Parasites and allergy: Observations from Africa

Summary Population studies from the African continent have observed a marked increase in the prevalence of allergy‐related diseases over the past few decades, but the cause of this rise is not fully understood. The most investigated potential risk factor has been the relationship between exposure to helminths and allergy‐related outcomes. Immunologically, parallels exist between responses to helminths and to allergens as both are associated with elevated levels of immunoglobulin E, increased numbers of T helper 2 cells and other immune cells. However, epidemiological studies from the African continent have found inconsistent results. In this review, observations from population studies carried out in Africa over the last decade that focus on the relationship between helminth infections and allergy‐related outcomes are examined. How these findings advance our understanding of the complex interactions between helminths and allergies at the population level is also explored as well as some of the underlying immune mechanisms involved. This knowledge is important for better diagnosis, treatment and prevention of allergy‐related diseases and has wider global significance.


| INTRODUC TI ON
The global burden of helminth infections remains high with an estimated 1.5 billion people worldwide chronically infected with at least one soil-transmitted helminth 1 and 240 million having the waterborne helminth disease schistosomiasis. 2 It is also estimated that 92% of those with schistosomiasis live in sub-Saharan Africa. 3 Helminth infections are strongly linked to poverty, poor hygiene and inadequate sanitation and are more common in rural locations. 4,5 Reduced exposure to helminths is thought to be one of the factors driving the rise in the incidence of allergic disease worldwide. 6 Increases in the prevalence of allergy-related diseases over time have also been seen on the African continent especially among children, 7,8 particularly in urban compared to rural areas. [8][9][10][11][12][13][14] Aside from helminths, other factors have been linked to this rise including urbanization, less exposure to childhood pathogens in general and lifestyle changes. 15 Helminths are of specific interest to investigations on allergyrelated diseases due to the immunological parallels between these two conditions. Both are associated with T helper 2 (Th2) cell induction, high levels of immunoglobulin E (IgE) and the involvement of immune cells such as mast cells, basophils and eosinophils. 16,17 While immunological parallels exist, the resultant clinical outcomes are dissimilar and chronic helminth infections can induce an immune regulatory network in the host resulting in an anti-inflammatory environment as well as general T-cell hyporesponsiveness. 18 Although some epidemiological studies conducted in helminth-endemic countries have reported an inverse association between helminths and allergy-related outcomes, other investigations have observed either no association or even a positive association. In this review, we highlight findings from population studies conducted in Africa that explore the relationship between helminths and allergy-related outcomes with a particular emphasis on studies conducted within the past decade.
We identified relevant publications through searches conducted in PubMed using keywords related to "allergy" or "hypersensitivity" in combination with "helminths," "parasites," or "worms" and "Africa (South of the Sahara and Northern Africa)." Our search was restricted to human studies reported in the English language that were published between January 2008 and March 2018. The search was limited to the last decade to highlight new and recent findings in the literature given that older observations have been covered extensively elsewhere. [19][20][21][22][23] In this review, we first explore findings on helminths and reported/clinical allergy outcomes from epidemiological studies conducted in the past decade. Our reported/clinical allergy outcomes included asthma, wheeze, exercise-induced bronchospasm, rhinitis, conjunctivitis, eczema, reported adverse reactions to peanut and skin prick test (SPT) reactivity to allergens. We then examine population study findings related to immune mechanisms including helminths and IgE sensitization, cellular immune mechanisms and other pathways through which helminths may protect or increase susceptibility to allergy-related diseases.

| General characteristics of studies
Epidemiological studies that investigated the relationship between helminths and allergy-related outcomes (self-reported, doctordiagnosed or SPT), between January 2008 and March 2018, are shown in Table 1. A flow diagram showing the selection of publications for inclusion is displayed in Figure 1. In general, studies were (a) mostly observational, as either cross-sectional or case-control studies, (b) conducted among school-age children or adults, (c) conducted in either urban or rural areas or both and (d) one study was an intervention of anthelmintic treatment in pregnancy with subsequent follow-up of the offspring into childhood as part of a birth cohort.

