Published ahead of print on June 15, 2006, doi:10.1164/rccm.200603-331OC
© 2006 American Thoracic Society doi: 10.1164/rccm.200603-331OC
Asthma and Current Intestinal Parasite InfectionSystematic Review and Meta-AnalysisDivision of Epidemiology and Public Health and School of Pharmacy, University of Nottingham, Nottingham, United Kingdom Correspondence and requests for reprints should be addressed to Jo Leonardi-Bee, Ph.D., Division of Epidemiology and Public Health, University of Nottingham, Clinical Sciences Building, Nottingham NG5 1PB, UK. E-mail: jo.leonardi-bee{at}nottingham.ac.uk
Rationale: Epidemiologic studies suggest that intestinal parasite infections may protect against asthma. Objectives: A systematic review and meta-analysis of epidemiologic studies to determine whether total or species-specific current parasite infection is associated with a reduced risk of asthma or wheeze. Methods: We searched MEDLINE, EMBASE, and CINAHL (up to January 2006); reviews; and reference lists from publications, with no language restrictions. We included studies that reported asthma or wheeze as an outcome measure and ascertained parasite infection by fecal examination. We estimated pooled odds ratios (OR) and 95% confidence intervals (CI) using data extracted from published papers, or where available, original data provided by authors, using random effect models. Measurements and Main Results: Thirty-three studies met the inclusion criteria. Infection with any parasite was associated with a small, nonsignificant increase in asthma risk (OR, 1.24; 95% CI, 0.98–1.57; 29 studies). In species-specific analysis, Ascaris lumbricoides was associated with significantly increased odds of asthma (OR, 1.34; 95% CI, 1.05–1.71; 20 studies), while hookworm infection was associated with a significantly strong reduction (OR, 0.50; 95% CI, 0.28–0.90; 9 studies) that was directly and significantly related to infection intensity (p < 0.001; OR for highest tertile of infection, 0.34; 95% CI, 0.19–0.62). Other species had no significant effects on asthma. Infection effects on wheeze were derived from smaller numbers, but revealed a broadly similar pattern of results. Conclusions: Parasite infections do not in general protect against asthma, but infection with hookworm may reduce the risk of this disease.
Key Words: asthma • meta-analysis • parasites • systematic review Asthma is one of the most common chronic diseases of developed countries, but is relatively rare in the developing world, especially in rural subsistence communities (1–3). Infection with intestinal parasites has been postulated as an important factor that could produce these geographic differences in prevalence (4–7) through various mechanisms that may suppress or inhibit the immune processes leading to the asthma and other allergic diseases (8–10). Although a review of the epidemiologic evidence in 1985 concluded that the data "neither refute nor support the theory that parasite infection protects against asthma" (11), the relation between parasite infections and asthma risk has continued to attract interest (1, 12–19). A further potentially relevant observation from some of our own work in Ethiopia (13, 20) that is consistent with earlier findings on Schistosoma infection in Gabon (15) is that protection against asthma may be greater with parasite species that have a host systemic phase in their life cycle, and may also be related to the intensity of infection (13). We have therefore reviewed and performed a meta-analysis of the epidemiologic literature to determine whether, in pooled data from the published literature to date, parasite infection is associated with a reduced risk of asthma or wheeze, and whether there are any significant species-specific or intensity related effects. The preliminary results from this study have been previously reported in the form of an abstract (21).
Systematic Review Methods A comprehensive literature search in MEDLINE (1966 to January 2006), EMBASE (1980 to January 2006), and CINAHL (1982 to January 2006) was performed according to standard guidelines (22) to identify all epidemiologic studies with no restrictions on language. We also searched published reviews, reference lists from identified publications, and abstracts from major conference proceedings. Studies were included if they fulfilled the following criteria: (1) the design was a comparative epidemiologic study; (2) asthma or wheeze was listed as an outcome; (3) current parasite infection was measured by direct fecal microscopy. Studies that met the inclusion criteria were scored for methodologic quality using the Newcastle-Ottawa Quality Assessment Scale (23). A score of 7 or more was chosen a priori to indicate high methodologic quality. We attempted to contact at least one author from each identified study to invite them to share original study data. Additional detail on the methods used for the systematic review is provided in an online data supplement.
