We observed a prevalence of asthma symptoms of 26 percent in indigenous Warao children aged two to ten years. This prevalence is within the range of 8.6 - 32 percent of prevalence rates found in the ISAAC study performed in 2000 in Latin America . Lai et al. reported a prevalence of asthma symptoms of 10 to 20 percent in Venezuelan schoolchildren. They did not, however, report whether included children lived in rural or urban areas . The only other study assessing the prevalence of asthma symptoms in rural indigenous Latin American children was performed in the highlands in Guatemala, where a low prevalence of asthma symptoms of 3.3 percent was observed in children aged four to six years . We observed an increased prevalence of asthma symptoms in children living in households using wood fires for cooking compared with children living in households where gas was used for cooking (AOR 2.12, 95% CI 1.18 - 3.84). Other studies addressing the relationship of asthma symptoms and biomass smoke showed similar results. Schei et al. observed a higher prevalence of asthma symptoms in children exposed to smoke from open fires compared to children exposed to cooking with planchas, a local type of chimney stove used in rural Guatemala (AOR 3.4, 95% CI 1.3 - 8.5) . Melsom et al. observed a higher prevalence of asthma symptoms in Nepalese children 11 to 17 years of age who were exposed to cooking on open fires compared to children who were exposed to cooking on gas or kerosene stoves (AOR 2.2, 95% CI 1.0 - 4.5) . In a case–control study in Nairobi a higher prevalence of asthma symptoms in children sleeping in bedrooms with damage from dampness compared to children sleeping in bedrooms without damage from dampness was observed (AOR of 4.9, 95% CI 2.0 - 11.7) . However, several other studies did not observe a significant association between asthma symptoms and indoor wood smoke. Noorhassim et al. performed a cross-sectional study in 1007 children aged one to 12 years in Malaysia and did not observe an association of asthma symptoms to wood stove cooking, use of mosquito coil or smoking parents . No significant associations of asthma symptoms to cooking fuel type were observed in Malaysian and South African children, while exposure to environmental tobacco smoke at home was associated with an increased likelihood of asthma symptoms [14, 18].
Living in a house without walls was positively associated with asthma symptoms in our survey. Furthermore, a significant association of the variable “house without walls” with the variable “cooking on wood” was observed. We think that this association can be explained by the difference in the degree of acculturation between households. Warao people traditionally live in houses without walls. Over the past 20 years, acculturation has led to more Warao families building houses with walls for protection from the weather or for protection of newborns . Thereby, cooking on gas has become increasingly popular among Warao people as this makes heavy wood gathering unnecessary. Less acculturated families usually cook on wood fire and live in houses without walls while more acculturated Warao more often have a house with walls and gas stoves. There are some studies showing a relationship between the level of acculturation and asthma symptoms, but none were performed in indigenous populations. A study performed in the United Kingdom and Scotland showed a higher prevalence of persistent wheezing in children whose fathers’ social class was low and in those living in high poverty index score areas compared to children whose fathers belonged to a higher social class and children from low poverty index score areas (p < 0.001) . Furthermore, a study performed in New York City showed that asthmatics were five times more likely to live in public housing than non-asthmatics .
We observed a higher prevalence of asthma symptoms in the children exposed to more than ten cigarettes per day compared to children not exposed to cigarettes (AOR 2.69, 95% CI 1.11 - 6.48). There is ample evidence of a strong positive association of environmental tobacco smoking to asthma symptoms as stated in the Global Asthma Report . Maternal smoking seems to be particularly influential on the prevalence of asthma symptoms in children . The mother was the primary caregiver in 92 percent of the cases in our population, and only four percent of the children in our study had a smoking mother. In our study, fathers were responsible for the highest exposure to tobacco smoke, namely in 67 percent of the children, while they were the primary caregiver in only one percent of the children. In contrast to the findings in other studies, the prevalence of asthma symptoms in our study was not higher in children exposed to one to ten cigarettes daily compared to non-exposed children. We hypothesize that this is due to the finding that the primary caregivers were smokers in very few families in our study and exposure to tobacco smoking is probably not substantial when the people who smoke are not the caregivers.
There is evidence that environmental exposure to triggers, including particulate air pollution, is associated with a non-specific lung irritation effect rather than with asthma. The results of an experimental study using a mouse model of allergic airway inflammation suggest that woodsmoke exposure can exacerbate rather than cause allergic airway inflammation . As no prospective studies assessing the influence of woodsmoke on the development of asthma have been performed, it is not known whether indoor combustion is associated with the development of asthma or with exacerbation of symptoms among asthmatic individuals . The latter hypothesis would suggest that woodsmoke has a non-specific lung irritation effect rather than an etiologic contribution to the development of asthma. An experimental study performed by Riddervold et al. supports this hypothesis since they did not observe a significant effect of wood smoke exposure on lung function or cytokine levels in non-smoking volunteers without bronchial hyper-responsiveness .
