INTRODUCTION
Smoking and tuberculosis (TB) are two major public health problems worldwide. The WHO estimated in 2019 that 23% of the world population smokes.1 Tobacco smoking is recognized as being associated with an increased risk of TB infection and disease.2,3 In addition, it is reported that active smoking is associated with poor TB treatment adherence, delayed smear conversion, TB treatment failure, relapse, and death during treatment.4,5 It is also reported that TB patients who smoke exhibit more radiological findings, cavitary disease, and higher sputum smear grading than nonsmokers during initial presentation.6 Furthermore, some reports suggest that TB diagnostic delay is associated with smoking.7 The current approach to TB control focuses on case detection and treatment. Recent studies suggest that to achieve TB control, it is necessary to reduce risk factors that contribute to the occurrence of TB infection and/or disease.8 The prevalence of tobacco use in Gabon reported by the WHO is 22.4%.9 Smoking prevalence among TB patients could be higher than that in the general population. The prevalence of smoking among TB patients and the associated adverse effects are under-investigated in African countries that have a high incidence of TB, such as Gabon. To our knowledge, no previous studies have been performed to assess the prevalence and the burden of cigarette smoking among TB patients in Gabon. Gaining insights into the double burden of TB and tobacco use could help to establish the need for cessation services as a component of TB program. This study aimed to determine the proportion of tobacco smokers among TB patients and to examine the disease characteristics of smoking pulmonary TB patients in Lambaréné and the surrounding villages in Moyen-Ogooue Province.
METHODS
Study design.
A cross-sectional study conducted among TB patients enrolled between March 2016 and April 2019 at the Centre de Recherches Médicales de Lambaréné (CERMEL).
Study site and population.
Patients were recruited at five different sites in Lambaréné: 1) in- and outpatient departments of the Albert Schweitzer Hospital, 2) in- and outpatient departments of Georges Rawiri Regional Hospital, 3) the local outpatient HIV clinic (Centre de Traitement Ambulatoire), 4) the local outpatient TB clinic (Base d’épidémiologie), and 5) CERMEL, thus representing all “ports of entry” for TB patients into the local healthcare system. Centre de Recherches Médicales de Lambaréné is the TB national reference laboratory, and all presumptive TB patients in Moyen-Ogooue Province and beyond are referred to the CERMEL. All patients older than 15 years visiting the CERMEL, with signs and symptoms compatible with TB, were invited to participate in the study. Those unable to answer our questionnaire and those with extrapulmonary disease were excluded.
Data collection, definitions, and procedures.
Demographic and clinical data were collected by study investigators. Sociodemographic data on education were also collected during the visit, and self-reported cigarette smoking status was collected using the questionnaire. For the purposes of this study, the definition of a current smoker was expanded to include former smokers who had reported smoking cessation within the 2-month period before TB diagnosis. This expansion was to minimize misclassification of patients who temporarily quit smoking at the onset of symptoms. The duration of the illness was based on patient self-declaration of the onset of the first sign of TB to the date he/she presented for consultation. Patients who consumed alcohol were invited to provide information of the kind of alcoholic beverage and the approximate quantity consumed per week. The quantity of alcohol consumed per ml was converted to mg using standard drink conversion.10 Men who consumed regularly more than 60 g/day of alcohol and women who consumed more than 40 g/day were considered as high-risk alcohol users.10 Samples for mycobacterial investigation were analyzed by auramine fluorochrome staining or GeneXpert assay according to the national guideline. Bacteriological confirmation of TB was based on the examination of two spot sputum samples.
Patients were considered to be smokers if they reported having smoked ≥ 1 cigarette(s)/day continuously. The burden of cigarette smoking was classified according to pack-years, calculated by multiplying the number of cigarettes smoked per day by the number of years the person has smoked divided by 20. Patient-related delay in diagnosis was defined as “yes” if time of consultation was more than 21 days (>) and “no” if less (≤) than 21 days. Subsequent TB signs and symptoms were systematically assessed for each participant: cough, fever, dyspnea, chest pain, hemoptysis, night sweats, weight loss, and asthenia. A symptom score (0–8) was calculated on the basis of presence of symptoms or sign (1 point for each). The symptom score was categorized as “yes” if the patient presented or reported more than four (> 4) signs and/or symptoms or no if four or less. The WHO scale was used to categorize the severity of sputum mycobacterial load.11
Data management and statistical analysis.