Statistical Analysis Heterogeneity between study estimates was assessed using established methods (25). To explore reasons for heterogeneity between the studies, subgroup analyses were used to assess the differences between developed and developing countries, timing of when the studies were published (before 1980 or from 1980 onwards), the definitions of asthma used within the studies (clinician assessed, self reported, exercise-induced bronchoconstriction/spasm), and the effect of species-specific parasite associations with asthma or wheeze for parasites infecting 1% or more of the available study population. To assess the effect of a dose–response relationship between intensity of species-specific parasite infection and asthma, we analyzed original data from studies that recorded intensity of infection in eggs per gram of feces. Original data were also used to explore the possibility that the effect of parasite infections on asthma would be different for infection at early versus later age (dichotomized into above or below 2 yr). Publication bias was assessed using a funnel plot and Egger's Asymmetry test (26). Data were analyzed using STATA, version 8.0 (Stata Corporation, College Station, TX) and SAS for Windows, version 8 (SAS Institute, Inc., Cary, NC). p values less than 0.05 were considered statistically significant. Additional detail on the methods for statistical analysis is provided in the online supplement. The work was performed in accordance with the MOOSE (Meta-Analysis of Observational Studies in Epidemiology) guidelines (27).
Overview of the Included Studies Our search identified 57 studies that assessed the association between parasites and asthma or wheeze published between 1966 and January 2006, including one published only in abstract form (28) and one unpublished study (Muniz PT, de Castro TG, de Araújo TS, Nunes NB, Santiago FS, Elefant GR, Nunes MdS, Hoffmann EHE, Ferreira MU, and Cardoso MA. Saúde e nutrição infantil na Amazônia Ocidental Brasileira: inquérito de base populacional em dois municípios acreanos. Unpublished manuscript, 2004) (Figure 1). Thirty-three of the 57 studies met our inclusion criteria. Three of the eligible studies were excluded because they either did not present outcomes in a form appropriate for use in meta-analysis (14, 29) or presented parasite infection as a geometric mean of egg count for individuals with and without asthma (30) (Figure 1). A total of 29 studies (7, 13, 18–20, 28, 31–52, and Muniz et al.) were available for analysis of asthma as an outcome, and 7 for analysis of wheeze (13, 18–20, 34, 53, and Muniz et al.). The studies included 25,753 individuals (Table 1). Methods of defining or measuring asthma or wheeze varied substantially between the studies, but the most widely used method was based on a clinical diagnosis of asthma (7, 31–33, 35, 37–40, 42, 43, 45–49, 52). Some surveys used questionnaire ascertainment of a self-reported (or for children, parent-reported) diagnosis of asthma (13, 18–20, 34, 36, 41, 44, and Muniz et al.), and three studies used a measure of exercise-induced bronchoconstriction (28, 50, 51). Several case-control studies used clinically defined asthma cases drawn from hospital patient populations.
Of the 30 studies included in the meta-analyses, 5 included only adults (13, 33, 39, 41, 43), 14 only children (18–20, 28, 31, 35–37, 42, 44, 46, 50, 51, and Muniz et al.), and 11 studies included both adults and children (7, 32, 34, 38, 40, 45, 47–49, 52, 53). Thirty-nine percent of the individuals included in the studies were currently infected with parasites. We were able to establish contact with investigators from 20 of the original 57 identified studies, including 13 of the studies included in the present analysis (Table 1), and nine of these provided original data to enable adjusted analyses to be performed (13, 18–20, 34, 42, 44, 46, and Muniz et al.). The original data from most of the remaining studies had been discarded.
