Lethal interaction: the colliding epidemics of tobacco and tuberculosis

Abstract

Tobacco consumption ranks high among the leading health risks and tuberculosis (TB) is a major public health issue in countries where the smoking problem has reached epidemic proportions. Given that both smoking and TB are major health concerns and are widely prevalent in several countries, it is surprising that the association between smoking and TB is still a matter of debate and controversy. Although several studies have evaluated the effect of smoking on TB, the association has been largely overlooked by the TB and public health communities at large. Three recent reviews, including two meta-analyses, have summarized a large body of published literature on the association between smoking and various TB outcomes. These reviews show that there is considerable evidence that tobacco smoking is associated with TB. The evidence is strong for TB disease but less strong for TB infection and mortality. Even if the effect is relatively modest, the population-attributable risk is likely to be substantial due to the widespread nature of tobacco exposure. TB control programs must begin to address tobacco control as a potential preventive intervention. Since tobacco control will have multiple health benefits, it is likely to be a highly cost-effective intervention from a societal perspective.

Keywords::

• mortality
• smoking
• tobacco
• tuberculosis

Figure 1. Global prevalence of smoking among men.

Reproduced with permission from [3].

Figure 2. Estimated numbers of new tuberculosis cases in 2005.

Reproduced with permission from [6].

Figure 3. Forest plot of results from studies that evaluated the risk of mortality due to tuberculosis for smokers compared with nonsmokers.

Reproduced with permission from [12].

Cl: Confidence interval.

Figure 4. Forest plot of results from studies that examined the association between smoking and tuberculosis disease.

Reproduced with permission from [13].

The twin epidemics of tobacco & tuberculosis

According to the WHO, tobacco consumption ranks fourth among the top ten leading risks to human health [1]. There are an estimated 1.3 billion smokers in the world, of whom 1 billion are men [2]. This makes up approximately a third of the global population aged 15 years or older. An estimated 15 billion cigarettes are smoked worldwide every day [3]. Tobacco is the second major cause of death in the world. According to WHO estimates, tobacco is currently responsible for the death of one in ten adults worldwide (∼5 million deaths each year) [101]. If current smoking trends persist over time, it may cause approximately 10 million deaths each year by 2020 [2,101]. In countries such as China, Russia and India, tobacco consumption has reached epidemic proportions (Figure 1), raising significant concerns regarding its health effects in these countries [4].

An estimated 8.8 million cases of tuberculosis (TB) occur each year and almost 2 million deaths are attributed to TB [5,6]. Much of the morbidity and mortality due to TB occur in developing countries where smoking is widely prevalent (Figure 2)[6]. For example, China accounts for 1.32 million of the estimated 8.8 million TB cases that occur annually [6] and almost two-thirds of Chinese men smoke [7,8]. India accounts for 1.85 million TB cases each year [6] and more than half of rural Indian men are estimated to smoke regularly [9,10].

Association between tobacco & TB: results from recent systematic reviews

Given that both smoking and TB are major health concerns and are widely coprevalent in developing countries, it is rather surprising that the association between smoking and TB is still a matter of debate and controversy. Although several studies have evaluated the effect of smoking on TB, the association has been largely overlooked by the TB and public health communities at large [11]. What is the evidence for the association between tobacco and TB, and what can be learned from the published studies on this topic?

In the past few months, three reviews have examined this association [12–14], two of which used meta-analysis methods [12,13]. These reviews summarized the evidence on the association between active smoking and three TB outcomes: TB infection (detected using tuberculin skin testing), active TB disease and mortality due to TB.

In a review published in PLoS Medicine, Lin and colleagues reported a systematic review and meta-analysis of observational studies reporting effect estimates and 95% confidence intervals (CIs) on how tobacco smoking, passive smoke exposure and indoor air pollution was associated with TB [12]. They identified 33 papers on tobacco smoking and TB and found substantial evidence that smoking is positively associated with TB, regardless of the specific TB outcomes or the subgroups examined. Compared with people who did not smoke, smokers had an increased risk of having a positive latent TB infection, of having active TB and of dying from TB [12]. Figure 3 summarizes the data from this meta-analysis on the association between smoking and mortality due to TB. The plot shows a positive association with mortality in many studies, with studies from India demonstrating a stronger association than others. Although data from India are limited, available studies do show a significant and strong association between tobacco and TB mortality [15,16]. In a large study, Gajalakshmi and colleagues estimated that among Indian male smokers and nonsmokers together, smoking causes half of all deaths from TB [15]. A more recent study estimated that approximately a third of TB deaths may be attributable to smoking [17]. Large-scale studies are underway in India to confirm these observations [18].

