Throughout the study, we were not capable of demonstrating the hygiene measures’ effect on asthma, but we found other factors to be associated with it. The multivariable analysis findings showed an association between placing a heating system only in the bedroom and higher asthma occurrence in preschool children, whereas installing the heating system in the whole house was associated with lower asthma, in agreement with previous findings 41. Indeed, insulation of the house provides a warmer and drier indoor environment, therefore leading to amelioration in wheezing, school absenteeism, and trips to the hospital caused by respiratory conditions 42.
In addition, a significant association was found between smoking cigarettes and waterpipe inside the house, and the increased asthma odds in children. These results consolidate ones from previous studies in Lebanon 27,29. Amid children with asthma, the ones exposed to second-hand smoke are about twice as likely hospitalized for an acute episode of asthma, having poorer results of pulmonary function test, as estimated by a systematic review 43. Asthma prevalence is higher among children living with smokers in their homes, and the likelihood of asthma development increases proportionally with the number of smokers at home 44. Despite this available evidence, waterpipe is still wrongly considered by smokers to be less dangerous and less toxic than cigarettes 45. In fact, pregnant and nursing women may stop smoking cigarettes but not waterpipe, and this was proven to lead to higher risks of allergies and asthma among Lebanese children 29,46. Furthermore, children are frequently allowed to play in closed rooms with ongoing waterpipe smoking. There is a misconception about waterpipes compared to cigarettes, waterpipes being perceived as less addictive and containing less nicotine.
The likelihood of parents having a history of seasonal allergy or asthma was significantly associated with increased asthma odds in preschoolers, in accordance with preceding studies demonstrating that a risk factor associated with atopic wheeze in children is parental asthma 47,48.
As for the hygiene hypothesis, this is the first study using a validated scale to assess the association between the hygiene hypothesis related behaviors and asthma in preschool children; to the best of our knowledge. We investigated cleaning behaviors (related to cleaning the house) and individual cleanliness habits (related to washing the child’s hands, showering, etc.) using an extensive questionnaire enfolding over 40 relevant items. After adjusting all sociodemographic, socioeconomic and other factors described in the literature as being associated with asthma, the results of the multivariable analysis showed that hygiene hypothesis related scale was not significantly associated with asthma among this population group.
This could be explained by several factors. Firstly, the hygiene hypothesis’s different aspects could not be fully investigated in our study since we could not classify the area of living into rural or urban: in fact, there is no official classification in Lebanon for rural and urban regions. Moreover, the rather small total sample size may not be representative of the entire population; also, this small sample size can result in a power problem (false negative findings). The refusal rate, as well as the fact that the study was conducted in 3 out of 5 districts in Lebanon, can cause a plausible selection bias. The use of a questionnaire in surrogate responders (parents) might not be accurate all the time due to difficulties in comprehending questions, recalling, and symptoms’ over- or under-evaluation; leading to a plausible information bias. Due to the retrospective nature of this investigation, a recall bias may be entertained; defining asthma using questions about a physician’s diagnosis is acceptable according to the literature, but may nevertheless present a risk of over or under-diagnosis. Various toxics exposure was subjective and quantified as reported by the parents’ estimation. For known risk factors of asthma, the recall bias effect can be differential, leading to effects overestimation by parents of children with the disease. Nevertheless, the bias is mainly non-differential for substances that are not acknowledged to be associated to asthma, and the expected association with asthma is underestimated. The hygiene hypothesis being relatively unknown in Lebanon, we expect the bias to be non-differential in this case.
Finally, some other factors related to hygiene may have not been taken into consideration because of the limited funds to accomplish this project. In infants developing allergic disease, differences in the gastrointestinal flora were identified by some investigators 49,50. Even changes in pregnant women’s vaginal microflora were recently associated in early childhood with wheezing outcomes 51. The association between some parasitic infections and less asthma and allergic diseases in children was also shown 52,53, and the immune deviation in early life is somehow influenced by prenatal farm environment exposure 54 leading to a protective outcome when it comes to asthma development in schoolchildren 6,7. Unfortunately, not being able to measure these factors could be a limitation to this study.
Limitations and strengths
The number of children with asthma is rather small, which raises the possibility of false positive findings. The findings cannot be extended to imply causation and cannot be generalized to the whole population; in addition to false positive findings, the associations we describe may have arisen by confounding or be reverse causation (the latter especially may explain the association with playing outdoors and having a heater in the bedroom). In all cases, prospective studies of higher sample size that override the majority of the drawbacks would be anticipated to ameliorate our results’ precision. Future studies that assess the intensity and exposure time for each toxic are also suggested to confirm our findings. Despite all the limitations, we have no reason to doubt the overall validity of our results since our methodology corresponds to other cross-sectional research studies, comprising ISAAC ones that are widely used for international comparisons.