Methodologic Quality of Studies and Publication Bias
Effects of Current Infection with Any Parasite on Asthma
In a subgroup analysis of studies that were performed in developed countries, high levels of heterogeneity remained between studies (I2 = 68%, 5 studies) however this was attributed to the study by Tullis (49), which was extremely unusual in that it found that over 95% of asthma cases recruited from a Canadian sanatorium were infected with parasites. These findings are so atypical of the other evidence that they suggest a local outbreak of parasite infection at the sanatorium may have been responsible. Excluding this study from the analyses significantly reduced heterogeneity (I2 = 1%, 4 studies). In the developing countries, the results from studies published before 1980 were less heterogeneous (I2 = 53%, 11 studies) than those published in the last 25 yr (I2 = 83%, 14 studies). In a subgroup analysis based on the method of ascertainment of asthma, cases of clinician-diagnosed asthma were significantly more likely to be infected (OR, 1.47; 95% CI, 1.02–2.12; 17 studies). Consistent pooled results were seen in studies that used self-reported ascertainment (OR, 1.24) and exercise-induced bronchoconstriction ascertainment (OR, 1.42). High levels of heterogeneity were seen among studies that used clinician-diagnosed (I2 = 79%, 17 studies) or self-reported (I2=72%, 9 studies) definitions of asthma; however, considerably less heterogeneity was seen for studies that used exercise-induced bronchoconstriction/spasm (I2 = 16%, 3 studies).
Effects of Infection with Individual Parasite Species on Asthma
Current infection with A. lumbricoides was associated with a significant increase in the risk of asthma (OR, 1.34; 95% CI, 1.05–1.71; p = 0.02; 20 studies; Figure 3A), whereas current infection with hookworm was associated with significant reduction in risk (OR, 0.50; 95% CI, 0.28–0.90; p = 0.02; 9 studies; Figure 3B). Infection with T. trichuria, E. vermicularis, or S. stercoralis was not associated with significant effects on asthma risk (Figures 3C–3E). The pooled estimate for current infection with S. mansoni indicated a numerically high but nonsignificant increase in asthma risk (OR, 2.77; 95% CI, 0.54–14.28; 2 studies). High levels of heterogeneity among study findings were seen for A. lumbricoides and hookworm (74 and 67%, respectively); however, little heterogeneity was seen among the studies for T. trichuria, E. vermiculari, or S. stercoralis.
Effects of Current Infection on Wheeze
Effects of Infection Intensity
Interactions with Age There was no evidence of age by species-specific interaction for the relation between A. lumbricoides and asthma (interaction p = 0.38) or wheeze (interaction p = 0.30), or for hookworm and wheeze (interaction p = 0.60). There were no cases of concurrent hookworm infection and asthma at less than 2 yr of age.
Parasite infections are very common, particularly in the developing world, and systematic eradication programs are being introduced in many areas. Although eradication is undoubtedly beneficial to many aspects of individual and public health, the hypothesis that parasite infections may protect against allergic disease raises the possibility that eradication may increase asthma and other manifestations of allergy. For this reason, and also because any protective effect is likely to provide important insights into pathogenesis and possible therapeutic opportunities in allergic disease, it is important to establish as clearly as possible whether parasite infections and asthma are indeed related. This systematic review and meta-analysis indicates that any relation between intestinal parasite infection and asthma risk is likely to be species specific, and for hookworm, intensity related. Within the overall finding of no significant effect of infection on asthma, there were two opposing effects of individual species, with A. lumbricoides being associated with an increased risk, and hookworm (predominantly N. americanus in these studies) a more marked decrease. The other intestinal parasite species we were able to study individually (T. trichuria, E. vermicularis, and S. stercoralis) had no clear effect. The methodologic quality of studies included in this review, particularly those reported more recently, was generally good. We are also confident that our literature search, combined with the results of contacting investigators from most of the research groups known to be engaged in this area, identified the great majority of potentially suitable studies. Our funnel plot analysis indicates that publication bias is unlikely to have had a marked effect on our findings. We maximized the validity of our exposure measure by restricting analysis to studies using direct fecal examination for eggs or parasites, because measures based on serologic tests do not distinguish current from past infection. Most studies available to us used asthma as an endpoint, which was defined in several ways. Although asthma is an outcome measure potentially susceptible to both