Definition of the Agent and Means of Exposure

Cats are kept as pets in 27% of U.S. households. The major cat allergen, Fel d I, is a glycoprotein structured as a heterodimer with two chains of amino acids, which have been defined by polymerase chain reaction (PCR) and subsequent DNA sequencing (Griffith et al., 1992; Morgenstern et al., 1991; Schou, 1993). It is found on cat hair and is produced in cat sebaceous, salivary, and anal glands (De Andrade et al., 1996). In male cats, Fel d I glandular production is under hormonal control and decreases after castration (Zielonka et al., 1994). As discussed in the Third International Workshop report (Platts-Mills et al., 1997), the clinical significance of this decrease in allergen production is not certain, since many symptomatic cat-allergic asthmatics in the United States have neutered cats. Further investigation of hormonal and genetic control of Fel d I production could be relevant to the control of allergen levels in homes with cats. Although 90% of patients allergic to cats make immunoglobulin E (IgE) to Fel d I (de Groot et al., 1988; Schou, 1993), making Fel d I a marker for the immune response to cat allergens, at least eight other cat allergens have been identified (Duffort et al., 1987), suggesting that protection from Fel d I exposure may not be the equivalent of protection from cat allergen exposure. This conclusion is supported by the findings that 66% of the histamine-releasing activity of cat hair and dander extract, and about 60% of the cat dander radioallergosorbent test (RAST) activity, was carried by Fel d I (Schou, 1993).

Touching the cat is only one mode of contact that may result in airborne suspension of allergen and potential direct hand-to-nose deposition of allergen-associated particles. In contrast with cockroach allergen, which is airborne only transiently during the disturbance of household dust, cat allergen can remain airborne for long periods of time, in part because Fel d I is associated to a significant extent with smaller particles of less than 5 µm (Custovic et al., 1998b). Particles on which cat allergen is carried, coming primarily from cat dander, are also very adherent. Consequently cat allergen is spread easily throughout a house, even when cats are kept out of certain rooms. Moreover, cat allergen is easily carried from home to home, office, school, or day care center by those who touch cats or visit households with cats (Custovic et al., 1998a; Dybendal and Elsayed, 1994; Warner, 1992). At trace or small amounts that may be significant for sensitization or exacerbation of disease in sensitized individuals, Fel d I in settled dust is found in most homes without cats (Bollinger et al., 1996; Chew et al., 1998), although allergen levels are generally higher in homes with cats. A Baltimore, Maryland, study found measurable levels of airborne Fel d I in 37 homes with cats (range, 1.8–578 ng/m³; median, 45.9 ng/m³) and in 10 of 40 homes without cats (range for detectable samples, 2.8–88.5 ng/m³; median, 17 ng/m³) (Bollinger et al., 1996). In 38 of 40 homes without cats, Fel d I was present in the settled dust (range, 39–3,750 ng/g; median, 258 ng/g); dust levels were weakly correlated with airborne levels. Carpeting, bedding, and upholstered furniture can be reservoirs for deposited cat allergen (Wood et al., 1989); shaking the bedding in rooms with cats resuspends cat allergen in the air.

As well as being detected in homes without cats, cat allergen has also been detected in public places such as hospitals and schools. In one British study that measured both settled and airborne cat allergen in hospitals, the amount of cat allergen in settled dust in upholstered chairs was as high as in homes with cats (geometric mean, 23 µg/g dust), but airborne levels were low (0.22 ng/m³) (Custovic et al., 1998a).

Evidence Regarding Asthma Exacerbation

In cat-sensitized asthmatics, cat allergen can induce allergic symptoms, asthmatic symptoms, and decrements in lung function. Exposure to inhaled cat allergen in an experimental cat room led to significant decreases in forced expiratory volume in one second (FEV_1, range, 6–57%; mean, 25%) in a study of 13 adults with cat allergy. The percentage decrease in FEV₁ did not correlate significantly with either the intradermal titration end point with cat allergen or the magnitude of the RAST response with cat allergen. Those cat-allergic subjects classified as asthmatic by methacholine challenge testing experienced almost identical responses to environmental allergen challenge in an experimental cat room and inhalation challenge with cat allergen (Sicherer et al., 1997). Norman and colleagues documented progressive increases in both nasal and lung symptom scores during a 60-minute period in a cat room (Norman et al., 1996).

While initial cat room studies involved very high airborne cat allergen levels, a later experimental cat room exposure study by Bollinger and colleagues (1996) evaluated symptom and lung function responses of cat-sensitive subjects to low-level airborne cat allergens. They demonstrated that cat-sensitive individuals can have increases in upper- (congestion, rhinorrhea, pruritus) and lower- (chest tightness, wheezing) respiratory symptoms and decrements in lung function at levels of cat allergen occurring in homes without cats. The median FEV₁ change was 15% in the seven challenges with a Fel d I level less than 100 ng/m³.

In a Delaware case-control study, Gelber and colleagues (1993) studied allergen predictors of emergency room visits for asthma. This study compared 93 patients, 15 to 55 years of age, who presented with breathlessness and airway obstruction, to 93 patients presenting without breathlessness. For cat and cockroach, the combination of sensitization and the presence of allergen in the house was associated with asthma presenting to hospitals (14/93 asthmatics versus 1/93 controls). Whether other exposures potentiate the response of cat-allergic asthmatics to cat exposure is unknown. In a cross-sectional study of children in New Mexico with asthma or bronchial hyperresponsiveness, cat sensitization and exposure to cat allergen were common (Sporik et al., 1995). Among children with asthma (defined as symptomatic bronchial reactivity), 13/19 were sensitized to cat. Numbers were too small to compare symptoms in cat-sensitized asthmatics with and without significant home exposure to cat allergen (Ingram et al., 1995).

Evidence Regarding Asthma Development

Insufficient data are available to assess whether exposure to cats influences the development of asthma. Cross-sectional studies of children suggest an association between sensitization to cats and home exposure in the first six months of life (Suoniemi et al., 1981; Warner et al., 1991). These studies may be subject to recall bias. A longitudinal birth cohort study from the Isle of Wight found that the presence of a cat in the home predicted a greater risk of skin test reactivity to cat and a greater risk of any skin test reactivity by 2 years of age (Hide et al., 1994). Sensitization to cat predicts asthma development, but this may simply be a confirmation of the well-documented fact that atopic individuals are more likely to develop asthma than nonatopic individuals. In a New Zealand birth cohort study, the development of asthma by 13 years of age was associated with sensitization to cats and dogs at age 13 (Sears et al., 1989). Sensitization to cats predicted the development of bronchial hyperresponsiveness in a longitudinal study of adults in Boston, Massachusetts (Litonjua et al., 1997). However, neither of these studies provides evidence of whether exposure to cats predicts either asthma or bronchial hyperresponsiveness. Since cat allergen exposure can potentially take place either at home or in schools and public places, the relative importance of home versus community-wide exposure to cat allergen in the risk of specific sensitization to cats is unknown.

Although the field of the genetics of asthma is in its infancy, preliminary studies suggest that certain genetic phenotypes are associated with allergy to specific insects or animals (Fukuda et al., 1995; Hizawa et al., 1998; Young et al., 1992, 1993). Understanding the genetics of allergy and asthma, including understanding the phenotypes associated with allergy specificity, may eventually prove useful in understanding gene-by-environment interaction in the development of asthma.

Conclusions: Asthma Exacerbation and Development

In cat-sensitive asthmatics, cat allergen exposure leads to worsening of respiratory symptoms and to a decline in lung function. Although sensitization to cats is a prerequisite to reactivity to cat exposure, the level of airborne cat allergen that exacerbates asthma varies by individual and is not necessarily predictable by the size of the skin test reaction to cat or the titer of IgE antibody. However, specific sensitive subgroups have not been defined. The relationship between cat allergen in the home and asthma development is uncertain. Because cat allergen is frequently found outside the home and in households without cats, the assessment of individual exposures to cats is difficult, making evaluation of the association between cat allergen exposure and asthma development difficult as well. In summary:

• There is sufficient evidence of a causal relationship between cat allergen exposure and exacerbation of asthma in individuals specifically sensitized to cats.
• There is inadequate or insufficient evidence to determine whether or not an association exists between cat allergen exposure and the development of asthma.

Evidence Regarding Exposure Mitigation and Prevention: Homes

Removal of the cat from the home will decrease exposure to cat allergen and is widely recommended for symptomatic cat-sensitive asthmatics. However even when the owner removes the cat, cat allergen levels may remain elevated for 20 weeks or more (Wood et al., 1989). Removal of carpets and upholstery, with encasement of mattresses and pillows, may be required for diminishing cat allergen to levels commonly measured in homes without cats (Wood et al., 1989).

No studies are available that evaluate symptoms or lung function in cat-sensitive asthmatics before and after removal of the cat from the home. Nor are there studies of change in symptoms or lung function after moving from a home with a cat to a home without a cat. However, exposure studies suggest that in cat-sensitive subjects the decline in lung function associated with low-level airborne exposure to cat allergen, which can be present in a home with no cats, tends to be less extreme than the decline in lung function associated with higher-level airborne exposure (Bollinger et al., 1996). The experiments demonstrating entry into a room with a cat as a source of exacerbation of asthma in catsensitive individuals also suggest that removal of the cat from the household may decrease symptoms in cat-allergic asthmatic.

Because of the reluctance of cat-allergic symptomatic asthmatics to get rid of their cats, a number of studies have focused on the potential for lowering cat allergen levels by washing the cat. In eight households with cats, de Blay and colleagues (1991a) found that the combination of washing the cat weekly, reducing furnishings, vacuum cleaning, and air filtration reduced airborne cat allergen levels. In a second study, however, Avner and colleagues (1997) from the Platts-Mills group found that while washing cats by immersion will transiently remove significant allergen from the cat and reduce the quantity of airborne Fel d I, this reduction in allergen is not maintained by one week. Klucka and colleagues (1995) found no significant reduction of Fel d I by washing, use of Allerpet-C (a widely advertised topical spray), or acepromazine, a tranquilizer advocated as efficacious in subsedating doses. While removal of the cat from the living room and bedroom areas of the home and use of a High-Efficiency Particulate Air (HEPA) filter reduced airborne levels of cat allergen in homes with cats, the reduction was not evenly spread across the particle size range (Custovic et al., 1998b). It is unclear that the level of reduction of allergen obtained in this study is sufficient to influence symptoms in cat-sensitized asthmatics. No studies are available to assess the efficacy of recommendations to wash cats in reducing symptoms in cat allergic-asthmatics. Although the combination of HEPA filter use, mattress and pillow covers, and exclusion of cats from the bedroom reduced airborne cat allergen levels, a Maryland study detected no improvement in daily symptom scores, peak flow rates, medication use, monthly spirometry, pre-and post-study cat-specific IgE levels, and methacholine challenge studies in cat-allergic subjects (Wood et al., 1998). On the other hand, in a double-blind, placebo-controlled, cross-over study of twenty asthmatic children sensitized to cat or dog allergens, and living in homes with these animals, airway hyperresponsiveness was improved and peak flow variation was decreased during the use of air cleaners in the living room and bedroom of the child (van der Heide et al., 1999). The authors report that substantial amounts of cat and dog allergen were captured by the air cleaners; floor cat and dog allergen levels were unchanged by air cleaner use.

Evidence Regarding Exposure Mitigation and Prevention: Schools and Hospitals

Even if the cat is removed from the home, continued low-grade exposures may occur in public places or via clothes from cat owners. Since cat allergen is everywhere, there is little potential for absolute avoidance (Dybendal and Elsayed, 1994; Warner, 1992). Norwegian investigators have demonstrated the presence of Fel d 1 in schools on both smooth and carpeted floors, with approximately 11 times more allergen on the carpeted floors (Dybendal et al., 1991, 1989a, 1989b). The frequency of cleaning floors and furniture was believed to influence the level of cat and dog allergen, which were higher on chairs than in floor dust (Warner, 1992). Upholstered chairs and mattresses in hospitals are also demonstrated reservoirs for cat and dog allergen (Custovic et al., 1998a).

To lessen the risk of exacerbation of asthma in cat- or dog-sensitized asthmatics in public buildings, Warner and others (1992) have recommended the use of smooth floors and frequently cleaned wooden or plastic chairs. In their study demonstrating the presence of significant cat and dog allergen levels in a hospital in Manchester, England, Custovic and colleagues (1998a) questioned the introduction of soft furnishings and carpets into hospitals where highly cat- or dog-allergic asthmatics may come for care. Where upholstered chairs were present, they demonstrated that vacuuming three times a week significantly reduced allergen levels.

Conclusions: Exposure Mitigation and Prevention

Cat allergen levels can be reduced to levels found in homes without cats by removal of the cat from the home, but the reduction in allergen levels may require a prolonged period of time. The combination of HEPA filter use, mattress and pillow covers, and exclusion of cats from the bedroom may not reduce airborne cat allergen levels sufficiently to improve symptoms in cat-sensitive asthmatics. The absence of carpet, the use of plastic or wooden rather than upholstered chairs, and of frequent vacuuming in schools and hospitals may decrease the levels of cat allergen in public places. No studies are available to evaluate whether these measures improve symptoms or lung function in cat-sensitized asthmatics or whether they decrease the potential for sensitization in nonsensitized individuals. In summary:

• There is sufficient evidence of an association between removal of a cat from the home and a decrease in levels of cat allergen in the home; this decrease in levels of allergen may be slow if reservoirs of cat allergen are not simultaneously removed from the home.
• There is limited or suggestive evidence of an association between removal of a cat from the home and improvement of symptoms or lung function in cat-allergic asthmatics.
• There is limited or suggestive evidence of an association between measures short of removal of a cat from the home (e.g., washing the cat, HEPA filter use) and some transient reduction in cat allergen levels in the home.
• There is inadequate or insufficient evidence to determine whether or not an association exists between measures short of removal of a cat from the home (e.g., washing the cat, HEPA filter use) and improvement in symptoms in cat-allergic asthmatics.