Study data were collected and managed using the research electronic data capturing tool.12,13 To ensure the quality of data, double entry was performed by trained and independent data clerks. A clean database was extracted, and the analysis was conducted using R software version 3.6.1 (R Foundation for Statistical Computing, Vienna, Austria). Proportions were used to describe qualitative variables and mean or median for quantitative variables. The characteristics of smokers versus those of nonsmokers were compared first, using Pearson’s χ2 or Fisher’s exact test. This analysis was followed by multivariate logistic regression. Factors known to be associated with smoking from the literature review and those identified during the univariate analysis were included in the multivariate model. We conducted manual backward stepwise elimination to choose the final model. Because gender was found to be associated with HIV infection, education level, and alcohol consumption, the interaction effect of gender variables was assessed in the model by adding the interaction term (gender: HIV infection, gender: high-risk alcohol consumption, and gender: education level). Variables with an adjusted ratio that had a P-value less than 0.05 at 95% CI were considered significant.
Ethics approval and consent to participate.
The study was approved by the Institutional Ethics Committee of the CERMEL. Written informed consent was obtained from all participants before study enrollment. The participants between the age of 16 and 18 years provided written assent in addition to written consent from their parent or legal guardian to participate in the study.
RESULTS
A total of 295 TB patients were included, with a mean age of 31 years (±13). Of those 295 patients, 60% (177/295) had obtained secondary level of education, 30% (86/295) previously shared a household with a TB patient, and 80% (234/295) had never been treated for TB. High-risk alcohol consumption was reported in 45% (133/295) of individuals. The prevalence of HIV and TB coinfection was 31% (92/295).
Distribution of smoking status and associated factors.
The prevalence of smoking was 30% (89/295), and the median pack-years of smoking was 8 (IQR: 3–17). As presented in Table 1, bivariate analysis to detect factors associated with smoking indicated that smokers were predominantly male, and they consumed alcohol more often than nonsmokers. HIV infection prevalence was higher among nonsmokers than smokers. Smokers presented with more signs and or symptoms suggestive, or in line with, TB, and exhibited delays in time to diagnosis as compared with nonsmokers.
Characteristics of patients and factors related to smoking at the time of TB diagnosis at bivariate level
| Variable | Non-cigarette smokers, n (row %) 206 (70) | Cigarette smokers, n (row%) 89 (30) | Total (N = 295) | P-value |
|---|---|---|---|---|
| Age-group (years) | 0.204 | |||
| 16–24 | 76 (76) | 24 (24) | 100 | |
| 25–34 | 58 (72) | 23 (28) | 81 | |
| 35–44 | 36 (62) | 22 (38) | 58 | |
| 45–54 | 23 (66) | 12 (34) | 35 | |
| ≥ 55 | 13 (62) | 8 (38) | 21 | |
| Gender | < 0.001 | |||
| Female | 126 (95) | 7 (5) | 133 | |
| Male | 80 (49) | 82 (51) | 162 | |
| Education level | 0.325 | |||
| None | 9 (75) | 3 (25) | 12 | |
| Primary | 49 (63) | 29 (37) | 78 | |
| Secondary | 133 (71) | 54 (29) | 187 | |
| University | 15 (83) | 3 (17) | 18 | |
| Household contact of TB | 0.931 | |||
| No | 139 (70) | 60 (30) | 201 | |
| Yes | 64 (67) | 32 (33) | 94 | |
| History of TB treatment | 0.783 | |||
| No | 174 (69) | 77 (31) | 251 | |
| Yes | 32 (73) | 12 (27) | 44 | |
| High-risk alcohol consumption | < 0.001 | |||
| No | 143 (89) | 18 (11) | 161 | |
| Yes | 63 (47) | 71 (53) | 134 | |
| More than four TB symptoms | 0.391 | |||
| No | 70 (74) | 25 (26) | 95 | |
| Yes | 136 (68) | 64 (32) | 200 | |
| Patient-related diagnostic delay | < 0.001 | |||
| No | 158 (84) | 31 (16) | 189 | |
| Yes | 48 (45) | 58 (55) | 106 | |
| HIV coinfection | 0.047 | |||
| No | 134 (66) | 69 (34) | 203 | |
| Yes | 72 (78) | 20 (22) | 92 | |
| Sputum grading | 0.020 | |||
| 3+ | 100 (64) | 57 (36) | 157 | |
| 2+ | 70 (82) | 15 (18) | 85 | |
| 1+ | 27 (68) | 13 (32) | 40 | |
| Rare | 9 (69) | 4 (31) | 13 | |
TB = tuberculosis.