diagnostic and ascertainment bias, the fact that our findings were qualitatively broadly similar for the relatively small number of studies using self-reported wheeze as an outcome suggests that these biases are unlikely to be a major influence on our findings. We were able to obtain original data from several investigators, allowing us to control for sex and the potential confounding effect of age, but the data from many studies (particularly the older ones) were unavailable. However, the data available to us did not allow us to adjust for some other potentially important confounders or sources of heterogeneity, particularly socioeconomic status and urban/rural residence (13). Because only two of the included studies provided parasite-specific IgE data, we were unable to explore these effects in asthma any further. Our analysis of species-specific effects was performed to test the hypothesis arising in part from previous work (13, 20) that any protective effect was more likely to arise from intestinal parasites such as A. lumbricoides and hookworm, which include a host systemic phase in their lifestyle. The evidence, however, indicated that these parasites were associated with different effects on asthma risk, being greater for A. lumbricoides and lesser for hookworm. If our findings in relation to these species are in fact correct, the reason for this discrepancy is unclear but may relate to the relatively high allergenicity of A. lumbricoides (54), which unlike hookworm, is a recognized cause of tropical pulmonary eosinophilia. The different effects may also arise from species-specific differences in the immune polarization and other potentially antiallergic effects that have been suggested as the mechanism by which parasite infection might protect against allergic disease (10). It is true that the atopic state seems to confer an element of resistance to worm infections. We have previously shown that high levels of IgE and eosinophilia result in reduced worm fecundity and postchemotherapy reinfection rates (55). However, this does not detract from the hypothesis that hookworms may have evolved to survive by suppressing the allergic state, and hence have an effect on atopic disease. The relative timing of allergen exposure in relation to parasite infection is important as parasites have a documented ability to potentiate responses to allergen, and can even convert nonallergenic proteins into allergens during concomitant infection. Another possibility is that the effect is dependent on age at infection, because children tend to develop A. lumbricoides infection in the first years of life, whereas hookworm infection is acquired after children become old enough to be mobile and independent (56). We found no evidence in support of this in the age interactions we explored, but this remains a possible explanation. Our finding for A. lumbricoides is supported, however, by the results of a study of the effects of parasite eradication in asthma, in which treatment actually improved asthma control in participants from a population with predominant A. lumbricoides infection (57).
Parasites have a built-in antiinflammatory defense, and there is increasing evidence that these defense mechanisms include effects on regulatory T cells (for reviews, see References 10 and 58), of which there are three populations, some secreting antiinflammatory cytokines such as transforming growth factor- An alternative explanation is that the associations we have observed arise from reverse causation, and that, in particular, allergic individuals are less likely to acquire hookworm infection. The cross-sectional nature of the data available to us precludes any further insight into this possibility. Overall, however, our study indicates that different species of parasite infection may have important effects on the pathogenesis of asthma, and that in particular, the potential individual or public health benefits of hookworm infection merit further investigation.
Collaborating Investigators: Maria Helena D'Aquino Benício (Brazil), Philip Cooper (Ecuador), Damtew Dagoye (Ethiopia), Gail Davey (Ethiopia), Marcelo Urbano Ferreira (Brazil), Peter Hotez (United States), Peter LeSouef (Australia), Sarah Lewis (United Kingdom), Isabel Hagel (Venezuela), Marta Minvielle (Argentina), Maria Teresa Nascimento (Brazil), Sarah Scrivener (United Kingdom), Neda Sharghi (United States), Scott Weiss (United States), Andrea Venn (United Kingdom).
This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org. Originally Published in Press as DOI: 10.1164/rccm.200603-331OC on June 15, 2006 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of the manuscript. Received in original form March 7, 2006; accepted in final form June 12, 2006
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
and interleukin-10, and others that seem to depend on contact with activated T cells to survive. There is mounting evidence that this population of cells is expanded during parasite infection, but the pathologic impact of this expansion is not yet clear. 
