Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Mar 27.
Published in final edited form as: Int J Infect Dis. 2022 Jul 16;122:685–692. doi: 10.1016/j.ijid.2022.07.031

Possible sex difference in latent tuberculosis infection risk among close tuberculosis contacts

Paul Y Wada a,*, Allyson G Costa b, Mariana Araújo-Pereira c, Beatriz Barreto-Duarte c, Alexandra B Souza b,d, Michael S Rocha e, Marina C Figueiredo f, Megan M Turner f, Valeria C Rolla g, Afrânio L Kritski h, Marcelo Cordeiro-Santos b,d,i, Bruno B Andrade c,f,g,j,k,l,m,#, Timothy R Sterling a,f,#,**, Peter F Rebeiro a,n; Regional Prospective Observational Research in Tuberculosis (RePORT) Brazil Consortium
PMCID: PMC10041676  NIHMSID: NIHMS1875107  PMID: 35843494

Abstract

Objectives:

More men than women develop and die of tuberculosis (TB). Fewer data exist on sex differences in latent TB infection (LTBI). We assessed for potential sex differences in LTBI acquisition among close TB contacts.

Methods:

Regional Prospective Observational Research for TB-Brazil is an observational multi-center cohort of individuals with culture-confirmed pulmonary TB and their close contacts. Participants were enrolled from five sites in Brazil from June 2015 - June 2019. Close contacts were followed for 24 months after enrollment, with LTBI defined as a positive interferon- γ release assay (IGRA; QuantiFERON 3rd or 4th generation) at baseline or 6 months. We performed univariate, bivariate, and multivariable logistic regression and propensity-score weighted models to assess odds ratios (OR) and 95% confidence intervals (CI) for LTBI acquisition by birth sex among close contacts.

Results:

Of 1093, 504 (46%) female close contacts were IGRA positive compared to 295 of 745 (40%) men. The unadjusted OR for IGRA positivity among women vs men was 1.31 (95% CI: 1.08–1.58). Bivariate adjustments yielded ORs in women vs men ranging from 1.19 to 1.33 (P-value range: < 0.01–0.07). Multivariable regression and weighted models yielded similar ORs in women vs men, of 1.14 (95% CI: 0.92–1.41) and 1.15 (95% CI: 0.94–1.40), respectively.

Conclusion:

The point estimate for LTBI among close TB contacts in Brazil was higher in women, though less pronounced in multivariable models. If the sex difference in LTBI is confirmed in additional settings, studies of possible underlying differences in socio-behavioral factors or TB pathogenesis are warranted.

Keywords: Latent tuberculosis, Sex difference, Interferon-Gamma release assay, Health status disparities

Introduction

It is well known that there are sex differences in active tuberculosis (TB), with men more likely to contract and die of the disease (World Health Organization, 2021). This is particularly true in low- and middle-income countries (Horton et al., 2016). While the disproportionate active TB prevalence may be due in part to under-reporting in female populations (Saunders et al., 2019), this alone is unlikely to explain such trends, which have been seen in multiple settings (Neyrolles and Quintana-Murci, 2009; Rhines, 2013). Instead, behavioral and physiologic differences have been posited as probable drivers (Nhamoyebonde and Leslie, 2014; Dodd et al., 2016; Horton et al., 2020).

There are fewer data on sex differences in latent TB infection (LTBI) testing, treatment, and outcomes. While some studies have found no significant sex differences (Ting et al., 2014; Teklu et al., 2018; Ncayiyana et al., 2016), others appear to show higher rates of LTBI in men (Reichler et al., 2020; Sabri et al., 2019; He et al., 2015; de Souza et al., 2014; Chen et al., 2015). Further analyses found that men were more likely to receive (Fiske et al., 2014) and complete LTBI treatment (Hirsch-Moverman et al., 2015) and have less antitubercular medication toxicity (Pettit et al., 2013). Of note, several of these studies were in specific populations such as healthcare workers, and only the Tuberculosis Epidemiologic Studies Consortium Task Order-2 Team evaluated large numbers of close contacts with active TB disease (Reichler et al., 2020; Fiske et al., 2014).

Several known factors may interact with sex in increasing the risk for LTBI. Lower socio-economic status groups appear to have a higher prevalence of LTBI (Lule et al., 2020). Heavy alcohol use is associated with LTBI risk (Puryear et al., 2021) and incomplete treatment (Hirsch-Moverman et al., 2010), while active and passive smoking alters host immunity (Bai et al., 2018) and leads to higher infection rates (Lindsay et al., 2014; Patra et al., 2015). Comorbidities such as diabetes (Lee et al., 2017) and renal disease on dialysis (Shu et al., 2015) play a role, with diabetes, in particular, contributing to an ineffective or exaggerated immune response (Magee et al., 2020). Even after accounting for these separate factors, sex differences may still impact LTBI, LTBI conversion, and LTBI treatment success.

We sought to fill the current knowledge gap by evaluating potential sex differences in interferon-γ release assay (IGRA) positivity among close TB contacts in Brazil. To our knowledge, no prior studies have directly assessed a sex difference in IGRA conversion after prolonged TB exposure. Brazil is a highly diverse upper-middle income country with large gaps in income equality and decreasing yet substantial TB burden. Our well-characterized multi-center cohort closely represents the TB patient population in Brazil and is well positioned to address this question (Arriaga et al., 2021a). Based on preliminary data within our cohort (Souza et al., 2021; Arriaga et al., 2021b), we hypothesized that female close contacts would have higher rates of IGRA positivity than men, even after adjusting for potentially confounding factors.