| Prevalence of helminths and allergyrelated outcomes
As shown in Table 1, studies conducted across the African continent show a wide range for the prevalence of helminth infections with some investigations having high burdens (>50%), 11,31,37,38 while others had low percentages (<10%) 25,27,30,36 according to WHO classification. 40 Although the vast majority of investigations used the Kato-Katz method for the detection of soil-transmitted helminths and Schistosoma mansoni, and urine filtration for the detection of S. haematobium, a few studies used other methods such as formolether concentration. In a few studies, methods used were not specified. The type of helminth infections among participants varied from study to study, but the most common helminths were Schistosoma, hookworm, Ascaris and Trichuris species. The most commonly studied allergy-related outcomes were reported wheeze, eczema and asthma, all with a prevalence of less than 10%. The percentage of positive SPT responses varied from 2% 33 to 73%. 34 Generally, food allergens such as peanut elicited fewer SPT responses compared to environmental allergens such as house dust mite and cockroach.

| Association between helminths and allergyrelated outcomes from observational studies
Information on associations between helminth infections and allergy-related outcomes is summarized in Table 2 TA B L E 1 (Continued) not a meta-analysis, the total numbers of participants in the various studies are shown. Two of the three studies that showed a positive association were of Ascaris and respiratory symptoms of wheeze 38 or exercise-induced bronchospasms, 11 and the third between Trichuris and SPT. 36,38 Although migration of Ascaris larva through the lungs is mostly asymptomatic, on rare occasions this may be associated with respiratory symptoms such as wheezing, dyspnoea and bronchospasm. 41 Among rural and urban South African children, Calvert and Burney observed a positive association between A. lumbricoides infection and exercise-induced bronchospasms, but an inverse association between Ascaris and SPT positivity, 11 suggesting that the underlying mechanisms for SPT and bronchospasm are different.
Of the six studies that showed an inverse association between helminths and allergy-related outcomes (in Table 2), five of them were conducted in areas with a high proportion of individuals infected with helminths, 11,24,31,35,37 while the sixth study was conducted in an area with low proportion of individuals infected with helminths. 36 It is also important to note that although a study conducted among island communities in Lake Victoria where the proportion of individuals with helminth infections was high, no inverse associations with allergy-related outcomes were reported. 38 Tables 1 and 2 show that a large proportion of comparisons found no association between helminth infections and allergy-related outcomes. There are several possible explanations for this observation.
First, for most of these studies that showed no association, the prevalence of clinical allergy-related outcomes was low, 25,29,32,33,38,39 and/or the prevalence of helminths was low, particularly in urban areas. 25,27 This may have resulted in wide confidence intervals and thus reduced power to detect a statistically significant difference.
Some studies had small sample sizes and are likely to have been underpowered. 26,30,34 Although the prevalence of SPT was generally higher in many investigations, its association with helminths was not consistent for reasons that are not well understood.
Secondly, misclassification of helminth status may have occurred since many of these studies relied on the examination of a single parasitological sample using microscopy, which is not very sensitive. 42 Misclassification of allergy-related outcomes may have occurred since most studies relied on self-reported allergy-related outcomes from a questionnaire rather than doctor-diagnosed outcomes. Aside from questionnaires being a problem in settings where clinical diagnosis and translation of questionnaires into the local languages are a challenge, there may have been an over-estimation or under-estimation of self-perceived allergy. 43 Lastly, there is evidence that the risk of allergy-related outcomes is established early in life, 44 yet most investigations shown in Table 1 were conducted among school-age children or adults. Therefore   allergy-related disease risk. This study, conducted in Uganda, was a randomized placebo-controlled trial of albendazole vs placebo and praziquantel vs placebo (2 × 2 factorial design) in pregnancy, followed by albendazole vs placebo for the offspring between 15 months to 5 years of age. 45 This investigation showed that anthelmintic treatment during pregnancy resulted in increased risk of eczema in infancy 28 in the first 5 years 46 and that maternal hookworm during pregnancy was associated with reduced risk of eczema in infancy and in the first 5 years of life. 35 Albendazole treatment in early childhood was not associated with allergy-related outcomes. 35,46 However, the effects of anthelmintic treatment in pregnancy on clinical asthma could not be measured due to the small number of children who developed asthma in this cohort. 39 This birth cohort also showed that the children's own helminth infections in early life (hookworm and Trichuris) were associated with a reduced risk of eczema in later childhood. 35