From the multivariable analysis (Table 2), the following factors were independently associated with smoking: high-risk alcohol consumption (odds ratio [OR] = 6.26; 95% CI = 2.97–13.83), patient-related delayed diagnosis (adjusted OR [AOR] = 8.18; 95% CI = 3.67–19.56), higher (3+) sputum mycobacterial loads (AOR = 3.18; 95% CI = 1.33–8.11), and more than four TB signs and symptoms (AOR = 2.74; 95% CI = 1.18–6.73).
Characteristics of smoking in TB patients
| Variable | Bivariable | Multivariable | |||
|---|---|---|---|---|---|
| Odds ratio (95% CI) | P-value | Adjusted odds ratio (95% CI) | P-value | ||
| Education level | Secondary | Reference | Reference | ||
| None | 0.82 (0.18–2.87) | 0.774 | 0.92 (0.11–6.26) | 0.932 | |
| Primary | 1.46 (0.83–2.54) | 0.185 | 1.58 (0.69–3.64) | 0.282 | |
| University | 0.49 (0.11–1.57) | 0.278 | 0.46 (0.08–2.19) | 0.350 | |
| High-risk alcohol consumption | No | Reference | Reference | ||
| Yes | 8.95 (5.03–16.65) | < 0.001 | 6.26 (2.97–13.83) | < 0.001 | |
| Diagnostic delay | No | Reference | Reference | ||
| Yes | 6.16 (3.61–10.71) | < 0.001 | 8.18 (3.67–19.56) | < 0.001 | |
| Sputum grading | 2+ | Reference | Reference | ||
| 3+ | 2.66 (1.42–5.22) | 0.003 | 3.18 (1.33–8.11) | 0.012 | |
| 1+ | 2.25 (0.94–5.36) | 0.067 | 3.19 (0.96–10.84) | 0.060 | |
| rare | 2.07 (0.51–7.34) | 0.273 | 3.56 (0.53–22.28) | 0.178 | |
| HIV coinfection | No | Reference | Reference | ||
| Yes | 0.54 (0.30–0.94) | 0.035 | 1.19 (0.50–2.83) | 0.699 | |
| More than four signs and symptoms | No | Reference | Reference | ||
| Yes | 1.32 (0.77–2.30) | 0.321 | 2.74 (1.18–6.73) | 0.022 | |
OR = odds ratio; TB = tuberculosis. Factors related to smoking at the time of TB diagnosis at multivariate logistic regression level. Adjusted for age, gender, and history of TB treatment.
DISCUSSION
The results show a high prevalence of smoking among TB patients diagnosed in Lambaréné between 2016 and 2019. Smoking was associated with alcohol consumption, more TB symptom, higher sputum mycobacterial load, and consultation delay.
Similar to studies conducted in Senegal and Iran,14,15 we found a 30% prevalence of smokers among the included patients. This prevalence is higher than that reported in the general Gabonese population, especially in men.9 However, the smoking prevalence observed in our cohort was lower than that in other countries, with approximately 40% (Malaysia) and 50% (South Africa) having been reported.16,17 This difference could be due to a background of tobacco-smoking habit from each country. The smoking prevalence in Malaysia’s and South Africa’s general population is higher than that in Gabon.18 On the other hand, one explanation could be the choice of inclusion criteria in our study. As mentioned, we only included bacteriologically confirmed pulmonary TB patients, and daily smoking was used to categorize a smoker.
Effective TB control requires early diagnosis and prompt treatment. In our study, smoking was associated with consultation delay. Similar results were reported from Spain and Morocco.19,20 Although the most common symptom of TB is cough, it may be difficult to perceive coughing as a hint to an underlying condition such as TB in smoking patients who are prone to coughing anyway. As in studies from Tunisia and Senegal,14,21 smokers in our study experienced more signs and symptoms. Smoking severely impedes the pulmonary immune system’s function mainly via the inhibition of innate immune responses and pulmonary T-cell recruitment.22,23 This could explain why smoking exposes one to more severe TB.