Definition of the Agent and Means of Exposure

Dogs are present in 31% of U.S. households and are also sources of allergens (Schou, 1993). Allergy to cats is reported to be about twice as common as allergy to dogs, despite the fact that dogs are as common in U.S. households as cats (Bollinger et al., 1996). Can f I and Can f II are purified dog allergens that have been identified (Schou, 1993). Can f I is a polypeptide whose molecular weight and structure have been partially but not fully defined (Schou, 1993). It is present in dander, pelt, hair, and saliva, but not in the urine or feces of dogs (Schou, 1993). Though there are likely to be other dog allergens, no others have been found to have clinical importance. Can f I is considered a major allergen, because it accounts for at least half of the allergenic activity in dog hair and dander. In addition, 92% of dog-allergic patients had a positive skin prick test to Can f I (Schou, 1993; Yman et al., 1973). It is still a controversial matter whether true breed-specific dog allergens exist or whether the differences observed between breeds are quantitative rather than qualitative (Schou, 1993). Hair is not the only source of dog allergen, and it is not known whether short-haired dogs are less allergenic. Cross-reactivity can be found between dog and cat allergen (Vanto and Koivikko, 1983).

Dog allergen, like cat allergen and unlike cockroach, is easily aerosolized and widely disseminated throughout the community (Custovic et al., 1997). In a Baltimore study of 42 homes, dog antigen was demonstrated in more than half of households (Lind et al., 1987; Schou, 1993). In Sweden, dog allergen has been measured in homes that have never had dogs (Munir et al., 1992). Like cat allergen, dog allergen has been found in significant amounts in public buildings such as schools (Berge et al., 1998; Dybendal et al., 1989a; Schou, 1993; Warner, 1992) and hospitals (Custovic et al., 1998a). In dust from upholstered English hospital chairs, Can f I levels (geometric mean = 22 µg/g, range, 4–63) were as high as levels in settled dust from households with dogs (Custovic et al., 1998a; Munir et al., 1994). Hospital airborne Can f I levels were detectable in 7 of 10 testing days but were lower (range 0.09–0.22 ng/m³) than those often found in homes with dogs (range 0–100 ng/m³ Can f I) (Custovic et al., 1998a; Hodson et al., 1999).

Evidence Regarding Asthma Exacerbation

The asthmatic response to bronchial provocation test (PT) with dog allergen was evaluated in a cross-sectional Finnish study of 203 asthmatic children selected from the Children's Asthma Registry (Vanto and Koivikko, 1983). Of those with a positive PT, 64% had kept dogs, whereas only 36% with a negative PT had kept dogs. A positive PT was correlated with a positive skin prick test to dog (correlation coefficient = 0.8), but not with the frequency of reported asthma symptoms. In immunotherapy trials, positive response to bronchial provocation with dog allergen has also been associated with elevated levels of IgE to dog allergen in asthmatic subjects (Hedlin et al., 1995; Valovirta et al., 1984; Vanto et al., 1980). Some investigators consider symptomatic and bronchial response to animal allergen in an experimental animal room to be more definitive proof that animal allergen triggers asthma than allergen bronchial PT. They question whether the airway response to bronchial provocation with an allergen is always an allergic rather than an irritant response. The committee could find no published studies of the response of dog-sensitized asthmatics to exposure to dogs in an experimental dog room analogous to the cat room set up by the Hopkins group (Sicherer et al., 1997).

Evidence Regarding Asthma Development

There is insufficient evidence regarding the role of dog allergen in the development of asthma. In keeping with the ecology of Los Alamos, New Mexico, which is high and dry, with less dust mite, the allergens to which asthmatics are sensitized tend to be the predominant allergens in the indoor and outdoor environment. These allergens include dog and Alternaria, as well as cat (Ingram et al., 1995; Sporik et al., 1995). A cross-sectional retrospective Finnish study suggested that dog allergy was more prevalent in children from homes where dogs were present in the first few months of life, compared to homes where the dog was introduced after the child had reached the age of 1 (Vanto and Koivikko, 1983). As mentioned above, in a New Zealand birth cohort study the development of asthma by 13 years of age was associated with sensitization to dogs at age 13 (Sears et al., 1989). In contrast, the presence of a dog in the home in childhood was negatively associated with asthma (OR = 0.85, 95% CI = 0.78–0.92) among adults reporting no parental allergy, in the cross-sectional European Community Respiratory Health Survey of 13,932 20- to 44-year-old subjects from 36 areas in Europe, New Zealand, Australia, and the United States (Svanes et al., 1999). Among 1,649 Swedish school children aged 7–13 years, a report of keeping a cat or a dog in the first year of life was also negatively associated with asthma and allergic rhinitis (Hesselmar et al., 1999).

Conclusions: Asthma Exacerbation and Development

• There is sufficient evidence of an association between dog allergen exposure and exacerbation of asthma in individuals specifically sensitized to dogs.
• There is inadequate or insufficient evidence to determine whether or not an association exists between dog allergen exposure and development of asthma.

Evidence Regarding Exposure Mitigation and Prevention

Because the aerodynamic properties, carrier material, and chemical composition of dog and cat allergens are similar, issues related to mitigation of dog allergen are likely to be similar. However fewer studies are available related to the mitigation of dog allergen exposure and its health consequences. In a cross-sectional study of 203 asthmatic children listed in the Finnish Asthma Register, 68 of 203 had kept a dog, but 59 of 68 (87%) had removed the dogs from their homes. Parents tended to report that removal of the dog had improved asthma symptoms and had not been detrimental because of emotional deprivation (Vanto and Koivikko, 1983). The Finnish study also demonstrated that homes with dogs had higher dog antigen levels in dust than homes without dogs, some of which had a past history of a dog in the home. Homes in which occupants had indirect contact with dogs had more dog allergen than homes in which no one reported contact with dogs. Despite reported avoidance of dogs, a rising or steadily high level of dog-specific IgE was observed in follow-up of 24 dog-allergic subjects. The authors hypothesized that dog allergen encountered outside the home might be sufficient to boost IgE synthesis in most sensitive subjects (Vanto and Koivikko, 1983).

In a study of 25 homes with dogs, Custovic found that dogs had to be washed at least twice a week to maintain a reduction in recoverable Can f I from the hair (Hodson et al., 1999). Airborne dog allergen levels were not significantly affected by washing the dog. No studies are available regarding the effect of this mitigation measure on symptoms or lung function in dog-sensitive asthmatics.

Conclusions: Exposure Mitigation and Prevention

• There is limited or suggestive evidence of an association between removal of the dog from the home and reduction of dog allergen levels. This evidence comes from an association between the absence of a dog in the home and the measurement of low dog allergen levels in a study including homes that had a history of keeping dogs (Vanto and Koivikko, 1983).
• There is inadequate or insufficient evidence to determine whether or not an association exists between removal of a dog from the home and improvement in symptoms or lung function in dog-sensitized asthmatics. The one epidemiologic study reporting an association between dog removal and symptom improvement in asthmatic children relies on retrospective parental reporting without measures of sensitization at the time of dog removal or measures of symptoms or lung function before and after dog removal (Vanto and Koivikko, 1983).

Definition of the Agent and Means of Exposure

Exposure to rodents can come either from keeping pets or from their presence as pests in the home. Rodents (mouse, rat, and guinea pig) can also be found in school settings. They have been studied as sources of allergens, particularly because of their extensive use as laboratory animals (Schou, 1993). Hair and epithelial fragments carry allergenic molecules, the allergens measured are believed to be from urine, saliva, or skin. The relative importance of these allergen sources has been debated (Karn, 1994; Longbottom and Austwick, 1987; Walls and Longbottom, 1985). As described by Schou (1993), rodents have permanent proteinuria; allergenic protein from sprayed urine dries up and becomes airborne on dust particles. Airborne rodent allergen has been measured in laboratory facilities (Sakaguchi et al., 1990a; Swanson et al., 1985; Twiggs et al., 1982) and, in one study, in inner city apartments (Swanson et al., 1985). Clinically important allergens that have been identified include Mus m I and Mus m II for mouse, Rat n I for rat, and Cav p I and II for guinea pig (Schou, 1993). Assessment of rodent exposure in the home has been limited, to some extent, by limitations in the ability to measure allergens from all species of wild mice potentially present in the home.

Evidence Regarding Asthma Exacerbation

A number of cross-sectional studies document the association between handling animals in a laboratory setting and allergy (Beeson et al., 1983; Cockcroft et al., 1981; Cullinan et al., 1994; Davies and McArdle, 1981; Gross, 1980; Newman-Taylor, 1982; Schumacher et al., 1981; Venables et al., 1988). Hollander and colleagues (1996) conducted a prospective panel study of self-reported symptoms and peak flow in Dutch laboratory animal workers. Workers who reported asthmatic symptoms (chest tightness) due to working with rats had significant decreases in peak expiratory flow on days they worked with the animals; 86% of them were sensitized to rat allergens (Hollander et al., 1996). A cross-sectional study of British laboratory workers demonstrated an association between positive skin tests to animal extracts and asthmatic symptoms (Cockcroft et al., 1981). Whereas IgE antibody to rat allergen was present in only 2 of 135 laboratory workers without asthmatic symptoms, it was present in 12 of 18 laboratory workers with symptoms (Platts-Mills et al., 1987).

In the U.S. National Cooperative Inner City Asthma Study, 19% of asthmatic children were allergic to rats and 15% were allergic to mice, suggesting exposure to rat or mouse allergens in the home (Kattan et al., 1997). No data are available on rat or mouse allergen levels in the home and the exacerbation of asthma among rodent-sensitized asthmatics.

Evidence Regarding Asthma Development

Although there are retrospective reports of the incidence of asthma symptoms after beginning laboratory work with rodents (Platts-Mills et al., 1987), the committee could find no relevant studies on rodent allergen exposure and the development of asthma.

Conclusions: Asthma Exacerbation and Development

• There is sufficient evidence of an association between exposure to rodents in a laboratory setting and exacerbation of symptoms or lung function in rodent-sensitized asthmatics.
• There is inadequate or insufficient evidence to determine whether or not an association exists between exposure to rodents (wild or as pets) in the home and exacerbation of symptoms or lung function in rodent-sensitized asthmatics.
• There is inadequate or insufficient evidence to determine whether or not an association exists between exposure to rodents and the development of asthma.

Evidence and Conclusions: Exposure Mitigation and Prevention

No studies are available to document the influence of rodent eradication on the level of rodent allergen or on the severity of asthma in sensitized asthmatics.

• There is inadequate or insufficient evidence to determine whether or not an association exists between removal of rodents from the home and the reduction of rodent allergen levels.
• There is inadequate or insufficient evidence to determine whether or not an association exists between removal of rodents from the home and improvement in symptoms or lung function in rodent-sensitized asthmatics.

Definition of the Agent and Means of Exposure

Relatively few people in the United States live on farms where exposure to cows, horses, or pigs can be significant; only 1.5% of U.S. households keep horses as pets. Cow hair and dander contain at least 17 antigens, 4 of which have been identified as allergens and 3 of which have been purified (Bos d I, II, and III). Horse hair and dander have been demonstrated to contain three important allergens (Equ c I, II, and III) (Schou, 1993).

Evidence and Conclusions: Asthma Exacerbation and Development

Allergies to cows and horses are considered occupational diseases of farm workers and veterinarians (Prahl and Roedpetersen, 1979; Schou, 1993). Data on the effect of non-occupational exposures is lacking. In summary:

• There is inadequate or insufficient evidence to determine whether or not an association exists between cow or horse allergen in the home and the exacerbation of asthma in sensitive children or the development of asthma.

Living on a Farm and Development of Asthma

The epidemiologic literature on allergy in farmers and children from farming families was reviewed in a 1999 article demonstrating a lower prevalence of hay fever and allergic sensitization in farmer's children compared to peers from nonfarming families living in the same rural Swiss community (BraunFahrlander et al., 1999). While farm animals can be allergenic, lower rates of sensitization to pollen and animal dander have been reported in adult farmers compared to other occupational groups (Iversen and Pedersen, 1990; Kohler et al., 1983; Rautalahi et al., 1987; Sigsgaard et al., 1996). Swedish conscripts and Finnish university students raised on farms have reported fewer allergic symptoms than students from nonagricultural backgrounds (Åberg, 1989; Kilpeläinen et al., 1997). This may be the result of self-selection out of the farming community by individuals and families with a genetic predisposition toward allergy (von Mutius et al., 1994).

These findings have led investigators to question whether the farm environment itself might play a protective role in the development of allergy and allergic asthma. One hypothesis is that contact with farm animals and their bacterial products (including endotoxin, which is discussed later in this chapter) may be protective against allergy or asthma through early-life stimulation of TH1 immunity, particularly for individuals with specific genetic characteristics (Baldini et al., 1999). However, no additional evidence is available to confirm or reject the hypothesis that for some children, exposure to farm animals early in life might be protective against either allergy or asthma. The farm is a complex environment that varies by country and culture, and many other aspects of farm living may contribute to the observed epidemiologic differences between asthma or atopy in children from farming versus nonfarming families.

