Respiratory symptoms in relation to indoor exposure to mite and cat allergens and endotoxins

Study design

To enrich the population-based sample with atopic subjects, this study on indoor factors and genetics in asthma (INGA) was designed as a nested case-control study following a cross-sectional study performed from 1990–1992 in the two German cities of Erfurt and Hamburg within the European Community Respiratory Health Survey (ECRHS) 10–12. The methods of the ECRHS are described in detail elsewhere 12. On the basis of the study results of the ECRHS in 1990–1992, cases and controls were defined. Every subject willing to participate attended the centre again for a detailed questionnaire, including respiratory symptoms and determination of specific IgE. The study was performed from June 1995–November 1996 and an identical methodology was followed in the two centres. The study protocol had been approved by the local ethics committees and all subjects gave their informed written consent.

Study population

The study population consisted of a subset of the subjects who had participated within the ECRHS in Erfurt and Hamburg. Cases for the present study were defined as subjects fulfilling at least one of the following four conditions according to the results from the ECRHS (1990–1992): 1) asthma diagnosed by a physician according to the long questionnaire of the ECRHS 12; 2) at least one positive specific serum IgE (>0.35 kU·L⁻¹, CAP system, Pharmacia and Upjohn, Stockholm, Sweden) against grass, birch, cat dander, D. pteronyssinus or Cladosporium herbarum; 3) at least one positive skin-prick test reaction (mean wheal diameter ≥3 mm) against the allergens mentioned above; and 4) provocative dose causing a 20% fall in forced expiratory volume in one second (FEV₁) (PD₂₀) ⪕2.0 mg methacholine or positive bronchodilator test. The subjects of the control group did not show any of these conditions at the ECRHS.

Based on these criteria, 107 cases and 106 controls were recruited from a random list of 538 cases and 621 controls in Hamburg, and 115 cases and 109 controls out of 363 cases and 368 controls in Erfurt. Four-hundred and five of these subjects (204 in Erfurt and 201 in Hamburg, 205 cases and 200 controls) agreed to measurements of indoor exposure to mite (D. pteronyssinus antigen 1, Der p 1, and D. farinae antigen 1, Der f 1) and cat (cat antigen d1 Fel d 1) allergens and endotoxins in their apartments.

Questionnaire

The questionnaire used in the INGA study to asses a history of asthma attacks and respiratory symptoms within the past 12 months, smoking, and social status had been developed according to the validated long questionnaire with 71 items used in the ECRHS 12. The validated German version of the original ECRHS questionnaire was shortened to 40 items without changing the questions or their wording.

Allergic sensitization

Allergic sensitization in the INGA study was assessed by measurement of specific IgE against Alternaria alternata (m6), Aspergillus fumigatus (m3), tree pollen (tx5), birch (t3), C. herbarum (m2), D. pteronyssinus (d1), D. farinae (d2), grass pollen (gx1, gx4), Phleum pratense (g6), weed pollen (wx3), Blatella germanica (i6), dog (e5), and cat dander (e1) using the IgE CAP system. Allergic sensitization was defined as at least one of the specific IgE listed above >0.70 kU·L⁻¹. This differs from the definition used for definition of cases, but the cut-off point of 0.70 kU·L⁻¹ seemed to be more reliable.

Lung function

Spirometric measurements were performed using pneumotachograph-based electronic spirometers. For details, see Richter et al. 13.

Measurement of allergen and endotoxin content of house dust

The homes were visited from June 1995–November 1996 by trained personnel. In each apartment, dust samples were taken from the living room floor, from the bedroom floor, and from the mattress surface according to a standardized protocol 14. All dust samples were taken using the same type of vacuum cleaner (Flüsterjet Vitall 371, Philips, Hamburg, Germany) by vacuuming an area of 1 m² for 2 min at any of the three locations. Dust was collected on cellulose filters using sampling nozzles and filter boxes (ALK Laboratories, Hørsholm, Denmark). Samples were stored in filter boxes at −20°C until extraction. Dust samples were not sieved before extraction.

