochnau,
2-10
-------
2004).2 Unlike pollen, mold is not localized to specific regions of the country; it can be found
throughout the United States, except in the coldest regions, but it can be found in higher
concentrations in some regions due to specific environmental conditions, most notably humidity
(Phipatanakul, 2005) (Table 2-8). Mold requires a consistently high relative humidity, ranging
between 70 and 85% (Burge, 2002; Hamilton and Eggleston, 1997).
Table 2-8. Geographic distribution of major clinically relevant mold in the
United States
Latin name
Alternaria alternata
Aspergillus fumigatus
Cladosporium (C. cladosporioides;
C. herbarum)
Drechslera or Bipolaris type (e.g.,
Helminthosporium solani)
Epicoccum nigrum
Penicillium (P. chrysogenum;
P. expansum)
Geographic region3
Grain-growing areas
Warm climates
(>40°C)
Temperate zones
N/A
N/A
N/A
Reference
Corden and Millington, 2001;
Targonski et al., 1995
Hamilton and Eggleston, 1997
Hamilton and Eggleston, 1997
N/A
N/A
N/A
aStudies detailing common mold aeroallergens do not address their distribution within the United States. The
literature has hinted at mold being found ubiquitously in the United States. Areas or regions of the United States
are included for those types of mold where information was available.
Mold is primarily located outdoors, but unlike pollen, can colonize indoor materials
(Burge, 2002). Alternaria and Cladosporium are universally dominant outdoor fungal species
that are detected indoors, while Penicillium and Aspergillus are universally dominant indoors
(Hamilton, 2005). Burge et al. (2002) found that the concentrations of outdoor fungal species in
indoor environments are driven by outdoor concentrations. Indoors, the distribution of fungal
concentrations throughout the aboveground living space of a home is fairly consistent, with the
highest concentrations being found in basements due to ideal growing conditions, but the types
found in basements are usually not related to those found outdoors (Burge, 2002).
2Hamilton and Eggleston (1997) state, although fungal counts are substantially larger than those observed for pollen,
it is currently unclear if the viable spore colony count or the total (viable and nonviable) spore count is a better
indicator for clinically relevant mold allergens in the environment.
2-11
-------
The literature focuses primarily on Alternaria, one of the more common atmospheric
mold spores in the United States (Corden and Millington, 2001). Alternaria flourishes in warm,
humid environments (Hamilton and Eggleston, 1997). It grows well on fruits and tomatoes, as
well as textiles, allowing it to flourish in indoor environments; however, it is usually not found
indoors (Corden and Millington, 2001; Hamilton and Eggleston, 1997). Alternaria is found in
highest concentrations in cultivated areas, such as the Midwest, in which grasslands and grain
fields predominate (Bush and Prochnau, 2004). In studies conducted in Derby, UK by Corden
and Millington (2001), and in Chicago by Targonski et al. (1995), seasonal Alternaria
concentrations were observed primarily from June to October, and July to October, respectively,
periods which coincide with harvest time, although spores were occasionally found at other times
throughout the year. Cladosporium is the most abundant mold spore in temperate parts of the
world. Alternaria counts are 10 to 100 times lower than Cladosporium.
Unfortunately, information for the other clinically relevant molds is limited. Specific
regions of growth and periods of highest concentration have been identified for only a few mold
types, as shown in Table 2-8. Cladosporium thrives in temperate zones, and Aspergillus thrives
in warm climates (>40°C), while Penicillium grows on stale bread, citrus fruits, and apples
(Hamilton and Eggleston, 1997). It is unclear if a specific time of year is associated with
increased concentrations of Cladosporium and Penicillium, but Aspergillus concentrations do
have seasonal peaks, provided they penetrate indoor environments when heating is the highest,
during autumn and winter (Hamilton and Eggleston, 1997).
There are many other fungal species that have been studied for allergenicity,3 including
molds, mushrooms, and yeasts; however, few of the fungal allergens have been
well-characterized, possibly due to the complexity and large number of fungal spores (Horner et
al., 1995). Thus, this report limits itself to the well-studied mold allergens, noting that there may
be many more important fungal allergens discovered in the future.
2.1.3. Indoor Allergens
Similar to mold, indoor allergens are not particularly associated with specific regions of
the United States. Indoor environments have been found to be the main determinant influencing
the level of indoor allergens. It has been postulated that an increase in the price of energy has
3Allergenicity refers to the degree to which a protein is likely to elicit an allergic response. However, the term is
periodically used in the literature in reference to pollen protein concentrations.
2-12
-------
resulted in an increase in insulation and a decrease in ventilation in buildings, providing ideal
growth conditions for the most prevalent indoor allergen, house dust mites (Nielsen et al., 2002).
House dust mites are ubiquitous throughout the United States except in very dry climates and at
higher elevations (Phipatanakul, 2005). They have also been found to thrive in warm conditions
where the relative humidity is approximately 70% (Hamilton, 2005). Cockroaches, on the other
hand, are found more predominantly in urban areas, particularly in inner city, low-income
environments, but are also more common than previously thought in suburban middle-class
homes (Hamilton, 2005; Phipatanakul, 2005). The concentrations of all indoor allergens do not
vary with season as is observed for pollen and some mold, but are instead found perennially.
Indoor allergens are not confined to specific regions of the United States, but factors have been
identified that influence the level of allergens found indoors. Major clinically relevant indoor
allergens include cat epithelium, dog epithelium, domestic mites, and German cockroaches
(Table 2-9).
Table 2-9. Geographic distribution of major clinically relevant indoor
allergens
Latin Name
Penicillium
Felis domesticus
Canis familiaris
Dermatophagoides farinae ;
Dermatophagoides pteronyssinus
Blattella germanica
Common name
Penicillium mold
Cat (epithelium)
Dog (epithelium)
Dust mites
German cockroach
Geographic region(s)a
N/A
N/A
N/A
N/A
N/A
alndoor allergens are not confined to specific regions of the United States, but factors have been
identified that influence the levels of allergens found indoors.
2.2. ASSOCIATED ALLERGIC DISEASES
Exposure to allergens results in allergic diseases in approximately 20% of the United
States population (AAAAI, 1996-2006). The development of allergic diseases occurs through a
two-stage process. In the first stage, an immunologically naive individual is sensitized to an
allergen, resulting in the production of IgE antibodies. In the second stage, renewed exposure to
the allergen elicits a disease response due to the presence of IgE antibodies and the associated
cellular response (Nielsen et al., 2002). Currently, three main allergic diseases have been
2-13
-------
associated with exposure to aeroallergens: allergic rhinitis (hay fever), asthma, and atopic
dermatitis (eczema).
The initial sensitization to an aeroallergen can occur during any period of an individual's
life. Wood (1986) cites a study by Ziering and Klein (1982), which found that respiratory
allergies develop by 2 years of age in 40% of those affected and by 6 years of age in the
remaining 60%. Wood (1986) also cites a study by Kemp (1979) that found that the sensitivity
to grass pollen of children who reached the age of 3 months during a time of high environmental
exposure to grass pollen was significantly greater than the sensitivity to grass pollen of children
born at other times of the year. Although sensitization and the subsequent development of
allergic diseases can occur during childhood, sensitization to common aeroallergens increases
with age and with the length of the exposure period (Nielsen et al., 2002). The German
Multicenter Allergy birth cohort study, for example, observed rates of sensitization to grass
pollen and dust mites of 6.2 and 3.0%, respectively, before the age of 2, but as the children grew
older, the rates of sensitization to both outdoor and indoor allergens increased (Phipatanakul,
2005). The incidence of allergic rhinitis was observed in the study to increase by as much as 3 to
4% each year after the age of three (Phipatanakul, 2005).
Underlying genetic factors have been found to have a strong influence on the process of
sensitization and the subsequent development of allergic diseases during the course of an
individual's life. Individuals classified as atopic (a probably hereditary allergy) are inheritably
predisposed to produce elevated amounts of IgE antibodies upon exposure to allergens, and as a
result, are more easily sensitized to allergens than are nonatopic individuals (Nielsen et al.,
2002). The hereditary association between aeroallergen exposure and allergic illness
development has been identified as a primary risk factor for the development of allergic rhinitis
in children, especially if both parents are affected by the illness (Phipatanakul, 2005). Although
there is a major hereditary contribution to the development of these allergic diseases,
environmental factors, specifically exposure to aeroallergens, play a significant role in their
manifestation (Nielsen et al., 2002).
The degree to which an aeroallergen causes an allergic illness in a sensitized individual
depends on multiple factors, one of the primary factors being the aeroallergen to which the
individual is exposed. Galant et al. (1998) performed skin prick tests for different allergens on
individuals in California with allergic rhinitis and asthma and found that some allergens are more
2-14
-------
prone to result in the development of allergic diseases than others. The study showed the
following rank order of positive responses: pollen (grasses > weeds > trees) and house dust mites
> pets (cat > dog) > cockroach and mold (Galant et al., 1998). These findings are consistent with
the findings in other studies that have examined the association between aeroallergen exposure
and the development of allergic diseases. For example, Nielsen et al. (2002) reported that allergy
to mold alone has low projective value for the development of asthma and allergic rhinitis. In
one study, 15% of subjects sensitized exclusively to mold had allergic symptoms, but subjects
sensitized to mold and pollen and/or house dust mites had a prevalence of allergic symptoms of
about 50%, suggesting that sensitization to mold alone is not as important in causing allergic
symptoms as sensitization to the other aeroallergens (Nielsen et al., 2002). In addition, a study
conducted in central Indiana found the sensitization rate to mold was only about half the
sensitization rate for pollen (Nielsen et al., 2002).
Not only the type, but also the amount of aeroallergen to which an individual is exposed
is influential in the development of an allergic illness. Similar to what is observed in most
disease causation scenarios, a dose-response relationship between aeroallergen exposure and
sensitization and exacerbation of disease has been observed—i.e., sensitized patients are more
likely to have more severe disease if exposure to allergens is high (Nielsen et al., 2002). This
relationship was observed in a study conducted in France, which looked at hay fever and grass
pollen sensitivity. The study found the prevalence of allergy to a given allergen is higher in
communities that are heavily exposed to allergens than those that are not (Burr, 1999). Although
the probability of an allergic response increases with increasing levels of exposure to
aeroallergens, a large exposure is not required to initiate allergic symptoms. Comtois and
Gagnon (1988) found that it only takes a small amount, 9 to 23 grains/m3 of tree pollen and 4 to
12 grains/m3 of grass pollen, to cause allergic symptoms in an already-sensitized individual.
The observation of a dose-response relationship between aeroallergen exposure and the
development of allergic illness is not specific to pollen exposure; such dose-response
relationships have also been observed for individuals sensitized to indoor allergens, specifically
house dust mites. Nielsen et al. (2002) found the level of indoor allergen exposure highly
influenced the severity of asthma. Because of this, exposure reduction is one of the main
methods used to control the development of allergic illness in sensitized individuals (Nielsen et
al., 2002).
2-15
-------
Although there may be a dose-response relationship between aeroallergen exposure and
the development of allergic illness, other confounding factors may make this relationship
difficult to observe. If, for example, the proportion of the population that is genetically
predisposed to develop allergic diseases happens to decrease as the level of the aeroallergen
increases, a dose-response relationship could be masked. Although multiple studies have shown
a correlation between aeroallergen levels and disease development, this is not the case for all
such studies. White et al. (2005) found no association between regional pollen levels and the
frequency of skin test reactivity to specific tree pollen allergens in a study conducted in
Southwestern Ohio. These researchers noted that their findings might be specific to
Southwestern Ohio; however, these findings call into question whether increased exposure to
aeroallergens elicits the same disease response throughout the United States.
All of the factors discussed above influence the development of allergic diseases in
individuals exposed to aeroallergens. Table 2-10 shows the allergic diseases associated with
exposure to each of the clinically relevant aeroallergens listed in Table 2-1.4 The allergic
diseases associated with exposure to aeroallergens are discussed more fully below, including the
evidence supporting causal relationships between aeroallergen exposure and disease
development.
2.2.1. Allergic Rhinitis
The most common allergic illness associated with exposure to aeroallergens is allergic
rhinitis (hay fever). Allergic rhinitis is also commonly referred to as rhinoconjunctivitis—
because the clinical manifestations associated with the condition may include not only sneezing,
itching rhinorrhea, or nasal congestion, but, also, itchy, red, and watery eyes (conjunctivitis)
(Phipatanakul, 2005). It is also sometimes called pollinosis, because seasonal allergic rhinitis is
primarily caused by airborne pollen (Nielsen et al., 2002). In some cases, the symptoms of
allergic rhinitis may also affect the ears and throat and include postnasal drainage (Phipatanakul,
2005). All of these symptoms result from exposure to aeroallergens after an initial sensitization;
hence allergic rhinitis is termed a type 1 or immediate hypersensitivity reaction (Wood, 1986).
4Table 2-10 does not include atopic dermatitis (eczema) because the literature to date has not definitively concluded
that there is a casual association between aeroallergen exposure and atopic dermatitis development (Whitmore et al.,
1996).
2-16
-------
Table 2-10. Allergic diseases correlated with the major clinically relevant
aeroallergens
Latin name
Common name
Allergic illness
Reference
Tree pollen
Acer negundo
Acer rubra
Alnus rubra
Betula papyrifera
Carya illinoensis
Fraxinus americana
Juglans nigra
Juniperus ashei
Morus alba
Olea europaea
Plantanus
occidentalis
Populus deltoids
Quercus alba
Quercus rubra
Ulmus Americana
Ulmus parvifolia
Ulmus pumila
Box-elder
Red maple
Alder
Paper birch
Pecan
White ash
Black walnut
Mountain cedar
Mulberry
Olive
American
sycamore
Eastern
cottonwood
White oak
Red oak
American elm
Chinese elm
Siberian elm
Asthma, Allergic rhinitis
Allergic rhinitis
Allergic rhinitis3
Asthma, Allergic rhinitis
Allergic rhinitis3' d
Asthma, Allergic rhinitis
Allergic rhinitis
Asthma, Allergic rhinitis
Asthma, Allergic rhinitis
Asthma, Allergic rhinitis
Asthma, Allergic rhinitis
Asthma, Allergic rhinitis
Allergic rhinitis
Allergic rhinitis
Asthma, Allergic rhinitis
Allergic rhinitis3
Allergic rhinitis3
Phadia, 2002; White et al.,
2005
White et al., 2005
Nielsen et al., 2002
White etal., 2005; White
and Bernstein, 2003;
Emberlin et al., 2002
White et al., 2005
Phadia, 2002; White et al.,
2005
White et al., 2005
Phadia, 2002
Phadia, 2002
Phadia, 2002
White etal., 2005; White
and Bernstein, 2003
Phadia, 2002; White et al.,
2005
White etal., 2005; White
and Bernstein, 2003
White etal., 2005; White
and Bernstein, 2003
Phadia, 2002
Nielsen et al., 2002
Nielsen et al., 2002
Grass pollen"
Cynodon dactylon
Festuca elatior
Holcus halepensis
Lolium perenne
Paspalum notatum
Phleum pretense
Bermuda
Meadow fescue
Johnson
Rye
Bahia
Timothy
Asthma, Allergic rhinitis
Asthma, Allergic rhinitis
Asthma, Allergic rhinitis
Asthma, Allergic rhinitis
Asthma, Allergic rhinitis
Asthma, Allergic rhinitis
Nielsen et al., 2002
Nielsen et al., 2002
Nielsen et al., 2002
Nielsen et al., 2002
Nielsen et al., 2002
Nielsen et al., 2002
2-17
-------
Table 2-10. Allergic diseases correlated with the major clinically relevant
aeroallergens (continued)
Latin name
Common name
Allergic illness
Reference
Weed pollen
Amaranthus
retroflexus
Ambrosia
artemisiifolia
Artemisia vulgaris
Kochia scoparia
Plantago lanceolata
Rumex acetosella
Salsola kali
Red root pigweed
Short ragweed
Mugwort
Burning bush
English plantain
Sheep sorrel
Russian thistle
Asthma, Allergic rhinitis
Allergic rhinitis
Asthma, Allergic rhinitis
Asthma, Allergic rhinitis
Asthma, Allergic rhinitis
Allergic rhinitis3
Allergic rhinitis3
Phadia, 2002
White and Bernstein, 2003
Phadia, 2002
Phadia, 2002
Phadia, 2002
Nielsen et al., 2002
Nielsen et al., 2002
Mold
Alternaria alternata
Aspergillus
fumigatus
Cladosporium (C.
cladosporioides; C.
herbarum)
Drechslera or
Bipolaris type (e.g.,
Helminthosporium
solani)
Epicoccum nigrum
Penicillium (P.
chrysogenum; P.
expansum)
N/A
N/A
N/A
N/A
N/A
N/A
Asthma, Allergic0
Asthma
Asthma
Asthmab
Asthmab
Asthmab
Halonen et al., 1997; Corden
and Millington, 2001;
Andersson et al., 2003
Nielsen et al., 2002
Nielsen et al., 2002
Nielsen et al., 2002
Nielsen et al., 2002
Nielsen et al., 2002
aThe literature did not detail a specific allergic illness or diseases associated with exposure to these pollen types.
Exposure to all pollen types is known to cause pollinosis (i.e., allergic rhinitis); therefore, allergic rhinitis was
listed as the associated allergic illness for these pollen types (Nielsen et al., 2002).
bThe literature did not detail a specific allergic illness(es) associated with exposure to these types of mold. Nielson
et al. (2002) states exposure to mold is a primary risk factor for the development of asthma; as a result, asthma was
defined as the associated allergic illness for these types of mold.
°Mold can cause both asthma and allergic rhinitis (Nielsen et al., 2002). Allergic rhinitis is only associated with
exposure to Alternaria in this table because the literature did not provide definitive evidence that the other types of
mold detailed in the table can also cause allergic rhinitis.
dThe literature does not associate a specific allergic illness with exposure to pecan. It only states pecan is highly
allergic (White et al., 2005).
2-18
-------
Allergic rhinitis annually affects approximately 20 to 40 million people in the United
States, including 10 to 30% of adults and up to 40% of children (Gilmour et al., 2006;
O'Connell, 2004). Although exposure to the majority of aeroallergens can result in the
development of allergic rhinitis (Table 2-10), sensitization to pollen is a primary risk factor for
its development (Nielsen et al., 2002). Pollen from wind-pollinated plants are of particular
concern because they are lighter and can become airborne without difficulty, allowing for
individuals to be easily exposed (Wood, 1986; White et al., 2005). The significance of pollen
exposure in the development of allergic rhinitis was highlighted in a study conducted by the
Spanish Society of Clinical Allergy and Immunology. The study found that 65% of pollinosis
cases reported in city hospitals were caused by grass pollen (Gonzalez Minero et al., 1998).
Numerous studies have found that exposure to specific pollen types increases the risk of
developing allergic rhinitis, but it remains unclear which pollen types are more highly associated
with the development of allergic rhinitis. It has been estimated that ragweed pollen is
responsible for 50 to 75% of all allergic rhinitis cases in the United States (American College of
Allergy, 2006; Nielsen et al., 2002), while 20 to 25% of hay fever sufferers are allergic to birch
(Emberlin et al., 2002). A study conducted in Tucson, Arizona, however, found that children
who had immediate skin test responses to Bermuda grass were more prone to develop allergic
rhinitis (Halonen et al., 1997). These findings are consistent with those of Levetin and Van de
Water (2003), who classify Bermuda, Johnson, and bahia as important allergic grasses, but they
also contribute to the puzzle about which plant species has the largest influence on the
development of allergic rhinitis.
The literature on the development of allergic rhinitis in response to aeroallergen exposure
focuses primarily on pollen, but studies have found that exposure to both indoor allergens and
mold can also contribute to the development of allergic rhinitis in sensitized individuals.
Multiple studies have shown a causal relationship between sensitization for hay fever and
exposure to indoor allergens, such as dust mites and cockroaches, as well as exposure to mold.
Although exposures to indoor allergens and pollen both result in the development of allergic
rhinitis, a difference has been observed in the symptom pattern. Unlike allergic rhinitis
symptoms associated with exposure to pollen, which follow the months of the pollen season, the
symptoms associated with exposure to indoor allergens are perennial (Phipatanakul, 2005).
The (minimal) literature on the development of allergic rhinitis associated with exposure
to mold focuses specifically on Alternaria. It has been hypothesized that the smaller spores of
2-19
-------
Alternaria (2 to 8 um) would result in a more potent cause of allergic rhinitis than other types of
mold, such as Cladosporium, which has much larger spores (Andersson et al., 2003). In a study
examining the association between exposure to Alternaria in sensitized children and the
development of allergic rhinitis, Andersson et al. (2003) concluded that sensitized individuals in
regions of the United States with high concentrations of fungal spores are at risk of developing
allergic rhinitis.
2.2.2. Asthma
Second only to allergic rhinitis in prevalence, asthma is one of the primary allergic
diseases associated with exposure to aeroallergens. Unlike allergic rhinitis, which is primarily
associated with exposure to pollen, asthma has been found to be more strongly associated with
exposure to indoor allergens and mold. The Centers for Disease Control and Prevention (CDC)
estimated the prevalence of asthma in the United States adult population as of 2004 to be 7.5%,
or 16 million people, with the overall prevalence in the entire population ranging from 5 to 8%
(Gilmour et al., 2006; O'Connell, 2004). It is unclear, however, what percent of the asthma
cases identified each year can be attributed solely to exposure to aeroallergens. In
epidemiological studies, the proportion of asthmatics who showed an allergic reaction in a skin
prick test to one or more common aeroallergens was usually less than one half (Nielsen et al.,
2002). Therefore, the estimated prevalence of asthma within the United States may not
accurately reflect the prevalence of asthma attributed specifically to exposure to aeroallergens.
Although there is a perceived association between exposure to pollen and asthma
development (Table 2-10), pollen exposure has been historically considered to lead primarily to
hay fever (Burge, 2002). However, recent data suggest a supporting role for exposure to pollen
in the development of asthma (Burge, 2002). In a prospective study conducted in England,
detailed by Burr (1999), most patients with grass pollen sensitivity and a history of seasonal
exacerbations experienced an asthma attack following a rise in pollen count. White and
Bernstein (2003) also found that sensitization to plant aeroallergens is associated with significant
morbidity caused by symptoms of seasonal asthma. Although there is mounting evidence that
suggests that exposure to pollen can lead to asthma, overall sensitization to pollen remains a low
risk factor for asthma development (Nielsen et al., 2002).
The majority of studies examining the development of asthma in response to aeroallergen
exposure have focused on the role of indoor allergens and mold. This is primarily because mold,
allergens from pets, and cockroaches have shown strong associations with asthma development,
2-20
-------
unlike common tree, weed, and grass pollen, which have not shown strong independent
associations (Halonen et al., 1997; Nielsen et al., 2002; Hamilton, 2005; Henderson et al., 2000).
As with exposure to all aeroallergens, including indoor allergens, sensitization influences the
allergic illness an individual will develop. For example, an increased risk of asthma sensitization
in atopic individuals has been associated with house dust mite levels higher than 2000 ng/G of
fine dust (Hamilton, 2005). After sensitization, exposure to house dust mite levels higher than
10,000 ng/G has been associated with an increased risk of asthma symptoms (Hamilton, 2005).
The association between high indoor allergen levels and an increase in asthma severity suggests
a dose-response relationship (Nielsen et al., 2002). The National Co-operative Inner City
Asthma Study clearly implied such a dose-response relationship between indoor allergen levels
and asthma severity when it concluded that children allergic to cockroach allergens and exposed
to high levels had a greater severity of asthma (Custovic et al., 2002).
