Epidemiological clues: tracing the impact of Mtb on human coinfections

Areas with high TB incidence frequently overlap with other infectious diseases in endemic regions, yet coinfections remain understudied. This is especially relevant in high-burden countries, such as India, Indonesia, China, the Philippines, Pakistan, Nigeria, Bangladesh, and the Democratic Republic of the Congo, which together account for over two-thirds of global TB cases.23 In these environments, populations are also affected by HIV, malaria, helminths, enteric pathogens, and opportunistic bacterial and fungal pathogens.75 These infections can alter TB pathogenesis, diagnosis, and treatment outcomes. However, current data on coinfections are fragmented, often limited to single pathogens and specific localities. The urgency to adopt a more integrated research approach is heightened by global shifts such as climate change, urbanization, and increased human mobility, which facilitate the emergence and spread of infectious diseases.76

In this review, we have compiled sixty-six cohort and analytical studies addressing coinfections with Mtb, summarized in Table S3. This descriptive table gathers all publications discussed in this section, providing the foundation for the comparative discussion that follows. Moreover, given the frequent lack of precision in defining TB states, we used only the terminology explicitly provided by each study: we refer to LTBI when specified, and otherwise use the term Mtb infection. Building on this, we aim to understand the real-world landscape of Mtb coinfections in humans, including their distribution, impact on TB outcomes, and potential contributions to increased case numbers or disease severity.

The synergistic threat of Mtb/HIV coinfection

Mtb and HIV coinfection is widely recognized as the most significant infectious risk factor for TB progression. In 2023, approximately 24% of the 660,000 individuals with TB and HIV died compared to 11% of those without HIV who died from TB.77 TB remains the leading cause of hospitalization among HIV-infected adults (18%) and children (10%), and TB-related in-hospital mortality worldwide was 25% among adults and 30% among children with HIV infection.78 HIV increases risk of TB progression to active disease, primarily by disrupting local immune responses within granulomas.79 Notably, people living with chronic HIV infection show significantly reduced Mtb-specific Th1 cells.80 This disruption can double the incidence of TB within the first year of HIV infection81,82 and increases the risk of LTBI reactivation.83,84 Conversely, active TB may enhance HIV-1 replication, promoting early viral dissemination.85 Both latent and active TB have been associated with strong immune activation in patient samples, which can accelerate HIV progression.86,87 HIV also predisposes individuals to further opportunistic infections. Notably, coinfection between Mtb and Pneumocystis carinii has been linked to more severe pneumonia in HIV-infected patients,88 and Mtb/Streptococcus pneumoniae coinfections have been documented in HIV-positive individuals.89,90 Mtb infection is also a suspected factor in exacerbation in HIV/Leishmania donovani coinfection.91

While most studies underscore the exacerbating nature of TB/HIV coinfection, recent findings suggest that asymptomatic TB may confer some degree of immune modulation: Kusejko et al. reported lower HIV viral loads in individuals with LTBI compared to TB-negative patients.92 Additionally, Tepekule et al. showed that transcriptomic changes in peripheral blood mononuclear cells (PBMCs) from HIV-positive individuals with LTBI were associated with baseline immune activation and reduced viremia, a phenomenon that was absent in active TB.93 These results suggest a complex interplay between TB status and HIV progression, raising the possibility that asymptomatic Mtb infection may confer a degree of heterologous protection. However, these findings remain preliminary and are derived from limited cohorts; further investigation is needed to assess their robustness and generalizability. Together, these data highlight the need for nuanced and context-specific studies of Mtb/HIV coinfection, as well as dedicated in vitro investigations to elucidate the underlying immunological mechanisms; these topics will be explored in more detail later in Section “bridging scales: from population patterns to animal models and cellular mechanisms of coinfection” of this review.