Sample allergen content was measured by means of a monoclonal enzyme-linked immunosorbent assay (ELISA) 6 with standards UVA 93/03, UVA 93/02, and UVA 94/01 (Indoor Biotechnologies, Clwyd, UK). Allergen concentrations are expressed as ng·gram⁻¹ of dust. The lower limit of detection was 10 ng·g dust⁻¹ for Der p 1 and Der f 1, and 15 ng·g dust⁻¹ for Fel d 1. Endotoxin content in living room floor dust was quantified using a chromogenic kinetic Limulus amoebocyte lysate test described in 15. Escherichia coli endotoxin (lot no. 5L570, Bio Whittaker, Walkersville, UK) was used as the standard endotoxin. The potency of this standard was 14.5 endotoxin unit (EU)·g⁻¹. Endotoxin concentrations are expressed as ng·gram⁻¹ of dust.

Definition of exposure

Exposure to allergens is considered of major importance while sleeping, since people come into closest contact with mites in their beds. Therefore, the analysis of the relationship between allergens in house dust and respiratory symptoms was restricted to allergen concentrations in mattress dust.

Allergen levels were classified into low and high, defining subjects exposed to >10 µg group I dust mite allergen per gram of dust or >8 µg cat allergen per gram of dust as highly exposed, which are the threshold levels supposed to increase the risk of sensitization and asthmatic symptoms 16, 17. Exposure to endotoxins was categorized into three categories defining exposure to <1,100 ng·g dust⁻¹ as low, 1,100–4,700 ng·g dust⁻¹ (25th–75th percentile) as medium and >4700 ng·g dust⁻¹ as high.

Statistical methods

Because of the lognormal distribution of the allergen and endotoxin levels, geometric means (GM) and 95% confidence intervals (CI) are presented to characterize distributions. Concentrations less than the detection limit were assigned a value of one-half of the detection limit. Correlations between allergens and endotoxins were expressed as Spearman correlation coefficients.

The definition of “case” in the present study was found to be inadequate as it was not a well-defined clinical entity. Consequently outcomes other than case/control were to be analysed (i.e. case/control data was not analysed as designed, but as cross-sectional data). In order to do this, it was necessary to show that a relationship between exposure and the respective outcome was not just a consequence of the selection process. This was approached by adjusting for “caseness”.

Associations between allergen concentrations in mattress dust, endotoxin concentrations in living room floor dust and the potentially confounding factors of region, season of dust sampling, age, education, and active smoking were analysed by means of the Wilcoxon two-sample test and the Kruskal-Wallis test, respectively. For the age variable, tertiles were used as cut-off points.

Multiple logistic regression analysis was used to estimate the effects of mite and cat allergens in mattress dust and endotoxins in living room floor dust on the outcomes listed above. Adjusted odds ratios (OR) with 95% CI are presented. Statistical significance was set at a conventional 5% level, and marginal significance at a 10% level.

Description of the study population

A description of the studied sample of subjects is given in table 1⇓. Concerning respiratory symptoms, highest prevalence was observed for wheeze. Prevalence of the remaining respiratory symptoms was <10%. Allergic sensitization (based on the specific IgE measurements conducted in the INGA study) to at least one of the allergens was found in 31.8% of the subjects. Sensitization to mites occurred more frequently than sensitization to cat.

Allergen and endotoxin concentrations in house dust

Table 2⇓ presents the distributions of allergen and endotoxin levels for the different sampling locations. Highest concentrations of mite allergens were found in mattress dust and lowest concentrations in living room floor dust, whereas Fel d 1 concentrations were highest in living room floor dust and lowest in mattress dust. Levels of Der f 1 were generally higher than levels of Der p 1. Weak, but statistically significant (p<0.05) correlations were found between Der p 1 and Der f 1 (r=0.35), Der p1 and Fel d 1 (r=0.12), Der f 1 and Fel d 1 (r=0.15) and endotoxins and Der p 1 (r=0.20), Der f 1 (r=0.15), and Fel d 1 (r=0.21).