Multiple studies have found that exposures to mold, including Alternaria, Aspergillis
fumigatus, and Cladosporium, are also risk factors for the development of asthma (Halonen et
al., 1997; Nielsen et al., 2002; Lin and Williams, 2003). In a study conducted in Tucson,
Arizona, Halonen et al. (1997) found that children who had an immediate skin test response to
Alternaria were more prone to develop asthma. Bush and Prochnau (2004) noted that in the
United States, up to 80% of individuals with confirmed asthma have demonstrated positive
reactivity to one or more species of mold. Although there is evidence of associations between
asthma development and exposure to all mold (Table 2-10), the literature focuses primarily on
the development of asthma in response to Alternaria exposure.
Exposure to Alternaria, and subsequently sensitization, has been increasingly recognized
as a risk factor for the development and persistence of asthma, increased asthma severity, and
potentially fatal asthma exacerbations (Nielsen et al., 2002; Bush and Prochnau, 2004).
Similarly, a study conducted in Chicago by Targonski et al. (1995) found that mean mold spores,
rather than tree, grass, or ragweed pollen, was associated with asthma-related deaths. Targonski
et al. (1995) also found the risk of asthma-related deaths increased 2.16 times when the total
Alternaria spore count was about 1,000 spores/m3. Overall, individuals sensitized to Alternaria
appear to be more at risk for developing severe asthma compared to individuals with sensitivities
to other aeroallergens (Bush and Prochnau, 2004).
Although there is evidence to support causal relationships between asthma development
and exposure to both indoor allergens and mold, it is still unclear which class of aeroallergens is
2-21
-------
the greater risk factor for asthma development. Some data, such as those provided by Halonen et
al. and Targonski et al., suggest that mold may have a larger impact on asthma development.
However, a study conducted on children in Virginia and New Mexico found that hypersensitivity
to indoor allergens (e.g., cat and house dust mites) has a stronger association with asthma than
hypersensitivity to mold (Lin and Williams, 2003).
Although these studies suggest that exposure to either indoor allergens or mold can cause
asthma, other researchers have found the evidence for such associations inconclusive (Tortolero
et al., 2002). As a result, some members of the scientific community feel they cannot
definitively state that a direct relationship exists between indoor allergen or mold exposure, and
asthma development. Overall, however, most of the literature suggests that exposure to indoor
allergens and mold in sensitized individuals can result in a strong disposition to both the
development of asthma and subsequent asthma exacerbations.
2.2.3. Atopic Dermatitis
Exposure to aeroallergens has also been implicated in the development of atopic
dermatitis (eczema), and its development has commonly been found to predate the development
of the more prevalent allergic diseases, allergic rhinitis and asthma (O'Connell, 2004). It has
been estimated that atopic dermatitis affects 15 to 20% of the population of children worldwide
(O'Connell, 2004).5 Studies examining the association between aeroallergen exposure and the
development of atopic dermatitis have focused on individual responses to allergens by way of
skin patch or skin prick tests. Most studies have found that 30 to 40% of patients with atopic
dermatitis have positive skin patch tests to allergens (Whitmore et al., 1996). Clark and Adinoff
(1989) found that the most common responses in skin patch tests on individuals with atopic
dermatitis were for animal danders (53%), mites or dust (37%), mold (32%), and tree, grass, and
weed pollen (14 to 35%). Adinoff et al. (1988) also observed positive skin prick tests for
aeroallergens: 30% positive for pollen, 20% for mold, and 75% for dust, mites, and animals.6
Although these studies hint at an association between exposure to aeroallergens and the
development of atopic dermatitis, there is conflicting evidence. Studying patients presenting
with contact dermatitis, Whitmore et al. (1996) found that regardless of whether or not they were
atopic, those suspected of having allergic contact dermatitis had a low incidence of presently
5O'Connell (2004) was the only study that provided a prevalence rate for atopic dermatitis. Unfortunately, the rate
provided is worldwide, although the rest of the paper focuses on United States allergic illness rates.
6This study was not an epidemiological study. It was conducted to examine the possibility that atopic dermatitis may
be triggered by aeroallergens in some individuals; therefore, the findings cannot be used to infer a causal association.
2-22
-------
relevant allergic dermatitis when exposed to aeroallergens (Whitmore et al., 1996). Powell and
Smith (1978), studying individuals sensitized to Russian thistle, observed dermatitis only in
individuals who came into direct contact with the plant, rather than by way of exposure to its
pollen.
Because of this contradictory evidence, the role of aeroallergens in the development of
atopic dermatitis remains controversial (Whitmore et al., 1996). Whitmore et al. (1996) explain
that the uncertainty surrounding the association is due partly to the fact that most of the studies
do not include nonatopic control subjects. As a result, it is unclear if aeroallergens are the
primary culprit in atopic dermatitis (Whitmore et al., 1996).
2.2.4. Cross-Reactivity
There is substantial evidence suggesting a causal relationship between aeroallergens and
allergic diseases, but it remains unclear which aeroallergens are more highly associated with
causing sensitization and subsequent disease development. The inability to develop a hierarchy
of specific aeroallergens and their role in initiating an allergic response is primarily due to the
cross-reactivity of aeroallergens—the ability of two or more aeroallergens, due to biochemical
similarities, to elicit an allergic response in an individual who may be sensitized to only one of
them. Multiple studies have found cross-reactivity among the aeroallergens implicated in
causing allergic diseases (Table 2-11). Some aeroallergens not identified as being clinically
relevant have shown cross-reactivity with those that are, which further complicates the ability to
identify allergens associated with causing allergic diseases. For example, short ragweed is
identified as a major cause of allergic rhinitis, but false and western ragweed all cross-react with
short ragweed, which could result in an allergic response in a sensitized individual exposed to
any of the ragweeds (White and Bernstein, 2003). Cross-reactivity is not specific to pollen; it
has also been observed in mold and among asthma-related indoor allergens as well (Andersson et
al., 2003; Halonen et al., 1997).
2-23
-------
Table 2-11. Cross-reactivity of major clinically relevant aeroallergens
Latin name
Common
name
Cross reactive
aeroallergen(s)
Reference
Tree pollen
Acer negundo
Acer rubra
Alnus rubra
Betula papyrifera
Carya illinoensis
Fraxinus americana
Juglans nigra
Olea europaea
Quercus alba
Quercus rubra
Ulmus americana
Ulmus parvifolia
Ulmus pumila
Box-elder
Red maple
Alder
Paper birch
Pecan
White ash
Black walnut
Olive
White oak
Red oak
American elm
Chinese elm
Siberian elm
Red maple
Box-elder
Paper birch, white oak, red
oak
Alder, white oak, red oak
Black walnut
Olive
Pecan
White ash
Paper birch, alder, red oak
Paper birch, alder, white oak
Chinese elm, Siberian elm
American elm, Siberian elm
American elm, Chinese elm
Phipatanakul, 2005
Phipatanakul, 2005
White and Bernstein, 2003
White and Bernstein, 2003
White and Bernstein, 2003
White and Bernstein, 2003
White and Bernstein, 2003
White and Bernstein, 2003
White and Bernstein, 2003
White and Bernstein, 2003
Phipatanakul, 2005
Phipatanakul, 2005
Phipatanakul, 2005
Grass pollen
Cynodon dactylon
Festuca elatior
Holcus halepensis
Lolium perenne
Paspalum notatum
Phleum pretense
Bermuda
Meadow fescue
Johnson
Rye
Bahia
Timothy
Johnson
Bahia
Bermuda
Bahia
Timothy, meadow fescue, rye
Bahia
White and Bernstein, 2003
White and Bernstein, 2003
White and Bernstein, 2003
White and Bernstein, 2003
White and Bernstein, 2003
White and Bernstein, 2003
Weed pollen
Amaranthus
retroflexus
Ambrosia
artemisiifolia
Artemisia vulgaris
Salsola kali
Red root
pigweed
Short ragweed
Mugwort
Russian thistle
Russian thistle
Mugwort
Ragweed
Red root pigweed
Phadia, 2002
White and Bernstein, 2003
White and Bernstein, 2003
Phadia, 2002
Mold
Alternaria alternata
Epicoccum nigrum
N/A
N/A
Epicoccum nigrum
Alternaria alternata
Levetin, 2006
Levetin, 2006
Indoor allergens
Dermatophagoides
farinae;
Dermatophagoides
pteronyssinus
Arthropods
(domestic
mites)
Cross reactive with one
another
Phipatanakul, 2005
Note: This table includes only those aeroallergens that have been implicated as being cross-reactive with another
aeroallergen.
2-24
-------
3. HISTORICAL TRENDS IN AEROALLERGENS AND ALLERGIC DISEASES
IN THE UNITED STATES
The amount and distribution of aeroallergens, as well as the prevalence of allergic
diseases in the United States, is likely to change over time. This section examines past trends
and current levels of both aeroallergens and allergic diseases.7
3.1. AEROALLERGENS
During approximately the last 30 years, numerous studies have examined historical trends
in aeroallergen production and distribution, most notably for pollen and some types of mold.
Most studies observing pollen levels over time have found year-to-year fluctuations but no major
trends. Observing pollen trends in Philadelphia and Southern New Jersey, Dvorin et al. (2001)
found that although all pollen levels fluctuate yearly, tree pollen demonstrates a larger
fluctuation than either grass or weed pollen. In a 21-year study of airborne pollen levels in
Switzerland, Clot (2003) observed no major change in the yearly pollen abundance for the
majority of pollen species studied.8
In some studies, the overall abundance of pollen in an area did change dramatically over
time, but this was due to specific nonclimatic factors. For example, Emberlin et al. (1999)
reports a study observing pollen trends conducted at three sites within the UK—London, Cardiff,
and Derby—which found pollen levels decreased in Derby and London while they significantly
increased in Cardiff. The substantial changes in pollen levels at each site were attributed to
changes in land use that occurred during the study period (Burr, 1999). Similarly, Sneller et al.
(1993) found a dramatic increase in pollen levels over 5 decades in Tucson, Arizona as a result
of the importation of certain tree species to the city due to changing architectural and landscape
preferences. With the exception of cases of anthropogenic changes, which altered the abundance
of aeroallergen levels observed in Emberlin (1994) and Sneller et al. (1993), however, pollen
levels have tended to remain fairly consistent on a year-to-year basis.
7This section of the report does not address other factors that have been implicated in affecting the overall trends of
aeroallergens, specifically climate change.
8Clot (2003) found ihatAlnus, Taxus/Cupressaceae, and. Artemisia pollen were significantly higher at the end of the
20-year study then at the beginning, but it is unclear why the pollen levels for these four species increased over time.
3-1
-------
Studies have shown that the duration of the pollen season has remained fairly stable over
time. In their 6-year study of pollen levels in Philadelphia and Southern New Jersey, Dvorin et
al. (2001) observed that the pollen season did not change significantly. They found late April to
early May and early September consistently represented the significant spring and fall periods,
respectively, of airborne pollen prevalence (Dvorin et al., 2001). These findings agree with those
of Kosisky and Carpenter (1997), who found in a study observing tree pollen over a 5-year
period in Washington, D.C. that April remained the month with the highest weekly average
concentrations over the study period (Kosisky and Carpenter, 1997). We conclude that the time
frame for these studies (5 and 6 years, respectively) may not be long enough to examine trends in
pollen season.
Although these studies suggest that the duration of the pollen season has been relatively
stable, other studies suggest a trend towards an earlier initiation of the pollen season. Emberlin
et al. (2002) and Clot (2003) both observed a shift in the timing of the pollen season during
long-term pollen observation studies. Clot (2003) observed strong trends towards an earlier
pollen season for tree pollen and a less remarkable shift for grass and weed pollen. Emberlin et
al. (2002) observed a trend towards an earlier start date for the Betula (Birch) pollen season by
about 6 days, but ranging up to 30 days. Preliminary data suggest that a change in the initiation
of the pollen season may not influence its overall duration (Clot, 2003).9 It should be noted that
warming has occurred mainly over the past 30 years and that interannual variations make it hard
to detect trends over time. We would need a 2-3 decade record of pollen counts measured in a
consistent way to have a chance of observing a trend.
While pollen levels have remained fairly consistent over time, this is not the case for
mold. The evidence suggests a possible increase in the concentration of some types of mold.
Epicoccum nigrum has recently been sprayed onto sunflowers to control sunflower head rot
(Burge, 2002). A continued increase in the use of Epicoccum nigrum and other types of mold as
biocontrol agents might increase the proportion of outdoor mold that is associated with allergic
diseases in the environment (Burge, 2002). Although the increased use of mold commercially
could result in an increase in mold in the environment, an increase in the abundance of
Alternaria has already been observed. In a study conducted in Derby, UK from 1991-1998,
9Emberlin et al. (2002) did not study the duration of the pollen season.
3-2
-------
Corden and Millington (2001) found a dramatic increase in the number of days with Alternaria
spore counts above 50 spores/m3.
3.2. HISTORICAL TRENDS OF ALLERGIC DISEASES
If aeroallergen levels remain fairly stable, the prevalence of allergic illness would be
expected to remain stable as well. The evidence suggests, however, that this is not the case. The
prevalence of allergic diseases in the United States has increased over the last 30 years
(Figure 3-1 and Figure 3-3). This upward trend in the United States appears to be mirrored in
other countries, too. In the Copenhagen Allergy Study, Linneberg et al. (2000) found the
prevalence of specific IgE antibodies to at least one allergen in the cohort increased significantly
from 1990 to 1998, which coincided with an increase in the prevalence of allergic rhinitis.
Linneberg et al. (2000) also cite Nakagomi et al. (1994), who found an increase in IgE positivity
from 21.4% in 1978 to 39.4% in 1991 to one or more of 16 allergens in schoolgirls in Japan.
These findings and data collected via surveys by the CDC suggest the prevalence of allergic
diseases has increased over time, but that finding is suspect, given the lack of sufficiently long
records using consistent measurement methodologies. We conclude that allergic illness trends
described here are consistent with the overall body of scientific literature but not of adequate
duration to support certain trends.
The perceived increase in allergic diseases over time has not been adequately explained.
It might be expected that an increase in the prevalence of allergic diseases would imply a
corresponding increase in the levels of their associated aeroallergens, but, as noted above, we do
not yet have sufficient data and research to determine whether this has occurred. Therefore, it is
possible that there are other factors to explain the increase in the prevalence of allergic diseases.
The rate at which the prevalence of respiratory allergies has been increasing argues against the
trend being solely attributed to genetic factors. One theory, known as the "hygiene hypothesis,"
suggests that larger family size, exposure to respiratory infections, microbial exposure, and
exposure to other bacterial components such as endotoxin have a protective effect against the
development of hay fever and other allergic diseases (Phipatanakul, 2005). The smaller family
sizes now observed in Western countries have reduced children's exposure to cross infections,
which may prevent the development of hay fever (Von Hertzen, 1998). Evidence supporting this
hypothesis comes from studies that have found a negative correlation between the prevalence of
3-3
-------
allergic rhinitis and the number of older siblings, implying that increased family size reduces a
child's risk of developing allergic rhinitis (Phipatanakul, 2005). A similar protective effect has
also been observed for children who have early exposure to day care after 1 year of age
(Phipatanakul, 2005). Unfortunately, recent studies have been unable to identify single or
multiple determinants in lifestyle or home environment that could significantly affect disease
development (Linneberg et al., 2000).
3.2.1. Asthma
Over the last 30 years, there has been a significant increase in the prevalence of asthma
(Figure 3-1). It is not yet clear what is driving the observed increase in asthma prevalence,
because many factors may influence its development. As noted above, aeroallergens have a
significant impact on asthma development. It is unclear what percentage of asthma cases each
year can be attributed to exposure to aeroallergens, and aeroallergens have also not shown a
corresponding increase that could potentially account for some of the increase in asthma
prevalence (Nielsen et al., 2002).
§
ns
Pe
Prevalence pe
0 0 8
Prevalence of Asthma in the United States (1970-1 996)
O
1*.
tM
00
a>
h-
a>
00
00
to
00
a>
O>
00
tM
O>
tO
O>
Year(s)
Figure 3-1. Prevalence of asthma in the United States from 1970 through
1996.
Sources: National Center for Health Statistics (NCHS, 1973, 1985, 1989, 1992, 1993,
1994, 1999).
3-4
-------
The increase in the prevalence of asthma has been particularly acute among individuals
of low socioeconomic status (Phipatanakul, 2005; Hamilton and Eggleston, 1997). This has been
believed to be primarily the result of higher levels of exposure to indoor allergens, especially
cockroaches, in this population (Phipatanakul, 2005; Hamilton and Eggleston, 1997). This trend
may extend beyond the inner city; recently, cockroach allergen has been found to be more
common in suburban middle class homes with asthmatic children than previously thought
(Hamilton and Eggleston, 1997; Hamilton, 2005).
Figure 3-2 shows that asthma prevalence has increased in all regions of the United States,
with the most significant increase occurring in the Northeast. Some recent studies have shown a
possible stabilizing of asthma prevalence, but it is unclear if this is a true effect or a result of
multiple definitions being used to identify asthma. These studies have not shown consistent
results across geographic regions or demographic characteristics, but instead have shown a
heterogeneity of patterns of asthma diagnosis, symptoms. Allergic sensitization of the
assessment of the trends observed in asthma prevalence is thus difficult because of the
heterogeneity of the disease and the fact that there is no recognized standard used to make a
diagnosis (Lawson and Senthilselvan, 2005).
Prevalence of Asthma in the United States by
Geographic Region (1970-1996)
Q1970
• 1982
D1989
D1992
• 1996
Northeast
Midwest South
Region
West
Figure 3-2. Prevalence of asthma in the United States by geographic
region from 1970 through 1996.
Sources: NCHS (1973, 1982, 1985, 1989, 1990, 1992, 1994, 1999).
3-5
-------
3.2.2. Allergic Rhinitis
Consistent with the findings for asthma, there has also been an increase in the prevalence
of allergic rhinitis (hay fever) in industrialized countries over the last 30 years (Figure 3-3).
Similarly, because this increase is not accompanied by a corresponding increase in pollen
abundance, its origins remain unclear (Clot, 2003).
Prevalence of Allergic Rhinitis (Hay Fever) in the United
States (1970-1 996)
«) 190 T
c i^u
o
2 100
a)
Q.
O QQ _
o_
*~ 60
0)
Q.
O
c
0) 20
5
2 o
Q. UH
^-^*^—
^ «^^ ^^»
/^~
*^
O T- CM 00 O> CM tO
1^. CO CO CO CO O) O)
CD O) 0) O) 0) 0) 0)
en 0>
Year
Figure 3-3. Prevalence of allergic rhinitis (hay fever) in the United
States from 1970 through 1996.
Sources: NCHS (1973, 1982, 1985, 1989, 1990, 1992, 1994, 1999).
Numerous studies have shown that, unlike asthma, allergic rhinitis has a higher
prevalence in individuals of higher socioeconomic status (Phipatanakul, 2005). It is unclear why
this might be the case, but according to the hygiene hypothesis, a decrease in exposure to certain
infections may account for this observation.
As shown in Figure 3-4, there have been significant increases in the prevalence of
allergic rhinitis in all regions of the United States, with the greatest number of cases consistently
occurring in the West. The sampling protocols used to obtain prevalence rates, however, may
not accurately reflect the true prevalence. Studies have observed the prevalence through two
3-6
-------
avenues: questionnaires/interviews and physician diagnosis, both of which tend to underestimate
the actual prevalence of the disease (Phipatanakul, 2005).
Prevalence of Allergic Rhinitis (Hay Fever) in the
United States by Geographic Region (1970-1996)
• 1970
• 1982
D1989
D1992
• 1996
Northeast
Midwest
South
West
Region
Figure 3-4. Prevalence of allergic rhinitis (hay fever) in the United
States by geographic region from 1970 through 1996.
Sources: NCHS (1973, 1982, 1985, 1989, 1990, 1992, 1994, 1999).
3.2.3. Atopic Dermatitis
There is only limited information on the historical trends of atopic dermatitis.
Approximately 15 to 20% of the worldwide childhood population is currently afflicted with the
illness, but considerable evidence suggests the prevalence of atopic dermatitis may be increasing
above the 15 to 20% now observed (O'Connell, 2004). Because of the controversy surrounding
the diagnosis of atopic dermatitis, as discussed by Whitmore et al. (1996), it will be possible to
accurately reflect the prevalence of atopic dermatitis over time only when studies are conducted
with adequate controls.
3-7
-------
3-8
-------
4. IMPACTS OF CLIMATE CHANGE ON AEROALLERGENS
Climate change, caused in part by increased atmospheric CC>2 and other greenhouse gas
concentrations, may result in increases in temperature, precipitation, humidity, and extreme
weather events. These factors, including CC>2 concentration, can impact the production,
distribution, dispersion, and allergen content of aeroallergens and the growth and distribution of
organisms that produce them (i.e., weeds, grasses, trees, and fungus). Shifts in aeroallergen
production and, subsequently, human exposures may result in changes in the prevalence and
severity of symptoms in individuals with allergic diseases. This section reviews the potential
and observed impacts of climate change on aeroallergen production, distribution, dispersion, and
allergen content, and discusses how climate-related changes in aeroallergen production may lead
to indirect impacts on allergic diseases.
4.1. PRODUCTION OF AEROALLERGENS
It has generally been observed that the presence of elevated CO2 concentrations and
higher temperatures stimulate plants to increase photosynthesis, biomass, water use efficiency,
and reproductive effort (The Center for Health and the Global Environment, 2005; Jablonski et
al., 2002). However, these relationships are complex and likely differ among taxa and species.
Short-term responses to climate change (i.e., over 10 to 20 years) might involve changes in plant
phenology and biochemistry. This is consistent with a recent meta-analysis that indicates the
current rate of phenologic advance is 5 days per decade for numerous species of plants (Root et
al., 2003). A key finding of the National Assessment Synthesis Team (NAST; Melillo et al.,
2001) was that over the next few decades, climate change is very likely to lead to increased plant
productivity and carbon storage for many parts of the country, especially those areas that become
warmer and wetter (Melillo et al., 2001; Joyce et al., 2001). We might infer, then, that pollen
production in these areas, on average, would be expected to increase. The NAST also found that
areas where soils dry out during the growing season, such as the Southeast under certain
scenarios, are likely to see reduced productivity and carbon storage, and hence, less pollen
production. The following subsections review specific studies on how climate change may alter
plant and fungal reproductive responses in the United States.
4-1
-------
4.1.1. Pollen
4.1.1.1. Tree Pollen
Research on the potential effects of climate change on tree pollen production in the
United States is limited. Researchers have tried to identify important climatic variables for
seasonal forecasting of tree pollen seasons, but often these models do not account for climate
change and are specific to species and geographic locales outside of the United States. However,
to the extent that this research successfully identifies strong projections of pollen season severity,
one can infer that changes in those projectors may directly impact pollen production. Overall,
research shows preseason temperature and precipitation to be the most consistent projectors of
tree pollen seasons. The relevant details of studies on tree pollen production are presented below
using the framework of start date and pollen season severity.
Levetin (2001) reported that cumulative season total pollen for Jimiperus (cedar),
Quercus (oak), Carya (hickory and pecan), and Betula (birch) increased significantly during a
14-year period beginning in 1987 in Oklahoma. Meteorological data also showed a significant
increase in average winter temperatures. Correlations between winter temperatures and pollen
counts were not significant. The author noted that the cause of the increasing pollen count trend
was unknown and could include climate change, urbanization, or evolving landscaping patterns.
Lo and Levetin (2007) found increased cumulative season total pollen counts for Cupressaceae
pollen increased from 1986-2007 in Tulsa, OK, although there was no significant change in
Ulmus pollen counts, nor was the observed seasonal start date significant. Pollen counts for both
taxa were significantly positively correlated with daily minimum, mean, and maximum
temperatures and significantly negatively correlated with precipitation.
