Outdoor Allergens

CitationBurge, Harriet A., and Christine A. Rogers. 2000. Outdoor allergens. Environmental Health Perspectives 108(Sup 4): 653-659.

Published Versiondoi:10.2307/3454401

Permanent linkhttp://nrs.harvard.edu/urn-3:HUL.InstRepos:4569473

Terms of UseThis article was downloaded from Harvard University’s DASH repository, and is made available under the terms and conditions applicable to Other Posted Material, as set forth at http://nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of-use#LAA

Share Your StoryThe Harvard community has made this article openly available.Please share how this access benefits you. Submit a story .

Accessibility

Outdoor AllergensHarriet A. Burge and Christine A. RogersSchool of Public Health, Harvard University, Boston, Massachusetts, USA

Outdoor allergens are an important part of the exposures that lead to allergic disease. Understandingthe role of outdoor allergens requires a knowledge of the nature of outdoor allergen-bearing particles,the distributions of their source, and the nature of the aerosols (particle types, sizes, dynamics ofconcentrations). Primary sources for outdoor allergens include vascular plants (pollen, fern spores,soy dust), and fungi (spores, hyphae). Nonvascular plants, algae, and arthropods contribute smallnumbers of allergen-bearing particles. Particles are released from sources into the air by wind, rain,mechanical disturbance, or active discharge mechanisms. Once airborne, they follow the physicallaws that apply to all airborne particles. Although some outdoor allergens penetrate indoor spaces,exposure occurs mostly outdoors. Even short-term peak outdoor exposures can be important ineliciting acute symptoms. Monitoring of airborne biological particles is usually by particle impactionand microscopic examination. Centrally located monitoring stations give regional-scalemeasurements for aeroallergen levels. Evidence for the role of outdoor allergens in allergic rhinitis isstrong and is rapidly increasing for a role in asthma. Pollen and fungal spore exposures have bothbeen implicated in acute exacerbations of asthma, and sensitivity to some fungal spores predicts theexistence of asthma. Synergism and/or antagonism probably occurs with other outdoor air particlesand gases. Control involves avoidance of exposure (staying indoors, preventing entry of outdooraerosols) as well as immunotherapy, which is effective for pollen but of limited effect for spores.Outdoor allergens have been the subject of only limited studies with respect to the epidemiology ofasthma. Much remains to be studied with respect to prevalence patterns, exposure and diseaserelationships, and control. Key words: asthma, exposure, fungal spores, outdoor allergens, pollen,predictive models. - Environ Health Perspect 1 08(suppl 4):653-659 (2000).http.//ehpnet 1. niehs. nih.gov/docs/2000/suppl-4/653-659burge/abstract html

People are exposed throughout life to outdoorallergens either directly or after the allergen-bearing particles peretrate interiors. Themost widely recognized and abundant sourcesfor these outdoor allergens are pollen grainsand fungal spores (1). At least some allergen-bearing particles are present in outdoor airthroughout the world, although in very lowconcentrations during periods where snowcovers their sources (2). They also penetrateinteriors, and some outdoor fungi colonizeindoor substrates and become essentiallyindoor allergens. Pollen allergens are com-monly considered to play a role in allergicrhinitis (3), but the particle size of pollen hasbeen considered too large to penetrate thelower airways, and therefore too large to leadto asthma. However, evidence is increasingfor a relationship between exposure to pollen(4), fungal (5,6), and other airborne allergenssuch as soy (7,8) and exacerbation of asthma.

We present here a brief review of thenature and patterns of outdoor allergens andthe evidence for an association between out-door allergen exposure and allergic disease,particularly asthma.

Characteristics of OutdoorAllergensNature ofPollen and Pollen AliergensPollen grains are the male gametophyte inthe sexual reproduction of flowering plants

(angiosperms) and conifers (gymnosperms).Pollination, the transfer of pollen grains frommale to female reproductive structures, canbe accomplished via three vectors-wind,water, or animals. In wind-pollinated plants,pollen grains are released into the atmos-phere to passively find their way onto anappropriate receptive female stigma. Becausethis is a less efficient transfer than in insectpollination, anemophilous plants producecopious amounts of pollen to ensure success-ful fertilizations. In addition, the flowers ofthese plants often have no petals, and theanthers (pollen sacs) are exposed to air move-ment. Hence, pollen from anemophilousplants is the most abundant in the atmos-phere and is also the most important interms of human exposure. Anemophily is acommon strategy for plants in the temperateregions of the world, whereas tropical plantsoften produce insect-pollinated flowers.Pollen grains are usually more or less spheri-cal, at least when hydrated, with a rigid cellwall formed of a complex polysaccharide-based substance called sporopollenin. Pollengrains are identified using light microscopyby the shape and size of the grain, and itswall structure. Many grains have apertures(pores and/or furrows) that aid in identifica-tion. Some have sculptured wall surfaces, andothers have distinctive inclusions. Based onmorphological features, some grains can beassigned to very specific taxonomic categories

(e.g., Typha latifolia-broad-leafed cattail).Other less distinctive types are assigned onlyto relatively large groupings (e.g., grasspollen). The size range within a genus is typi-cally small and can often be used as a diag-nostic feature (9). Most airborne pollengrains are 15-50 pm in diameter, althoughthe overall range for pollen may be as broadas 10-100 pim.

Isolation of pollen allergens has shownthat they are typically low-molecular-weightproteins or glycoproteins (5-60 kDa) that arereleased quickly upon contact with aqueoussolutions (10). Speculations on the functionof these proteins include cell recognition fac-tors, enzymes involved in pollen germination,or reserve storage proteins for pollen-tubegrowth (10). However, only sparse evidenceexists that suggests these proteins can beinvolved in recognition systems for incompat-ibility responses within and between plants(11), or that they have enzymatic activity(12-14). Therefore, the functional role ofpollen allergens in the plant has still not beenclearly established.

The potency of pollen allergens is notsimply a matter of protein abundance as, forexample, comparable amounts of two aller-gens in rye grass pollen produce widely differ-ing allergenicities based on radioallergosorbenttest (RAST) inhibition (15). Hence, structuraland/or compositional differences occur thatconfer allergenicity (16). In addition, a con-siderable degree of cross-reactivity of allergensoccurs between taxa (17-19). Immunogoldlabeling experiments have localized allergenson or as part of the pollen grain wall (exine)and in the cytoplasm (20). A moderateamount of allergen is also found associatedwith apertures.

Pollen allergens have been recovered fromsmall particle fractions of outdoor air (inde-pendent of the respective pollen grains).Schappi et al. (21) recovered birch pollenallergen 1.2 ng Bet v 1/m3, equivalent to 200birch pollen grains (typically 25 pm in dia-meter) from the particle fraction less than

This article is part of the monograph on Environmentaland Occupational Lung Diseases.

