METHODS

Patients

Thirty eight patients aged 11–44 years with a history of asthma and house dust mite allergy were recruited from the outpatients department of Asthmacenter Heideheuvel in Hilversum, The Netherlands, from January 1996 to December 1998. Informed consent was obtained from the patients or their parents. The patients were selected on the basis of increased bronchial responsiveness to histamine inhalation (PC₂₀ < 4 mg/ml, 30 minute method), positive skin tests and/or raised specific IgE to house dust mite allergen, and relevant HDM exposure on the mattress (>1 μg Der p 1/g dust). All patients had FEV₁ values >60% (predicted value). Patients had no history of respiratory tract infections in the previous 6 weeks or severe asthma attacks in the previous 6 months. None had received oral corticosteroids in the previous 6 months.

All patients gave informed consent. The medical ethics committee of Asthmacenter Heideheuvel approved the study.

Study design

The study was of a randomised, placebo controlled, double blind, parallel group design, comparing the effect of allergen impermeable encasings on the mattresses, pillows and bedcovers during 1 year with matching placebo encasings.

At the start of the study a trained respiratory nurse visited the patients to collect dust samples from the mattresses of the patients for Der p 1 measurement and to note the allergen avoidance measures already present in the house. All patients included in the study had smooth bedroom floors. Patients were instructed to wash their sheets each week at 60°C. Apart from the mattress encasings, no other allergen avoidance measures were taken. At the end of the study the same nurse visited the houses again to collect dust from the mattress covers.

The patients were included during the entire year; the inclusion period was 2 years. Pollen allergic patients were tested outside the pollen season.

At the first visit patients underwent clinical evaluation. FEV₁ and vital capacity (VC) values were measured, skin tests performed, and a PC₂₀ histamine was assessed. Medication was withheld before the study period: inhaled steroids and sodium cromoglycate for 1 week before the bronchial histamine provocation test; theophyllines, oral β₂ adrenergic agents, long acting inhaled β₂ adrenergic agents, and antihistamines for 48 hours, and inhaled short acting β₂ adrenergic drugs for 6 hours before the tests.

Collection and extraction of house dust

Before the intervention, at 4 and 8 months, and at the end of the intervention mattress dust was collected by the same vacuum cleaner (Philips Vitall 377, 1300 watt, Philips, Eindhoven, The Netherlands) from the whole mattress during 2 minutes with a special filter device (ALK, Horsholm, Denmark). At the start of the study dust was collected directly from the mattresses; at the end of the study dust was collected on top of the encasings. The filters were stored in the freezer at –20°C until analysis at the end of the study.

Der p 1 antigen was measured using enzyme linked immunosorbent assay (ELISA). Monoclonal antibodies against Der p 1 were immobilised on a 96-well plate. Incubation with the dust extracts was followed by a second incubation step with a polyclonal antibody (horseradish peroxidase, HRP). After adding 1,2-phenyldiamine HCl (OPD) as substrate, absorption at 490 nm was measured using an ELISA reader.

Histamine challenge

Histamine phosphate solutions (doubling concentrations from 0.25 to 32 mg/ml) were administered through a De Vilbiss 646 nebuliser with a gauged output of 0.13 mg/ml. The nebuliser was mounted on a valve box with an aerosol filter. The nebulisation time was 30 seconds, during which the patient was instructed to breath quietly. The test was started with inhalation of a phosphate buffer aerosol. Before inhalation three measurements of VC and FEV₁ were performed (Jaeger Masterscreen). FEV₁ was measured after each concentration. PC₂₀ histamine was derived by linear interpolation.

Mattress encasings

Mattresses, pillows, and bedding in the intervention group were encased with covers supplied by Cara C'air (Allergy Control AC btm Velserbroek, Netherlands). The matched placebo covers were made by the same company. The encasings were placed in position by a research nurse and left in situ for 1 year.

Quality of life

Quality of life was assessed by the Quality of Life for Respiratory Illness Questionnaire (QoL-RIQ).¹⁷ The QoL-RIQ is a disease specific quality of life questionnaire for patients with asthma and COPD which consists of 55 items divided into seven domains: breathing problems (9 items), physical problems (9 items), emotions (9 items), situations triggering/enhancing breathing problems (7 items), general activities (4 items), daily and domestic activities (10 items), social activities, relationships and sexuality (7 items). To focus the questions on the patients' experiences, items are formulated in terms of “how much trouble” they had experienced from the mentioned symptom or emotion. In the case of activity related items, questions are stated in terms of “how much they were impeded” by their disorder in carrying out that specific activity. Patients are asked to give their answer on a 70 point Likert scale ranging from “not at all” to “extremely” troubled or impeded. Reliability (test-retest, internal consistency) and validity have been proven.¹⁷

Clinical parameters

During the 14 day period before the intervention and at the end of the 12 month intervention period the patients were asked to keep diary cards in which asthma and nasal symptoms, peak flow values, and medication use were recorded twice daily. Asthma symptoms included dyspnoea, cough, expectoration, and wheezing. Nasal symptoms included nasal blockage, rhinorrhoea, sneezing, and itching. Each item was scored on a scale from 0 (no symptoms) to 4 (severe symptoms).

