Table 6 Research in context

Evidence before this study

After reviewing primary studies and systematic reviews, the Lancet Respiratory Commission on household air pollution (HAP) judged the evidence linking HAP exposure and risk of developing tuberculosis (TB) as inadequate (Gordon et al., 2014). Recently, Lee et al (Lancet GH, 2020), summarised 53 studies and concluded that HAP was independently associated with TB (RR 1·26, 95%CI 1·08-1·48). Designs included mainly surveys, and results relied mostly on questionnaires with no history of cumulative HAP exposure. Smith and colleagues (2016) were the first to link individual CO measurements to TB in a large cohort from California. However, no such comparable study was conducted in sub-Saharan Africa nor in any country outside Africa with a high TB burden. Strikingly, although people living with HIV (PLHIV) are highly susceptible to TB, no studies have investigated whether exposure to HAP contributes to the risk of acquiring TB in this vulnerable population.

Added value of this study

This case-control study, conducted in a high HIV and TB burden region, is the largest study of HAP exposure from biomass fuel smoke and pulmonary TB risk among PLHIV. The study was conducted in an extremely resource-poor population in a post-conflict area of the DRC, where almost all households burn biomass fuels for cooking; ≥95% have no access to consistent electricity. Data obtained from 1277 PLHIV, 435 of whom also had TB (HIV+/TB+), allowed us to construct an index of HAP exposure intensity (hours cooking/day, hours cooking/week, lifetime cooking “ugali”). Moreover, 24h personal CO average levels helped us determine an exposure‑dependent relationship between HAP and TB in a random subsample of 255 participants.

Implications of all the available evidence

Time spent cooking (among women) and personal CO exposure were independently associated with increased risk of TB. In low-income settings, millions of deaths are ascribed to the convergent challenges of the HIV and TB epidemics and exposure to HAP. Longitudinal studies (e.g., community cluster randomised trials) are needed to confirm our findings and assess interventions to reduce incidence of TB attributable to long-term exposure to HAP in PLHIV.