INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Children living in damp homes have been reported to have an excess of respiratory problems, compared with children living in dry homes, in studies carried out in different climatic conditions (1).

The mechanisms behind the relationship between dampness and respiratory problems are still poorly understood. One explanation could be that home dampness favors growth of various microorganisms and that these microorganisms produce compounds that cause respiratory problems. So far, house dust mite is the best established residential allergen determinant of respiratory symptoms and asthma (11). In some areas, house dust mites are present in a large proportion of homes and could represent the link between home dampness and respiratory conditions (11). However, in the cold climate, the moisture and dampness problems differ from those in the warmer climates. The relative air humidity indoors can be low (10 to 20%) for several weeks in winter. Consequently, the most frequent causes of moisture problems in dwellings are water damages, pipe leaks, construction defects, and insufficient ventilation. Low humidity reduces the occurrence of house dust mites (11), and the level of house dust mites has been reported to be lower in areas of cold climate (14) than in warmer areas (12, 13). Despite an expected low level of house dust mites, a study in Finland (2) has shown an association between dampness problems and respiratory conditions.

The aim of the present study was to assess the role of exposure to home dampness problems and house dust mites in the development of bronchial obstruction in young children living in the relative cold climate of Oslo, Norway.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Children within the Oslo Birth Cohort (17) diagnosed according to symptoms and signs as having bronchial obstruction during their first 2 yr of life and age-matched healthy control subjects were included in a matched case-control study. The study was approved by the Regional Ethical Committee for Medical Research.

The Oslo Birth Cohort

The Oslo Birth Cohort is a cohort of 3,754 children born in Oslo during 1992 and 1993. Details of this study have been presented elsewhere (17, 18). Briefly, the main eligibility criteria were the following: permanent address within the city of Oslo, no plans of moving from Oslo within the near future, birth weight > 2,000 g, no serious illness that might impair respiration, at least one family member able to speak and write Norwegian, at least one biologic parent living together with the child, and no known drug abuse in the family. Questionnaire information on the child's health and on environmental exposures was collected at birth and when the child was 6 mo (follow-up rate: 95%), 12 mo (92%), 18 mo (92%), and 24 mo (81%) of age.

Definition and Selection of Cases

The outcome of interest was bronchial obstruction defined as two or more episodes with symptoms and signs of obstruction or one episode lasting more than 1 mo. Participating families were instructed to contact the project pediatrician when their child had respiratory obstructive problems. Also, outpatient clinics were instructed to refer children to the project pediatrician. The parents were provided with cards to be filled in by physicians whenever the child was examined for any respiratory problems. Additionally, questionnaires with positive responses to questions on respiratory problems were extracted and parents contacted for verification. At least one episode of bronchial obstruction had to be diagnosed by physicians by direct observation, and the guidelines were that at least three out of five symptoms or signs (wheezing, chest recession, rhonchi during auscultation, forced expiration, and rapid breathing) should be observed. The final diagnosis of bronchial obstruction was made by a consensus decision of three senior pediatricians based on data from the pediatric clinical examination and from family physician, hospital, and outpatient records (18).

Selection of Control Subjects

The child born next in time to a case was chosen as control, providing he/she had no medical history suggestive of bronchial obstruction at the time of inclusion. Only children still living in Oslo and with the same home address during the last 3 mo before becoming candidates for being case or control were included.

Data Collection

Exposure assessment and interview of parents were undertaken during site inspection by two trained persons. The measurements were performed at the same time for the matched case/control pair. When a case was defined and an age-matched control was selected, both were contacted for planning of the parallel measurements. The families were urged not to perform any environmental changes in the home until the measurements had been finished. The initial visits were mainly conducted within 1 wk after the first contact. Exposure assessment was not conducted during the summer season (June, July, and August).

Exposure Assessment

Home dampness problems. Home dampness problem was defined as the presence of water damage, damp stain, or visible mold/mildew growth during the last 2 yr, confirmed by data collectors during site inspection. A repaired dampness problem was also considered as confirmed, but only if it could be verified by a receipt or otherwise by contacting the executive builder. Parental report of home dampness problem(s) in the interview was also recorded.

House dust mite exposure. The child's bed mattress was vacuumed for 4 min (2 min on the left part, 2 min on the right) by a nuzzle containing a cellular filter (19). The dust was collected on two equal filters, producing two dust samples for each bed mattress, one for analysis of mite allergen concentration and one for microscopic examination. Mite allergens: Dermatophagoides pteronyssinus (Der p 1) and Dermatophagoides farinae (Der f 1) concentrations were determined by enzyme-linked immunosorbent assay, using monoclonal antibodies (purchased from Dr. Martin Chapman and Indoor Biotechnologies Ltd, Clewyd, UK), following mainly the manufacturer's instructions (20). Results are given as µg allergens per g sieved dust. The assay sensitivity was 5 ng/ml for Der p 1 and 2.5 ng/ml for Der f 1, corresponding to 0.2 µg/g and 0.1 µg/g sieved dust, respectively. However, some beds had very low dust levels, leading to poorer sensitivity when expressed as µg/g. Microscopic count of mites is given as number of mites per bed based on examination of a sample from half of the bed.

