Tobacco Documents Online

I I I I I I I I I I I I 1 I I I I I STUDY OF THE RELATION BETWEEN SMOKING AS A LIFESTYLE FACTOR AND LUNG CANCER IN BEIJING AREA OF CHINA Fan Ruo-lan*, Zheng Su-hua*, Wu Zhao-su**, Wu Zhao-ru*, Zhang Rui-song**, Cao Li-hua* and Li Yu-zhen* * Beijing Tuberculosis and Thoracic Tumor Research Institute, Beijing, China ** Beijing Heart, Lung and Blood Vessel Medical Center, Beijing, China A case-control study involving interviews with 403 (252 male and 151 female) primary lung cancer patients and 1151 (734 male and 417 female) population-based controls from the Sino-MONICA- Beijing Project (involving about 750,000 residents) showed that cigarette smoking, as a lifestyle factor, is the highest risk factor for lung cancer. The overall odds ratio (OR) was 2.65 (95 % CI 2.04-3.44) for all patients. The OR was 2.84 (95% CI I.90-4.28 for males and 3.92 (95% CI 2.59-5.94) for females. The Population Attributable Risk (PAR) was 55.5% for males and 40.5% for females. It is clear that risk trends rise with increasing smoking intensity, duration and degree of deep smoking (inhalation). About 70% of all patients were examined by pathological and/or cytological techniques. The association between smoking and lung cancer was the strongest in squamous cell carcinoma (PAR 87.3 %). Small cell carcinoma ranked second (PAR 44.5%). The correlation was not statistically significant for adenocarcinoma (PAR 14.3 % and P> 0.05). Introduction Disease patterns in China have changed greatly from the 1950s to the 1990s. The mortality rate from malignant neoplasms was 63.9/100,000 in the 1950s. It increased rapidly to 128.0/100,000 in the 1990s and became a major cause of death. The percentage of total deaths from cancer was 5.17% in the 1950s and it increased to 21.88 % in the 1990s. Cancer ranked as the leading cause of death in the 1990s; it was only the seventh cause of death in the 1950s. Lung cancer mortality increased in China from 1973-1975 to 1990. Nationwide survey figures for lung cancer mortality in 1973-1975 were 5.45/100,000; it increased to 32.89/100,000 in 1990. The percentage of deaths from lung cancer increased for all of China each year; the average annual increase rate was 11.9%. The increase in the lung cancer death rate is the highest among the selected sites of malignant neoplasms: 8.1 % for intestinal and rectal cancer, 5.1 % for breast cancer, 4.0% for liver cancer and 1.4% for stomach cancer. Only esophageal cancer decreased from 1973-1975 to 1990 with an average annual reduction rate of 3.3 %. Generally speaking, lung cancer mortality was higher in large cities such as Beijing, Shanghai, Tianjin and Guangzhou and along the east coast. The mortality from lung cancer in Beijing men was 29.6/100,000 in 1977-1978 (urban areas) and 20.6/100,000 in females. It increased to 33.0/100,000 in males and 23.3/100,000 in females in 1986. It is obvious that the trend is for lung cancer mortality, in Beijing, to continue to increase in recent years. Thus, it is important and necessary to study the etiology of lung cancer in the Capital city. W I

I 1 Using data from the Sino-MONICA-Beijing Project in the People's Republic of China from January 1, 1990 to December 31, 1991, Odds Ratios (OR), Attributable Risk (AR), and Population Attributable Risk (PAR) were measured in this study. Methods A case-control study of primary lung cancer patients (International Classification of Disease Ninth Revision, Code 162) in this study came from the monitoring system for the Sino-MONICA-Beijing Project during a period of 24 months. A three-level monitoring•system for the Sino-MONICA-Beijing Project was formally started on January 1, 1984 in six scattered urban districts and one rural county of Beijing. The WHO-MONICA Project is a worldwide monitoring system for cardiovascular disease from 1984 to 1993. From January 1, 1990 to December 31, 1993 the epidemiology of lung cancer was an integral part of Sino-MONICA-Beijing Project. The three-level monitoring system consisted of the following: a) The Beijing Heart, Lung and Blood Vessel Medical Center served as the coordinating Center (the first level); b) forty-two districts and regional hospitals formed the actual monitoring units (the intermediate level) and c) the residence committee and residence health stations were the basic units of the monitoring system which involved 335 units in urban areas and 188 villages in rural areas in 1990. There were 742,198 residents in the Sino-MONICA-Beijing monitoring system in 1990 (369,427 males and 372,771 females). There were 580,973 residents in urban areas and 161,225 in rural area. The cases in this study were adults aged from 18 to 80 years old. Controls were selected from the same residence committee or village of the general population of the monitoring system. All of the controls were of the same sex as the cases and their ages were within two years more or less than the cases. Interviews were sought with all living cases and controls. Family members were visited if the case was deceased. The OR, AR and PAR were the measures of the association between smoking and primary lung cancer. Results A total of 252 male and 151 female cases of primary lung cancer were identified over 24 months. Interviews were completed with 734 male and 417 female controls. (Table 1) I I I I I I I I I I I I I -2- 0 00 ~ ~ co W W W A I I I I

