Association Between Passive Smoking and Infection With Mycobacterium tuberculosis in Children
OBJECTIVE. Tuberculosis and smoking are both significant public health problems. The association between passive smoking and Mycobacterium tuberculosis infection is not well documented. The objective of this study was to examine the influence of passive smoking on M tuberculosis infection in children.
METHODS. A community survey was conducted in 15% of addresses in 2 adjacent low-income suburbs in Cape Town, South Africa. All children (<15 years of age) and their adult household members residing at these addresses were included in the study. Children underwent tuberculin skin testing. An induration of ≥10 mm was considered to define M tuberculosis infection. Passive smoking was defined as living in the household with at least 1 adult who smoked for at least 1 year. Random-effects logistic regression analysis was performed, and odds ratios were adjusted for age, presence of a patient with tuberculosis in the household, average household income, and clustering at the household level.
RESULTS. Of 1344 children, 432 (32%) had a positive tuberculin skin test. Passive smoking was significantly associated with M tuberculosis infection in the unadjusted analyses but not in the adjusted analyses. In the 172 households with a patient with tuberculosis, passive smoking was significantly associated with a positive tuberculin skin test (but not in the 492 households without a patient with tuberculosis.
CONCLUSIONS. Passive smoking is associated with M tuberculosis infection in children living in a household with a patient with tuberculosis. More studies are needed to confirm this observation, but the possible association is a cause of great concern, considering the high prevalence of smoking and tuberculosis in most developing countries.
Recently, we reported that cigarette smoking is associated with Mycobacterium tuberculosis infection in adults,1 a finding confirmed in several other studies.2–6 Few studies have investigated the association between environmental tobacco smoke exposure (passive smoking) and M tuberculosis infection.
Passive smoking may have particularly harmful effects in children compared with adults because children's respiratory and immune systems are not fully developed.7–10 In addition, children spend more time at home and are, therefore, likely to experience more intense and prolonged smoke exposure if adult household members smoke. Passive smoking increases a child's risk to develop asthma, bronchitis, pneumonia, otitis media, and sudden infant death syndrome and to undergo procedures such as tympanostomies, tonsillectomies, and adenoidectomies.7,11–13 At present, there is insufficient information to establish whether there is an association between passive smoking and M tuberculosis infection in children. We conducted a cross-sectional community survey allowing an assessment of the association between M tuberculosis infection and passive smoking in children. We hypothesized that passive smoking was associated with M tuberculosis infection in children.
Study Area and Study Population
The study was performed in 2 adjacent urban low-income to middle-income communities (Ravensmead and Uitsig) in Cape Town, South Africa, with a population of 36334 in 2001.14 The tuberculosis notification rate was 341 per 100000 for new smear-positive tuberculosis and 841 per 100000 for total adult tuberculosis (pulmonary and extra-pulmonary tuberculosis) in 2002.14,15 Childhood tuberculosis cases comprise ∼18% of the total tuberculosis case load.16 The prevalence of human immunodeficiency virus infection in antenatal women in the district in which Ravensmead and Uitsig are located increased from 7.9% in 2001 to 15.1% in 2004.17
This study was part of a comprehensive lung-health survey and was conducted between July 1 and December 15, 2002. Eight hundred thirty-seven (15%) residential addresses were selected randomly from a Geographical Information System containing the exact location of all 5592 addresses in the study area. If the head of the address did not give consent, the adjacent address was selected, first to the right and then to the left. All residents living at the selected addresses were eligible for participation in the study. Signed informed consent was obtained from all adult (≥15 years of age) participants and from the parents or legal guardians of the children (<15 years of age). All children received an intradermal tuberculin skin test (TST) of 2 TU (0.1 mL) purified protein derivative RT 23 on the ventral aspect of the left forearm.18 The induration was measured after 48 to 120 hours by trained nurses. A reaction of ≥10 mm was considered to represent M tuberculosis infection. All adult household members completed a questionnaire containing questions on smoking behavior, income, and past diagnosis of tuberculosis.