Other building characteristics. Detailed registration of building characteristics including building material was conducted. The registered building characteristics included information on type of building, type of ventilation, polyvinyl chloride (PVC) flooring in one or more rooms, textile wall paper in one or more rooms, crowdedness (m² per occupant), and use of humidifier (Table 1). The relative air humidity in the central living room was measured by diffusion of water vapor through a tube with lithium chloride monohydrate (LiCl · H₂O) for 14 d and determined by differential weight (21). Based on these measurements and on outdoor temperature, absolute humidity (g/kg dry air) was calculated (22). The residential air change rates (h¹) were measured with a passive tracer gas method, perfluorocarbon tracer gas method (23), over the same 14 d. Measurements of humidity and air change rate were halted because of economic reasons after 186 pairs. Because of technical errors, correct air change rates were not obtained from one or both dwellings in 14 pairs.

Questionnaire information. Information on gender, birth weight, maternal age, breastfeeding, having siblings, maternal age, parental history of atopy, day care attendance at age of 1 yr, socioeconomic conditions (maternal education and family annual income), environmental tobacco smoke (ETS) exposure, and presence of pets was extracted from the questionnaires. The categorization of these covariates is presented in Table 2. Children having at least one parent reporting a history of asthma or hay fever were considered to have atopic parents. Children living in homes in which one or more of the inhabitants smoked as reported in the birth questionnaire were considered to be exposed to ETS.

Statistical Methods

Differences in levels of exposure between cases and control subjects were compared in bivariate analyses and tested for statistical significance by paired t tests or paired Wilcoxon tests. The odds ratio was used as a measure of the relative risk of bronchial obstruction according to level of exposure. Odds ratios were estimated in conditional logistic regression analyses (24). When estimating adjusted odds ratios, gender, siblings, day care attendance at the age of 1 yr, breastfeeding, ETS exposure, and parental atopy were considered as the core covariates and were always included in the model. Other covariates were included only if they contributed to a better accuracy and precision of the studied relation (25). For some covariates, values were missing for either a case and/or a control subject, mainly for the house dust mite variables. In this situation, the information was assumed equal in cases and control subjects and thus all pairs were retained in the analyses to maximize the precision. Parallel analyses were performed excluding the pairs with missing values. Both approaches gave similar results for both exposure to dampness problems and mites.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

A total of 304 children were diagnosed as cases. Of these, 256 children were still living in Oslo at the time of the diagnosis and could be included in the home visits. A total of five pairs were lost because the parents of three cases were not willing to have a home visit and the homes of two pairs were not visited by mistake. Table 2 shows the distribution of variables of the children and families participating in the cohort and separately for those participating in the case-control study. The case group had a higher proportion of boys, of children with siblings, atopic parents, and exposure to ETS and a lower proportion of children breastfed after 6 mo than the control group. Compared with the cohort, the selected control subjects did not show any substantial differences in the distribution of variables. No differences between the groups were found for most of the registered characteristics except for the floor materials containing PVC and decoration with textile wall paper, which were both more often seen in the homes of cases than control subjects (Table 1).

In total, dampness problems in the homes during the last 2 yr were confirmed in 27% of cases and 14% of the control subjects. The risk of bronchial obstruction in children was associated with home dampness both if the problems were confirmed by trained data collectors or professional builders or if reported by the parents (odds ratio = 2.6, 95% confidence interval [CI] = 1.6-4.2 and odds ratio = 2.6, 95% CI = 1.7-4.0, respectively) (Table 3). Dwellings were stratified into three groups: dwellings with no report of dampness problems (72%), with parent-reported dampness problems not confirmed by professionals (8%), and with parent-reported dampness problems confirmed by professionals (21%). Children living in dwellings with confirmed dampness problems had an increased odds ratio of bronchial obstruction of 2.7 (95% CI = 1.7-4.5), and those living in dwellings with only parent-reported dampness problems an odds ratio of 2.3 (95% CI = 1.2-4.7) compared with those with no reported dampness problems. The relative humidities were in general low, and there was no difference between homes of cases and control subjects (mean = 37%, SE = 0.69% and mean = 36%, SE = 0.58%, respectively). The mean values for absolute humidity were 6.0 g/kg dry air, (SE = 0.11) and 5.9 g/kg dry air (SE = 0.10) and for the air change rate were 0.67 h¹ (SE = 0.03) and 0.69 h¹ (SE = 0.03) for cases and control subjects, respectively.

A concentration of Der p 1 allergens > 2 µg/g dust from the children's beds was only found in beds of 11 cases (4.5%) and three control subjects (1.2%). More than 5 D. pteronyssinus mites per bed were found in beds of 20 cases (8.2%) and nine control subjects (3.7%) by microscopic examination of bed dust. Mites were found both in dwellings with and without confirmed dampness problems. The mite exposure assessed in both ways was related to the risk of getting bronchial obstruction (odds ratio = 3.7, 95% CI = 1.0-13.1 and odds ratio = 2.2, 95% CI = 1.0-4.9, respectively) (Table 3). Modest numbers of other house dust mite species were also found. Exposure to these mites was not associated with the outcome.