Our results showed higher mycobacterial loads among smoking TB patients at the time of diagnosis. These findings are similar to those reported by other studies.6,24 Several studies in animal models and humans have shown that exposure to tobacco smoke causes immunological changes, acting on alveolar macrophages by decreasing the production of tumor necrosis factor alpha (TNF-α), interferon gamma (IFN-γ), and mucociliary clearance, thus promoting disease progression.22,23,25 The increased severity of TB due to cigarette smoking is also seen on chest radiography, often exhibiting advanced lung lesions and cavitations in smokers, which could explain the capacity to produce more mycobacteria in the sputum by smoking patients.26,27
Another finding in our study is the high prevalence of HIV coinfection and alcohol consumption. The HIV and TB coinfection rate was 31% in our study, and this is similar to a previous report on adults in Lambaréné.28 This rate is higher than the 18% estimated by the WHO in Gabon TB country profile in 201829; a difference that could be attributed to the number of TB patients tested for HIV in the WHO country report. Indeed, only 48% of TB cases were tested for HIV according to the data reported by the WHO. This underlines the importance to improve the screening of HIV in all TB patients in Gabon.
More than 40% of TB patients in our cohort were high-risk alcohol users, and alcohol consumption was associated with smoking. These two factors are classically associated with TB and have been reported in many studies.19,26,30 However, the prevalence of high-risk alcohol consumption in our study is higher than that reported from Tanzania and Nigeria.30,31 Compared with other regions in Africa, Gabon had a higher general population alcohol consumption level and pattern.32 These alcohol consumption levels are indeed relevant to a number of countries with high TB burden.31 To that end, in Gabon, interventions can be implemented to reduce the harm due to alcohol consumption. Such interventions include implementing regulations of the environment in which alcohol is marketed, as well as individually directed interventions for those with at-risk levels of alcohol consumption.
We covered all TB diagnosis and treatment sites in the Moyen-Ogooue region. All pulmonary TB patients diagnosed during the study period were given the chance to be included. This is the first study in Gabon to investigate the burden of smoking among TB patients. Our study shows for the first time that in Gabon, the epidemiological profile of smoking TB pulmonary patients is characterized by the delay of consultation and the higher mycobacterial load. Therefore, TB smoking patients going undiagnosed could increase the community transmission of TB. A more structured study would be necessary to determine the impact of smoking on the community transmission of TB.
This study has some limitations such as a potential misclassification of exposure. This could be because of the fact that the questionnaire used contained no information on other forms of tobacco intake. More so, we did not consider former smoking, and, most importantly, we had to confine our study to a single center. The restriction to a single center was due to limited resources and infrastructure not permitting us to conduct such research across the country at this stage. The logical next step is to perform such a study in Gabon and beyond across the Central African region in a multicenter prospective study, to understand the bigger epidemiological picture. Furthermore, the self-report approach was used to record tobacco consumption, and this can raise the issue of the credibility of participant answers. However, a study conducted by Brunet et al.17 in South Africa showed that self-reported tobacco consumption is an accurate measure of smoking status. Because of the cross-sectional design, the study showed only association between smoking and TB from the epidemiological point of view, and cannot show causality. That notwithstanding, we provide valuable information illustrating the prevalence of cigarette smoking among TB patients in Lambaréné and its association with TB manifestation.
CONCLUSION
In summary, our findings indicate a high prevalence of smoking among TB patients in Lambaréné. Tuberculosis patients who smoke demonstrate a greater delay in TB diagnosis, more TB-associated signs and symptoms, and a higher sputum mycobacterial load. National TB control programs in an endemic setting such as Gabon should include investigations of smoking behavior and offer smoking cessation interventions. Community-based activities with intention to screen TB among smokers should be envisaged, to facilitate early TB diagnosis among such populations and reduce TB transmission.
ACKNOWLEDGMENTS
We are grateful to the field-workers of the tuberculosis research team at the CERMEL for their hard work. We thank Matthew McCall, Bertrand Lell, and Francis Mhimbira, the Ifakara Health Institute, Bagamoyo, Tanzania, for their thoughtful comments and review. We also wish to thank the TB nurses at the health facilities for their support.