Definition of the Agent and Means of Exposure

Birds are kept in 5% of U.S. households. While it is clear that hypersensitivity pneumonitis can be associated with antigens from bird excreta, serum protein, and proteinaceous material in dispersed dust from birds (Christensen et al., 1975; Hendrick et al., 1978), specific bird antigens associated with allergy and asthma have not been defined with certainty. Tauer-Reich and colleagues (1994) studied five bird fanciers who complained of asthmatic symptoms during contact with their birds and also had documented bronchial hyperreactivity to acetylcholine. These individuals had positive IgE antibody reactions to bird sera as well as to extracts of feathers. Skin prick tests to mites, molds, pollen, and domestic animals other than birds were negative in all five patients.

Evidence Regarding Asthma Exacerbation and Development

A few additional case reports describe bird handlers with the combination of asthmatic symptoms in the presence of birds, symptoms of egg hypersensitivity, and specific IgE antibodies against blood serum proteins of chicken, parrot, budgerigar, or pigeon serum (de Blay et al., 1991b).

A portion of what is called bird allergy may be an expression of allergy to dust mites. Bird feathers can harbor mites (Kemp et al., 1996). The material used in skin prick testing for bird allergy may also be contaminated with mite allergen. Although clinicians have traditionally advised asthmatic patients not to use feather pillows, an epidemiologic study found an increased risk of wheeze in children using foam pillows compared to children using feather pillows (Strachan and Carey, 1995). This association could occur if parents of symptomatic children tend to provide their children with synthetic rather than feather pillows because of advice from health professionals. On the other hand, this association may relate to increase mite exposure from synthetic pillows. New Zealand researchers found that after four months of use, dust mite (Der p I) levels were significantly higher in synthetic than in feather pillows (Crane et al., 1997; Rains et al., 1999). To prevent the feathers from coming out, feather pillows may have more impermeable covers than synthetic pillows; which could result in less dust mite infestation.

Conclusions: Asthma Exacerbation and Development

• There is limited or suggestive evidence of an association between bird exposure and exacerbation of symptoms in birdsensitized asthmatics. This association may be confounded by the allergic asthmatic response to mites harbored by birds.
• There is inadequate or insufficient evidence to determine whether or not an association exists between bird allergen exposure and the development of asthma.
• There is inadequate or insufficient evidence to determine whether or not an association exists between down pillows and exacerbation of symptoms or lung function in asthmatics. Pillows are believed to be a risk factor for asthma because of their documented mite content, rather than because of the presence of bird allergen.

Evidence and Conclusions: Exposure Mitigation and Prevention

There is inadequate or insufficient evidence to determine whether or not an association exists between removal of a bird from the home and reduction in bird allergen levels or improvement in symptoms or lung function in bird-sensitized asthmatics.

Agent Definition and Biology

Cockroach is an important source of indoor allergen worldwide. Although more than 60 species have been identified, the most common indoor species in North America are the German (Blattella germanica), American (Periplaneta americana), and Oriental (Blatta orientalis). There are multiple allergens from cockroach that have been identified (Bla g I or Per a I, Bla g II, IV, and V) and cloned (Bla g II, IV, and V). Schou and colleagues (1990) purified an allergen from B. germanica and P. americana extract, which resulted in positive reaction to skin tests in 50% of patients who were allergic to cockroaches and was designated Bla g I (or Per a I). Monoclonal antibodies have been produced against extracts of both cockroach species and used for cockroach allergen identification and purification. Enzyme-linked immunosorbent assays have been developed that can be used to estimate exposure to some of these cockroach allergens (Bla g I, Per a I, and Bla g II) in the environment. Bla g I is a 25-kD (kilodaltons, a unit of molecular weight; also abbreviated kDa) cross-reacting antigen from both the German and the American cockroach. Bla g II is a 36-kD species-specific antigen derived from the German cockroach. Specific allergens have not been purified from Oriental cockroaches.

Neither the antigenic relationships between cockroach species nor the precise source of cockroach allergens are well understood. The source of the allergens derived from cockroaches is not known, but there are speculations to suggest that they may come from feces, parts of the body, or other sources in the body. Sequence homology searches are useful tools for investigation of the biologic function of cockroach allergens, and as more sequences become available it will be possible to make comparisons of biologic function and allergenicity, to compare allergen expression in different species, and to localize the source of the allergens in cockroach tissues.

At present, cockroach extracts are not standardized; however, the Food and Drug Administration's (FDA's) Center for Biologics Evaluation and Research has embarked on a program to standardize B. germanica extracts on the basis of skin testing, protein content, RAST inhibition, and specific allergen assays. Most patients in the United States appear to be sensitized to B. germanica, although there is a cross-reactivity between B. germanica and P. americana on skin tests and serum IgE antibody assay. The prevalences of IgE antibodies to Bla g I and Bla g II in patients allergic to cockroaches are 40% and 60–80%, respectively (Pollart et al., 1991a). Approximately 20% of patients lack detectable IgE antibody to either allergen, which suggests that cockroaches produce other important allergens (Chapman, 1993). However, the levels of the two allergens have been found to be highly correlated (r = 0.92; p < .01) in dust samples (Pollart et al., 1991a).

Factors Influencing Exposure

Studies have suggested that cockroach sensitization is an affliction of the inner city poor, but the complex interrelationship of race, poverty, and residence has been difficult to unravel (Bernton and Brown, 1964; Call et al., 1992; Garcia et al., 1994; Gelber et al., 1993; Koehler et al., 1987). Sarpong and colleagues (1996a) examined race and socioeconomic status (SES) as risk factors for cockroach allergen exposure and sensitization. In their cohort of 48 white and 39 African-American children, they found that both factors were independent predictors of cockroach sensitization. Among low-SES subjects, sensitization was common with 50% (4 of 8) of white participants and 75% (15 of 20) of African-American participants exhibiting a positive skin test. Cockroach antigens are widely distributed in homes and schools, and the kitchen is the most common source of cockroach allergen (Rosenstreich et al., 1997; Sarpong et al., 1996b, 1997). The cockroach allergen in school dust was similar to that reported in homes. The level of antigen reported in school dust is of concern because it may constitute a very important occupational risk to students, teachers, and other school workers. The source of the allergen is not known, but it is likely that the schools may have been infested with cockroaches since there was evidence of dead cockroaches in the schools examined (Sarpong et al., 1997).

In kitchens, food and water sources may be important factors for the proliferation of cockroaches. The humidity in the home may be an important factor for increased cockroach allergen in infested homes. However, in a study evaluating allergen levels in schools, the presence of air conditioners, which may lower the humidity, did not affect the distribution of allergen levels (Sarpong et al., 1997). Despite the evidence that cockroach allergen levels are higher in the kitchens of both asthmatic and nonasthmatic individuals, the concentration of cockroach allergen in the bedroom has been used as the surrogate marker of exposure (Eggleston et al., 1998; Rosenstreich et al., 1997; Sarpong and Han, 1999; Sarpong et al., 1996a, 1996b).

Exposure to cockroach allergens are dependent on their aerodynamic properties and the characteristics of the surface on which they are deposited. In an Ohio study, there was no difference between dust mite allergen concentrations in low-pile carpet and smooth floors, but allergen levels were significantly higher in high-pile carpets (Arlian et al., 1982). In a school study in the Baltimore metropolitan area, there was no difference in cockroach allergen, Bla g I between low-pile carpet and uncarpeted floors (Sarpong et al., 1997). However, kitchen areas are generally noncarpeted, and this may have distorted the level of allergen in uncarpeted compared to carpeted areas.

Cockroach allergens may behave like the dust mite antigen; that is, they are carried on large particles that become airborne for short periods of time during active disturbance. High concentrations of cockroach allergen are found in the kitchen, compared to the living room or bedroom. However, some studies have reported similar levels of cockroach allergen in all these sites. Taken together, the cockroach allergen may be more relevant in the bedroom than the kitchen or the living room because of close contact with the pillow while in bed. Interestingly, about 20% of homes with no evidence of cockroach infestation have significant levels of cockroach allergen in settled dust (Pollart et al., 1991b). Although the possibility of reporting bias cannot be excluded, it is likely that cockroach allergen may be present in a home long after the infestation has been controlled.

Cockroach hypersensitivity is a unique risk factor for asthma among the urban poor. Many case-control studies have documented that cockroach allergen exposure and sensitization are significantly more common in patients with asthma living in urban homes compared to those living in suburban homes (Bernton et al., 1972; Kang et al., 1993; Sarpong et al., 1996a). However, despite higher rates of exposure and sensitization among urban subjects, place of residence was not independently associated with cockroach exposure after correction for socioeconomic status; families of lower SES were likely to be exposed wherever they lived (Sarpong et al., 1996a). This agrees with the observations of Gelber and colleagues (1993), who used lack of health insurance as a marker of poverty and showed that this was a more important correlate of sensitization than urban residence per se as well as with others who have categorized patients on the basis of access to private medical care (Garcia et al., 1994). A 1999 report from Morgantown, West Virginia, documented the role of sensitization in infantile asthma (Wilson et al., 1999). These children who were at least 3 years of age with documented wheezing episodes demonstrated cockroach hypersensitivity at a rate of 25%. The rate of sensitivity in this predominantly white population was not related to low socioeconomic status.

Biologic Plausibility

Bernton and colleagues (1972) reported that inhalation of cockroach extract could induce an immediate asthmatic reaction in cockroach-sensitive asthmatic subjects. Subsequently, controlled inhalation challenge confirmed the production of antigenspecific, acute and late bronchospasm in cockroach-sensitive asthmatics (Kang, 1976). These antigen-induced asthmatic reactions were blocked by premedication with disodium cromoglycate. Challenge of these subjects with inhaled cockroach extract produced a significant antigen-specific peripheral eosinophilia, which progressed to its peak 24–36 hours after exposure and often more than doubled the baseline value. In animal studies there is evidence to suggest that cockroach antigen can induce lung eosinophilia (Campbell et al., 1998). Previous data in guinea pigs have shown that aerosolized cockroach antigen can be utilized to induce airway inflammation and alter airway physiology (Kang et al., 1996). The development of these models will allow the evaluation of mediators involved in both stages of cockroach allergen challenge, as well as the testing of specific therapeutic modalities.

Consistency

There is now good evidence from epidemiologic studies in several parts of the world which demonstrates that the development of immediate sensitivity to cockroach allergens is associated with asthma morbidity and that sensitization is related to the degree of allergen exposure (Eggleston et al., 1998; Rosenstreich et al., 1997; Sarpong and Han, 1999; Sarpong et al., 1996a; Sastre et al., 1996). In the original description of cockroach sensitivity in 1967, Bernton and Brown (1967) found higher rates of sensitization among Puerto Ricans and African Americans than among Jews and Italians in New York City and believed that the difference was related to the improved economic conditions of the latter groups. Shulaner (1970) also found a relationship to poverty and overcrowding, and a report by Gelber and colleagues (1993) relates exposure and sensitization to urban housing, race, and poverty as indicated by lack of medical insurance.

Strength of Association

Sensitization, with production of specific IgE antibodies to cockroach, is a strong risk factor for acute severe asthma, especially when sensitized persons are exposed to high concentration of allergen in their homes. R.P. Nelson and colleagues (1996) compared 29 children with acute asthma, ages 3–16 years, first seen in a Florida emergency department, to 25 control subjects. They found that sensitization to cockroach allergens was associated with acute asthma that required emergency treatment. Gelber and colleagues (1993) and Call and colleagues (1992) showed that cockroach sensitization was a risk factor for acute asthma in patients visiting emergency departments in Wilmington, Delaware, and Atlanta, Georgia, respectively. The multicenter National Cooperative Inner City Asthma Study (NCICAS) of the home environment of asthmatic children aged 4–9 years living in seven major urban areas and eight centers reported that cockroach allergen was the predominant indoor allergen in these homes. In the NCICAS, it was suggested that sensitization to cockroach allergen and exposure to high levels of cockroach allergen may explain the asthma-related health problems of inner city children (Rosenstreich et al., 1997).

Dose–Response

A dose–response relationship between cockroach allergen exposure and sensitization has been established in asthmatic children. Sarpong and colleagues (1996a) report that children who were exposed to Bla g I or Bla g II of 1 unit per gram (U/g) or higher demonstrated skin sensitivity to cockroach allergen. For Bla g I, all the children who were exposed to more than 10 U/g were sensitized to cockroach allergen. Similarly, 100% of the children who were exposed to more than 5 U/g of Bla g II were sensitized to cockroach allergen. Following this report, the NCICAS has suggested that cockroach allergen (Bla g I) exposure is related to sensitization in a dose–response manner (Eggleston et al., 1998). Preliminary data from a population of pregnant women have suggested that total cockroach allergen (Bla g I and Bla g II) exposure and sensitization were related in a dose–response fashion (Sarpong and Han, 1999). This relationship was demonstrated in both asthmatic and nonasthmatic controls. However, the dose of cockroach allergen concentration required to induce sensitization rates in the asthmatic population was at least a factor of ten lower than in the nonasthmatic population. A dose–response relationship was also demonstrated in the inner city asthma study between cockroach allergen (Bla g I) exposure and morbidity due to asthma. Children who were exposed to more than 8 U/g of Bla g I were more likely to be hospitalized for asthma symptoms than those who were exposed to less. In a prospective study by Gold and colleagues (1999), infants who were born in homes with Bla g I levels greater than 2 U/g reported increased risk of wheezing by the age of 1 year. A study by Litonjua and colleagues available as a conference abstract (1998) evaluated the relationship between home allergen levels and the prevalence and incidence of asthma 16 months later among 215 children younger than 5 years old. The researchers found that, among the children with new onset of asthma, 7 (87.5%) lived in homes with high Bla g I or II levels while 1 (12.5%) lived in a home with low levels (p = .028).

