Tobacco Documents Online

AN ANALYSIS OF RISK FACTORS OF LUNG CANCER IN NANJING AND AN ESTIMATE OF ATTRIBUTABLE RISKS By Shen Xio-bing, Wang Guo-long, Xiang Long-Sheng, Wang Xing-he, Nanjing Railway Medical College Abstract: This is a case-control study of 83 cases of primary pulmonary squamous cell carcinoma and 180 cases of primary pulmonary adenocarcinoma in Nanjing. Multi- variate conditional logistic regression analysis shows five risk factors for pulmonary squamous carcinoma. These were: smoking, cooking fume pollution, family tumor history, type of fuel used in the home, and use of coal stove for heating in winter. The relative risks (RR) for these five risk factors were 1.0340, 3.8138, 5.6135, 4.9665, 3.7187, and the respective population attributable risks (PAR) were 0.6823, 0.5156, 0.2772, 0.5465, 0.3611. The four risk factors of pulmonary adenocarcinoma were smoking, cooking fume, chronic bronchitis, family tumor history. The respective RRs were 1.0124, 2.9943, 2.4925, 4.7693, and the respective PARs were 0.1987, 0.4699, 0.1763, 0.1844. The combined PAR of the five risk factors for pulmonary squamous cell carcinoma was 0.9431 and the combined PAR of risk factors for pulmonary adenocarcinoma was 0.7895.

Key words: Lung cancer, case-control study, tobacco smoke, indoor air pollution, bronchitis. Among malignant tumors, lung cancer has become one of the most threatening to human health. Lung cancer death rate in China shows an apparent rising trend in recent years.1 Squamous cell carcinoma and adenocarcinoma are the two major histopathological types of lung cancer. To examine the risk factors of squamous carcinoma and adenocarcinoma we have conducted a matched case- control study of risk factors for pulmonary squamous carcinoma and adenocarcinoma in Nanjing. Materials and Methods 1. selection of Cases: Diagnosed primary lung cancer cases were obtained from Nanjing Municipal Hospitals and were Nanjing residents for longer than 20 years. A total of 83 pulmonary squamous carcinoma and 180 adenocarcinoma cases was included in the study. 2. selection of Controls: Healthy controls were selected from over-20 year, tumor-free Nanjing residents. They were matched 1:1 with cases by sex, age (± 5 years), nationality, street address. - 2 -

3. Data Collection: * Shallow: information in the literature and incorporated features uniquely found in local population and environments. Table 1 contains coding information methods for and assign values. 4. Method of Analysis: Squamous carcinoma and adenocarcinoma risk factors were analyzed by conditional logistic regression; the risk factors were then subjected to analysis risk.234. to estimate attributable Table 1. Coding and Value Assignment of Data Variable Factor Assigned,Value X1 Smoking index Amount smoked (cigarettes per day) x years of smoking/20 X2 Degree of inhalation* No: 0; Shallow: 1-Medium: 2-Deep: 3 X3 Passive smoking No: 0; Yes: 1 X4 Occupational exposure to cooking fume No: 0; Yes: 1 X5 History of bronchitis No: 0; Yes: 1 X6 History of tuberculosis No: 0; Yes: 1 X7 Family tunor history No: 0; Yes: 1 X8 Living space average living area in last 20 years Non-solid fuel: 0; Solid fuel: 1 X9 Type of fuel in the home (fuel index)** No: 0; Yes: 1 X10 Coat stove foF winter heating consumption per person per month X11 Oil consumption in cooking No: 0; Yes: 1 X12 Kitchen cooking fume pollution No: 0; Yes: 1 X13 Regular consumption of fried food Average times of cooking per week X14 Cooking index Note: The standardized questionnaire utilized exhale by mouth; Medium: Deep: swallow smoke. ** Based on data of last 20 years. exhale by nose; - 3 -