I I I I I I I I I I I I I I I I I I Table 1. The status of interviews with primary lung cancer and population-based controls in the Sino-MONICA-Beqing monitoring system (1990-1991) Sex Number of Lung Cancer Cases .. . Number of Controls . Male 252 - - - 734 - -- - Female 151 417 Total 403 1151 Table 2 shows the odds ratio from a case-control study to evaluate the relationship between cigarette smoking and lung cancer. The reported frequency of cigarette smoking in lung cancer patients (85.7% (216/216+36) in males and 54.3% (82/82+69) in females) was much higher than that in the controls. The corresponding figures were 67.8% (498/498+236) in males and 23.3% (97/97+320) in females. Smokers experienced an increased risk of lung cancer (OR 2.84 in males, 3.92 in females). Table 2. Odds Ratios from a case-control study of the relationship between cigarette smoking and lung cancer among monitoring system for Sino-MOIVICA-Beijing (1990-1991) Cigarette Smoking Male Case Control Female Case Control Total Case Control Yes 216 498 82 97 298 595 No 36 236 69 320 105 556 OR 2.84 3.92 2.65 95%CI 1.904.28 2.59-5.94 2.04-3.44 The results of the indices of risk measured by OR, AR and PAR are shown in Table 3. The AR of 64.8 % among males and 74.5 % among females was attributed to smoking. A PAR of 55.5 % for males and 40.5 % for females was calculated. -3- ~ 0 0 ~ -4 0o W W W tJf I

I Table 3. Indices of risks between cigarette smoking and lung cancer measured by OR, AR and PAR Indices of Risk Male Female Total 1. OR 2.84 3.92 -. 2.65 2. AR 64.8 74.5 62.3 3. PAR 55.5 40.5 46.0 Diagnosis of 68.2% of the lung cancers (403) were based on pathological examination of tissue specimens and/or based on cytology examinations. Table 4 shows the risk values measured by OR, AR and PAR in different cell types. According to the PAR, 87.3 % of squamous cell carcinoma and 44.5 % of SCLC and 14.3% of adenocarcinoma were attributed to smoking. Table 4. OR, AR and PAR for squamous, adenocarcinoma and SCLC associated with cigarette smoking Pathology andtor Cytology Eaam. No. of Cases OR AR PAR Squamous cell carcinoma 81 *12.18 91.8 87.3 Adenocarcinoma 112 #1.39 28.1 14.3 SCLC 55 **2.48 59.7 44.5 . .* P = 0.00000001 P=0.008 N P = 0.139 I I I I I I I I I I I I I , IV O O ~ ' V 00 W 4- W W O ' , I

I I I I I I I I I I I I I I Tables 5, 6 and 7 show that risk trends increased with intensity (daily dose), duration and degree of inhaling. Table 5. OR for lung cancer associated with intensity (daily dose) of cigarette smoldng Mzle r<mLle . . . .. .. Tunl Nwvba of Cigmctle.clday rM Cadro] OR Cx.u Cwtral- DR Cam C®ttol OR 0 36 236 1.00 69 320 IAO 105 556 1.0 1 13 121 070 17 48 1.64 30 169 0.9 10 53 171 2.03 30 37 3.75 83 208 2.1 20 111 183 3,98 31 12 11.94 142 195 3.8 30 39 23 11.12 4 1 18.49 43 24 9.8 Table 6. OR for lung cancer associated with duration of cigarette smoldng Dmmen of Meic Frndc TUW Smoking by YeUs Cue CmtrN OR C.ce Ctmrcal OR Ctx Covtrol OR 0 36 236 1.00 69 320 1.00 105 556 I 1 29 135 1,41 8 15 2.47 37 150 1 30 44 122 2.36 19 23 3.83 63 145 2 40 143 2A1 3.98 55 59 8.38 198 300 3 ' ! N 0 00 , V Co W W ' S V !