United States researchers examined pollen counts in New England before and after the
occurrence of an El Nino event that started in mid-1997 and continued until the summer of 1998
(Freye and Litwin, 2001). While El Nino is a cyclical climatic event that is not necessarily
associated with climate change, it can serve as an example of the impact of short-term variability
on pollen production. This El Nino was similar to projected climate change in that precipitation
was 2 to 8 inches greater than normal during the winter and spring of 1998, and temperature was
4 to 6°F higher than normal during winter of 1998. The authors observed that, relative to 1997
and 1999, maximum pollen counts were higher and occurred about 2 to 4 weeks earlier for most
tree types during 1998, but a statistical analysis of the difference is not provided. Similarly,
4-2
-------
Reiss and Kostic (1976) found strong correlations (r2 range 0.85 to 0.94) between pollen season
severity and spring and summer minimum temperatures and mid-spring precipitation amounts in
New Jersey, but they did not specify pollen types. Oak pollen counts in the San Francisco Bay
Area were strongly correlated with total rainfall during the previous year (Weber, 2003a, b).
In Cordoba, Spain, researchers studied the influence of meteorological parameters on
O. europaea L pollen and found that cumulative variables for temperature and sunlight hours
were the most common significant (Students t < 0.05) projectors of pollen concentration in
regression analysis (Vazquez et al., 2003). In Poland, researchers found positive significant
(p < 0.05) correlations between air temperature and birch pollen concentration but negative
nonsignificant correlations with poplar pollen, indicating the need for species-specific analysis
(Puc and Wolski, 2002). A study of birch pollen in two sites in Denmark found that the pollen
season started earlier, peaked earlier, and ended earlier, and had increased seasonal total pollen
counts, higher peak count levels, and more days with pollen counts greater than zero. The study
attributed the earlier start of the pollen season to higher winter and spring temperatures
(Rasmussen, 2002).
International research to identify trends in pollen season start dates, using databases,
including species relevant to the United States, such as birch (Betuld) and olive (Oka
europaea L.), is also informative. Even in this case, however, there is limited assessment of
changes in pollen production. Long-term pollen monitoring data are available for several
locations in Europe, and researchers have analyzed these data for changes in pollen season start
dates. Overall, while several analyses show earlier start dates, there is an indication that the
observed effect may be specific to species and geography. Clot (2003) analyzed time series of
21 years of data in Switzerland. Using the Seasonal Pollen Index (SPI), Clot (2003) found that
there was no major change in the abundance of pollen among most of the 25 taxa studied. There
were a few exceptions to this. Linear trend analysis showed increases (p < 0.05) of pollen
quantities were observed forAlnus (alder), Ambrosia (ragweed), Artemisia (mugwort), and
Taxus/Cupressacea (yew/cypress), and decreases were seen in Ulmus (elm). Clot (2003) also
observed that the duration of the pollen season did not appear to change but that 71% of the start
or end dates of the pollen season occurred significantly earlier in the year. The average observed
advance was 0.84 days/year, and a stronger reaction was found in trees to climate change than in
weeds and grasses (Clot, 2003).
4-3
-------
Frenguelli (2002) reviewed 20 years of data, from 1982 to 2001, on airborne pollen and
mean air temperature in Perugia (central Italy) and reported an increase in annual mean
temperature of 0.7°C, with the months of February, May, June, and August experiencing the
greatest increases. Results show the pollen seasons of most taxa starting earlier, and for several
taxa, the season duration is shorter as well. An exception is Urticaceae (nettle), which
experienced an increased seasonal duration.
Emberlin et al. (2002) investigated relationships between changes in start dates of birch
pollen seasons and changes in spring temperatures, using daily birch pollen counts from six
metropolitan cities in Europe from 1982-1999. London, Brussels, Zurich, and Vienna showed
trends towards earlier start dates, and a regression analysis indicated the mean start dates at these
sites would advance by about 6 days over the next 10 years (Emberlin et al., 2002). In Kevo,
Finland the opposite effect was observed, with cooler springs and therefore later starts of 6 days
on average per decade. While the data are suggestive of changes in the timing of pollen season
starts, there was no assessment of whether or not this would lead to greater pollen production or
allergen content.
In Andalusia, Spain, Galan et al. (2005) compared the start of O. europaea L. pollen
season and heat accumulation over a selected temperature threshold while investigating the
influence of topography on the results. The authors used pollen and meteorological data from
1982-2001 for five sites in central and eastern Andalusia. The study found that all of the study
sites had increasingly earlier start dates during the study period. The authors used the Hadley
Climate Model (a regional model developed by the Hadley Meteorological Centre, UK) to
estimate the impact of projected climate change on the olive tree's flowering phenology. Their
results indicated an advance of 1-3 weeks by the end of the century. Garcia-Mozo et al. (2006)
found that Quercus species trees showed earlier start dates in fourteen different locations
throughout Spain from 1992-2004, most likely due to the increased temperatures seen at those
sites over the same time period. Using projected meteorological changes from Regional Climate
Models and an externally validated growing degree days forecasting model, they projected that
the Quercus pollen season could start on average one month earlier in Spain and pollen counts
could increase by up to 50%, with the largest increases in inland Mediterranean locations. As
with previous research, however, it is not clear how a shift in pollen season start may affect
production. Researchers have also found advances in start dates in Japanese cedar (Cryptomeria
4-4
-------
japonica), as well as an increased pollen count and an extended pollen season (Teranishi et al.,
2000).
Glassheim et al. (1995) examined the short-term (i.e., 1-day lag) relationship between
observed tree pollen counts from elm, juniper, maple, cottonwood, and pine in Denver, Colorado
and a selection of independent meteorological variables. With the exception of pine, none of the
tree pollen studied for the 5-year period appeared affected by temperature. However, this
short-term type of analysis may have limited relevance to seasonal pollen production and climate
change. The authors did observe modest negative correlations with both precipitation and
relative humidity, which is likely due to "scrubbing" or particle adsorption (to raindrops), which
removes pollen from the air.
In summary, preseason temperature and precipitation are important projectors of tree
pollen production. To the extent that climate change results in changes in these two
meteorological variables, then, we would expect corresponding changes in tree pollen
production, all else equal, although the evidence to date suggests that the nature of the changes
may be region and species-specific. One United States study observed a trend of increasing
pollen production in Oklahoma (Levetin, 2001). Changes in phenology (start date) appear to be
a relatively consistent finding, especially for European species. However, in most studies, the
change in start date did not correspond to a lengthening of the pollen season. Additionally, it is
unclear whether the phenologic changes have an effect on total pollen production or allergen
content. The literature does not provide clear evidence of changes in phenology in United States
species; however, this may be due to limitations in data.
4.1.1.2. Grass Pollen
As with tree pollen, research on the potential effects of climate change on grass pollen
production in the United States is limited in scope at present. Overall, forecast models show
temperature and precipitation to be the most consistent projectors of grass pollen seasons, but
these models do not directly take into account climate change and are again specific to
geographic locales, most of which are not in the United States.
As described above, United States researchers Freye et al. (2001) examined pollen counts
in New England before, during, and after the occurrence of an El-Nino event. With the
4-5
-------
exception of an earlier peak in 1998, overall grass pollen concentrations did not appear to be
affected. The details of selected research are reviewed below.
Contrary to general projections of increased production, one of the longest data series for
grass pollen suggests earlier starts but declining annual counts and severity in England. These
changes are most likely due to land use trends, such as declining agriculture and pasture areas
(Emberlin, 1994). In England, Emberlin et al. (1999) used data from 1961-1993 at Cardiff,
Derby, and London to project total seasonal catches, the severity of seasons in terms of the
number of days with high counts, and the start dates of seasons. The authors found that at two of
the sites (Derby and London) the annual counts and severity declined but at different rates, while
at the third site (Cardiff), annual counts and severity increased in the 1960s, declined in the
1970s, and rose again in the 1980s. There was a trend towards earlier start dates at the Derby
site, a less pronounced trend at Cardiff, and a trend towards later starts in London. In models,
the most important climatic variables influencing the broad features of grass pollen seasons (e.g.,
seasonal cumulative pollen counts and peaks) were cumulative rain and temperature, but the
importance of these variables differed by site and was overshadowed by the influence of land
use. The authors conclude that the contrasting patterns both in pollen records and land use
changes among the three sites underscore the need for regional data. In a prior analysis, they
reached similar conclusions, suggesting that changes in pollen production will vary by region
such that many central areas north of the Alps could have longer grass pollen seasons, while
grass pollen concentrations are likely to decrease in the southern Mediterranean area during
summer months (Emberlin, 1994). The regional differences reflect the interaction of climate
change at different latitudes and topography, i.e., reduced snow cover in the Alps and increasing
drought in the Mediterranean; both of which must be evaluated against land use trends as well.
In contrast, Clot (2003) reviewed 21 years of grass pollen data from a single trap in Switzerland
and found an earlier start date (-14 days) but no significant change in the duration or intensity of
the pollen season.
Puc and Puc (2004) analyzed grass pollen seasons in western Poland from 2000 to 2003
to evaluate relationships between meteorological parameters and Poaceae (grass) pollen counts.
The authors found that air temperature and relative humidity were most consistently correlated
with pollen counts.
4-6
-------
In Australia, grass pollen accounts for 71% of the total atmospheric pollen counts (Green
et al., 2004). Green et al. (2004) evaluated grass pollen counts and associations with
meteorological parameters in Brisbane, Australia from 1994 to 1999. The authors found that
daily grass pollen counts were positively associated (p < 0.0001) with maximum and minimum
temperature each sampling year. Precipitation was observed to "scrub" or remove pollen grains
from the atmosphere during significant periods of rainfall.
Glassheim et al. (1995) examined the short-term (i.e., 1-day lag) relationship between
meteorological variables and grass pollen in Denver, Colorado. The authors found that grass
pollen counts during the period 1987-1991 were correlated with high temperature (r = 0.305,
p < 0.001) and less so with percent daily sunshine (r = Q.149,p < 0.006) and were negatively
associated with precipitation (r = -0.227, p< 0.001) and relative humidity (r = -0.430,
p < 0.006). Glassheim et al. (1995) also found that correlations were not consistent from year to
year, suggesting the intraseasonal meteorological conditions that determine pollen counts may
vary from year to year or that preseason conditions are more important.
Research in Spain also indicates that preseason meteorological variables are more
important and consistent determinants of seasonal pollen load than are day-to-day weather
conditions (Gonzalez Minero et al., 1998). Declines in grass pollen were observed for the period
1987-1996. This was attributed to several years of drought, a potentially important but less
projectable feature of climate change. Preseasonal rainfall, temperature, and average monthly
humidity in Spain were strong projectors of total grass pollen count (Burr, 1999;
Gonzalez Minero et al., 1998).
In summary, temperature and precipitation are important factors in grass pollen
production, but more so in terms of preseason conditions than day-to-day meteorological
conditions during the pollen season. The correlation with precipitation is not straightforward as
preseason precipitation may increase pollen counts but in-season precipitation tends to "scrub" or
remove pollen from the air. To the extent that climate change results in changes in these two
meteorological variables, we would expect some changes in grass pollen production. In Europe,
earlier start dates have been observed, as well as declines in production, but this is attributed to
changes in land use. The literature does not provide clear evidence of changes in start dates or
production in United States species.
4-7
-------
4.1.1.3. WeedPollen
Common ragweed (Ambrosia artemisiifolia L.) is recognized as a significant cause of
allergic rhinitis in the United States, and there is relatively more research on the response of this
weed to climatic variables, especially in the context of climate change. Specifically, several
researchers have used controlled environments to examine ragweed response to carbon dioxide
levels and temperature, the two covariates for which models reliably project increased levels in
the future. The following section details the studies that observed the association between
climatic variables and ragweed production.
Ziska and Caulfield (2000) tested whether the increase in atmospheric CC>2
concentrations since the Industrial Revolution and projected future increases may alter growth
and pollen production of common ragweed. Experiments were conducted using a controlled
environmental chamber to measure the growth and pollen production of common ragweed from
preindustrial levels of CC>2 (280 umol/mol) to current concentrations (370 umol/mol) to a
projected 21st century concentration of 600 umol/mol. The experiments showed that pollen
production increased approximately 132% from preindustrial levels to current levels of carbon
dioxide and approximately 90% for current to projected future levels of carbon dioxide. The
observed increase of pollen production from the preindustrial CC>2 concentration was due to an
increase in the pollen per floral spike (at 370 umol/mol) and number of floral spikes (at
600 umol/mol).
Wayne et al. (2002) found similar results using environmentally controlled greenhouses
to grow stands of ragweed plants from seed through flowering stages at CC>2 concentrations of
350 vs. 700 uL/L. The authors found that stand level pollen production was 61% higher in
elevated versus ambient CC>2 environments (F= I5.l6,p = 0.005). The authors comment that
previous studies with ragweed have shown that adding essential resources to stands (e.g.,
nitrogen) results in plants investing in proportionally more male pollen-generating reproductive
structures versus female pollen-accepting reproductive structures, consistent with the
observations of Ziska and Caulfield (2000).
Ziska et al. (2003) followed up on the chamber studies conducted by Ziska and Caulfield
(2000) and Wayne et al. (2002) with field studies. The authors used existing temperature/CO2
concentration gradients between urban and rural areas in Maryland to examine the quantitative
and qualitative aspects of ragweed growth and pollen production. In addition, pollen was
4-8
-------
subjected to immunochemical analysis to quantify content of the allergen protein AMB A 1.
Average daily (24-hour) values of CC>2 were 30 and 31% higher in 2000 and 2001, respectively,
within an urban environment vs. at a rural site; air temperature was 1.8 and 2.0°C higher in 2000
and 2001, respectively, within the urban environment. Overall, the results demonstrated small
but measurable phenologic differences as a function of both temperature and CC>2 concentration.
Ragweed grew faster, flowered earlier, and produced significantly greater aboveground biomass
and ragweed pollen at urban locations, which have a higher CC>2 concentration and temperature
than at rural locations. However, a significantly (p < 0.01) higher quantity of antigenic protein
was extracted from pollen at the rural site relative to other sites.
Ziska et al. (2007) documents the continued evaluation of the urban to rural transect in
Maryland that they reported on previously in Ziska et al. (2003). The authors continued to
monitor the three plot locations through 2005 and found that the rural and suburban plots
continued to increase in ragweed biomass, but, after 2003, the urban plot declined significantly
despite the continued higher temperature and CC>2 concentrations at the urban location. They
document that with the increased biomass, which was observed in the urban plots associated with
higher temperatures and CC>2, there is also higher litter accumulation, which can hinder the
growth of new seedlings, and the suburban plot started to show biomass decline in 2005, which
could signal a similar mechanism after high biomass. Another possible reason for the decline
could be a decrease in soil disturbance, which ragweed requires for growth.
Rogers et al. (2006) designed a study to examine the potential impact of earlier arrival of
spring and the interaction with CC>2 concentrations on ragweed pollen productivity. The authors
used climate controlled greenhouses to test (1) whether variability in the onset of spring alters
the rate and magnitude of ragweed development, flowering phenology, and seasonal pollen
production; and (2) whether atmospheric CC>2 concentrations directly alter ragweed development
and productivity, and influence plant responses to climatic variability. Cohorts of ragweed seeds
were released from dormancy at three 15-day intervals and grown at ambient concentration or
700 ppm CC>2 concentration. Carbon dioxide treatment did not significantly affect days to
anthesis or anthesis date. The authors found the timing of spring onset was the primary factor in
a model fit for indicators of plant growth and, thus, pollen production. At ambient CC>2
concentration, the earlier cohort had 54.8% greater pollen production than the latest cohort.
However, in the early cohort, pollen production was similar under ambient and high CC>2
4-9
-------
concentrations, but in the middle and late cohorts, high CC>2 concentration increased pollen
production by 32 and 55%, respectively compared to ambient CC>2 levels (see Figure 4-1). Thus,
at elevated CO2 concentrations, pollen productivity appears less sensitive to variability in season
onset. The authors project that in future climates with elevated CC>2 concentrations, pollen
production will be just as robust in years with late springs as those with early springs.
5 1-°
If 08
=-ss.
f £ O.G
0.4
0.2
o £
0.0
Early Middle
Dormancy release
Late
Figure 4-1. Pollen production in A artemisiifolia
for three springtime dormancy release cohorts
grown at two CO2 concentrations (380 ppm and
700 ppm).
Notes: Error bars indicate 95% confidence intervals.
Source: Rogers et al. (2006, Figure 4).
Wan et al. (2002) experimented with changes in temperature and clipping ragweed plants
to determine if either or both of these factors made a difference in ragweed plant growth and
pollen production. The warmed plots had 1.2°C air temperature, and the soil temperature was
1.8°C higher in the heated plots without clipping and 2.7°C in the heated plots with clipping, as
clipping showed to increase the soil temperature possibly by exposing more of it to the heat
source. The ragweed plants in the warmed plots showed many significant changes, including
increases in the number of stems, total biomass, percent coverage, pollen diameter, and total
pollen production. The authors hypothesized many possible reasons for these changes, including
that ragweed is well adapted to disturbed warm areas and is taller than other plants; that warming
4-10
-------
extends the growing season and therefore enhances survivorship; and that experimental warming
may have increased nitrogen availability in the soil, which promotes the growth of ragweed.
Whatever the reason, this experiment implied that warmer temperatures may increase the growth
and pollen production of ragweed.
Glassheim et al. (1995) calculated correlation coefficients between observed pollen
counts in Denver, Colorado and a selection of independent meteorological variables. The
prevalent weeds analyzed were ragweed, sage, and the chenopod/amaranth group (pigweed).
High and low temperature were most strongly correlated with total weed pollen counts during
1987-1991 (r = 0.603, p< 0.001). This is consistent with the work by Rogers et al. (2006),
which indicates that an early start to the growing season, as indicated by minimum temperature,
results in larger, more productive plants. Similar observations of increased biomass have been
observed in CC>2 enrichment experiments with poison ivy (Mohan et al., 2006). While poison
ivy does not produce aeroallergens per se, the smoke generated from burning poison ivy can be
highly allergenic.10 Stefanic et al. (2005) found similar results in the Republic of Croatia,
reporting that mean and minimum annual air temperatures were significantly correlated with the
amount of ragweed pollen in the air during 2001-2003. Similar to Glassheim et al. (1995),
however, the authors found inconsistency in relationships from year to year.
Overall, studies of ragweed in controlled environments and in field studies show that
pollen production can be expected to increase with increased temperature and carbon dioxide
levels. The experimental results have demonstrated that doubling carbon dioxide levels from
current (350 umol/mol) to projected future levels (i.e., 700 umol/mol) would result in a 60 to
90% increase in ragweed pollen production (Ziska and Caulfield, 2000; Wayne et al., 2002).
Field studies of differences between rural and urban growth patterns also clearly show that
ragweed flowers earlier and produces greater amounts of pollen at urban locations compared to
rural locations (Ziska et al., 2003). Rogers et al. (2006) confirmed this effect by showing that the
timing of spring onset (i.e., early start) was the primary factor in a model fit for indicators of
ragweed growth and, thus, pollen production. The higher allergen concentration in pollen at the
rural site, however, highlights the need for caution in making inferences about public health
implications.
10Personal communication with J. Patz, June 9, 2006.
4-11
-------
4.1.2. Mold
Assessment of mold production in response to climate change is mainly derived from
observational analyses of long-term data sets. Other assessments of mold production are
essentially short-term forecasts of intraseasonal spore counts and are strongly dependent on
whether the mold is a wet- or dry-weather type. Details of these studies are presented below.
Katial et al. (1997) analyzed 8 years of spore count data for Cladosporium, Alternaria,
and Epicoccum in Denver, Colorado. The authors found a statistically significant year effect
(p < 0.01), indicating a positive linear trend in Cladosporium spore counts over time. No trends
were observed for Alternaria or Epicoccum. In addition, there were no trends in annual
temperature, precipitation, or humidity to account for the trend in Cladosporium spore counts.
The authors suggest urbanization of Denver as a potential explanation for the increase in
Cladosporium, but the mechanism for the increase (e.g., soil disturbance, changing land use, etc.)
is not clear. They found that for Cladosporium, average temperature (p < 0.02) and humidity
(p < 0.01) were positively associated with spore counts, while precipitation was negatively
associated with spore counts (p < 0.01). Neither Alternaria nor Epicoccum showed correlations
with meteorological parameters.
Corden and Millington (2001) examined Alternaria concentrations during 1970-1998 in
Derby, UK and found an upward trend, which increased markedly after 1992. Their analysis
also showed an earlier start date and a longer season over time. A further analysis by Cordon et
al. (2003) comparing Alternaria counts in Derby, UK and Cardiff, UK hypothesized that the
earlier start date, longer season, and higher counts in Derby compared to Cardiff could be due to
increased cereal production in fields near Derby and possibly to changes in climate as evidenced
by low Alternaria counts during a drought year in Derby compared to high counts in a much
wetter year. This is in contrast to the analysis of grass pollen by Emberlin et al. (1999), which
demonstrated earlier start dates but declining annual counts and severity, an effect that was
attributed to changes in land use patterns such as declining agriculture and pasturelands.
However, it is not clear if this explanation is also consistent with increasing trends in mold
counts observed by Cordon and Millington (2001), who note that bursts of Alternaria follow
grass mowing and harvest time. Hollins et al. (2004) found that summer temperature was the
strongest projector of the number of days that Cladosporium spore concentrations exceeded
4,000 spores/m3, while there was a negative relationship between precipitation and spore counts.
4-12
-------
In Tulsa, Oklahoma, Troutt and Levetin (2001) attempted to correlate fungal spore
concentrations with meteorological data during May 1998 and May 1999. These 2 months were
selected because they represented climatic extremes—May 1998 was exceptionally dry, and
May 1999 had unusually high precipitation. The spore types studied were Cladosporium,
Alternaria, Epicoccum, Curvularia, Pithomyces, Drechslera, smut spores, ascospores, and
basidiospores. Dry air spora (i.e., Cladosporium) were much more prevalent during May 1998
(the dry year). No single multiple regression model successfully projected all spore
concentrations, but temperature and dew point were important indicators.
Recent cyclic and extreme weather events have also been implicated in increased mold
production. Research in New England found maximum mold counts to be higher and 2 to
4 weeks earlier after the occurrence of an El Nino event (Freye and Litwin, 2001). An
examination of New Orleans housing stock after Hurricane Katrina revealed extensive mold
growth (Ratard et al., 2006). The CDC assessed the extent of mold growth in a sample (N=\ 12)
of households in the area. Almost half the homes had "visible mold growth," and 17% had
"heavy mold coverage," defined as ">50% coverage on [the] interior wall of most-affected
room." Indoor and outdoor air sampling indicated Aspergillus spp. and Penicillium spp. were the
predominant populations (Ratard et al., 2006). To the extent that extreme storm events such as
Katrina are more likely to occur or more severe in the future due to climate change, poststorm
mold problems may increase.
In summary, there is limited, but inconsistent evidence of increasing trends in mold
production. Short-term forecasts indicate that while temperature and humidity can be strong
predictors of mold concentrations, the effect varies by mold species and geography. At least one
United States study observed an upward trend in Cladosporium but not for co-occurring mold
such as Alternaria or Epicoccum. Another U. S. study observed increases in mold counts after an
El Nino event, while in the U.K., an analysis showed increasing trends in Alternaria. After
Hurricane Katrina, large portions of the housing stock were shown to have extensive mold
growth (Ratard et al., 2006). Overall, the relationship between climate factors and mold species,
extent, and geography suggests a complex multifactor mechanism. To the extent that extreme
storm events such as Katrina are more likely or more severe in the future, due to climate change,
poststorm mold problems may increase.