Methods

Study population and follow-up

Regional Prospective Observational Research for TB in Brazil (RePORT-Brazil) is a prospective observational cohort study of culture-confirmed pulmonary TB cases and their close contacts (Arriaga et al., 2021a). There were five sites across three high TB-burden regions in Brazil: three in the southeast (Rio de Janeiro – Rio de Janeiro), one in the northeast (Salvador – Bahia), and one in the north (Manaus – Amazonas). Study participants were enrolled from June 2015 through June 2019 and followed for 2 years, with follow-up completed in June 2021. Close contacts were defined as individuals spending at least 4 hours per week in proximity to an index case within the 6 months preceding active TB diagnosis (Loredo et al., 2014). Close contacts were excluded if they had no available IGRA result at baseline or 6 months and index cases without culture-confirmed pulmonary disease. Index cases were not included if the individuals were ≤18 years of age, pregnant, breast-feeding, or already on antitubercular therapy for > 7 days.

Close contacts were encouraged by phone or text to present at study enrollment sites. Baseline evaluation included questionnaires, in-person physical examination, blood work, and chest imaging. At 6 months, close contacts returned for clinical assessment and repeat blood work and chest imaging. Subsequent evaluations occurred at 6-month intervals by phone throughout the 2-years follow-up period.

Exposure, outcome, and potential confounders

Close contacts were characterized by biological sex assigned at birth (male or female) to define the exposure of interest. No information was available on participants’ self-reported gender. IGRA testing was performed with QuantiFERON 3rd or 4th generation assay at baseline evaluation and again at six months in case of negative or indeterminate results at baseline to define the outcome of interest. IGRA positivity was defined as a positive result at baseline or 6 months; IGRA indeterminate results were considered IGRA negative. For this study, index case data were limited to sex, X-ray characteristics, and degree of smear positivity (scanty, 1 +, 2 +, and 3 +). Later in the study, groups of close contacts were also asked whether they slept in the same room or bed as the index case or had at least 5 hours of indoor exposure per day. Potential confounders of the exposure-outcome relationship were as follows: close contact baseline age in years, self-identified race, region of enrollment, education level, household income, body mass index, number of household members, and index case sex, smear positivity, and X-ray cavitary disease.

Data and statistical analysis

We used Chi-square and Wilcoxon Rank Sum tests to identify unadjusted differences between male and female close contacts for categorical and continuous variables, respectively. After multiply imputing variables with > 10% missingness over 10 iterations, we performed unadjusted, bivariate-adjusted, and multivariable logistic regression modeling to obtain the odds ratios (OR) and 95% (CI) for IGRA positivity by sex. Variance inflation factors were utilized within multivariable models to determine the possible collinearity of included covariates. We assessed the outcome association for sex by adjusting for single confounders in bivariate analyses. In contrast, we adjusted for all confounders simultaneously in the multivariable analysis. We performed propensity-score-matched models utilizing all available confounders, with exposure weights constructed and assigned according to the methods of Li and Greene (Li and Greene, 2013). We conducted sensitivity analyses accounting for possible false-positive IGRA conversion utilizing known specificities of QuantiFERON 3rd and 4th generation tests (Takasaki et al., 2018). Finally, we assessed for a potential sex difference in quantitative assay measurements (nil, mitogen, TB antigen tube 1 [TB1 antigen], and TB antigen tube 2 [TB2 antigen]) using unadjusted linear regression models stratified by IGRA positivity and clustered by participant. All analyses were conducted in R version 4.1.0 (R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/).

Ethical conduct of research and role of funders

The study was approved by the Institutional Review Boards at all enrollment sites and Vanderbilt (CAAE: 25102412.3.1001.5262); all participants provided written informed consent before inclusion. Funding was obtained through the Brazilian Ministry of Health, Department of Science and Technology (DECIT), as well as the United States National Institutes of Health (NIH). Funders played no role in cleaning data, conducting analyses, or interpreting results.

Results

The source population included 1917 close contacts and 1181 index cases, of which there were 1840 close contacts of 601 culture-confirmed pulmonary TB index cases. Two close contacts were excluded because IGRA results were unavailable at baseline or 6 months. Of the 1838 remaining close contacts, 1093 (59%) were female and 745 (41%) were male. Female close contacts had a median age of 36 years (interquartile range [IQR]: 20, 50), compared with 26 years for males (IQR: 13, 42). Both female and male contacts had similar percentages in self-reported race categories: 19% vs 20% White, 21% vs 20% Black, and 59% vs 59% Pardo/Mixed, respectively. Very few close contacts identified as Asian, Indigenous, or unknown (1%). Female close contacts were more often living in Rio de Janeiro (38% vs 34%) or Salvador (19% vs 13%) and residing within a poor community (favela, 25% vs 20%) than males. Female close contacts had similar exposure to index case cavitary disease (53% vs 50%; P = 0.40) and high smear positivity (3 + : 29% vs 28%; P = 0.24) than males. Furthermore, female close contacts described similar household exposure – sleeping in the same room (35% vs 31%; P = 0.22) and sleeping in the same bed (19% vs 17%; P = 0.58) – but more daily five-hour indoor exposure (87% vs 78%; P < 0.001) than males. 46 participants were living with HIV, including 22 females (2%) and 24 males (3%); 13 participants living with HIV were IGRA positive (28%). Two participants had end-stage renal disease, eight were on chemotherapy, and 20 received immunosuppressive medications.