| Dissociation between allergen-specific IgE and SPT reactivity in Africa
Observations from Africa have shown that allergen-specific IgE and SPT are not as strongly associated with allergy-related outcomes as seen in high-income countries and that this association is even weaker in rural compared to urban areas. For example, a study in

| Helminth-induced IgE cross-reactivity
The role of helminths in stimulating the production of IgE antibodies was first postulated in a study published in the late 1960s in which a group of Ethiopian pre-school children infected with the helminth A. lumbricoides were found to have 28 times higher levels of total IgE compared to Swedish children of the same age. 47 Since then, elevated levels of allergen-specific IgE associated with helminth infections that do not translate into allergy symptoms have been seen in numerous studies from Africa. 33,36,48 This may partly be due to the phenomenon of IgE cross-reactivity in which antibodies directed against one epitope recognize similar epitopes in homologous molecules. 49 Research studies over the past few decades have linked two kinds of cross-reactivity to allergens; cross-reactivity due to proteins and cross-reactivity due to the carbohydrate structures on glycoproteins known as cross-reactive carbohydrate determinants (CCDs).
With regard to helminths and protein IgE cross-reactivity, a number of allergens from invertebrate sources such as house dust mite, shrimp and cockroach have been shown to cross-react with helminth antigens. 50 In Africa, a role for helminths in cross-reactivity involving carbohydrates was demonstrated by an investigation conducted among schoolchildren in Ghana in which 18% were peanut IgEsensitized, but 92% of those sensitized were peanut SPT negative. 33 Additionally, current S. haematobium infection was positively associated with peanut IgE sensitization and a strong correlation was observed between IgE to CCDs and IgE to whole peanut extract. In a subset of children in that study, inhibition assays showed that S. haematobium soluble egg antigen and the CCD marker bromelain were strong inhibitors of IgE binding to peanut extract. 33 In a follow-up study that used a synthetic glycan microarray to identify the specific glycan motifs associated with carbohydrate-related IgE crossreactivity, 64 Schistosoma infection was linked to IgE cross-reactivity involving the core xylose glycan motif. 64 Although further investigations are needed to examine how CCD-specific IgE may inhibit allergic effector responses, Doenoff and colleagues have explored the role of IgG antibodies as blocking antibodies. 65,66 Through in vitro studies, rabbit anti-schistosome IgG was found to cross-react with allergens such as latex 65 and peanut extract. 66 The authors proposed that anti-schistosome IgG antibodies may block IgE-induced allergic responses and therefore prevent the manifestations of allergic disease. 66 Generally, IgE to CCDs are not thought to be biologically active, but in recent years, responses to the mammalian carbohydrate epitope galactoseα-1,3-galactose (α-gal) have been of interest in the field of allergy since this epitope has been linked to two forms of anaphylaxis. 67 The first being anaphylactic reactions following the infusion of the monoclonal antibody cetuximab among cancer patients undergoing therapy in the Southeastern United States 68 and the second being delayed-onset reactions hours after mammalian meat product consumption. 69 Reactions to α-gal have been documented from the Southeastern US, Central America, Australia and East Asia. 67 A history of tick-bite exposure has been shown to be strongly correlated with high levels of IgE to α-gal. 70  Although no direct link has been made between α-gal and helminths in Africa, elevated levels of IgE to α-gal have been seen in individuals from helminth-endemic areas in Ghana, 64 Kenya 73 and Zimbabwe. 74 In Ghana, α-gal sensitization has been strongly linked to rural residence rather than current helminth infection, 64 but further studies are needed to identify the specific factors associated with rural living that underlie this sensitization.