Although it seems clear that exposure to cockroach allergen precedes disease, it has not been shown that sensitization consistently precedes disease. However, it is the combination of strong association, biological plausibility, dose–response, and provocation experiments that creates the strength of the argument.

Agent Definition and Biology

Purification of indoor allergens is dependent on the quality of source materials. Before the discovery of dust mites, several unsuccessful attempts were made to purify allergens from dust obtained from carpets and bedding. Even when dust mite cultures became available, the early attempts at purification used only a small amount of material, ~40 g. The first purification of the major allergen Der p I was by classical immunochemistry from 400 g of “spent” culture (Chapman and Platts-Mills, 1980). The spent culture was very rich in fecal material. Purification made it possible to measure the quantity of allergen in extracts, in dust samples, and airborne (Tovey et al., 1981a). In addition, the purified allergens have been used to study the immune response to allergens, as well as the properties of these proteins. Each of these approaches has relevance to the causes of lung symptoms.

Epidemiologic evidence about the relationship between sensitization to indoor allergens and asthma is based on skin tests or on serum assays of IgE antibodies. However, the immune response to inhalant allergens also includes IgG antibodies, IgA antibodies, and T cells. These T cells in allergic individuals are predominantly CD4+ cells of the TH2 type (Romagnani, 1992; Wierenga et al., 1990). There are two important questions about this response: (1) Can differences in the immune response predict which skin test-positive individuals will develop asthma? Most of the evidence suggests that skin test-positive asymptomatic individuals have very similar IgE antibodies and T cells (i.e., TH2) compared to symptomatic patients. (2) What, if anything, is the immune response to mite allergens in skin test-negative individuals? Here, the evidence is conflicting. Several groups are proposing that nonallergic individuals have made a TH1 response. This is supported by evidence that T cells in the cord blood of individuals who are going to become allergic produce less interferon (IFN-γ) in response to nonspecific stimuli, and also by evidence that it is possible to clone mite-specific TH1 T cells from the peripheral blood of nonallergic individuals (Miles et al., 1996; Prescott et al., 1998; Warner et al., 1997; Wierenga et al., 1990). On the other hand, most investigators have not found evidence of an immune response to mite proteins in nonallergic individuals. For example, most skin test-negative individuals have no detectable IgG antibodies to purified antigens, have poor or no in vitro T cell responses, and have no immediate or delayed skin responses. These results suggest (1) that the primary issue is who makes an IgE antibody response and (2) that most individuals who are nonallergic have not made any immune response to dust mites.

Almost all of the well-defined allergens are proteins or glyco-proteins, and it is not surprising that many of them have amino acid sequence homology with known enzymes (Arruda et al., 1997; Stewart and Thompson, 1996). In some cases these proteins have enzymatic activity that could play a role in their immunogenicity. For example, Der p I is a potent protease that has been shown to cleave CD23 and CD25 on the surface of lymphocytes in vitro (Hewitt et al., 1995; Schulz et al., 1998). Furthermore, in animal experiments, it appears that the enzymatic activity of Der p I can influence immunogenicity (Comoy et al., 1998). However, enzymatic activity is not a prerequisite for a protein to be a significant allergen, and there are considerable doubts about whether these enzymes are active in vivo. Some of the allergens that have homology with enzymes have not been shown to have enzymatic activity. More significantly, it has not been established that any of these proteins have enzymatic activity under the conditions that occur in the human respiratory tract. In addition, many important allergens do not have homology with enzymes, including Der p II. It is certainly possible that enzymatic activity is relevant to the effects of some allergens; however, it is unlikely that this property plays an important role either in inducing IgE antibody responses or in causing symptoms.

Factors Influencing Exposure

The quantities of dust mite allergen that have been found in the air of houses range from <0.2 to ≥100 ng/m³. Thus, accurate determination of the quantity and particle size of airborne allergen is dependent on immunoassays capable of accurately measuring quantities as small as 1 ng (Chapman et al., 1987; Luczynska et al., 1989; Sakaguchi et al., 1990b). After it became clear that Der p I was concentrated in mite fecal particles, experiments were designed to answer whether the allergen became airborne in this form. Using various approaches to air sampling, several conclusions became clear: (1) there is very little or no airborne mite allergen in an undisturbed room; (2) the allergen that becomes airborne during disturbance is predominantly on particles ≥10 µm; and (3) allergen falls rapidly after disturbance, in keeping with an aerodynamic size of ~10 µm (Tovey et al., 1981b). In addition, allergen-containing particles identified using a microimmunodiffusion technique have been shown to be mite fecal particles (Tovey et al., 1981a). These properties are strikingly different from those of other indoor allergens (e.g., of cat or dog). However, the size of the particle is very similar to pollen grains. A small proportion (i.e., 5–15%) of the fecal particles that are breathed through the mouth would be expected to enter the lungs, and we assume that these particles are assumed to produce the inflammatory response (Bates et al., 1966; Svartengren et al., 1987). At present the question of how allergen particles enter the lungs is not resolved, and this issue is of considerable importance since it may well define the distribution of “inflammation” and airway obstruction.

Association with Sensitization

Many studies have shown an association between dust mite sensitization and asthma (Platts-Mills et al., 1997). These studies include case-control studies in clinics, emergency rooms, and hospitals (Gelber et al., 1993); population-based studies in schools (Peat et al., 1996; Squillace et al., 1997); and prospective studies (Sears et al., 1989; Sporik et al., 1990). The results consistently show odds ratios for asthma of 6 to >12 in individuals with dust mite sensitization. However, in all studies there are a significant number of individuals who are skin test positive but not symptomatic. Using multiple regression analysis, a prospective study in New Zealand demonstrated that mite sensitization was an independent risk factor for asthma (Sears et al., 1989). In the same analysis, pollen sensitization was not significantly associated with bronchial reactivity. This pattern has now been seen in many studies and in several different countries (Table 5-1). Furthermore, similar results have been found in areas where other indoor allergens are most important (Sporik et al., 1995). That is, the relationship between sensitization and asthma is not for allergens in general but specifically for those allergens to which patients are exposed perennially in the environment. Although most of the perennial allergens are found indoors, the airborne fungus Alternaria is also an important factor in areas of the world where exposure persists for many months of the year (Halonen et al., 1997).

TABLE 5-1

Sensitization as a Risk Factor for Asthma (symptomatic bronchial hyperreactivity).

In many of these studies the odds ratios for asthma for individuals who are allergic to dust mites are very high (i.e., ≥6) (Table 5-1 ). However there are several features of the prospective and case control studies that need to be emphasized. In all reported studies, there were a significant number of individuals who were skin test positive to dust mites (or other indoor allergens) but did not have symptoms. In many countries or areas, dust mites are the most important source of allergens. In the south, the south-east, and the west coast of the United States, as well as in New Zealand, Australia, Japan, and the UK, mites of the genus Dermatophagoides are dominant. By contrast in areas where dust mites do not thrive, other allergens are important (e.g., animal dander in the mountain states of the United States and in Scandinavia; or cockroach in the apartments of Chicago, Boston, New York, and Philadelphia). In addition the studies shown in Table 5-1 illustrate the fact that in most areas sensitization to outdoor pollens is not significantly associated with asthma.

Sensitization to storage mites (Lepidoglyphus destructor, Tyrophagus putrescentiae, Acarus siro, and others) is an important risk factor in occupational asthma in agricultural settings (ATS, 1998). These mites may also play a role in nonoccupational asthma from exposures in indoor environments in rural and perhaps other areas, although research characterizing exposure levels and responses is lacking.

Association with Exposure

Sampling. The measurement of allergen entering the lungs or even the measurement of dust mite allergen inhaled has proved very difficult. The essential problem is that the allergen is carried on particles that behave aerodynamically as if they are 10–25 µm in diameter. This means that in still air, they remain airborne only for a few minutes. Thus, all measurements of airborne dust mite have required artificial disturbance, and it has not proved possible to standardize disturbance (Sakaguchi et al., 1990b; Tovey et al., 1981b). Because of this, other measurements of mite allergen have been used as an index of exposure. The accepted measurement is the concentration of Group 1 mite allergen, which is Der p I + Der f I in house dust expressed as micrograms per gram of dust. The immunoassays available use monoclonal antibodies in a two-site enzyme-linked immunosorbant assay (ELISA). The significance of this measurement has been endorsed by three international workshops (Platts-Mills et al., 1997). While there are many different ways of obtaining samples, the most widely used technique is to sample reservoir dust from bedding, bedroom floors, and the living room with a hand-held vacuum cleaner.

In areas of the world that are humid persistently or for at least eight months of the year, mites flourish inside houses, and it is not unusual to find 100 or even 500 mites per gram of dust. This translates to 2 or 10 µg of Group 1 allergen per gram of dust, respectively. In areas where most houses (i.e., at least 80%) have greater than 2 µg of mite allergen per gram of dust, sensitization to these allergens has consistently been found in a large proportion (45–85%) of children with asthma. In some studies the concentration in individual houses has been shown to be an important predictor of sensitization (Peat et al., 1996; Sporik et al., 1990). In Germany, Kuehr and his colleagues demonstrated that the concentration necessary to convert “non-atopic” children to positive skin tests to dust mite was ~60 µg/g. In the same study, children who already had a positive skin test to another allergen were at risk of becoming sensitized with exposures greater than 2 µg/g (Kuehr et al., 1994). Taken together, the results strongly support a dose–response relationship between exposure and development of sensitization, with an approximate threshold of 2 µg/g. This threshold is a statistical concept representing the concentration at which exposure to mite allergen becomes a risk for sensitization. However, there is no sense in which the threshold for mite exposure is comparable to thresholds for airborne toxic gases: (1) high concentrations of mite allergen are not toxic to nonallergic individuals; (2) concentrations below this level may cause symptoms for highly allergic or highly reactive asthmatics; and (3) the measurement is an index of exposure, which may not reveal high levels of exposure elsewhere in the house and has only an indirect relationship to exposure of the lungs.

Experimental Evidence

The earliest experiments on bronchial provocation confirmed that allergens inhaled into the lungs could produce an asthmatic response (Blackley, 1873). With the development of the nebulizer it became much easier to challenge the lungs. The exposure represents ~10⁸ droplets with 0.1–10 µg Der p I per milliliter in a 2-minute challenge giving approximately 0.5 mL inhaled. In allergic individuals this challenge consistently gives rise to an immediate fall in FEV₁ and often produces a late response as well (Cockcroft et al., 1979). Following a bronchial challenge, eosinophils are recruited into the lungs in keeping with the events that are thought to occur in asthma. These changes can also be produced by a segmental challenge, and in these experiments the eosinophils may persist in the lung for up to two weeks after challenge (Shaver et al., 1997).

The lungs of patients with asthma are characterized by inflammation and nonspecific BHR. In several different types of studies it has been shown that these changes can be reversed under conditions that include reduction of exposure to mite allergens. Mite allergic children moved from the Netherlands; Marseille, France; or Verona, Italy to sanatoriums in the Alps have consistently shown clinical improvement and decreased nonspecific BHR. These sanatoriums have many features that are different from the children's homes, including very low mite allergen levels (Platts-Mills and de Weck, 1988). Piacentini and his colleagues (1996) in Italy have also shown that the children who spend three months in the Dolomite Mountains have a decrease in eosinophils and eosinophil products in induced sputum in parallel with decreased BHR. With adults, similar data have been obtained by moving patients from their homes in London to mitefree hospital rooms. Again, there were many other changes associated with the move to mite-free rooms (e.g., no animals, very low spore counts, increased physical activity as patients felt better). However, it is clear that BHR is reversible in many mite-allergic patients.

Distribution of Mites in Houses

Most houses contain at least three of the four requirements for mite growth: (1) there are multiple sites that can provide a nest for mites (i.e., carpets, sofas, mattresses, pillows, bedding); (2) the presence of humans guarantees an abundant food source in the form of skin scales; (3) in the latter half of the twentieth century, most houses are kept close to the optimal temperature for mite growth. For these reasons, the fourth requirement—humidity—is the major and often the only factor that determines whether a house has high concentrations of mites and mite allergen. In areas where humidity is high for most of the year, mites will grow well and may be found in almost any fabric including drapes and clothing, as well as traditional nests. By contrast, in truly dry areas (e.g., the mountain states [≥5,000 feet elevation]), and in Chicago in winter, mites cannot grow and houses are generally devoid of mites (Rosenstreich et al., 1997; Sporik et al., 1995). It is in areas where humidity is marginal or raised for a prolonged season that the structural features of a house may make a large difference in the concentration of mites (see Box 5-2). The three biggest factors are the position of the living area within a building; the presence of carpeting on unventilated floors, which will both trap water leaks and lead to local condensation; and buildings whose ventilation rates are so low that humidity produced by the occupants will accumulate even though the outdoor humidity is not high. The scale of these effects is well illustrated by the fact that a sample of houses and duplexes in Boston was between two and three times more likely to have “high” (in excess of 10 µg/g) dust mite allergen concentrations than apartments in the same city (Chew et al., 1998).