The patients were trained in performing peak flow manoeuvres with the mini-Wright meter. They were instructed to perform three readings and to record the highest value in the morning when waking and in the evening before sleeping.

Patients were asked to continue their normal inhalation medication and to record extra rescue medication in case they needed it.

Data analysis

Statistical analyses were performed with SPSS. Comparisons within groups (before and after intervention) were made with the Wilcoxon signed rank test. Logarithmic data were analysed using the sign test. The Mann-Whitney U test was used for between group comparisons. p values of <0.05 were considered significant. Values are expressed as mean (SE) or as median (range).

Power calculations were performed. We expected a 20% increase in the PC₂₀ histamine level. Based on the mean PC₂₀ of the complete group of 1.59 (0.38) mg/ml, an increase to 1.91 (0.38) was expected. With a power of 99% of achieving a significant result at the 5% level, the calculated sample size was 29.3 patients. As 30 patients completed the study, a 20% change in PC₂₀ histamine would have a probability of 99% of being noticed.

RESULTS

Thirty eight atopic non-smoking patients with asthma and house dust mite allergy entered the study. Eight patients did not finish the study, five from the placebo group and three from the treated group. In the placebo group three dropped out because of asthma instability, one because of moving to another city, and one because the recording of symptoms, peak flow and rescue medication was not sufficient to make an accurate analysis. In the treated group one individual dropped out because the study was too much of a burden, and two because of insufficient diary keeping. Drop out due to disease instability was significantly higher in the placebo group. 30 patients completed the entire study, 16 in the treated group and 14 in the placebo group (fig 1). The clinical characteristics of these patients are presented in table 1. Although both groups had severe hyperresponsiveness (geometric mean values of PC₂₀ histamine 1.75 and 1.27 mg/ml in the placebo and treated groups, respectively), three patients in each group did not regularly use inhaled steroids. The demographic characteristics of the two treatment groups were similar (table 1). There were no significant differences between the two groups in PC₂₀ histamine, FEV₁, PEF, and medication used.

Der p 1 concentrations on mattresses

Figure 2 shows mean log Der p 1 concentrations on the mattresses before and 12 months after the start of the study in both groups. In the treated group Der p 1 concentrations on the mattresses were significantly lower after 1 year (26.19 (8.58) v 2.79 (0.88) μg/g fine dust, p=0.004). In the placebo group there was no significant reduction in Der p 1 (23.28 (10.44) v 25.11 (11.98) μg/g fine dust, p=0.18). A significant difference in the treatment induced change in Der p 1 concentration was seen between the two groups (p=0.04). The significant reduction in the Der p 1 concentration was present after 4 months and persisted throughout the year.

Histamine challenge

At the start of the study the mean PC₂₀ histamine was 1.45 (0.44) mg/ml and after 1 year this had risen to 1.66 (0.35) mg/ml in the treated group; this difference was not significant (p=0.64, fig 3). In the placebo group there was no significant change in mean PC₂₀ over 1 year (from 1.75 (0.32) mg/ml at the start of the study to 1.57 (0.28) mg/ml after 1 year (p=0.97)). Treatment induced changes in PC₂₀ histamine were not significantly different between the two groups (p=0.77).

Quality of life

Overall quality of life scores were comparable at baseline in the placebo and in the treated group. The same was true for the subdomains. Clinically relevant improvements (difference >0.5)¹⁸ were seen in breathing problems, physical problems related to chest problems, triggering/enhancing circumstances, and total score in both the treated and the placebo groups. Although the size of the improvements did not significantly differ between the two groups, improvements within the treated group were significant.

Clinical parameters

Baseline values of asthma symptom scores showed no significant differences between the groups (table 2). The median pulmonary symptom score did not change significantly over 1 year in either group. There was a significant decrease in the nasal symptom score in the treated group (p=0.04) but not in the placebo group; the difference between the two groups was not significant. Baseline PEF values (morning and evening) were comparable for both groups (table 3). No significant changes in morning and evening PEF, peak flow variability, or the use of rescue medication occurred in either group following the 1 year intervention.

DISCUSSION

This study was performed to investigate the effect of anti-allergic mattress encasings in carpet free bedrooms on Der p 1 exposure in the bed and on clinical parameters in patients with moderate to severe asthma with allergy to house dust mite. We found a significant reduction in Der p 1 concentration in the dust collected from the mattresses in the actively treated group compared with the placebo group. PC₂₀ histamine did not improve during the 1 year intervention period. Although a significant improvement in nasal symptoms and quality of life was observed only in the actively treated group, we found no significant difference between the placebo and actively treated groups in the change in pulmonary and nasal symptoms, quality of life, peak flow values, and use of rescue medication.