I Table 7. OR for lung cancer associated with degree of inhalation of cigarette smoldng Degree Of Ma1e Fema(c 7op1 Inhalatirm Caae Cmtrnl OR Ca.u Contrul OR Gse Cootroi OR No amkmg 36 236 1.00 69 " 320 1.00 IOS 556- I. I" 7 89 0.52 14 25 2.59 21 114 0. I1"" E 61 171 2.34 9 27 1.54 70 198 1. IR'•• 36 72 3.28 5 7 3.30 41 79 2. SIWIow mialafioa "seuW geu m and an af mwlh Middle iWalaficn: "smolse" geu hn aM mLL of nose Oeep mhaledo¢ 'smoke' gets m and out of Nurax Discussion A case-control study is a suitable design for an epidemiological study of lung cancer. The investigation in this paper was carried out in the Sino-MONICA-Beijing Project. Cigarette smoking is very common in China. Nationwide random survey data show that smoking rates are 69.7% among men more than 20 years old and 8.2% in women more than 20 years old. The residents of the Sino-MONICA-Beijing Project are a defined population. Because both lung cancer cases and the population-based controls came from the defined residents, this can avoid the Berkson bias. The correlation between smoking and lung cancer was measured by OR, AR and PAR. The results were an OR 2.84 for males and an OR 3.92 for females in this paper. It is not clear why the OR in males was lower than that in females; further study is required. The risk of lung cancer increased with the daily dose (intensity) and duration of cigarette smoking and with the degree of inhalation. It is obvious that cigarette smoking, as a lifestyle, is the highest risk factor for lung cancer among Chinese men. The PAR among men (55.0%) is larger than that among women (40.5 %). By comparison, the PAR value for men in Shanghai (69.0%) was higher than that in Beijing men (55.5%). On the contrary, the PAR among women in Shanghai (24.0%) was lower than that in Beijing women (40.5%). But the PAR among American men (89.5 %) and among American women (85.5 %) was much higher than that in Chinese men and women. Perhaps cigarette smoking is less hatmful to Chinese than to people elsewhere. I I I I I I I I I I I I I I I O O R y 00 -6- 0 W 1 W O I

I I I I I I I I I I I In this paper, the percentage between males (160 cases) and females (88 cases) was different for the principal types of lung cancer (248 cases). Among males, the percentage for squamous cell type was 79.0 %(21.0% in female), 69.1 % for small cell lung carcinoma and 51.8 % for adenocarcinoma. There also was a difference in the risk observed in different cell types. All of the principal types of lung cancer (squamous cell, small cell lung cancer and adenocarcinoma) were affected. OR (12.18), AR (91.8 %) and PAR (87.3) for squamous cell have been reported. The association between smoking and squamous cell type was the strongest (P=0.00000001). The association between smoking and SCLC was the second (P =0.008). The figures for SCLC were 2.48. 59.7% and 44.5%, respectively, and 1.39. 28.1% and 14% for adenocarcinoma, respectively. The correlation between smoking and adenocarcinoma was not statistically significant (P=0.139). Acknowledgements: Thanks to the Beijing Scientific Association and Beijing Bureau of Public Health as sponsors. I I I I I N O O ~ -4 00 I I _'j_ W W W CO

I I References I 1. Y-T, Gao et al. Lung Cancer and Smoking in Shanghai. I Int. J. Enidemiol. (1988);17(2):277-280. 2. Samet, Jonathan et al. Cigarette Smoking and Lung Cancer in New Mexico. Am. Rev. Respir. Dis. (1988); 137:1110-3. I 3. Beckett, W.S. Epidemiology and Etiology of Lung Cancer. Clinics Chest Medicine (1993);14(1):1-15. I 4. Schlesselman, J.J. Case-Control Studies. Oxford University Press (1982). , 5. Hennekens, C.H. et al. Epidemiology in Medicine, Little Brown & Company (1987) 1 1 ' I I I I I N O ' ~ ~ ti C* w ' w .p O -8- I I