| Several studies have reported that socioeconomic status (SES) is inversely associated with the risk of lung cancer. This thesis investigates different aspects of the relationship between SES and lung cancer risk in order to improve our understanding of the factors influencing the risk of this disease. Its overall objective is to explore associations between SES and risk of lung cancer according to: (1) different indicators of SES, (2) types of adjustments for smoking, and (3) potential intermediate risk factors of the association between SES and lung cancer.The first article compares how different indices of financial availability, i.e., an index based on residential values, the median income from census data, and the self-reported household income, each recoded into 5 categories, classified study participants. When compared to the self-reported household income, very good concordance was observed for the residential value index and the census income (73% and 64%, respectively). The weighted kappa was higher when comparing self-reported household income with the residential value index (Kappap=0.36, IC 95%: 0.31--0.40) than with the census income (Kappap=0.22, IC 95%: 0.17--0.27). The choice of the financial availability index, resides, however, on the study objectives and feasibility aspects in the study population.The second article describes associations between the three indices of financial availability and the risk of lung cancer. The strong associations observed initially with crude models disappeared after adjustment for the multiple dimensions of smoking. Smoking therefore played an important role in the relationship between SES and risk of lung cancer.Finally, potential intermediate risk factors of the association between three different dimensions of SES---education, income, occupational class---were studied. The results suggest that the three potential intermediate risk factors examined, i.e., smoking, intake of carotenoid-containing fruit and vegetables, and occupational exposures, had an independent mediating effect on the relationship between SES and lung cancer. The most important mediator was smoking, followed by diet, and, to a much lesser extent, occupational exposures.Data from a large case-control study conducted in Montreal between 1996 and 2002, and assessing the role of environmental factors in the etiology of lung cancer, were used. The sample analyzed consisted of 1,203 cases (738 men and 465 women) diagnosed in 1996--1997 across 18 major hospitals in the Greater Montreal and 1,513 controls (899 men and 614 women). Controls, recruited randomly from the electoral list, were frequency-matched to cases by age, sex and electoral district. The subjects were interviewed to collect information on a wide range of factors, including socio-economic and demographic characteristics, smoking and occupational history, and diet. Additional data collection was conducted specifically for this thesis. Indeed, property value assessments were obtained from the public database of the City of Montreal using the participants' addresses provided at the time of interview.Future studies on the relationship between SES and lung cancer risk would benefit from considering various SES indicators when possible, and should apply a detailed adjustment for the different smoking dimensions. A better understanding of the various intermediate risk factors is essential for formulating preventive measures for lung cancer.Keywords: Socioeconomic status, education, income, occupation, diet, lung cancer, intermediate risk factors... |