Address correspondence to H.A. Burge,Environmental Health, Harvard School of Public Health,665 Huntington Ave., Boston, MA 02115 USA.Telephone: (617) 432-4638. Fax: (617) 432-3349.E-mail: hburge@hsph.harvard.edu

Received 15 December 1999; accepted 24 May2000.

Environmental Health Perspectives * Vol 108, Supplement 4 * August 2000 653

BURGE AND ROGERS

7.5 pm. Ragweed allergens have also beenrecovered from small particles (22-24). Grasspollen allergen (Lol p 5) has been recoveredon particles < 5 pm during rainfall, and rup-tured pollen grains are often seen on air sam-ple slides. Grass allergens have been measuredin fine-partide aerosols and attached to starchgrains and combustion (diesel exhaust) parti-cles (25,26). The role of pollen-derived aller-gens associated with small particle fractionsremains speculative. However, the ability ofthese particles to penetrate the lower airwaysmay play a role in asthma exacerbations. Inaddition, longer airborne residence times ofsmaller partides, due to lower settling veloci-ties, potentially increase the risk of exposureto the associated allergens.

Pollen Prevalence PatternsPollen source distributions. Distribution ofpollen-producing plants is naturally a resultof floristic patterns. For example, the north-ern boreal forests produce large amounts ofpine, spruce, hemlock, and birch pollen.Oak-hickory forests cover large areas in theUnited States, and maples are abundant inthe East. Some of the Southwestern moun-tains support forests of mountain cedar thatshed copious amounts of pollen. Grass is animportant pollen source throughout theworld. Ragweed, long the dominant pollenallergen source in the Midwestern UnitedStates, is increasing in importance in manyparts of the world, especially in EasternEurope (27).

Landscaping has significantly changed theair biota in many parts of the world. The his-torical planting of elm trees throughout east-ern U.S. cities probably led to early springpollen peaks in these population centers.Dutch elm disease has done much to remedi-ate this problem. In the Southwestern UnitedStates, irrigation of the desert and planting oflawns and street trees has destroyed what wasonce a refuge for allergy sufferers. In particu-lar, mulberry trees have been widely plantedand have become major allergen sources.Their planting has been outlawed in somecommunities (28).

Pollen production and release. The pres-ence of pollen in the air depends on the abun-dance of source vegetation, and factorscontrolling release and dispersal. Many differ-ent factors affect production of pollen, and thesubject cannot be treated in depth here. A fewespecially important aspects of this interestingfield are presented. The reader is referred toWodehouse (29) for further information.

Plants produce pollen in response tointernal, genetically controlled cyclesimpacted by environmental factors such astemperature, available moisture, and light. Intemperate regions of the world, most treesproduce pollen in the spring following winter

dormancy. Some trees (e.g., birch) producereproductive structures (catkins) in the fall,and the pollen matures in the spring. Others(e.g., maple) form flowers in the spring. Dueto the need to coordinate reproductive efforts,particularly in wind-pollinated plants, pollenrelease is signaled by unambiguous environ-mental cues. To maximize the likelihood ofsuccessful fertilizations, flowering periods oftree types are typically short (1-2 weeks) butintense. In addition, some tree taxa (e.g.,birch, pine, beech) exercise mast cycling inreproduction, where a particularly bountifulyear in pollen production and seed set is fol-lowed by 1 or 2 years of gready reduced pro-duction. This can present particularly badyears for allergy sufferers sensitive to thepollen of these taxa.

All annual plants and perennials that dieback each year require several months for veg-etative structures to grow before the plants aremature enough to produce pollen. For exam-ple, in cold climates grass seeds germinateearly, but several months (April-June) arerequired for growth before flowering in thesummer. Flowering and pollen seasons forsome grasses and many herbaceous taxa oftenlast 1 month or more.

Nature ofFungl Spores and TheirAllergensFungi are saprobic or (more rarely) parasiticorganisms that occupy a kingdom of theirown. They are responsible for most of theaerobic decay of plant materials (e.g., deadgrass, leaves, etc.) and are present in airthroughout the world, often as the dominantbiological component (2,30).

The fungal cell is eukaryotic, containingwell-developed membrane systems (indudingmitochondria) and one to several nuclei. Therigid cell wall is composed of acetyl glu-cosamine polymers (chitin) and j-glucans, ormannans. The wall also often contains waxes(dry spores), and most are coated with extra-cellular polysaccharides.

Fungi are identified primarily by themethod of spore production, including thenature of the spore production process andthe morphology of the spores (31,32). Fungalspores may be colorless to nearly black, withbrown melanin pigments commonly present.Each spore may include one to many cellsarranged in lengthwise chains or in two orthree-dimensional arrays (33). Some fungalspores can be assigned to a genus and specieson the basis of their morphology, with noother reproductive information (e.g.,Epicoccum nigrum). Most, however, can cur-rently be assigned only to larger categories(i.e., Claosporium, the Penicillium/Aspergillusgroup, and basidiospores), although effortsare underway to expand the number of differ-entiable spore categories (34).

Although airborne fungal spores canrange in diameter from < 2 to > 50 pm, mostfall into the range of 2-10 pm and thus read-ily penetrate into the lower airways. However,some very common types implicated inasthma are pollen sized (e.g., E. nigrum). Fewfungal spores are spherical, and some haverather asymmetrical shapes for which aerody-namic diameters are difficult to estimate.

The current state of knowledge of theisolation and characterization of fungal aller-gens was recently reviewed by Horner et al.(35). Advances in the characterization of fun-gal allergens has been hampered by produc-tion and stability variations under differingconditions, the difficulty in standardizingextraction and isolation protocols, and theoverall enormous diversity of fungi. Morethan 80 fungal genera have been associatedwith allergic respiratory symptoms. However,few fungal allergens have been characterizedto date. The allergens from A. fumigatus,C herbarum, and Alternaria alternata havebeen the best characterized. Most isolatedallergens have been found to be proteins orglycoproteins with molecular weights of 6-90kDa. Although in some cases carbohydrateportions of fungal extracts show allergenicactivity, most IgE-binding activity is associ-ated with the protein component. Often sev-eral allergens (up to 20) are detected in eachof the fungal types examined. This empha-sizes that exposure involves complex allergenmixtures. In addition, cross-reactivity is com-mon among phlyogenetically related taxa.Shared allergenic and antigenic epitopes havebeen found among various ascomycetes. Themolecular, biochemical, and functional char-acterizations of fungal allergens are enormousendeavors, and progress has been slow.Application of molecular biological tech-niques should rapidly expand our knowledgein this field.

Fungl Prvalence PattnsSpore source distributions. Floristic patternsalso play some role in the types of fungalspores that enter the air. Many fungi groweither on or in dose association with specificplant hosts. For example, many mushroomspecies are associated with very specific treespecies, and in the absence of the tree, themushroom is not found. Likewise, the rustfungi can be highly host specific. Thecedar/apple rust, for example, only occurswhere cedar (juniper) trees occur nearmembers of the apple family.

