Cockroach allergy and exposure to cockroach allergen in Polish children with asthma


Iwona Stelmach
M Curie Hospital
Department of Pediatrics and Allergy
35 Parzeczewska Str.
95–100 Zgierz


Background:  Asthma morbidity increases every year, especially among children, and exposure to high levels of indoor allergens is a very important factor. We evaluated the prevalence and exposure to cockroach (CR) allergen in asthmatic children in Poland, and also tested the hypothesis that asthma with allergy to CR is more severe than with allergy to other antigens.

Methods:  One hundred and sixty children with asthma were examined, had skin prick tests (SPT) with common and CR allergens, underwent spirometry, and provocation tests to histamine. Children with positive SPT to CR had measured specific IgE levels to this antigen and Bla g 2 concentrations were measured in their homes.

Results:  The most common allergen, was dust mite 51.3%, followed by pollen 48.8% and CR allergen 24.3%. In children with CR sensitivity, 13% had mild asthma, 26% moderate and 61% had severe asthma. Their levels for forced expiratory volume in one second (FEV 1 ), and the provocative concentration of histamine that caused a 20% fall in FEV 1 (PC 20 ), were statistically lower than in the group of children with other than CR allergies. Bla g 2 antigen was detected in 55.13% samples. The highest levels of Bla g 2 were found in old houses, without central heating, and in houses with lower income.

Conclusion:  In Polish children, CR allergen is a very important factor of sensitivity. Concentrations of Bla g 2 in homes are higher than previously reported in other European countries, and are strongly related to the houses' characteristics. Also, children with CR hypersensitivity have severe asthma more often than children with other allergies.

Morbidity due to asthma is disproportionately high among inner-city residents, for reasons that are not completely understood (1). Proposed explanations include increased exposure to allergens, poor air quality, psychosocial problems, and inadequate access to good medical care (2).

Allergens involved in causing asthma include those derived from house-dust mites, animal dander, and mold spores (3.4). Recent studies have highlighted the clinical relevance of cockroach (CR) hypersensitivity in allergic subjects, in particular those living in urban areas and in crowded multifamily dwellings (5–9).

In the USA almost 50% of the urban atopic population, belonging to a particular group (overall poor, Hispanic or black), have positive skin prick tests (SPTs) to CR allergens (6). In addition, it was reported that asthmatic patients who are CR positive (CR+) are sicker and more steroid-dependent than CR negative (CR–) subjects. However, a clear causal relation among allergy to CR, increased levels of CR allergen, and asthma has not been demonstrated.

It is known that CR components such as body parts and feces disintegrate and become incorporated into respirable house dust, acting as inhalant allergens (7, 10, 11). In the USA and Europe the most common CR species are the German Blattella germanica and the American Periplaneta americana (12–14).

So far, few data have been reported about the prevalence of CR hypersensitivity in European countries and the study populations were mainly represented by adult patients. In this paper we investigated the prevalence of CR allergy and exposure to CR allergen in a population of Polish children with atopic asthma (15). We also tested the hypothesis that asthma with allergy to CR is more severe then with allergy to other antigens.

Material and methods


We recruited 174 child patients (aged 4–17 years; mean 9.89 ± 4.03 years) from our Allergy Clinic in Zgierz, Poland. Only patients with confirmed diagnosis of atopic asthma (15) (mild to severe) were included in the study. Asthma severity was defined according to Highlights of the Expert Panel Report 2(16): “Guidelines for the diagnosis and management of asthma” (NIH Publication no. 97–4051 A, May 1997).

The study took place in fall of the year 2000. All subjects were skin prick test (SPT) positive to at least one of the tested allergens. They underwent pulmonary function testing and bronchial provocation tests. In children with positive skin test to CR allergen, specific IgE level to this antigen was also measured.

Dust samples from homes of patients with positive SPT to CR allergen were collected and analyzed for CR allergen Bla g 2.

A questionnaire addressing familial and personal history of atopic diseases, age at the onset of atopic disease, place of residence (urban or rural), housing characteristics (age of building older or less than 50 years, central heating or other), and family income (low income less than 350 PLN/month/person in the family) was used.

Skin testing

At the time of evaluation all of the subjects were healthy and had not had drugs interfering with skin testing for at least 15 days. SPT to common inhaled allergens and CR allergen (Nexter-Allergopharma, Germany), were performed. A mean wheal diameter ≥ 3 mm and ≥ the histamine control was defined as a positive response. The following allergens were tested: Dermatophagoides pteronyssinus and D. farinae, skin derivatives of cat and dog, moulds, pollen of grasses, weeds, trees and Blattella germanica.