Increased susceptibility and severity in coinfections with Mtb

In this section, we examine how TB infection may alter susceptibility to, or the clinical course of, other infectious diseases. Because TB and other infections are associated with immune perturbations, coinfections are often reported as increased susceptibility and/or exacerbated outcomes, an observation that also applies to TB. In addition to HIV, other chronic viral infections, such as cytomegalovirus (CMV) and hepatitis B and C virus (HBV and HCV), are established TB risk factors. CMV seropositivity is linked to LTBI94 and increased progression to active TB,95,96 particularly in children infected with CMV within their first year.97 With an overall prevalence of 7% among Mtb-infected patients,98 HCV infection is associated with higher TB disease risk.99,100 In a UK cohort, 18.4% of patients with newly diagnosed TB had HBV or HCV markers, higher than the national average, with coinfection prevalence greater in active TB compared to LTBI.101 Nonetheless, rates of coinfection among individuals with LTBI remain elevated, as also observed in Southern California between 2008 and 2019.102

Regarding acute viral infections, influenza viruses provide a notable model for studying the impact on Mtb infection. Epidemiological data often show an increase in influenza incidence among patients with TB103,104 or a worsening of either disease during coinfection,105,106,107 which is consistent with the historical data described in Subsection “mutual shaping of TB and influenza pandemics”. However, many studies also refrain from drawing a clear conclusion, instead suggesting a neutral or inconclusive interaction between the two diseases.108,109,110,111,112,113

Turning to bacterial coinfections, they are frequently observed in settings with a high TB burden. In Far East Russia, where TB incidence reached 105.7 per 100,000 in 2008, children with Mtb infection were more frequently colonized by Haemophilus influenzae than healthy controls, leading to acute upper and lower respiratory tract infections.114 Similarly, Gram-negative bacteria, such as Klebsiella spp. and Pseudomonas spp., were found in 33% of Mtb-infected patients’ sputum, compared to 9% of all patients, in a Cambodian cohort.115 These findings suggest that individuals with TB experience an increased burden and more severe pathological outcomes related to bacterial superinfections. A similar trend can be observed for some fungal infections, with a high prevalence of Candida spp. in Mtb-infected patients’ sputum.116,117,118

Finally, parasitic coinfections further illustrate the complexity of TB-associated susceptibility. In countries with high parasitic prevalence, parasitic species are found to be concurrent with Mtb infection in multiple organs, which increases antibacterial therapy intolerance and deteriorates TB disease prognosis.119 These patterns illustrate how parasitic coinfections can amplify disease severity rather than merely coexist. Despite the large geographical overlap between endemic regions, few studies document the cross-sectional prevalence of malaria in Mtb-infected patients, with rates ranging from 5% to 37% depending on the regions and parameters considered.120,121 However, an ecological study based on epidemiological surveillance data from the Brazilian Amazon revealed a spatial association between malaria and TB, independently of socio-economic factors, suggesting potential synergistic drivers of disease burden at the population level.122 Similarly, helminths and Mtb infections largely overlap at the population and geographic level, with a pooled prevalence of 29.69%.123 In Brazil and Ethiopia, active TB was significantly associated with intestinal helminth infections.124,125 However, in HIV-positive populations in Uganda and Brazil, no association was found with gastrointestinal parasites or Mansonella perstans,126 although schistosomiasis was linked to TB progression in Uganda.127 Taken together, these studies highlight the heterogeneous impact of parasitic coinfections on TB outcomes, with effects ranging from increased susceptibility and disease progression to the absence of detectable associations, underscoring the importance of pathogen-specific and context-dependent interactions.

In some of these studies, additional parameters challenge the causal link between TB and coinfections. For instance, the higher prevalence of Gram-negative bacteria observed in Cambodian Mtb-infected patients may be attributed to lung cavitation, which independently favors bacterial colonization.115 Likewise, factors such as living conditions, limited access to healthcare, and poor sanitation represent additional confounders that could bias the observed associations; however, these risk factors are rarely acknowledged and thoroughly investigated in epidemiological coinfection studies. The example of influenza also highlights that not all interactions follow a straightforward exacerbation pattern, with several studies reporting neutral or inconclusive outcomes.103,104,105,106,107,108,109,110,111,112,113

Overall, coinfections in the context of TB are frequently associated with increased susceptibility to, or greater severity of, secondary infections, as a wide range of pathogens appear more common among Mtb-infected individuals. However, establishing a causal link remains challenging, as in many cohorts, additional factors can independently facilitate colonization or symptomatic disease. Altogether, these variables highlight the challenge of distinguishing the direct immunological impact of Mtb infection from the secondary consequences of tissue damage and chronic inflammation, thereby underscoring the necessity for enhanced surveillance of patients.