| Cell-mediated and Humoral Immune Mechanisms
Studies on immune mechanisms from population studies illustrate the role of helminths in protection from or susceptibility to allergy-related outcomes. The anti-inflammatory environment associated with helminths in the human host is characterized by elevated levels of the cytokines IL-10 and TGFβ, and general immune hyporesponsiveness. 18 In addition, the importance of IL-10 in helminth-induced regulation of the allergic immune response has been shown in some population studies. 48 in Lake Victoria, Uganda, found that in a subset of participants, reported wheeze was negatively associated with S. mansoni-specific cytokine responses. 77 These observations are in line with the protective effects of helminth infection on allergy-related outcomes.
When it comes to other immune mechanisms, a study among Ghanaian schoolchildren found that high expression of innate immune gene Toll-like receptor 2 (TLR2) and suppressor of cytokine signalling (SOCS)-3 messenger RNA (mRNA) was positively associated with SPT reactivity to house dust mite. 24  Allergen-specific IgG4 is often considered as a marker of immune modulation while allergen-specific IgE as a marker of allergic disease. 77 The IgG4 to IgE ratio can be utilized to determine immune modulation over the allergic response. The aforementioned investigation conducted on the islands in Uganda also looked at the ratio of allergen-specific IgG4 to IgE and found that house dust mite-specific IgG4 to IgE ratio was significantly lower in those reporting wheeze (P = 0.032). 77 In a study conducted in Zimbabwe, Rujeni and colleagues looked at the ratio of mite-specific (Der p 1) IgE to IgG4 in a low S. haematobium transmission community as well as in a high transmission community. 31 This mite-specific IgE to IgG4 ratio provides an indicator of allergic response over immune modulation and the study found that Schistosoma infection intensity was negatively associated with Der p 1-specific IgE to IgG4 ratio in the high transmission area. 31 Following treatment for schistosomiasis, in the high transmission area there was no change in ratio of Der p 1-specific IgE to IgG4 although the overall anti-Der p 1 IgE responses declined. 80 In the same province in Zimbabwe, antibody responses among pre-school children between the ages of 3 and 5 years before and 6 weeks after treatment were investigated; treatment had no effect on Der p1-specific IgE or IgG4 levels. 81 The low-affinity CD23 receptor is thought to be involved in the regulation of IgE synthesis. 82  Another immune mechanism that has been examined is the association between helminth infection and basophil cell suppression. A study in Uganda investigated the hypothesis that immunoregulatory responses that characterize helminth infections reduce the response of IgE effector cells, such as basophils, to IgE-mediated activation resulting in suppression of responses to parasite and non-parasite antigens. 37 For this investigation, changes in specific and non-specific histamine release in whole blood after anthelmintic treatment in schoolchildren from an area endemic for both S. mansoni and hookworm were examined. 37 The study findings were that among children without detectable hookworm infection, a significant positive association was observed between circulating levels of house dust mite-specific IgE and histamine release. 37  Studies conducted in Africa have observed a strong correlation between helminths and rural environments. For instance, studies that have reported an inverse association between helminths and allergy-related outcomes were predominantly from rural areas. 24,31,37 Since helminths are predominantly found in rural areas, observational studies may be faced with confounding from other factors in the rural environment that may not be fully adjusted for in statistical analysis. This makes it difficult to tease out whether helminths influence the risk of allergy-related outcomes in their own right, or through interaction with other environmental factors in rural areas. This is exemplified by the two studies conducted in two different fishing communities along the shores of Lake Victoria in Uganda; both had high percentages of helminth infections but one showed an inverse association with wheeze 37 while the other showed a positive association. 38 Therefore, future studies should explore immune mechanisms related to rural environments aside from helminths that influence the pathogenesis of allergy-related outcomes.

| CON CLUS ION
Studies from Africa investigating the association between helminths and allergy-related outcomes have found a number of inconsistent observations, some showing inverse, positive or no associations.
Since most of these studies are observational and conducted among school-age children, establishing causation is impossible. However, these studies have illustrated how IgE cross-reactivity between environmental or food allergens and helminth antigens has limited the diagnostic value of measuring IgE responses to whole allergen extracts in populations from helminth-endemic areas. They also raise pertinent questions about the classification of asthma as "atopic"/"allergic" or not based on the measurement of allergenspecific IgE to whole allergen extracts.
Although not consistent, studies on immune mechanisms have illustrated how helminth infections can induce immune regulatory responses that may protect against allergic immune responses.
The observed lower burden of allergy-related outcomes in rural compared to urban areas deserves further investigation. There is also evidence to suggest that the critical period worth investigating is early life. Therefore, studies in early life, either as anthelmintic trials during pregnancy and early childhood or as birth cohorts investigating a broad range of risk factors in both rural and urban areas, are needed. The current epidemiological transition in Africa offers an excellent opportunity to conduct such studies, so as to identify the primary factors driving the global rise in allergic disorders.

ACK N OWLED G EM ENTS
The authors would like to thank Professor Alison Elliott and Professor Maria Yazdanbakhsh for their critical comments and insightful feedback.

D I SCLOS U R E S
None.