4-13
-------
4.2. DISTRIBUTION OF AERO ALLERGENS
Long-term responses to climate change (over 50 to 100 years) are likely to include
changes in species' ranges or distributions. In some cases, extinction may occur. Generally,
species are expected to migrate poleward and uphill as temperatures increase with climate
change. The Bernard et al. (2001) evaluated continental level shifts in forest and vegetation
distribution in the United States using various models and scenarios. Climate change scenarios
were based on two atmospheric general circulation models (GCMs)—the Hadley model and the
Canadian model. These models were selected because they represented the higher and lower
halves of the range of temperature sensitivity among the GCMs available when the analysis was
conducted. For both models, shifts in the distribution of vegetation types were projected with
significant variation across geographic regions (Melillo et al., 2001). In studies of the fossil
records in the United States, other researchers, Davis and Shaw (2001), project distribution shifts
and extinctions based on extensive range shifts. In Europe, Emberlin (1994) also used computer
models of future climatic changes resulting from increased CC>2 emissions and discussed the
potential impact on the distribution of major allergenic pollen types.
Joyce et al. (2001) conducted a continental-scale analysis, for forest vegetation, of
climate-induced changes in the distributions of biomes, community types, species richness, and
individual tree and shrub species. Species interactions and the physiological response of species
to carbon dioxide are not included in these models. The baseline scenario was the average
climate for the 1961-1990 period. Comparisons were made to the transient Canadian and
Hadley scenarios for the period 2070 to 2100. The results of these ecological models suggest
that the potential habitats (i.e., distribution) for many tree species in the United States are likely
to change, in some cases dramatically, by the end of the 21st century. Potential habitats for trees
favored by cool environments are likely to shift northward. The habitats of alpine, subalpine
spruce/fir, and aspen communities are likely to contract dramatically in the United States and
largely shift into Canada. Potential habitats are likely to increase in the United States for
oak/hickory, oak/pine, ponderosa pine, and arid woodland communities.
In a related review and analysis, Melillo et al. (2001) used biogeography model outputs to
simulate shifts in the geographic distributions of major plant species by 2090-2099. The authors
assume biogeochemical (i.e., production) changes will dominate ecological response to climate
change in the next few decades, while species shifts will dominate by the end of the 21st century.
4-14
-------
Unlike the models used for tree distribution, these models include CC>2 effects. Some of the
major regional changes projected by the biogeography models for both Hadley and Canadian
scenarios are as follows:
• Northeast: Forest will remain the dominant natural vegetation, but winter deciduous
forest may expand at the expense of mixed conifer-broadleaf forest (Hadley). There
could be a modest increase in savannas and woodlands (Canadian).
• Southeast: Forest remains the dominant natural vegetation, but the forest mix changes
(Hadley). Alternatively, there could be significant expansion of savannas and grasslands
at the expense of forest (Canadian).
• Midwest: Under both simulated climates (Hadley and Canadian), forest remains the
natural vegetation, but the mix of forest types changes.
• Great Plains: Two of three models project an increase in woodiness, while one (Hadley)
does not. The Canadian model suggests no change or a slight decrease in woodiness.
• West: Forest ecosystems grow at the expense of desert ecosystems (Hadley and
Canadian).
• Northwest: Forest area grows slightly (Hadley and Canadian).
How well plants and trees actually track changes in potential habitats will be influenced
by their dispersal abilities and disturbances in their environments. Davis and Shaw (2001) note
that changes in geographic distribution are so frequently documented in the fossil record that
range shifts are seen as the expected plant response to future climate change. These authors use
fossil records of trees and cite evidence of genetic adaptation to climate to argue that the
interplay of adaptation and migration has been central to the botanic response to climate change.
The authors conclude that unprecedented rates of climate change anticipated to occur in the
future, coupled with land use changes that impede gene flow, could result in extinctions of many
taxa (see Figure 4-2). The complexity of range shifts is evidenced in a study of the observed
increase in Juniperus occidentalis (Western Juniper) and coincident decrease in two herbaceous
species in central Oregon, which the researchers did not find to be influenced by variability in
climate, fire, grazing, or pathogens individually; however, they cannot exclude the cumulative
effects of a favorable climate, decline in fire frequency, or increases in CO2 concentrations as
contributing to the observed land-cover change (Knapp and Soule, 1998).
4-15
-------
B A
deb
phenotvpic gradient
~
climate gradient
B
1. Stable climate
e
E
d
D
c
C
b
B
climate gradient
2. Changing climate
3. Rapidly changing climate
X X
b
D
pollea'seed dispersal
a rrean phenotype
A fitness optimum
a
C
Figure 4-2. (A) Schematic depiction of phenotypic frequencies (mean
phenotype) for a population at a location along a climate gradient where
fitness maximum is C. (B) Schematic depiction of fitness optima (red) for a
species that ranges across a climate gradient.
Notes: Adaptive differentiation of population phenotypes is shown in black; arrows
indicate gene flow through pollen and seed dispersal. Spatial distributions
of the climate gradient, fitness optima, and phenotypic frequencies are
shown for three conditions: 1, stable climate; 2, slowly changing climate;
and 3, rapidly changing climate.
Source: Davis and Shaw (2001, Figure 5).
Other research has relied on computer model projections of future climatic changes
resulting from increased CC>2 emissions to gauge the potential impact on the distribution and
abundance of major allergic pollen types in Europe (Emberlin, 1994). The results suggest an
extension of the northern limit of birch by several hundred kilometers and a corresponding
increase in height of the altitudinal tree line and contraction of the range in the south. Emberlin
indicates that olive trees and ragweed could also experience a northward expansion.
In summary, long-term responses to climate change (over 50 to 100 years or more) are
likely to involve range or distribution shifts in species, with possible extinction in some cases.
Trees favoring cool environments, such as maple and birch, are likely to shift northward out of
the United States entirely, thus dramatically altering the pollen distribution associated with them.
4-16
-------
Under certain scenarios, the Southeast will experience significant warming trends, leading to an
expansion of savannas and grasslands at the expense of forest, again altering the presence of
major aeroallergens in large regions of the country.
4.3. DISPERSAL
There has been only limited research on how climate change could affect the dispersal of
pollen and mold. However, there are cases of both pollen and dust being dispersed long
distances from their release sites. For example, long distance dispersion ofJuniperus ashei
pollen has been routinely observed in Tulsa, Oklahoma (Van de Water et al., 2003). The nearest
upwind sources of J. ashei pollen are 200 to 600 km from their deposition site (Tulsa). Analysis
of the statistical relationship between wind direction and ragweed pollen counts in Pistoia and
Florence, Italy, where ragweed plants do not grow, indicates that these pollen grains likely are
the result of long distance transport from southern Hungary (Cecchi et al., 2006). Emberlin
(1994) suggests that in Europe, increased strength of westerly winds due to climate change will
enhance the long-range transport of birch pollen already observed to take place from north and
central Europe to Scandinavia. Transcontinental transport of dust particulates has also been
observed (Husar et al., 2001). Griffin et al. (2001) found Cladosporium cladosporioides in a
sample of African dust in the air over St. John in the Virgin Islands, and Griffin et al. (2006)
found various fungi species, some of which are associated with allergic diseases, in samples
taken of the mid-Atlantic atmosphere believed to come from African desert dust. Additionally,
many genera of fungus, including Cladosporium, Alternaria, Penicillium, and Aspergillus were
found in air samples taken over Barbados in 1996-1997 of long-range transport dust from Africa
(Prospero et al., 2005). During April 1998, two large dust storms occurred over the Gobi desert
(Mongolia and north central China). The dust plume crossed the Pacific Ocean and resulted in
strong spikes in particulate matter concentrations 10 days later (April 29) along the west coast of
the United States (Husar et al., 2001). To the extent that climate change results in altered wind
patterns and increased extreme weather events, one might expect corresponding changes in
dispersion patterns of pollen and mold.
4-17
-------
4.4. ALLERGEN CONCENTRATION
Allergic symptoms are related to pollen in a dose-response manner (Singer et al., 2005).
While pollen concentration has been taken as the indicator of potential dose, the underlying
mechanism for allergic symptoms comes in part from the protein allergens (antigen) in the pollen
(Singer et al., 2005; Ahlholm et al., 1998; Beggs, 1998). Recent research has examined the
influence of meteorological variables such as temperature and precipitation and air pollutants,
such as carbon dioxide levels, on the concentration of allergen protein, or the allergenicity of
pollen produced by ragweed and birch. The major allergen proteins in ragweed and birch are
AMB A 1 (Antigen E) and BET V 1, respectively.
Using controlled environmental chambers, Singer et al. (2005) evaluated how AMB A 1
allergen concentrations changed in response to rising carbon dioxide concentrations. The
authors used an enzyme-linked immunosorbent assay to quantify AMB A 1 in protein extracted
from pollen of A. artemisiifolia grown at different CC>2 concentrations in a previous experiment.
The CC>2 concentrations were 280, 370, and 600 umol/mol. A key finding was that, while total
pollen protein remained unchanged, AMB A 1 concentrations increased as a function of CC>2
concentrations. Relative to pollen grown at current CC>2 concentrations (i.e., 370 umol/mol),
pollen grown at 700 umol/mol of CC>2 contained 1.6 times more AMB A 1 allergen (p < 0.01).
The authors note that recent and projected increases in CC>2 concentrations could directly
increase the allergen concentrations in ragweed pollen, and consequently, the prevalence and/or
severity of seasonal allergic disease. They also point out, however, that genetic and abiotic
factors governing allergen expression will need to be better established to fully understand these
data and their public health implications.
Ahlholm et al. (1998) investigated the impact of genetics and temperature on the allergen
content of birch (B. pubescens) pollen by studying trees of 10 half-sib families. Pollen samples
were collected from two tree line gardens where the daily mean temperatures were different
during the growing season due to altitudinal differences between the gardens. The temperature
difference was approximately 1.0 to 2.5°C. After controlling for descendant group, the authors
found that IgE-immunoblotting responses were stronger in sera exposed to pollen grown at the
higher temperature. It is unclear whether the effect originated during the previous or the current
growing season. Differences in allergen concentration were also seen between different
4-18
-------
progenies of trees. The authors suggest the lower soil temperature, which limits nutrient (i.e.,
nitrogen) uptake and, thus, the rate of allergen synthesis, as a possible mechanism.
The research on allergen content of pollen/mold is limited but does suggest that even if
pollen production remained unchanged, allergic illness could increase due to increasing levels of
allergenic protein within pollen grains.
4.5. CLIMATE VARIABILITY AND ALLERGIC DISEASES
Climate change caused by elevated greenhouse gases, including carbon dioxide, is
expected to lead to increases in global mean temperature, a stronger hydrologic cycle, and an
increase in the number and severity of extreme weather events. These changes may lead to
alterations in the production, distribution, and dispersion of aeroallergens, as well as changes in
allergen protein concentration. It is possible that production (both timing and amount) and
protein content of aeroallergens will increase, and with time, plant distributions will shift as well.
If changes in aeroallergen production occur as a result of climate change, then the patterns of
seasonal allergic disorders, such as allergic rhinitis (hay fever), asthma, and possibly atopic
dermatitis could be affected as well.
The development of allergic illness is a multistage process in which a genetically
predisposed and immunologically naive individual is first sensitized to an allergen, resulting in
the production of IgE antibodies (Nielsen et al., 2002), and then subsequent exposures elicit a
disease response due to the presence of IgE antibodies and the associated cellular response.
Furthermore, there appears to be a dose-response relationship between allergen exposure,
sensitization, and exacerbation of disease (Neilsen et al., 2002). Thus, there are at least three
causal pathways for climate change-induced impacts on aeroallergens to alter the severity and
possibly the prevalence of allergic diseases. First, a longer exposure during sensitization may
lead to greater likelihood of the development of allergy (increased prevalence). Second, a higher
dose during sensitization may lead to a greater likelihood of development of an allergy
(increased prevalence). Third, a higher dose during subsequent exposures (postsensitization)
may lead to a more severe allergic response (Nielsen et al., 2002). Figure 4-3 outlines this
process using asthma as an example (Beggs and Bambrick, 2005).
Definitive statements on the impact of climate change on aeroallergens and subsequent
allergic illness, however, are rarely found in the literature. This is in part because studies are of
4-19
-------
necessity often narrowly defined, and a single
study is unlikely to encompass the broad subject
of weather, aeroallergens, and allergic illness.
There is also an inherent uncertainty as to how the
climate will change, especially at a regional level.
The etiology of allergic diseases, especially
asthma, is complex and has a gene-environment
interaction that is poorly understood. In addition,
there are numerous other factors that come into
play, such as changes in land use, air pollution,
adaptive responses, and modifying factors that are
difficult to assess (see Figure 4-4). The
interaction between atmospheric pollutants and
aeroallergens adds a further complication, as there
is evidence of mediating effects of air pollution
on the allergic responses in exposed individuals
Human activities
Almospheric CO,
T
'Temperature
T
Earlier poll en season
Pollen season duration
"Pollen allergen! city
Altered distribution
T Plant g rovyth. germination.
biomass
Pollen quantity
T
T Expo sure to allergens
_
\ T Sen sitization to al lerge ns [
3
L
\ Severity
of asthma
episodes
I Incidence of allergic
asthma
\
T Asthma prevalence
I
H
* Frequency
of asthma
episodes
"Burden of asthma
>
Figure 4-3. Schematic diagram of the
relationship between global climate
change and the rise in asthma
prevalence and severity, via impacts of
climate change on plant and pollen
(D'Amato et al., 2002). However many studies of attributes.
the health effects of air pollutants do not take into Source: Beggs and Bambrick (2005,
account the possible confounding, synergistic, or Figure 4).
antagonistic effect of aeroallergens. It is important that more studies, particularly ones
investigating the role of air pollution on asthma exacerbations, take into account aeroallergens to
further understanding of this complex interaction.
This section reviews the evidence and provides a qualitative assessment of the likely
impact of climate change on allergenic illnesses based on the expected changes in production,
distribution, and dispersion of aeroallergens and allergen content of aeroallergens in response to
climate change. It then reviews a limited number of studies on weather, aeroallergens, and
allergic disease.
4.5.1. Timing of Aeroallergen Production and Subsequent Illness
Shifts in phenology are one of the most consistent findings in studies of plant pollen
production (Root et al., 2003). Alterations in the timing of aeroallergen production in response
4-20
-------
to weather variables have been clearly demonstrated for certain tree species, but less so for grass
and weed pollens and mold (Clot, 2003; Emberlin et al., 2002; Katial et al., 1997). This is
consistent with the observation that the flowering of many trees is regulated by temperature,
whereas photoperiod determines the flowering of many weeds in late summer. Evidence is
mixed for grass pollens, with trend studies showing substantial differences by region in England
(Emberlin, 1994), earlier start dates in Switzerland (Clot, 2003), but no apparent effect after an
El Nino event in New England (Freye and Litwin, 2001).
Climate change
(natural and
human-
caused)
Regional
weather
changes
•Heatwaves
• Extreme
weather
•Temperature
• Precipitation
Atmospheric
concentrations
of pollutants
•PM
•SQ2
•NO,
•CO"
Increased respiratory
symptoms and illness
Exacerbated chronic
heart and lung
disease
Accelerated lung
aging
Increased lung
cancer risk
Increased risk of
premature death
Aeroallergens
ia mount, timing.
and distribution)
I/
Research
s-
j**^ J
Allergic diseases
Asthma
Allergic rhinitis
Adaptation
measures'1
Figure 4-4. Potential air pollution-related health effects of climate change.
Notes: "Moderating influences include nonclimate factors that affect
climate-related health outcomes, such as population growth and
demographic change, standards of living, access to health care,
improvements in health care, and public health infrastructure.
6 Adaptation measures include actions to reduce risks of adverse health
outcomes, such as emission control programs, use of weather forecasts to
project air quality levels, development of air quality advisory systems, and
public education.
Source: Bernard et al. (2001, Figure 1).
4-21
-------
Ragweed has been shown to flower earlier in urban environments where temperature and
CC>2 concentrations were higher compared to rural areas (Ziska et al., 2003). There was limited
evidence on the start dates for the emergence of mold, although the El Nino event in New
England indicated an earlier start (Freye and Litwin, 2001). Some mold such as Alternaria is
associated with agriculture, and therefore, the timing of production will be associated with the
harvest (Corden and Millington, 2001). The concentrations of indoor allergens (e.g., dust mites,
cockroaches) do not vary seasonally.
Analyses of trends in allergenic illness, however, are based on annual prevalence and
generally do not document the seasonal timing of these illnesses within the year. Nevertheless,
in sensitized individuals, exposure clearly leads to allergic response; thus, it is reasonable to
expect that changes in the timing of production of seasonal aeroallergens would result in
corresponding changes in the timing of the associated seasonal allergenic illness (i.e., rhinitis).
There is not clear evidence that the timing of mold emergence has shifted, and indoor allergens
are generally not seasonal. In addition, the relationship between indoor allergens and climate
change is unclear. Shifts in the timing of asthma and atopic dermatitis in response to changes in
phenology are not as predictable.
4.5.2. Aeroallergen Production, Allergen Content, and Subsequent Illness
Increases in aeroallergen production and/or protein concentration could impact the
prevalence or severity of allergic illness via sensitivity and response pathways. A key
conclusion of the NAST (Melillo et al., 2001) was that over the next few decades, climate
change is likely to lead to increased plant productivity and carbon storage for many parts of the
country, especially those areas that become warmer and wetter. Therefore, pollen production
and possibly mold (e.g., Cladosporidium) in these areas would be expected to increase, on
average. The NAST also concluded that areas where soils dry out during the growing season,
such as the Southeast under certain scenarios, are likely to see reduced productivity and carbon
storage, and hence, less pollen production.
These conclusions are supported by experimental and field studies that have
demonstrated increased pollen production in ragweed and other species in conditions similar to
those expected with climate change (Ziska et al., 2003, Jablonski et al., 2002; The Center for
Health and the Global Environment, 2005). There are several examples where regional weather
4-22
-------
patterns, i.e., increased precipitation and temperature, lead to stronger pollen production (Freye
and Litwin, 2001; Reiss and Kostic, 1976; Weber, 2003a, b). One study conducted in the United
States showed increasing trends of total pollen production in cedar, oak, hickory, pecan, and
birch in Oklahoma that may have been attributable to warmer winters (Levetin, 2001). In
addition, studies of birch and ragweed provide evidence of increasing allergen content under
similar conditions.
However, while the prevalence of both hay fever and asthma have increased in recent
years (see Section 3.2), the limited observational data on aeroallergen trends in the United States
present some difficulty in making an association to the observed increases in these allergic
diseases. While there is at least one regional example of increasing trends in mold (Katial et al.,
1997) and tree pollen (Lapidus, 2001), the observational studies of United States pollen levels do
not appear to have sufficient data (i.e., >10 years) to conduct trend analyses. The increases in
allergen content observed in experiments, however, may provide an alternative explanation for
increasing allergic illness prevalence in the absence of documented increases in pollen levels.
On the basis of model projections by the NAST, pollen production in many areas of the
country may increase until mid-21st century. It is also possible but less clear if allergen content
and mold production may increase as well. Exposure to elevated pollen and mold concentrations
during sensitization may lead to a greater likelihood of development of an allergy such as rhinitis
or asthma—i.e., the prevalence of allergic disease might increase. In addition, exposures to
higher concentrations of aeroallergens or allergen proteins may lead to more severe allergic
responses (Nielsen et al., 2002; Singer et al., 2005). It is unclear how indoor allergen
concentrations might change, but there may be changes in exposure patterns. For example, more
time could be spent indoors during summer heat waves, but less time could be spend indoors
during the winter as minimum temperatures rise (Patz et al., 2000).
These inferences are similar to the findings of a recently published report, Climate
Change Futures, sponsored by Swiss Re and the United Nations Development Programme and
conducted by The Center for Health and the Global Environment at Harvard Medical School.
The Climate Change Futures (CCF) project relied on two scenarios of gradual warming with
growing variability and more weather extremes. These scenarios were then applied to case
studies, one of which included asthma. Both scenarios are based on business-as-usual, which, if
unabated, would lead to a doubling of atmospheric CO2 concentrations by the mid 21st century.
4-23
-------
The first impact scenario (CCF-1) is based on gradual warming, with increasing variability and
escalating impacts. The second impact scenario (CCF-2) is also based on gradual warming with
increasing variability but includes surprise impacts due to abrupt climate change. The CCF-1
envisions a perceptible impairment of public health as a result of higher concentrations of
aeroallergens whether measured by morbidity and mortality, disability adjusted life years
(DALY)11 lost, or the value of the incremental medical resources devoted to the emerging
medical problems. The CCF-2 suggests that the combination of more aeroallergens, more heat
waves, photochemical smog, greater humidity, more wildfires, and more dust and particulates
could considerably compromise respiratory and cardiovascular health in the near term.
Widespread respiratory distress is plausible for large parts of the world, bringing with it
increasing disability, productivity losses, school absences, and rising costs for health care and
medications.
Ecological models indicate climate change will likely lead to increased plant productivity
and carbon storage in many parts of the country. Experimental and observational analyses
support model assessments, but production changes may be species- and region-specific. Data
gaps limit assessment of trends in United States pollen and mold. Increases in aeroallergen
production and/or allergen content could lead to increased prevalence and severity of allergic
diseases. A recent report by Harvard Medical School envisions perceptible impairments in
public health as a result of higher concentrations of allergens due to climate change (The Center
for Health and the Global Environment, 2005).
A study in Genoa, Italy observed that the total yearly pollen counts for Parietaria pollen
significantly increased from 1981 to 1997 but that pollen sensitivity in the population did not
show much year-to-year variability. However, sensitivity to Parietaria was highest in the
population living in the area where Parietaria had the highest pollen count of the pollens studied.
For the other pollens they studied, counts of Poaceae and Artemisia did not show any significant
upwards trends, and while Ambrosia pollen counts have been increasing during the time period
studied, very few people are sensitized to it, which could be attributed to its low overall count
compared to the other pollens studied (Voltolini et al., 2000).
nDALY = The sum of years of potential life lost due to premature mortality and the years of productive life lost due
to disability.
4-24
-------
4.5.3. Distribution and Dispersion of Aeroallergens and Subsequent Illness
Changes in the geographic distribution of plants and mold may alter the distribution of
allergic illness. Long-term responses to climate change (over 50 to 100 years) are likely to
involve range or distribution shifts in species and, in some cases, extinction of species (Joyce et
al., 2001; Melillo et al., 2001; Davis and Shaw, 2001). The results of ecological models indicate
that the potential habitats (i.e., distribution) for many tree species in the United States are likely
to change, in some cases dramatically, by the end of the 21st century. Potential habitats for trees
favored by cool environments are likely to shift northward (Joyce et al., 2001). The habitats of
alpine, subalpine spruce/fir, and aspen communities are likely to contract dramatically in the
United States and largely shift into Canada. Potential habitats are likely to increase in the United
States for oak/hickory, oak/pine, ponderosa pine, and arid woodland communities. Projections
for (nonforest) vegetation redistribution suggest that savannahs and grasslands are likely to
expand, especially in the Southeast, where hot and dry climate conditions are projected in
response to climate change.
The models developed for the NAST are supported by fossil record evidence. Davis and
Shaw (2001) note that changes in geographic distribution are so frequently documented in the
fossil record that range shifts are seen as the expected plant response to future climate change.
These authors cite evidence of genetic adaptation to climate and argue that the interplay of
genetic adaptation and migration has been central to the biotic response to climate change.
Assessing the potential impact of vegetation range shifts on allergic illness is difficult.
