In-home coal and wood use and lung cancer risk: a pooled analysis of the International Lung Cancer Consortium

Abstract

Background: Domestic fuel combustion from cooking and heating is an important public health issue because roughly 3 billion people are exposed worldwide. Recently, the International Agency for Research on Cancer classified indoor emissions from household coal combustion as a human carcinogen (group 1) and from biomass fuel (primarily wood) as a probable human carcinogen (group 2A).

Objectives: We pooled seven studies from the International Lung Cancer Consortium (5,105 cases and 6,535 controls) to provide further epidemiological evaluation of the association between in-home solid-fuel use, particularly wood, and lung cancer risk.

Methods: Using questionnaire data, we classified subjects as predominant solid-fuel users (e.g., coal, wood) or nonsolid-fuel users (e.g., oil, gas, electricity). Unconditional logistic regression was used to estimate the odds ratios (ORs) and to compute 95% confidence intervals (CIs), adjusting for age, sex, education, smoking status, race/ethnicity, and study center.

Results: Compared with nonsolid-fuel users, predominant coal users (OR = 1.64; 95% CI, 1.49-1.81), particularly coal users in Asia (OR = 4.93; 95% CI, 3.73-6.52), and predominant wood users in North American and European countries (OR = 1.21; 95% CI, 1.06-1.38) experienced higher risk of lung cancer. The results were similar in never-smoking women and other subgroups.

Conclusions: Our results are consistent with previous observations pertaining to in-home coal use and lung cancer risk, support the hypothesis of a carcinogenic potential of in-home wood use, and point to the need for more detailed study of factors affecting these associations.

Figure 1

Lung cancer risk (ORs and 95% CIs) in all solid-fuel users (A), predominant wood users in North America and Europe (B), and predominant coal users in Asia (C) by sex and smoking status compared with subjects who used nonsolid fuels (gas, oil, electric), adjusted for age, sex, education, race/ethnicity, and smoking status. The gray squares represent the adjusted ORs, with the size proportional to the number of cases in that subgroup of analyses (i.e., overall, wood in Europe/North America, coal in Asia), and the horizontal lines represent the 95% CIs. p-Value for heterogeneity measures heterogeneity between studies. (A) All solid-fuel users (n = 1,548 unexposed cases and 2,732 unexposed controls). p-Value for heterogeneity between subgroups: men versus women, 0.18; ever- versus never-smokers, 0.14; whites versus Asians, 6.5 × 10⁻⁵; studies carried out in Asian versus North American and European countries, 1.9 × 10⁻⁷. (B) Predominant wood users in North America and Europe (n = 3,146 unexposed cases and 4,081 unexposed controls): includes only studies from North America and Europe (CEE, California, New England, Toronto), because the number of exposed cases from Asian countries was small (n = 94). The risk of lung cancer associated with wood use among all seven studies was 1.24 (95% CI, 1.10–1.41). Models also include interaction terms between the covariates and study. (C) Predominant coal users in Asia (n = 165 unexposed cases and 511 unexposed controls): includes only studies from Asia (Singapore, Xuanwei1, Xuanwei2). The risk of lung cancer associated with coal use was 1.64 (95% CI, 1.49–1.81) among all seven studies and 1.15 (95% CI, 1.02–1.30) among the four studies in North America and Europe.