Result 1. Multi-variate risk analysis. By conditional logistic regression model, risk factors for pulmonary squamous carcinoma and pulmonary adenocarcinoma were analyzed respectively, at a=0.05. Results are shown in Table 2 and Table 3. Table 2. Results of Conditional Logistic Regression Analysis in Pulmonary Squamous Carcinoma Group Standard Error of Regression Regression Relative Factor Coefficient Coefficient P Value Risk Cooking furne 1.3386 0.6536 0.0203 3.8138 Smoking index 0.0334 0.0132 0.0056 1.0340 Family tumor history 1.7275 0.7767 0.0132 5.6135 Fuet type 1.6027 0.9324 0.0428 4.9665 Coal Stove for heating 1.3134 0.7352 0.0370 3.7187 Table 3. Results of Conditional Logistic Regression Analysis in Pulmonary Adenocarcinoma Group Standard Error of Regression Regression Factor Coefficient Coefficient P Value Relative Risk Smoking index 0.0123 0.0065 0.0300 1.0124 Chronic bronchitis 0.9133 0.3411 0.0037 2.4925 Cooking fume 1.0967 0.2956 0.0001 2.9943 Family tumor history 1.5622 0.4627 0.0004 4.7693 2. Estimate of population attributable risk. - 4 -

After various risk factors were identified, the RR of the factors was calculated and the respective PAR and SAR were estimated. (Table 4) Table 4. Population Attributable Risks of Lung Cancer Risk Factors. Pulmonary Squamous Cell Carcinoma Pulmonary Adenocarcinoma Factor Level RR Number of Cases PAR RR Number of Cases PAR Smoking index Cooking fume Chronic bronchitis Family turnor history Coal stove for heating Measurement 0 1 0 1 0 1 0 1 1.0340 1 2.8138 1 5.6135 1 3.7187 83 25 58 55 28 42 41 0.6823 0.5156 .2722 0.9431 1.0124 1 2.9943 1 2.4925 1 4.7693 53 127 127 53 138 42 0.1987 0.4699 0.1763 0.1844 Fuel index Measurement 4.9665 83 0.5465 I SAR 0.9431 0.7895 Population attributable risk (PAR) is a measure of the relative risk of a given factor. For a given risk factor, PAR depends on the relation risk of the magnitude of exposure to that factor. As seen in Table 4, the main risk factors for pulmonary squamous cell carcinoma were smoking, type of fuel used in the home, cooking fumes, coal stove use for winter heating, and family tumor history; the risk factors for pulmonary adenocarcinoma were, in their order of significance, cooking fume, family tumor history, chronic bronchitis and smoking. - 5 -

Discussion Most studies show that smoking has different effects on various types of lung cancer. Our study shows that smoking is a major risk factor for squamous carcinoma in the lung, with a PAR of 68.23%. But the PAR of smoking for pulmonary adenocarcinoma was 15.6%. Passive smoke was not found to be one a risk factor. Cooking fume is the vapor-phase product of cooking and resulted from pyrolysis of cooking oil and food under high heat. Since Chinese traditionally cook with high temperatures, cooking fume is one of the major indoor pollutants. We have tested the chemical composition of cooking fume for its genetic toxicity and found it to contain benzo(a)pyrene, benz(a)thracene. Toxicology experiments also confirmed the mutagenicity of the contents of cooking fume. Our study found cooking fume to be a common risk factor for both squamous carcinoma and adenocarcinoma in the lung. The respective relative risk was 3.8138 and 2.9943 and the respective PAR was 51.56% and 46.99%. Therefore, control of cooking oil-food pyrolysis and improvement of kitchen exhaust should be emphasized. At present, many urban residents of China still use coal and charcoal as fuel for daily living, and in winter coal stoves are used for heating. These are primary causes of indoor air - 6 -

pollution. Our study shows a close relationship between coal stove use and lung squamous cell carcinoma. Family tumor history is also a common risk factor for both squamous cell carcinoma and adenocarcinoma of the lung. Those with family history of tumor should avoid being exposed to other risk factors. Chronic bronchitis is related to the occurrence of pulmonary adenocarcinoma with RR value of 2.4124, and PAR value of 0.1724. Therefore, timely treatment of respiratory disease may help to reduce incidence of pulmonary adenocarcinoma.

References 1. National Environment Protection Bureau, "Environment of China 1991," J. of Chinese Environment 6(2): 1992. 2. Whittemore, A.S. "Estimating Attributable Risk from Case- Contro7. Studies," Am. J. Epidemiol. 117: 76-85, 1983. 3. Walter, S.D. "Effects of Interaction, Confounding and Observational Error on Attributable Risk Estimation," Am. J. Epidemiol. 117: 598-604, 1983. 4. Bruzzi, P., et al. "Estimating the Population Attributable Risk for Multiple Risk Factors Using Case-Control Data," Am. J. Epidemiol. 122: 904, 1985. 5. Levin, M.L. "The Occurrence of Lung Cancer in Man," Acta Un Intern Cancer 19: 531-41, 1953.