Agricultural practices have significantlychanged floristic patterns throughout theworld. Much of the central United States, forexample, is covered with field crops through-out most of the growing season. The cropsthemselves are not considered major allergensources, but they support massive populations

Environmental Health Perspectives * Vol 108, Supplement 4 * August 2000654

OUTDOOR ALLERGENS

of fungi that have come to dominate the airspora, especially during the harvest season(36). Grain storage and handling are otheragricultural practices that may introduce verylarge concentrations of specific spore typesinto the air locally (37,38).

Disposal of organic waste is an increasingconcern in the world. Dumping in water sup-plies and the ocean is no longer an ecologi-cally sound practice. Incineration requiresenergy and contributes to air pollution withcombustion products. Composting is aprocess that uses microorganisms to reduceorganic waste to (essentially) carbon dioxideand water. Sewage waste is mixed with woodchips and stored in piles that are periodicallyturned to provide oxygen to the microorgan-isms using the mixture for food. Celluloseand wood-decaying fungi and bacteria playan especially important role in this type ofcomposting. A succession of organisms occu-pies the piles. Many of these produce sporesthat become airborne in enormous numberswhen the piles are moved. Among the spore-forming organisms occupying compost areAspergillus fumigatus, several mushroomspecies, and the thermophilic actinomycetes(bacteria that produce airborne spores).Exposure to any of these can lead to asthmaor hypersensitivity pneumonitis, and A. fimi-gatus is a well-recognized human pathogen.Very little research exists that can guide deci-sions regarding citing of composting facilitiesand handling of the compost to control expo-sure of surrounding occupants. Occupationalexposures, of course, are also of concern (3y),but exposure control probably would have toindude personal respiratory protection.

Spore production and release. Productionof fungal spores is also controlled by internalfactors impacted by the environment. Manyfungi have seasonal patterns of spore produc-tion coinciding with availability of host mate-rial to colonize. Thus, many of the plantpathogens (e.g., the leaf spot fungi such asVenturia) produce spores in the spring whenplants are young and vulnerable. Those thatdecay dead plant material (e.g., the commonspecies of Penicillium and Aspergillus) are pro-duced in response to temperature and mois-ture conditions. Many mushrooms require aseason of mycelial growth to accumulateenergy for fruiting-body production.

Fungal spores are released from the spore-bearing cells either by active or passive mech-anisms. The active mechanisms all depend onchanges in moisture conditions. Ascosporesand basidiospores are released as the spore-bearing cell absorbs water, either duringrainfall or as humidity increases. Some dry-weather spores (e.g., Cladosporium) areshaken loose as the spore-bearing cell twists asit dries. In other cases, air movements aloneare sufficient to cause release of spores (2,30).

Other Allergen-Bearing PartidesOther organisms produce airborne spores thatcan be locally abundant. Algae and lowerplants (mosses, liverworts, club mosses, andferns) also release reproductive units (spores)into the air. Lichens are formed of algae andfungi in a symbiotic relationship. Theyrelease typical fungal spores as well as algalcells that may be entwined with fungalhyphae.

Any living organism can release fragmentsinto the air. Humans, for example, are con-stantly releasing skin scales that can be foundabundantly in indoor air, but also are surelypresent outdoors. Arthropod fragments arenoticeable components of the outdoorbioaerosol. Some occupational activities causethe release of masses of biological particles.The classic example of this aerosol is the soy-bean allergen clouds identified in Barcelona,Spain, during unloading of container ships(40,41). They caused 26 outbreaks of asthmainvolving 687 subjects and resulting in 1,155emergency room admissions. Remediation,involving the installation of bag filters in silos,reduced outdoor soy allergen levels signifi-cantly, and asthma outbreaks disappeared.However, half of the patients remained sensi-tized to soy, and no further improvement inasthma episodes was seen after the initial 2years (42). Weak associations between soy-bean unloading and asthma outbreaks havealso been reported for Valencia and Coruna,Spain (43). Soybean dust inhalation allergyhas also been reported in a child who playedwith a soybean-filled beanbag (44).

Transport and RemovalAirborne allergen-bearing particles follow thesame physical rules as any particle of the sameaerodynamic diameter. They are dispersed viaair movements and settle and impact in rela-tion to their aerodynamic diameter, availableimpaction surfaces, and factors such as rainthat enhance removal.

Spores are usually released by air movementwithin a laminar boundary layer surround-ing their sources. Many remain in the layerand eventually settle near the source. Manydispersion models predict that the majorityof particles of the size of pollen and sporeswill be deposited close to the source (< 100m) (2,45). Others are carried aloft with tur-bulence and may be transported with windfor long distances. Wind gusts may be espe-cially important in dislodging spores fromsurfaces, either by direct sweeping of the sur-faces or by causing adjacent surfaces to rubtogether (46). Wind and gusts also affectremoval by bringing the spores nearimpaction surfaces and by increasing theirinertia so that impaction occurs. Impactionis related to the aerodynamic particle diame-ter (as for nonbiological particles). Most

pollen and spores have a density near unity,so that this diameter is primarily dependenton the shape and size of the spore. Manyspores are hygroscopic, and aerodynamicdiameter may increase with increasing humid-ity. Pollen diameters are also probably affectedby humidity, traveling as collapsed units whendry and as inflated cells when moist. Theextent of this humidity effect has not beenreported. Most deposition occurs on narrowsurfaces such as leaf edges or thin fibers (47).

Rainfall is well known to cause release ofspores by splash and by so-called tap and puffmechanisms (48). Rain also removes particlesfrom the air by both rainout and washouteffects. Rainout involves spores acting as con-densation nuclei and falling with the resultantdroplet. In washout, raindrops capture sporesand pollen as they fall. Frontal rains are moreefficient at capturing particles than long driz-zle (49). Because rainfall both disperses andremoves spores, it is difficult to predict air-borne spore concentrations during rainfall.During long gentle rains, release mechanismsstrongly exceed washout, often leading tomuch higher spore concentrations during rainthan on sunny days without rain. However,the spores released during rain are differentfrom those that were in the air before rainfallbegan, hence there exists qualitatively distinctwet-air spora and dry-air spora.

