Infectious disease incidence is often male-biased. Two main hypotheses have been proposed to explain this observation. The physiological hypothesis (PH) emphasizes differences in sex hormones and genetic architecture, while the behavioral hypothesis (BH) stresses gender-related differences in exposure. Surprisingly, the population-level predictions of these hypotheses are yet to be thoroughly tested in humans.
Methods and Findings
For ten major pathogens, we tested PH and BH predictions about incidence and exposure-prevalence patterns. Compulsory-notification records (Brazil, 2006–2009) were used to estimate age-stratified ♂:♀ incidence rate ratios for the general population and across selected sociological contrasts. Exposure-prevalence odds ratios were derived from 82 published surveys. We estimated summary effect-size measures using random-effects models; our analyses encompass ~0.5 million cases of disease or exposure. We found that, after puberty, disease incidence is male-biased in cutaneous and visceral leishmaniasis, schistosomiasis, pulmonary tuberculosis, leptospirosis, meningococcal meningitis, and hepatitis A. Severe dengue is female-biased, and no clear pattern is evident for typhoid fever. In leprosy, milder tuberculoid forms are female-biased, whereas more severe lepromatous forms are male-biased. For most diseases, male bias emerges also during infancy, when behavior is unbiased but sex steroid levels transiently rise. Behavioral factors likely modulate male–female differences in some diseases (the leishmaniases, tuberculosis, leptospirosis, or schistosomiasis) and age classes; however, average exposure-prevalence is significantly sex-biased only for Schistosoma and Leptospira.
Our results closely match some key PH predictions and contradict some crucial BH predictions, suggesting that gender-specific behavior plays an overall secondary role in generating sex bias. Physiological differences, including the crosstalk between sex hormones and immune effectors, thus emerge as the main candidate drivers of gender differences in infectious disease susceptibility.
Citation: Guerra-Silveira F, Abad-Franch F (2013) Sex Bias in Infectious Disease Epidemiology: Patterns and Processes. PLoS ONE 8(4): e62390. doi:10.1371/journal.pone.0062390
Editor: Hiroshi Nishiura, University of Hong Kong, Hong Kong
Received: March 15, 2012; Accepted: March 25, 2013; Published: April 24, 2013
Copyright: © 2013 Guerra-Silveira, Abad-Franch. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: Fiocruz-Fapeam agreement, Brazil. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Infectious diseases rarely affect males and females equally, despite even demographic sex ratios. This has been observed in humans – and other animals, from birds to invertebrates –. Two major, not mutually exclusive hypotheses have been put forward to explain such sex-biased patterns . The physiological hypothesis (PH) posits that the interactions between sex hormones and the immune system render one sex more susceptible to infection and disease, with genetic (chromosome) differences likely playing also a role , , , , –. Since sex and immunity are both critical to fitness and energetically costly, immune effectors and sex hormones engage in an intense physiological crosstalk –; in mammals, sex steroid immunomodulation has been linked to higher infection rates in males, with testosterone down-regulating and estrogen promoting T-helper(Th)1- and antibody-dominated responses , –, . On the other hand, the behavioral hypothesis (BH) posits that sex-biased infection rates emerge from sex-specific exposure to contagion , . Higher exposure usually arises from differences in behavior (foraging, combat), although sexual size dimorphism may also be important , . In some human populations, gender-related behavioral differences may render one sex more exposed to certain pathogens –, , , .
Several studies have addressed human male-female differences in overall mortality (e.g., ), susceptibility to allergic and autoimmune diseases (e.g., ), or individual infectious disease risk (e.g., –); yet, and surprisingly, a critical, comprehensive test of the major hypotheses outlined above is currently unavailable (but see ref. ). Aiming to address this gap, here we use a hypothesis-driven approach and two complementary, large datasets to investigate the roles of physiology and behavior as potential drivers of sex-biased infectious disease incidence in humans.
We consider the following major infectious diseases: American cutaneous and visceral leishmaniasis, schistosomiasis, pulmonary tuberculosis, lepromatous and tuberculoid leprosy, typhoid fever, leptospirosis, meningococcal meningitis, hepatitis A, and severe dengue fever. For each, we test the key predictions of the PH and BH as outlined below (see also Table 1).
Table 1. Testing hypotheses on sex-biased infection rates: pathogen traits and main expectations on sex-biased incidence under the physiological hypothesis (PH) and the behavioral hypothesis (BH).doi:10.1371/journal.pone.0062390.t001
Hypotheses and Predictions
Under the PH, infectious disease incidence is predicted to be similar among male and female children, who have similarly low levels of sex hormones; importantly, however, the transient rise of sex steroid levels in the first year of life, known as “minipuberty” , should yield patterns resembling more those seen after puberty than those seen in childhood. In age classes in which sex hormones are physiologically active (infancy, puberty, and reproductive period), the PH predicts that disease incidence should be male-biased for the leishmaniases, pulmonary tuberculosis, lepromatous leprosy, typhoid fever, leptospirosis, meningococcal meningitis, and hepatitis A; in contrast, female bias is predicted when strong immune responses (i) enhance pathogenesis (such as in severe dengue and Manson’s schistosomiasis) or (ii) favor a certain form of disease presentation (such as the milder, tuberculoid forms of leprosy, vs. the more severe, lepromatous forms, which are predicted to be male-biased). Finally, male-female differences are overall expected to decrease among the elderly, when estradiol levels sharply decrease in women but androgen levels remain high among men , ; more specifically, sex bias should (i) disappear for diseases in which female bias is expected to result from immune response-related pathogenesis, and (ii) decrease slightly to moderately for diseases in which male bias is expected to result from androgen down-regulation of immune responses.
On the other hand, the BH predicts that male-female differences should not appear in same-behavior age classes (mainly infancy) and should be overall absent for diseases whose transmission (typhoid fever, hepatitis A) or clinical progression (tuberculoid vs. lepromatous leprosy, meningococcal meningitis, severe dengue forms) do not depend upon host behavior. Male-biased patterns are expected for the leishmaniases, schistosomiasis, tuberculosis, and leptospirosis – particularly after puberty, when behavior-related exposure is usually higher among males. In addition, bias patterns are predicted to differ in populations with contrasting sociological-behavioral profiles, which can lead to different levels of gender-biased exposure. Thus, the risk of infection with vector-borne (Leishmania) and water-borne (Leptospira, Schistosoma) pathogens should be particularly male-biased in rural settings, where men engage more often in risky activities such as agriculture. Since the PH predicts similar outcomes for individuals of the same age class, regardless of sociological background, any difference in male bias between rural and urban populations would reflect, at least partially, behavioral risk factors.
Finally, the PH and the BH make contrasting predictions regarding indices of exposure to pathogens: they should be sex-unbiased under the PH, and sex-biased under the BH, for age classes whose members engage in gender-specific behavior (mainly after puberty). Such indices are provided by exposure-without-disease surveys that make use of antibody detection (via serology) and, in some instances, other measures of sensitization (e.g., delayed-type hypersensitivity skin tests) or direct evidence of infection in apparently healthy subjects (e.g., S. mansoni egg shedding in stools or nasopharyngeal Neisseria meningitidis carriage). The main predictions of each hypothesis are summarized in Table 1.
Our disease incidence data consist of sex- and age-stratified compulsory-notification records published by the Brazilian Ministry of Health (http://dtr2004.saude.gov.br/sinanweb) for the 2006–2009 period. For general comparisons across age classes, we used countrywide data and the official population estimates for each age class, sex, and year provided by the Brazilian Institute for Geography and Statistics (IBGE) (http://www.ibge.gov.br) (Dataset S1). For some diseases, we selected cases with specific characteristics so that clear-cut predictions could be derived in each instance. Thus, for leprosy we considered only new, confirmed cases of the two polar forms of the disease – lepromatous and tuberculoid leprosy. For the analyses of tuberculosis, we considered only new, confirmed cases of community-acquired pulmonary disease in HIV-negative subjects. For hepatitis A, we selected only new, confirmed acute cases. For dengue, we selected only new, confirmed cases of the severe forms of the disease (i.e., dengue hemorrhagic fever and dengue shock syndrome). This latter choice follows from two facts: first, immunity to dengue virus depends on whether the case is a primary or a secondary infection, and this information is unavailable in the datasets we used; in contrast, severe forms of dengue are epidemiologically associated with secondary infections with a heterologous viral serotype, with antibody-dependent enhancement probably playing a key role in pathogenesis , . Therefore, we were able to make clear-cut predictions for the severe forms of dengue, but not for dengue in general.
To compare sociologically distinct populations, we built rural/urban contrasts by selecting the Brazilian states with highest incidence (>200 new, confirmed, autochthonous cases notified in 2006) of cutaneous and visceral leishmaniasis, schistosomiasis, and leptospirosis, and used the detailed demographic data (age- and sex-stratified rural and urban population by municipality) available from the population count carried out in 2006–2007 by the IBGE. For cutaneous leishmaniasis, we separately analyzed data from the geographic areas where either Leishmania guyanensis or Le. braziliensis are the predominant etiological agents; see Dataset S2 for details on geographic coverage.