BOX 5-2

Features of Houses That Can Produce Major Changes in Mite Growth in “Marginal” Climates. Structural Features Ground floor level

Strategies to Control Mite Growth According to Climate

Humid Climates (i.e., eight months of the year or more with outdoor air water content of ≥5 g/g). In these climates, controlling mite growth can be achieved only by air conditioning or reducing the nests for mite growth. Air conditioning to maintain indoor relative humidity below 50% requires tight housing and is expensive in terms of energy costs. Controlling or minimizing nests can be achieved by removing carpets; reducing furniture or using surfaces such as leather; and avoiding unnecessary fabrics such as drapes, soft toys, and excess unenclosed clothing. In these areas of the country, the normal practices for bedding are effective (e.g., covering mattresses and pillows; regular hot [>130°F or 55°C] washing of all bedding).

Areas of Moderate or Seasonal Humidity. Many strategies can be helpful in controlling mite growth in these areas (see Box 5-3). During dry seasons, simply opening windows for one hour per day will ensure removal of humidity from the house (Harving et al., 1994). In addition, moving to an upper-level apartment can dramatically decrease mite exposure. The simple physical strategies normally advised are effective. In particular, the recommendations for bedding, carpets, and reducing furnishings can all help. Chemical treatment of a carpet using benzyl benzoate can be effective in controlling mite growth for a period of months (Hayden et al., 1992).

Critical issues include the nature of the fabrics used to cover mattresses and pillows and whether fabrics that breathe but block passage of mites can prevent colonization of mattresses, pillows, or furniture (Vaughan et al., 1999).

Dry Areas. Here, control of mites can be achieved simply by ensuring ventilation on a daily basis. In these areas, only very tight housing will allow accumulation of humidity within the building. Thus, this is more likely to be a problem in cold areas where the temptation to control heat loss may prevent loss of water produced by the inhabitants of a house. In the Mountain States and the Southwest, mite growth in houses is unusual.

Agent Definition and Biology

Endotoxin is the substance responsible for certain characteristic toxic effects of gram-negative bacteria. The toxic compound, lipopolysaccharide (LPS), is a structural component of the outer membrane of these bacteria, the polysaccharide portion of which represents the antigenic surface (Sonesson et al., 1994). The lipid portion of the molecule (lipid A) is essential for its characteristic toxicity. The outer portion of the polysaccharide (O-specific antigen) varies among serotypes of a single bacterial species. The core polysaccharide and lipid A are conserved within species but vary in structure and composition between species and to a greater extent between genera. Variations in lipid A structure are associated with variations in toxic potency over a wide range, and there is some evidence for qualitative variations in toxicity. Gram-negative bacterial endotoxin should not be confused with Bacillus thuringensis delta endotoxin (Du et al., 1999; Potekhin et al., 1999), a protein from a gram-positive bacterium that has recently been genetically engineered into certain crops.

Factors Influencing Exposure

Because gram-negative bacteria are the natural surface flora for plants and are abundant in soil (Edmonds, 1979), endotoxin is ubiquitous in the outdoor environment, particularly during the growing season. High-level exposures to endotoxin occur when organic dust is generated in agriculture and related industries such as animal feed production and in cotton mills (Rylander and Morey, 1982). Recirculating water systems can also be sources of endotoxin, and high level exposures have been recorded in industries where machining fluids and recirculated wash water are used (Milton and Johnson, 1995; Walters et al., 1994). Humidification systems are also potentially abundant reservoirs for gram-negative bacteria and thus for endotoxin (Flaherty et al., 1984; Rylander et al., 1978). Therefore, even home humidifiers can generate high levels of endotoxin exposure (Tyndall et al., 1995).

Measurement of endotoxin exposure in homes is usually performed with a Limulus amebocyte lysate assay because this bioassay correlates well with endotoxin measured in the rabbit pyrogen assay. However, Limulus-based assays are prone to interference (Milton et al., 1997) and likely underestimate exposures in organic dusts, including house dust, compared to chemical assay (Saraf et al., 1999). Furthermore, endotoxin in organic dust contains a wide variety of lipid A structures, not all of which are equally reactive in the Limulus-based assay (Saraf et al., 1997). These compounds may not stimulate significant production of the inflammatory cytokine interleukin-1 (IL-1), but may be able to stimulate other effects such as reversing tolerance to certain antigens (Baker et al., 1990, 1992) that may be important in directing immune responses.

Endotoxin exposures in homes have not been extensively studied. A doctoral thesis described the association of airborne endotoxin levels with home characteristics recorded on questionnaires (Park, 1999). The strongest predictors of increased endotoxin in living room air were the presence of a dog, signs of mice, a concrete floor in the living room, and mold or mildew in the bedroom during the past year. Dehumidifiers were associated with reduced airborne endotoxin. However, airborne endotoxin levels in homes were similar to those in outdoor air. The particle size distribution of endotoxin in ambient and indoor air is not known. Inhaling typical home air, containing one endotoxin unit (EU) per cubic meter, for 24 hours (a total daily dose of approximately 1 EU in a small child and 10 EU in an adult) may be the source of the low levels of endotoxin present in bronchoalveolar lavage fluid (0.1 ng/ml) (Dubin et al., 1996).

In addition to endotoxin exposure from environmental sources, endogenous endotoxin exposure may arise from infection. It has been suggested that a strong association between periodontal disease and premature birth may be caused by endotoxin exposure from gram-negative bacteria present in subclinical infections (Damare et al., 1997). While periodontal disease is not likely a source of endotoxin exposure in children, other infectious agents common in children, such as Hemophilus sp., may be important sources of exposure, as may commensal coliforms.

Biologic Evidence

It has long been recognized that endotoxin is a potent stimulus for macrophage production of (TNFα), IL-1, and a variety of other cytokines; arachidonic and linoleic acid metabolites, and reactive oxygen species (Rietschel and Brade, 1992). The receptors responsible for binding endotoxin and triggering responses are still in the process of being described. The pathway for endotoxin binding and cell activation includes an opsonin, LPS binding protein (LBP), which presents LPS to CD14 (Wright et al., 1990). After binding LPS–LBP complexes, membrane-bound CD14, attached to myeloid cells via a glycosylphosphatidyl–inositol (GPI) anchor, activates myeloid cells and soluble CD14 activates nonmyeloid (endothelial or epithelial) cells (Pugin et al., 1993). The signal is apparently transduced via the human toll-like receptors TLR4 and TLR2 in myeloid and nonmyeloid cells, respectively (Chow et al., 1999; Kirschning et al., 1998; Ulevitch, 1999; Wright, 1999; Yang et al., 1998). There may also be a role in signaling for heterotrimeric G proteins associated with the GPI anchors for CD14 (Solomon et al., 1998).

CD14 and the toll-like receptors are pattern receptors and are also involved in the recognition of peptidoglycan and lipoteichoic acid, and other microbial products (Cleveland et al., 1996; Dziarski et al., 1998; Schwandner et al., 1999). However, opsonization by LBP appears to confer approximately a hundredfold greater sensitivity to LPS compared with peptidoglycan or lipoteichoic acid (Kirschning et al., 1998; Schwandner et al., 1999; Sugawara et al., 1999).

The monocyte lineage, including macrophages and dendritic cells, has abundant membrane-bound CD14 and exhibits sensitive and prolific responses to endotoxin. Cytokines produced include IL-1, 6, 8, 10, 12; TNFα, INF-β, TGF-β; MIP-1α; CSF-1; and GM-CSF. Expression of MHC-II and B7.1 are also upregulated (Medzhitov et al., 1997; Santiago-Schwarz et al., 1989; Weatherstone and Rich, 1989). LPS also is a B cell mitogen and promotes isotype switching from IgM to IgE in the presence of IL4 (Snapper et al., 1991). Thus, endotoxin not only serves as a potent stimulus to innate immune responses but also serves as a stimulus and bridge to cognitive immunity.

The gene for CD14 maps to chromosome 5q31.1, a candidate region for loci regulating IgE expression. Baldini and colleagues (1999) observed that a C to T transposition in the promoter region of CD14 (bp-159) was associated, in TT homozygotes, with significantly higher sCD14 levels. Among white children with positive skin tests to local antigens, TT homozygotes had lower total IgE levels. Among those with any positive skin test, the TT homozygotes had significantly fewer positive tests.

Much experimentation suggests that the net effect of endotoxin exposure is to promote TH1-type immune responses such as those typically seen in bacterial infections (Baldini et al., 1999; Fearon and Locksley, 1996) (see also Chapter 4). However, endotoxin is noted for the peculiar patterns of response to repeated exposures known as the Schwartzman phenomenon and endotoxin tolerance, for which cellular mechanisms have been described (Berg et al., 1995; Bohuslav et al., 1998). It is clear from mechanistic studies of repeated exposure that endotoxin stimulation of production of IL-10, which suppressed the TH1 response, is at least as important a phenomenon as stimulation of IL-12, which enhances this response. Furthermore, much of the data indicating that endotoxin promotes TH1 responses, except Baldini et al. (1999), come from single-exposure in vitro studies. Human data on TH1/TH2 responses to endotoxin and studies of repeated experimental exposure in vivo are sparse. Two studies suggest that under at least certain circumstances, endotoxin promotes TH2 immune responses. A study of human volunteers injected with 4 ng/kg of endotoxin (Zimmer et al., 1996) demonstrated that IL-12 levels were unchanged following injection but IL-10 levels increased, suggesting that systemic endotoxin exposure promoted a TH2 immune response. A mouse model for periodontal disease, produced by repeatedly injecting LPS at 48-hour intervals, found that after the first several injections, gingival T cells were primarily of the TH1 subgroup, based on in situ hybridization. However, after 10 injections, T cell subgroups changed from a TH1 to a TH2 predominance (Iwasaki et al., 1998). Thus, chronic exposure may produce qualitatively different responses from a single acute exposure. The route of exposure as well as dose rate may be important factors, so that the net effect on development of immune responses following chronic low-level, airborne endotoxin exposure may be difficult to predict.

Epidemiologic Evidence

Several laboratory and cross-sectional epidemiologic studies suggest that asthmatics may be sensitive to the proinflammatory effects of endotoxin at lower levels of exposure than nonasthmatics. After endotoxin exposure, increased bronchial responsiveness to histamine challenge was observed among asthmatics, but not among nonasthmatic adults (Michel et al., 1989). In a cross-sectional study of asthmatic adults, the endotoxin content of house dust was associated with increased asthma severity (Michel et al., 1996).

High levels of endotoxin in agriculture and industry are associated with both acute and chronic effects on respiratory symptoms and lung function, regardless of preexisting lung disease. The increasing body of epidemiologic evidence for respiratory effects of occupational exposure to endotoxin at 10 or more times the ambient outdoor levels has been reviewed by Douwes and Heederik (1997) and Milton (1999). One occupational study suggested that the chronic airways disease associated with daily exposure to high levels of endotoxin is characterized by an increased peak flow amplitude and thus may represent a form of asthma (Milton et al., 1996). However, except for the use of contaminated humidifiers, most homes have airborne endotoxin levels similar to those found in ambient urban and suburban air. Thus, it is not clear what effect if any exposure at these levels can have on persons who do not already possess inflamed airways with increased LBP and sCD14.

If endotoxin influences the development of the asthmatic inflammatory response, it is unclear whether the influence will be as a risk factor or a protective factor; basic laboratory research (see above) suggests that either is possible. Both German and Swiss studies suggest associations between living on a farm and decreased risk of asthma (Braun-Fahrlander, 1999; von Mutius, 1994). The authors suggest that exposure to endotoxin early in life may be protective against asthma development and that there may be a gene-by-environment interaction in creating tolerance. At this time there are no data on home endotoxin exposure and the risk of asthma development in children.

Introduction

There are more than 1,000,000 species of fungi, 200 different types to which people are routinely exposed. Exposure occurs universally, both outdoors and indoors, and is impossible to avoid completely. Fungal exposure (even to one type of fungus) is complex with respect to disease agents and usually includes allergens, irritants, toxins, and sometimes potentially infectious units.

These factors have led to confusion, poorly constructed studies on the role of fungi especially in the area of allergic disease, inconclusive results, and avoidance of the field by the best investigators. However, clearly and unequivocally, fungal exposure does cause allergic, toxic, and infectious disease, and it remains only to document the extent of the problem, factors leading to disease, and approaches for control.

Fungi may play a role in asthma in several ways. The most obvious of these is via fungal allergen exposure that leads to sensitization, perhaps leads to the development of asthma, and exacerbates symptoms in sensitized people. Fungi also contain and release irritants that may enhance the potential for sensitization, potentiate allergen-induced symptoms, and (possibly) exacerbate asthma in nonsensitized people. Finally, fungal toxins could play a role in modulating the immune response and may cause direct lung damage leading to pulmonary diseases other than asthma.

Definition of Agents

Nature of Fungi Fungi are eukaryotic organisms characterized primarily by their filamentous morphology and saprobic life-style. Fungal cells are bounded by rigid cell walls that are usually composed of chitin fibrils embedded in a matrix of (1→3) β-D-glucans and/or mannans. Cell walls may be coated externally with waxes or hydrophilic extracellular polysaccharides that carry various levels of antigenic specificity. To digest food, fungi excrete enzymes into the environment initially as probes to evaluate food availability, then to digest complex carbon compounds. These enzymes are some of the major fungal allergens. While processing organic material, the fungi produce many ancillary metabolites, some of which are highly toxic (antibiotics, mycotoxins). These compounds may accumulate in the fungal body, in spores, and in the environment. The primary mode of fungal reproduction is by airborne spores, which form a major fraction of both the outdoor and the indoor large-particle aerosol. Fungi also colonize manmade environments, releasing both spores and metabolic materials.

