Background: Data on the association between exposure to environmental tobacco smoke (ETS) and Mycobacterium tuberculosis infection in children are limited.
Objective: To examine the dose-response effect of ETS exposure on the risk of M. tuberculosis infection in children in a high tuberculosis (TB) burden setting.
Methods: This cross-sectional study included healthy South African children from impoverished urban communities. Data were collected on household ETS and M. tuberculosis exposure, demographics, socio-economic and anthropometric data, M. tuberculosis infection, human immunodeficiency virus and TB disease status.
Results: Among 196 children (median age 6.8 years, range 0.3-15.9), 97 (49.5%) were M. tuberculosis - i nfected (tuberculin skin test [TST] ≥ 10 mm) and 128 (65.3%) reported ETS exposure; of these, 81/128 (63.3%) were exposed to ≥ 2 household smokers. The presence of ≥ 2 household smokers was associated with M. tuberculosis infection in univariate analysis, irrespective of TST cut-off point. In analysis adjusting for M. tuberculosis exposure, socio-economic status, age and previous TB treatment, ETS exposure remained associated with M. tuberculosis infection. In univariate and multivariate analysis, pack-years of exposure were associated with risk of TB infection.
Discussion: Exposure to ETS is associated with M. tuberculosis infection in children after adjustment for multiple variables, with a dose-response relationship between the degree of ETS exposure and risk of infection. Public health interventions to reduce exposure to tobacco smoke among children in high TB burden settings are urgently needed.