Our meta-analyses of published exposure-prevalence surveys are based on non-systematic article searches (Text S1) in PubMed, ISI Web of Knowledge, Scopus, and Google Scholar; query terms included the disease/pathogen name (e.g., “leishman*”) and different combinations of keywords – mainly “prevalence”, “seroprevalence”, “serology”, and “population-based”. No time limits were set. To keep results maximally comparable with our incidence data, we prioritized Brazilian studies, but surveys carried out elsewhere were also considered. We screened titles and abstracts to select population-based exposure-prevalence surveys in which the results for males and females were reported separately (the only inclusion/exclusion criterion), and noted the numbers of tested and positive subjects, the outcome measure, the diagnostic methods used, the study setting, and the age range of the subjects. Considering Brazilian official incidence records and our complementary appraisal of the literature on exposure-prevalence, our analyses made use of data from about 497,000 individual disease cases or diagnostic test results. We note that, in a few studies, individual subjects were tested with more than a single procedure, and emphasize that, even if document searches were non-systematic, none of the retrieved studies was excluded because it did or did not reveal sex biases – only those not reporting sex-specific results were.
To assess male-female differences in disease incidence, we estimated male:female incidence rate ratios (IRRs) and their 95% confidence intervals (CIs); for each disease, year, and age class, the ♂:♀ IRR is given by IRR = (C♂/N♂)/(C♀/N♀), where C is the number of disease cases notified to the Brazilian Ministry of Health and N is the population estimate provided by the IBGE. The variance of the logeIRR was estimated as Var(logeIRR) = 1/C♂ +1/C♀ − 1/N♂ − 1/N♀. IRRs for each disease, year, and age class were then combined through a meta-analytic approach using DerSimonian-Laird random-effects models, with inverse-variance weighting and α set to 0.05 , . In a few cases, we used cumulative incidence (total number of cases during a given time-period divided by the population at the start of that period) to derive male:female IRRs (see also Text S2).
For the assessment of exposure-prevalence in apparently healthy subjects, we used the data from each of the published reports we retrieved to estimate the male:female odds ratio (OR) and its 95%CI; the ♂:♀ OR is given by OR = [C♂/(N♂−C♂)]/[C♀/(N♀−C♀)], with Var(logeOR) = 1/C♂ +1/(N♂−C♂) +1/C♀ +1/(N♀−C♀); here, C is the number of subjects classified as infected/positive and N is the number of subjects tested. These study-specific ORs were also summarized using DerSimonian-Laird random-effects models with inverse-variance weighting and α = 0.05 , . The analyses described thus far were conducted using Review Manager 5.1 (The Cochrane Collaboration). We intentionally avoided formal null hypothesis significance testing ; IRR/OR 95%CIs including 1 indicate that male-female differences are not statistically significant at the 5% level. To compare simple proportions, we estimated their 95%CI limits using the Agresti-Coull method . Sensitivity analyses examining the consistency of results across years and their robustness to incomplete reporting and the choice of analytical procedures are provided in Text S2.
Crude incidence data (Figure 1) revealed patent post-pubertal male bias in cutaneous leishmaniasis, tuberculosis, lepromatous leprosy, and leptospirosis. More subtle differences appeared also in visceral leishmaniasis, schistosomiasis, and meningococcal meningitis, whereas no clear sex-specific pattern seemed to emerge from hepatitis A and typhoid fever data. Incidence appeared to be slightly female-biased after puberty for tuberculoid leprosy, and overall for severe dengue fever. Figure 1 also suggests that male bias is larger among infants than across childhood for cutaneous and visceral leishmaniasis, schistosomiasis, meningococcal meningitis, and leptospirosis. Incidence patterns were remarkably consistent across years except for typhoid fever, the least common among the diseases we studied; data on laboratory confirmation of hepatitis A cases were unavailable for 2006, which inflates incidence in both sexes. While overall suggestive of support for the PH, these data do not allow for a detailed quantitative appraisal of sex effect-sizes, which we therefore estimated using random-effects models; below we present detailed results by disease.
Figure 1. Infectious disease incidence in Brazil: sex- and age class-stratified incidence profiles (cases/100,000 population).
Diseases: American cutaneous (CL) and visceral leishmaniasis (VL); schistosomiasis (SCH); pulmonary tuberculosis (TB); lepromatous leprosy (LL); tuberculoid leprosy (TL); typhoid fever (TF); leptospirosis (LE); meningococcal meningitis (MM); hepatitis A (HA); and severe dengue fever (SDF). Incidence (2006–2009) was computed from Brazilian compulsory-notification records and official demographic data for males (M, blue lines) and females (F, orange-red lines). Age classes (in years) are given on the x axes. Insets present overall annual incidence for 2006–2009 (blue, males; orange, females). See main text and Dataset S1 for details.doi:10.1371/journal.pone.0062390.g001
Male infants are twice as likely to develop clinical cutaneous leishmaniasis as females (Figure 2, Table 2); this bias was observed in all years from 2006 to 2010 (Text S2), and was present in both rural and urban settings except for rural areas north of the Amazon river (Figure 3B). Infant male bias shrinks by ~48% in early and by ~45% in late childhood. With the onset of puberty, the magnitude of male bias reaches values slightly larger (~17%) than those of infancy. It then soars to a peak IRR value >3.5 in the reproductive-age population, and declines back to puberty/infancy values among the elderly (Figure 2, Table 2). Male:female IRR values are higher in urban than in rural settings in late childhood, puberty, and adulthood. This was consistently observed in areas where Le. guyanensis is the dominant agent of cutaneous leishmaniasis and in areas where Le. braziliensis predominates, but 95%CIs were much wider for Le. guyanensis due to the smaller number of records (Figure 3A, 3B). Published prevalence survey data suggest that exposure to the parasites causing cutaneous leishmaniasis is sex-unbiased (Figure 4). Surveys based on both leishmanin skin tests and antibodies yield consistent results: in Brazil (random-effects OR 1.18, 95%CI 0.61–2.28; Figure 4) – and elsewhere (e.g., ref. : OR 1.20, 95%CI 0.94–1.53), male and female subjects are at similar risk of exposure (overall random-effects OR 0.99, 95%CI 0.72–1.36, N = 3566 tests).
Figure 2. Infectious disease incidence in Brazil: male:female incidence rate ratios (IRRs) and 95% confidence intervals (CIs) computed from compulsory-notification records and official demographic data.
Diseases: American cutaneous (CL) and visceral leishmaniasis (VL); schistosomiasis (SCH); pulmonary tuberculosis (TB); lepromatous leprosy (LL); tuberculoid leprosy (TL); typhoid fever (TF); leptospirosis (LE); meningococcal meningitis (MM); hepatitis A (HA); and severe dengue fever (SDF). Circles are random-effects point estimates computed from Brazilian compulsory-notification annual incidence records (2006–2009). IRR >1 indicates male-biased incidence; the vertical line at IRR = 1 indicates no sex bias. When CIs include 1, sex bias is not statistically significant at the 5% level. Age classes are given on the y axes; a few IRRs could not be estimated due to small numbers of incident cases. The last Panel (labeled ‘Infants’) compares cumulative incidence (2006–2009) among infants (<1 year old); despite the likely absence of sex-related behavior/exposure differences, significant male bias is seen in several diseases. See main text and Dataset S1 for details.doi:10.1371/journal.pone.0062390.g002
Figure 3. Infectious disease incidence in Brazil: sociological contrasts.
Panels A-E: rural and urban male:female incidence rate ratios (IRRs) and 95% confidence intervals (CIs) for American cutaneous leishmaniasis in Brazilian sub-regions where Leishmania braziliensis (Panel A) or Le. guyanensis (Panel B) are the primary etiological agent; American visceral leishmaniasis (Panel C); schistosomiasis (Panel D); and leptospirosis (Panel E). Each age class (y axes) is represented by a gray or white band, with rural and urban estimates given as the upper and lower value within each band (as illustrated for infants under 1 year of age in Panel A); IRR >1 indicates male-biased incidence; the vertical line at IRR = 1 indicates no sex bias; when CIs include 1, sex bias is not statistically significant at the 5% level. Panel F: percentage of males (with 95%CIs) among 9498 incident leptospirosis cases (Brazil, 2006–2010); for each age class (grey/white band), the overall value is followed by infection site-specific estimates for the home (all age classes) and work environments (age classes >10); the vertical line at 50% indicates even demographic sex ratios. IRRs and CIs were computed from compulsory-notification records and official demographic data. See main text and Dataset S2 for details.doi:10.1371/journal.pone.0062390.g003
Figure 4. Sex bias in exposure to human pathogens: published exposure-without-disease surveys.