A total of 799 close contacts had a positive IGRA result, including 504 (46%) females and 295 (40%) males (Table 1). 433 (40%) female and 246 (33%) male close contacts were IGRA positive at baseline, with 71 (6%) female and 49 (7%) male close contacts IGRA positive at 6 months. IGRA-positive close contacts were older (median age 36 years [IQR: 18, 50]) in comparison to IGRA-negative close contacts (median age 29 years [IQR: 15, 44]). Black race (53%), Rio de Janeiro region (55%), Salvador region (52%), community settings (favela, 63%), and exposure to cavitary disease (54%) or high smear positivity (2 + : 54%; 3 + : 53%) resulted in high IGRA positivity. There were two pregnant female participants at the time of enrollment, one of whom had a positive IGRA result at baseline. See Supplemental Tables 1 and 2 for detailed close contact demographics and exposure characteristics.

Table 1.

The association of female sex with IGRA positivity, accounting for various demographic, contextual, and clinical characteristics, among close contacts of an index pulmonary TB source case.

Factor IGRA negative N(%) or median(IQR)a IGRA positive N(%) or median(IQR)a Unadjusted OR (95% CI)b Sex-specific, bivariate adjusted OR (95% CI)c Sex-specific, multivariable-adjusted OR (95% CI)d
Sex
 Male 450 (60) 295 (40) Referent
 Female 589 (54) 504 (46) 1.31 (1.08–1.58) 1.14 (0.92–1.41)
Age 29 (15, 44) 36 (18, 50) 1.02 (1.01–1.02) 1.19 (0.98–1.45)
Race 1.31 (1.08–1.58)
 White 227 (64) 130 (36) Referent
 Black 173 (47) 199 (53) 2.01 (1.49–2.70)
 Asian 1 (20) 4 (80) 6.98 (0.77–63.25)
 Pardo/Mixed 629 (58) 460 (42) 1.28 (1.00–1.63)
 Indigenous 9 (64) 5 (36) 0.97 (0.32–2.96)
Region 1.23 (1.01–1.49)
 Rio de Janeiro 301 (45) 371 (55) Referent
 Manaus 590 (69) 270 (31) 0.37 (0.30–0.46)
 Salvador 145 (48) 156 (52) 0.87 (0.66–1.15)
Household Income 1.24 (1.02–1.50)
 > Minimum 405 (60) 269 (40) Referent
 <= Minimum 375 (52) 353 (48) 1.38 (1.13–1.69)
Education Level 1.33 (1.10–1.61)
 Primary 425 (53) 380 (47) Referent
 Secondary 299 (55) 240 (45) 0.90 (0.73–1.12)
 Higher 195 (64) 108 (36) 0.63 (0.48–0.83)
BMI 1.32 (1.09–1.59)
 Normal Weight 473 ( 57) 355 (43) Referent
 Underweight 48 (59) 34 (41) 0.94 (0.59–1.49)
 Overweight 277 (54) 232 (46) 1.12 (0.89–1.39)
 Obese 217 (56) 172 (44) 1.05 (0.82–1.34)
Household 3 (2, 5) 3 (2, 5) 0.98 (0.95–1.02) 1.31 (1.08–1.58)
Memberse
Index Case Sex 1.31 (1.08–1.58)
 Female 393 (56) 310 (44) Referent
 Male 646 (57) 489 (43) 0.96 (0.79–1.16)
Index Case 1.28 (1.06–1.56)
Cavitary Disease
 No 424 (67) 211 (33) Referent
 Yes 442 (46) 511 (54) 2.32 (1.89–2.86)
 Indeterminate 173 (69) 77 (31) 0.89 (0.65–1.22)
Index Case Smear 1.28 (1.06–1.56)
Positivity
 Scanty 124 (72) 49 (28) Referent
 1+ 322 (59) 222 (40) 1.63 (1.13–2.33)
 2+ 163 (46) 192 (54) 2.79 (1.91–4.07)
 3+ 203 (47) 230 (53) 2.81 (1.96–4.03)
Open in a new tab

IGRA positivity was defined as a single positive QuantiFERON 3rd or 4th generation result at baseline or 6 months from enrollment. IGRA indeterminate results were considered IGRA-negative.

Abbreviations: BMI = body mass index; CI = confidence interval; IGRA = interferon-γ release assay; IQR = interquartile range; N = number; OR = odds ratio.

a

Percentages are row percentages. Interquartile ranges denote the 25th and 75th percentiles.

b

Unadjusted odds ratios were obtained for IGRA positivity and the following specified factor.

c

For bivariate-adjusted odds ratios, IGRA positivity was adjusted for sex and the following specified factor.

d

For multivariable-adjusted odds ratios, IGRA positivity was adjusted for sex, age, race, region, household income, education level, BMI, number of household members, index case sex, index case cavitary disease, and index case smear positivity.

e

Number of members living in the same household, excluding the participating close contact.

The unadjusted odds of IGRA positivity were significantly higher for females compared with males (OR = 1.31; CI: 1.08–1.58). Sensitivity analyses accounting for possible false-positive IGRA results yielded the same unadjusted OR of 1.31 (CI: 1.09–1.59). Bivariate ORs for IGRA positivity, adjusted for single potential confounding factors, resulted in sex-specific effect estimates ranging from 1.19–1.33. Of these, only the bivariate OR adjusting for age had a CI including the null (OR = 1.19; CI: 0.98–1.45) (Table 1).

After performing multiple imputations for household income, education level, and index case smear positivity, multivariable logistic regression yielded a sex-specific OR of 1.14 (CI: 0.92–1.41). Propensity-score-matched weighted regression within and across multiple imputed datasets found similar ORs of 1.15 (CI: 0.94–1.40) and 1.15 (CI: 0.94–1.40), respectively. The multivariable-adjusted OR and propensity-score-matched ORs had CIs that included the null. However, they all yielded effect estimates that remained in the same direction as the unadjusted and bivariate-adjusted models.

For IGRA-negative participants, Wilcoxon Rank Sums revealed slight differences in IGRA quantitative assay measurements between male and female sex (Table 2). This was not seen, however, for IGRA-positive participants. Univariate linear regression modeling demonstrated no differences in quantitative assay measurements by birth sex when stratified by IGRA positivity.