Passive smoking was defined as living in the household with at least 1 adult who smoked for at least 1 year. Average income of the adult household members was calculated as the total income of all household members together, divided by the number of adult household members. Children who were active smokers and children for whom no data on passive smoking was available were excluded from the analysis. A random-effects logistic regression was used to model M tuberculosis infection on passive smoking. The clustering at the household level was accounted for by including household as the random effect in the model. Unadjusted and adjusted odds ratios (ORs) and their 95% confidence intervals (CIs) were estimated. Confounding factors that were taken into consideration were age of the child, average income of the adult household members, and the presence of a patient with tuberculosis in the household. We also used a random-effects logistic regression model to determine the association between M tuberculosis infection in children and the number of smokers in the household and the average number of cigarettes smoked per day in the household. Separate analyses were conducted for children living in households where at least 1 of the adults ever had tuberculosis and for children living in households where none of the adults ever had tuberculosis. The TST distributions of children who had the induration measured after 48 to 72 hours and those who had it measured after 96 to 120 hours were compared by using the Mann-Whitney test.
Ethics approval was obtained from Stellenbosch University and the University of Cape Town. The study was conducted in accordance with the ethical standards of the World Medical Association's Helsinki Declaration.
A total of 1811 children were eligible for the study. For 1593 (88%) children, consent was obtained from the parents or legal caregivers. Three children were excluded because no data on passive smoking was available, and 220 children were excluded because the TST was not done or read within 48 to 120 hours. There was no difference in TST size distribution between those who had the indurations read between 48 and 72 hours and those who had it read up to 120 hours after administration (P = .4242). Therefore, all children with TST read between 48 and 120 hours were used in the analyses. Twenty-six children were excluded from the analyses because they were active smokers. All these children were also passive smokers. A total of 1344 (84%) of 1593 children from 664 households were included in the analyses. The children who were included in the analyses did not differ in gender or age from the children who were not included in the analyses or those for whom no consent was obtained (Table 1).
Four hundred thirty-two children (32%) had a TST induration of ≥10 mm. Older children were more often infected than younger children (Table 2). A total of 1170 (87%) children were passive smokers in the household, and these children more often had a positive TST (34%) than children who were not passive smokers (21%) (unadjusted OR: 1.89; 95% CI: 1.24–2.86). However, the significance of this finding was not maintained after adjustment for the age of the child, the average income of all adults in the household, and the presence of a patient with tuberculosis in the household (OR: 1.35; 95% CI: 0.86–2.12) (Table 2). There was a significant relationship between M tuberculosis infection in children and both the number of smokers in the household (OR for each additional smoker: 1.12; 95% CI: 1.03–1.20) and the number of cigarettes smoked per day in the household (OR for each additional cigarette: 1.009; 95% CI: 1.003–1.014). However, this was no longer significant after adjustment for the variables age of the child, average income of the adults in the household, and the presence of a patient with tuberculosis in the household (adjusted OR for each additional smoker: 1.03; 95% CI: 0.94–1.13) and the number of cigarettes smoked per day in the household (adjusted OR for each additional cigarette: 1.005; 95% CI: 0.999–1.011).
Passive smoking was significantly associated with a positive TST in the 172 households with a patient with tuberculosis (adjusted OR: 4.60; 95% CI: 1.29–16.45) but not in the 492 households without a patient with tuberculosis (adjusted OR: 1.20; 95% CI: 0.75–1.93) (Tables 3 and 4). The OR for the interaction term of presence of smoker and presence of a patient with tuberculosis in the household was 4.09 (95% CI: 0.95–17.54; P = .058). In the analysis stratified by the presence of a patient with tuberculosis in the household, we could not detect a quantitative effect for the number of smokers in the household or for the average number of cigarettes smoked per day in the households.