The data reveal no sex bias for Leishmania spp. (CL, cutaneous forms; VL, visceral forms); Mycobacterium leprae (LEP); Salmonella enterica serovar Typhi (TF); Hepatitis A virus (HA); Dengue virus (DEN); or Mycobacterium tuberculosis (TB). Exposure to Schistosoma mansoni (SCH 1, as determined by the Kato-Katz technique; SCH 2, as determined by immunological tests); and Leptospira interrogans (LE) is male-biased. Neisseria meningitidis (MM) is more often carried by adult men, likely because of male-biased risk factors such as smoking; MM 2 is a subset analysis of surveys involving children or high-school students, which reveals no sex bias; furthermore, many papers reporting a “non-significant” gender difference do not present the actual figures. Estimates (x axis) are random-effects odds ratios with 95% confidence intervals; when these include 1, sex bias is not statistically significant at the 5% level. The numbers of individual tests and published studies (in parentheses) analyzed are also given to the right of each estimate.doi:10.1371/journal.pone.0062390.g004
Table 2. Sex bias in infectious disease epidemiology: age-stratified male:female incidence rate ratios and 95% confidence intervals computed from Brazilian compulsory-notification records and official demographic data (see Dataset S1).doi:10.1371/journal.pone.0062390.t002
Visceral leishmaniasis incidence is slightly but significantly male-biased during the first year of life, with a 15% higher risk among males; subsequently, risk becomes sex-unbiased in early and late childhood (Table 2). Male adolescents are at a 63% higher risk than females of developing visceral leishmaniasis. Across the reproductive age class, men are ~3 times as likely as women to be diagnosed with the disease (Table 2). “Minipuberty” effects are observed only in the urban population; this result is not readily explicable, but the otherwise consistent results suggest that it might have arisen from noise in the 2006 records for this age class. IRRs are similar in rural and urban settings for the rest of age classes (Figure 3C). Exposure indices (serology and delayed-type hypersensitivity tests) again showed that contact with Le. infantum is sex-unbiased in Brazil (random-effects OR 0.96, 95%CI 0.79–1.15; data from 5866 individual tests; Figure 4) (– plus two reports ,  presenting the same data) and elsewhere (e.g., –); the overall random-effects OR for these studies is 0.95 (95%CI 0.83–1.09; N = 14,971 tests).
Schistosomiasis risk is overall male-biased during infancy (Figure 2, Table 2) yet the magnitude of the effect is small and it was not detected neither in the 2006 data (Text S2) nor when separately analyzing urban and rural populations from high-incidence states (Figure 3D). Male-female differences wane in early childhood and reappear among 5-9-year-olds, albeit the effect is again relatively small (a 23% increase in risk for males). Afterwards, male:female IRRs stabilize at ~1.5, with 95%CI upper limits always <1.77 (Table 2). These patterns were fairly stable across years (Text S2), and reappeared in rural/urban comparisons, albeit male bias was larger in urban than in rural settings among adolescents and reproductive-age adults (Figure 3D). Serological evidence from apparently healthy Brazilian subjects shows that exposure to Schistosoma mansoni is male-biased among children ,  and in the general population , with OR point estimates consistently above 2. The overall OR including all Latin American studies we retrieved (Brazil and Puerto Rico –) was 1.90 (95%CI 1.38–2.64; N = 5395 subjects; Figure 4). Brazilian prevalence surveys using the Kato-Katz method to detect S. mansoni eggs in stool samples from apparently healthy subjects – revealed male bias values similar to those seen in serological surveys; the random-effects OR was 2.12 (95%CI 1.42–3.18; N = 6891 subjects; Figure 4).
During infancy, we observed an overall non-significant male bias (Table 2, Figure 2). The small number of cases among infants (122 notifications in 2006–2009) and the relatively large between-year variation (Text S2) need however be considered; for instance, the random-effects IRR reached significance (1.47, 95%CI 1.12–1.94) when the 90 cases reported in 2010 (retrieved November 7th 2011) were included in the analysis, and male bias is even slightly larger if the three cases reported in 2006 are disregarded (IRR 1.50, 95%CI 1.13–1.98). These results suggest that there is a relatively small “minipuberty” effect in tuberculosis. Male bias shrinks during childhood and then rises progressively up to a peak IRR value of 2.98 in the population over 60 years of age (Table 2, Figure 2). Our meta-analysis of published Brazilian infection-without-disease tuberculosis surveys, including tuberculin skin tests (TST) – and, in one case, TST plus an IFN-γ release assay , revealed a marginally non-significant male bias in exposure to Mycobacterium tuberculosis (random-effects OR 1.16, 95%CI 1.00–1.36; N = 6725 subjects; Figure 4).
Both adolescent and reproductive-age women are consistently more likely than men to be diagnosed with the milder, tuberculoid form of the disease; for the more severe lepromatous forms, the opposite pattern emerged, with males being at much higher risk, particularly after puberty and throughout the reproductive age class (Figure 2, Table 2). More specifically, average tuberculoid leprosy risk is even for both sexes in early and late childhood, although the precision of the estimate is low for younger children (Figure 2, Table 2). With the onset of puberty, girls become about 1.2 times as likely as boys to get a tuberculoid leprosy diagnosis; this female-biased pattern persists in the 20–59 age class, but disappears among the elderly (Table 2). During early childhood, the few reported cases of lepromatous leprosy (22 between 2006 and 2010) are not sex-biased (Table 2), paralleling the absence of any sex effect in tuberculoid disease. In late childhood, lepromatous leprosy risk shows a slight trend towards male bias; although 95%CIs included 1 in all year-specific analyses (Text S2), the random-effects mean IRR was estimated at 1.50 (95%CI 1.09–2.05) (Table 2). Male-biased lepromatous leprosy risk becomes evident during puberty and reproductive life, with males being about 2–3 times as likely to be diagnosed with this condition as females. In contrast with tuberculoid leprosy results, the risk of lepromatous forms remains markedly male-biased in the older population (Figure 2, Table 2). Exposure surveys conducted in Brazil – show that, even though results varied across studies, anti-Mycobacterium leprae seropositivity is overall sex-unbiased (random-effects OR 1.19, 95%CI 0.77–1.86; N = 12,444 subjects; Figure 4).
We did not observe any consistent male-female difference in our typhoid fever data (Figure 2, Table 2), but rather a somewhat erratic incidence pattern in which uncertainty was exacerbated by the relatively small number of records. Thus, the apparently male-biased risk in the 10–19 years old class (Table 2) emerges from inconsistent (and noticeably imprecise) year-specific estimates (Text S2). Out of 25 year- and age-specific comparisons, only in five was there significant male bias, whereas in 20 cases 95%CIs encompassed 1 (Text S2); this suggests that typhoid fever risk is overall sex-unbiased. We did not find any Brazilian study, but antibody surveys among apparently healthy adults conducted in Nigeria  and Vietnam  reveal no sex bias in exposure (random-effects OR 1.22, 95%CI 0.62–2.39; N = 3394 subjects; Figure 4).
“Minipuberty” effects are overtly evident in leptospirosis, with similar male bias estimates in urban (IRR 2.00, 95%CI 1.20–3.34) and rural settings (IRR 2.04, 95%CI 1.67–2.50); subsequently, male-female differences disappear throughout childhood in both populations (Figure 3E). Male-biased risk soars with the onset of puberty to a peak IRR of 4.95 (95%CI 2.47–9.95) for rural youngsters aged 10–19, with a somewhat lower IRR for the same age class in urban environments. Reproductive-age rural men are at higher risk than women (IRR 4.22, 95%CI 3.09–5.76), with a much larger sex bias than that observed in urban settings (IRR 1.83, 95%CI 1.52–2.20). IRRs become again comparable in the older population (Figure 3E). For the economically active age classes (10–59 years old), the overall urban IRR is 1.98 (95%CI 1.75–2.25), vs. 4.33 (95%CI 3.49–5.38) in rural environments. To further explore the relative contributions of behavior and physiology to leptospirosis risk, we examined the proportion of males among 9498 incident cases (2006–2010 period) for which the most likely site of infection was traced to the working environment (91.7% males, 95%CI 91.2–92.1%; N = 3291 cases) or the home environment (69.6% males, 95%CI 69.0–70.2%; N = 6207 cases) (Figure 3F). Serological evidence of exposure to Leptospira spp. in Brazil – shows a moderately higher risk among males: a random-effects meta-analysis of four surveys (N = 3996 subjects) yielded an OR of 1.41 (95%CI 1.18–1.69) (Figure 4).
Our data show a relatively small but significant male bias in infancy and among reproductive-age adults (risk ~26% and ~39% higher for males, respectively), whereas incidence is unbiased in late childhood and in people over 60 (Table 2, Figure 2). Meta-analyses of 17 published Neisseria meningitidis carrier surveys – showed a significant male bias in the general population (OR 1.52, 95%CI 1.28–1.81; N = 31,351 subjects) but a non-significant bias before adulthood (random-effects OR 1.16, 95%CI 0.97–1.38; N = 18,315 subjects) (Figure 4).
The data reveal a small, non-significant male bias among infants, which reaches significance in early childhood; hepatitis A risk then becomes unbiased or slightly female-biased among 5–9 years old children (Table 2, Figure 2). Pubertal boys are at higher risk (~25%) than girls; male bias peaks among reproductive-age adults (IRR 1.39), then wanes in older people, with all year-specific 95%CIs encompassing 1 except for 2007 (Text S2). Serological surveys conducted in Brazil – reveal sex-unbiased exposure patterns (random-effects OR 1.01, 95%CI 0.90–1.14; N = 8642 subjects; Figure 4). Except for a single Iranian study showing male-biased exposure , results from several further settings – reveal the same unbiased pattern (overall random-effects OR 1.01, 95%CI 0.90–1.13; N = 20,702 subjects).
Severe Dengue Fever
Women between 20 and 59 years old are at a significantly higher risk (about 37%) than their male counterparts of developing severe forms of dengue (Table 2, Figure 2); this was consistent for all years analyzed except for 2009, when female bias was marginally non-significant (Text S2). In contrast, the overall female-biased pattern observed among 1-4-year-olds seen in Figure 2 results from year-specific data showing no significant sex bias except for 2007, when a slight female bias was observed (Text S2). Severe dengue fever risk is sex-unbiased for the rest of age class comparisons. A random-effects meta-analysis of the results of serological surveys conducted in Brazil – reveals that exposure to dengue virus is, on average, sex-unbiased (OR 0.96, 95%CI 0.80–1.14; N = 5920 subjects; Figure 4).