Shifts in vegetation distribution are likely to occur over relatively long periods of time, i.e.,
decades. Furthermore, cross-reactivity between species implies that the range of a species (e.g.,
birch) could contract or move northward and another (e.g., white oak) could take its place
without any appreciable difference in allergic illness. However, one can look to examples of
invasive and cultivated species to assess the potential impacts on allergic illness. Ragweed, for
example, has spread throughout Europe in recent decades and is now regarded as a major
allergen in France, north Italy, Hungary, and Croatia (Stefanic et al., 2005) and is beginning to
spread into Switzerland (Taramarcaz et al., 2005). Pollen counts for ragweed are increasing in
Europe as are the number of people allergic to ragweed in Hungary, northern Italy, the Rhone
area of France, Prague and Brno in the Czech Republic, and Vienna, Austria (Rybnicek and
Jaeger, 2001); however, no statistical link was made between the pollen counts and prevalence of
4-25
-------
ragweed allergy. In desert regions such as the southwestern United States, the natural vegetation
is primarily animal- or insect-pollinated (Sneller et al., 1993). However, urban development and
landscape preferences for grasses and shade trees (i.e., wind-pollinated plants) in areas such as
Tuscon, Arizona have led to dramatic changes and increases in the pollen burden (Sneller et al.,
1993).
There is some indication that the introduction of new pollens into a region can lead to
increased allergic illnesses in the population, even among people who did not have prior allergies
to other pollens. The increase in pollen allergies to birch and ragweed pollens has been
attributed to the introduction of birch trees into the suburbs north of Milan in the 1970s and
1980s and the natural spread of ragweed into the area in the 1980s, and an investigation into
people with new allergies solely to birch or ragweed pollen found that they had an older age of
onset of allergies and were less likely to have a family history of allergic illnesses compared to
people allergic to other pollens naturally found in the area (Asero, 2002).
There has been only limited research on how climate change could effect the dispersal of
pollen and mold. Dispersion has the potential via shifts in long-term weather patterns and
extreme weather events to expose human populations (sensitize) to novel allergens and to create
severe and possibly life threatening exposures. There are cases of both pollen and dust being
dispersed long distances from their release sites. For example, long distance (200-600 km)
dispersion ofJunipems ashei pollen has been routinely observed in Tulsa, Oklahoma and is
associated with allergic illness in that community (Van de Water et al., 2003). Transcontinental
transport of dust particulates has also been observed (Husar et al., 2001). During April 1998,
two large dust storms occurred over the Gobi desert (Mongolia and north central China). The
dust plume crossed the Pacific Ocean and resulted in strong spikes in particulate matter
concentrations 10 days later (April 29) along the west coast of the United States (Husar et al.,
2001). It is unclear if there were any health impacts associated with the dust in the United States,
but state health agencies issued air pollution advisory warnings to the general public
(Husar etal., 2001).
In summary, shifts in vegetative distribution are expected to occur but over relatively
long periods of time. There does not appear to be any literature estimating the impact of climate
change on the distribution of aeroallergens and subsequent illness. The impact of climate change
on aeroallergen dispersion and subsequent illness does not appear to be well studied either.
4-26
-------
There are specific examples of dispersion, indicating that exposure to novel aeroallergens or
unusually high concentrations of allergens are distinct possibilities. Overall, however, it is
difficult to project how changes in dispersal patterns and geographic distribution of plants and
mold may impact allergic illness.
4.5.4. Observational Studies of Weather, Aeroallergens, and Illness
There are several examples of observational studies that provide a linkage between
weather, aeroallergens, and health outcomes. These studies provide limited evidence of the
seemingly obvious but difficult to demonstrate link between weather, aeroallergen production,
and subsequent illness. A study by Epton et al. (1997) is one of the few examples of a
prospective design that integrates the three variable categories (i.e., weather, aeroallergens,
illness) and can serve as a model for future studies.
Epton et al. (1997) conducted a 1-year prospective study to explore relationships between
weather, fungal spore counts, pollen counts, and peak expiratory flow rates (PEFRs) and asthma
in a group of asthmatic subjects. A small positive association was found between PEFR and
mean temperature. The study also found an association between days with high basidiospore
counts and nocturnal wakening and medication use to relieve asthma. The authors concluded
that the effects of weather and aeroallergens on PEFR and asthma symptoms in the studied
population were small and that other causes needed to be sought out to explain variations in
asthma severity and exacerbations. However, there were no control subjects, and 75% of the
cases were users of prescribed inhaled anti-inflammatory medications—usually corticosteroids.
Steroid use combined with the low to moderate pollen levels during the study may explain why
the authors did not find a more substantial role of aeroallergen influence on asthma.
A time-series study of the link between grass and weed pollen and emergency department
(ED) visits for asthma in Montreal found that significant increases in grass pollen concentrations
corresponded to increases in first time ED visits for asthma with a lag of 3 and 5 days, whereas
readmissions for asthma increased 4 days after an increase in grass pollen concentrations.
Adding meteorological or air pollution variables, however, did not significantly affect the
relationships found between pollen and asthma (Heguy et al., 2008).
Another study in Montreal investigated the links between weather, Ambrosia (ragweed)
pollen, and medical consultations for allergic rhinitis. Ambrosia pollen concentrations were
4-27
-------
significantly positively correlated with maximum, mean, and minimum temperatures but
negatively correlated with precipitation. The authors analyzed the number of over-consultations
each day, defined as the number of consultations greater than the average for that day of week
during pollen season, with Ambrosia pollen concentrations while controlling for meteorological
variables and pollen season. Logistic regression showed strong correlations between pollen
count and over-consultations for allergic rhinitis on the day of high pollen exposure as well as 1,
2, 3, and 5 days after high pollen exposure (Breton et al., 2006).
Lewis et al. (2000) investigated the joint effects of aeroallergens, rainfall, thunderstorms,
and outdoor air pollutants on daily asthma admissions and Accident and Emergency (A&E)
attendance using routinely collected data between 1993 and 1996 in Derby, England. The
authors found a significant interaction between the effects of grass pollen and weather conditions
on A&E attendance, such that the increase in attendance with grass pollen count was most
marked on days of light rainfall. Asthma admissions also increased significantly with
Cladosporium count.
Severe weather events also provide intriguing evidence of an association between
weather, aeroallergens, and allergic illness. Dales et al. (2003) explored the hypothesis that
thunderstorms, by increasing aeroallergen levels, cause exacerbations in asthma. The analysis
was done using 6 years of emergency department visit data with approximately 4,000 asthma
hospital admissions yearly. Air pollution, meteorological factors, and aeroallergen levels were
accounted for simultaneously. The authors found an average daily rate of 8.6 asthma visits on
days without thunderstorms and a 15% increase to 10 visits (p < 0.05) on days with
thunderstorms. The concentrations of total fungal spores almost doubled during thunderstorms
(from 1,512 to 2,419 m3). A time series analysis was used to test the association between
changes in daily concentrations of aeroallergens and changes in the daily number of emergency
visits irrespective of thunderstorms; there was a significant association with fungal spores but not
pollen. Air pollution was also higher on days with thunderstorms compared to days without, but
the time series analysis detected no significant effect of these pollutants (63, SOx, NOx, haze) on
asthma.
Analysis of cases of asthma hospital admissions during a thunderstorm compared to
asthmatic controls found that there were high correlations between the timing of the
thunderstorm, concentrations ofDidymella, Cladosporium, and broken Alternaria spores, and
4-28
-------
asthma hospital admissions. They also found that the odds of being sensitive to either Alternaria
and/or Cladosporium spores was significantly increased for the thunderstorm-induced asthma
admissions compared to controls (Pulimood et al., 2007).
Hurricane Katrina provides a recent example of extreme weather in the United States and
the potential impact on aeroallergens and allergic illness. Large sections of New Orleans were
flooded for weeks, resulting in extensive mold growth in buildings. The CDC assessed the
extent of mold growth in a sample (N=\ 12) of households in the area (Orleans, Jefferson,
Plaquemines, and St. Bernard Parishes) and collected indoor (N= 20) and outdoor (N= 11) air
samples. Almost half the homes had "visible mold growth," and 17% had "heavy mold
coverage," defined as ">50% coverage on [the] interior wall of most-affected room" (Ratard et
al., 2006). Indoor and outdoor air sampling indicated Aspergillus spp. and Penicillium spp. were
the predominant populations (Ratard et al., 2006). Geometric mean glucan levels were 1.6 ug/m3
inside homes and 0.9 ug/m3 outside. Geometric mean endotoxin levels were 23.3 EU/m3
(endotoxin units per cubic meter) inside and 10.5 EU/m3 outside. Solomon et al. (2006) also
took indoor (N= 8) and outdoor (N= 2 3) mold samples in October and November 2005 in New
Orleans and found that the main mold spores found in outdoor air were Cladosporium,
Aspergillus, and Penicillium species, whereas indoors, more than 70% of the identified spores
were Aspergillus and Penicillium. Their samples showed much higher spore concentrations in
outdoor flooded areas (mean = 66,167 spores/m3) compared to unflooded areas (mean =
33,179 spores/m3), which were also higher than the background spore concentration for the
region (mean = 23,835 spores/m3), and that inside spore concentrations (mean =
320,005 spores/m3) were higher than concentrations just outside. The authors also found that
homes that had been partially or fully remediated for mold had lower spore concentrations than
the homes that had not been touched since the flooding. They found no appreciable difference in
endotoxin concentrations for flooded and unflooded areas.
Hospitals in the area have reported seeing an increased number of patients with allergy
and cold symptoms, and doctors have suggested that allergy to the mold and dust circulating in
New Orleans is making residents susceptible to respiratory illness (Wilson, 2006). There are
also reports of a nagging cough throughout New Orleans that has been nicknamed "Katrina
cough," believed to be caused by high levels of "dust" in the air—particles from construction
debris and dried mud, coupled with high spore counts from mold and mites that feed on mold
4-29
-------
spores (Bennett, 2006). This a particular concern for workers removing debris (Wilson, 2006).
Overall rates of asthma in Louisiana children have also increased post-Katrina from 14% (2003)
to 18% (2006) according to results from the Louisiana Child & Family Health Study, and may be
even higher for minority and underprivileged children or children residing in certain
geographical areas that were affected by post-Katrina flooding (The Center for Health and the
Global Environment, 2005).
Studies examining the relationship between weather, aeroallergens, and health outcomes,
provide intriguing evidence of potentially serious impacts on health. For example, asthma
prevalence is reportedly higher in post-Katrina Louisiana; spikes in mold spore concentrations
and asthma have been observed on days with thunderstorms. Additionally, light rain and grass
pollen counts were associated with asthma admissions in the United Kingdom. However, for
diseases with complicated etiologies, such as asthma, more rigorous prospective designs as
conducted by Epton et al. (1997) may be required to better understand the relationship between
weather, aeroallergens, and illness.
4.5.5. Linkages Among Air Pollution, Aeroallergens, and Allergic Diseases
Some recent studies have shown that the links between air pollution, aeroallergens, and
allergic diseases are complex and that air pollution may play a significant role in the etiology of
some allergic diseases. D'Amato et al. (2002) hypothesize that the reason for the increase in
urban allergic disease could be due to the role that air pollutants play in mediating the health
effects of aeroallergens. The authors summarize the literature that demonstrates that the
inflammatory effects of ozone, particulate matter, and sulfur dioxide allow for easier penetration
of pollen allergens into the airways, that air pollutants can increase the release of antigens in
pollen grains that lead to allergic responses, and that pollutants can also absorb pollen grains and,
thus, prolong their retention in the body. Given that climate change may increase
temperature-dependent air pollutants as well as aeroallergens, the interactions between criteria
air pollutants and aeroallergens are important for further study.
Knowlton et al. (2007) mapped the locations of ragweed pollen prevalence and the areas
with at least one ozone exceedance day per year on the same map of the United States, thereby
indicating the locations where people would be at an increased risk of asthma symptoms from
two sources known to affect asthma. Notably, 14 of the 15 cities listed by the Asthma and
4-30
-------
Allergy Foundation of America as the "most challenging places to live with asthma" fell in areas
that had overlapping risks of ozone and ragweed. These regions of overlap also highlight regions
of future vulnerability, given that both ragweed prevalence and ozone levels are projected to
increase with climate change.
In observational studies of asthma, where aeroallergens and air pollutants could both be
the cause, it is important that analysis be done on the potential for confounding of one by the
other. Of the few studies that have investigated such confounding, differing results have been
found, which may be due to the statistical methods used, the location of the study, or other
factors. For example, in a study investigating the link between air pollution and emergency
department visits for asthma in Alberta, Canada, the researchers found that adding aeroallergen
data to their time stratified case-crossover study design did not change their risk estimates for air
pollution's effects on ED visits for asthma (Villeneuve et al., 2007). However, a study focusing
on the relationship between grass pollen and ED visits for asthma in Melbourne, Australia found
that the air particle index and NC>2 were independently associated with asthma ED visits but that
grass pollen was still associated with asthma ED visits even with all of the air pollutants added to
the model (Erbas et al., 2007). These are just two examples of studies that investigate the
complex interactions of impacts of air pollutants and aeroallergens on respiratory health effects.
The interrelationships between climate variability, air pollution and aeroallergens are
very complex, so that projections of climatic changes for any of these gets complicated by the
lack of complete understanding and the many variables involved. A recent review of these issues
by D'Amato and Cecchi (2008) explains this complexity and demonstrates how projected
climate changes could have both positive and negative effects on respiratory allergies given the
many multifaceted ways in which climate changes and other global changes of urbanization and
increased energy uses may impact air pollutants, aeroallergens, and, therefore, allergic illnesses.
4-31
-------
4-32
-------
5. ECONOMIC AND QUALITY-OF-LIFE IMPACTS OF ALLERGIC DISEASES
This section of the report focuses on the costs, both monetized and nonmonetized, of
allergic diseases. Unless stated otherwise, all costs are in 2005 dollars.12 The incidence of
allergic disease has grown substantially in recent years, affecting millions of people annually.
Allergic reactions can involve several organ systems, including the respiratory tract, skin,
cardiovascular system, and the gastrointestinal tract. A recent nationwide survey reported that
54.6% of people in the United States test positive for one or more common allergens (AAAAI,
1996-2005); among specific allergens, dust mite, rye, ragweed, and cockroach caused
sensitization in approximately 25% of the population (Arbes et al., 2005). Allergies are the sixth
most costly chronic disease category in the United States, costing the health care system
approximately $21 billion annually (AAAAI, 1996-2005).
Although there are several different types of allergic disease affecting the respiratory
tract, skin, and other organ systems, this section discusses the costs of those allergic diseases that
have been associated with common aeroallergens in the United States—primarily, allergic
rhinitis/rhinoconjunctivitis (hereafter referred to as "allergic rhinitis"), asthma, and atopic
dermatitis/eczema (hereafter referred to as "atopic dermatitis"). Table 5-1 shows nationwide
hospital statistics for the conditions of interest.
The AAAAI reports that allergic rhinitis affects approximately 40 million people in the
United States each year, 40% of whom are children. Estimated total direct costs of treatment are
$6.2 billion per year. Indirect costs include 3.8 million missed days of school and work per year.
Allergic rhinitis seldom results in hospitalization. In 2003, the total number of hospital
discharges with allergic rhinitis listed as the principal diagnosis was 293; the total number
including those with allergic conjunctivitis in addition was 368. The aggregate charges totaled
$2.1-$2.5 million (see Table 5-2 and Table 5-3). For the most part, direct medical costs of
allergic rhinitis treatment can be attributed to outpatient services and medications (Schoenwetter
et al., 2004).
12Medical costs were inflated to 2005 U.S. dollars using Bureau of Labor Statistics Consumer Price Index (BLS
CPI) for Medical Care. All other costs were inflated to 2005 dollars using BLS CPI.
5-1
-------
Table 5-1. National hospital statistics, 2003—principal diagnosis only (all conditions)
Asthma
(ICD-9: 493)
Allergic rhinitis3
(ICD-9: 477)
Allergic conjunctivitisb
(ICD-9: 372.05, 372.13, 372.14)
Allergic rhinoconjunctivitis
AR+AC
Atopic dermatitis/eczema0
(ICD-9: 691.8 692.9 373.3)
Total number
of discharges
469,738
293
75
368
2,582
LOS (length of
stay), days
(mean)
3.4
1.9
2.1
1.9
3.2
Mean hospital
charge (2005$)
12,623
7,192
5,629
6,870
9,163
Aggregate charges,
(the "national bill")
(2005$)
5,931,347,575
2,109,848
420,298
2,530,146
23,801,038
Admitted from
emergency
department
338,659
(72.10%)
178
(60.90%)
45
(60.56%)
224
(60.83%)
1,550
(60.03%)
In-hospital
deaths
1,669
(0.36%)
0
(0.00%)
0
(0.00%)
0
(0.00%)
0
(0.00%)
aAllergic rhinitis or AR.
bAllergic conjunctivitis or AC (defining ICD codes adopted from (Ray et al., 1999).
"Defining ICD codes adopted from (Ellis et al., 2002).
Notes: 2003 dollar values were inflated to 2005 dollars using BLS CPI for Medical Care.
Source: Weighted national estimates from HCUP Nationwide Inpatient Sample (NIS), 2003, AHRQ, based on data collected by individual states and
provided to AHRQ by the states. Total number of weighted discharges in the United States based on HCUP NIS = 38,220,659.
(http://hcup.ahrq.gov/HCUPnet.asp).
-------
Table 5-2. Allergic rhinitis national hospital statistics, 2003—principal
diagnosis only
All discharges
Age
group
Region
<1
1-17
18-44
45-64
65-84
85+
Northeast
Midwest
South
West
Total number
of discharges
293 (100.00%)
*
71 (24.17%)
100 (34.29%)
52 (17.93%)
55 (18.90%)
*
89 (30.54%)
*
133 (45.27%)
*
LOS (length
of stay),
days (mean)
1.9
*
1.6
1.8
2.8
1.6
*
1.7
*
2.2
*
Mean
hospital
charge,
(2005$)
7,192
*
5,948
7,372
*
6,855
*
4,809
*
9,224
*
Aggregate
charges, (the
"national bill")
(2005$)
2,109,848
*
420,884
740,138
*
379,407
*
429,975
*
1,222,748
*
Admitted
from
emergency
department
178 (60.90%)
*
41 (57.76%)
58(57.51%)
24 (46.62%)
46 (83.33%)
*
65 (73.04%)
*
82(61.64%)
*
In-
hospital
deaths
0 (0.00%)
*
0 (0.00%)
0 (0.00%)
0 (0.00%)
0 (0.00%)
*
0 (0.00%)
*
0 (0.00%)
*
Notes: 2003 dollar values were inflated to 2005 dollars using BLS CPI for Medical Care; ICD-9 code 477;
statistics based on 70 or fewer unweighted cases in the nationwide statistics (NIS and KID) are not
reliable. These statistics are suppressed and are designated with an asterisk (*).
Source: Weighted national estimates from HCUP Nationwide Inpatient Sample (NIS, 2003), Agency for
Healthcare Research and Quality (AHRQ), based on data collected by individual states and provided to
AHRQ by the states. Total number of weighted discharges in the United States based on HCUP NIS =
38,220,659 (http://hcup.ahrq.gov/HCUPnet.asp).
5-3
-------
Table 5-3. Allergic rhinoconjunctivitis national hospital statistics, 2003—
principal diagnosis only
All discharges
Age
group
Region
<1
1-17
18-44
45-64
65-84
85+
Northeast
Midwest
South
West
Total number
of discharges
368 (100.00%)
*
95 (25.71%)
121 (32.80%)
58 (15.87%)
75 (20.52%)
*
114(31.04%)
61 (16.72%)
142 (38.56%)
*
LOS
(length
of stay),
days
(mean)
1.9
*
1.6
1.8
2.7
2.1
*
1.7
1.4
2.3
*
Mean
hospital
charge,
(2005$)
6,870
*
5,311
6,894
9,225
7,688
*
5,214
4,231
9,142
*
Aggregate
charges, (the
"national
bill")
(2005$)
2,530,146
*
501,866
831,111
530,748
579,817
*
594,857
260,040
1,295,354
*
Admitted
from
emergency
department
224 (60.83%)
*
45 (47.84%)
78 (64.61%)
30(51.94%)
56 (74.44%)
*
85(74.51%)
31(50.26%)
86 (60.96%)
*
In-
hospital
deaths
0 (0.00%)
*
0 (0.00%)
0 (0.00%)
0 (0.00%)
0 (0.00%)
*
0 (0.00%)
0 (0.00%)
0 (0.00%)
*
Notes: 2003 dollar values were inflated to 2005 dollars using BLS CPI for Medical Care; ICD-9 codes 477 (AR), and 372.05,
372.13, 372.14 (AC, see [Ray et al., 1999]); statistics based on 70 or fewer unweighted cases in the nationwide statistics
(NTS and KID) are not reliable. These statistics are suppressed and are designated with an asterisk (*).
Source: Weighted national estimates from HCUP Nationwide Inpatient Sample (NTS, 2003), Agency for Healthcare Research and
Quality (AHRQ), based on data collected by individual states and provided to AHRQ by the states. Total number of
weighted discharges in the United States based on HCUP NIS = 38,220,659 (http://hcup.ahrq.gov/HCUPnet.asp).
Asthma is estimated to affect approximately 15 million Americans (AAAAI,
1996-2005). The condition often begins in childhood, and it has been estimated that 30% of all
patients are children. There were 1,669 deaths due to asthma in 2003 (see Table 5-4). The age-
adjusted death rate for asthma has been in the neighborhood of 5 deaths per 100,000 during the
past decade (see Figure 5-1). In addition, asthma is indicated as a "contributing factor" for
nearly 7,000 other deaths in the United States each year (NCHS/CDC, 2001). Asthma was given
as the primary diagnosis in about 500,000 hospitalizations in 2003 and was listed as a secondary
diagnosis in over 1 million hospitalizations (see Table 5-2).
5-4
-------
Table 5-4. Asthma national statistics, 2003—principal diagnosis only
All discharges
Age
group
Region
<1
1-17
18-44
45-64
65-84
85+
Missing
Northeast
Midwest
South
West
Total number of
discharges
469,738 (100.00%)
16,631(3.54%)
148,170(31.54%)
104,400 (22.23%)
111,670(23.77%)
74,650 (15.89%)
13,007 (2.77%)
1,211 (0.26%)
128,928 (27.45%)
98,392 (20.95%)
171,441 (36.50%)
70,976(15.11%)
LOS
(length
of stay),
days
(mean)
3.4
2.5
2.2
3
4
4.9
5.3
2.1
3.4
3.2
3.5
3.3
Mean
hospital
charge
(2005$)
12,623
8,655
8,201
11,748
15,626
18,099
17,949
13,197
14,979
9,188
11,250
16,777
Aggregate
charges, (the
"national
bill")
(2005$)
5,931,347,575
143,854,405
1,216,728,121
1,228,750,528
1,744,778,735
1,348,215,082
233,476,765
15,543,940
1,931,237,122
905,320,891
1,928,278,846
1,166,510,716
Admitted from
emergency
department
338,659(72.10%)
9,528 (57.29%)
97,712 (65.95%)
83,191 (79.68%)
83,997 (75.22%)
53,867(72.16%)
9,645(74.16%)
719 (59.36%)
108,523(84.17%)
65,535 (66.61%)
114,966(67.06%)
49,635 (69.93%)
In-hospital
deaths
1,669 (0.36%)
5 (0.03%)
34 (0.02%)
130(0.12%)
404 (0.36%)
829(1.11%)
268 (2.06%)
0 (0.00%)
429 (0.33%)
330 (0.34%)
558 (0.33%)
352 (0.50%)
Notes: 2003 dollar values were inflated to 2005 dollars using BLS CPI for Medical Care; ICD-9 code 493; statistics
based on 70 or fewer unweighted cases in the nationwide statistics (NIS and KID) are not reliable. These
statistics are suppressed and are designated with an asterisk (*).