Table 2.

Close contact IGRA quantitative assay measurements by sex and participant IGRA positivity.

Male median (IQR)a Female median (IQR) a Wilcoxon P-value Beta Coefficient (SE) Beta P-value
IGRA Negative (N) 818 1,076
 QFT Nil 0.05 (0.02, 0.08) 0.05 (0.03, 0.1) 0.003 0.006 (0.029) 0.842
 QFT Mitogen 10 (7.07, 10) 10 (7.11, 10) 0.783 −0.029 (0.139) 0.833
 QFT TB1 0.05 (0.03, 0.12) 0.07 (0.04, 0.14) <0.001 0.02 (0.022) 0.382
 QFT TB2 0.05 (0.02, 0.11) 0.07 (0.03, 0.13) <0.001 0.028 (0.027) 0.308
IGRA Positive (N) 347 578
 QFT Nil 0.06 (0.03, 0.14) 0.07 (0.04, 0.15) 0.073 0.033 (0.038) 0.383
 QFT Mitogen 10 (7.71, 10) 10 (8.22, 10) 0.457 −0.02 (0.157) 0.897
 QFT TB1 1.28 (0.5, 4.38) 1.51 (0.6, 4.45) 0.398 0.215 (0.213) 0.313
 QFT TB2 1.09 (0.46, 3.31) 1.32 (0.56, 3.64)  0.268 0.329 (0.23) 0.153
Open in a new tab

IGRA positivity was defined as a single positive QuantiFERON 3rd or 4th generation result at baseline or 6 months from enrollment. IGRA indeterminate results were considered IGRA-negative. Among 1039 IGRA-negative participants, there were 1894 IGRA results: 1853 negative and 41 indeterminate. Among 799 IGRA-positive participants, there were 925 IGRA results: 114 negative, seven indeterminate, and 804 positives. Quantitative assay measurements were clustered by participants to account for repeat testing. Wilcoxon Rank Sums and univariate linear regression models were performed to assess for a possible sex difference in quantitative assay measurements for IGRA-negative and positive groups.

Abbreviations: IGRA = interferon-γ release assay; IQR = interquartile range; QFT = QuantiFERON; N = number; SE = standard error; TB1 = TB antigen tube 1 (CD4 response); TB2 = TB antigen tube 2 (CD4 and CD8 response).

a

Interquartile ranges denote the 25th and 75th percentiles.

Discussion

For all analyses, the odds of LTBI acquisition (i.e., IGRA positivity) among female close contacts were higher than among males (Figure 1). Bivariate adjustments had a limited effect on the sex-specific effect size. Of these, only the adjustment for age resulted in a CI containing the null. Unweighted and weighted multivariable regression modeling resulted in more attenuated ORs, again with CIs containing the null. These findings suggest a small possible difference in sex-specific IGRA positivity within our close contact population.

Figure 1.

Figure 1.

Open in a new tab

Forest plot for unadjusted, adjusted, and propensity-score weighted sex-specific IGRA positivity odds ratios among close contacts of an index pulmonary TB source case.

Abbreviations: BMI = body mass index; CI = confidence interval: IGRA = interferon-γ release assay; OR = odds ratio; TB = tuberculosis

IGRA positivity was defined as a single positive QuantiFERON 3rd or 4th generation result at baseline or 6 months from enrollment. IGRA indeterminate results were considered IGRA-negative

*For bivariate-adjusted OR, IGRA positivity was adjusted for sex and the following factor. Multiple imputations were performed for household income, education level, and index case smear positivity before bivariate regression modeling.

ᶧFor multivariable-adjusted OR, IGRA positivity was adjusted for sex, age, region, race, household income, education level, BMI, household members, index case sex, index case cavitary disease, and index case smear positivity. Multiple imputations were performed for household income, education level, and index case smear positivity before multivariate regression modeling.

ᶴMultiple imputations were performed for all missing variables before propensity-score weighting with univariate regression modeling.

If replicated in other settings, a sex difference in LTBI would be informative from a health equity standpoint. TB disease stigma persists in Brazil, with higher stigmatization paradoxically among those with knowledge of LTBI (Rebeiro et al., 2020). Identifying and dismantling structural barriers that reinforce stigma can decrease the active TB disease burden – this can be achieved partly through early recognition, education, and treatment of LTBI. While there have been advances in sex and gender health equity in Brazil, there remains a divide in wages and labor force participation, resulting in downstream healthcare access disparities (Chant, 2006). Knowledge of LTBI risk in female close contacts can provide clinicians with the impetus for broader outreach and education within vulnerable populations.

Excess unadjusted LTBI acquisition was at least partially attributable to age, region, and income. Age is known to correlate with IGRA positivity and may indirectly represent the duration of exposure (Zelner et al., 2014). The median age of our female cohort was 10 years older than that of the male cohort, with proportionately more female close contacts aged 25 years and older (Figure 2). Accordingly, female close contacts in all age groups were found to have slightly higher IGRA positivity rates compared to males. Region and household income are likely to be correlated with other factors such as unstable housing, lower health awareness, overcrowding, and air pollution in contributing to LTBI and active TB disease. We measured residence in rural, suburban, and community settings. However, given collinearity with other confounders, we did not include this in the multivariable model. Instead, we utilized the number of household members as this was thought to be a better proxy for overcrowding.

Figure 2.

Figure 2.

Open in a new tab

Histogram with number of close contacts by age and sex. There were proportionately more female close contacts from age 25 and on, as compared with males.