In a review published in the Archives of Internal Medicine, Bates and colleagues performed a meta-analysis of 24 studies on active smoking and TB [13]. Separate analyses were performed for TB infection (six studies), TB disease (13 studies) and TB mortality (five studies). For TB infection, the pooled relative risk (RR) estimate was 1.73 (95% CI: 1.46–2.04); for TB disease, estimates ranged from 2.33 (95% CI: 1.97–2.75) to 2.66 (95% CI: 2.15–3.28). Figure 4 summarizes the results from this meta-analysis on the association between active smoking and risk of TB disease. The plot shows that most studies found a positive association between smoking and TB disease, with RR estimates in the range of approximately 1.5–3.0. Thus, this meta-analysis showed that smoking is a risk factor for TB infection and TB disease [13]. For TB mortality, the results were highly heterogeneous across studies and no pooled estimate was calculated. Studies that used verbal autopsy to determine cause of mortality showed a larger effect than studies that used routine death certificates [13]. Further studies are required to resolve this issue and also to understand the exact mechanism that results in death.

In a narrative, qualitative review, Chiang and colleagues summarized the evidence for the association between tobacco and several TB outcomes, including infection, active disease, delay in diagnosis, bacteriological conversion during TB therapy, relapse, drug resistance and mortality [14]. The authors concluded that smoking is causally associated with TB disease and recommended that patients with TB should receive counseling and assistance in stopping smoking [14].

Biological mechanisms underlying the association between tobacco & TB

The mechanisms that underlie the mechanistic link between smoking and TB are unclear, although putative pathways have been suggested [19,20]. A large body of scientific evidence suggests that macrophages, CD4 T cells, apoptosis of infected cells, IFN-γ, IL-12 and TNF-α are essential for host immunity in human TB [21]. Cigarette smoke inhibits the production of TNF-α by macrophages in the lungs via interaction with α7 nicotinic acetylcholine receptors [22] and nicotine enhances the replication of intracellular organisms and selectively downregulates the production of IL-12 and TNF-α [23]. Cigarette smoke prevents pathogen-specific expansion and activation of CD4 T cells [24] and reduces IFN-γ-producing adenoid-specific CD4 and CD8 T-cell numbers [25]. Moreover, cigarette smoke reduces macrophage apoptosis [26], which promotes the killing of virulent mycobacteria [27] and reduces murine macrophage cytotoxicity [28], and expression profiling reveals that smoking modulates a variety of key macrophage functions [29]. Other potential mechanisms by which smoking may attenuate mycobactericidal activity include oxidative stress in the lung tissues and mechanical disruption of cilia function and other clearance mechanisms in the tracheobronchial system [13]. Further research is needed to understand the exact biological mechanisms and pathways through which tobacco smoke adversely affects TB outcomes. Research is also needed to better understand the effect of environmental tobacco smoke and air pollution (indoor and outdoor) on TB outcomes.

If smoking is indeed causally associated with TB disease, the clinical manifestations of TB could well be influenced by the underlying mechanism of its action. Examination of the effect of smoking on the clinical manifestations of TB might therefore help to substantiate such a causal relationship and provide some useful hints on the possible mechanisms. Although none of the three recent reviews specifically addressed this important aspect, there is some evidence pointing to the preponderance of pulmonary TB, more severe pulmonary disease and predominant upper-zone involvement among smokers [30,31], all of which are compatible with a site-specific action of the inhaled tobacco smoke.

Although the postulated mechanisms are highly plausible, it is necessary to examine whether the observed associations could have been due to bias and confounding. For example, alcohol use and socioeconomic status could have confounded the association between tobacco and TB disease. Published studies have controlled for bias and confounding to varying degrees, and studies that did adjust for confounding found significant effects, even after adjusting for several risk factors. Even so, at least some of the observed effect could be explained by confounding, and future studies will need to control for all the key confounders and thoroughly explore the mechanisms underlying tobacco-associated TB pathogenesis.