Predictive ModeingBecause the presence and abundance ofoutdoor allergens cannot be controlled,avoidance of outdoor allergen exposure is themajor strategy for allergic individuals.However, this requires adequate forewarningof potentially high aeroallergen levels. Severalinvestigators have published models designedto forecast concentrations of specific allergen-bearing particles. Arizmendi et al. (50), usedneural network technology on historical 2-hrtotal pollen concentrations, without account-ing for the effects of meteorological variables,to forecast near-future concentrations.Stephen et al. (51) combined an estimateddiurnal rhythm of spore concentrations witha one-parameter time-series model that pro-vided good short-term forecasts up to 24 hrin advance. The one-step (2-hr) predictionerror variance was reduced by 88% forCladosporium and by 98% for basidiospores.Moseholm et al. (52) forecast grass pollenconcentrations using time-series analysis andregression using meteorological parameters.They found high predictive capability using a2-day lag in temperature. Stark et al. (53)used Poisson regression models to forecastragweed pollen concentrations. Seasonalitywas modeled by day of the year and the influ-ence of the weather was incorporated throughanalysis of temperature trends, wind, andrainfall. They accurately forecast 79% of the

Environmental Health Perspectives * Vol 108, Supplement 4 * August 2000 655

BURGE AND ROGERS

days into the appropriate high, moderate, orlow categories set by the American AcademyofAllergy Asthma and Immunology (AAAAI)Aeroallergen Monitoring Network. Norris-Hill (54) used accumulated weather variablesto model seasonality of airborne grass pollenand a multiple regression model with maxi-mum temperature, relative humidity, andrainfall to forecast daily values. The modelexplained 59% of the variation in the datawith a forecast accuracy of ±25% that wasachieved on 71% of the days.

Although forecasting models continue toimprove, published models fail at least 25%of the time and are available for a very fewparticle types. Also, none predict the magni-tude of any pollen or spore season. However,because both biological and environmentalfactors affect prevalence, once the spatio-temporal scale and magnitude of these rela-tionships are determined, these types ofmodels can be developed and a much higherlevel of accuracy can be expected.

ExposureParticle size considerations. The mechanismwhereby large particles cause asthma remainsspeculative. Pollen and large fungal spores areinhalable (i.e., they penetrate into the respira-tory tract) but are considered too large to berespirable (i.e., penetrating into the lower res-piratory tract). However, Michel et al. (55)demonstrated that some pollen grains do pen-etrate the distal lower airways. In addition,many fungal spores are well within the res-pirable size range, and pollen allergens (asmentioned above) have been shown to bepresent in outdoor air on particles smallerthan intact pollen (22,23,25,56).

Personal exposure. At this point, little ifany data exist on personal exposure to outdoorallergens. The methods usually used for out-door monitoring involve rooftop collections atcentral sites. Indoor exposures are rarely ana-lyzed with respect to the contributions of theoutdoor aerosol, and the effects of the intenseexposures that can be found locally at thebreathing zone have not been well studied.Gautrin et al. (57) have studied personal expo-sure in lawn cutters. They conclude that theseworkers are more heavily exposed to fungalspores than the general population and thatsuch exposure results in an increased rate ofsensitization to fungi. There are likely to besimilar differences in exposure for other occu-pational and leisure categories, and personalexposure measures are essential for evaluatingthe relationship between exposure and diseasein individuals. In epidemiological studies, per-sonal exposure monitoring would improve therepresentativeness of exposure measures, butmethods currently are unavailable that canreasonably be used in this context (seemonitoring methods section below).

Indoor exposure. It is well recognized thatpeople in developed countries spend most oftheir time indoors (58). Outdoor allergens dopenetrate indoor environments. In fact,indoor spore concentrations in uncontami-nated environments closely parallel those out-doors. In naturally ventilated environments,concentrations are also similar. Where pene-tration barriers exist, indoor levels are usuallymuch lower than those outdoors (555. Pollenallergens also are found indoors (60).

Models that estimate overall averageexposures use factors to account for bothindoor and outdoor exposures, althoughthese have yet to be applied to outdoorspores and pollen (61). However, thisapproach is only appropriate if average, orcumulative, exposure is the important para-meter. With allergen exposure, it is likelythat short-term peaks in exposure exacerbatesymptoms. Thus, during the ragweed season,the 10% of time spent outdoors in the rela-tively intense aerosol may be sufficient toinduce serious symptoms, which may persistduring the 90% of time spent indoors withlow exposure.

Monitoring for outdoor allergens.Microscopic identification and counting ofallergen-bearing particles from either rotat-ing arm impactors (pollen) or from suctionspore traps located generally on urban/sub-urban rooftops has been the standardmethod for assessing outdoor allergen expo-sure (62,63). Although such methods con-tinue to be useful for assessing exposure torecognizable allergen-bearing particles,spores and pollen are only indicators of thepresence of allergens. Recognizable particlesof different kinds may contain similar aller-gens, all particles of a particular type maynot contain the same amount of any specificallergen, and as previously mentioned someallergens may not be readily associated withan identifiable particle and may be present asunrecognizable and possibly very small parti-cles. In addition, suction spore traps, whilemore efficient than rotating arm impactors,collect fewer than 50% of particles smallerthan about 5 pm, thereby providing littleinformation about many small, potentiallyimportant fungal spores.

The collection of high-volume samples onfilters, which have high collection efficiencieswell below the smallest fungal spore, is ofincreasing interest. Immunochemical meth-ods can be used for analysis of these samples,and these methods have been compared topollen or spore counting by others.Specifically, Johnsen et al. (64) found astrong correlation between pollen counts andimmunochemical measures of birch, grass,and mugwort allergens. D'Amato et al. (4)report similar results for Parietaria judaica.Agarwal et al. (65), used immunochemical

measures on high-volume air samples forAlt a 1, one of the principal allergens ofAlternaria. However, immunochemical assaysare restricted to well-characterized allergensfor which specific antibodies are available,and do not reveal exposure to unknown aller-gens or to allergen mixtures.

Methods for outdoor allergen monitoringare currently restricted to central site moni-toring, although interest is increasing in per-sonal monitoring. Poulos et al. (66) havedeveloped a nasal sampler that showspromise, and other studies have investigatedthe utility of portable filtration devices (67).

Outdoor Allergens and AsthmaExisting studies that have sought to relateoutdoor allergen exposure and asthma gener-ally fall into two categories: studies of the asso-ciation between skin test sensitivity and thepresence of asthma, and studies comparingsymptom data and exposure.