Earlier studies using several different types of mattress encasings have also shown a reduction in Der p 1 exposure on top of the mattress (table 4).^8–10,19,20 However, other studies did not show a reduction in Der p 1 concentrations and, remarkably, the carpets in the bedroom were not removed in these studies.^21,22 We excluded the problem of Der p 1 contamination from the floor¹³ by including only patients who had uncarpeted floors in their bedroom. This may have contributed to the fact that we could reach a significant reduction in the actively treated group even though we had higher baseline Der P 1 concentrations than other studies.^22,23 The reduction in allergen concentration was reached after 4 months and remained during the whole study period.

Although Der p 1 concentrations were significantly reduced in the actively treated group compared with the placebo group, we did not find a significant reduction in bronchial hyperresponsiveness. Other studies have also failed to demonstrate an improvement in bronchial hyperresponsiveness.^10,11,22 Two studies^22,23 did not find a substantial reduction in allergen concentrations in dust, which explains the lack of improvement in bronchial hyperresponsiveness. Frederick et al¹⁰ stated that all patients were reasonably controlled on regular prophylactic treatment, so little or no change in clinical parameters could be expected. Even Cloosterman and coworkers,¹¹ who tried to avoid this treatment effect by including only patients who either did not use inhaled steroids or were able to stop them, did not find a significant improvement in bronchial hyperresponsiveness nor in any of the clinical parameters used such as symptom score, PEF variability, and reversibility of FEV₁.

How can we reconcile these observations? Patients participating in our study had severe hyperresponsiveness despite relatively high doses of inhaled corticosteroids of > 800 μg (for comparison, PC₂₀ <4 mg/ml using the 30 minute method is comparable to 1 mg/ml in the 2 minute method). Thus, despite suppression of airway inflammation by use of inhaled steroids for years, severe hyperresponsiveness remained. Airway inflammation and bronchial hyperresponsiveness are induced by repeated inhalation of low doses of allergen.²⁴ However, a 1 year reduction in exposure to HDM might be too short in patients with a life time exposure before the intervention. One can hypothesise that the persistence of severe hyperresponsiveness is related to airway remodelling, the structural changes in bronchial architecture as a result of chronic airways inflammation.²⁵ No further improvement can therefore be expected with allergen reduction that most probably affects acute inflammation in this stage of already established disease.

Until now there have been no data available on the effect of allergen avoidance measures on quality of life. Clinically relevant improvements in quality of life were found in both groups. The instrument we used is a questionnaire specific for asthma and COPD. Our study focused on the allergic component of asthma which was represented by only three of the 55 items. This may partially explain the lack of difference between the two groups. Alternatively, allergen reduction may not affect quality of life, and the improvement in both the intervention and placebo groups may reflect the special attention received by the patient during the study period.

The important question in allergen avoidance studies is whether the avoidance technique used improves asthma control in sensitised patients.²⁷ Only a few double blind placebo controlled studies on allergen avoidance in relation to symptoms have been performed (table 4). When reviewing these studies it is clear that improvement in symptoms occurred in four studies.^14,24,26,28 Two studied children,^26,28 one studied patients with mild asthma, and one studied allergic patients who had not yet developed asthma (subclinical). Other studies,^11,22 even in children,^10,23 did not find a reduction in symptoms. PEF was recorded in seven studies; in two a significant increase in PEF occurred in patients with mild and preclinical asthma^14,29 while in five there was no significant increase in PEF.^{10,12,19,22,23} Medication was recorded in one study without a positive result.¹⁰ Taken together, the data suggest that the contribution of allergen avoidance measures is ineffective in patients with moderate to severe asthma. The severity of the clinical manifestation is influenced by more factors such as other allergens, viruses and air pollution, which are not influenced by the avoidance measures.

Surprisingly, there are no published reports of controlled trials of the effects of allergen avoidance on nasal symptoms. In our study rhinitis symptoms were scored in addition to asthma symptoms and we found a significant improvement in nasal symptoms in the treated group, although the difference between the two groups was not significant. The nose therefore seems to be more responsive to avoidance measures than the lower airways. A controlled trial of avoidance measures in patients with rhinitis using subjective and objective parameters might be of interest.

In conclusion, the use of anti-allergic mattress covers results in a reduction in Der p 1 concentrations in carpet-free bedrooms. In patients with moderate to severe asthma no change occurred in airways hyperresponsiveness and clinical parameters. This lack of effect may be due to the chronic stage of the asthma and/or limiting the avoidance measures to the bedroom. Future studies should explore whether night time and daytime avoidance measures in the early stages of the disease are more effective.