Expert commentary & five-year view

There is now considerable evidence that tobacco smoking is associated with TB. The evidence is strong for TB disease, but less strong for TB infection and TB mortality. Although the effect is quite consistent across studies, the estimated risk is modest. However, the impact of any risk factor is determined by the magnitude of the association (i.e., RR), and the prevalence of the exposure in the population. Therefore, even if the true effect is modest, the population-attributable risk is likely to be substantial because of the widespread nature of tobacco exposure. For example, if the RR for TB disease was only 1.5, with 30% of the population exposed to tobacco smoke, the population-attributable risk percentage would be approximately 15%. In other words, 15% of the TB cases in the world each year may be attributable to tobacco exposure.

In reality, the impact is likely to be greater, because the risk estimates for TB infection, disease and mortality are not, independent [13]. If smoking increases the risk of TB infection, this will increase the proportion of smokers who acquire TB infection and are at risk of TB disease. Consequently, if smoking increases the risk of TB disease among those already infected, this will increase the number of smokers at risk of mortality due to TB. Thus, the overall impact of smoking on TB is likely to be substantial, even without considering the effect of passive smoking and indoor air pollution. While published reviews have mainly focused on active smoking, the results may potentially be applicable to passive smoking and indoor air pollution as well, although the effects are likely to be smaller with the latter exposures [12].

In resource-poor countries, there are now three interacting epidemics: smoking, TB and HIV. In addition to the interactions between smoking and TB already reviewed and the known interaction between TB and HIV (HIV increases TB mortality and susceptibility to TB, and TB accelerates the progression of HIV), smoking may also be associated with HIV infection. Smoking may be an independent risk factor for the acquisition of HIV [32] and is associated with worse outcomes in HIV-associated opportunistic infections, including pneumonia [33] and HIV-related noninfectious complications (e.g., lung cancer and chronic obstructive pulmonary disease), and is an independent risk factor for non-AIDS-related mortality [34]. Thus, the interaction between smoking and TB–HIV coinfection deserves further study. Research over the next 5 years should address this issue and also explore causal and biological mechanisms underlying tobacco-associated TB pathogenesis, with and without the HIV interaction.

In conclusion, tobacco consumption is an important risk factor for TB and TB control programs must begin to address tobacco control as a potential preventive intervention [35]. Since smoking is a modifiable and preventable risk factor, there is considerable potential to design and implement tobacco cessation and prevention programs as part of global TB control efforts. Since tobacco control will have multiple health benefits, it is likely to be a highly cost-effective intervention from a societal perspective. Lastly, in developing countries where TB is common, TB may offer a more convincing argument for investments in tobacco control rather than cancer or cardiovascular disease.

Information resources

•	Tobacco Free Initiative www.who.int/tobacco/en/
•	The Tobacco Atlas www.who.int/tobacco/statistics/tobacco_atlas/en/
•	US CDC: Smoking & Tobacco Use www.cdc.gov/tobacco/index.htm
•	International Union Against Tuberculosis and Lung Disease www.iuatld.org/
•	Stop Tuberculosis Partnership www.stoptb.org/
•	WHO Stop Tuberculosis Department www.who.int/tb/en/
•	American Thoracic Society www.thoracic.org/
•	American Lung Association www.lungusa.org

Key issues

•	Tobacco consumption ranks fourth among the top ten leading risks to human health; there are an estimated 1.3 billion smokers in the world, of whom 1 billion are men.
•	Tobacco is the second major cause of death in the world; it is currently responsible for the death of one in ten adults worldwide.
•	An estimated 9 million cases of tuberculosis (TB) occur each year and almost 2 million deaths are attributed to TB.
•	Tobacco consumption and TB are highly prevalent in many countries; therefore, the interaction between tobacco and TB epidemics is an important public health issue.
•	Three recent reviews, including two meta-analyses, have summarized the published evidence on the association between smoking and various TB outcomes.
•	These reviews show that there is now considerable evidence that tobacco smoking is associated with TB. The evidence is strong for TB disease but less strong for TB infection and TB mortality. Even if the effect is relatively modest, the population-attributable risk is likely to be substantial because of the widespread nature of tobacco exposure.
•	TB-control programs must begin to address tobacco control as a potential preventive intervention.
•	Future studies will need to determine the causal effect of tobacco on TB by controlling for all the key biases, thoroughly explore mechanisms underlying tobacco-associated TB pathogenesis and consider the effect of tobacco on TB–HIV coinfection.