- Top of page
- Material and methods
Background: Higher prevalences of allergic diseases and IgE antibodies to inhalant allergens have been reported for persons living in urban areas than for persons living in rural areas.
Methods: Associations between cumulative incidences of allergic diseases in 1878 children aged 13–14 years and their place of residence (urban, semiurban, or rural) from birth were assessed by questionnaire (ISAAC), in order to find out whether there is a period of increased sensitivity to external influences during the first few years of life. Family history and exposure to pets, tobacco smoke, and damp were considered in multiple regression.
Results: There was a significantly higher prevalence of allergic diseases with urban residence than with rural residence during the first 2 years of life (e.g., for bronchial asthma, relative risk (RR) for the first year 2.1, 95% CI 1.2–3.7). An increased risk was still found after multiple regression (RR=1.7). Semiurban residence was associated with an intermediate cumulative incidence of allergic diseases. Maternal smoking during pregnancy was associated with asthma (RR=1.4, 95% CI 1.0–2.0).
Conclusions: The findings support a period of increased susceptibility during the first years of life. Whether rural lifestyle protects against allergy or whether urban pollutants contribute to allergy has to be elucidated.
Material and methods
- Top of page
- Material and methods
The population was selected according to the protocol developed by the International Study of Asthma and Allergies in Childhood (ISAAC) ( 17). The ISAAC questionnaires were filled in by 3377 children; i.e., 95.7% of all 3529 children aged 13–14 years living in a defined part of the county of Östergötland, including Linköping and Motala, in southeastern Sweden. Of these, 2339 children also answered an additional questionnaire on matters such as living habits. There was no significant difference in reported atopic diseases (bronchial asthma, atopic dermatitis, allergic rhinoconjunctivitis) between the 3377 and 2339 children (data not shown). Place of residence could be defined by the GIS system (see below for definition) for 1878 children during their first year of life, for 1959 during their second year, for 2115 during their seventh year, and for 2334 during their 13/14th year.
The parents of the children completed questionnaires on family history of allergy, smoke exposure, and home characteristics including living area, damp indoors, and presence of furred pets indoors. The ISAAC question used for asthma was, “Have you ever had asthma?”; for eczema, “Have you ever had an itchy rash which came and went for at least 6 months?”; and for allergic rhinoconjunctivitis, “Have you ever had hay fever?”
A population register maintained by the tax authorities with information on personal identification number, address, and property type of the home was linked to a property register comprising coordinates of the individual's residence in the Swedish national grid system. The resulting database thus included individuals' records with the x and y coordinates of the center (centroid) of their home (property/building where they lived) ( 18). This was repeated for the first, second, and seventh year, and the year of the investigation. The subjects were then matched by coordinates in a geographic information system (MAPINFO™ 4.0 software package, MAPINFO Corporation, Troy, NY, USA, 1996). Digital maps of the region were obtained from the Swedish land survey organization. These were used as background map, but they also included a layer of polygons representing urban areas according to the definition by Statistics Sweden, and containing thematic information about the number of residents in each polygon. By a structured query language (SQL) search routine, all subjects taking part in the investigation were allocated to a rural, semiurban, or urban setting for each year studied. Rural setting was defined as an area with fewer than 50 households within a distance of 200 m from each other and with fewer than 200 residents in total. As semiurban were defined those areas that comprised a cluster of more than 50 houses within a distance of 200 m from each other and a total number of inhabitants ranging from 200 to 10 000. Urban areas were defined as those where the number of inhabitants was more than 10 000. In this way, the place of residence of 1878 children was identified in the first year, and these children constitute the basis for this paper. The children studied did not differ significantly from the 2339 children in allergic diseases or sex.
Relative risk (RR) values with 95% confidence intervals were used in the univariate and multivariate analyses. The calculation was performed by the GENMOD procedure (SAS System, Version 6.12) with log as the link function and binomial as the distribution. This form of logistic regression estimates the RR instead of the odds ratios. In Table 1, we used multivariate models to determine whether any confounding was present on the risk of type of residence in the first year. The chi-square test and the chi-square test for trend were employed.
