DOMESTIC EXPOSURE TO COOKING FUMES AND RISK OF LUNG CANCER AMONG CHINESE WOMEN IN SINGAPORE

Abstract

Despite their low smoking prevalence, Chinese women in Singapore and elsewhere have among the highest rates of lung cancer in the world. There is some evidence that factors in the domestic environment may account for part of this risk. The objective of the present study was to examine the association between lung cancer risk and cooking exposure in this population. In the light of evidence for the carcinogenicity of heterocyclic and other aromatic amines, we were interested specifically in the effect of exposure to fumes from cooking meat, and in the genetic polymorphisms involved in the metabolic pathway of these compounds. Data on demographic background, smoking status, domestic cooking exposures and other relevant variables were obtained by in-person interview of 303 Chinese women with pathologically confirmed lung cancer, and 765 age-matched controls admitted to three major hospitals in Singapore between 1 April 1996 and 30 September 1998. The proportion of current and ex-smokers among cases and controls was 41.7% and 13.1% respectively. Adenocarcinomas constituted 31.5% of cancers among smokers and 71.6% among non-smokers. The risk estimate for adenocarcinoma of the lung among current smokers was 2.2 (95% C.I. 1.5 - 3.7), and for squamous or small cell carcinoma it was 19.8 (10.8- 36.4). Among smokers, women who reported stir-frying daily 20-30 years prior to diagnosis had a significantly elevated risk of lung cancer (adjusted odds ratio (OR) 1.9, 95% C.I. 1.0 - 3.7). This was enhanced if they stir-fried meat daily (adjusted OR 2.5, 1.2 -4.8), and further raised if the kitchen was filled with oily fumes during cooking (adjusted OR 3.1, 1.5 - 6.4). Among non-smokers, we did not observe a relationship between frequent cooking and risk of lung cancer. However, women who reported habitually re-using oil for frying in adolescence or 20-30 years previously had elevated risks, the adjusted OR being 2.2 (95% C.I.1.2 - 4.0) and 1.4 (95% C.l. 0.8 - 2.3) respectively. Among non-smokers, we also observed weak but significant associations between lung cancer and environmental tobacco smoke exposure, past history of tuberculosis and family history of cancer. A genetic analysis of 153 cases and 141 controls for N-acetyltransferase 2 genotype revealed that among non-smokers, slow acetylator status conferred a higher risk of lung cancer. The age- and birthplace-adjusted OR was 2.1 (95% C.I. 1.1 - 4.0) for adenocarcinomas, and 1.7 (95% C.I. 0.5 - 6.5) for squamous or small cell carcinomas. The number of smokers in this study was small, and no association was observed. Our results suggest that the pattern of risk factors for lung cancer differs markedly between smokers and lifetime non-smokers in our population. Smoking and cooking appear to have a multiplicative effect. Our results are consistent with a role for heterocyclic amines from cooking of meat in causing this effect, and reveal the need for more studies in this area. Among non-smokers, it is possible that the dose from cooking alone is not sufficient to affect risk, but the relationship between risk of the disease and NAT2 activity again suggests that the total exposure to arylamines in the environment may play a role in lung carcinogenesis in this population.