Results

Table 1 shows the characteristics of the cases and controls. Compared with the entire ISAAC study population (n=2330), there were slightly more boys and more individuals from the higher Indo-Aryan ethnicity in the asthmatic group. The asthmatic group was equally divided between the urban (Kathmandu) and rural area (in Kathmandu Valley). Sixty per cent of the questionnaires were answered by the mother and 40% by the father, an elder sister, or the schoolchild itself. One hundred children lived in traditional houses (brick, stone and mud) and 147 lived in modern houses (cement). The modern houses were mostly located in Kathmandu city and the traditional ones in the rural area. The mean number of family members and elder siblings was 6.1 and 2.0, respectively.

Neither building construction, number of family members, number of elder siblings, bed sharing, nor pet ownership was associated with having asthma (table 2). Ninety families (36%) kept cattle (defined as buffalo, cow, horse, goat or pig) at home. Forty of these families kept their cattle inside the house at night. The animals were housed on the ground floor and often in the same room as the cooking place. From this room there was an opening in the ceiling with stairs leading to the second floor and sleeping rooms. Keeping cattle inside the house was inversely related to the risk of having asthma (tables 2 and 3). Keeping cattle in general (OR 0.9 (95% CI 0.5 to 1.6)) or keeping cattle outside the house at night was not related to asthma (OR 1.4 (95% CI 0.7 to 2.9)).

Keeping poultry inside the house seemed to increase the risk for asthma (table 2) and this association was upheld, although was not statistically significant, on multivariate analysis (table 3). Stratified analysis revealed that keeping poultry inside the house was a risk factor for asthma only in children who had two or more family members who smoked (unadjusted OR 3.4 (95% CI 1.0 to 11.9)).

For cooking, 75 families (34%) used open fire or burning stoves without a flue (with solid fuel such as wood and grass), 43 (18%) used gas stoves, and 106 (43%) kerosene stoves; 63 children (26%) reported that they had been bothered by smoke coming from the cooking place into the sleeping room. The risk for asthma was higher among children exposed to smoky fuels (defined as living in a house with an open fire or burning stove without a flue) than for those not exposed (table 3).

In 63% of the cases and 60% of the controls one or more family member smoked daily inside the house. Seven children (four cases and three controls) reported being smokers. Smoking by one family member (n=86) did not increase the risk of having asthma, but in families where two or more family members smoked (n=63) the risk of asthma was significantly increased in the index child (tables 2 and 3).

In analyses stratified by sex we found that passive smoking and the use of smoky fuels increased the risk for asthma in boys (OR_smoking 3.0 (95% CI 1.3 to 6.7) and OR_{smoky fuels} 2.4 (95% CI 1.2 to 5.1)) but not in girls (OR_smoking 1.0 (95% CI 0.3 to 3.1) and OR_{smoky fuels} 0.8 (95% CI 0.3 to 2.2)) (estimates not mutually adjusted).

No families in the city of Kathmandu kept cattle inside the house, few houses had open fires or burning stoves without a flue, and few families had more than one family member who smoked. In stratified analysis we found that living in an area with observed heavy traffic was the only significant risk factor for asthma in the urban population (matched OR 3.0 (95% CI 1.1 to 9.2)). In the rural population asthma was strongly related to tobacco smoking (OR 3.5 (95% CI 1.4 to 8.7)), indoor biosmoke (OR 3.3 (95% CI 1.1 to 9.9)), and keeping cattle inside (OR 0.2 (95% CI 0.1 to 0.6)) (mutually adjusted odds ratios from logistic regression analysis including building construction).

Discussion

Most studies on indoor climate and asthma have been carried out in developed countries. Asthmatic patients in these countries often have knowledge about allergen removal which might confound the association between possible risk factors and asthma. It is less likely that families involved in our study actively try to avoid potential causes of asthma in their homes.

Cases and controls were selected from a questionnaire based screening survey, so the diagnosis of asthma was based on symptoms rather than the measurement of lung function or bronchial hyperreactivity. Both a written questionnaire and a video questionnaire were used according to the ISAAC protocol. A validation study in Chinese schoolchildren concluded that the ISAAC international video questionnaire provided a simple and valid tool for the diagnosis of bronchial hyperreactivity in this population.19

We found a relatively strong inverse relationship between keeping cattle inside the house and having asthma. This finding was consistent after controlling for building construction, age, sex, keeping poultry, tobacco smoke, and the use of smoky fuels. A better nutritional status in families with household animals could be a confounding factor. However, this seems unlikely as keeping cattle outside the house had no protective effect. Furthermore, keeping cattle inside the house at night represents a source of heating, so a lower use of smoky fuels for heating could bias our results. However, the proportion of children bothered by smoke coming into the sleeping room was not reduced in families who kept cattle inside the house.

Cattle kept inside represent both a reservoir of microorganisms and a source of allergens—directly through ectoparasites living on the animals, excrement, and dander and indirectly through mites in fodder.20 ,21 The cattle were kept on the ground floor, usually in the same room or next door to the kitchen/cooking place. Most houses had openings in the ceiling or the walls that separated the rooms. Family members could be exposed to microorganisms and allergens both through inhalation and ingestion. There is little migration in the area and we assume that the families had kept their animals in a similar way for several years. Thus, early life exposure could be as important as exposure at the time of the investigation.