Behavioral factors that affect sex differences in active TB disease may play a role in LTBI. Women may be at increased risk for LTBI as they represent a growing percentage of the total workforce in Brazil (Bruschini, 2007). Furthermore, women spend more time performing indoor household tasks on a weekly basis (Bruschini, 2007), which could affect exposure dynamics for domestic close contacts. Female-headed households have increased over time within Brazil, with female single-parent families often remaining within the lowest income quintile (Gukovas et al., 2016). The intersection between female sex and poverty has shown to be exceedingly complex (Chant, 2006), yet this may provide some explanation for LTBI risk, at least among female population subsets.

Physiologic mechanisms may be important in driving sex differences in LTBI among close contact populations. Macrophages are the primary targets of Mycobacterium tuberculosis upon inhalation into the lung, with multiple surface receptors to facilitate entry via phagocytosis (Glickman and Jacobs, 2001). At the same time, macrophages within females have enhanced phagocytic activity (Nhamoyebonde and Leslie, 2014). Once within phagosome vacuoles, M. tuberculosis alters compartmental characteristics for intracellular survival. This may not result in higher rates of active TB disease in females, as hormonal exertion of various proapoptotic effects counteracts mycobacterial proliferation. However, sex differences in macrophage activation could contribute to higher initial mycobacterial uptake among female close contacts.

There were limitations to our study. We could not provide detailed commentary on M. tuberculosis transmission dynamics. We did not assess the impact of multiple index cases or whether there were social or occupational exposures, nor did we have information on the total duration of exposure, of which 250 hours may predict risk for LTBI (Reichler et al., 2020). Nevertheless, we obtained information on exposure via chest X-ray findings and smear positivity results that could only be ascertained with a parallel enrollment of index cases and close contacts, a strength of this cohort. Furthermore, 5 -hour daily indoor exposure was likely a strong proxy for the total duration of exposure, as this represented high average exposure in the 6 months preceding index case diagnosis.

Our second limitation was that IGRA testing remains a surrogate marker for LTBI, measuring disease via immune reactivity. Individuals with immunodeficiency are more likely to test falsely negative or indeterminate. Only small subsets of our population had end-stage renal disease, HIV, or the use of chemotherapy or immunosuppressive medications; due to their effects on model fit, none of these variables were included in our regression analyses. False-positive results were likely uncommon because all close contacts had culture-confirmed pulmonary TB exposure. This was reflected in our sensitivity analysis to account for false-positive testing, which did not alter our results. Although not described in the literature, IGRA reactivity could theoretically differ between male and female individuals – mouse models have shown a more robust Th 1 response with increased interferon-γ production in the presence of estradiol (Klein and Flanagan, 2016). An inverse relationship may occur following testosterone exposure, with lower levels of interferon-γ secretion in natural killer cells of female-treated mice. In our cohort, there were slightly higher quantitative assay measurements among female populations compared with males, with significant Wilcoxon Rank Sums for nil, TB1 antigen, and TB2 antigen by sex among IGRA-negative participants. Even though no linear relationships were found between IGRA quantification and sex, future studies would help to tease out subtle differences.

Our last limitation involved losses across the cascade of care, which have been reported previously (Souza et al., 2021). This included attrition from initial outreach to informed consent, baseline and follow-up evaluations, and treatment start and completion. We could not assess whether the individuals who did not participate in our study differed significantly from those who enrolled. This could have led to sampling or selection biases, although less likely as we identified all new TB diagnoses for each site for enrollment over the study period.

Our findings suggest a possible sex difference in LTBI among close contacts in Brazil. Effect estimates were consistent across multiple analyses, although adjusted and weighted regression models resulted in attenuated sex-specific associations with IGRA positivity. If these findings are confirmed in other study populations, an investigation of underlying socio-behavioral and biological mechanisms to explain apparent sex disparities would be warranted.