REFERENCES
- 1.↑
World Health Organization, 2019. Global Report on Trends in Prevalence of Tobacco Use 2000–2025, 3rd edition. Geneva, Switzerland: WHO. Available at: http://apps.who.int/iris/. Accessed February 29, 2020.
- 2.↑
Bates MN, Khalakdina A, Pai M, Chang L, Lessa F, Smith KR, 2007. Risk of tuberculosis from exposure to tobacco smoke: a systematic review and meta-analysis. Arch Intern Med 167: 335–342.
- 3.↑
Lugos MD, Hill PC, Hammond AS, Kendall L, Donkor S, Brookes RH, Owiafe PK, Ota MOC, Adetifa IMO, 2017. Mycobacterium tuberculosis infection in close childhood contacts of adults with pulmonary tuberculosis is increased by secondhand exposure to tobacco. Am J Trop Med Hyg 97: 429–432.
- 4.↑
Tachfouti N, Nejjari C, Benjelloun MC, Berraho M, Elfakir S, El Rhazi K, Slama K, 2011. Association between smoking status, other factors and tuberculosis treatment failure in Morocco. Int J Tuberc Lung Dis 15: 838–843.
- 5.↑
Maciel EL et al. 2013. Smoking and 2-month culture conversion during anti-tuberculosis treatment. Int J Tuberc Lung Dis 17: 225–228.
- 6.↑
Rathee D, Arora P, Meena M, Sarin R, Chakraborty P, Jaiswal A, Goyal M, 2016. Comparative study of clinico-bacterio-radiological profile and treatment outcome of smokers and nonsmokers suffering from pulmonary tuberculosis. Lung India 33: 507–511.
- 7.↑
Das S, Basu M, Mandal A, Roy N, Chatterjee S, Dasgupta A, 2017. Prevalence and determinants of delay in diagnosis of pulmonary tuberculosis in Darjeeling district of West Bengal. J Fam Med Prim Care 6: 627–635.
- 8.↑
Rangaka MX, Cavalcante SC, Marais BJ, Thim S, Martinson NA, Swaminathan S, Chaisson RE, 2015. Controlling the seedbeds of tuberculosis: diagnosis and treatment of tuberculosis infection. Lancet Lond Engl 386: 2344–2353.
- 9.↑
World Health Organization, 2019. Report on the Global Tobacco Epidemic, Gabon Country Profile. Geneva, Switzerland: WHO. Avaialble at: https://www.who.int/gho/countries/gab/country_profiles/en/. Accessed February 29, 2020.
- 10.↑
World Health Organization, 2000. International Guide for Monitoring, Alcohol Consumption and Related Harm. Geneva, Switzerland: WHO. Available at: https://apps.who.int/iris/handle/10665/66529/. Accessed March 29, 2019.
- 11.↑
World Health Organization, 2007. TB Diagnostics and Laboratory Strengthening. Definition of a New Sputum Smear-Positive TB Case. Geneva, Switzerland: WHO. Available at: https://www.who.int/tb/laboratory/policy_sputum_smearpositive_tb_case/en/. Accessed April 29, 2020.
- 12.↑
Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG, 2009. Research electronic data capture (REDCap)–a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 42: 377–381.
- 13.↑
Harris PA, Taylor R, Minor BL, Elliott V, Fernandez M, O’Neal L, McLeod L, Delacqua G, Delacqua F, 2019. The REDCap consortium: building an international community of software platform partners. J Biomed Inform 95: 103208.
- 14.↑
Kombila UD, Mbaye FBR, Dia Kane Y, Ka W, Toure Badiane NO, 2018. Particularités cliniques et radiologiques de la tuberculose pulmonaire chez le tabagique. Rev Mal Respir 35: 538–545.
- 15.↑
Aryanpur M, Masjedi MR, Hosseini M, Mortaz E, Tabarsi P, Soori H, Emami H, Heydari G, Baikpour M, 2016. Cigarette smoking in patients newly diagnosed with pulmonary tuberculosis in Iran. Int J Tuberc Lung Dis 5: 679–684.