Allergen and endotoxin levels and potentially confounding factors

Concentrations of all indoor factors were significantly higher in Hamburg compared to Erfurt (Der p 1: GM 344 versus 84 ng·g⁻¹, Der f 1: 1,114 versus 161 ng·g⁻¹; Der p 1 plus Der f 1: 3,621 versus 551 ng·g⁻¹; Fel d 1: 854 versus 213 ng·g⁻¹; endotoxin: 2,713 versus 1,911 ng·g⁻¹; p<0.01). There was a tendency towards higher Der f 1, Der p 1 plus Der f 1, and endotoxin levels in homes of subjects with the lowest educational level (p<0.10, data not shown).

Multiple logistic regression analysis

Results of multiple logistic regression analyses are shown in table 3⇓. Statistically significant associations (p<0.05) were found between Der f 1 and the sum of Der p 1 and Der f 1 in mattress dust and wheeze and breathlessness, whereas a level of >8 µg Fel d 1 per gram of dust increased the risk of cough at night. No statistically significant association was found between endotoxin levels in living room floor dust and asthma attacks and respiratory symptoms; however, the risk tended to be higher in highly exposed subjects.

Distinguishing between exposure to single allergens and mixtures of several allergens, the risk of respiratory symptoms (except wheeze) was found to be highest among subjects exposed to elevated concentrations of at least two allergens (Der p and/or Der f 1 and/or Fel d 1). The respective ORs and 95% CIs are presented in figure 1⇓ where estimable. Subjects exposed to elevated concentrations of at least two of the regarded allergens had a six- to seven-fold increase in the risk of wheeze and breathlessness and cough without cold in winter compared to nonexposed subjects. Because of sample size limitations, it was not possible to differentiate between different types of allergen mixtures.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.—

Adjusted odds ratios (OR) and 95% confidence intervals (CI) for selected respiratory symptoms comparing low exposure to mite and cat allergens to high exposure to one single allergen and mixtures of several allergens. ⋄: Dermatophagoides pteronyssinus antigen 1 (Der p 1) >10 µg·g⁻¹ (total number of subjects (N)=32); •: D. farinae antigen 1 (Der f 1) >10 µg·g⁻¹ (N=55); ▪: cat antigen d1 (Fel d 1)>8 µg·g⁻¹ (N=49); ▴: Der p 1>10 µg·g⁻¹ and/or Der f 1>10 µg·g⁻¹ and/or Fel d 1>8 µg·g⁻¹ (N=34). The horizontal line represents the null hypothesis that the respective OR=1.

Prevalence of asthma attacks and respiratory symptoms related to allergic sensitization are shown in figure 2⇓. Sensitized subjects were found to have an increased risk of asthma attacks and respiratory symptoms compared to nonsensitized subjects. Interactions between allergic sensitization and exposure to elevated concentrations of Der f 1 and Fel d 1 are presented in table 4⇓. Because of sample size limitations, it was not possible to classify sensitization into sensitization to related (Der p 1/Der f 1 or Fel d 1) and nonrelated allergens. Sensitized subjects exposed to elevated concentrations of Der f 1 or Fel d 1 were found to have the highest risk of asthma attacks and respiratory symptoms. In addition, sensitized subjects exposed to ⪕10 µg Der f 1 per gram of dust were found to have a five-times higher risk of wheeze and breathlessness (p<0.05), compared to nonsensitized subjects exposed to ⪕10 µg Der f 1 per gram of dust.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.—

Prevalence and 95% confidence intervals (CI) of a) asthma attacks, b) wheeze, c) wheeze and breathlessness, d) wheeze without cold, e) shortness of breath at rest, f) breathlessness at night, g) cough at night, and h) cough without cold in winter, for subjects with (□) and without (▪) allergic sensitization. Allergic sensitization refers to ≥1 serum specific immunoglobulin-E (IgE) ≥0.70 kU·L⁻¹ against Alternaria alternata (m6), Aspergillus fumigatus (m3), tree pollen (tx5), birch (t3), Cladosporium herbarum (m2), Dermatophagoides pteronyssinus (d1), D. farinae (d2), grass pollen (gx1, gx4), Phleum pratense (g6), dog dander (e5), cat dander (e1), weed pollen (wx3), and Blatella germanica (i6).