Using data from nearly 0.5 million individual cases or diagnostic test results, we conducted the first quantitative, comprehensive test of the predictions of the two main hypotheses formulated to explain sex-biased infectious disease risk in humans (see also ref. ). Three major findings seem to clearly contradict key BH predictions while closely matching those of the PH. First, male-biased infectious disease risk among infants cannot be accounted for by behavioral differences. Second, as predicted by the PH, risk is female-biased in tuberculoid leprosy but male-biased in lepromatous leprosy, and female-biased in severe dengue fever. And, third, exposure-prevalence indices are, with a few exceptions, sex-unbiased. The data, however, also suggest a role for gender-related behavior in diseases in which it is expected to affect exposure (e.g., schistosomiasis, leptospirosis, the leishmaniases) or result in contrasting risk factors for disease progression (e.g., smoking or alcohol use for tuberculosis among older adults).
Our findings are remarkably robust to inter-annual variation in official rates, to underreporting, and to the modeling procedure used to derive summary effect-size measures (see Text S2). Although our age categories are based on well-established physiological criteria , , we also tested whether an alternative definition of the ‘reproductive age class’, with two groups (20–39 and 40–59) instead of one, would qualitatively change our conclusions; the only difference was that no significant sex bias was found for typhoid fever risk (details not shown). Several study limitations must however be considered when interpreting our results. First, we used notification records of uncertain quality; still, we do not think it likely that quality issues may affect male and female data differentially: since we assess a relative measure (IRR), the results will hold if both are equally poor. Second, and as in other broad-scale appraisals, we do not have data on putative confounders; our strategy of (i) stratifying by age class, which allowed us to assess sex bias in infancy and childhood, and (ii) examining both disease-incidence and exposure-prevalence data alleviates, albeit does not eliminate, the problem of confounding. Third, although we analyzed data from a fairly diverse set of pathogens, these were not selected to test our hypotheses – they were given by data availability, by their public health importance, and by our ability to specify clear-cut predictions under each hypothesis. It is conceivable that a different set of diseases might lead to different conclusions; the relatively few studies addressing male-female differences in exposure and disease do suggest, however, that our findings are not confounded by taxonomy (see Text S2 for examples). Finally, we ignored pathogen and host variability and considered key physiological parameters only coarsely. For instance, our interpretation of age-stratified results relies on typical, average sex steroid levels for each age class . More generally, while overall suggestive of support for the PH and partially at odds with the BH, our findings do not allow us to make direct causal claims. With these caveats in mind, we now proceed to discuss our results for each age class in relation to the key predictions of each major hypothesis.
Risk is male-biased during infancy for cutaneous and visceral leishmaniasis, schistosomiasis, leptospirosis, and meningococcal meningitis; in tuberculosis and hepatitis A, infant male bias is non-significant, and a similarly non-significant female bias is observed for severe dengue forms (Figure 2). The small number of incident cases precluded the assessment of typhoid fever and leprosy incidence during infancy. Since behavioral exposure to insect vectors (leishmaniases), contaminated water (schistosomiasis, leptospirosis), or meningococci-carrying respiratory droplets can be assumed equal for infants of both sexes, these results are clearly at odds with the BH.
In the case of pulmonary tuberculosis, we have seen (Results section) that male bias reaches significance when a few more cases are considered, and this could also be the case for hepatitis A (Table 2), suggesting that an adequately powered analysis is necessary to detect “minipuberty” effects in these diseases. Severe dengue fever is associated with a phenomenon known as antibody-dependent enhancement , : upon secondary infection with a heterologous dengue virus serotype, cross-reacting antibodies developed during the primary infection event cannot neutralize the newly arrived virus, resulting in enhanced viral replication within monocytes and, consequently, more severe forms of the disease , . Since infants are unlikely to get infected twice in their first year of life, both the PH and the BH predict a similar outcome (no bias) in this age class (Table 1).
What mechanisms may underlie these patterns of sex-biased infectious disease risk in infants? As mentioned in the Introduction, the PH predicts that the transient rise of sex steroid levels in the first year of life (“minipuberty” ) should yield the observed patterns. Other possibilities seem far less likely. For instance, genetic (chromosome) differences must apply roughly equally across age classes, and would not explain by themselves drastic changes in sex bias between infants and children. Differences in access to healthcare are another plausible explanation, but a study based on a country-representative sample of 110,000 Brazilian households revealed no gender-related difference during infancy and childhood . It seems conceivable, however, that the reduction of infant mortality brought about by Cesarean delivery and neonatal intensive care units disproportionally benefited male infants, and particularly those with low weight and/or premature . Since Cesarean delivery is particularly common in Brazil , we might speculate that this could have increased the pool of ‘frail’ male infants with a predisposition for infection. Yet, since the ‘male infant disadvantage’ these improvements helped reduce  is obviously not due to behavioral factors, we see this possibility as a further argument against the BH.
As expected for similar exposure and negligible sex steroid activity, bias either shrinks (cutaneous leishmaniasis, tuberculosis) or disappears (visceral leishmaniasis, schistosomiasis, leptospirosis) in children aged 1–4. Male bias is significant in meningococcal meningitis and hepatitis A, and female bias in severe dengue fever, among 1-4-year-olds. Male-biased risk emerges in late childhood for cutaneous leishmaniasis, schistosomiasis, lepromatous leprosy, leptospirosis, and, to a lesser extent, visceral leishmaniasis and meningococcal meningitis; a non-significant female bias is observed in typhoid fever, severe dengue fever, and hepatitis A, while risk is unbiased in tuberculosis and tuberculoid leprosy. Hence, the general pattern is one of even or slightly male-biased risk across childhood (Figure 2, Table 2). Since both the PH and the BH predict similar outcomes for this age group (Table 1), these results add little to our hypothesis-testing exercise. We also note that defining discrete age-classes may result in some confounding from cross-class effects in some individuals. For instance, physiological effects might be ‘carried over’ from infancy into early childhood, and puberty effects, physiological and behavioral, may already be apparent in older children.
Male-female differences are overtly seen during puberty; yet, their magnitude is nearly always smaller than across the reproductive period (Figure 2), mirroring average sex steroid levels in those age classes . This is observed in cutaneous and visceral leishmaniases, tuberculosis, lepromatous leprosy, hepatitis A, and severe dengue fever; the difference is smaller in tuberculoid leprosy and meningococcal meningitis, and absent in schistosomiasis, leptospirosis, and typhoid fever (Figure 2). Reproductive-age men endure higher disease risk than women for intracellular pathogens, including protozoa (cutaneous and visceral leishmaniases), bacteria (tuberculosis, lepromatous leprosy), and viruses (hepatitis A), as well as for schistosomiasis and extracellular bacteria (leptospirosis and meningococcal meningitis). Particularly striking are the patterns in leprosy, a polymorphic disease whose clinical progression is largely driven by host immunity , . Under the PH, different forms of leprosy are predicted to yield sharply contrasting sex bias patterns: while the milder tuberculoid form is linked to efficient Th1 responses, severe lepromatous forms are related to less protective, Th2-dominated responses (see , ). Since exposure to M. leprae is most likely sex-unbiased (cf. Figure 4), and disease progression behavior-independent, the BH predicts unbiased risk of leprosy, whereas the PH predicts tuberculoid forms to be female-biased and lepromatous forms to be male-biased (Table 1); this is precisely what we observed (Figure 2). The PH also predicts that severe dengue fever, linked to antibody-dependent enhancement , , should be female-biased in the reproductive-age population; our results also match this prediction (Figure 2). Among the elderly, male bias decreases visibly in cutaneous and visceral leishmaniasis, leptospirosis, meningococcal meningitis, and hepatitis A, whereas no clear changes are apparent for schistosomiasis, lepromatous leprosy, and typhoid fever; as predicted by the PH, female bias disappears in tuberculoid leprosy and severe dengue fever (Figure 2). Strongly male-biased tuberculosis risk is likely to stem from male-biased risk factors, such as smoking and alcohol intake .
We note that the results discussed so far not only provide partial support for the PH: they are also at odds with several key predictions of the BH, especially regarding disease progression/severity (leprosy, severe dengue) and sex-biased risk in same-behavior age classes. To provide a test of the third BH prediction, we built sociological contrasts comparing incidence rates in rural and urban populations for diseases in which differential risk-exposure is likely. The age-stratified patterns of male-female differences are comparable in both settings, although effect-sizes often differ: somewhat in contrast with our predictions (see Methods), post-pubertal male bias is larger in urban than in rural areas for cutaneous leishmaniasis and schistosomiasis; for visceral leishmaniasis, estimates are similar (Figure 3C). This suggests that urban women may be more protected from cutaneous leishmaniasis vectors (associated with forest/cropland) than rural ones, but both are equally exposed to the peridomestic vectors of visceral leishmaniasis (see also Dataset S2). Incidence profiles also suggest that cutaneous and visceral leishmaniasis transmission dynamics differ sharply (Figure 1). These data indicate that behavior, probably related to agriculture, exacerbates male-biased cutaneous leishmaniasis risk in some populations. Similarly, urban women seem more protected than rural ones from schistosomiasis (see Dataset S2), further supporting a role for behavior in modulating male-female differences. Since pathogenic (i.e., persistent) Th1 responses are overall stronger in women ,  but exposure is most likely male-biased, we predicted schistosomiasis incidence to be less male-biased than Schistosoma mansoni-exposure indices (Table 1; see also below). For leptospirosis, the larger post-pubertal male bias in rural settings is likely related to higher male exposure in the working environment (Figure 3F), again suggesting a link with agriculture.