Table 1. Relative risk (95% CI) of allergic disease among 1878 children in relation to their type of residence in years 1, 2, 7, and 13/14 (univariate analyses and upper and lower confidence intervals). Multivariate analyses were also performed for first year including family history of allergy, sex, current parental smoking, maternal smoking during pregnancy, furred pets indoors, dampness indoors, and living area (more or less than 31 m2/person). No significant chi-squarefor trend was found at 7 and 13/14 years of age
| ||First year||Second year||Seventh year||13/14th year|
| Semiurban||1.5 (0.8, 2.9)||1.3 (0.6, 2.6)||1.4 (0.8, 2.7)||1.0 (0.6, 1.8)||0.7 (0.4, 1.1)|
| Urban||2.1 (1.2, 3.7)||1.7 (0.9, 3.2)||2.1 (1.2, 3.6)||1.4 (0.9, 2.3)||0.9 (0.6, 1.4)|
| Semiurban||1.1 (0.8, 1.6)||1.1 (0.8, 1.5)||1.1 (0.8, 1.6)||1.0 (0.7, 1.3)||0.9 (0.6, 1.2)|
| Urban||1.2 (0.9, 1.6)||1.1 (0.7, 1.6)||1.2 (1.0, 1.7)||1.1 (0.8, 1.4)||1.0 (0.8, 1.3)|
| Semiurban||1.1 (0.8, 1.5)||1.2 (0.9, 1.7)||1.2 (0.9, 1.6)||1.3 (1.0, 1.7)||1.1 (0.8, 1.4)|
| Urban||1.2 (0.9, 1.5)||1.3 (0.9, 1.7)||1.2 (1.0, 1.6)||1.2 (0.9, 1.6)||1.0 (0.8, 1.3)|
| Semiurban||1.1 (0.9, 1.4)||1.1 (0.9, 1.4)||1.2 (0.9, 1.4)||1.1 (0.9, 1.4)||1.0 (0.8, 1.2)|
| Urban||1.2 (1.0, 1.5)||1.2 (1.0, 1.5)||1.2 (1.0, 1.5)||1.1 (0.9, 1.4)||1.0 (0.8, 1.2)|
| Semiurban||1.4 (0.8, 2.3)||1.5 (0.9, 2.7)||1.4 (0.9, 2.3)||1.2 (0.8, 1.8)||1.0 (0.6, 1.4)|
| Urban||1.6 (1.0, 2.5)||1.5 (0.9, 2.5)||1.6 (1.1, 2.5)||1.3 (0.8, 1.9)||1.1 (0.7, 1.5)|
The study was approved by the Human Research Ethics Committee of the Medical Faculty at the University of Linköping (Sweden).
- Top of page
- Material and methods
The cumulative incidence at 13–14 years of age was 44.4% (831/1871) for any atopic disease, 9.3% (173/1855) for bronchial asthma, 23.2% (427/1837) for allergic rhinoconjunctivitis, and 28.0% (518/1849) for atopic dermatitis. Only small preponderances for boys compared to girls were seen for bronchial asthma (RR=1.1 [95% CI 0.8–1.5]), and allergic rhinoconjunctivitis (1.1 [95% CI 0.9–1.4]), but fewer boys had histories of atopic dermatitis (RR=0.6 [0.5–0.8]).
There was a significantly higher proportion of children with any allergic disease in the group living in an urban district for the first 2 years of life than in the group of children living in a rural district for the same period ( Fig. 1, Table 1).
Figure 1. Allergic diseases at 13/14 years of age in relation to living in rural, semiurban, or urban area during first and second years of life. Chi-square test for trend for rural, semiurban, and urban residence is added (*P<0.05, **P<0.01, ***P<0.001).
Download figure to PowerPoint
The difference was even greater in the case of two or more allergic diseases (Table 1). Urban residence, as compared to rural residence, during the first 2 years was significantly associated with bronchial asthma, but urban living during the seventh and last year before the investigation was not (Table 1). The RR of asthma associated with urban residence vs rural residence during the first year of life was 2.1 (95% CI 1.2–3.7), and for the second year it was 2.1 (95% CI 1.2–3.6). There was also an association with place of residence during the first 2 years for atopic dermatitis (Table 1).