Nonsensitized subjects exposed to >10 µg Der f 1 per gram of dust were not found to have an increased risk of asthma attacks or respiratory symptoms compared to nonsensitized subjects exposed to ⪕10 µg Der f 1, but if subjects were sensitized in addition to high exposure, they were found to have an increased risk of asthma attacks (p<0.05), and wheeze and breathlessness (p<0.05). Sensitized subjects exposed to ⪕8 µg Fel d 1 per gram of dust showed a statistically significant increase in the risk of wheeze and breathlessness and breathlessness at night compared to nonsensitized subjects exposed to ⪕8 µg Fel d 1 per gram of dust. Nonsensitized subjects exposed to elevated concentrations of Fel d 1 had a significantly higher risk of breathlessness at night, and cough at night (p<0.05) compared to nonsensitized subjects exposed to low concentrations of Fel d 1. The risk of wheeze, wheeze and breathlessness, and shortness of breath at rest was strongly increased if subjects were exposed to >8 µg Fel d 1 per gram of dust in addition to sensitization.

Allergen and endotoxin levels

The occurrence of mite species varies between countries, and even in the same country it may vary between regions. D. pteronyssinus and D. farinae were analysed in the present study. The concentrations of major mite allergen Der f 1 in floor and mattress dust were 2.5–3.5 times higher than Der p 1 levels. This is similar to findings of other German studies 18, 19 in which D. farinae was found to be the predominant species. Concentrations of Der p 1 and Der f 1 in mattress dust were 5- and 6.5-times higher than in living room floor dust. Der p 1 and Der f 1 levels in mattress dust were higher than in floor dust from bedrooms and living rooms. This is in accordance with the results of other studies 20–22.

Although only 15.6% of the subjects had a cat during the past 12 months, the major cat allergen Fel d 1 was found in 97.5% of the homes. This supports the fact that cat allergen is ubiquitous due to passive transport 23. Bollinger et al. 24 found Fel d 1 levels capable of causing upper and lower respiratory symptoms in settled dust of homes without cats. Therefore, they concluded that the assessment of cat exposure should not be based solely on the presence or absence of a cat in the home. Fel d 1 levels >8 µg·g⁻¹ were found in 11% of the homes where no cat was kept. Cat allergen content was lower in bedroom floor dust than in living room floor dust. This was also reported by Custovic et al. 25. Endotoxin concentrations in living room floor dust were in line with the concentrations reported by Michel et al. 26.

Mite allergens and respiratory health

Since allergen exposure plays a major role in the pathogenesis of asthma and allergic sensitization in early childhood, or even prenatally, most studies include children. There are only a few studies on adults analysing the impact of allergen exposure on asthma and allergies. Chan-Yeung et al. 27 did not find a relationship between total allergen levels (sum of Der p 1 and Der f 1) and severity of asthma in adults either with or without house dust mite allergy, whereas Custovic et al. 28, as well as Tunnicliffe et al. 29, demonstrated a relationship between severity of asthma and exposure to Der p 1 in asthmatics sensitized to house dust mite. In the present study, the risk of asthma attacks and respiratory symptoms was not found to be increased continuously with exposure to mite allergens. For this reason, allergen levels were not used as continuous variables in the regression analysis, but were classified into two categories. The levels of 2 µg (data not shown) and 10 µg group I dust mite allergen, which have been proposed to increase the risk of acute attacks of asthma and symptoms, were used as cut-off values. Stronger effects (except for cough without cold in winter) were found using 10 µg as the cut-off value. Exposure to >10 µg mite allergen (Der p 1 plus Der f 1) and in particular, exposure to >10 µg Der f 1 per gram of dust, was found to be a risk factor for respiratory symptoms. Although Der p 1 and Der f 1 are highly cross-reactive, the effect estimates for the sum of Der p 1 and Der f 1 were weaker than the presented effects of Der f 1. No effect was found for Der p 1 alone, although the two antigens seem to be biologically rather similar. One possible explanation for this might be that Der f 1 is the predominant species in the region under study and that Der f 1 levels are much higher than Der p 1 levels. A relationship between presence of house dust mites and asthma-related respiratory symptoms in adults has also been demonstrated by Björnsson et al. 30. In this Swedish study, the presence of house dust mites was assessed by applying the semiquantitative ACAREX test. Since this test is not species specific, it was not possible to clarify the role of different species of mites. Alvarez et al. 31 found asthma symptoms to be at least partially dependent on current exposure to Der p 1.