The inverse association between keeping cattle inside the house and asthma could be explained by the “infection hypothesis”.22 This theory suggests that reduced microbial exposure during the first years of life, particularly exposure to infections with intracellular pathogens, may be disadvantageous in terms of sensitisation to allergy. The development of resistance or sensitivity to environmental antigens depends to a large extent on the nature of the immunological memory which is generated during infancy and early childhood.23 T cell selection that favours the production of T helper 2 (Th2) cells increases the likelihood of atopic disease.24 Many intracellular pathogens generate a predominantly Th1 cell cytokine profile that is thought to downregulate the Th2 cell mediated immune mechanism. Animal experiments have shown that Th2 responses to an inhaled allergen in highly responsive young rats could be suppressed by exposure to infections and microbial adjuvants concomitantly with the allergen.17 The protective effect of childhood exposure to animals seen in this study, in studies of early exposure to pets,6 ,7 and in studies in the farm environment8-10 might result from a contribution by the animals to the microbial environment which promotes the development of a non-allergic immunity. The hypothesis of microbial stimulation may thus provide a unifying explanation for the inverse association between allergy and family size6 ,22 ,25 and the protective effect of childhood exposure to animals. Alternatively, the protective effect of early exposure to cattle could be related to the development of tolerance.

Keeping cattle inside the house, in particular, could provide an exposure to endoparasites in cattle excrement and to ectoparasites living on the cattle.21 It has been suggested that a high frequency of parasitism during early childhood may be associated with a lower prevalence of asthma26 ,27 and the role of parasitic infections in the development of asthma has been the subject of several studies.28 Improved sanitary conditions in developed countries have reduced exposure to parasites and may be a contributing factor to the increasing prevalence of allergic disease.

Although not statistically significant, keeping poultry seemed to increase the risk for asthma in our study. Exposure to indoor mite and cat allergens has been associated with increased sensitisation and asthma.29 ,30 Likewise, poultry allergens may be more potent in inducing asthmatic symptoms than allergens from other domestic animals such as dogs and cattle. Exposure to poultry has been associated with wheezing and bronchial hyperreactivity in other studies,31 ,32 and high concentrations of mites are found in dust from poultry in Nepal and elsewhere.33 ,34

Smoking inside the house by at least two family members was an important risk factor for asthma in our study. Unfortunately, we did not specify maternal or paternal smoking. We assume, however, that one smoker (prevalent in the city) means that the father smoked and that “two or more family members” (prevalent in the rural areas) in most cases implied that both father and mother smoked. In the rural area of Kathmandu Valley Pandey et al found a high prevalence of smoking in both men (85%) and women (62%). Few of the female smokers were concerned about the harmful effects of smoking on their own health (18.6%) or on the health of others (1.3%).12 Most of the female smokers in our study probably smoked during pregnancy. As no harmful effect was found in households with only one smoker, we speculate that maternal smoking in pregnancy could be of importance, as well as postnatal passive smoking. Increased asthma risk related to both these exposures has been observed in previous studies in the Western world.11 ,12

We found no effect of passive smoking in the urban area. This could be related to the limited exposure as in most families there was only one smoker. Furthermore, passive smoking seemed to be more harmful in homes with smoky fuels (OR_smoking 4.1 (95% CI 1.4 to 12.5)), which applied mostly in the rural area, than in those without (OR_smoking 1.1 (95% CI 0.4 to 2.8)).

The use of smoky fuels was associated with asthma in our study. Domestic biomass fuel combustion has been associated with reduced lung function or asthma in some,15 ,16 but not all,35 ,36 studies. Carbon monoxide, nitrogen dioxide, and suspended particulate matter are important air pollutants produced by combustion of biofuels and very high concentrations of total suspended particles have been measured in kitchens with open fires or non-vented stoves in Nepal and in other countries.37 ,38 Women in Nepal spend a considerable time near the fireplace. Direct exposure of children to smoky fuels, as well as maternal exposure during pregnancy, could explain our findings.

The harmful effects of passive smoking and of smoky fuels were of greatest significance in boys. This could be related to the fact that boys have relatively narrower airways than girls during childhood18 and, furthermore, boys tend to have more airway infections and more asthma than girls during this age period.18 The literature on passive smoking gives conflicting results with regard to sex differences in susceptibility,39 with some studies showing a greater effect of passive smoking in boys. A possible sex difference in the susceptibility to damage by airway irritants needs to be addressed in larger studies.

In conclusion, we found that keeping cattle inside the house during the night was related to less asthma in Nepalese children. There was no effect when cattle were kept outside the house, which suggests a particular effect of close exposure to livestock. Our findings support the concept that the microbiological environment during early childhood influences the prevalence of asthma. Stimulation of the Th1 pathway early in life by microbes from cattle could be a possible mechanism; induction of tolerance by early exposure to allergens from cattle is an alternative explanation. The fact that our findings from Nepal agree with studies of the effects of pets and a farm environment on the prevalence of asthma in the West suggests that the underlying mechanisms are of universal importance in the physiology of asthma and underlines the need for further research in the area.

Our study further suggests that passive smoking and exposure to smoky fuels increases the risk of asthma in childhood. This is also in agreement with reports of asthma and passive smoking in the West and points to a general role of airway irritants in promoting asthma. As well as direct exposure, maternal tobacco smoking and exposure to smoky fuels in pregnancy may also have been important in this study.