Fungal Allergens Fungi produce an enormous array of compounds that are potentially allergenic. Each fungus produces many different allergens of a range of potency. Table 5-2 lists the major defined allergens isolated from fungi. Others have been identified, but they are generally “minor” (i.e., few patients react to them). Many others remain to be identified.

TABLE 5-2

Major Defined Allergens Isolated from Fungi.

Sources and Variability Fungal allergen production varies by isolate (strain), species, and genera (Burge et al., 1989). Different allergen amounts and profiles are contained within spores, mycelium, and culture medium (Cruz et al., 1997; Fadel et al., 1992). In addition, substrate (growth) medium strongly influences the amount and patterns of allergen production. For example, the allergen content of Alternaria spores produced on ceiling tiles probably differs from that of spores produced on dead grass. Fungi release proteases during germination and growth, and fungal extracts contain sufficient protease to denature other allergens in mixtures. This has been demonstrated clearly for Alternaria extracts (Nelson HS et al., 1996).

Cross-Reactivity Patterns Patterns of cross-reactivity among fungal allergens have been examined using in vitro methods in which inhibition of heterologous assays is assumed to indicate cross-reactivity (although nonspecific assay inhibition must be ruled out). RAST is the most often used assay (O'Neil et al., 1990), although allergens derived from many genera of fungi have also been shown to cross-react using IgE immunoblot techniques (Verma et al., 1995). In addition, cross-reactivity has also been inferred from patterns of skin reactivity to fungal extracts. Although this latter case could result from cross-reactivity, it could also happen with multiple exposures leading to (separate) multiple sensitizations (O'Neil et al., 1990). Unfortunately, data on fungal allergen cross-reactivity are inconsistent and appear to depend on the specific strains used and on methods of allergen extraction.

Other Fungal Agents Exposure to some kinds of fungal spores induces inflammatory changes in the lung independent of allergy or sensitization (Rao, 1999; Shahan et al., 1998). MIP-2, lactase dehydrogenase, and myeloperoxidase are released in response to fungal exposure, and blood cell patterns change. These effects are dependent on the type of fungus and on the concentration of spores administered. The role these processes might play in the development or exacerbation of asthma remains unknown.

Glucans Fungal cell walls are composed of acetylglucosamine polymer (chitin) fibrils embedded in a matrix of glucose polymers ((1→3) β-D-glucans). The glucans may be chemically bound to the chitin or may form a soluble matrix (Sietsma and Wessels, 1981). Potent T cell adjuvants, the (1→3) β-D-glucans have been investigated as antitumor agents (Kiho et al., 1991; Kitamura et al., 1994; Kraus and Franz, 1991). They increase resistance to gram-negative bacterial infection by stimulating macrophages and affecting the release of TNFα mediated by endotoxin (Adachi et al., 1994a, 1994b; Brattgjerd et al., 1994; Saito et al., 1992; Sakurai et al., 1994; Zhang and Petty, 1994). Soluble glucans have an effect in the lung similar to that of endotoxin (Fogelmark et al., 1994). Glucans may be involved in the development of fungal-induced hypersensitivity pneumonitis by affecting the inflammation-regulating capacity of airway macrophages. They also probably play a role in organic dust toxic syndrome in workers exposed to dust that includes high concentrations of fungal spores. The possible role of glucans in the development and/or exacerbation of asthma has not been studied.

Mycotoxins Most mycotoxins are cytotoxic and interfere with protein synthesis, causing cell lysis and death. Some mycotoxins are potent carcinogens, and a few affect cell division (cytochalasins) or are estrogenic (zearalenone) or vasoactive (ergot alkaloids). Some cross the blood–brain barrier and affect the central nervous system. Some mycotoxins selectively kill macrophages (Gerberick et al., 1984; Jakab et al., 1994; Nikulin et al., 1996, 1997; Richard and Thurston, 1975; Sorenson and Simpson, 1986; Sorenson et al., 1985, 1986). A toxin produced by Aspergillus fumigatus inhibits macrophage functioning and may play a role in allergic bronchopulmonary aspergillosis (ABPA) by facilitating colonization of the airways of asthmatics (Amitani et al., 1995; Murayama et al., 1996). Gliotoxin, another A. fumigatus toxin, causes fragmentation of DNA, especially in thymocytes, and may facilitate tissue invasion leading to aspergillosis in immunosuppressed patients (Sutton et al., 1996; Waring and Beaver, 1996; Waring et al., 1997). Some fungal components directly cause the release of mediators of inflammation, including cytokines, reactive oxygen metabolites, and chemotactic factors (Shahan et al., 1998). The role of mycotoxins in the development and exacerbation of asthma has not been studied.

Means of Exposure

Measurement of Fungal Exposure Visual observation is the most frequently used approach to estimate potential for fungal exposure, with observational data often obtained from occupant questionnaires (e.g., Brunekreef et al., 1989). Observational data are limited by the fact that fungi are microscopic and are not visible until growth is extensive. Culture of air or dust samples is also often used indoors (e.g., ACGIH, 1999; Burge and Solomon, 1987; Su et al., 1992; Verhoeff et al., 1992). Since the presence of allergens does not always depend on culturability, this type of measure is likely to underestimate actual allergen exposure, as well as (possibly) confounding results with high levels of nonallergenic types. Microscopic identification and counting of spores from air samples (e.g., Delfino et al., 1997) represent the usual approach for outdoor samples and constitute the only currently available method that assesses nonculturable spores. The method is limited by the relatively few species that can be identified and the time commitment required for analysis. Although a few assays are available, no epidemiological studies have emerged that use fungal allergen measures of exposure. Other methods have been proposed that evaluate total fungal biomass (ergosterol, glucan assays) or that identify the presence of specific fungi (PCR techniques). None of these has been used in studies of asthma.

Factors Influencing Exposure

Nature of the Particles Spore walls may be hydrophobic with a waxy outer coat, or hydrophilic with an outer surface of water-soluble polysaccharides. Most fungi release single spores that range in size from 2 to 10 µm. Some spores can exceed 100 µm in length, although aerodynamic diameters are usually much smaller. Some spores are released as chains or clumps. Cladosporium spores are frequently encountered as large branching chains, as well as individual spores, so that particle sizes of Cladosporium units can range from near 1 µm to many hundreds of micrometers in diameter. However, air sampling with particle size-separating cascade impactors indicates that most fungal spores are <5 µm in aerodynamic diameter. Fungal hyphae also become airborne with disturbances such as high winds. Studies are not available that document the natural presence (or absence) of fungal allergens on particles other than intact fungal spores.

Studies of Fungal Aerosol Release Fungal aerosols may be produced through intrinsic spore discharge mechanisms or mechanical agitation (Ingold, 1971). Most indoor release relies on active disturbance. Few studies document actual types and intensities of mechanical disturbance necessary to release spores from surface growth. At least for some types, strong agitation or even direct abrasion is necessary (e.g., Stachybotrys chartarum), whereas for others, air movement such as that produced by a fan may be sufficient (Madelin and Johnson, 1992).

Outdoor Exposure to Fungi Many studies document the almost continuous presence of fungal spores in outdoor air, and the factors affecting prevalence for different types (e.g., AAAAI, 1998; Cross, 1997; Li and Kendrick, 1995; Munuera et al., 1998; Takahashi, 1997). Fungal spores are always present in outdoor air, although levels can be very low during periods of snow cover (Cross, 1997). Patterns of prevalence depend on seasonal and climatic factors, geography, and to some extent, human activity, especially that associated with agriculture. Because fungi are continuously present outdoors, often in concentrations far higher than those indoors, it is difficult to rule out outdoor exposure as the primary determinant of sensitization and symptoms. It is also difficult to document the presence of indoor growth using air sampling.

How Does the Indoor Environment Affect or Influence Exposure?

Penetration of Outdoor Aerosols Fungal particles probably penetrate indoor environments following the same physical principles as other types of particles. Published studies that compare indoor–outdoor relationships for fungi have not clearly addressed the relative contribution of the outdoor aerosol. Although outdoor fungal spores readily enter through open windows, few penetrate into closed environments (e.g., buildings and automobiles) (Muilenberg et al., 1991; Solomon et al., 1980).

Indoor Sources Fungi are always present in dust and on surfaces. Fungal growth occurs only in the presence of moisture. Food materials and temperature affect the amount of water required, as does the strain of fungus. Several reports present data on laboratory conditions that lead to fungal growth on building materials (Chang et al., 1996; Grant et al., 1989). With relatively concentrated inocula, some xerophilic fungi increase at relatively low substrate water activities. However, extensive growth was reported only at humidities near 100%, and most fungi require very wet conditions (near saturation), lasting for many days, to extensively colonize an environment. Human habits may affect the numbers and types of active fungal sources in buildings. Poorly maintained water-based appliances (e.g., humidifiers, vaporizers) harbor and release fungi. Wood stored indoors contains many fungi, although exposure from this source has not been studied. Moldy food is an obvious source, and indoor recoveries from air may be dominated by these small sources if recently disturbed.

Sensitization to Fungal Allergens

Exposure to fungi clearly plays a role in asthma. Good-quality studies have been reported that document the sensitizing potential of fungal allergens and relate fungal sensitization to the existence of asthma. Challenge studies have documented asthmatic responses in sensitized patients, and several studies have begun to make the connection between natural exposure and symptom development.

Prevalence of Sensitization

It has been estimated that about 6–10% of the population and 15–50% of atopics are sensitized to fungal allergens (Table 5-3). In studies where Alternaria extracts were used alone (i.e., not in a mixture), overall rates for allergy patients range from 3 to 36% and for asthmatics, 7 to 39%.

TABLE 5-3

Immediate Skin Reactivity to Fungi.

Skin test surveys most often focus on broad panels of allergens, with fungi included either as a single representative extract (usually Alternaria), two or three extracts, or mixtures of several different fungi. However, Galant et al. (1998), in a survey of California allergy patients, revealed that nine different fungal extracts were necessary to detect 90% of mold allergy patients. Interestingly, most studies that focus specifically on fungal sensitivity also use only a few extracts, with Alternaria again being the dominant type. One study (Szantho et al., 1992) suggests that the prevalence of sensitivity to fungi is increasing and attributes the increase to an increase in concentrations of outdoor fungi due to growth on pollution-weakened plants. However, no monitoring data support this hypothesis. Fungal skin sensitivity rates increase with age (Erel et al., 1998). Production of IgG antibodies as a result of allergen exposure may block the skin test response even in patients clearly experiencing symptoms with exposure (Witteman et al., 1996). Fungal allergens can produce a strong IgG response, possibly making reported incidences of skin reactivity underestimates.

Fungal Sensitization and Asthma

In a population of adults in Sweden, sensitivity to Cladosporium or Alternaria (but not dust mites) was associated with a current asthma odds ratio (OR) of 3.4 (1.4–8.5) (Norbäck et al., 1999). In an Arizona population, responses to A. alternata were the most frequent among asthmatic children, and a positive Alternaria skin test was the only independent association with asthma (Halonen et al., 1997). In a University of Virginia study, Alternaria sensitivity was also a significant independent risk factor for asthma in school children in Charlottesville, Virginia, and Los Alamos, New Mexico, but not in Albemarle County, Virginia (Perzanowski et al., 1998). Of the 1,218 children born on the Isle of Wight, 6% of the 918 4-year-old children tested had positive skin tests to Alternaria and/or Cladosporium. In these 61 children, a positive test to Alternaria was associated with a diagnosis of asthma (Tariq et al., 1996). In Costa Rica, where the prevalence of childhood asthma is 20–30%, dust mite, cat, Alternaria, and Cladosporium-specific IgEs predict asthma (Soto-Quiros et al., 1998). A case-control study in Denver (acute asthmatics 3–16 years old) versus matched controls used RAST to document specific IgE levels: 45% of asthmatics versus 4% of controls had high IgE to Alternaria (Nelson RP et al., 1996). In a large (6,394 children) questionnaire-based study, Peat et al. (1995) reveal that among asthmatic children, Alternaria sensitivity rates are higher inland in New South Wales, but in the damp coastal climate, sensitivity to dust mites is most prevalent. In 4,295 6- to 25-year-olds in the second National Health and Nutrition Examination Survey (NHANES II) cohort, asthma was associated with sensitivity to dust mite (OR = 2.9 [1.7–5]); and Alternaria (OR = 5.1 [2.9–8.9]) (Gergen and Turkeltaub, 1992). Retrospective investigation of asthma deaths in teenagers revealed that a large majority had positive skin tests to Alternaria, and the authors concluded that sensitivity to Alternaria is a risk for severe asthma and death from an attack (O'Hollaren et al., 1991). (The authors concluded that exposure to Alternaria is a risk factor, although they did not measure exposure.) Many studies focusing on asthma ignore fungi completely (e.g., Gottlieb et al., 1996).

Immunotherapy Data

Further evidence for the role of fungal sensitization in symptomatic asthma is provided in immunotherapy data. Bousquet and Michel (1994) have reviewed the overall literature on immunotherapy. The literature on mold immunotherapy has been reviewed by Dhillon (1991) and Bonifazi (1994). Immunotherapy with partially characterized and standardized A. alternata allergens in a small case-control study has been shown to decrease asthma symptoms overall and during challenge with relevant allergens (Cantani et al., 1988; Horst et al., 1990). In both of these studies, children who were sensitive only to Alternaria (by skin prick test) were treated. Dreborg et al. (1986) selected 30 children who were skin test, RAST, and provocation positive to Cladosporium. Most had other sensitivities as well. He treated 16 of these children; the remaining 14 served as controls. Although symptom scores remained similar for the two groups (probably because of multiple sensitivities), nasal and bronchial challenge responses and medication usage were significantly lower in the treated versus the control group.

