Discussion
We have tested the hypothesis that early life exposures to geohelminths - through an infected mother during pregnancy or early childhood, or both - protect against wheeze/asthma and atopy at school-age. To do this, we did a birth cohort study to measure the effects of maternal and early childhood geohelminths on the development of atopy (measured as SPT), wheeze/asthma, and airways reactivity and inflammation at 8 years. Our findings indicate that maternal geohelminths have persistent effects on childhood SPT but that this effect was strongest among children of infected mothers who also acquired infections during early childhood. A maternal effect on increased wheeze and airways inflammation was seen among non-atopic children, the dominant phenotype in non-affluent societies.21,22 The maternal effect on SPT was not associated with a specific parasite species, while that on wheeze appeared to be mediated by T. trichiura infection. In contrast, early childhood T. trichiura protected against wheeze irrespective of atopy.
There are few previous longitudinal analyses of the effects of early geohelminth infections on development of allergy, and none of these have adequately addressed effects of maternal or childhood geohelminths on asthma or atopy: 1) a birth cohort in Ethiopia that did not measure maternal gehelminths and in which the prevalence of geohelminths (<4%) in early childhood was too low to explore effects on allergy at 5 years;23 and 2) a longitudinal study in Brazil, with no data on maternal geohelminths, showed that T. trichiura infections in early childhood, particularly at higher parasite burdens were associate with a reduced risk of SPT in later childhood.24 To our knowledge, the only other study to show effects of maternal geohelminths on allergy-related outcomes was an observational analysis in Uganda showing maternal hookworm to be associated with a reduced risk of eczema in children to 5 years.25
Previous cross-sectional studies have provided evidence that childhood geohelminths might protect against wheeze/asthma: 1) a study in Ethiopia in 1-4 year olds showed a negative association between A. lumbricoides infection and wheeze;26 2) a study among schoolchildren in a rural region in Ecuador showed an inverse association between heavy infections with T. trichiura and atopic wheeze27 - most previous cross-sectional studies, however, showed no effects of T. trichiura on asthma symptoms;8,21,28,29 and 3) three separate studies in Ethiopia showed an inverse relationship between hookworm infection and asthma symptoms.8 With respect to A. lumbricoides infections in school-age children, several studies have shown a positive association between A. lumbricoides infection or allergic sensitization to Ascaris antigens and asthma symptoms8,28,30,31 and airways reactivity,29,31,32 an effect that was strongest in non-atopics.28 Our data showed positive associations between greater parasite burdens with A. lumbricoides in mothers and risk of asthma and markers of airways inflammation in non-atopic children, while A. lumbricoides in children was associated with elevated FeNO.
Our observation that maternal infections protect against atopy are consistent with observations of inverse associations between geohelminths and SPT from cross-sectional studies of schoolchildren.21,24,33 A protective effect of maternal geohelminth (against mite) was present from 3 years of age.15,16 Childhood infections appeared to protect against SPT to perennial allergens from 5 years,16 but the observation that the effect was only seen among children with infected mothers irrespective of childhood infection status, indicate the maternal effect is key. Maternal geohelminths were strongly associated with childhood infections to 5 years of age – reflecting a shared risk of infection in the household environment – a child growing up in a household where one or more family members are infected, is at greater risk of infection.34 The previous observation from Brazil showing a protective effect of early life T. trichiura infections against SPT at school age24could have been mediated by maternal infections which were not measured but with which early childhood infections are likely to be strongly associated. A maternally-mediated effect on SPT could explain two previous observations from Ecuador: 1) bimonthly anthelmintic treatments in schoolchildren showed no treatment effect on allergen SPT;10 and 2) community mass drug administrations with the broad-spectrum anthelmintic, ivermectin, over 15 years for the elimination of onchocerciasis, was associated with an increase in SPT prevalence in schoolchildren. Long-term ivermectin started before most children were born, likely resulted in reduced geohelminth infections in mothers.35
We have shown previously in this population that newborns of mothers infected with A. lumbricoides have evidence of sensitization of CD4+ T cells to Ascaris antigens.36 The same is likely to be true for T. trichiura that, although purely enteric, has an intimate relationship with the mucosal immune system.7 Certainly, geohelminth antigens are present in the blood37 of infected mothers and can cross the placenta to sensitize the foetus. Immunological sensitization of the foetus could increase or decrease immune responsiveness. Decreased responsiveness could be associated with tolerization to parasite allergens including those that are cross-reactive with aeroallergens. Extensive cross-reactivity has been demonstrated between helminth parasites and aeroallergens such mite allergens,38 and such cross-reactivity can mediate cross-sensitization in immediate hypersensitivity skin reactions in murine models.39The suppressive effect of maternal geohelminths on SPT, particularly to mite allergens, in children could occur through tolerization to cross-reactive allergens.
The effect of maternal T. trichiura on childhood wheeze likely occurs through a distinct non-allergic mechanism, perhaps through the interconnected mucosal immune system.40 The evolutionary significance of such an effect resulting in increased mucosal responsiveness could, for example, increase chemical signals leading to airways reactivity and wheeze symptoms (as observed here) but the same signals in the gut might enhance peristalsis and expulsion of parasites. The type of immune response generated in the foetus likely will be affected by a number of factors such as host genetics, the ‘intensity’ of exposure, and geohelminth parasite species and could be parasite antigen-specific. The maternal effect of T. trichiurawas evident only among children of infected mothers who did not acquireT. trichiura infections during the first 5 years of life. The acquisition of childhood T. trichiura abrogated the maternal effect indicating that in utero effects could be modified by exposure during childhood, presumably by modulation of the same mechanisms.
Asthma is a highly heterogeneous disease for which several phenotypes and endotypes have been described.41 Numerous traits have been described for asthma (e.g. airflow limitation and airway inflammation) caused by distinct causal mechanisms.42Similarly, geohelminths are a diverse group of parasites with distinct life cycles and niches within (and outside) the intestine in humans. The human host has developed a wide variety of inflammatory mechanisms, primarily mediated by Th2 cytokines, with which to kill and expel geohelminths.7 Balanced parasitism requires an accommodation between host and parasite to allow parasites to survive without severely debilitating their host. Such an accommodation includes the modulation of host anti-parasite Th2 responses.7Our observations of parasite species-specific effects on different ‘traits’ (e.g. wheeze symptoms, airways reactivity and elevated FeNO) likely reflects this complex interaction.
Strengths of the study include prospective design with follow-up from birth, stool data on maternal geohelminths during pregnancy, and collection of large number of sociodemographic and lifestyle variables allowing us to control for potential confounders. Potential biases were reduced by using objective measures of geohelminth infections, performing all evaluations blind to the child’s exposure status, and high retention in the cohort to 8 years (~80%). Repeated exposure measures for childhood geohelminths during the first 5 years of life provided more precise estimates of infection rates but children with positive stools were treated thus reducing prevalence and parasite burdens.