CR-specific IgE antibodies

Serum IgE to CR-specific antibodies (B. germanica) were measured by the commercially available immunoassay technique according to the instructions of the manufacturer (Pharmacia CAP, Uppsala, Sweden).

Measurements of lung function and bronchial responsiveness

Lung function measurements were performed at least 6 h after the last inhalation of short acting β2-agonists. Forced expiratory volume in one second (FEV1) was measured by spirometer (Lungtest 1000), and the highest of three measurements was taken. Reference values were computed, according to the recommendations of the American Thoracic Society standards of acceptability and reproducibility (16).

Histamine inhalation provocation tests were performed (if the prebronchodilator FEV1 equalled 70% of predicted value). This was performed according to the method described by Cockroft (17), using a DeVilbiss jet nebulizer (DeVilbiss Health Care Inc., Somerset, PA) with an output of 0.33 ml/min. The result was expressed as the provocative concentration of histamine that caused a 20% fall in FEV1 (PC20).

Home visits and dust collection

All families were asked for permission for home visits as they entered the study. At the home visit, personnel inspected the home according to a standardized protocol.

Measurement of allergen levels in settled dust

Dust samples were collected from the home with a hand-held vacuum (Cobra 2000, 1400 W, Zelmer, Poland), in a filter (Dustscreen, Heska, Switzerland). Samples were collected from kitchen and bedroom by using standardized methods, as previously described (18, 19). The kitchen floor was vacuumed for 5 min. In the bedroom the sample was collected from a 1-m2 area near to, underneath, and on the bed for 5 min. Dust samples were removed from the filter, sieved, and then stored at −30°C until they were extracted.

An aqueous extract of 100 mg of sieved dust was prepared in 2 ml of borane-buffered saline. The extracts were stored at −30°C until they were assayed for CR allergen (Bla g 2). A sandwich ELISA (Indoor Biotechnologies, UK) was performed with affinity-purified monoclonal antibodies against Bla g 2 for allergen capture, and rabbit anti-CR antibodies for allergen detection. Allergen concentrations are reported in µg/g of dust.

Statistical analysis

The results were analyzed by StatSoft Statistica for Windows 5.0 (StatSoft, Inc., Tulsa, OK). Data were presented as mean with 95% confidence intervals (CI). On demographic data means with standard deviation (SD) were used. Significant differences were determined by Student t-test for independent variables. P-values < 0.05 were considered to be significant. Multiple regression analysis was used to show correlations between different independent variables with Bla g 2 concentrations.


The study was approved by the Medical Ethics Committee of the Medical University. All parents or guardians gave written consent for participation in this study.


One hundred and sixty subjects completed the study. Fourteen patients were withdrawn, 10 of whom were lost to follow-up, and four who withdrew consent later in the study.

The most common allergens (based on SPT results) were dust mite 51.3%, followed by pollen 48.8% and CR allergen 24.3%. Characteristics of the patients who completed the study are given in Table 1.

Table 1.  Patients characteristics. Data are presented as means ± standard deviation
 CR– atopic population
n = 121
CR+ population
n = 39
Age (years)9.7 ± 0.3610. ± 0.66
Asthma duration (years)3.7 ± 0.143.9 ± 0.22

Characteristics of CR-sensitive and CR-negative patients

The number of children with positive family history of atopic disease was almost the same among CR+ and CR– patients, 26/39 (66.7%) and 90/121 (61.9%), respectively.

The average time of asthma duration in patients sensitive to CR antigen was similar to that of patients sensitive to other allergens.

In children with CR sensitivity (39 patients), 5 (13%) had mild asthma, 10 (26%) moderate and 24 (61%) had severe asthma. In CR– children (121 patients), 35 (29%) had mild, 42 (35%) moderate and 44 (36%) severe asthma (Fig. 1).

Figure 1.

Asthma severity in CR+ and CR– patients.

Specific IgE > 0.35 kU/l against German CR was found in 36/39 (92.3%) of CR+ children. Class 3 and 4 in specific IgE had 17/39 (44%) children and these levels correlated with severity of asthma; 16 out of these 17 children (94%) had severe and one had mild asthma (Fig. 2).

Figure 2.

Correlation between asthma severity and level of IgE specific to Bla g 2 (using radioallergosorbant assay RAST); r  = 0.4445.

FEV1 values in CR+ and CR– children were 82.3% (95% CI, 77.1–87.6) (range 52%−124%) and 91.3% (95% CI, 88.7–94.0) (range 63%−124%) of predicted values, respectively, and the difference between both groups was statistically significant (P = 0.001).