Preventive measures should be taken to avoid health damage related to the use of open wood fires. Harris et al. observed a 40–73 percent decrease in clinic visit rates for lower respiratory infections in children 0 to 5 years of age in the year 2006 compared to the year 2002, when chimney stoves were installed (p < 0.01) . The introduction of two types of improved stoves in rural communities in Peru resulted in a 42 - 54 percent reduction of personal exposure to particle matter (PM) in women (p < 0.05) . Albalak et al. measured PM in 30 households in rural Guatemala over a period of 8 months. They also found a significant reduction of PM when open fires were replaced by chimney stoves in a rural setting in Guatemala (p < 0.05) .
The indoor levels of PM and carbon monoxide (CO) found in studies performed in rural areas are much higher than WHO recommendations of indoor air quality [38, 39]. The content of the smoke produced by biomass containing PM and CO is thought to cause the respiratory problems such as asthma and respiratory infections . Bruce et al. investigated the relationship between the time period that children younger than 18 months of age were present in the kitchen during cooking and CO exposure. During two observations, they observed a significantly higher exposure to CO when the child was in the kitchen twice compared with when the child was in the kitchen only once (p < 0.01) . This means that education about the risks of wood smoke, improved chimney stoves and keeping the children out of the kitchen during cooking could play a role in asthma prevention.
There are a number of limitations to our study, some of which are related to the challenging logistics of conducting an epidemiological study in an area with a poor infrastructure characterized by low literacy and poor access to health care. First, we added clarifications adjusted to local terminology to the ISAAC questionnaire. In order to minimize errors in reporting of asthma symptoms in these populations, questionnaires must use terminology familiar to participants . Although these clarifications as well as the study aims and definitions were discussed with community elders, local health care workers and local translators prior to initiation of the survey, it remains possible that the verbal translation of the questions into the native Warao language was not always accurate. This could have led to under- or over-reporting of symptoms when parents did not speak Spanish.
Second, although we performed a multivariable analysis taking into account possible confounders such as age and sex, unmeasured factors may have caused residual confounding. Overcrowding, high parental education levels, a family history of asthma, shorter duration of breastfeeding and having pets in the home have been positively associated with self-reported asthma in other studies using the ISAAC questionnaire [43–48]. The prevalence of asthma symptoms increased in children that were breastfed for less than 6 months in a population-based prospective cohort study of Sonnenschein-van der Voort et al. . As virtually all Warao children are breastfed until at least 12 months of age, it is not likely that the lack of information on duration of breastfeeding has influenced our study results. The same accounts for parental education, as Warao parents generally have not received any formal education. However, the lack of information on a family history of asthma and the presence of pets in the household may have caused residual confounding. A positive association between parental asthma and asthma symptoms in children has been observed in other studies [43, 44, 48]. It is, however, questionable whether the answers to questions related to a family history of asthma would have been reliable in our study setting. As Warao women usually have eight or more children, families are large and the constant migration of mainly male adults to and from other communities means that families live scattered across the Delta Amacuro. Due to the poor infrastructure and the lack of telephone or internet access, there is very little contact between family members, if any. The role of pets in the development of asthma symptoms is controversial . Although positive associations between pets in the household and asthma symptoms have been described in single-center studies [43, 44], no significant association of current cat or dog exposure to asthma symptoms was observed in children under ten years of age in a third phase ISAAC study including children from 98 countries in all parts of the world .
A third limitation to our study is the cross-sectional study design. As we measured the prevalence rather than the incidence of asthma symptoms, it remains unknown whether woodsmoke is a risk factor involved in the etiology of asthma symptoms. Belanger et al. performed a review including studies assessing the association of asthma symptoms with indoor combustion sources and concluded that no studies measuring asthma symptom incidence have been published . Prospective birth cohort studies are needed to determine the role of woodsmoke exposure in asthma etiology and the severity of asthma symptoms over time.
Fourth, we measured the prevalence of asthma symptoms anamnestically without objectively assessing lung capacity, for example by spirometry. However, spirometers are not available in the small health posts in the Delta Amacuro and no personal medical records in which spirometry results or asthma episodes could be recorded are kept. As spirometry values are usually normal between attacks of asthma symptoms, the performance of spirometry in our cross-sectional survey would not have been of additional value to the diagnosis of asthma. Furthermore, the ISAAC questionnaire is a validated and widely used instrument for measuring asthma symptoms in rural settings. Yeh et al. observed a significant advantage of the ISAAC video questionnaire compared with the ISAAC written questionnaire for the accurate diagnosis of asthma symptoms compared to the golden standard spirometry . In our survey it was, however, not feasible to use the ISAAC video questionnaire due to limited power supplies.
Finally, the sample size of our study was small in comparison with the ISAAC studies including more than 50,000 children [1, 3, 52, 53]. The ISAAC study performed in Latin America included 88,813 children from nine different Latin American countries . There are, however, several limitations to the interpretation of the ISAAC Latin American data: a) most of the study sites were concentrated in large, mainly urban areas, b) many of the participating countries have levels of socioeconomic development comparable to industrialized nations and c) sampling of participants was not representative of the different socioeconomic strata within each country. There is a relative absence of epidemiologic studies in rural, poor, indigenous populations in Latin America while ten percent of the Latin American population consists of indigenous people . Our study provides estimates of the prevalence rate of asthma symptoms in children in a rural indigenous population in Latin America as well as insight into the factors associated with asthma symptoms in these children.