Source: Weighted national estimates from HCUP Nationwide Inpatient Sample (NIS, 2003), Agency for Healthcare
Research and Quality (AHRQ), based on data collected by individual states and provided to AHRQ by the
states. Total number of weighted discharges in the United States based on HCUP NIS = 38,220,659
(http://hcup.ahrq.gov/HCUPnet.asp).
5-5
-------
Deaths/100,000 Population
ICD.7
ICD/8
ICD/10
0
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000
Year
Black Male* - - White Male —Black Female' ----White Female
* Xomvhite from 1950 to 1967.
Figure 5-1. Age-adjusted death rates for asthma by race and
sex, United States (1951-2002).
Source: NHLBI/NIH (2004). Chartbook on Cardiovascular, Lung,
and Blood Diseases (p. 70).
According to a 2000 study (AAAAI, 1996-2005), the direct costs of asthma totaled
nearly $12.5 billion (with hospitalizations the single largest portion of direct cost), and indirect
costs (lost earnings due to illness or death) totaled $9.1 billion. In 2003, the national hospital bill
for asthma was $5.9 billion (see Table 5-4). For the past decade, the number of physician office
visits has fluctuated around 10 million per year (see Table 5-3 and Figure 5-2). For adults,
asthma is the fourth leading cause of work absenteeism and "presenteeism" (significant lowering
of on-the-job productivity) resulting in nearly 15 million missed or "reduced productivity"
workdays each year (Mannino et al., 2002). Among children ages 5 to 17, asthma is the leading
cause of school absences from a chronic illness. It accounts for an annual loss of more than
14 million school days per year (approximately 8 days for each student with asthma) and more
hospitalizations than any other childhood disease (see Figure 5-3). It is estimated that children
with asthma spend nearly 8 million days per year restricted to bed (Asthma and Allergy
Foundation of America, 2000).
5-6
-------
20
15
10
Number (Millions)
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001
Year
Figure 5-2. Physician office visits for asthma, United States
(1989-2001).
Source: NHLBI/NIH (2004). Chartbook on Cardiovascular, Lung,
and Blood Diseases (p. 65).
Number (Thousands)
1200
10GO
800
600
400
200
1980
1985 1990 1995
Year
— Primary —Secondary
2000
Figure 5-3. Hospitalizations for asthma, United States
(1980-2002).
Source: NHLBI/NIH (2004). Chartbook on Cardiovascular,
Lung, and Blood Diseases (p. 66).
5-7
-------
As noted previously, atopic dermatitis is one of the most common skin diseases,
particularly in infants and children. According to the AAAAI (1996-2005), 10-15% of the
population is affected during childhood, and there is considerable evidence that the prevalence is
increasing. It often precedes other allergic disorders—up to 50% of patients with atopic
dermatitis develop asthma. A recent estimate of the direct medical costs associated with atopic
dermatitis is $1.2-$5.9 billion per annum (Ellis et al., 2002). As in the case of allergic rhinitis,
atopic dermatitis seldom results in hospitalization. The total number of hospital discharges with
atopic dermatitis listed as the primary diagnosis was 2,582 in 2003, while the aggregate hospital
charges totaled $23 million (see Table 5-5).
Table 5-5. Atopic dermatitis national statistics, 2003—principal diagnosis
only
All discharges
Age
group
Region
<1
1-17
18-44
45-64
65-84
85+
Missing
Northeast
Midwest
South
West
Total number of
discharges
2,582 (100.00%)
212 (8.20%)
755 (29.24%)
465 (17.99%)
594 (23.01%)
474 (18.36%)
67 (2.60%)
*
651 (25.20%)
551 (21.35%)
964 (37.35%)
415 (16.09%)
LOS
(length of
stay),
days
(mean)
3.2
3
2.9
2.9
o
J
3.9
4.9
*
3.2
3.1
3.3
2.8
Mean
hospital
charge
(2005$)
9,163
10,797
8,086
8,379
8,147
12,009
10,877
*
12,268
6,864
8,057
9,983
Aggregate
charges, (the
"national
bill")
(2005$)
23,801,038
2,285,682
6,134,619
3,892,126
4,840,298
5,689,869
845,183
*
7,982,066
3,783,047
7,769,804
4,266,119
Admitted from
emergency
department
1,550 (60.03%)
109(51.28%)
366 (48.48%)
354(76.31%)
392 (66.06%)
274 (57.72%)
44 (65.52%)
*
508 (78.05%)
333 (60.50%)
549 (56.95%)
159 (38.36%)
In-
hospital
deaths
0 (0.00%)
0 (0.00%)
0 (0.00%)
0 (0.00%)
0 (0.00%)
0 (0.00%)
0 (0.00%)
*
0 (0.00%)
0 (0.00%)
0 (0.00%)
0 (0.00%)
Notes: 2003 dollar values were inflated to 2005 dollars using BLS CPI for Medical Care; ICD-9 codes 691.8, 692.9,
373.3 (Ellis et al., 2002); statistics based on 70 or fewer unweighted cases in the nationwide statistics (NIS
and KID) are not reliable. These statistics are suppressed and are designated with an asterisk (*).
Source: Weighted national estimates from HCUP Nationwide Inpatient Sample (NIS, 2003), Agency for Healthcare
Research and Quality (AHRQ), based on data collected by individual states and provided to AHRQ by the
states. Total number of weighted discharges in the United States based on HCUP NIS = 38,220,659
(http://hcup.ahrq.gov/HCUPnet.asp).
5-8
-------
The impacts these allergic diseases impose on the United States economy and the
nonmonetized quality-of-life impacts they impose on the individuals who suffer from them are
discussed more fully below. Because cost-of-illness (COI) studies are the primary means by
which the direct (medical) and indirect (opportunity) costs of diseases are assessed, an
introduction to COI methodology is provided in Subsection 5.1. Recent COI estimates available
for asthma, allergic rhinitis, and atopic dermatitis are given in Subsection 5.4.
5.1. COST OF ILLNESS—METHODOLOGY
COI studies are a type of economic study common in the medical literature, particularly
in specialist clinical journals. COI studies were pioneered in the late 1950s and early 1960s and
have proliferated over the past 30 years. The aim of a COI study is to identify and measure the
costs of a particular disease, including the direct (medical) costs, the indirect (opportunity) costs,
and the intangible costs (e.g., pain and suffering). A COI study, thus, attempts to estimate the
total cost to society of a particular disease and by implication, the amount that would be saved if
the disease were abolished. It also identifies the different components of cost and the size of the
contribution of each.
The COI study is one of several
types of economic evaluation of clinical
care, as shown in Figure 5-4. While the
COI study focuses on the identification
of costs, cost-effectiveness analysis
focuses on the relative cost-effectiveness
of different treatments, and cost-benefit
analysis compares the costs of treatment
with the benefits. Economic studies also
vary with respect to the perspective
("points of view") for cost evaluation:
society, patient, payor, or provider.
Finally, the studies may include
TYPES OF
COST
INCLUDED
Intangible
Indirect
Direct
1 TYPE OF COST ANALYSIS
Figure 5-4. Three dimensions of economic
evaluation of clinical care.
different cost components. Thus, even if Source: Bombardier and Eisenberg (1985).
the studies belong to the same type (e.g.,
5-9
-------
COI), there still may be substantial variability along the other two dimensions (perspective and
cost components), which may potentially affect comparability of the estimates. The key
distinctions are summarized below.
5.2. COST COMPONENTS
Figure 5-5 summarizes the types of costs that may be subject to evaluation by a given
study. COI studies measure the economic burden resulting from disease and illness across a
defined population, including both direct and indirect costs. Direct costs are the value of
resources used in the treatment, care, and rehabilitation of persons with the condition under study
and are, therefore, unavailable to produce other goods and services. Indirect costs represent the
value of economic resources lost because of disease-related work disability or premature
mortality. In addition, a disease typically involves deterioration in the quality of life of the
patient (and his or her family) through its impacts on physical, social, and emotional health—i.e.,
intangible costs (Kirschstein, 2000).
[Total costs
I
Medical
1
>irect costs
Nonmedical
Diagnostic tests and Program
procedures administration
1
Indirect costs
Productivity losses
Time spent by patient
seeking core
I
Intangible costs
Emotional anxiety
and fear
DIIKJS .nul niedic.il
supplies
Pliysicion office
visits
Hospitaliz.ition
Physic.il |>.ice y care givers
|- iih ,in l suffering
Mi(|nnri:.Mi • n
Figure 5-5. Cost inventory diagram.
Source: CDC Economic Evaluation of Public Health Preparedness and Response
Efforts (http://www.cdc.gOv/owcd/EET/Cost/Fixed/2.html).
5-10
-------
5.2.1. Direct Medical Costs
Direct medical costs are the costs connected with the use of medical care in the
prevention, diagnosis, and treatment of disease and in the continuing care, rehabilitation, or
terminal care of patients. Examples include expenditures for hospitalization, outpatient clinical
care, nursing home care, and home health care; services of primary physicians and specialists,
dentists, and other health practitioners; drugs and drug sundries; and rehabilitation counseling
and other rehabilitation costs, such as for prostheses, appliances, eyeglasses, hearing aids, and
other devices to overcome impairments resulting from illness or disease. Collectively, these
expenditures represent the personal health care component of the United States National Health
Expenditures Accounts (Kirschstein, 2000).
5.2.2. Direct Nonmedical Costs
Direct nonmedical costs are the costs borne by patients or other payers that are not
included in the National Health Expenditures Accounts. Examples of such costs are
expenditures for transportation to hospitals, to physicians' offices, or to other health providers;
certain household expenditures (e.g., help for cleaning, laundering, and cooking); special diets
and clothing; and relocation and moving expenses (Kirschstein, 2000).
5.2.3. Indirect Costs
Indirect costs are the value of time that patients lose from employment or other
productive activity due to mortality or morbidity. These costs also include reduced productivity
once the patient returns to work, including unwanted job changes and loss of opportunities for
promotion or education, and the value of time lost from work, housekeeping, etc., by family
members or friends who transport, visit, and care for patients (Kirschstein, 2000).
5.2.4. Intangible Costs
COI studies rarely attempt to evaluate the intangible costs of disease—the associated
pain, suffering, and changes in the quality of life. This issue is of particular importance in the
case of chronic diseases (such as those considered here), where there can be a substantial impact
on the quality of life over a long period of time (Kirschstein, 2000).
5-11
-------
5.2.5. Hidden Costs
There are often, in addition, some "hidden costs" associated with diseases, which are
usually neglected by COI studies (Schoenwetter et al., 2004). A disease or condition may
contribute to increased costs as a secondary diagnosis, or as a risk factor for other diseases and
conditions. For instance, inadequately treated or untreated allergic rhinitis can be associated
with a dramatic increase in the cost of caring for comorbid conditions such as asthma, recurrent
nasal polyps, sinusitis, and chronic otitis media (Halpern et al., 2004).
5.3. SOURCES OF VARIABILITY IN COST ESTIMATES FROM COI STUDIES
The literature on COI studies documents substantial variation in the methods and data used to
estimate the overall costs of illness. Attempts to compare cost data across disease categories
should consider the conceptual and methodological issues that may lead to variations in cost
estimates. The following are the issues that should be taken into account when considering the
COI estimates within and across conditions (see Kirschstein, 2000).
5.3.1. Reference Year
COI estimates are expressed in dollars for a particular reference year. To express all
estimates in a common reference year, it is necessary to adjust for changes in the disease burden
over time, patterns of treatment and care, and the purchasing power of the dollar for health care
services (Kirschstein, 2000).
5.3.2. Cost Components
The comprehensiveness of the estimates of direct and indirect costs differs across studies
because of the difficulty and cost required to estimate the nonmedical costs, and the indirect
costs related to reduced productivity after returning to the job and the value of services of unpaid
care providers. Studies often make a number of specialized assumptions that may drive their
results (Kirschstein, 2000).
5.3.3. Discount Rate
In some cases, the present discounted value of the expected stream of lost earnings or
medical expenditures incurred over future years is calculated for a base or reference year using a
5-12
-------
discount rate intended to reflect people's rate of time preference—i.e., the tradeoff between the
value of a dollar received today versus one received next year. The choice of an appropriate
discount rate remains controversial and may vary considerably between studies (Kirschstein,
2000).
5.3.4. Definition of Disease
Because the interrelationships among disease categories or causal agents are complex and
patients often present more than one disease or condition, it is not always feasible or appropriate
to construct mutually exclusive disease categories and associated cost estimates. Cost estimates
depend on how narrowly or broadly the disease is defined, whether it includes related conditions
beyond its narrowly defined or primary ICD-10 code (International Classification of Diseases
version 10), whether the estimate includes identifiable extra costs attributable when the disease is
listed as a secondary diagnosis or comorbidity, and whether the estimate includes costs
attributable to the disease or condition as an underlying cause or risk factor for other diseases
(Kirschstein, 2000).
5.3.5. Prevalence vs. Incidence Approach
COI studies approach cost estimation from either of two perspectives. Most COI studies
use the prevalence-based (or annual cost) approach that measures the costs that accrue during a
base year due to all existing (or prevalent) cases of disease in that year. In estimating the
economic burden resulting from the prevalence of disease, the present discounted value of future
losses due to mortality occurring in the base year is calculated. The conventional methodology
attributes the future losses to the year in which the death occurs (Kirschstein, 2000).
The incidence-based (or lifetime cost) approach measures the present value of the
lifetime costs of the disease for all new (incident) cases with onset of disease during the given
base year (Weiss and Sullivan, 1993). Estimation of incidence-based costs requires knowledge
of the likely course of a disease and its duration, survival rates, onset, patterns of medical care,
and the impact of disease on employment, so it is generally more difficult than estimation of
prevalence-based costs. However, the incidence-based approach is sometimes more useful for
comparing the effects of alternative interventions to prevent, treat, or manage a particular
disease.
5-13
-------
5.3.6. Scope and Perspective of Estimation
COI estimates may focus on the total United States resident population, or they may be
specific to particular geographic areas or ethnic groups. They may cover all ages, or they may be
limited to certain age groups. Similarly, COI studies may estimate costs to the total society,
regardless of who bears the costs, or they may estimate the costs to patients, payers, or providers
(Kirschstein, 2000).
5.4. COST OF ILLNESS—ESTIMATES
The COI studies discussed below were based on United States data, but varied with
respect to scope, perspective, reference year, cost components, and, in some cases, the definition
of disease. None of the studies used the incidence approach or applied discount rates to the
stream of lost earnings over future years. The majority of COI studies were for asthma, followed
by allergic rhinitis and atopic dermatitis.
5.4.1. Allergic Rhinitis
Reed et al. (2004) and Schoenwetter et al. (2004) are two recent comprehensive allergic
rhinitis burden-of-disease literature reviews. Two older review papers by Kozma et al. (1996)
and Blaiss (2000) discuss the economic and quality-of-life consequences of allergic rhinitis.
Finally, O'Connell (2004) and Weiss and Sullivan (2001) discuss allergic rhinitis in the context
of atopic diseases, in general.
Table 5-6 summarizes the key features and findings of the original research papers on
allergic rhinitis. Direct medical costs and/or indirect costs were estimated by this body of
research. However, no attempt was made to monetize the intangible costs of allergic rhinitis. In
addition, studies vary in the way they define the condition. Some create estimates for allergic
rhinitis only, while others define the disease as "allergic rhinoconjunctivitis" by combining
ICD-9 codes for allergic rhinitis and a set of conjunctivitis-related codes (Ray et al., 1999). The
direct medical costs of allergic rhinitis range from $1.7 billion to $6.2 billion, while indirect
costs are estimated to range from $0.1 billion to $6.6 billion. Variation in estimates comes
largely from different assumptions about prevalence, inclusion of over-the-counter drugs, and
partial productivity losses.
5-14
-------
Table 5-6. Annual cost of allergic rhinitis/rhinoconjunctivitis estimates, in
2005 $ by cost category
Study name
Period
Data sources
Methodology
Malone et al., 1997
1987
Various U.S. national, NMES, 1987
S, preval. 39 mil
McMenamin, 1994
1985-1990
Various U.S. national
S, preval. 22 mil
Baraniuk et al., 1996
1990
Various U.S. national
S, meta-study
Mackowiak, 1997
Various yrs., used 1997 for inflation
adjustments
Various U.S. national
S, preval. 12 mil
Law et al., 2003
1996
MEPS
S, SR, preval. 12 mil
Ray et al., 1999
1994
Various U.S. national, NCHS
D, S, SR, preval. 27 mil
Storms et al., 1997
1993
Population-based survey
Incl. OTC, preval. 36 mil
Kessler et al., 2001
1996-1997
Repr. diary survey of 739
Partial productivity, preval. 13 mil
Crystal-Peters et al., 2000
1995
Various U.S. national, NHIS/BLS
S, preval. 26 mil
Ross, 1996
1983-1994
Various U.S. national
S, preval. 13 mil
Direct medical
costs
1,741 mil
2,285 mil
2,393 mil
6,199 mil
4,787 mil
2,611 mil
5,331 mil
NA
NA
NA
Direct
nonmedical
costs
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Hidden
costs
NA
NA
NA
NA
NA
5,665 mil
NA
NA
NA
NA
Indirect costs
Cost of lost Cost of lost
school days work
126 mil
1,168 mil
2,335 mil
4,137 mil
NA
NA
NA
10,296 mil
6,687 mil
5,838 mil
Intangible
costs
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Notes: Bold type indicates a national estimate. Medical costs were inflated to 2005 dollars using BLS CPI for
Medical Care; all other costs were inflated to 2005 dollars using BLS CPI. Methodology abbreviations: S =
society perspective, SR = serf-reported data, OTC = over-the-counter drugs, D = Delphi method used.
5-15
-------
There are several methodological issues specific to allergic rhinitis COI studies. First,
inadequately treated or untreated allergic rhinitis can be associated with a dramatic increase in
the cost of caring for comorbid conditions such as asthma, recurrent nasal polyps, sinusitis, and
chronic otitis media (Schoenwetter et al., 2004). These are among the hidden costs of allergic
rhinitis. A survey by Halpern et al. (2004) of over 27,398 patients with asthma demonstrated
that costs for those with allergic rhinitis and asthma were roughly twice those for patients with
asthma alone. Ray et al. (1999) estimate these hidden costs were $5.7 billion. Blaiss (2000)
reports that 58% of patients with asthma, 25% of patients with sinusitis, and 35% of children
with otitis media have allergic rhinitis. Second, very few allergic rhinitis COI studies consider
the cost of over-the-counter medications. Reed et al. (2004) estimate that 69% of individuals
with symptoms of allergic rhinitis used over-the-counter medications in 1993, compared with
45% who used prescription medications. Storms et al. (1997) estimate that the cost of over-the-
counter medications was $90 per patient per year. Thus, excluding the cost of over-the-counter
medications will result in a substantial underestimate of the direct medical costs of allergic
rhinitis.
Third, the symptoms of allergic rhinitis and sedating side effects of some allergic rhinitis
medications are typically not severe enough to cause work absence. However, the symptoms
may significantly lower on-the-job productivity ("presenteeism"). Thus, the studies that rely
only on estimates of days lost from work are likely to significantly underestimate the indirect
costs of allergic rhinitis. In addition, assigning monetary values to decreased work productivity
and performance at school is difficult.
Finally, studies by Tripathi and Patterson (2001) and Meltzer (2001) discuss the impact
of allergic rhinitis on the quality of life. They point out that poorly controlled symptoms of
allergic rhinitis may contribute to loss of sleep, secondary daytime fatigue, learning impairment,
decreased cognitive functioning, and decreased long-term productivity. Pharmacological
therapies in some cases have considerable adverse side effects, affecting attention, working
memory, vigilance, and speed (via sedation mechanism). However, to date no studies have
attempted to assign monetary value to the deterioration of quality of life resulting from allergic
rhinitis.
5-16
-------
5.4.2. Asthma
There is international concern about growing asthma morbidity, and the literature on
asthma is very extensive. The review papers that discuss the burden of asthma are Bousquet et
al. (2005), Milton et al. (2004), O'Connell (2004), Gergen (2001), Weiss and Sullivan (2001),
and Weiss and Sullivan (1993). In addition, there are a number of comprehensive COI studies
(Weiss et al., 1992; Weiss et al., 2000; Smith et al., 1997; Farquhar et al., 1998; Birnbaum et al.,
2002; Cisternas et al., 2003), including a recent analysis of willingness to pay (WTP) to avoid
asthma (Zillich et al., 2002).
Table 5-7 provides a summary of the available asthma COI studies conducted in the
United States. As with allergic rhinitis, the studies differ substantially in cost estimates and
methods employed; however, importantly, efforts have been made to estimate all known cost
components. Direct medical costs range from $2.7 billion to $16.9 billion in total (and from
$1,340 to $3,600 per patient) per annum. Cisternas et al. (2003) estimated $579 per patient in
direct nonmedical costs. Hidden costs were estimated to be $2,450 per patient in extra direct
medical costs for asthma comorbidities and $373 per patient in work loss costs due to
exacerbating effects of asthma on related conditions (Birnbaum et al., 2002).
Total indirect costs are not always comparable due to differences in components
included. Estimates that include loss of life are substantially higher ($2.6 billion to $6.1 billion
as compared to $0.9 billion). In addition to these estimates, a recent comprehensive study of
productivity loss by Ward et al. (2002) reports that 25.2% of asthma patients in their sample
were unable to work, 17.5% were limited in kind or amount of work, and 47.2% attributed the
limitation in their ability to work to asthma.
Weiss and Sullivan (2001) noted that (1) asthma imposes a considerable financial burden
on the family, which may adversely affect access to care by poorer individuals; and
(2) emergency department visits and hospitalizations are the key components of asthma care,
with estimated costs per family of $2,784-$4,057 per annum. Hospitalization and medication
represent two thirds to three quarters of total direct asthma-related costs (Gergen, 2001).
Stanford et al. (1999) conclude that nursing accounts for the largest portion of hospital costs
(43.6%), followed by respiratory therapy (13.6%), and medications (10.4%). Based on
international comparisons, the percent of direct costs associated with hospitalization appears to
be inversely correlated with the percent associated with medications (Gergen, 2001). This
5-17
-------
Table 5-7. Annual cost of asthma estimates, in 2005 $ by cost category
Study name
Period
Data sources
Methodology
Vance and Taylor, 1971
1967-1969
3-year panel of 21 families with active
asthmatics
F
US NHLI, 1972
1967
Various U.S. National data sources
S
Marion et al., 1985
1977-1980 used 1980 for inflation
adjustments
1-year panel of 30 families with
asthmatic child
F
Weiss et al., 1992
1985
Various U.S. National data sources
S, CHG, SR
Smith et al., 1997
1987
NMES
S, SR
Farquhar et al., 1998
1996
1987 NMES,
S
Weiss et al., 2000
1994
Various U.S. National data sources
S, CHG, SR
Birnbaum et al., 2002
1996-1998 used 1998 for inflation
adjustments
Claims data for Fortune 100 national
company
E, case-control
Zillich et al., 2002
2002 (used 2002 for inflation
adjustments)
Survey 100 asthmatics from
community pharmacies in KY / P,
WTP
Cisternas et al., 2003
1998-1999
MEPS, NCS, panel of 401 adults from
a sample of CA providers
S
Direct
medical
costs
2,9037
family
2,784 mil
4,0577
family
7,633 mil
7,883 mil
16,995
mil
9,357 mil
1,3407
patient
NA
3,6007
patient
Direct
nonmedical
costs
NA
NA
NA
NA
NA
NA
NA
NA
NA
5797
patient
Hidden
costs
NA
NA
NA
NA
NA
NA
NA
2,450b 7
373C/
patient
NA
NA
Indirect costs
Cost of lost Cost of Loss
schooldays lost work of life
NA
1,591 mil
3487
family
2,624 mil
911 mil
866
mil
847
mil
890 mil"
257 mil
555
mil
NA
2,489 mil
6,116 mil
1,261 mil
NA
2,725
mil
1387
patient
2,130
mil
NA
NA
NA
2,0747
patient
NA
Intangible
costs
NA
NA
NA
NA
NA
NA
NA
NA
2,102d7
1,5997
patient
NA
Components may not add up to the total because other categories of indirect costs included bed days for children under 4.
bExtra direct medical costs for other asthma-related conditions.