Cat allergens and respiratory health

Roost et al. 32 found positive associations between the community prevalence of cat and the community prevalence of respiratory symptoms, physician-diagnosed asthma, and current asthma medication thus demonstrating the importance of cat allergens with respect to asthma and respiratory symptoms. Fel d 1 concentrations >8 µg·g⁻¹ were found to be a risk factor for cough at night. The risk of respiratory symptoms was strongly increased when subjects were sensitized in addition to high exposure. The authors conclude that exposure to Fel d 1 levels >8 µg·g dust⁻¹ is a risk factor for asthma attacks and respiratory symptoms in sensitized subjects. These findings are in line with the findings of Noertjojo et al. 33, who found cat owners to have a significantly higher risk of having current asthma and asthma-like symptoms compared to persons without cats. In the subset with positive skin-prick tests, they found that those who were allergic to cat dander had a significantly higher risk of current asthma than those not allergic to cat dander and not keeping a cat. Because of sample size limitations, it was not possible to do stratified analyses on cat-sensitized subjects in this study. In contrast, Chan-Yeung et al. 27 did not find a relationship between Fel d 1 and severity of asthma, either in adults with or without allergy to cats. Not only sensitized, but also nonsensitized subjects were found to have an increased risk of nocturnal breathlessness, nocturnal cough, and cough without cold in winter when exposed to >8 µg Fel d 1 per gram of dust. The reason for this is not yet clear. Active smoking was regarded as one possible factor explaining this dose-dependent effect in nonsensitized subjects, but smoking was not found to be associated with respiratory health, neither by itself nor as a cofactor in combination with a high allergen load.

Endotoxins and respiratory health

Michel et al. 26 found that the concentration of endotoxin measured in house dust is an important determinant of asthma severity in house dust mite-sensitized subjects. In the present study, endotoxins were not measured in mattress dust, but only in living room floor dust. Endotoxins in mattress dust were not found to affect respiratory health, neither by themselves nor as cofactors with high allergen levels (data not shown). One possible explanation for this lack of association might be that exposure to endotoxins, like exposure to allergens, is of major importance while sleeping, since people come into very close contact with it in their beds. A tendency towards higher frequencies of asthma attacks and respiratory symptoms was found among subjects exposed to elevated concentrations of endotoxins. Weak, but statistically significant, positive correlations were found between mite and cat allergens in mattress dust on the one hand and endotoxin concentrations in living room floor dust on the other. It is assumed that correlations would be stronger if both allergens and endotoxins were measured in mattress dust and that the correlation between allergens and endotoxins may explain the increase in risk of symptoms, even in nonsensitized subjects.

Mixtures of mite and cat allergens and respiratory health

Studies on the impact of exposure allergens in house dust on asthma-like symptoms focused either on mite or on cat allergen. Simultaneous exposure to elevated levels of more than one allergen was not reported in literature. In the present study, subjects exposed to elevated concentrations of at least two of the regarded allergens were found to have up to a seven-fold increase in the risk of respiratory symptoms in comparison with nonexposed subjects.