Sensitivity and AsthmaPollart et al. (60), reported that elevated IgEto grass allergens was associated with emer-gency room visits for asthma [X2 = 69;p < 0.0001; odds ratio (OR) = 69] in aCalifornia population. Pollart et al. (68)reported an association between asthma andelevated IgE to grass pollen (X2 = 8.8;p < 0.005) in a Virginia population. In a studyof allergy to laboratory animals, Newill et al.(69) incidentally report a positive associationbetween hyperreactive airways disease and skinsensitivity to (ragweed) pollen allergens. Inthis study, 17 of 36 lab animal workers withone or more positive methacholine challengeswere skin-test positive to ragweed extract,whereas only 2 and 7 of these patients hadpositive reactions to lab animals and house-hold allergens, respectively (p < 0.004). Lehreret al. (6), report a significant association (x2;p < 0.005) between skin reactivity tobasidiospore extracts and atopy, asthma, andasthma with rhinitis (but not rhinitis alone).Basidiospores are generally not producedindoors and are often the most abundant out-door allergen-bearing particle. The incidenceof asthma has also been associated with skinreactivity to Alternaria (OR 5.1; 95% confi-dence interval [C.I.] 2.9-8.9). Thirty-ninepercent of inner city asthmatic children hadpositive skin test to Alternaria (far more abun-dant outdoors than in), and an additional 4%were sensitive to Penicillium (common bothout and in) (70). O'Hollaren et al. (71) sug-gest that sensitization to Alternaria allergens inyoung asthmatics is a risk factor for respiratoryarrest. It should be noted that a significantrelationship between symptoms and skin testreactivity is not proof of a direct relationshipbetween exposure to a specific allergen andsymptoms. Many patients have skin reactivity

Environmental Health Perspectives * Vol 108, Supplement 4 * August 2000656

OUTDOOR ALLERGENS

to allergens that do not induce symptoms onexposure. Specific skin test/symptom relation-ships may reflect multisensitization.

Outdoor Allergen Exposure andAsthmaExposure to outdoor allergens has beenrelated to respiratory symptoms in severalstudies. In the California study discussedabove by Pollart et al. (60), exposure to grasspollen was measured using a gravity collector,the data from which was strongly correlatedwith counts from a rotorod. They presentemergency room admissions for asthma andpollen counts graphically, and the peak ofasthma visits is near the center of the pollenpeak. However, they do not present a statisti-cal analysis of these data. D'Amato et al. (4)used symptom diaries and spore trapping torelate symptoms and exposure to P. judaicapollen. Exposure to grass pollen allergens hasbeen associated indirectly with epidemicasthma (72,73). These studies used filter col-lections and immunoassays for grass allergensfor exposure assessment. Johnsen et al. (64)report relationships between exposure tobirch, grass, and mugwort pollens (Burkardspore-trap measurements) and allergens (mea-sured immunochemically from a high-volumefilter sampler) and symptom scores but donot discuss the kinds of symptoms recorded.Neas et al. (74) related changes in peak expira-tory flow rate (PEFR), a measure of lung func-tion and bronchoconstriction, in 108 childrenand incremental exposure (using spore trapdata) to several outdoor fungal spore types(Cladosporium, 10,000 spores/m3; Epicoccum,60 spores/m3). Delfino et al. (75) revealedpositive relationships between outdoor con-centrations of total fungal spores (spore trapdata), spores for which skin test material wasavailable (primarily Cladosporium), spores forwhich skin test material was not available (pri-marily basidiospores and ascospores), anddaily asthma symptoms and inhaler use in atime-series diary study of 12 children withasthma. Although symptoms were associatedwith exposure to the comparable fungal sporetype in children with positive skin tests, mostassociations were found for spore types forwhich skin-test materials were unavailable.

Associations between Asdtma, AirPoliution, and Meteorological FactorsAtmospheric pollution has been proposed as apossible factor in the continuing increase inasthma mortality and morbidity. Kesten et al.(76) report an association between NO2,SO2, ozone, air pollution and air qualityindices, and emergency room visits for acuteasthma when the data are lagged by 1 and/or7 days but no relationship for any exposureindex for the day of the visit. Marzin et al.(77) report an association between emergencyservice for asthma and both SO2 and high

atmospheric pressure. Forsberg et al. (78)report a relationship between relatively lowlevels of black smoke particulate matter andasthma symptoms as reported by diary.Schwartz et al. (79) reported an associationbetween particulate air pollution and emer-gency room visits for asthma in patientsunder 65 years of age. Although exposure tooutdoor allergens clearly is related to allergicdisease, few studies control for allergen expo-sure when studying effects of air pollutants.Delfino et al. (75) related PEFR decrementsand personal exposure to ozone while con-trolling for fungal spore exposure in 12 chil-dren, and controlled for ozone exposure withrespect to the relationship between sporeexposure and asthma.

Interrelationships among air pollution,meteorological factors, and fungal allergenswith respect to their effects on asthma remainlargely unexplored. Such relationships couldinvolve interactive effects on the nature ofexposure as well as synergistic effects on thedisease process itself. Clearly, meteorologicalfactors influence concentrations of outdoorallergen-bearing partides [e.g., see Stark et al.(53)]. Small (< 3 pm) particles released fromgrass pollen during thunderstorms are thereported link between thunderstorms andasthma epidemics in the grass pollen season(72). Using these small particles, Suphioglu etal. (73) elicited IgE-mediated responses inpatients with asthma and produced broncho-constriction with inhalation challenge in fourpatients. A few studies suggest direct effects ofair pollutants on the nature of allergens.Ruffin et al. (80), exposed oak, grass, and elmpollen to CO, SO2, and NO2 and notedchanges in amino acid and molecular-weightprofiles and in precipitin banding patterns.

Controlling Outdoor Aliergen-InducedDiseaseAvoidance. As discussed above, penetrationby outdoor allergens into interiors is relatedto the types of barriers to their entrance. Allbuildings have inadvertent openings thatallow penetration of gases, vapors, and evenparticles. However, the recognized allergen-bearing particles do not readily penetratebuildings that are mechanically ventilatedwith filtered air. Thus, in otherwise clean,modern office buildings, exposure to outdoorallergens is minimal (81). In residences, asimilar effect is seen, with air-conditionedhomes having lower outdoor particle levelsthan naturally ventilated homes (84.

The indoor environment is only protec-tive if it remains relatively allergen free.Removing indoor sources (i.e., outdoor aller-gens in settled dust, fungal growth in reser-voirs) is an obvious approach but one that hasyet to be tested with respect to improvementin allergy symptoms.

Another potentially effective exposurecontrol approach is to stay indoors during thetimes when the allergen-bearing particle isabundant outdoors. This approach requiresan in-depth knowledge of the patterns ofprevalence for each particle of concern,including seasonal, spatial, and diurnal varia-tion patterns (83). The development of pre-dictive models may increase the precision ofthese avoidance approaches.

Outdoor source control is anotherapproach that has been tried for pollen reduc-tion. Laws now exist that forbid the plantingof pollen-rich trees (e.g., mulberry in theSouthwestern United States). Ragweed controlhas been attempted, but is difficult because ofthe aggressiveness of the plant in colonizingdisturbed ground (83). No attempts have beenmade to control the airborne fungi that areimportant allergen sources.

The use of masks to reduce particleexposure is well accepted in the industrialenvironment. However, few people are will-ing to wear respirators that adequatelyprevent exposure to fungal-size particles.

Some symptom relief can be obtainedfrom some allergens by moving to an envi-ronment where the allergen is absent. Thisused to be the case for ragweed, which wasisolated to the eastern United States. Thereare still some pollen types for which this typeof prevention could be used. However, inmost cases, new allergies will develop to what-ever allergens are abundant in the new envi-ronment. In addition, fungal exposureoutdoors is essentially universal.