Supplementary Material

Appendix; Supplementary Material

Acknowledgments

The RePORT-Brazil Consortium consists of 12 partner institutions from Brazil represented by the following members: Amanda Araújo da Costa, Instituto de Pesquisa Clínica Carlos Borborema, Fundação de Medicina Tropical Doutor Heitor Vieira Dourado, Manaus, Brazil; André Luiz Bezerra, Faculdade de Medicina, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil; Anna Cristina Calçada Carvalho, Faculdade de Medicina, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil, and Laboratório de Inovações em Terapias, Ensino e Bioprodutos, Fundação Oswaldo Cruz, Rio de Janeiro, Brazil; Anna Karla Silveira, Faculdade de Medicina, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil; Betânia M. F. Nogueira, IBIT, Fundação José Silveira, Salvador, Brazil, Multinational Organization Network Sponsoring Translational and Epidemiological Research (MONSTER) Initiative, Salvador, Brazil, and Faculdade de Medicina, Universidade Federal da Bahia, Salvador, Brazil; Bruna da Costa Oliveira Lima, Instituto de Pesquisa Clínica Carlos Borborema, Fundação de Medicina Tropical Doutor Heitor Vieira Dourado, Manaus, Brazil; Brenda Karoline Souza Carvalho, Instituto de Pesquisa Clínica Carlos Borborema, Fundação de Medicina Tropical Doutor Heitor Vieira Dourado, Manaus, Brazil; Bruna da Costa Oliveira Lima, Fundação de Medicina Tropical Doutor Heitor Vieira Dourado, Manaus, Brazil; Bruna Pires de Loiola, Instituto de Pesquisa Clínica Carlos Borborema, Fundação de Medicina Tropical Doutor Heitor Vieira Dourado, Manaus, Brazil; Carolina Arana Schmaltz Stanis, Instituto Nacional de Infectologia Evandro Chagas, Fundação Oswaldo Cruz, Rio de Janeiro, Brazil; Eline Naiane de Freitas Medeiros, Instituto de Pesquisa Clínica Carlos Borborema, Fundação de Medicina Tropical Doutor Heitor Vieira Dourado, Manaus, Brazil; Francine Peixoto Ignácio, Instituto Nacional de Infectologia Evandro Chagas, Fundação Oswaldo Cruz, Rio de Janeiro, Brazil; Hayna Malta-Santos, Faculdade de Medicina, Universidade Federal da Bahia, Salvador, Brazil, and Laboratório de Inflamação e Biomarcadores, Instituto Gonçalo Moniz, Fundação Oswaldo Cruz, Salvador, Brazil; Jéssica Rebouças Silva, Faculdade de Medicina, Universidade Federal da Bahia, Salvador, Brazil, and Laboratório de Inflamação e Biomarcadores, Instituto Gonçalo Moniz, Fundação Oswaldo Cruz, Salvador, Brazil; João Marine Neto, SMS-RJ, Rio de Janeiro, Brazil, and Hospital Federal do Andaraí, Ministério da Saúde, Brazil; Leandro Sousa Garcia, Instituto de Pesquisa Clínica Carlos Borborema, Fundação de Medicina Tropical Doutor Heitor Vieira Dourado, Manaus, Brazil; Maria Luciana Silva-Freitas, Laboratório Interdisciplinar de Pesquisas Médicas, Instituto Oswaldo Cruz, FIOCRUZ, Rio de Janeiro, Brazil; Mayla Gabriele Miranda de Melo, Faculdade de Medicina, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil; Rosa Maria Placido-Pereira, Laboratório Interdisciplinar de Pesquisas Médicas, Instituto Oswaldo Cruz, FIOCRUZ, Rio de Janeiro, Brazil; Samyra Almeida-Da-Silveira, Laboratório Interdisciplinar de Pesquisas Médicas, Instituto Oswaldo Cruz, FIOCRUZ, Rio de Janeiro, Brazil; Vanessa de Souza Nascimento, IBIT, Fundação José Silveira, Salvador, Brazil, MONSTER Initiative, Salvador, Brazil, and Bahiana School of Medicine and Public Health, Bahia Foundation for the Development of Sciences, Salvador, Brazil. Team members of clinical and laboratory platforms of RePORT-Brazil include: Aline Benjamin, Quezia Medeiros, Francine Ignacio, Adriano Gomes, Elisangela Silva, Jamile Garcia, Renta Spener-Gomer, Martinelle Godinho, Adriana Rezende, Andre Bezerra, Alice Andrade, and others. A special thanks to the following members for their administrative and logistical support: Elze Leite, FIOCRUZ, Salvador, Brazil; Eduardo Gama, FIOCRUZ, Rio de Janeiro, Brazil; Elcimar Junior, Fundação de Medicina Tropical Doutor Heitor Vieira Dourado, Manaus, Brazil; Leticia Linhares, Vanderbilt University Medical Center, Nashville, USA; Hilary Vansell, Vanderbilt University Medical Center, Nashville, USA.

Funding

This work was supported by the Departamento de Ciência e Tecnologia–Secretaria de Ciência e Tecnologia–Ministério da Saúde, Brazil (25029.0 0 0507/2013-07) and the National Institute of Allergy and Infectious Diseases (NIAID) at the National Institutes of Health (U01 AI069923, R01 AI147765, R01 AI20790, K01 AI131895).

Footnotes

Declaration of Competing Interest

We declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

VCR: The author has served as a consultant for ONU-OMS for the HIV response in Myanmar and received payment from Glaxo-SmithKline, Qiagen, and Virology Education for educational events

PFR: The author has served as a consultant for Gilead (2020) and Johnson & Johnson (2021)

All the other authors have no competing interests to declare.

Supplementary materials

Supplementary material associated with this article can be found, in the online version, at doi: 10.1016/j.ijid.2022.07.031.

Ethical approval statement

The study was approved by the Institutional Review Boards at all enrollment sites and at Vanderbilt (CAAE: 25102412.3.1001.5262); all participants provided written informed consent before inclusion. Funding was obtained through the Brazilian Ministry of Health, Department of Science and Technology (DECIT), as well as the United States National Institutes of Health (NIH). Funders played no role in cleaning data, conducting analyses, or interpreting results.