Finally, we tested the sharply contrasting PH and BH predictions about male-female differences in exposure-prevalence. Random-effects meta-analyses of male-female differences in published exposure-prevalence surveys showed that, as predicted by the PH and contrary to BH predictions, average exposure to the pathogens causing leishmaniases, leprosy, typhoid fever, hepatitis A, dengue, and, marginally, tuberculosis is sex-unbiased (Figure 4). Male bias among N. meningitidis carriers most likely arises from male-biased risk factors and from reporting bias (see Figure 4 caption). In agreement with our previous analyses, exposure to Schistosoma and Leptospira is male-biased, again suggesting a role for gender-specific behavior. As predicted taking both behavior and physiology into account, exposure-prevalence male bias is larger (about 60–70%) than disease-incidence male bias in schistosomiasis.
We have presented a comprehensive test of the two major competing hypotheses proposed to explain male-female differences in infectious disease risk. Our findings suggest that sex-related physiology plays a major role in this phenomenon, albeit gender-specific behavior probably modulates infection risk in some instances. These results have potentially important implications from various perspectives. From an academic viewpoint, they provide novel insight into the population-level consequences of the evolutionary balance between sex and immunity to infection, suggesting that, in our species, the fitness trade-offs involved in such balance often manifest themselves as clear male-female differences in morbidity , , , , . From a more practical stance, our results warn against unthinking extrapolation of biomedical research results, including vaccine development, across ages and genders , , , , . Our findings also highlight sex as a major disease risk determinant that should not be treated (as is often the case in epidemiology) as a mere confounder; this, in turn, calls for better reporting standards in the epidemiological literature: sex-and-age-stratified results should always be provided for publication. Finally, our results suggest that, by considering immune-endocrine interactions , , –, –, insight could be gained not only into the epidemiology, but also into the pathogenesis, management, and prognosis of many infectious diseases . Although specific research is obviously required to explore this possibility, a recent E. coli O104:H4 outbreak  provides a hint of the potential predictive power of the PH. Intriguingly, adult women were more often (and more seriously) affected in this outbreak ; behavior, in the form of food preferences, was proposed as a likely explanation . Under the PH, however, this ‘atypical’ epidemiological pattern ,  is the one to be expected if immunity plays a key role in pathogenesis. The subsequent demonstration that IgG antibody depletion has therapeutic value in the most severe forms of the disease  suggests that this prediction may indeed prove accurate.
In conclusion, and from a broader perspective, our analyses show that the development of a truly coherent view of infectious disease epidemiology demands an inter-disciplinary approach drawing from the diverse fields of microbiology, immunology, endocrinology, evolutionary biology, and sociology. Further research is needed to see to what extent our conclusions hold and can be generalized, but a growing body of evidence suggests that physiological sex differences likely underlie gender inequality in many infectious diseases.
Sensitivity analyses and examples involving pathogens not included in our study. Includes Tables S1–S11 (sensitivity to analytical procedure) and Figure S1 (year-specific results for each disease, 2006–2010).
Raw data used for the male:female incidence rate ratio calculations in sociological (rural/urban) contrasts (Figure 3 of the main text).
We thank RS Rocha, the Editor, and the anonymous referees for their useful comments on previous versions of this manuscript. This work is dedicated to the memory of Ana Patrícia Paz Veras.
Contributed to the final version of the manuscript: FG-S FA-F. Conceived and designed the experiments: FA-F FG-S. Performed the experiments: FG-S FA-F. Analyzed the data: FA-F FG-S. Contributed reagents/materials/analysis tools: FA-F FG-S. Wrote the paper: FA-F.
- 1. Wizeman TM, Pardue M-L, editors (2001) Exploring the Biological Contributions to Human Health: Does Sex Matter? Washington, DC: National Academy Press. 288 p.
- 2. World Health Organization (2007) Addressing Sex and Gender in Epidemic-Prone Infectious Diseases. Geneva: WHO. iv +40 p.
- 3. Brabin L, Brabin BJ (1992) Parasitic infections in women and their consequences. Adv Parasitol 31: 1–81. doi: 10.1016/s0065-308x(08)60020-2
- 4. Tolhurst R, Koning K, Price J, Kemp J, Theobald S, et al. (2002) The challenge of infectious disease: time to take gender into account. J Health Manage 4: 135–151. doi: 10.1177/097206340200400204
- 5. Sen G, Östlin P, George A (Women and Gender Equity Knowledge Network) (2007) Unequal, Unfair, Ineffective and Inefficient – Gender Inequity in Health: Why it Exists and How We Can Change it. Final Report to the WHO Commission on Social Determinants of Health. WHO website. Available: http://www.who.int/social_determinants/resources/csdh_media/wgekn_final_report_07.pdf. Accessed 2011 Jun 19.
- 6. Zuk M, McKean KA (1996) Sex differences in parasite infections: patterns and processes. Int J Parasitol 26: 1009–1024. doi: 10.1016/s0020-7519(96)80001-4
- 7. Sheridan LAD, Poulin R, Ward DF, Zuk M (2000) Sex differences in parasitic infections among arthropod hosts: is there a male bias? Oikos 88: 327–334. doi: 10.1034/j.1600-0706.2000.880211.x
- 8. Moore SL, Wilson KW (2002) Parasites as a viability cost of sexual selection in natural populations of mammals. Science 297: 2015–2017. doi: 10.1126/science.1074196
- 9. Klein SL (2004) Hormonal and immunological mechanisms mediating sex differences in parasite infection. Parasite Immunol 26: 247–264. doi: 10.1111/j.0141-9838.2004.00710.x
- 10. Roberts ML, Buchanan KL, Evans MR (2004) Testing the immunocompetence handicap hypothesis: a review of the evidence. Anim Behav 68: 227–239. doi: 10.1016/j.anbehav.2004.05.001
- 11. Roberts CW, Walker W, Alexander J (2001) Sex-associated hormones and immunity to protozoan parasites. Clin Microbiol Rev 14: 476–488. doi: 10.1128/cmr.14.3.476-488.2001
- 12. Bouman A, Heineman MJ, Faas MM (2005) Sex hormones and the immune response in humans. Human Reprod Update 11: 411–423. doi: 10.1093/humupd/dmi008
- 13. Klein SL, Jedlicka A, Pekosz A (2010) The Xs and Y of immune responses to viral vaccines. Lancet Infect Dis 10: 338–349. doi: 10.1016/s1473-3099(10)70049-9
- 14. Muehlenbein MP, Bribiescas RG (2005) Testosterone-mediated immune functions and male life histories. Am J Hum Biol 17: 527–558. doi: 10.1002/ajhb.20419
- 15. Fish EN (2008) The X-files in immunity: sex-based differences predispose immune responses. Nature Rev Immunol 8: 737–744. doi: 10.1038/nri2394
- 16. Klein SL, Roberts CW, editors (2010) Sex Hormones and Immunity to Infection. Berlin: Springer Verlag. x+319 p.
- 17. Ober C, Loisel DA, Gilad Y (2008) Sex-specific genetic architecture of human disease. Nature Rev Genet 9: 911–922. doi: 10.1038/nrg2415
- 18. Restif O, Amos W (2010) The evolution of sex-specific immune defences. Proc R Soc B 277: 2247–2255. doi: 10.1098/rspb.2010.0188
- 19. Brei B, Fish D (2003) Comment on “Parasites as a viability cost of sexual selection in natural populations of mammals”. Science 300: 55a. doi: 10.1126/science.1079746
- 20. Krieger N (2003) Genders, sexes, and health: what are the connections – and why does it matter? Int J Epidemiol 32: 652–657. doi: 10.1093/ije/dyg156
- 21. World Health Organization (2009) Women and Health: Today’s Evidence Tomorrow’s Agenda. Geneva: WHO. ix +91 p.
- 22. Owens IPF (2002) Sex differences in mortality rates. Science 297: 2008–2009. doi: 10.1126/science.1076813
- 23. Garenne M (1994) Sex differences in measles mortality: a world review. Int J Epidemiol 23: 632–642. doi: 10.1093/ije/23.3.632
- 24. Crabtree TD, Pelletier SJ, Gleason TG, Pruett TL, Sawyer RG (1999) Gender-dependent differences in outcome after the treatment of infection in hospitalized patients. JAMA 282: 2143–2148. doi: 10.1001/jama.282.22.2143
- 25. Acuna-Soto R, Maguire JH, Wirth DF (2000) Gender distribution in asymptomatic and invasive amebiasis. Am J Gastroenterol 95: 1277–1283.
- 26. Varkevisser CM, Lever P, Alubo O, Burathoki K, Idawani C, et al. (2009) Gender and leprosy: cases studies in Indonesia, Nigeria, Nepal, and Brazil. Lepr Rev 80: 65–76.