Immunotherapy. Historically, immuno-therapy has been the treatment of choice foroutdoor allergens, and when the allergen iswell defined, the approach works well.Unfortunately, although many pollen aller-gens are well recognized, the specific fungithat cause allergic diseases are unknown forthe most part. In cases where fungal allergenshave been clearly identified, and purifiedforms have been used in immunotherapytrials, outcomes have been good (84).

Areas for Future ResearchPrevalence and ExposureGlobal climate change is generally consid-ered to be imminent, but its potentialimpact on biological systems is not wellunderstood. In the near future, changes inclimate and climate variability could poten-tially alter the timing and abundance of air-borne allergens. Preliminary studies on theimpact of increased atmospheric CO2 onragweed reproduction have shown a 60%increase in pollen production with a dou-bling of CO2 (85). Over longer time scales,climate change could also induce shifts inspecies ranges, potentially expanding areas of

Environmental Health Perspectives * Vol 108, Supplement 4 * August 2000 657

BURGE AND ROGERS

favorable growth for allergen-producingtaxa. Expansion of agricultural practices tofeed the expanding world population will alterland use patterns and increase source areas forfungal spore release. Hence, in some areasincreased exposure to pollen or fungal aller-gens could occur. The continued developmentof accurate forecasting systems will be neces-sary to compensate for exposure uncertaintydue to increased climate variability.

Large-scale documentation of the magni-tude and importance of outdoor allergens as acause of acute asthma exacerbations in the gen-eral population of asthmatics is required. Inparticular, knowledge of the subpopulations ofasthmatics for which outdoor allergen levels areimportant is needed. Currently no publishedstudies document the connection betweenasthma and outdoor allergen exposure and din-ical sensitivity to specific allergens. No existingstudies evaluate the role of indoor allergen sen-sitivity on outdoor allergen-induced asthma orcontrol for other potentially confounding vari-ables. The synergistic effect ofoutdoor allergensand air pollutants on asthma must be mea-sured. Priming effects and subsequent expo-sures to allergens or pollutants are also criticalareas of future study.

MonitoringMonitoring of outdoor allergen levels willremain a key aspect of assessing the impact ofglobal climate change on airborne allergenconcentrations and on potential exposure.Cost-efficient and less labor-intensive techno-logical advancements in detection and identi-fication of allergen-bearing particles for use inbroad temporal and spatial scale monitoringprograms are desperately needed. Thereremains a suite of unknown fungal sporesthat have defied classification and detection.Although microscopic identification of fungalspores has advanced greatly in recent decades,there are practical limitations to identificationthrough microscopy. Continued developmentof specific allergen determination and detec-tion methods (particularly for small particlefractions) is required for accurate assessmentofoutdoor allergen exposure.

Personal exposure monitoring remains anintractable problem of balancing cost versussensitivity. Current technology prevents theuse of personal monitoring on a wide scale,which would allow a broader interpretationfor exposure. Continued development of newlow-cost personal monitors would be benefi-cial for determining more exact indoor versusoutdoor allergen exposures.

Disease RelationshipsCertainly, much remains to be accomplishedin clarifying the role of outdoor allergenexposure in human disease. To date, a clearrelationship between exposure, sensitization,

and symptoms has not been made for any ofthe outdoor allergens. The role of pollenexposure in asthma and the mechanism of theeffect are important research areas. Fungalallergy remains one of the most frustratingand poorly studied areas in allergic disease.The kinds of fungi and the nature of theirallergens that lead to asthma developmentand exacerbation need intensive study. Inaddition the need for relevant and potentfungal allergens for skin testing andimmunotherapy trials is crucial.

REFERENCES AND NOTES

1. Maunsell K. The impact of aerobiology on allergy. Acta Allergol25:329-350 (1971).

2. Gregory P. The microbiology of the atmosphere. London:LeonardHill Books, 1961.

3. Mygind N, Weeke B. Allergic and nonallergic rhinitis. In: AllergyPrinciples and Practice (Middleton E, Reed C, Ellis E, eds). St.Louis, MO:CVMosby, 1983;1101-1117.

4. D'Amato G, Gentili M, Russo M, Mistrello G, Saggese M,Liccardi G, Falagiani P. Detection of Parietaria judaica airborneallergenic activity: comparison between immunochemical andmorphological methods including clinical evaluation. Clin ExpAllergy 24:566-574 (1994).

5. Gergen P, Turkeltaub P. The association of individual allergenreactivity with respiratory disease in a national sample: data fromthe second National Health and Nutrition Examination Survey,1976-80 (NHANES II). J Allergy Clin Immunol 90:579-588 (1992).

6. Lehrer S, Hughes J, Altman L, Bousquet J, Davies R, Gell L, Li J,Lopez M, Mailing H, Mathison D. Prevalence of basidiomyceteallergy in the USA and Europe and its relationship to allergicrespiratory symptoms. Allergy 49:460-465 (1994).

7. Rodrigo M, Morell F, Helm R. Identification and partial character-ization of the soybean-dust allergens involved in the Barcelonaasthma epidemic. J Allergy Clin Immunol 85:778-784 (1990).

8. Codina R, Ardusso L, Bertoya N. Sensitization to soybean hullallergens in subjects exposed to different levels of soybeandust inhalation in Argentina. J Allergy Clin Immunol105:572-576 (2000).

9. Erdtman G. Pollen morphology and plant taxonomy-Angiosperms. New York:Hafner, 1966.

10. Howlett B, Knox R. Allergic interactions. In: Cellular Interactions:Encyclopedia of Plant Physiology. New Series, Vol 17 (LinskensH, Heslop-Harrison J, eds). 1984;655-673.

11. Howlett R, Knox R, Paxton J, Heslop-Harrison J. Pollen wallproteins: physiochemical characterization and role in self-incompatibility in Cosmos bipinnatus. Proc R Soc London188:166-182 (1975).

12. Howlett B, Vithanage H, Knox R. Pollen antigens, allergens andenzymes. Curr Adv Plant Sci 35:1-17 (1979).

13. King T, Norman P. Isolation studies of allergens from ragweedpollen. Biochemistry 1:709-720 (1962).

14. Weeke B, Lowenstein H, Neilsen L. Allergens in Timothy pollenidentified by crossed radioimmunoelectrophoresis. Acta Allergol29:409-417 (1974).

15. Howlett B, Clarke A. Isolation and partial characterization oftwo antigenic glycoproteins from rye grass pollen. Biochem J197:695-706 (1981).

16. King T. Immunochemical properties of some atopic allergens. JAllergy Clin Immunol 64:159-163 (1979).

17. Moller C, Dreborg S. Cross-reactivity between deciduous treesduring immunotherapy. I: In vivo results. Clin Allergy16:135-143(1986).