The cumulative incidences of bronchial asthma, allergic rhinoconjunctivitis, and atopic dermatitis were associated with the number of family members with allergies ( Table 2). However, family histories of allergic diseases were evenly distributed among the groups with urban, semiurban, and rural residence during the first year. Women living in urban districts slightly more often reported smoking during pregnancy (287/1196, 24.0%) than women living in rural districts (58/292, 19.9%; NS). Smoking during pregnancy was associated with an increased cumulative incidence of bronchial asthma and allergic rhinoconjunctivitis; that is, an RR of 1.4 (95% CI 1.0–2.0) and 1.2 (95% CI 1.0–1.5) (Table 2), respectively, and in the multivariate model 1.3 (0.8–1.9) and 1.2 (0.9–1.6). For asthma, we investigated whether there was any interaction between smoking during pregnancy and urban residence and we found a risk of 3.6 (0.8–15) for this interaction compared to the expected 2.9 (univariate risk for smoking during pregnancy, 1.4, times univariate risk for urban living, 2.1). However, the interaction was nonsignificant. Furred pets were more common in families living in rural districts (239/331, 72.2%) than in families living in urban districts (515/969, 53.1%, P<0.001), and damp was more common in urban than rural districts (P=0.003).
Table 2. Relative risk (95% CI) of allergic disease among 1878 children in relation to various risk factors (univariate analysis)
| || ||Family history of allergy|
| ||Sex (M/F) ||None||Single||Double||Parental current smoking ||Maternal smoking during pregnancy ||Furred pets indoors |
|Bronchial asthma||1.1 (0.8, 1.5)||1.0||2.1 (1.5, 2.8)||3.2 (1.9, 5.0)||1.0 (0.7, 1.4)||1.4 (1.0, 2.0)||0.5 (0.4, 0.7)|
|Allergic rhinoconjunctivitis||1.1 (0.9, 1.4)||1.0||1.6 (1.3, 1.9)||2.2 (1.6, 2.9)||1.0 (0.8, 1.3)||1.2 (1.0, 1.5)||0.8 (0.6, 0.9)|
|Atopic dermatitis||0.6 (0.5, 0.8)||1.0||1.4 (1.1, 1.6)||1.6 (1.1, 2.1)||1.0 (0.8, 1.1)||1.1 (0.8, 1.3)||0.8 (0.7, 1.0)|
|Any disease||0.9 (0.7, 0.9)||1.0||1.4 (1.2, 1.6)||1.6 (1.3, 2.0)||1.0 (0.9, 1.2)||1.2 (1.0, 1.4)||0.9 (0.8, 1.0)|
|≥2 diseases||0.9 (0.6, 1.1)||1.0||1.6 (1.3, 2.0)||3.4 (2.2, 4.8)||0.9 (0.6, 1.2)||1.1 (0.7, 1.4)||0.5 (0.4, 0.7)|
The higher risk of occurrence of bronchial asthma, atopic dermatitis, any allergic disease, and more than one allergic disease among children living in urban districts than among children living in rural districts during the first years of life was still apparent after multiple regression, taking parental history of allergy, passive smoking, sex, indoor pets, living area, and dampness in the home into account (Table 1). The largest confounding was found for bronchial asthma, where the RR for urban living decreased by 24% or from 2.1 univariate to 1.7 multivariate. Nevertheless, the RR multivariate was nearly significant.
- Top of page
- Material and methods
Allergic diseases were less common in teenagers who had spent their first years in rural rather than urban areas. In particular, children who had spent their first 2 years in a rural district had less bronchial asthma and were less likely to have more than one allergic disease than children who had lived in an urban district during their first 2 years of life. After adjusting for parental history of allergy, living space, passive smoking, sex, dampness, and pets in a regression analysis model, the effect of urban residence during the first year of life on allergic disease remained evident (Table 1). Previous studies ( 3) have shown that sensitization early in life is associated with later atopic manifestations, and that exposure during the first few months of life is of particular importance ( 16).