Of the 16 children treated in the Dreborg (1986) study, 13 experienced general reactions during treatment. Kaad and Ostergaard (1982) report that 19% of 38 children treated with fungal extracts had to be withdrawn from immunotherapy because of apparent type III reactions. All had slightly elevated IgG (precipitating antibodies) to the relevant extracts before hyposensitization was begun and developed increased titers during therapy.

Overall, carefully controlled studies of mold immunotherapy show good effect. Negative studies are small, with too few patients and nonstandardized allergens. These data support the notion that there is a relationship between fungal allergen sensitization and symptoms of asthma. However, there are too few studies to confirm that mold immunotherapy is an effective public health intervention.

Fungal Exposure and Asthma

Challenge Experiments Challenge tests with Stemphylium in children with positive skin and RAST tests to Stemphylium extracts resulted in bronchial responses in 13 of 59 children (12 others had nasal responses) (Lelong et al., 1986). Malling (1986) used positive bronchial challenge in adult asthmatics with Cladosporium as a patient selection criterion for his studies on relationships between natural exposure and symptoms. Licorish et al. (1985) reproduced symptoms of asthma with Penicillium spore challenges.

Measured Natural Exposure

Outdoors Neas and colleagues (1996) associated pulmonary function changes in Pennsylvania school children with measured outdoor concentrations of several kinds of fungal spores. Peak flows, symptoms and time spent outdoors were recorded daily, and 24-hour average outdoor spore concentrations were measured using a Burkard spore trap. Changes in morning peak flow of –1 liter/min (CI –1.9 to 0.2) were associated with increments of 10,000 Cladosporium spores in unselected school children. Likewise, increases of 50 Epicoccum spores was associated with decrements in morning peak flows of 1.5 liter/min (CI –2.8 to –0.2). Epicoccum exposure was also associated with incidence of morning cough (OR = 1.8, CI = 1.0–3.2). Delfino et al. (1997) report associations between asthma severity (symptom scores and peak expiratory flow rates) and exposure to outdoor fungal spores measured with a spore trap. Symptoms were associated with total fungal spore increases of 4,000 spores per cubic meter, with a decrease in evening peak flow of 12 L per minute. Analysis of specific fungal genera increased the associations. The most important fungal correlates were Alternaria, basidiospores, and hyphal fragments. Malling (1986) related daily symptom scores to Cladosporium counts during the autumn mold season. He studied 24 adult asthmatics with positive provocation tests to Cladosporium extracts. Significant associations with spore counts were found for symptom scores and medication use. Although some patients were also sensitive to Alternaria, Alternaria counts did not relate to health outcomes in this population.

Indoors Su et al. (1992) describe relationships between indoor measured levels of culturable Aspergillus and asthma in Topeka, Kansas schoolchildren. They used culture plate impactors to collect short “grab” samples in homes of children recruited through the public schools. The relationship between Aspergillus and asthma symptoms was non-linear, possibly reflecting the different species of Aspergillus that are common in indoor environments. Garrett et al. (1998) report associations between measured concentrations of fungal spores (counted) and culturable fungi, observed indicators of dampness and molds, fungal sensitivity, and respiratory symptoms. Eighty households with 148 children (36% asthmatic) were sampled 6 times for airborne fungi. Penicillium exposure was associated with asthma, and Aspergillus exposure with atopy. Cladosporium and Penicillium exposure were associated with the presence of fungal allergy. No associations were seen for total fungal counts. Exposure measures to specific fungi showed a stronger association with symptoms than did dampness indicators.

Other studies of indoor fungi and asthma report comparisons between fungal levels in homes of asthmatics compared to homes of control subjects. Li and Hsu (1997) compared culturable fungi in homes of 46 asthmatic children, 20 atopic (presumably nonasthmatic) children, and 26 nonatopic controls. Although fungal concentrations were highest in the asthmatic and control homes, Cladosporium concentrations were higher in asthma homes than in control homes. In a study by Horak et al. (1996), Bacillus, Aspergillus, and total fungal concentrations were higher in homes of asthmatics than in control homes. These kinds of studies are compromised by the fact that families of asthmatic children often take measures to reduce exposure to allergens.

Visible Mold, Odors, and Dampness

The majority of studies that attempt to associate fungal exposure and asthma rely on reports of visible mold or other dampness indicators. Dampness is clearly associated with a variety of respiratory symptoms.

In addition, Williamson et al. (1997) report a correlation between mold growth indicators and asthma (196 age- and sex-matched subjects); r = 0.23, p < .035). These researchers also identified a relationship between severity of asthma and total dampness (r = 0.3, p = .006), and mold growth (r = 0.23, p = .35). A reduction in FEV₁ of 10.6% (CI 1.0–20.3) was associated with living in a damp home.

Dampness and even visible mold growth could be indicators for dust mite allergen exposure as well as fungal exposure. However, Garrett et al. (1998) report that homes with musty odor, water intrusion, high humidity, limited ventilation, and visible mold growth had higher fungal spore concentrations than dry homes. Visible mold was associated with high concentrations of Cladosporium spores, but not total spores. Pasanen et al. (1992) report that airborne Cladosporium and yeast counts were significantly higher in damp than in dry residences. Cladosporium, in particular, appeared to be derived from indoor sources. Nicolai et al. (1998) report that the relationship between dampness and bronchial hyperreactivity in adolescents in Munich remains when controlled for dust mite allergen exposure, providing indirect evidence for a possible role of fungi.

Definition of the Agent and Means of Exposure

Rhinovirus is the medical term used to designate a large group of viruses responsible for a variety of respiratory infections including the common cold. There are more than 120 serotypes of the rhinovirus, of which few have been associated with lower-respiratory abnormality. Rhinovirus is transmitted through the respiratory route, with virus particles shed in nasal secretions and spread by coughing and sneezing. It can also be transmitted through direct contact with tissues or hands.

Evidence Regarding Asthma Exacerbation and Development

Widespread clinical experience and epidemiologic research indicate that rhinovirus is associated with wheezing and exacerbations of asthma in established asthmatics (Johnston, 1997; Johnston et al., 1995; Lemanske et al., 1989; Micillo et al., 1998; Rakes et al., 1999). The evidence of this is especially strong for children. The mechanism by which the virus traffics from the upper airway to the lower airway to cause asthma symptoms is not known. It is speculated that individuals with asthma may make more of a specific cytokine than those without asthma, leading to increased airway inflammation and the development of symptoms. Alternatively, individuals with asthma and those without asthma may make the same amount of cytokines in response to viral infections, but these cytokines may react differently in individuals with asthma and produce different effects in the airway. Individuals with asthma have a different cytokine profile from those without asthma, which may result in a different clinical phenotype.

Studies have not identified an association between rhinovirus infection and asthma development. Martinez (1995) observes that most infants who wheeze during viral infections become symptom free later in life and suggests that viral infections in infants are likely to play a minor role in the subsequent development of asthma. There is burgeoning research interest in this issue, however, with investigators suggesting both the possibility of a mechanistic association and the prospect that certain viruses may have a protective effect (Grunberg and Sterk, 1999; Martinez, 1994).

Definition of the Agent and Means of Exposure

Respiratory Syncytial Virus (RSV) belongs to the Paramyxoviridae family and to the genus Pneumovirus. Human paramyxoviruses are a common cause of respiratory disease in children, with RSV particularly important in infants. RSV has been found in every geographical area examined for evidence of infection (Glezen et al., 1981; Hall et al., 1979), and RSV infection appears to have similar characteristics in areas with widely differing climates. Outbreaks of infection occur yearly, typically in the spring and winter. Indeed, RSV is the only viral respiratory agent that can be relied on to produce a sizable crop of infection each year. Socioeconomic status and race or ethnicity appear to influence the risk of RSV bronchiolitis (Glezen et al., 1981).

Evidence Regarding Asthma Exacerbation and Development

The study of RSV and asthma exacerbations is complicated by the problematic diagnosis of asthma in infants and young children. However, viruses in general have been identified as responsible for most wheezing illnesses and asthma exacerbations occurring in childhood, with RSV among the predominant organisms recognized (Hegele, 1999; Pattemore et al., 1992). An association has also been noted for adults (Nicholson et al., 1993).

Significant attention has been paid to the possibility of an association between RSV bronchiolitis in infancy and later development of asthma. RSV bronchiolitis is associated with the development of high titers of virus-specific IgE in respiratory secretions during both the acute and the convalescent stages of illness in some patients (Robinson et al., 1993; Welliver and Duffy, 1993). An exaggerated RSV IgE response at the time of RSV bronchiolitis in infancy is associated with recurrent wheezing in later childhood (Scott et al., 1984). Various mechanisms may be proposed to explain this association. These include direct induction of release of inflammatory mediators from pulmonary mast cells, participation in a cascade of mediators of airway obstruction, reflection of general IgE hyperresponsiveness to multiple allergens, and induction of persistent T helper type 2 responses (Roman et al., 1997). Although an exaggerated RSV IgE response to RSV infection during infancy may correlate with recurrent wheezing during a child's early years, this may not represent asthma per se but rather repeated similar responses to multiple infections with RSV and other respiratory pathogens (Welliver et al., 1986). A study by Sigurs and associates (1995) tends to support the hypothesis that RSV infection may cause or promote asthma by inducing atopy in infected individuals. Since the rate of sensitization among controls with and without IgG antibodies to RSV is similar, an encounter with the virus without development of bronchiolitis does not seem sufficient to increase the risk of sensitization. Other investigators have reached different conclusions about the relationship between RSV infection in infancy and the development of atopy and asthma later in childhood. In a study by Pullan and Hey (1982), the frequency of positive skin tests for common allergens was similar for individuals with previous bronchiolitis and for controls. Murray and colleagues (1992) found no difference in skin test reactivity to dust mites between bronchiolitis patients and controls.

A precise mechanism by which RSV bronchiolitis might induce allergies and asthma has not been elucidated (Dezateux et al., 1997). However, one possible pathogenic mechanism could be through increasing the synthesis of IgE since it has been demonstrated that RSV bronchiolitis can induce an IgE antivirus response and that RSV-specific IgE responses in infancy are associated with later recurrence of wheezing. Welliver and colleagues (1981) found that IgE titers to RSV in nasal secretions were highest in patients with evidence of airway obstruction. Chang and colleagues (1990) prospectively followed 38 of these infants for 48 months after an initial episode of RSV bronchiolitis. Only 20% of infants with undetectable titers of RSV IgE had subsequent episodes of documented wheezing. In contrast, 70% of children with high RSV IgE antibody titers experienced wheezing. In a four-year prospective study of 13 infants of history-positive bilaterally allergic parents, Frick and colleagues (1979) noted that viral infections including RSV occurred one to two months before the onset of allergic sensitization in 10 of the 11 children who subsequently became atopic.

Despite the strong correlation between virus-specific IgE, mediator release, and the clinical pattern of acute and recurrent disease, it is premature to conclude that virus-specific IgE is causal in this process. Moreover, whether those infants who had persistent wheezing and/or airway obstruction were concurrently both sensitized and exposed to indoor allergens is not known.

Definition of the Agent and Means of Exposure

There are two forms of the bacterium Chlamydia that have been examined in relation to asthma. Chlamydia trachomatis is a sexually transmitted infectious disease that may be spread from mother to child during birth and perhaps through close contact with secretions of an infected individual. Chlamydia pneumoniae is an important cause of adult respiratory disease including pneumonia, bronchitis, sinusitis, and pharyngitis and may be associated with atherosclerosis (Kalman et al., 1999). A 1991 editorial in the Journal of the American Medical Association suggested that C. pneumoniae infection might provide an explanation for the increased incidence of asthma (Bone, 1991).

Evidence Regarding Asthma Exacerbation and Development

An anecdotal report suggests that C. trachomatis produces asthma-like symptoms in infants that may be treated successfully with antichlamydial antibiotic therapy (Bavastrelli et al., 1992). Björnsson and colleagues (1996) report that serological signs of a previous C. trachomatis infection were found significantly more often in subjects who reported having had asthma at some time, asthma during the past year, wheezing during the past year, and bronchial hyperresponsiveness. It is unclear, however, whether such reports uniquely implicate C. trachomatis or are unrecognized C. pneumoniae infections since the two species are cross-reactive.

Chlamydia pneumoniae infection has been associated with asthma exacerbations in both adults and children. Björnsson and colleagues (1996) found a statistically significant relationship between current or recent C. pneumoniae infection and wheezing. Johnston (1997) found that immune responses to C. pneumoniae were positively associated with an increased frequency of asthma exacerbations in a group of 9- to 11-year-olds. Cunningham and colleagues (1998) report increased numbers of asthma exacerbations in children with higher local antibody responses to C. pneumoniae. Miyashita and colleagues (1998) found a significantly higher frequency and geometric mean titer of C. pneumoniae antibodies in adult patients with exacerbations of asthma. However, Cook and colleagues (1998) did not find an association between asthma exacerbations (termed “acute asthma” in the study) and seropositivity in a two-year study of 123 adult asthmatics and 1,518 controls.