"Extra loss of work costs for other asthma-related conditions.
dObjective willingness to pay; e = subjective willingness to pay.
5-18
-------
Notes'. Bold type indicates a national estimate. Medical costs were inflated to 2005 dollars using BLS CPI for Medical Care; all
other costs were inflated to 2005 dollars using BLS CPI. Methodology abbreviations: F = family perspective, S = society
perspective, E = employer perspective, P = patient perspective, CHG = includes hospital charges and not costs, WTP =
willingness to pay, SR = self-reported data.
5-19
-------
relationship may reflect the well-known fact that adequately managed asthma can reduce
hospitalizations. The cost of asthma can be substantially nonuniform across asthmatics. Smith
et al. (1997) noted that less than 20% of the individuals with asthma in their sample accounted
for more than 80% of the total direct costs.
Finally, Zillich et al. (2002) estimated WTP to avoid asthma, the only measure that would
include the intangible costs of the illness. Their survey of one hundred patients with asthma
(recruited from Kentucky pharmacies) suggested that WTP was significantly related to both
objective disease severity (as defined by a physician) and disease severity subjectively assessed
by the patient. For objective disease severity, the mean monthly WTP was $97 for mild asthma,
$142 for moderate asthma, and $359 for severe asthma. For subjective disease severity, the
mean monthly WTP was $52 for mild asthma, $180 for moderate asthma, and $262 for severe
asthma. A weighted annual average is $2,102 for objective WTP and $1,599 for subjective WTP
per patient.
5.4.3. Atopic Dermatitis
A recent review paper by Carroll et al. (2005) on the burden of atopic dermatitis on
patients, family, and society indicates that COI estimates for atopic dermatitis are very limited in
the United States. This is despite the fact that atopic dermatitis is widespread and is generally
considered to be associated with substantial deteriorations in quality of life for patients and their
families. In addition, O'Connell (2004) notes that atopic dermatitis can have a large
social/emotional and financial effect on the family and often predates the development of allergic
rhinitis and asthma.
Table 5-8 summarizes the available United States evidence. Lapidus et al. (1993) studied
emergency room visits and ambulatory care billing records of an urban hospital in Philadelphia
and extrapolated the direct costs to the United States to be $665 million annually. However, this
study, published in 1993, was thought to underestimate the true cost of atopic dermatitis because
it included only ER and physician visits (Carroll et al., 2005).
In a systematic review of third party claims data, Ellis et al. (2002) estimated the direct
cost of atopic dermatitis in the United States to be $1.2-$5.2 billion. This analysis used claims
from a managed care payor and state Medicaid program, with atopic dermatitis diagnoses based
on International Classification of Diseases (ICD-9-CM) codes. Claims were reviewed by a
5-20
-------
Table 5-8. Annual cost of atopic dermatitis estimates, in 2005 $ by cost
category.
Study name
Period
Data sources
Methodology
Lapidus et al., 1993
1991
Philadelphia Children Hospital
S, national estimate
Ellis et al., 2002
1997
Claims data (private provider and a
state Medicaid program)
PAY, D, case control
Fivenson et al., 2002
1997
Survey of 248 individuals in a
managed care population
P, OTC
Direct
medical
costs
665 mil3
1,239 mil-
5,234 mil
or
1,4607
personb
427/person
Direct
nonmedical
costs
NA
NA
NA
Hidden
costs
NA
NA
NA
Indirect costs
Cost
of lost
school
days
Cost
of
lost
work
NA
NA
4 Ill/person
Intangible
costs
NA
NA
NA
aLikely to underestimate direct medical costs because includes only ER and physician's office visits.
bEstimate for privately insured individuals was $799/person.
Notes: Bold type indicates a national estimate. Medical costs were inflated to 2005 dollars using BLS CPI for
Medical Care; all other costs were inflated to 2005 dollars using BLS CPI. Methodology abbreviations: S =
society perspective, PAY = payor perspective, P = patient perspective, OTC = over-the-counter drugs, D =
Delphi method.
panel, and comorbidities were classified as most likely related to atopic dermatitis and possibly
related to atopic dermatitis (using the Delphi method to create consensus, as explained in
[Powell, 2003]). The cost quoted included all atopic dermatitis claims for visits, prescription
drugs, and "likely" atopic dermatitis-related comorbidities. The estimate, however, did not
include the costs of over-the-counter medications or any indirect costs of lifestyle changes.
Fivenson et al. (2002) estimated the direct and indirect costs of atopic dermatitis at
$838 per patient annually, using a patient survey to determine the indirect costs (including time
lost from work) and managed care claims data to assess the direct costs. The direct medical costs
(not including over-the-counter medications) were found to be only 27% of the total, suggesting
the significant underestimation that occurs if only direct costs are used to estimate the economic
burden of atopic dermatitis. Additionally, as discussed in an editorial by Ellis et al. (2003), there
5-21
-------
may have been an unrepresentatively small number of severely affected patients in Fivenson's
study sample, which would lead to lower cost estimates (Carroll et al., 2005).
As noted above, atopic dermatitis is often associated with significant morbidity. Pruritus
(severe itching, often of undamaged skin) caused by atopic dermatitis can affect both sleep and
mood, and affected individuals often must modify several aspects of their lives because of
treatment regimens and associated lifestyle changes. Individuals with atopic dermatitis are also
at risk for psychosocial difficulties that may have long-lasting consequences, potentially
affecting career choices and personal relationships. Patients are thus affected both by the
condition itself and by the stigma associated with its visibility. A number of studies have shown
that people with atopic dermatitis tend to report lower health-related quality of life and greater
psychological distress than the general population (Carroll et al., 2005). In addition, the effects
of atopic dermatitis on the entire family can be extensive. Unfortunately, monetary assessments
of these intangible costs in the United States are yet to come.
5-22
-------
6. FUTURE RESEARCH
Further progress must be made in documenting and understanding aeroallergen response
to climate, the role of aeroallergens in disease development, and the willingness to pay to
avoid—the intangible costs of these allergic diseases.
A review of the literature indicates that there are limited data on aeroallergen trends in
the United States. Integrated long-term data series on all aeroallergens is necessary to clearly
document future changes in aeroallergen production and distribution, as well as allergen content.
Additional research on the response of mold and indoor allergens to climate change would be of
particular value. In addition, further experimental and field studies are needed to examine how
allergen content and distribution of aeroallergens may be altered in response to climate change.
Such studies could address a number of key issues, including (1) the combined effects of CC>2
and temperature, as well as interactions between these and other important variables, such as
water and nutrient availability, disturbance, and competition (Beggs, 2004); (2) within-species
genetic variation in response to changing CC>2 concentration availability and temperature (Beggs,
2004); and (3) effects of urban warming or land use changes, which may alter observed impacts
of climate change (Beggs, 2004).
There is a need for better understanding of the role of aeroallergens in disease
development, especially asthma, specifically, what the relative contribution of different
aeroallergens is to the development of asthma (Selgrade et al., 2006). There is a need to know
what levels of allergen exposure constitute a risk for development of asthma (Selgrade et al.,
2006). There is also a need for standardized approaches for measuring exposures and outcomes
in epidemiologic studies (Selgrade et al., 2006). Finally, the possible synergistic effects of
aeroallergens and air pollutants on the development of allergic diseases could be an important
area for future research. For example, changes in the timing of pollen seasons could result in
some overlap between the peak pollen period and the ozone season.
Based on a review of the COI literature on allergic rhinitis, asthma, and atopic dermatitis,
it is clear that an important research gap is the current lack of assessment of—and, in particular,
estimation of, willingness to pay to avoid the intangible costs of these diseases. In addition,
better methodologies are needed to address productivity losses, aeroallergen avoidance, and
over-the-counter medication use. Finally, as noted in Section 5, a disease or condition may
6-1
-------
contribute to increased costs as a secondary diagnosis, or as a risk factor for other diseases and
conditions. These hidden costs of comorbidity need to be properly estimated and, if possible,
included in future COI studies.
6-2
-------
7. SUMMARY AND CONCLUSIONS
This report reviewed the available literature on (1) aeroallergens and associated allergic
diseases prevalent in the United States; (2) the potential impacts of climate change on these
aeroallergens and, by inference, on the allergic diseases associated with them; and (3) the
economic and quality-of-life impacts of these diseases. Although some of the relevant research
cited was carried out in other countries, this report focuses on the United States.
Aeroallergens are distributed throughout the United States, but some are concentrated in
particular geographic regions. Three allergic diseases have been associated with aeroallergens in
the United States: asthma, allergic rhinitis (hay fever), and atopic dermatitis (eczema). Although
all aeroallergens have been linked to each of these three allergic diseases, the strongest
associations appear to be between pollen (tree, grass, or weed) and allergic rhinitis (hay fever),
and between house dust mites or mold and asthma.
Limited data suggest aeroallergen levels in the United States have remained relatively
constant (though the period of record may be too short to assess trends). While significant
increases in the prevalence of allergic diseases have been observed, the factors contributing to
this increase remain unclear. At the same time, experts have hypothesized that an increase in the
distribution and concentration of aeroallergens could further increase the economic and quality-
of-life burdens imposed by these diseases in the United States.
The literature does not provide definitive data or conclusions on how climate change
might impact aeroallergens and subsequently the severity and prevalence of allergic diseases in
the United States. There is also an inherent uncertainty as to how the climate will change,
especially at a regional level. In addition, the etiology of allergic diseases, especially asthma, is
complex and has a gene environment interaction that is poorly understood. Finally, there are
numerous other factors that affect aeroallergen levels and the prevalence of associated allergic
diseases, such as changes in land use, air pollution, adaptive responses, and modifying factors;
many of which are difficult to assess.
Nevertheless, some tentative inferences can be drawn about the potential impact of
climate change on aeroallergens and the associated allergic diseases by making reasonable
inferences about the links between (1) climate change and the characteristics of aeroallergens
and (2) those aeroallergen characteristics and the associated allergic diseases. Global climate
7-1
-------
models developed for the NAST suggest that many areas of the United States will become
warmer and wetter. In addition, research has shown that preseason temperature and precipitation
have been consistently important projectors of pollen and mold production. Moreover,
atmospheric carbon dioxide concentration has increased and will continue to increase, and this
alone will have impacts on aeroallergens, regardless of changes in precipitation and temperature.
Overall, experimental and observational data, as well as models, indicate the following likely
changes in aeroallergen production, distribution, dispersal, and allergen content as a result of
climate change in the United States:
• Pollen production is likely to increase in many parts of the United States, with the
possible exception of the Southeast
• Phenologic advance (i.e., earlier start of pollen season) is likely to occur for numerous
species of plants, especially trees (Root et al., 2003)
• There will likely be changes in the distribution of pollen-producing species, including the
possibility of extinction in some cases and expansion in others (Joyce et al., 2001)
• Intercontinental dispersal (e.g., of pollen) is possible, facilitating the introduction of new
aeroallergens into the United States (Husar et al., 2001)
• Increases in allergen content of some aeroallergens are possible (Beggs, 2004; Beggs and
Bambrick, 2005)
Aeroallergen (e.g., pollen) exposure in sensitized individuals is associated with allergic
rhinitis and less clearly with asthma and atopic dermatitis. Furthermore, some studies have
demonstrated links between weather, aeroallergen production, and subsequent increased illness.
Therefore, we can infer that changes in the timing, severity, and possibly the prevalence of
allergic rhinitis (hay fever) are likely, given the clear association between allergen exposure and
response in sensitized individuals. While recent research points to a link between aeroallergens
and asthma, the complex etiology of this illness and the unclear link between indoor
aeroallergens and climate change lead to greater uncertainty about how asthma severity or
prevalence might change in response to climate change and corresponding impacts on
aeroallergens.
7-2
-------
Because the economic and quality-of-life impacts of these allergic diseases are
substantial, the corresponding economic and quality-of-life impacts of increases in the
prevalence of these diseases could similarly be significant. It has been reported that 54.6% of
people in the United States currently test positive to one or more allergens. Consequently,
allergies are the sixth most costly chronic disease category in the United States, costing the
health care system approximately $21 billion annually (in 2005 dollars). Although the allergic
diseases discussed in this report—allergic rhinitis, asthma, and atopic dermatitis—are not the
only allergic diseases in the United States, they are among the most important ones, and the costs
associated with them account for a substantial component of the total costs of allergies in the
United States.
Allergic rhinitis affects approximately 40 million people each year in the United States,
40% of whom are children. Estimated total direct costs of treatment are $6.2 billion per year (in
2005 dollars). Indirect costs include 3.8 million missed days of school and work per year.
Asthma is estimated to affect approximately 15 million Americans, and 30% of all
patients are children. Asthma can be life-threatening—there were 1,669 deaths due to asthma in
2003. According to a 2000 study, direct costs totaled nearly $12.5 billion (in 2005 dollars) and
indirect costs (lost earnings due to illness or death) totaled $9.1 billion (in 2005 dollars). For
adults, asthma is the fourth leading cause of work losses, resulting in nearly 15 million missed or
"reduced productivity" workdays each year. Among children, asthma is the leading cause of
school absences from a chronic illness, resulting in an annual loss of more than 14 million school
days per year.
Atopic dermatitis is one of the most common skin diseases, particularly in infants and
children—10 to 15% of the population is affected during childhood, and there is considerable
evidence that the prevalence is increasing. The direct medical costs associated with atopic
dermatitis are estimated to be $1.2-$5.9 billion (in 2005 dollars) per annum.
The cost of illness studies for allergic rhinitis, asthma, and atopic dermatitis that
contributed to this review were all based on United States data, but varied with respect to scope,
perspective, reference year, cost components, and, in some cases, the definition of disease. None
of the studies used the incidence approach or applied discount rates to the stream of lost earnings
over future years.
7-3
-------
7-4
-------
REFERENCES
Adinoff, AD; Tellez, P; Clark, RA. (1988) Atopic dermatitis and aeroallergen contact sensitivity. J Allergy Clin
Immunol81(4):736-742.
Ahlholm, JU; Helander, ML; Savolainen, J. (1998) Genetic and environmental factors affecting the allergenicity of
birch (Betula pubescens ssp. czerepanovii [Orl.] Hamet-ahti) pollen. Clin Exp Allergy 28(1 1): 1384-1388.
AAAAI (American Academy of Allergy, Asthma & Immunology). (1996-2006) Allergy statistics. [Cited 2006
April 2006]. Available fromhttp://www.aaaai.org/media/resources/media_kit/allergy_statistics.stm.
AAAAI (American Academy of Allergy, Asthma & Immunology). (2002) Common Outdoor Allergens. [Cited 2006
April 11]. Available from: http://www.aaaai.org/nab/index.cfm?p=common_outdoor_allergens.
AAAAI (American Academy of Allergy, Asthma & Immunology). (1996-2005) The allergy report: science based
findings on the diagnosis & treatment of allergic disorders. [Cited. Available from
http://www.theallergyreport.com/reportindex.html.
American College of Allergy. (2006) Five most troublesome allergens. [Cited 2006 April 12]. Available from
http://www.acaai.org/public/facts/5allergens.htm.
Andersson, M; et al. (2003) Natural exposure to Alternaria spores induces allergic rhinitis symptoms in sensitized
children. Pediatric Allergy and Immunology 14(2): 100-105.
Arbes, S. J., et al. Prevalences of positive skin test responses to 10 common allergens in the US population: results
from the third National Health and Nutrition Examination Survey. J Allergy Clin Immunol, 2005. 1 16(2): p.
377-83.
Asero, R., 2002: Birch and ragweed pollinosis north of Milan: a model to investigate the effects of exposure to
"new" airborne allergens. Allergy, 57, 1063-6
Asthma and Allergy Foundation of America, A. 2000. The Costs of Asthma. [Cited. Available from:
http://www.aafa.org/display.cfm?id=6&sub=63
Baraniuk, J., E. Meltzer, and S. Spector. Impact of allergic rhinitis and related airway disorders. J Respir Dis, 1996.
17(Suppl.): p. 511-23.
Beggs, P. J.. Pollen and pollen antigen as triggers of asthma: what to measure. Atmospheric Environment. 1998.
32(10): 1777-1783.
Beggs, P. J. Impacts of climate change on aeroallergens: past and future. Clin Exp Allergy, 2004. 34(10): p.
1507-13.
Beggs, P. J. and H. J. Bambrick. Is the global rise of asthma an early impact of anthropogenic climate change?
Environ Health Perspect, 2005. 113(8): p. 915-9.
Bennett, J. W. The Molds of Katrina. New York Academy of Sciences, 2006. January/February 2006: p. 6-9.
Bernard, S. M., et al. The potential impacts of climate variability and change on air pollution-related health effects in
the United States. Environ Health Perspect, 2001. 109 Suppl 2: p. 199-209.
Birnbaum, H. G., et al. Direct and indirect costs of asthma to an employer. J Allergy Clin Immunol, 2002. 109(2): p.
264-70.
R-l
-------
Blaiss, M. S. Cognitive, social, and economic costs of allergic rhinitis. Allergy Asthma Proc, 2000. 21(1): p. 7-13.
Bombardier, C. and J. Eisenberg. Looking into the crystal ball: can we estimate the lifetime cost of rheumatoid
arthritis? J Rheumatol, 1985. 12(2): p. 201-4.
Bousquet, J., et al. The public health implications of asthma. Bull World Health Organ, 2005. 83(7): p. 548-54.
Breton MC, Garneau M, Fortier I, Guay F, Louis J. 2006. Relationship between climate, pollen concentrations of
Ambrosia and medical consultations for allergic rhinitis in Montreal, 1994-2002. Sci Total Environ, 370(1): 39-50.
Burge, H. A. An update on pollen and fungal spore aerobiology. J Allergy Clin Immunol, 2002. 110(4): p. 544-52.
Burr, M. L. Grass pollen: trends and projections. ClinExp Allergy, 1999. 29(6): p. 735-8.
Bush, R. K. and J. J. Prochnau. Alternaria-induced asthma. J Allergy Clin Immunol, 2004. 113(2): p. 227-34.
Carroll, C. L., et al. The burden of atopic dermatitis: impact on the patient, family, and society. Pediatr Dermatol,
2005. 22(3): p. 192-9.
Cecchi, L., M. Morabito, M. Paola Domeneghetti, A. Crisci, M. Onorari and S. Orlandini, 2006: Long distance
transport of ragweed pollen as a potential cause of allergy in central Italy. Ann. Allergy. Asthma. Immunol., 96,
86-91
Cisternas, M. G., et al. A comprehensive study of the direct and indirect costs of adult asthma. J Allergy Clin
Immunol, 2003. 111(6): p. 1212-8.
Clark RAF and AD Adinoff. Aeroallergen contact can exacerbate atopic dermatitis patch tests as diagnostic tests.
Journal of American Academy of Dermatology. 1989. 21(2)863-869
Clot, B. Trends in airborne pollen: An overview of 21 years of data in Neuchatel (Switzerland). Aerobiologia, 2003.
19(3-4): p. 227-234.
Corden, J. M. and W. M. Millington. The long-term trends and seasonal variation of the aeroallergen Alternaria in
Derby, UK. Aerobiologia, 2001. 17: p. 127-136.
Corden, J.M., W.M. Millington and J. Mullins, 2003: Long-term trends and regional variation in the aeroallergens in
Cardiff and Derby UK. Are differences in climate and cereal production having an effect? Aerobiologia, 19, 191
Comtois, P and L Gagnon. Pollen concentration and frequency ofpollinosis symptoms. French Review of Allergies
1988. 28(4): 279-286
Crystal-Peters, J., et al. The cost of productivity losses associated with allergic rhinitis. Am J Manag Care, 2000.
6(3): p. 373-8.
Custovic, A., et al. Controlling indoor allergens. Annals of Allergy, Asthma and Immunology, 2002. 88(5): p. 432 -
442.
Dales, R. E., et al. The role of fungal spores in thunderstorm asthma. Chest, 2003. 123(3): p. 745-50.
D'Amato G, Liccardi G, D'Amato M, Cazzola M. Outdoor air pollution, climatic changes and allergic bronchial
asthma. Eur Respir J 2002; 20:763-76.
D'Amato G and Cecchi L. 2008. Effects of climate change on environmental factors in respiratory allergic diseases.
Clinical and Experimental Allergy, 1-11. doi: 10.1111/j.l365-2222.2008.03033.x
R-2
-------
Davis, M. B. and R. G. Shaw. Range Shifts and Adaptive Responses to Quaternary Climate Change. Science, 2001.
292: p. 673-679.
Dvorin, D. I, etal. A comparative, volumetric survey of airborne pollen in Philadelphia, Pennsylvania (1991-1997)
and Cherry Hill, New Jersey (1995-1997). Ann Allergy Asthma Immunol, 2001. 87(5): p. 394-404.
Ellis, C. N., et al. Cost of atopic dermatitis and eczema in the United States. J Am Acad Dermatol, 2002. 46(3): p.
361-70.
Ellis, C. N., et al. Validation of expert opinion in identifying comorbidities associated with atopic dermatitis/eczema.
Pharmacoeconomics, 2003. 21(12): p. 875-83.
Emberlin, J. The effects of patterns in climate and pollen abundance on allergy. Allergy, 1994. 49(18 Suppl): p.
15-20.
Emberlin, J., et al. Responses in the start of Betula (birch) pollen seasons to recent changes in spring temperatures
across Europe. Int J Biometeorol, 2002. 46(4): p. 159-70.
Emberlin, J., et al. Regional variations in grass pollen seasons in the UK, long-term trends and forecast models. Clin
Exp Allergy, 1999. 29(3): p. 347-56.
Epton, M. J., et al. Climate and aeroallergen levels in asthma: a 12 month prospective study. Thorax, 1997. 52(6): p.
528-34.
Erbas B, Chang JH, Charmage S, Ong EK, Hyndman R, Newbigin E, Abramson M. 2007. Do levels of airborne
grass pollen influence asthma hospital admissions? Clin Exp Allergy, 37(11): 1641-7.
Farquhar et al. Cost estimates for environmentally related asthma, in Research in Human Capital and Development
(Vol. 12), A. Sorkinand I. Farquhar, Editors. 1998. p. 35-46.
Fivenson, D., et al. The Effect of Atopic Dermatitis on Total Burden of Illness and Quality of Life on Adults and
Children in a Large Managed Care Organization. J Managed Care Pharm, 2002. 8(5): p. 333-42.
Frenguelli, G. Interactions between climatic changes and allergic plants. Monaldi Arch Chest Dis, 2002. 57(2): p.
141-3.
Freye, K., Litwin, C.M. Variations of Pollen and Mold Concentrations in 1998 during the Strong El Nino Event of
1997-1998 and Their Impact on Clinical Exacerbations of Allergic Rhinitis, Asthma, and Sinusitis. Allergy and
Asthma Proceedings, 2001. 22(4): p. 239-247(9).
Galan, C., et al. Heat requirement for the onset of the Olea europaea L. pollen season in several sites in Andalusia
and the effect of the expected future climate change. Int J Biometeorol, 2005. 49(3): p. 184-8.