18. Bernstein I, Perera M, Gallagher J, Michael J, Johansson S. In-vitro cross-allergenicity of major aeroallergenic pollens by theradioallergosorbent technique. J Allergy Clin Immunol57:141-152 (1976).

19. Lowenstein H. Cross reactions among pollen antigens. Allergy35:198-200 (1980).

20. Grote M, Fromme H. Visualization of birch pollen allergens usingIgE containing sera from human atopic individuals in immuno-gold labelling experiments. Histochem J 18:24-28(1986).

21. Schappi G, Suphioglu C, Taylor P, Knox R. Concentrations of themajor birch tree allergen Bet v 1 in pollen and respirable fineparticles in the atmosphere. J Allergy Clin Immunol100:656-661 (1997).

22. Solomon W, Burge H, Muilenberg M. Allergen carriage byatmospheric aerosol. I: Ragweed pollen determinants in smallermicronic fractions. J Allergy Clin Immunol 72:443-447 (1983).

23. Habenicht H, Burge H, Muilenberg M, Solomon W. Allergencarriage by atmospheric aerosol. Il: Ragweed pollen determi-nants in submicronic atmospheric fractions. J Allergy ClinImmunol 74:64-67 (1984).

24. Agarwal M, Swanson M, Reed C, Yunginger J. Airborne rag-weed allergens: association with various particle sizes andshort ragweed plant parts. J Allergy Clin Immunol 74:687-693(1984).

25. Suphioglu C. Thunderstorm asthma due to grass pollen. Int ArchAllergy Immunol 116:253-260(1998).

26. Knox R, Suphioglu C, Taylor P, Desai R, Watson H, Peng J,Bursill L. Major grass pollen allergen lol p 1 binds to dieselexhaust particles: implications for asthma and air pollution. ClinExp Allergy 273:246-251 (1997).

27. D'Amato G, Spieksma F, Liccardi G, Jager S, Russo M, Kontou-Fili K, Nikkels H, Wuthrich B, Bonini S. Pollen-related allergy inEurope. Allergy 53:567-578(1998).

28. Solomon A, Hayes H. Impacts of urban development upon aller-genic pollen in a desert city. J Arid Environ 3:169-178 (1980).

29. Wodehouse R. Pollen Grains. New York:Hafner, 1959.30. Kendrick B. The Fifth Kingdom. Newburyport, MA:Focus

Information Group, 1992.31. Arx JV. The Genera of Fungi Sporulating in Pure Culture. Lehre,

Germany:J Cramer, 1970.32. Carmichael J, Kendrick W, Conners I, Sigler 1. Genera of

Hyphomycetes. Edmonton, Canada:University of Alberta Press,1980.

33. Barnett H, Hunter B. Illustrated Genera of Imperfect Fungi. NewYork:MacMillan, 1987.

34. Levetin E. Basidiospore identification. Ann Allergy 62:306-310(1989).

35. Horner WE, Helbling A, Salvaggio JE, Lehrer SB. FungalAllergens. Clin Microbiol Rev 8:161-179 (1995).

36. Burge H, Muilenberg M, Chapman J. Crop plants as a source offungal spores of medical importance. In: Microbial Ecology ofLeaves (Andrews J, Hirano S, eds). New York:Springer Verlag,1990;222-236.

37. Zejda J, Dosman J. Respiratory disorders in agriculture.Tubercle Lung Dis 74:74-86 (1993).

38. Smid T, Heederik D, Mensink G, Houba R, Boleij J. Exposure todust, endotoxins, and fungi in the animal feed industry. Am IndHyg Assoc J 53:362-368 (1992).

39. Epstein E. Neighborhood and worker protection for compostingfacilities: issues and actions. In: Science and Engineering ofComposting: Design, Environmental, Microbiological andUtilization Aspects (Hoitink H, Keener H, eds). Wooster,OH:Ohio State University, 1993;319-338.

40. Aceves M, Grimalt J, Sunyer J, Anto J, Reed C. Identification ofsoybean dust as an epidemic asthma agent in urban areas bymolecular marker and RAST analysis of aerosols. J Allergy ClinImmunol 88:124-134(1991).

41. Pont F, Gispert X, Canete C, Pinto E, Dot D, Monteis J. An epi-demic of asthma caused by soybean in L'Hospitalet deLlobregat (Barcelona). Arch Bronconeumol 33:453-456 (1997).

42. Anto J, Soriano J, Sunyer J, Rodrigo M, Morell F, Roca J,Rodriguez-Roisin R, Swanson M. Long term outcome of soybeanepidemic asthma after an allergen reduction intervention.Thorax 54:670-674 (1999).

43. Ballester F, Soriano J, Otero I, Rivera M, Sunyer J, Merelles A,Verea H, Marin J, Anto J. Asthma visits to emergency roomsand soybean unloading in the harbors of Valencia and Coruna,Spain. Am J Epidemiol 149:315-322 (1999).

44. Falleroni A, Zeiss C. Bean-bag allergy revisited: a case ofallergy to inhaled soybean dust. Ann Allergy Asthma Immunol77:298-302 (1996).

45. Raynor G, Hayes J, Ogden E. Areas within isopleths of ragweedpollen concentrations from local sources. Arch Environ Health19:92-98 (1969).

46. Aylor D, Parlange J. Ventilation required to entrain small parti-cles from leaves. Plant Physiol 56:97-99 (1975).

47. Chamberlin A, Little P. Transport and capture of particles by veg-etation. In: Plants and Their Atmospheric Environment (Grace J,Ford E, Jarvis P, eds). Oxford:Blackwell, 1981;147-173.

48. Hirst J, Stedman 0. Dry liberation of fungal spores by raindrops.J Gen Microbiol 33:335-344 (1963).

49. Starr J, Mason B. The capture of airborne particles by waterdrops and simulated snow crystals. 0 J R Meteor Soc92:490-499(1966).

50. Arizmendi C, Sanchez J, Ramos N, Ramos G. Time series pre-dictions with neural nets: application to airborne pollen fore-casting. Int J Biometeorol 37:139-144 (1993).

658 Environmental Health Perspectives * Vol 108, Supplement 4 * August 2000

OUTDOOR ALLERGENS

51. Stephen E, Raftery A, Dowding P. Forecasting spore concentra-tions: a time series approach. Int J Biometeorol 34:87-89(1990).

52. Moseholm L, Weeke E, Petersen B. Forecast of pollen concen-trations of Poaceae (grasses) in the air by time series analysis.Pollen Spores 29:305-322 (1987).

53. Stark P, Ryan L, MacDonald J, Burge H. Using meteorologicdata to model and predict ragweed pollen levels. Aerobiologia13:177-184 (1997).