Several authors ( 9, 12, 19) have shown that current urban residence, as compared to rural residence, is a risk factor for the prevalence of allergic diseases. For example, bronchial asthma, as manifested by a positive exercise challenge test, was more common in children from urban districts than in children from rural districts ( 9, 12). The mean peak flow rate was also significantly lower in urban than in rural areas in Sweden ( 8) (mean difference 10.6 l/min, 95% CI 2.2–19.0). This difference remained after adjustments were made for passive smoking, sex, and height. Asthma attacks were significantly more often reported for girls living within 1000 m of a freeway, when density of truck traffic was used as exposure variable, adjusted OR 4.3 (1.1–17) ( 20). In none of these studies, however, was the place of residence during the first years reported in relation to symptoms or findings at school age.
The cause or causes of the association between early place of residence and allergic disease have not been clarified. Pollution, e.g., with nitrogen dioxide with its adjuvant effect ( 21, 22), is more common in urban than in rural areas ( 23, 24). A higher risk of asthma was found in a polluted town than in a nonpolluted town ( 25). However, less allergic disease and sensitization to common allergens was found in the former East Germany with its high air-pollution levels than in the former West Germany ( 7).
Possible explanations of the urban/rural effect are differences in food (i.e., processed or fresh foods) and water intake (i.e., chlorinated water from the municipal water supply rather than spring water), more time spent indoors in urban areas ( 23), and socioeconomic factors, including different awareness of disease ( 26). Urban residence in infancy may be responsible for increased allergies because of a cleaner environment that deprives the immune system of adequate stimulus (pathogens or nonpathogens) ( 27).
Adjuvant factors may also explain some of the differences between urban and rural residents. Our previous study ( 28) found an increased risk of allergic rhinitis and birch-pollen-induced asthma among children living close to a paper pulp plant. An increased risk of allergic rhinitis for persons living in streets with high traffic density has also been found ( 29). In Denmark, the prevalence of allergic rhinoconjunctivitis was 19% in Copenhagen but only 6% and 11% in the rural areas of Fyn and Jutland, respectively ( 10).
We found a difference between place of residence during the first 2 years and cumulative incidence of atopic dermatitis. In a study from Venezuela by Lynch and et al. ( 11), no difference in atopic dermatitis was found in relation to place of residence. However, the rural residence group in Venezuela was younger than the urban residence group, and this might have affected the results. In contrast, an increased risk of atopic dermatitis for children living in streets with high traffic density has been found in Germany ( 30).
Development of bronchial asthma and allergic rhinoconjunctivitis was more pronounced in children who had been exposed to smoking during pregnancy. Current environmental tobacco smoking when the child was 13–14 years of age did not significantly influence the cumulative incidence of allergic diseases. This could be due to a selection bias; i.e., some parents probably stopped smoking because their child had obstructive bronchitis or atopic dermatitis. Several studies have shown that passive smoking is an important risk factor (summarized in ref. 31).
Other suspected risk factors, such as damp and keeping pets indoors, were not similarly distributed in urban and rural areas. Furred pets were more common in rural areas, and damp was more common in urban districts. However, only 2.6% of the families perceived indoor dampness problems. Keeping pets indoors was found to be less common in homes where there was a child with an allergic disease. This was probably due to the fact that some families removed their pet if they had a child with atopic manifestations or never kept a pet in the first place, as a prophylactic step ( 32, 33).
Some caution should always be exercised in deductions concerning causality with data on health outcome and exposure, in cross-sectional studies, because the temporal criterion is not fulfilled, and in ecological correlation studies, because data are available at different levels of resolution. While health outcome is provided at the individual level, exposure data are available at a more aggregated (group) level. For example, although all urban residents are exposed to an unspecified “urban factor”, this factor is an aggregate exposure measure comprised of several other “subexposures” where unmeasured ecological confounding including factors, such as variation in the underlying sociodemography of areas, may lead to variations in disease incidence ( 34).
The data indicate that early urban residence contributes to the development of allergic diseases, particularly bronchial asthma. The findings support a period of increased susceptibility during the first years of life. Whether higher levels of outdoor or indoor pollutants in urban homes cause these differences, or whether lifestyle factors associated with rural residence protect the child during the first years of life, has to be elucidated.