- 27. Neyrolles O, Quintana-Murci L (2009) Sexual inequality in tuberculosis. PLoS Med 6: e1000199. doi: 10.1371/journal.pmed.1000199
- 28. Labrie F, Cusan L, Gomez JL, Martel C, Bérubé R, et al. (2009) Comparable amounts of sex steroids are made outside the gonads in men and women: strong lesson for hormone therapy of prostate and breast cancer. J Steroid Biochem Mol Biol 113: 52–56. doi: 10.1016/j.jsbmb.2008.11.004
- 29. Guzmán MG, Kourí G (2002) Dengue: and update. Lancet Infect Dis 2: 33–42. doi: 10.1016/s1473-3099(01)00171-2
- 30. Halstead SB, Mahalingam S, Marovich MA, Ubol S, Mosser DM (2010) Intrinsic antibody-dependent enhancement of microbial infection in macrophages: disease regulation by immune complexes. Lancet Infect Dis 10: 712–722. doi: 10.1016/s1473-3099(10)70166-3
- 31. DerSimonian R, Laird N (1986) Meta-analysis in clinical trials. Control Clin Trials 7: 177–188. doi: 10.1016/0197-2456(86)90046-2
- 32. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR (2010) A basic introduction to fixed-effect and random-effects models for meta-analysis. Res Syn Meth 1: 97–111. doi: 10.1002/jrsm.12
- 33. Cohen J (1994) The Earth is round (p<.05). Am Psychol 49: 997–1003. doi: 10.1037//0003-066x.49.12.997
- 34. Brown LD, Cai TT, DasGupta A (2001) Interval estimation for a binomial proportion. Stat Sci 16: 101–117. doi: 10.1214/ss/1009213286
- 35. Naiff RD, Barrett TV, Arias JR, Naiff MF (1988) Encuesta epidemiológica de histoplasmosis, paracoccidioidomicosis y leishmaniasis mediante pruebas cutáneas. Bol Of Sanit Panam 140: 35–50.
- 36. Gomes AC, Yamamoto YI, Capinzaiki AN, Amaral NMM, Guimarães AJG (1992) Aspectos ecológicos da Leishmaniose Tegumentar Americana 9. Prevalência/incidência da infecção humana nos municípios de Pedro de Toledo e Miracatu, São Paulo, Brasil. Rev Inst Med Trop São Paulo 34: 149–158. doi: 10.1590/s0036-46651992000200011
- 37. Coimbra CEA Jr, Santos RV, do Valle ACF (1996) Cutaneous leishmaniasis in Tupí-Mondé Amerindians from the Brazilian Amazonia. Acta Trop 61: 201–211. doi: 10.1016/0001-706x(95)00145-5
- 38. Brandão-Filho SP, Campbell-Lendrum D, Brito MEF, Shaw JJ, Davies CR (1999) Epidemiological surveys confirm an increasing burden of cutaneous leishmaniasis in north-east Brazil. Trans R Soc Trop Med Hyg 93: 488–494. doi: 10.1016/s0035-9203(99)90346-2
- 39. Oliveira CCG, Lacerda HG, Martins DRM, Barbosa JDA, Monteiro GR, et al. (2004) Changing epidemiology of American cutaneous leishmaniasis (ACL) in Brazil: a disease of the urban-rural interface. Acta Trop 90: 155–162. doi: 10.1016/j.actatropica.2003.11.011
- 40. Muñoz G, Davies CR (2006) Leishmania panamensis transmission in the domestic environment: the results of a prospective epidemiological survey in Santander, Colombia. Biomédica 26(Suppl (1)) 131–144.
- 41. Nascimento MD, Souza EC, Silva LM, Leal PC, Cantanhede KL, et al. (2005) Prevalência de infecção por Leishmania chagasi utilizando os métodos de ELISA (rK39 e CRUDE) e intradermorreação de Montenegro em área endêmica do Maranhão, Brasil. Cad Saúde Pública 21: 1801–1807. doi: 10.1590/s0102-311x2005000600028
- 42. Moreno EC, Melo MN, Genaro O, Lambertucci JR, Serufo JC, et al. (2005) Risk factors for Leishmania chagasi infection in an urban area of Minas Gerais State. Rev Soc Bras Med Trop 38: 456–463. doi: 10.1590/s0037-86822005000600002
- 43. Gouvêa MV, Werneck GL, Costa CH, de Amorim Carvalho FA (2007) Factors associated to Montenegro skin test positivity in Teresina, Brazil. Acta Trop 104: 99–107. doi: 10.1016/j.actatropica.2007.07.010
- 44. Barão SC, de Fonseca Camargo-Neves VL, Resende MR, da Silva LJ (2007) Human asymptomatic infection in visceral leishmaniasis: a seroprevalence study in an urban area of low endemicity. Preliminary results. Am J Trop Med Hyg 77: 1051–1053.
- 45. Oliveira AL, Paniago AM, Sanches MA, Dorval ME, Oshiro ET, et al. (2008) Asymptomatic infection in family contacts of patients with human visceral leishmaniasis in Três Lagoas, Mato Grosso do Sul State, Brazil. Cad Saúde Pública 24: 2827–2833. doi: 10.1590/s0102-311x2008001200011
- 46. de Gouvêa Viana L, de Assis TS, Orsini M, da Silva AR, de Souza GF, et al. (2008) Combined diagnostic methods identify a remarkable proportion of asymptomatic Leishmania (Leishmania) chagasi carriers who present modulated cytokine profiles. Trans R Soc Trop Med Hyg 102: 548–555. doi: 10.1016/j.trstmh.2008.02.007
- 47. Falqueto A, Ferreira AL, dos Santos CB, Porrozzi R, da Costa MV, et al. (2009) Cross-sectional and longitudinal epidemiologic surveys of human and canine Leishmania infantum visceral infections in an endemic rural area of southeast Brazil (Pancas, Espírito Santo). Am J Trop Med Hyg 80: 559–565.
- 48. Crescente JA, Silveira FT, Lainson R, Gomes CM, Laurenti MD, et al. (2009) A cross-sectional study on the clinical and immunological spectrum of human Leishmania (L.) infantum chagasi infection in the Brazilian Amazon region. Trans R Soc Trop Med Hyg 103: 1250–1256. doi: 10.1016/j.trstmh.2009.06.010
- 49. Silveira FT, Lainson R, Pereira EA, de Souza AA, Campos MB, et al. (2009) A longitudinal study on the transmission dynamics of human Leishmania (Leishmania) infantum chagasi infection in Amazonian Brazil, with special reference to its prevalence and incidence. Parasitol Res 104: 559–567. doi: 10.1007/s00436-008-1230-y
- 50. Fakhar M, Motazedian MH, Hatam GR, Asgari Q, Kalantari M, et al. (2008) Asymptomatic human carriers of Leishmania infantum: possible reservoirs for Mediterranean visceral leishmaniasis in southern Iran. Ann Trop Med Parasitol 102: 577–583. doi: 10.1179/136485908x337526
- 51. Faye B, Bucheton B, Bañuls AL, Senghor MW, Niang AA, et al. (2011) Seroprevalence of Leishmania infantum in a rural area of Senegal: analysis of risk factors involved in transmission to humans. Trans R Soc Trop Med Hyg 105: 333–340. doi: 10.1016/j.trstmh.2011.02.009
- 52. Scarlata F, Vitale F, Saporito L, Reale S, Vecchi VL, et al. (2008) Asymptomatic Leishmania infantum/chagasi infection in blood donors of western Sicily. Trans R Soc Trop Med Hyg 102: 394–396. doi: 10.1016/j.trstmh.2008.01.011
- 53. Garrote JI, Gutiérrez MP, Izquierdo RL, Dueñas MA, Zarzosa P, et al. (2004) Seroepidemiologic study of Leishmania infantum infection in Castilla-León, Spain. Am J Trop Med Hyg 71: 403–406.
- 54. Agrela IF, Sánchez E (2009) Prevalencia de la infección por Leishmania spp. en un área periurbana de Altagracia de Orituco, estado Guárico, Venezuela. Bol Mal Salud Amb XLIX: 107–116.
- 55. Diza E, Kansouzidou A, Gerou S, Vezyri E, Metallidis S, et al. (2008) Leishmaniases in Northern Greece: seroprevalence of the infection and incidence of the disease during the period 2001–2006. Eur J Clin Microbiol Infect Dis 27: 997–1003. doi: 10.1007/s10096-008-0538-y
- 56. Camargo-Neves VL, Kanamura HY, Vellosa SA, Gargioni C, Dias LC (1998) IgM antibodies to Schistosoma mansoni gut–associated antigens for the study of schistosomiasis transmission in Ribeirão Pires, São Paulo. Mem Inst Oswaldo Cruz (Suppl. 1): 273–278.
- 57. Lima VL, Guercio VM, Rangel O, Kanamura HY, Dias LC (1998) Immunofluorescence test on Schistosoma mansoni worm paraffin sections (IgM-IFT) for the study of schistosomiasis transmission in Campinas, São Paulo, Brazil. Mem Inst Oswaldo Cruz (Suppl. 1): 283–287.
- 58. Zacharias F, de Carvalho ME, Gargioni C, Teles HM, Ferreira CS, et al.. (2002) Schistosomiasis mansoni in Bananal (State of São Paulo, Brazil). III. Seroepidemiological studies in the Palha District. Mem Inst Oswaldo Cruz (Suppl. 1): 19–22.
- 59. Hillyer GV, Tsang VC, Vivas-Gonzalez BE, Noh J, Ahn LH, et al. (1999) Age-specific decrease in seroprevalence of schistosomiasis in Puerto Rico. Am J Trop Med Hyg 60: 313–318.
- 60. Hillyer GV, Soler de Galanes M (1999) Seroepidemiology of schistosomiasis in Puerto Rico: evidence for vanishing endemicity. Am J Trop Med Hyg 60: 827–830.