Passive smoking was not in general associated with M tuberculosis infection. However, a stratified analysis by households with a patient with tuberculosis demonstrated a significant association between passive smoking and M tuberculosis infection in children. This finding suggests that passive smoking may increase the risk to acquire M tuberculosis infection, given household exposure to an adult index case. The association is a cause of great concern, because in many developing countries with a high burden of tuberculosis, the prevalence of smoking is rapidly increasing, especially among women.19,20
In 1995, 58% of men and 59% of women from the ethnic group that we studied were smokers.21 In our study, 87% of the children experienced passive smoking at home, which is higher than was reported by Reddy et al,21 who found that 67% of children had at least 1 household member who smoked. The difference may be explained by the fact that we included current smokers as well as ex-smokers. However, we think that few smokers in this community stop smoking. The high proportion of women who smoke is of particular concern because they expose their children to tobacco smoke.
Passive smoking might affect the immune system of the child, thus increasing the risk of getting infected. Exposure to tobacco smoke leads to alterations in the epithelial function, such as reduced mucociliary activity, decreased clearance of inhaled substances, and abnormal vascular and epithelial permeability.3,22–25 Furthermore, smoking can change the amount, consistency, and permeability of the mucous.3,26,27 The number of alveolar macrophages increases because of tobacco smoke exposure, but their ability to phagocytose and/or kill bacteria decreases.28–30 As a result, the innate immunity of the lung is compromised, and it is easier for infectious agents to reach the alveolar tissue. T cells are highly susceptible to cigarette smoke, which could impair their cytotoxic capacity to fight infections.26 Furthermore, cigarette smoking is associated with reductions in serum immunoglobulins, T-lymphocyte helper/suppressor cell ratios, and natural killer cytotoxic activity,31–33 which may result in a decreased immune response of the body to M tuberculosis infection.
Few other studies have looked at the association between passive smoking and M tuberculosis infection. Kuemmerer and Comstock34 noted that tuberculin reactions were larger in children where both parents smoked. A study in India reported a relative risk of 2.68 (95% CI: 1.52–4.71) for M tuberculosis infection in children under the age of 5 years who lived with an adult patient with tuberculosis and who were exposed to environmental tobacco smoke compared with children who were not exposed to environmental tobacco smoke.35 Multivariate analysis included sputum positivity of the index case and malnutrition of the child, but no adjustments were made for socioeconomic status, which may have biased the results. Current smoking status of the tuberculosis index case was a significant univariate risk factor for tuberculosis infection in adult and child contacts in a study by Gerald et al.36 However, they excluded smoking from their final model because of small numbers.
Our study has some limitations. In the analysis on the subgroup of children who were exposed to a patient with tuberculosis in the household, the number of children without a smoker in the household was very small. Although the OR was large, suggesting that there is an association between passive smoking and M tuberculosis infection, the small sample size is reflected in a wide CI and, therefore, we must interpret the findings with caution. Smoking status was based on self-reporting and might have been influenced by reporting bias. We did not measure biological markers such as urine cotinine levels to confirm exposure to tobacco smoke. Another possible bias is that smoking households may differ from nonsmoking households in aspects other than the smoking of tobacco. We tried to control for socioeconomic status by adjusting for average income. Education and crowding were not significantly associated with M tuberculosis infection and were, therefore, not included in the analyses. We could not adjust for nutritional status. We did control for the presence of a patient with tuberculosis in the household because it is an important confounder because smokers are more likely to have tuberculosis than nonsmokers.37–39 Cough in heavy smokers can also lead to delay in the diagnosis of tuberculosis, resulting in more time to transmit the disease. Patients with tuberculosis who smoke might also be more infectious because of more coughing. The TST as a measure of tuberculosis infection has its limitations. Small indurations may be because of exposure to environmental mycobacteria and/or cross-reactivity because of BCG. However, we do not think environmental mycobacteria are highly prevalent in the study area as we found an unimodal distribution of TST responses with hardly any small indurations. Furthermore, tuberculin reactivity after BCG vaccination is primarily affected by age at vaccination. If the vaccine is given in infancy, as is the case in our communities, tuberculin reactions wane rapidly in all individuals.40,41 Therefore, we think that the cutoff point of 10 mm used to indicate M tuberculosis infection is justified in children from this population.