Hints of protection in Mtb coinfections

The potential protective effect of Mtb infection against other pathogens has recently emerged as a surprising and largely understudied phenomenon. In addition to the historical data presented in Subsection “the TB/leprosy cross-immunity puzzle”, the possible cross-protection between TB and leprosy has been shown in 2016 in the Marshallese population in Arkansas, USA, where a negative association was found between the presence of anti-Mycobacterium leprae antibodies and positive IGRA tests.128 Similar observations have been reported more unexpectedly in unrelated infectious contexts. In a large population-based study in Tennessee, the incidence of Chlamydia trachomatis infection was found to be reduced within the first year post-TB diagnosis.129 In line with these findings, Kusejko et al. also reported reduced incidence of opportunistic infections such as candidiasis and oral hairy leukoplakia in individuals with LTBI, compared to healthy individuals.92 However, these observations are largely associative and do not establish causality.

In contrast to the studies mentioned earlier, in a Tanzanian cohort, pathogenic respiratory bacteria were less likely to be detected in Mtb-infected patients than in controls.130 Similarly, a study in Ethiopia found that asymptomatic helminth infection correlated with lower sputum smear positivity in active TB, suggesting reduced Mtb bacterial burden.131 These contrasting observations raise the possibility of pathogen-pathogen interactions, either affecting susceptibility to secondary infections or modulating Mtb burden.

Elsewhere, Lee et al. observed a significant decrease in the incidence of TB among people infected with measles in the 2000–2001 outbreak in Korea compared to the general population, but they did not conclude that measles reduces the risk of TB as no plausible mechanisms have been elucidated.132 Other explanations, such as improved healthcare access, have been proposed for low coinfection rates of Streptococcus pneumoniae and Mtb in tertiary health facilities in Nigeria.133

Recently, the COVID-19 pandemic has offered unexpected insights into Mtb-related immune modulation. While some case reports initially raised concerns about the reactivation of LTBI following SARS-CoV-2 infection, these were often based on hospitalized patients and may have been biased toward symptomatic coinfections.134 At a population level, Inoue et al. demonstrated an inverse correlation between the burden of past TB epidemics, particularly those of the 1950s, and COVID-19 incidence and mortality.135 Supporting this idea, Takahashi employed instrumental variable analysis to show a significant association between LTBI prevalence and reduced COVID-19 mortality.136 However, large population-based studies like these remain rare, and current data are insufficient to determine whether Mtb exacerbates, protects from, or has no impact on the outcome of secondary infections.

Interpretive limitations of current evidence

Despite these numerous reports suggesting either exacerbation or protection, most epidemiological studies available on coinfections with TB simply document their presence. Several studies have highlighted the occurrence of diverse bacterial coinfections, without drawing clear conclusions regarding the severity outcome.137,138,139,140 Among the coinfections discussed earlier in this review, studies on Streptococcus pneumoniae, Aspergillus spp. and Candida spp. report their presence in TB cohorts, but do not include comparisons with non-Mtb-infected populations.141,142,143,144,145 No association between the risk of infection with Leishmania donovani and Mtb coinfection was reported in a village in Eastern Sudan,146 and the same applies to non-tuberculous mycobacteria in a cohort in rural Mozambique147 or in Taiwan. For the latter, an increase in the prevalence has been shown between 2005 and 2013, but with no clear evaluation of the coinfection outcome.148 Moreover, we found no epidemiological studies on several prevalent past and present epidemics, for instance plague, syphilis and flavivirus infections.

Overall, these findings highlight the dual nature of TB coinfections: while many studies report the detrimental impact of coinfection on TB outcomes, a subset of epidemiological data suggest that Mtb infection may also exert protective effects in specific contexts. These findings in human data reflect the heterogeneity of host-pathogen interactions, as one pathogen may modulate the immune environment in ways that limit the other’s progression; mechanisms that are discussed further in Section “bridging scales: from population patterns to animal models and cellular mechanisms of coinfection” of this review. Importantly, the scarcity and inconsistency of such observations call for cautious interpretation, as a considerable number of studies find no clear association between Mtb infection and the coinfection outcome. Furthermore, epidemiological associations may be influenced by numerous confounding factors, including HIV status, malnutrition, diabetes, smoking, antimicrobial exposure, socioeconomic conditions, and access to healthcare, which are not always adequately accounted for across studies. This lack of consistent direction, combined with the broad spectrum of TB and its immunological manifestations, emphasizes the need for more systematic investigation, especially in TB-endemic regions, and the relevance of understanding the mechanisms behind these interactions.