- 16.↑
Awaisu A, Nik Mohamed MH, Abd Aziz N, Syed Sulaiman SA, Mohamad Noordin N, Muttalif AR, Ahmad Mahayiddin A, 2010. Tobacco use prevalence, knowledge, and attitudes among newly diagnosed tuberculosis patients in Penang state and Wilayah Persekutuan Kuala Lumpur, Malaysia. Tob Induc Dis 8: 3.
- 17.↑
Brunet L et al. 2011. High prevalence of smoking among patients with suspected tuberculosis in South Africa. Eur Respir J 38: 139–146.
- 18.↑
Global Health Observatory Data Repository, 2015. Prevalence - Most Recent Adult Survey Data by Country. Available at: http://apps.who.int/gho/data/node.main.TOB1249?lang=en/. Accessed February 29, 2020.
- 19.↑
Jiménez-Fuentes MÁ et al. TRWG, 2016. Factors associated with smoking among tuberculosis patients in Spain. BMC Infect Dis 16: 486.
- 20.↑
Janah H, Souhi H, Kouissmi H, Marc K, Zahraoui R, Benamor J, Soualhi M, Bourkadi JE, 2014. La tuberculose pulmonaire et le tabac: à propos de 100 cas [article in French]. Pan Afr Med J 19: 202.
- 21.↑
Fekih L, Boussoffara L, Abdelghaffar H, Hassene H, Fenniche S, Belhabib D, Zouari B, Megdiche ML, 2010. Effet du tabagisme sur la tuberculose pulmonaire. Rev Med Liège 65: 152–155.
- 22.↑
Shaler CR, Horvath CN, McCormick S, Jeyanathan M, Khera A, Zganiacz A, Kasinska J, Stampfli MR, Xing Z, 2013. Continuous and discontinuous cigarette smoke exposure differentially affects protective Th1 immunity against pulmonary tuberculosis. PLoS One 8: e59185.
- 23.↑
Aryanpur M, Mortaz E, Reza Masjedi M, Tabarsi P, 2016. Reduced phagocytic capacity of blood monocyte/macrophages in tuberculosis patients is further reduced by smoking. Iran J Allergy Asthma Immunol 3: 174–182.
- 24.↑
Prasad R, Suryakant, Garg R, Singhal S, Dawar R, Agarwal G, 2009. A case-control study of tobacco smoking and tuberculosis in India. Ann Thorac Med 4: 208–210.
- 25.↑
Feng Y et al. 2011. Exposure to cigarette smoke inhibits the pulmonary T-cell response to influenza virus and Mycobacterium tuberculosis. Infect Immun 79: 229–237.
- 26.↑
Altet-Gómez MN, Alcaide J, Godoy P, Romero MA, 2005. Clinical and epidemiological aspects of smoking and tuberculosis: a study of 13,038 cases. Int J Tuberc Lung Dis 9: 430–436.
- 27.↑
Leung CC et al. 2015. Smoking adversely affects treatment response, outcome and relapse in tuberculosis. Eur Respir J 45: 738–745.
- 28.↑
Niemann S et al. 2016. Tuberculosis treatment outcome and drug resistance in Lambaréné, Gabon: a prospective cohort study. Am J Trop Med Hyg 95: 472–480.
- 29.↑
World Health Organization, 2018. Tuberculosis Country Profiles. Geneva, Switzerland: WHO. Available at: http://www.who.int/tb/country/data/profiles/en/. Accessed June 22, 2018.
- 30.↑
Mwiru RS, Nagu TJ, Kaduri P, Fawzi W, Mugusi F, 2018. Prevalence and factors associated with alcohol drinking among HIV and tuberculosis co-infected patients in Dar es Salaam, Tanzania. AIDS Care 30: 173–177.
- 31.↑
Imtiaz S, Shield KD, Roerecke M, Samokhvalov AV, Lönnroth K, Rehm J, 2017. Alcohol consumption as a risk factor for tuberculosis: meta-analyses and burden of disease. Eur Respir J 50: 1700216.
- 32.↑
World Health Organization, 2018. Alcohol Consumption: Levels and Patterns; Data by Country Gabon: Country Profiles. Geneva, Switzerland: WHO. Available at: https://www.who.int/gho/countries/gab/country_profiles/en/. Accessed May 4, 2020.