In a conditional logistic regression analysis controlling for exposure to Der p 1 and potential confounding factors (Table 4), dampness problems confirmed by the investigators remained a significant determinant of bronchial obstruction (adjusted odds ratio = 3.8, 95% CI = 2.0-7.2). Also, parental reports of home dampness problems that were not confirmed by the investigators were significantly related to the outcome (adjusted odds ratio = 2.5, 95% CI = 1.1-5.5). Analyses including the missing values gave similar results (adjusted odds ratio = 3.4, 95% CI = 1.8-6.5 and adjusted odds ratio = 2.7, 95% CI = 1.2-6.3). The adjusted effect of exposure to house dust mites in beds became somewhat weaker compared with the crude figures. Analyses with and without missing values gave similar results. The association between dampness problems and bronchial obstruction was still evident after excluding cases and control subjects with positive mite findings from the analysis (adjusted odds ratio = 3.4, 95% CI = 1.7-6.5 for confirmed dampness problems and adjusted odds ratio = 2.4, 95% CI = 1.0-5.5 for only parentally reported problems).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Children living in dwellings with dampness problems had an increased risk of bronchial obstruction compared with children living in dwellings with no such problems, with an adjusted odds ratio of 2.5 (95% CI = 1.1-5.5) when the problems were reported by parents only and not confirmed by professionals and 3.8 (95% CI = 2.0-7.2) when confirmed by a professional. The effect was independent of exposure to house dust mites. Few children were exposed to house dust mites.

The case definition required symptoms and signs of two or more separate or recurrent episodes or one persisting episode of bronchial obstruction. Most likely, this increased the probability that children diagnosed as cases were truly suffering from obstruction of lower airways. Several measures were taken to detect all the cases in the cohort including use of the follow-up questionnaires, family physician reports, and medical records, and the final diagnosis was made by a committee of three senior pediatricians.

The control subjects had similar characteristics as the whole cohort, indicating no bias in the selection of controls. Home visits were conducted as soon as the child was diagnosed as a case or selected as a control in order to avoid the possible effects of any changes in the home environment made because of the studied health problems. The measurements were carried out concurrently for each case and his/her control to avoid the influence of any seasonal variation in exposure. In order to minimize information bias, the data collectors were advised not to ask the parents about the case status. Some parents may nevertheless have provided this information during site visits. Parental reports of home dampness in the interview could be influenced by their child's symptoms and introduce a bias in the relation between dampness and health. To avoid such information bias, the information on dampness problems was collected both by parental interview and site inspection. Assessment of mite exposure was based on two objective measures: allergen concentration and mite counts in the dust from children's beds.

The potential confounding by building characteristics and socioeconomic indicators was controlled for in the conditional logistic regression. The associations between dampness problems and bronchial obstruction were in general strengthened by the adjustment. Our results are consistent with previous studies in children, and the odds ratios are of the same size (1- 5, 7-10). Of these studies, only van Strien and colleagues (4) have focused especially on the association between home dampness and lower airway disease during early life. As in other studies, parent-reported dampness problems were more frequent and had a weaker association with respiratory symptoms than confirmed dampness problems (3, 8, 9). This has been explained by the "spot-sample" character of the observation by investigators (3). In order to get a more valid exposure indicator, the present study included earlier dampness problems confirmed and repaired by professional builders together with those confirmed by the investigators "on spot." Consequently, children exposed to confirmed dampness problems had a larger increase in the risk of bronchial obstruction compared with those exposed to dampness reported by parents only.

Low air humidity and cold climate could explain the low level of mite exposure among young Oslo children (11, 13). For those exposed to mites, the increase in the risk of respiratory problems was in accordance with other studies. The common Norwegian practice of having unheated bedrooms with open windows, even during winter season, could also have contributed to the modest mite growth. However, no earlier study has investigated the prevalence of the house dust mite exposure in Oslo dwellings. Avoidance of exposure assessment during summer and new mattresses in young children's beds could have contributed to a low mite exposure in the study population. On the other hand, a recent study demonstrated that house dust mite growth occurred rapidly in new mattresses (26).

In the present study, dampness problems seemed to be associated with the health outcome independent of mite exposure probably due to unfavorable conditions for mite growth (11, 13, 15, 27). Dampness in dwellings under such conditions probably has other causes (construction defects, water leakage) than in areas with mild and humid climate. It is possible that dampness problems experienced under such conditions could increase some other microbiologic exposures or increase children's susceptibility to microbiologic agents. Lower airway problems in early life are not necessarily symptoms of asthma and allergy but could also be caused by viral infections (28). The association between respiratory symptoms and home dampness problems could thus also be caused by nonspecific mechanisms as proposed by Dales and associates (1) or by increased viral exposure somehow enhanced by dampness.

This study indicates that residential dampness problems increase young Oslo children's symptoms and signs of bronchial obstruction mainly without increasing their exposure to house dust mites. The exposure assessment revealed a low house dust mite exposure among the participants probably due to unfavorable indoor growth conditions for house dust mites. However, for those exposed, house dust mite exposure seems to increase the risk of respiratory problems.