- 61. Rodrigues RN, Murta C, Teixeira Júnior MA, Cury GC, Rocha MO (1995) Clinical-epidemiologic study of schistosomiasis mansoni in Ponte do Pasmado, a village in the municipality of Itinga, state of Minas Gerais, Brazil, 1992. Rev Inst Med Trop São Paulo 37: 81–85. doi: 10.1590/s0036-46651995000100013
- 62. Guimarães ICS, Tavares-Neto J (2006) Transmissão urbana de esquistossomose em crianças de um bairro de Salvador, Bahia. Rev Soc Bras Med Trop 39: 451–455. doi: 10.1590/s0037-86822006000500006
- 63. Firmo JO, Lima Costa MF, Guerra HL, Rocha RS (1996) Urban schistosomiasis: morbidity, sociodemographic characteristics and water contact patterns predictive of infection. Int J Epidemiol 25: 1292–1300. doi: 10.1093/ije/25.6.1292
- 64. Coura-Filho P, Farah MW, Rezende DF, Lamartine SS, Carvalho OS, et al. (1995) Determinantes ambientais e sociais da esquistossomose mansoni em Ravena, Minas Gerais, Brasil. Cad Saúde Pública 11: 254–265. doi: 10.1590/s0102-311x1995000200009
- 65. Lima e Costa MF, Rocha RS, Leite ML, Carneiro RG, Colley D, et al. (1991) A multivariate analysis of socio-demographic factors, water contact patterns and Schistosoma mansoni infection in an endemic area in Brazil. Rev Inst Med Trop São Paulo 33: 58–63. doi: 10.1590/s0036-46651991000100011
- 66. Barreto ML (1991) Geographical and socioeconomic factors relating to the distribution of Schistosoma mansoni infection in an urban area of north-east Brazil. Bull World Health Organ 69: 93–102.
- 67. da Silva AA, Cutrim RN, de Britto e Alves MT, Coimbra LC, Tonial SR, et al. (1997) Water-contact patterns and risk factors for Schistosoma mansoni infection in a rural village of northeast Brazil. Rev Inst Med Trop São Paulo 9: 91–96. doi: 10.1590/s0036-46651997000200005
- 68. Coelho KR (2010) Infecção pelo Schistosoma mansoni em Estudantes como Indicador da Infecção na População em Área Rural de Alta Prevalência. MSc Thesis. Belo Horizonte: Universidade Federal de Minas Gerais. 97 p.
- 69. Roth VR, Garrett DO, Laserson KF, Starling CE, Kritski AL, et al. (2005) A multicenter evaluation of tuberculin skin test positivity and conversion among health care workers in Brazilian hospitals. Int J Tuberc Lung Dis 9: 1335–1342.
- 70. Teixeira EG, Menzies D, Comstock GW, Cunha AJ, Kritski AL, et al. (2005) Latent tuberculosis infection among undergraduate medical students in Rio de Janeiro State, Brazil. Int J Tuberc Lung Dis 9: 841–847.
- 71. Rabahi MF, Junqueira-Kipnis AP, Dos Reis MC, Oelemann W, Conde MB (2007) Humoral response to HspX and GlcB to previous and recent infection by Mycobacterium tuberculosis. BMC Infect Dis 7: 148. doi: 10.1186/1471-2334-7-148
- 72. Cailleaux-Cezar M, de A Melo D, Xavier GM, de Salles CL, de Mello FC, et al. (2009) Tuberculosis incidence among contacts of active pulmonary tuberculosis. Int J Tuberc Lung Dis 13: 190–195.
- 73. Machado A Jr, Emodi K, Takenami I, Finkmoore BC, Barbosa T, et al. (2009) Analysis of discordance between the tuberculin skin test and the interferon-gamma release assay. Int J Tuberc Lung Dis 13: 446–453.
- 74. Brasil MTLRF, Oliveira LR, Melo CS, Nakamura PM, Rimoli NS, et al. (1998) Aplicação do teste Elisa anti-PGL-1 em localidade com alta endemicidade de hanseníase na região norte do estado de São Paulo. Hansen Int 230: 35–48.
- 75. Calado KLS, Vieira AG, Durães S, Bührer-Sékula S, Oliveira MLW-D-R (2005) Positividade sorológica antiPGL-I em contatos domiciliares e peridomiciliares de hanseníase em área urbana. An Bras Dermatol (Suppl. 3): S301–S306.
- 76. Andrade ARC, Grossi MAF, Bührer-Sékula S, Antunes CMF (2008) Soroprevalência do teste ML Flow em contatos de hanseníase de Minas Gerais. Rev Soc Bras Med Trop 41(Suppl. II): 56–59.
- 77. Grossi MA, Leboeuf MA, Andrade AR, Bührer-Sékula S, Antunes CM (2008) Fatores de risco para a soropositividade do ML Flow em pacientes com hanseníase. Rev Soc Bras Med Trop 41(Suppl. II): 39–44.
- 78. Ibekwe AC, Okonko IO, Onunkwo AU, Donbraye E, Babalola ET, et al. (2008) Baseline Salmonella agglutinin titres in apparently healthy freshmen in Awka, South Eastern, Nigeria. Sci Res Essays 3: 425–430.
- 79. Gupta A, My Thanh NT, Olsen SJ, Sivapalasingam S, My Trinh TT, et al. (2006) Evaluation of community-based serologic screening for identification of chronic Salmonella typhi carriers in Vietnam. Int J Infect Dis 10: 309–314. doi: 10.1016/j.ijid.2005.06.005
- 80. Pereira MM, Andrade J (1990) Human leptospirosis in a slum area in the city of Rio de Janeiro, Brazil – a serological and epidemiological study. Mem Inst Oswaldo Cruz 85: 47–52. doi: 10.1590/s0074-02761990000100007
- 81. Aguiar DM, Cavalcante GT, Camargo LMA, Labruna MB, Vasconcellos SA, et al. (2007) Anti-Leptospira spp. and anti-Brucella spp. antibodies in humans from rural area of Monte Negro municipality, state of Rondônia, Brazilian Western Amazon. Braz J Microbiol 38: 93–96. doi: 10.1590/s1517-83822007000100019
- 82. Lacerda HG, Monteiro GR, Oliveira CC, Suassuna FB, Queiroz JW, et al. (2008) Leptospirosis in a subsistence farming community in Brazil. Trans R Soc Trop Med Hyg 102: 1233–1238. doi: 10.1016/j.trstmh.2008.05.010
- 83. Reis RB, Ribeiro GS, Felzemburgh RD, Santana FS, Mohr S, et al. (2008) Impact of environment and social gradient on Leptospira infection in urban slums. PLoS Negl Trop Dis 2: e228. doi: 10.1371/journal.pntd.0000228
- 84. Caugant DA, Høiby EA, Magnus P, Scheel O, Hoel T, et al. (1994) Asymptomatic carriage of Neisseria meningitidis in a randomly sampled population. J Clin Microbiol 32: 323–330.
- 85. Cartwright KA, Stuart JM, Jones DM, Noah ND (1987) The Stonehouse survey: nasopharyngeal carriage of meningococci and Neisseria lactamica. Epidemiol Infect 99: 591–601. doi: 10.1017/s0950268800066449
- 86. Blackwell CC, Weir DM, James VS, Todd WT, Banatvala N, et al. (1990) Secretor status, smoking and carriage of Neisseria meningitidis. Epidemiol Infect 104: 203–209. doi: 10.1017/s0950268800059367
- 87. Imrey PB, Jackson LA, Ludwinski PH, England AC 3rd, Fella GA, et al (1995) Meningococcal carriage, alcohol consumption, and campus bar patronage in a serogroup C meningococcal disease outbreak. J Clin Microbiol 33: 3133–3137.
- 88. Olsen SF, Djurhuus B, Rasmussen K, Joensen HD, Larsen SO, et al. (1991) Pharyngeal carriage of Neisseria meningitidis and Neisseria lactamica in households with infants within areas with high and low incidences of meningococcal disease. Epidemiol Infect 106: 445–457. doi: 10.1017/s0950268800067492
- 89. MacLennan J, Kafatos G, Neal K, Andrews N, Cameron JC, et al. (2006) Social behavior and meningococcal carriage in British teenagers. Emerg Infect Dis 12: 950–957. doi: 10.3201/eid1206.051297
- 90. Orr HJ, Gray SJ, Macdonald M, Stuart JM (2003) Saliva and meningococcal transmission. Emerg Infect Dis 9: 1314–1315. doi: 10.3201/eid0910.030344
- 91. Tayal SC, Rashid S, Muttu KM, Hildreth AJ (1997) Meningococcal carriage: prevalence and sex-related risk factors. J Infect 34: 101–105. doi: 10.1016/s0163-4453(97)92357-4
- 92. Gazi H, Surucuoglu S, Ozbakkaloglu B, Akcali S, Ozkutuk N, et al. (2004) Oropharyngeal carriage and penicillin resistance of Neisseria meningitidis in primary school children in Manisa, Turkey. Ann Acad Med Singapore 33: 758–762.