Although passive smoking, in general, was not associated with M tuberculosis infection, passive smoking was significantly associated with M tuberculosis infection in the stratified analysis by households with a patient with tuberculosis. More studies are needed to confirm this observation, but the possible association is a cause of great concern considering the high prevalence of smoking and tuberculosis in most developing countries.
We thank Dr Ivan Toms, Director of Health of the City of Cape Town, for the permission to conduct epidemiologic research in the communities of Ravensmead and Uitsig. We also thank Katherine Lawrence for data management.
- Accepted November 14, 2006.
- Address correspondence to Saskia den Boon, MSc, KNCV Tuberculosis Foundation, PO Box 146, 2501 CC, The Hague, Netherlands. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
- ↵den Boon S, Van Lill SWP, Borgdorff MW, et al. Association between smoking and tuberculosis infection: a population survey in a high tuberculosis incidence area. Thorax.2005;60 :555– 557
- Nisar M, Williams CS, Ashby D, Davies PD. Tuberculin testing in residential homes for the elderly. Thorax.1993;48 :1257– 1260
- ↵Cook DG, Strachan DP. Health effects of passive smoking-10: Summary of effects of parental smoking on the respiratory health of children and implications for research. Thorax.1999;54 :357– 366
- Pedreira FA, Guandolo VL, Feroli EJ, Mella GW, Weiss IP. Involuntary smoking and incidence of respiratory illness during the first year of life. Pediatrics.1985;75 :594– 597
- ↵DiFranza JR, Lew RA. Morbidity and mortality in children associated with the use of tobacco products by other people. Pediatrics.1996;97 :560– 568
- Dybing E, Sanner T. Passive smoking, sudden infant death syndrome (SIDS) and childhood infections. Hum Exp Toxicol.1999;18 :202– 205
- ↵Statistical Support and Informatics. Statistics South Africa: Western Cape. Census 2001. Camberwell Victoria, Australia: Space Time Research Pty Ltd; 2004
- ↵Western Cape Tuberculosis Program. Health Facility Report for Uitsig Clinic and Ravensmead Clinic 2002. Cape Town, South Africa: Western Cape Department of Health; 2003
- ↵Davies PD, Yew WW, Ganguly D, et al. Smoking and tuberculosis: the epidemiological association and immunopathogenesis. Trans R Soc Trop Med Hyg.2006;100 :291– 298
- Gupta KB, Gupta R. Association between smoking and tuberculosis. Indian J Tuberc.2003;50 :5– 7
- ↵Kum-Nji P, Meloy L, Herrod HG. Environmental tobacco smoke exposure: prevalence and mechanisms of causation of infections in children. Pediatrics.2006;117 :1745– 1754
- Aoshiba K, Tamaoki J, Nagai A. Acute cigarette smoke exposure induces apoptosis of alveolar macrophages. Am J Physiol Lung Cell Mol Physiol.2001;281 :L1392– L1401
- ↵Singh M, Mynak ML, Kumar L, Mathew JL, Jindal SK. Prevalence and risk factors for transmission of infection among children in household contact with adults having pulmonary tuberculosis. Arch Dis Child.2005;90 :624– 628
- ↵Kolappan C, Gopi PG. Tobacco smoking and pulmonary tuberculosis. Thorax.2002;57 :964– 966
- ↵Santiago EM, Lawson E, Gillenwater K, et al. A prospective study of Bacillus Calmette-Guérin scar formation and tuberculin skin test reactivity in infants in Lima, Peru. Pediatrics.2003;112(4) . Available at: www.pediatrics.org/cgi/content/full/112/4/e298
- Copyright © 2007 by the American Academy of Pediatrics