Galant, S., et al. Prevalence of Sensitization to Aeroallergens in California Patients with Respiratory Allergy. Annals
of Allergy, Asthma and Immunology, 1998. 81(3): p. 203-210.
Garcia-Mozo H, Galan C, Jato V, Belmonte J, de la Guardia C, Fernandez D, Gutierrez M, Aira M, Roure J, Ruiz L,
Trigo M, Dominguez-Vilches E. 2006. Quercus pollen season dynamics in the Iberian Peninsula: response to
meteorological parameters and possible consequences of climate change. Ann Agric Environ Med., 13(2): 209-24.
Gergen, P. J. Understanding the economic burden of asthma. J Allergy Clin Immunol, 2001. 107(5 Suppl): p.
S445-8.
Gilmour, M. I., et al. How exposure to environmental tobacco smoke, outdoor air pollutants, and increased pollen
burdens influences the incidence of asthma. Environ Health Perspect, 2006. 114(4): p. 627-33.
R-3
-------
Glassheim, J. W., et al. Analysis of meteorologic variables and seasonal aeroallergen pollen counts in Denver,
Colorado. Ann Allergy Asthma Immunol, 1995. 75(2): p. 149-56.
Gonzalez Minero, F. J., et al. Airborne grass (Poaceae) pollen in southern Spain. Results of a 10-year study
(1987-96). Allergy, 1998. 53(3): p. 266-74.
Green, B. J., et al. Atmospheric Poaceae pollen frequencies and associations with meteorological parameters in
Brisbane, Australia: a 5-year record, 1994-1999. Int J Biometeorol, 2004. 48(4): p. 172-8.
Griffin DW, Garrison VH, Herman JR, & Shinn EA. 2001. African desert dust in the Caribbean atmosphere:
microbiology and public health. Aerobiologia, 17, 203-213.
Griffin DW, Westphal DL, Gray MA. 2006. Airborne microorganisms in the African desert dust corridor over the
mid-Atlantic ridge, Ocean Drilling Program. Aerobiologia 22: 211-226.
Halonen, M, et al. Alternaria as a major allergen for asthma in children raised in a desert environment. Am J Respir
Crit Care Med, 1997. 155(4): p. 1356-61.
Halpern, M. T., et al. Allergic rhinitis: a potential cause of increased asthma medication use, costs, and morbidity. J
Asthma, 2004. 41(1): p. 117-26.
Hamilton, R. G. Assessment of indoor allergen exposure. Curr Allergy Asthma Rep, 2005. 5(5): p. 394-401.
Hamilton, R. G. and P. A. Eggleston. Environmental allergen analyses. Methods, 1997. 13(1): p. 53-60.
Heguy L, Garneau M, Goldberg MS, Raphoz M, Guay F, Valois MF. 2008. Associations between grass and weed
pollen and emergency department visits for asthma among children in Montreal. Environ Res. 106(2): 203-11.
Henderson, C. E., et al. Projecting asthma severity from allergic sensitivity to cockroaches in pregnant inner city
women. J Reprod Med, 2000. 45(4): p. 341-4.
Hollins, P. D., et al. Relationships between airborne fungal spore concentration of Cladosporium and the summer
climate at two sites in Britain. Int J Biometeorol, 2004. 48(3): p. 137-41.
Horner, WE, Helbing, A, Salvaggio, JE, and Lehrer SB. 1995. Fungal Allergens. Clinical Microbiology Reviews
8(2): 161-179.
Husar, R. B., et al. Asian dust events of April 1998. Journal of Geophysical Research, 2001. 106(D12): p.
18,317-18,330.
IPCC, 2007: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen,
M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge, United
Kingdom and New York, NY, USA, 996 pp..
Jablonski LM, X Wang and PS Curtis. Plant reproduction under elevated CO2 conditions: a meta-analysis of reports
on 79 crop and wild species. 2002. New Phyologist. 156: 9-26.
Joint Task Force on Practice Parameters, American Academy of Allergy, Asthma and Immunology. 2003. The
Major Clinically Relevant Aeroallergens of North America. Updated: Last Update January. [Cited Access 2003
2006]. Available from: http://www.jcaai.org/PP/images/lTT2.gif
Joyce, L., et al. Potential Consequences of Climate Variability and Change for the Forests of the United States, in
Climate Change Impacts on the United States: The Potential Consequences of Climate Variability and Change,
United States Global Change Research Program, Editor. 2001, Cambridge University Press: Cambridge.
R-4
-------
Katial, R. K., et al. Atmospheric mold spore counts in relation to meteorological parameters. Int J Biometeorol,
1997. 41(1): p. 17-22.
Kemp, AS. Relationship between the time of birth and the development of immediate hyper sensitivity to grass
pollen antigens. Medical Journal of Australia. 1979. 1(7)263-264
Kessler, R., et al. Pollen and mold exposure impairs the work performance of employees with allergic rhinitis. Ann
Allergy Asthma Immunol., 2001. 87(4): p. 289-95.
Kirschstein, R. Disease Specific Estimates of Direct and Indirect Costs of Illness and NIH Support. Fiscal Year 2000
Update, N. US DHHS, Editor. 2000.
Knapp PA. and Soule P T. 1998. Recent Juniperus occidentalis (western juniper) expansion on a protected site in
central Oregon. Global change Biology, 4: 347-357.
Knowlton K, Rotkin-Ellman M, Solomon G (2007) Sneezing and Wheezing: How Global Warming Could Increase
Ragweed Allergies, Air Pollution, and Asthma. Natural Resources Defense Council.
http://www.nrdc.org/globalWarming/sneezing/sneezing.pdf
Kosisky, S. E. and G. B. Carpenter. Predominant tree aeroallergens of the Washington, DC area: a six year survey
(1989-1994). Ann Allergy Asthma Immunol, 1997. 78(4): p. 381-92.
Kozma, C. M., M. K. Sadik, and M. L. Watrous. Economic Outcomes for the Treatment of Allergic Rhinitis.
PharmacoEconomics, 1996. 10(1): p. 4-13.
Lapidus, C., D. Schwarz, and P. Honig. Atopic dermatitis in children: who cares? Who pays? J Am Acad Dermatol,
1993. 28(5 Ptl): p. 699-703.
Lapidus, C. S. Role of social factors in atopic dermatitis: the US perspective. J Am Acad Dermatol, 2001. 45(1
Suppl): p. S41-3.
Law, A. W., et al. Direct costs of allergic rhinitis in the United States: estimates from the 1996 Medical Expenditure
Panel Survey. J Allergy Clin Immunol, 2003. 111(2): p. 296-300.
Lawson, J. A. and A. Senthilselvan. Asthma epidemiology: has the crisis passed? Curr Opin Pulm Med, 2005. 11(1):
p. 79-84.
Levetin, E. Effect of Climate Change on Airborne Pollen, in American Academy of Allergy, Asthma and
Immunology 57th Annual Meeting. 2001. New Orleans, LA: The Journal of Allergy and Clinical Immunology.
Levetin, E. 2006. Dr. Estellle Levetin's HomePage. [Cited 2006 May 12]. Available from: http://pollen.utulsa.edu/
Levetin, E. and P. K. Van de Water. Pollen count forecasting. Immunol Allergy Clin North Am, 2003. 23(3): p.
423-42.
Lewis, S. A., et al. Combined effects of aerobiological pollutants, chemical pollutants and meteorological conditions
on asthma admissions and A & E attendances in Derbyshire UK, 1993-96. Clin Exp Allergy, 2000. 30(12): p.
1724-32.
Lin, R. Y. and K. D. Williams. Hypersensitivity to molds in New York City in adults who have asthma. Allergy
Asthma Proc, 2003. 24(1): p. 13-8.
Linneberg, A., et al. Increasing prevalence of specific IgE to aeroallergens in an adult population: two cross-
sectional surveys 8 years apart: the Copenhagen Allergy Study. J Allergy Clin Immunol, 2000. 106(2): p. 247-52.
R-5
-------
Lo E and Levetin E. 2007. Influence of Meteorological Conditions on Early Spring Pollen in the Tulsa Atmosphere
from 1987-2006. J. Allergy Clinlmmun Vol 119 (1) S 101 (abstract).
Mackowiak, J. The health and economic impact of rhinitis. Am J Manag Care, 1997. 3: p. S8-S18.
Malone, D., et al. A cost of illness study of allergic rhinitis in the United States. J Allergy Clin Immunol, 1997. 99(1
Pt 1): p. 22-7.
Mannino, D. M, et al. Surveillance for asthma-United States, 1980-1999. MMWR Surveill Summ, 2002. 51(1): p.
1-13.
Marion, R., T. Creer, and R. Reynolds. Direct and indirect costs associated with the management of childhood
asthma. Ann Allergy, 1985. 54(1): p. 31-4.
McMenamin, P. Costs of hay fever in the United States in 1990. Ann Allergy, 1994. 73(1): p. 35-9.
Melillo, J., et al. Chapter 2: Vegetation and Biochemical Scenarios, in Climate Change Impacts on the United States:
The Potential Consequences of Climate Variability and Change, Foundation Report. United States Global Change
Research Program, Editor. 2001, Cambridge University Press: Cambridge.
Meltzer, E. O. Quality of life in adults and children with allergic rhinitis. J Allergy Clin Immunol, 2001. 108(1
Suppl): p. S45-53.
Milton, B., et al. The social and economic consequences of childhood asthma across the lifecourse: a systematic
review. Child Care Health Dev, 2004. 30(6): p. 711-28.
Mohan, J. E., et al. Biomass and toxicity responses of poison ivy (Toxicodendron radicans) to elevated atmospheric
CO2. Proceedings of the National Academy of Sciences, 2006.
Nakagomi T, H Itaya, T Tonsinaga, M Yamaki, SH Isamata, O Nakagomi. Is atopy increasing? Lancet. 1994.
343(8889): 121-122
National Center for Health Statistics (NCHS) and C.S. Wilder. Prevalence of Selected Chronic Respiratory
Conditions, United States-1970. Vital and Health Statistics, 1973. 10(84, DHEW Publication No. (HRA) 74-1511).
National Center for Health Statistics (NCHS). Current Estimates from the National Health Interview Survey, United
States, 1982. Vital and Health Statistics, 1985. 10(150, DHHS Pub No. (PHS) 85-1578).
National Center for Health Statistics (NCHS), P. F. Adams, and V. Benson. Current Estimates from the National
Health Interview Survey, 1989. Vital and Health Statistics, 1990. 10(176, DHHS Publication No. (PHS) 90-1504).
National Center for Health Statistics (NCHS) and J.G. Collins. Prevalence of Selected Chronic Conditions, United
States, 1986-88. Vital andHealth Statistics, 1993. 10(182, DHHS Publication No, (PHS) 93-1510).
National Center for Health Statistics (NCHS), C.S. Wilder, V. Benson, and M. A. Marano. Current Estimates from
the National Health Interview Survey, 1992. Vital and Health Statistics, 1994. 10(189, DHHS Publication No, (PHS)
94-1517).
NCHS/CDC. 2001. New Asthma Estimates: Tracking Prevalence, Health Care, and Mortality. [Cited. Available
from: http://www.cdc.gov/nchs/products/pubs/pubd/hestats/asthma/asthma.htm
National Center for Health Statistics (NCHS), et al. Current Estimates of the National Health Interview Survey,
1996. Vital and Health Statistics, 1999. 10(200, DHHS Publication No. (PHS) 99-1528).
R-6
-------
National Center for Health Statistics (NCHS). 2004. Geographic Region and Division. Updated: December 16,
2004. [Cited 2006 May 9]. Available from: http://www.cdc.gov/nchs/datawh/nchsdefs/region.htm
NHLBI/Nffl (National Heart Lung and Blood Institute/Nffl). Morbidity & Mortality: 2004 Chart Book on
Cardiovascular, Lung, and Blood Diseases. 2004, Washington, DC.
Nielsen, G. D., et al. IgE-Mediated Asthma and Rhinitis I: A Role of Allergen Exposure? Pharmacology &
Toxicology, 2002. 90: p. 231-242.
O'Connell, E. J. The burden of atopy and asthma in children. Allergy, 2004. 59 Suppl 78: p. 7-11.
Patz, J., et al. The Potential Health Impacts of Climate Variability and Change for the United States: Executive
Summary of the Report of the Health Sector of the U.S. National Assessment. Environmental Heath Perspectives,
2000. 108(4): p. 368-376.
Phadia. 2002. Updated: 2002. [Cited 2008 per access to website]. Available from:
http://www.immunocapinvitrosight.com/
Phipatanakul, W. Allergic rhinoconjunctivitis: epidemiology. Immunol Allergy Clin North Am, 2005. 25(2): p.
263-81, vi.
Powell, C. The Delphi technique: myths and realities. Journal of Advanced Nursing, 2003. 41(4): p. 376-382.
Powell, R. F. andE. B. Smith. Tumbleweed Dermatitis. Arch Dermatol, 1978. 114: p. 751-754.
Prospero JM, Blades E, Mathison G. and Naidu R. 2005. Interhemispheric transport of viable fungi and bacteria
from Africa to the Caribbean with soil dust. Aerobiologia 21: 1-19.
Puc, M. and M. I. Puc. Allergic airborne grass pollen in Szczecin, Poland. Ann Agric Environ Med, 2004. 11(2): p.
237-44.
Puc, M. and T. Wolski. Betula and Populus pollen counts and meteorological conditions in Szczecin, Poland. Ann
Agric Environ Med, 2002. 9(1): p. 65-9.
Pulimood TB, Corden JM, Bryden C, Sharpies L, Nasser SM. 2007. Epidemic asthma and the role of the fungal
mold Alternaria alternata. J Allergy Clin Immunol, 120(3):610-7.
Rasmussen A. The effects of climate change on the birch pollen season in Denmark. Aerobiologia 2002; 18:253-65.
Ratard, R. and e. al. Health Concerns Associated With Mold in Water-Damaged Homes After Hurricanes Katrina
and Rita - New Orleans, Louisiana, October 2005. MMWR, 2006. 55(2): p. 41-44.
Ray, N. F., et al. Direct expenditures for the treatment of allergic rhinoconjunctivitis in 1996, including the
contributions of related airway diseases. J Allergy Clin Immunol, 1999. 103(3 Pt 1): p. 401-7.
Reed, S. D., T. A. Lee, and D. C. McCrory. The Economic Burden of Allergic Rhinitis: A Critical Evaluation of the
Literature. PharmacoEconomics, 2004. 22(6): p. 345-61.
Reiss, N. M. and S. R. Kostic. Pollen season severity and meteorologic parameters in central New Jersey. J Allergy
Clin Immunol, 1976. 57(6): p. 609-14.
Rogers, C. A., et al. Interaction of the Onset of Spring and Elevated Atmospheric CO2 on Ragweed (Ambrosia
artemisiifolia L.) Pollen Production. Environmental Health Perspectives, 2006. 114(6): p. 865-869.
Root, T. L., et al. Fingerprints of global warming on wild animals and plants. Nature, 2003. 421: p. 57-60.
R-7
-------
Ross, R. The costs of allergic rhinitis. Am J Manag Care, 1996. 2: p. 285-90.
Rybnicek, O., and S. Jaeger, 2001: Ambrosia (ragweed) in Europe. ACI International, 13, 60-66
Saint Louis County, D. o. H. 2006. Pollen and Mold Center: Elm Family (Ulmaceae Family) - ELM. Updated: 2006.
[Cited 2006 May 8]. Available from: http://www.co.st-louis.mo.us/Doh/pollen_site/TreeElm.html
Schoenwetter, W. F., et al. Economic impact and quality-of-life burden of allergic rhinitis. Curr Med Res Opin,
2004. 20(3): p. 305-17.
Selgrade, M. K., et al. Induction of Asthma and the Environment: What We Know and Need to Know.
Environmental Health Perspectives, 2006. 114(4): p. 615-619.
Singer, B. D., et al. Increasing Amb a 1 content in common ragweed (Ambrosia artemisiifolia) pollen as a function
of rising atmospheric CO2 concentration. Functional Plant Biology, 2005. 32: p. 667-670.
Smith, D. H., et al. A national estimate of the economic costs of asthma. Am J Respir Crit Care Med, 1997. 156(3 Pt
1): p. 787-93.
Sneller, M. R., H. D. Hayes, and J. L. Pinnas. Pollen changes during five decades of urbanization in Tucson,
Arizona. Ann Allergy, 1993. 71(6): p. 519-24.
Solomon GM, Hjelmroos-Koski M, Rotkin-Ellman M, Hammond SK. 2006. Airborne mold and endotoxin
concentrations in New Orleans, Louisiana, after flooding, October through November 2005. Environ Health
Perspect, 114(9): 1381-6.
Stanford, R., T. McLaughlin, and L. J. Okamoto. The cost of asthma in the emergency department and hospital. Am
J Respir Crit Care Med, 1999. 160(1): p. 211-5.
Stefanic, E., V. Kovacevic, and Z. Lazanin. Airborne ragweed pollen concentration in north-eastern Croatia and its
relationship with meteorological parameters. Ann Agric Environ Med, 2005. 12(1): p. 75-9.
Storms, W., et al. The economic impact of allergic rhinitis. Allergy Clin Immunol, 1997. 99: p. S820-4.
Sutherst, R. W. Global change and human vulnerability to vector-borne diseases. Clin Microbiol Rev, 2004. 17(1):
p. 136-73.
Taramarcaz, P., B. Lambelet, B. Clot, C. Keimer and C. Hauser, 2005: Ragweed (Ambrosia) progression and its
health risks: will Switzerland resist this invasion? Swiss med. wkly., 135, 538-48
Targonski, P. V., V. W. Persky, and V. Ramekrishnan. Effect of environmental molds on risk of death from asthma
during the pollen season. J Allergy Clin Immunol, 1995. 95(5 Pt 1): p. 955-61.
Teranishi, H., et al. Possible role of climate change in the pollen scatter of Japanese cedar Cryptomeriajaponica in
Japan, dim Res, 2000. 14: p. 65-70.
The Center for Health and the Global Environment, H. M. S. 2005. Climate Change Futures Health, Ecological and
Economic Dimensions. Updated: November 2005. [Cited 2006. Available from:
http://www.climatechangefutures.org/pdf/CCF_Report_Final_10.27.pdf.
Tidwell, J. 2006. Fall Season Allergy Triggers. [Cited 2006 May 8]. Available from:
http://allergies.about.eom/cs/fall/a/aa091399.htm.
Tortolero, S. R., et al. Environmental allergens and irritants in schools: a focus on asthma. J Sch Health, 2002. 72(1):
p. 33-8.
R-8
-------
Tripathi, A. and R. Patterson. Impact of Allergic Rhinitis Treatment of Quality of Life. PharmacoEconomics, 2001.
19(9): p. 891-99.
Troutt, C. and E. Levetin. Correlation of spring spore concentrations and meteorological conditions in Tulsa,
Oklahoma. Int J Biometeorol, 2001. 45(2): p. 64-74.
US NHLI. Respiratory diseases: Task force report on problems, research approaches, and needs. Publication No.
(NIH) 76-432, E. a. W. Department of Health, Editor. 1972, US National Institutes of Health, Washington, DC.
Van de Water, P. K., et al. An assessment of projective forecasting of Juniperus ashei pollen movement in the
Southern Great Plains, USA. Int J Biometeorol, 2003. 48(2): p. 74-82.
Vance, V. and W. Taylor. The financial cost of chronic childhood asthma. Ann Allergy, 1971. 29(9): p. 455-60.
Vazquez, L. M, C. Galan, and E. Dominguez-Vilches. Influence of meteorological parameters on Olea pollen
concentrations in Cordoba (south-western Spain). Int J Biometeorol, 2003. 48(2): p. 83-90.
Villenueve PJ, Chen L, Rowe BH, Coates F. 2007. Outdoor air pollution and emergency visits for asthma among
children and adults: a case-crossover study in northern Alberta, Canada. Environ Health, 6:40.
Voltolini, S., P. Minale, C. Troise, D. Bignardi, P. Modena, D. Arobba, and A. Negrini, 2000: Trend of herbaceous
pollen diffusion and allergic sensitisation in Genoa, Italy. Aerobiologia, 16, 245-249.
Von Hertzen, L. C. The hygiene hypothesis in the development of atopy and asthma-still a matter of controversy? Q
JMed, 1998. 91: p. 767-771.
Wan, S.Q., T. Yuan, S. Bowdish, L. Wallace, S.D. Russell and Y.Q. Luo, 2002: Response of an allergenic species
Ambrosia psilostachya (Asteraceae), to experimental warming and clipping: implications for public health. Am. J.
Botany, 89, 1843-1846
Ward, M. M., et al. Lost income and work limitations in persons with chronic respiratory disorders. J Clin
Epidemiol, 2002. 55(3): p. 260-8.
Wayne, P., et al. Production of allergic pollen by ragweed (Ambrosia artemisiifolia L.) is increased in CO2-enriched
atmospheres. Ann Allergy Asthma Immunol., 2002. 88(3): p. 279-82.
Weber, R. W. Floristic zones and aeroallergen diversity. Immunol Allergy Clin North Am, 2003a. 23(3): p. 357-69.
Weber, R. W. Meteorologic variables in aerobiology. Immunol Allergy Clin North Am, 2003b. 23(3): p. 411-22.
Weiss, K., P. Gergen, and T. Hodgson. An economic evaluation of asthma in the United States. N Engl J Med.,
1992. 326(13): p. 862-6.
Weiss, K. B. and S. D. Sullivan. The economic costs of asthma: a review and conceptual model.
Pharmacoeconomics, 1993. 4(1): p. 14-30.
Weiss, K. B. and S. D. Sullivan. The health economics of asthma and rhinitis. I. Assessing the economic impact. J
Allergy Clin Immunol, 2001. 107(1): p. 3-8.
Weiss, K. B., S. D. Sullivan, and C. S. Lyttle. Trends in the costs of asthma in the United States, 1985-1994.
Journal of Allergy and Clinical Immunology, 2000. 106(3): p. 493-499.
White, J. F. and D. I. Bernstein. Key pollen allergens in North America. Annals of Allergy, Asthma and
Immunology, 2003. 91(5): p. 425-435.
R-9
-------
White, J. F., et al. Lack of correlation between regional pollen counts and percutaneous reactivity to tree pollen
extracts in patients with seasonal allergic rhinitis. Annals of Allergy, Asthma and Immunology, 2005. 94(2): p.
240-246.
Whitmore, S. E., et al. Aeroallergen patch testing for patients presenting to contact dermatitis clinics. J Am Acad
Dermatol, 1996. 35(5 Pt 1): p. 700-4.
Wilson, J. F. Health and the Environment After Hurricane Katrina. Annals of Internal Medicine, 2006. 144(2): p.
153-156.
Wood, S. F. Review of hay fever. 1. Historical background and mechanisms. FamPract, 1986. 3(1): p. 54-63.
Ziering, RW and GL Klein. Allergic disease: how to manage three common types. Postgraduate Medicine. 1982.
72(2) 163-174.
Zillich et al. Assessment of the Relationship between Measures of Disease Severity, Quality of Life, and
Willingness to Pay in Asthma. PharmacoEconomics, 2002. 20(4): p. 257-65.
Ziska, L. H. and F. A. Caulfield. Rising carbon dioxide and pollen production of common ragweed, a known
allergy-inducing species: Implications for public health. Australian Journal of Plant Physiology, 2000. 27: p.
893-898.
Ziska, L. H., et al. Cities as harbingers of climate change: common ragweed, urbanization, and public health. J
Allergy Clinlmmunol, 2003. 111(2): p. 290-5.
Ziska LH, George K, Frenz, DA 2007. Establishment and persistence of common ragweed (Ambrosia artemisiifolia
L.) in disturbed soil as a function of an urban-rural macro-environment. Global Change Biology. 13:266-274.
R-10
------- |