- 93. Patrick DM, Champagne S, Goh SH, Arsenault G, Thomas E, et al. (2003) Neisseria meningitidis carriage during an outbreak of serogroup C disease. Clin Infect Dis 37: 1183–1188. doi: 10.1086/378743
- 94. Pavlopoulou ID, Daikos GL, Alexandrou H, Petridou E, Pangalis A, et al. (2004) Carriage of Neisseria meningitidis by Greek children: risk factors and strain characteristics. Clin Microbiol Infect 10: 137–142. doi: 10.1111/j.1469-0691.2004.00750.x
- 95. Bakir M, Yagci A, Ulger N, Akbenlioglu C, Ilki A, et al. (2001) Asymptomatic carriage of Neisseria meningitidis and Neisseria lactamica in relation to Streptococcus pneumoniae and Haemophilus influenzae colonization in healthy children: apropos of 1400 children sampled. Eur J Epidemiol 17: 1015–1018.
- 96. García Rojas A, Bordes Benítez A, Lafarga Capuz B, Vázquez Moreno J, López Villarrubia E, et al. (2000) Encuesta de portadores de Neisseria meningitidis en el área de salud de Gran Canaria. Rev Esp Salud Pública 74: 419–424. doi: 10.1590/s1135-57272000000400014
- 97. Gilmore A, Jones G, Barker M, Soltanpoor N, Stuart JM (1999) Meningococcal disease at the University of Southampton: outbreak investigation. Epidemiol Infect 123: 185–192. doi: 10.1017/s0950268899002794
- 98. Parodi V, Allende F, Torres E, Macedo M, Maglione R, et al. (1998) Portadores de Neisseria meningitidis en una población de Montevideo. Rev Méd Uruguay 14: 221–225.
- 99. Jackson LA, Alexander ER, DeBolt CA, Swenson PD, Boase J, et al. (1996) Evaluation of the use of mass chemoprophylaxis during a school outbreak of enzyme type 5 serogroup B meningococcal disease. Pediatr Infect Dis J 15: 992–998. doi: 10.1097/00006454-199611000-00012
- 100. Fitzpatrick PE, Salmon RL, Hunter PR, Roberts RJ, Palmer SR (2000) Risk factors for carriage of Neisseria meningitidis during an outbreak in Wales. Emerg Infect Dis 6: 65–69. doi: 10.3201/eid0606.0006
- 101. Clemens SA, da Fonseca JC, Azevedo T, Cavalcanti A, Silveira TR, et al. (2000) Soroprevalência para hepatite A e hepatite B em quatro centros no Brasil. Rev Soc Bras Med Trop 33: 1–10. doi: 10.1590/s0037-86822000000100001
- 102. Zago-Gomes MP, Stantolin GC, Perazzio S, Aikawa KH, Gonçalves CS, et al. (2005) Prevalence of anti-hepatitis A antibodies in children of different socioeconomic conditions in Vila Velha, ES. Rev Soc Bras Med Trop 8: 285–289. doi: 10.1590/s0037-86822005000400001
- 103. Kozlowski AG, Motta-Castro AR, Nascimento LB, Silva AM, Teles SA, et al. (2007) Prevalence of hepatitis A virus infection in Afro-Brazilian isolated communities in Central Brazil. Mem Inst Oswaldo Cruz 102: 121–123. doi: 10.1590/s0074-02762007000100021
- 104. de Alencar Ximenes RA, Martelli CM, Merchán-Hamann E, Montarroyos UR, Braga MC, et al. (2008) Multilevel analysis of hepatitis A infection in children and adolescents: a household survey in the Northeast and Central-west regions of Brazil. Int J Epidemiol 7: 852–861. doi: 10.1093/ije/dyn114
- 105. Braga WS, Borges FG, Barros Júnior GM, Martinho AC, Rodrigues IS, et al. (2009) Prevalence of hepatitis A virus infection: the paradoxical example of isolated communities in the western Brazilian Amazon region. Rev Soc Bras Med Trop 42: 277–281. doi: 10.1590/s0037-86822009000300008
- 106. Taghavi SA, Asl MKH, Talebzadeh M, Eshraghian A (2011) Seroprevalence study of hepatitis A virus in Fars province, southern Iran. Hepat Mon 11: 258–288.
- 107. Fix AD, Martin OS, Gallicchio L, Vial PA, Lagos R (2002) Age-specific prevalence of antibodies to hepatitis A in Santiago, Chile: risk factors and shift in age of infection among children and young adults. Am J Trop Med Hyg 66: 628–632.
- 108. Chung GE, Yim JY, Kim D, Lim SH, Park MJ, et al. (2001) Seroprevalence of hepatitis A and associated socioeconomic factors in young healthy Korean adults. Gut Liver 5: 88–92. doi: 10.5009/gnl.2011.5.1.88
- 109. Verhoef L, Boot HJ, Koopmans M, Mollema L, Van Der Klis F, et al. (2011) Changing risk profile of hepatitis A in The Netherlands: a comparison of seroprevalence in 1995–1996 and 2006–2007. Epidemiol Infect 139: 1172–1180. doi: 10.1017/s0950268810003043
- 110. Vasconcelos PF, Mota K, Straatmann A, Santos-Torres S, da Rosa AP, et al. (2000) Epidemia de dengue em Ipupiara e Prado, Bahia. Inquérito soro-epidemiológico. Rev Soc Bras Med Trop 33: 61–67. doi: 10.1590/s0037-86822000000100009
- 111. Vasconcelos PF, Lima JW, Raposo ML, Rodrigues SG, da Rosa JF, et al. (1999) Inquérito soro-epidemiológico na Ilha de São Luis durante epidemia de dengue no Maranhão. Rev Soc Bras Med Trop 32: 171–179. doi: 10.1590/s0037-86821999000200009
- 112. da Cunha RV, Maspero RC, Miagostovich MP, de Araújo ES, Luz da C, et al. (1997) Dengue infection in Paracambi, State of Rio de Janeiro, 1990–1995. Rev Soc Bras Med Trop 30: 379–383. doi: 10.1590/s0037-86821997000500005
- 113. Vasconcelos PF, Travassos da Rosa ES, Travassos da Rosa JFS, Freitas RB, Degallier N, et al. (1993) Epidemia de febre clássica de dengue causada pelo sorotipo 2 em Araguaina, Tocantins, Brazil. Rev Inst Med Trop São Paulo 35: 141–148. doi: 10.1590/s0036-46651993000200005
- 114. Siqueira JB, Martelli CM, Maciel IJ, Oliveira RM, Ribeiro MG, et al. (2004) Household survey of dengue infection in central Brazil: spatial point pattern analysis and risk factors assessment. Am J Trop Med Hyg 71: 646–651.
- 115. Lima VL, Rangel O, Andrade VR, Silveira NY, Oliveira SS, et al. (2007) Dengue: inquérito populacional para pesquisa de anticorpos e vigilância virológica no Município de Campinas, São Paulo, Brasil. Cad Saúde Pública 23: 669–680. doi: 10.1590/s0102-311x2007000300025
- 116. Pessanha JEM, Caiaffa WT, Kroon EG, Proietti FA (2010) Dengue em três distritos sanitários de Belo Horizonte, Brasil: inquérito soroepidemiológico de base populacional, 2006 a 2007. Rev Panam Salud Pública 27: 252–258. doi: 10.1590/s1020-49892010000400003
- 117. Karlberg J (1987) On the modelling of human growth. Stat Med 6: 185–192. doi: 10.1002/sim.4780060210
- 118. Travassos C, Viacava F, Pinheiro J, Brito A (2002) Utilização dos serviços de saúde no Brasil: gênero, características familiares e condição social. Rev Panam Salud Pública 11: 365–373. doi: 10.1590/s1020-49892002000500011
- 119. Drevenstedt GL, Crimmins EM, Vasunilashorn S, Finch CE (2008) The rise and fall of excess male infant mortality. Proc Natl Acad Sci USA 105: 5016–5021. doi: 10.1073/pnas.0800221105
- 120. Meller FO, Schäfer AA (2011) Fatores associados ao tipo de parto em mulheres brasileiras: PNDS 2006. Cienc Saúde Coletiva 16: 3829–3835. doi: 10.1590/s1413-81232011001000018
- 121. Walker SL, Lockwood DNJ (2006) The clinical and immunological features of leprosy. Br Med Bull 77–78: 103–121. doi: 10.1093/bmb/ldl010
- 122. Modlin RL (2010) The innate immune response in leprosy. Curr Opin Immunol 22: 48–54. doi: 10.1016/j.coi.2009.12.001
- 123. Korenchevsky V, Dennison M, Schalit R (1932) The response of castrated male rats to the injection of testicular hormone. Biochem J 26: 1306–1314.
- 124. Dougherty TF (1952) Effect of hormones on lymphatic tissue. Physiol Rev 32: 379–401.
- 125. Grossman CJ (1985) Interactions between the gonadal steroids and the immune system. Science 227: 257–261. doi: 10.1126/science.3871252
- 126. Frank C, Werber D, Cramer JP, Askar M, Faber M, et al. (2011) Epidemic profile of Shiga-toxin-producing Escherichia coli O104:H4 outbreak in Germany. N Engl J Med 365: 1771–1780. doi: 10.1056/nejmoa1106483
- 127. Blaser MJ (2011) Deconstructing a lethal foodborne epidemic. N Engl J Med 365: 1835–1836. doi: 10.1056/nejme1110896
- 128. Greinacher A, Friesecke S, Abel P, Dressel A, Stracke S, et al. (2011) Treatment of severe neurological deficits with IgG depletion through immunoadsorption in patients with Escherichia coli O104:H4-associated haemolytic uraemic syndrome: a prospective trial. Lancet 378: 1166–2273. doi: 10.1016/s0140-6736(11)61253-1