References

  1. Arriaga MB, Amorim G, Queiroz ATL, Rodrigues MMS, Araújo-Pereira M, Nogueira BMF, Souza AB, Rocha MS, Benjamin A, Moreira ASR, de Oliveira JG, Figueiredo MC, Turner MM, Alves K, Durovni B, Lapa-E-Silva JR, Kritski AL, Cavalcante S, Rolla VC, Cordeiro-Santos M, Sterling TR, Andrade BBRePORT Brazil consortium. Novel stepwise approach to assess representativeness of a large multicenter observational cohort of tuberculosis patients: the example of Re-PORT Brazil. Int J Infect Dis 2021a;103:110–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Arriaga MB, Rocha MS, Nogueira BMF, Nascimento V, Araújo-Pereira M, Souza AB, Andrade AMS, Costa AG, Gomes-Silva A, Silva EC, Figueiredo MC, Turner MM, Durovni B, Lapa-E-Silva JR, Kritski AL, Cavalcante S, Rolla VC, Cordeiro-Santos M, Sterling TR, Andrade BB. The Effect of Diabetes and Prediabetes on Mycobacterium tuberculosis Transmission to Close Contacts. J Infect Dis 2021b;224:2064–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bai X, Aerts SL, Verma D, Ordway DJ, Chan ED. Epidemiologic evidence of and potential mechanisms by which second-hand smoke causes predisposition to latent and active tuberculosis. Immune Netw 2018;18:e22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bruschini MCA. Work and gender in Brazil in the last ten years. Cad Pesqui 2007;37:537–72. [Google Scholar]
  5. Chant SH. Re-thinking the “feminization of poverty” in relation to aggregate gender indices. J Hum Dev Capabil 2006;7:201–20. [Google Scholar]
  6. Chen C, Zhu T, Wang Z, Peng H, Kong W, Zhou Y, Shao Y, Zhu L, Lu W. High latent TB infection rate and associated risk factors in the Eastern China of low TB incidence. PLoS One 2015;10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. de Souza FM, do Prado TN, Pinheiro Jdos S, Peres RL, Lacerda TC, Loureiro RB, Carvalho JA, Fregona G, Dias ES, Cosme LB, Rodrigues RR, Riley LW, Maciel EL. Comparison of interferon-γ release assay to two cut-off points of tuberculin skin test to detect latent Mycobacterium tuberculosis infection in primary health care workers. PLoS One 2014;9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dodd PJ, Looker C, Plumb ID, Bond V, Schaap A, Shanaube K, Muyoyeta M, Vynnycky E, Godfrey-Faussett P, Corbett EL, Beyers N, Ayles H, White RG. Age- and sex-specific social contact patterns and incidence of Mycobacterium tuberculosis infection. Am J Epidemiol 2016;183:156–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Fiske CT, Yan FX, Hirsch-Moverman Y, Sterling TR, Reichler MR. Tuberculosis Epidemiologic Studies Consortium Task Order 2 Team. Risk factors for treatment default in close contacts with latent tuberculous infection. Int J Tuberc Lung Dis 2014;18:421–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Glickman MS, Jacobs WR Jr. Microbial pathogenesis of Mycobacterium tuberculosis: dawn of a discipline. Cell 2001;104:477–85. [DOI] [PubMed] [Google Scholar]
  11. Gukovas RM, Muller M, Pereira AC, Reimao ME. Washington: World Bank, A SNAP-SHOT OF GENDER IN BRAZIL TODAY: INSTITUTIONS, OUTCOMES, AND A CLOSER LOOK AT RACIAL AND GEOGRAPHIC DIFFERENCES GENDER ISSUES IN BRAZIL. World Bank open knowledge repository; 2016. https://openknowledge.worldbank.org/handle/10986/25976 (accessed 02/01/2022). [Google Scholar]
  12. He G, Li Y, Zhao F, Wang L, Cheng S, Guo H, Klena JD, Fan H, Gao F, Gao F, Han G, Ren L, Song Y, Xiong Y, Geng M, Hou Y, He G, Li J, Guo S, Yang J, Yan D, Wang Y, Gao H, An J, Duan X, Wu C, Duan F, Hu D, Lu K, Zhao Y, Rao CY, Wang Y. The prevalence and incidence of latent tuberculosis infection and its associated factors among village doctors in China. PLoS One 2015;10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Hirsch-Moverman Y, Bethel J, Colson PW, Franks J, El-Sadr W. Predictors of latent tuberculosis infection treatment completion in the United States: an inner city experience. Int J Tuberc Lung Dis 2010;14:1104–11. [PMC free article] [PubMed] [Google Scholar]
  14. Hirsch-Moverman Y, Shrestha-Kuwahara R, Bethel J, Blumberg HM, Venkatappa TK, Horsburgh CR, Colson PWTuberculosis Epidemiologic Studies Consortium (TBESC). Latent tuberculous infection in the United States and Canada: who completes treatment and why? Int J Tuberc Lung Dis 2015;19:31–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Horton KC, Hoey AL, Béraud G, Corbett EL, White RG. Systematic review and meta–analysis of sex differences in social contact patterns and implications for tuberculosis transmission and control. Emerg Infect Dis 2020;26:910–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Horton KC, MacPherson P, Houben RM, White RG, Corbett EL. Sex differences in tuberculosis burden and notifications in low- and middle-income countries: a systematic review and meta-analysis. PLoS Med 2016;13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Klein SL, Flanagan KL. Sex differences in immune responses. Nat Rev Immunol 2016;16:626–38. [DOI] [PubMed] [Google Scholar]
  18. Lee MR, Huang YP, Kuo YT, Luo CH, Shih YJ, Shu CC, Wang JY, Ko JC, Yu CJ, Lin HH. Diabetes mellitus and latent tuberculosis infection: a systematic review and metaanalysis. Clin Infect Dis 2017;64:719–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Li L, Greene T. A weighting analogue to pair matching in propensity score analysis. Int J Biostat 2013;9:215–34. [DOI] [PubMed] [Google Scholar]