Several groups of researchers have reported a serological association between C. pneumoniae and chronic adult asthma in studies that controlled for one or more known confounders. Hahn and colleagues (1991) found that patients with serologically confirmed C. pneumoniae infection were more likely than seronegative patients to develop bronchial asthma subsequent to their infection. A later study of 163 adolescents and adults by Hahn and McDonald (1998) had the same findings, leading the authors to conclude that acute C. pneumoniae respiratory tract infections in previously unexposed nonasthmatic individuals can result in chronic asthma. Cook and colleagues (1998) found a statistically significant association between severe chronic asthma and seropositivity in their two-year study of 123 adult asthmatics and 1,518 controls. Johnston (1999) reports a high prevalence of C. pneumoniae infection in a group of 9- to 11-year-old asthmatics. Von Hertzen and colleagues (1999) found asthma to be significantly associated with IgG antibody levels to C. pneumoniae, with the strongest association for nonatopic, long-standing asthma.

However, not all studies find this association. Kraft and colleagues (1998) did not find C. pneumoniae in the lungs or airways of 18 asthmatic subjects. Larsen and colleagues (1998) found that a group of 22 asthmatics did not differ from 55 controls with relation to two markers of C. pneumoniae infection. The authors note two weaknesses in many of the studies reporting a relationship: (1) a diagnosis of asthma was seldom well-established in these cases, and (2) no pathobiologic mechanism was proposed to account for initiation in adults (production of specific IgE was noted as a possible initiating mechanism in children).

At present (late 1999) there is a debate in the literature over the meaning of these studies. Although their results are consistent with an association between C. pneumoniae (and perhaps C. trachomatis) and asthma development, present data are insufficient to distinguish this premise from other reasonable hypotheses, notably whether asthma predisposes individuals to other chronic respiratory infections.

Definition of the Agent and Means of Exposure

Mycoplasma pneumoniae is a bacterial infection that is responsible for a number of respiratory diseases including tracheobronchitis, rhinitis, pharyngitis, otitis, and a form of pneumonia. Over the past several years there have been multiple anecdotal reports and epidemiologic studies suggesting an association between M. pneumoniae infection and exacerbations of asthma (e.g., Seggev et al., 1986). Among the more recent studies, Freymuth and colleagues (1999) found that C. pneumoniae and M. pneumoniae infections—alone or in combination—were present in nearly 82% of asthmatic exacerbations. However, Cunningham and colleagues (1998) and Johnston (1997, 1999) found no evidence of a role for M. pneumoniae in acute asthma exacerbations.

Evidence Regarding Asthma Exacerbation and Development

The literature regarding M. pneumoniae and asthma development is sparse. Yano and colleagues (1994) suggested in a case report that an adult male may have developed asthma as a result of M. pneumoniae infection. Kraft and colleagues (1998) found M. pneumoniae present in the lower airways of chronic stable asthmatics with significantly greater frequency (10 of 18) than in controls (1 of 11).

Houseplant Pollen

The data on houseplant pollen allergy is restricted to the occupational literature and to case reports. Allergy to some flower pollens has been clearly documented in occupational situations and could occur in residential environments with cut flowers (de Jong et al., 1998). Co-sensitization with some houseplant pollen allergens and some outdoor allergens is extremely common (e.g., mugwort, daffodil) and may indicate that these plants share allergens. In cases of this sort, asthma related to outdoor pollen exposure could be exacerbated by houseplant pollen, although very little pollen is produced by most houseplants and close contact would be required.

Other Plant Parts and Fluids

Exposure to plant materials during the production of plant extracts as medicines has led to occupational asthma (Giavina-Bianchi et al., 1997). The expanding use of herbal medicines may lead to an increase in these types of exposures. Several papers report cases of houseplant (in particular, Ficus) latex causing contact sensitization and even inhalation allergy. The available data are restricted entirely to the case study literature, and only two reports were retrieved (Diez-Gomez et al., 1998; Schenkelberger et al., 1998).

Arthropods

Arthropods that inhabit plants (particularly spider mites) may release allergens. However, again, all data are in the occupational case study literature, and no information is available on the extent of the problem (Orta et al., 1998). The potential for cross-reactivity with dust mite allergens has not been evaluated.

Fungi

Several reports implicate houseplants as sources of fungi in indoor environments, with a focus on human infectious agents rather than allergens. These studies, in general, rely on source (soil) sampling and do not relate exposure to any allergic disease (Staib, 1992, 1996; Staib et al., 1978a, b, 1980; Summerbell et al., 1989). Burge and colleagues (1982) found no relationship between the presence of houseplants and either concentration or type of fungal spores or pollen in the air.

Introduction

Pollen exposure has long been recognized as a stimulant for symptoms of allergic disease. Allergic rhinitis (hay fever) has been considered the primary outcome, and outdoor environments are the major source of exposure. The role of pollen in the development and exacerbation of asthma has received relatively little direct attention, and very few studies are available that document indoor exposure to pollen allergens.

Nature of the Agent

Structure and Function Pollen is the male reproductive structure of flowering plants and gymnosperms. Pollen grains are usually unicellular, with a rigid and highly resistant cell wall formed of a complex polysaccharide-based substance called sporopollenin. Each pollen grain contains the systems required for recognition of genetically relevant female flowers and the production of a pollen tube that grows into contact with the egg cell to effect fertilization.

Pollen Allergens

The list of pollens from which extracts have been derived that produce positive skin tests in some fraction of the population is long. In one California study, 57 skin test allergens were necessary to detect 90% of the atopic patients. This set included 2 grasses, 16 weeds, and 27 trees (Galant et al., 1998).

The list of purified allergens from pollens is also growing. These include Amb a I (short ragweed, Ambrosia artemisiifolia), Bet v I (birch pollen, Betula verrucosa), and Lol p I (ryegrass, Lolium perenne). One study has estimated that a single pollen grain of B. verrucosa contains 0.006 ng Bet v I (Schappi et al., 1997).

Other Pollen-Associated Agents

Serine proteases that could directly affect respiratory function apart from allergic reactions have been detected in ragweed and mesquite pollen (Bagarozzi and Travis, 1998; Travis et al., 1996).

Characteristics of Pollen Exposure

Particle Size Considerations Intact pollen grains range from about 10 to 100 µm, with the most common types in the range of 15–30 µm. However, pollen allergens have been documented in air on much smaller particles. During light rain, 1.2 ng Bet v I/per cubic meter (200 birch pollen equivalents) were present on the particle fraction less than 7.5 µm in outdoor air (Schappi et al., 1997). Grass pollen grains have been shown to rupture during rainfall, and allergen (Lol p V) has been recovered in particle fractions less than 5 µm both as free molecules and attached to other particles such as starch grains and combustion (diesel exhaust) particles (Suphioglu, 1998). Pollen allergens have been visualized on the surface of diesel particles under laboratory conditions (Knox et al., 1997). Ragweed allergens have also been recovered from small particles (Habenicht et al., 1984; Solomon et al., 1983).

Patterns of Pollen Prevalence

Outdoors An enormous body of literature documents prevalence patterns for pollens throughout the world. In the United States, the American Academy of Allergy, Asthma, and Immunology has been collecting pollen prevalence data for more than 30 years and currently publishes an annual report containing data from about 100 stations across the nation (e.g., AAAAI, 1998). In the United States, the pollen types most often considered important are mountain cedar, birch, and oak (trees); grasses (which cross-react broadly across genera); and ragweed (a late-summer weed).

Pollen is produced seasonally. In general, tree pollens are released early in the year, grasses during late spring and early summer, and weed pollens in the late summer and fall. Major exceptions occur. For example, some grass pollen is produced throughout the year in some areas.

Meteorological variables directly affect pollen concentrations in air (in addition to their effects on pollen production). Most pollen types are released during the morning hours, but dispersal may occur later in the day with increase in wind and other disturbances. Wind is a well-known dispersal agent for pollens, which may travel long distances (hundreds of miles) in traveling air masses (Levetin and Buck, 1986). Rain generally washes pollen from the air.

Indoors Outdoor infiltrate is the primary source of pollen in indoor environments. Birch pollen allergen (Bet v I) has been found associated with suspended particles indoors (Holmquist and Vesterberg, 1999; Ormstad et al., 1998).

Pollen Exposure and Asthma

Sensitization to Pollen Allergens In 3,371 Canadian allergy patients, skin tests indicated that 52% were sensitive to grass allergens and 45% to ragweed allergens (Boulet et al., 1997). In the Swiss population (random, more than 8,000 individuals studied) 32% were atopic, with sensitivity to grass allergen the most prevalent (12%), followed by dust mite (8.9%) and birch pollen (7.9%) (Wuthrich et al., 1995). In 1,159 random participants in Hamburg and Ehrfurt, respectively, grass and birch allergen sensitivities were 24 versus 19% and 19 versus 8% (Nowak et al., 1996). Of inner city asthmatics in Chicago, 45% were sensitized to ragweed pollen (76% were sensitive to indoor allergens) (Kang et al., 1993). In 7,079 retrospectively observed patients with asthma and/or rhinitis, 44% had one or more positive skin tests, with grass, cat, and birch allergens eliciting the greatest number of positive tests. Of the 35% of sensitized patients who were monosensitized, 7.4% reacted only to mite allergens and 7.0% to grass allergen (Eriksson and Holmen, 1996). On the other hand, Subiza and colleagues (1994) report that 92% of Madrid allergy patients are sensitive to grass allergens, 63% to olive allergens, and 56% to sycamore (Platanus) allergens.

Relationships to Asthma

Pollen Sensitization Related to Asthma In general, pollen sensitization has been associated primarily with hay fever rather than asthma, and this opinion is supported in several studies. In a study reported by Eriksson and Holmen (1996), patients with grass allergen sensitivity were more likely to have rhinitis than asthma. In a Tucson population, sensitivity to ryegrass and mulberry pollen independently predicted rhinitis but not asthma (Halonen et al., 1997). In a case-control study of 343 random children, 35% were atopic, and 90% of those with five or more episodes of wheezing had one or more positive skin tests (Henderson et al., 1995). Pollen sensitivity was not associated with wheezing in this population.

On the other hand, in an Australian population of 745 young adults, sensitivity to ryegrass pollen was among the four top allergen sensitivities that posed a risk for the existence of current asthma (Abramson et al., 1996). Winter respiratory symptoms (including asthma) have been related to positive prick, RAST, or endonasal challenge tests with alder and hazel pollen allergens (Laurent et al., 1994). Grass allergen elicited the most positive skin tests among black asthmatics in Johannesburg (Luyt et al., 1995).

Pollen Exposure and Asthma Ordaz and colleagues (1998) followed symptoms in 104 asthmatic patients and related exacerbation events to airborne pollen counts. When infectious disease-related events were excluded, there was a good correlation between event days and pollen counts (r = 0.7, p < .01). In a prospective study of mild to moderate asthmatics (N = 139), Epton and colleagues (1997) were unable to distinguish a relationship between measured pollen concentrations and measured peak flow.

Baraldi and colleagues (1999) documented a twofold increase in exhaled NO in grass-sensitive asthmatic children during the grass pollen season. Natural allergen exposure was shown to induce T cell, mast cell, and eosinophil inflammatory response in grass-sensitive patients undergoing seasonal exacerbation (N = 17) (Djukanovic et al., 1996). Rosas and colleagues (1998) report an association between grass pollen counts and asthma admissions in Mexico City in both dry and wet seasons. Celenza and colleagues (1996) related changes in grass pollen concentrations and decreases in temperature to epidemic asthma. Their patients, who were involved in a thunderstorm-associated epidemic in England, had high serum IgE levels specific for grass pollen allergens (Venables et al., 1997). Newson and colleagues (1997) counted asthma admissions in two age groups (0–14 and >14 years) and related these events to lightning episodes and grass pollen concentrations. Although there were excess densities of lightning and grass pollen counts in epidemics, many (in fact, most) epidemics were not preceded by thunderstorms. In a follow-up study, they report that thunderstorms following periods of high pollen counts are more likely to lead to asthma epidemics (Newson et al., 1998).

In an analysis of 59,624 asthma hospitalizations in Finland, a peak was observed in May that was attributed to the exposure of sensitized people to birch pollen allergens (Harju et al., 1997). As part of the increase in doctor-diagnosed asthma and episodes of breathlessness in children in Norway, symptoms with exposure to birch pollen increased from 3.7 to 6.1% of children between 1981 and 1993 (Skjonsberg et al., 1995).

Woodcock and Custovic (1998) suggest that exposure to ragweed allergen reduces corticosteroid binding capacity in sensitized asthmatics, contributing to poor asthma control.

In 20 subjects, airway responsiveness to histamine challenge was highest during the privet season, and symptom scores and bronchodilator use increased. However, challenge with privet pollen allergens induced no immediate asthmatic responses, and late responses occurred in only 6 of 17 tested patients (Richards et al., 1995).

Pollen–Air Pollutant Interactions

Laboratory Studies Strand and colleagues (1997) report that in a group of 18 pollen-sensitive asthmatics, short exposure in the laboratory to concentrations of NO₂ representative of those that occur in outdoor air enhances pollen allergen-induced late asthmatic responses. Repeated short exposures to NO₂ enhanced effects of otherwise nonsymptomatic doses of pollen allergen (Strand et al., 1998). However, in similar studies, Hanania and colleagues (1998) found no change in allergen response following ozone challenge.

Field Studies Anderson and colleagues (1998) report a synergistic effect for pollen and SO₂ with respect to hospital admissions of children (but not adults) for asthma. No other pollen–pollutant interactions were observed. In a comparison between randomly selected participants in Hamburg and Ehrfurt, Germany, the prevalence of atopy and pollen sensitization was highest in Hamburg, while air pollutant exposures were highest in Ehrfurt (Nowak et al., 1996).