PC20 for histamine values in CR+ and CR– children were 2.91 mg/ml (95% CI, 0.36–7.98) and 5.18 mg/ml (95% CI, 2.18–8.0), respectively, and the difference between both groups was statistically significant (P < 0.001). FEV1 and PC20(histamine) values for CR+ and CR– children are presented in Figs 3 and 4.

Figure 3.

Lung function (% FEV 1 ) in CR+ and CR– patients. Values are presented as means. Bronchial hyperreactivity to histamine (PC 20 H) in CR+ and CR– patients. Values are presented as means; t- test P  = 0.0015.

Figure 4.

Bronchial hyperreactivity to histamine (PC 20 H) in CR+ and CR– patients. Values are presented as means; t- test P  = 0.001.

Bla g 2 exposure

Thirty-nine children's homes were visited (78 rooms in 39 homes). Bla g 2 antigen was detected in 43 (55.13%) of 78 samples, in 62.8% kitchen samples (27 kitchens) and 37.2% bedroom samples (16 bedrooms). The range of Bla g 2 levels found in house dust was 0.1–389.26 µg/g dust. Mean levels of Bla g 2 in kitchens were higher than in bedrooms, 75.22 µg/g (95% CI, 43.03–107.4) and 9.37 µg/g (95% CI, 1.93–16.81), respectively (P < 0.001). There was a correlation in CR exposure between both rooms, r = 0.39, P = 0.01 (Fig. 5). In all homes with detected CR antigen, the levels were higher than sensitization value 8 µg/g (at least in one room), the proposed threshold for causing disease. The multiple regression analysis showed correlation between high levels of Bla g 2 and: old houses (P = 0.003); houses without central heating (P = 0.023); and houses with lower income (P = 0.007). Children living in urban areas were sensitized more often to CR than children living in the suburbs (P = 0.027).

Figure 5.

Correlation between Bla g 2 concentrations in kitchens and bedrooms of homes of children with allergy to CR; r  = 0.04; P  = 0.01.


These data demonstrate a clear relationship between current exposure to CR allergen Bla g 2 and current sensitization to CR allergen in asthmatic children.

The prevalence of CR allergy, defined by skin tests and/or blood tests, varies widely in different countries and from study to study but is usually high (1, 3, 4, 8, 9). Prevalence of CR hypersensitivity among atopic asthmatic children in our study was equal to 24.3% (positive skin test results and presence of specific IgE to CR). This prevalence is higher than that found in other European studies (13.14), but similar to that found in cities of the USA. Very recently, a multicenter study performed in the USA showed a prevalence of sensitization to CR allergens corresponding to 25–37% (3, 6, 20).

The high rate of sensitization to CR may be due to other factors. For example, some studies showed that a person might be sensitized to other allergens in the CR extracts that did not cross-react with the CR allergen measured in the environment (3, 10).

Gelber et al. (21) and Call et al. (2) showed that CR sensitization was more common in the asthmatics compared to the nonasthmatics presented to an emergency room. When sensitization was combined with exposure, a very high association with asthma occurred.

All of our children had atopic asthma but CR + children more often had severe asthma. They also had lower lung function measures and higher bronchial hyperreactivity than children with other allergies. It has been previously suggested by Kang et al. that these children may represent a subgroup of patients with more severe disease (6).

Many factors, such as exposure to allergen, place of residence, and socioeconomic status, may play an important role in the development of sensitization to CR allergens (12, 21–24). Our findings confirm the reports of others. In our study population the number of children living in urban areas, and from low income families were significantly higher among CR+ patients.

In 41–86% of inner-city homes in the USA Bla g 1 and Bla g 2, the major allergens of German CR, were found in concentrations above the proposed threshold for causing disease ( 8 µg/g dust) (2, 5, 23, 25). In our study concentrations of Bla g 2 were also found to be above this threshold in at least one of two samples taken from each house with detectable CR allergen. The concentration of Bla g 2 was lower in the bedroom than in the kitchen but correlation between both rooms was high. We also found significant correlations between allergen concentration in kitchens and in bedrooms and sensitization. Similar findings come from Eggleston et al. (20).

We found that 23% of children without detectable Bla g 2 allergen in house dust had positive SPT responses to CR extract. This could be due to previous exposure in another home or from current exposures at school or in day-care centres.

The highest levels of Bla g 2 were found in old houses, which in Poland do not have central heating, where humidity is high and cockroaches like to reside the most.

In summary, the data suggest that the prevalence of CR sensitization in Polish children is higher than among children from other European countries and similar to that among children in the USA. Children with CR hypersensitivity had severe asthma more often than children with other allergies. Furthermore, high concentrations of the CR allergen Bla g 2 were detected in kitchens and bedrooms from this region. We conclude that allergy to CR may be a more significant contributor to childhood asthma than previously appreciated.