  20. Lindsay RP, Shin SS, Garfein RS, Rusch ML, Novotny TE. The Association between active and passive smoking and latent tuberculosis infection in adults and children in the united states: results from NHANES. PLoS One 2014;9:e93137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Loredo C, Cailleaux-Cezar M, Efron A, de Mello FC, Conde MB. Yield of close contact tracing using two different programmatic approaches from tuberculosis index cases: a retrospective quasi-experimental study. BMC Pulm Med 2014;14:133. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Lule SA, Gupta RK, Krutikov M, Jackson C, Southern J, Abubakar I. The relationship between social risk factors and latent tuberculosis infection among individuals residing in England: a cross-sectional study. BMJ Glob Health 2020;5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Magee MJ, Trost SL, Salindri AD, Amere G, Day CL, Gandhi NR. Adults with Mycobacterium tuberculosis infection and pre-diabetes have increased levels of QuantiFERON interferon-gamma responses. Tuberculosis (Edinb) 2020;122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Ncayiyana JR, Bassett J, West N, Westreich D, Musenge E, Emch M, Pettifor A, Hanrahan CF, Schwartz SR, Sanne I, van Rie A. Prevalence of latent tuberculosis infection and predictive factors in an urban informal settlement in Johannesburg, South Africa: a cross-sectional study. BMC Infect Dis 2016;16:661. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Neyrolles O, Quintana-Murci L. Sexual inequality in tuberculosis. PLoS Med 2009;6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Nhamoyebonde S, Leslie A. Biological differences between the sexes and susceptibility to tuberculosis. J Infect Dis 2014;209:S100–6. [DOI] [PubMed] [Google Scholar]
  27. Patra J, Bhatia M, Suraweera W, Morris SK, Patra C, Gupta PC, Jha P. Exposure to second-hand smoke and the risk of tuberculosis in children and adults: a systematic review and meta-analysis of 18 observational studies. PLoS Med 2015;12 discussion e1001835. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Pettit AC, Bethel J, Hirsch-Moverman Y, Colson PW, Sterling TR. Tuberculosis Epidemiologic Studies Consortium. Female sex and discontinuation of isoniazid due to adverse effects during the treatment of latent tuberculosis. J Infect 2013;67:424–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Puryear SB, Fatch R, Beesiga B, Kekibiina A, Lodi S, Marson K, Emenyonu NI, Muyindike WR, Kwarisiima D, Hahn JA, Chamie G. Higher levels of alcohol use are associated with latent tuberculosis infection in adults living with Human Immunodeficiency Virus. Clin Infect Dis 2021;72:865–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Rebeiro PF, Cohen MJ, Ewing HM, Figueiredo MC, Peetluk LS, Andrade KB, Eakin M, Zechmeister EJ, Sterling TR. Knowledge and stigma of latent tuberculosis infection in Brazil: implications for tuberculosis prevention strategies. BMC Public Health 2020;20:897. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Reichler MR, Khan A, Yuan Y, Chen B, McAuley J, Mangura B, Sterling TRTuberculosis Epidemiologic Studies Consortium Task Order 2 Team. Duration of exposure among close contacts of patients with infectious tuberculosis and risk of latent tuberculosis infection. Clin Infect Dis 2020;71:1627–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Rhines AS. The role of sex differences in the prevalence and transmission of tuberculosis. Tuberculosis (Edinb) 2013;93:104–7. [DOI] [PubMed] [Google Scholar]
  33. Sabri A, Quistrebert J, Naji Amrani H, Abid A, Zegmout A, Abderrhamani Ghorfi I, Souhi H, Boucaid A, Benali A, Abilkassem R, Kmari M, Hassani A, Lahcen B, Siah S, Schurr E, Boisson-Dupuis S, Casanova JL, Lahlou A, Laatiris A, Louzi L, Ouarssani A, Bourazza A, Aouragh A, Mustapha B, Messaoudi N, Agader A, Cobat A, Abel L, El Baghdadi J. Prevalence and risk factors for latent tuberculosis infection among healthcare workers in Morocco. PLoS One 2019;14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Saunders MJ, Tovar MA, Collier D, Baldwin MR, Montoya R, Valencia TR, Gilman RH, Evans CA. Active and passive case-finding in tuberculosis-affected households in Peru: a 10-year prospective cohort study. Lancet Infect Dis 2019;19:519–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Shu CC, Hsu CL, Lee CY, Wang JY, Wu VC, Yang FJ, Wang JT, Yu CJ, Lee LN. Comparison of the prevalence of latent tuberculosis infection among non-dialysis patients with severe chronic kidney disease, patients receiving dialysis, and the dialysis-unit staff: a cross-sectional study. PLoS One 2015;10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Souza AB, Arriaga MB, Amorim G, Araújo-Pereira M, Nogueira BMF, Queiroz ATL, Figueiredo MC, Rocha MS, Benjamin A, Moreira ASR, Oliveira JG, Rolla V, Durovni B, Lapa E, Silva JR, Kritski AL, Cavalcante S, Sterling T, Andrade BB, Cordeiro-Santos MRePORT Brazil consortium. Determinants of losses in the latent tuberculosis infection cascade of care in Brazil. BMJ Glob Health 2021;6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Takasaki J, Manabe T, Morino E, Muto Y, Hashimoto M, Iikura M, Izumi S, Sugiyama H, Kudo K. Sensitivity and specificity of QuantiFERON-TB Gold Plus compared with QuantiFERON-TB Gold In-Tube and T-SPOT.TB on active tuberculosis in Japan. J Infect Chemother 2018;24:188–92. [DOI] [PubMed] [Google Scholar]
  38. Teklu T, Legesse M, Medhin G, Zewude A, Chanyalew M, Zewdie M, Wondale B, Haile-Mariam M, Pieper R, Ameni G. Latent tuberculosis infection and associated risk indicators in pastoral communities in southern Ethiopia: a community based cross-sectional study. BMC Public Health 2018;18:266. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Ting WY, Huang SF, Lee MC, Lin YY, Lee YC, Feng JY, Su WJ. Gender disparities in latent tuberculosis infection in high-risk individuals: a cross-sectional study. PLoS One 2014;9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. World Health Organization, Global tuberculosis report 2021. https://www.who.int/publications/i/item/9789240037021, 2021 (accessed 02/01/2022).
  41. Zelner JL, Murray MB, Becerra MC, Galea J, Lecca L, Calderon R, Yataco R, Contreras C, Zhang Z, Grenfell BT, Cohen T. Age-specific risks of tuberculosis infection from household and community exposures and opportunities for interventions in a high-burden setting. Am J Epidemiol 2014;180:853–61. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Appendix; Supplementary Material
NIHMS1875107-supplement-Appendix__Supplementary_Material.docx (99.2KB, docx)

RESOURCES