54. Norris-Hill J. The modelling of daily Poaceae pollen concentra-tions. Grana 34:182-188 (1995).

55. Michel F, Marty J, Quet L, Cour P. Penetration of inhaled polleninto the respiratory tract. Am Rev Respir Dis 115:609-613(1977).

56. Busse W, Reed C, Hoehne J. Where is the allergic reaction inragweed asthma? II: Demonstration of ragweed antigen in air-borne particles smaller than pollen. J Allergy Clin Immunol50:289-293 (1972).

57. Gautrin D, Vandenplas 0, DeWitte J, L'Archeveque J, LeblancC, Trudeau C, Paulin C, Arnoud D, Morand S, Comtois P.Allergenic exposure, 1gE-mediated sensitization, and relatedsymptoms in lawn cutters. J Allergy Clin Immunol 93:437-445(1994).

58. Spengler J, Sexton K. Indoor air pollution: a public health per-spective. Science 221:9-17 (1983).

59. Solomon W, Burge H, Boise J. Exclusion of particulate allergensby window air conditioners. J Allergy Clin Immunol 65:305-308(1980).

60. Pollart S, Chapman M, Fiocco G, Rose G, Platts-Mills T.Epidemiology of emergency room asthma in northern Califomia:association with IgE antibody to ryegrass pollen. J Allergy ClinImmunol 88:224-230(1988).

61. Sexton K, Letz R, Spengler J. Estimating human exposure tonitrogen dioxide: an indoor/outdoor modeling approach. EnvironRes 32:151-166 (1983).

62. Anonymous. Pollen and Spore Report. Milwaukee, WI:AmericanAcademy of Asthma, Allergy, and Immunology, 1999.

63. Solomon W, Burge H, Boise J, Becker M. Comparative particlerecoveries by the retracting rotorod, rotoslide, and Burkard

spore trap sampling in compact array. Int J Biometeorol24:107-116(1980).

64. Johnsen C, Weeke E, Nielsen J, Jensen J, Mosbech H, Frolundl, Madsen F, Poulsen l. Aeroallergen analyses and their clinicalrelevance. Il: Sampling by high-volume air sampler with immuno-chemical quantification versus Burkard pollen trap sampling withmorphologic quantification. Allergy 47:510-516 (1992).

65. Agarwal M, Swanson M, Reed C, Yunginger J.Immunochemical quantitation of airborne short ragweed,Alternaria, antigen E, and Alt 1 allergens: a two-year prospec-tive study. J Allergy Clin Immunol 72:40-45 (1983).

66. Poulos L, O'Meara T, Sporik R, Tovey E. Detection of inhaledDer p 1. Clin Exp Allergy 29:1232-1238(1999).

67. Palmgren U, Strom G, Blomquist G, Malmberg P. Collection ofairborne microorganisms on nuclepore filters: estimation andanalysis-CAMNEA method. J Applied Bacteriol 61:401-40611986).

68. Pollart S, Chapman M, Fiocco G, Rose G, Platts-Mills T.Epidemiology of acute asthma: IgE antibodies to commoninhalant allergens as a risk factor for emergency room visits. JAllergy Clin Immunol 833:875-880(1989).

69. Newill C, Eggleston P, Prenger V, Fish J, Diamond E, Wei 0, IIlRE. Prospective study of occupational asthma to laboratory ani-mal allergens: stability of airway responsiveness to metha-choline challenge for one year. J Allergy Clin Immunol95:707-715 (1995).

70. Eggleston P, Rosenstreich D, Lynn H, Gergen P, Baker D, KattanM, Mortimer K, Mitchell H, Ownby D, Slavin R, et al.Relationship of indoor allergen exposure to skin test sensitivityin inner city children with asthma. J Allergy Clin Immunol4:563-570 (1998).

71. O'Hollaren M, Yunginger J, Offord K, Somers M, O'Connell E,Ballard D, Sachs M. Exposure to an aeroallergen as a possibleprecipitating factor in respiratory arrest in young patients withasthma. N EngI J Med 324:359-363 (1991).

72. Bellomo R, Gigliotti P, Treloar A, Holmes P, Suphioglu C, SinghM, Knox B. Two consecutive thunderstorm associated epi-demics of asthma in the city of Melbourne. The possible role ofrye grass pollen. Med J Aust 156:834-837(1992).

73. Suphioglu C, Singh M, Taylor P, Bellomo R, Holmes P, Puy R,Knox RB. Mechanism of grass-pollen-induced asthma. Lancet339 (8793):569-572 (1992).

74. Neas L, Dockery D, Burge H, Koutrakis P, Speizer F. Fungusspores, air pollutants and other determinants of peak expiratoryflow rate in children. Am J Epidemiol 143:797-80711996).

75. Delfino R, Coate B, Zeiger R, Seltzer J, Street D, Koutrakis P.Daily asthma severity in relation to personal ozone exposureand outdoor fungal spores. Am J Respir Crit Care Med154:633-641 (1996).

76. Kesten K, Szalai J, Ozyngel B. Air quality and the frequency ofemergency room visits for asthma. Ann Allergy 74:269-273(1995).

77. Marzin C, LeMoullec Y, Ancelle T, Juhel J, Festy B, Pretet S.Asthme, pollution atmospherique urbaine et meteorologie. RevMal Respir 10:229-235(1993).

78. Forsberg B, Stjernberg N, Falk M, Lundback B, Wall S. Air pollu-tion levels, meteorological conditions and asthma symptoms.Eur Respir J 6:1109-1115 (1993).

79. Schwartz J, Koenig J, Slater D, Larson T. Particulate air pollu-tion and hospital emergency visits for asthma in Seattle. AmRev Respir Dis 147:826-831 (1993).

80. Ruffin J, Liu M, Sessoms R, Banerjee S, Banerjee U. Effects ofcertain atmospheric pollutants (SO2 NO2, and CO) on the solu-ble amino acids, molecular weight and allergenicity of someairborne pollen grains. Cytobios 46:119-129(1986).

81. Burge H, Pierson D, Groves T, Strawn K, Mishra S. Dynamics ofairborne fungal populations in a large office building. CurrMicrobiol 40:10-16 (20001.

82. Lebowitz M, O'Rourke M, Dodge R, Holberg G, Hoshaw R,Pinnas J, Barbee R, Sneller M. The adverse health effects ofbiologic aerosols, other aerosols, and indoor microclimate onasthmatics and nonasthmatics. Environ Int 8:375-380 (1982).

83. Fischbach FA. Biophysical factors in ragweed pollen: avoidancestrategies in a community. Grana 25:221-233 (1986).

84. Dhillon M. Current status of mold immunotherapy. Ann Allergy66:385-392(1991).

85. Wayne P, Foster S, Connolly J, Bazzaz F. Unpublished data.

Environmental Health Perspectives * Vol 108, Supplement4 * August 2000 659