Tobacco Documents Online

I I I I I I I I I I I I I I THE RELATIONSHIP BETWEEN SMOKING AND LUNG CANCER IN HUMANS Geng Guan-yi, Liang Zhong-hua, Xu Rui-heng, Liu Jing-ying and Shi Pei-ying Tianjin Medical University, Tianjin, China Abstract To attempt to clarify the relationship between smoking and lung cancer in humans, a case-control study was conducted. Ninety-nine male and 193 female lung cancer cases were included, matched 1:1 with controls on the basis of age, sex and area of residence. In males, the OR associated with smoking was 5.9 (95 % CI 2.65-13.50); ORs increased with the amount smoked. The average age of starting smoking was earlier for cases (22.4 years) than for controls (24.7 years). The OR for smoking was 3.31 (95% CI 1.96-4.42) in female cases. About 60% of cases were considered to be attributable to active smoking. The OR for lung cancer in a nonsmoking wife married to a smoking husband was 2.16 (95 % CI 1.03-4.53). 42 % of lung cancers in nonsmoking women were considered to be attributable to husband's smoking. ORs for passive smoking increased with the amount smoked by the husband, and with the length of time spent living with the husband. In total, about 60% of female lung cancer cases are believed to be attributable to active or passive smoking. The OR associated with occupational exposures was 3.1 (95% CI 1.58-6.02). The OR for cooking with coal (1x10° hours of exposure) was 1.54 (95% CI 1.20-1.96) and was 5.56 (95% CI 3.40- 9.10) for 4x10° hours of exposure, or approximately three hours per day for 37 years. The OR for pulmonary disease was 2.64. In conclusion, most lung cancer cases could be attributable to active or passive smoking in this population. Introduction Considerable controversy exists on the relationship between smoking and lung cancer, especially among females. Among all major cities in China, mortality rate of lung cancer in Tianjin is high; ranking second in males and first in females (28.3/105). We conducted a 1: 1 match case-control study to further investigate the relationship between smoking and lung cancer. Materials and Methods A total of 292 cases of lung cancer (99 male, 193 female) were analyzed. Of these, the majority p , (255 cases, 87.3 %) were confirmed histologically or cytologically. The other 37 cases were diagnosed ~ ~ to be lung cancer by CT, X-ray or bronchoscopy. All cases involved individuals residing in Tianjin for -4 more than 10 years. Cases were matched 1:1 with 292 controls, by age (± 2 years), sex, race, marital w w 0) I

I I status and location of home residence. The distribution of lung cancer histologic type is illustrated in Table 1. Table 1. Cell Type of Lung Cancer Cases Male . I pemate Cell Type No. % No. Group I(squamous cell and small cell) 42 42.4 83 43.0 Group II (adenocarcinoma) 32 32.3 73 37.8 Type unknown 25 25.3 37 19.2 Total 99 100.0 193 100.0 Results Age and residence distribution of the cases in this study is similar to other lung cancer studies conducted in Tianjin. The smoking rate of the controls is also typical of the population in Tianjin. Sutmnary of findings pertaining to active smoking The OR for active smoking and lung cancer is shown in Table 2. Table 2. OR of Active Smoking Controls Cases Smoker Nonsmoker Smoker Nonsmoker 116 102 27 47 102 3.78 OR 27 (102 - 27)2 ------------ 102 + 27 43.6 95%CI=2.17-4.30 P < 0.01 I I I I I I I I I I I I I O O V 0o -2- W N W V 1 I

I I I I I I I I I I I I I I I I I I 218 Exposure rate (cases) 74.7 % 292 143 Exposure rate (control) 49.0 % 292 .49 (3.78 - 1) PAR % = ------------------------ 57.7% (according the exposure rate of control) .49(3.78-1)+1 .747 (3.78 - 1) PAR % = _ ----.747----(3----.78------1)---+--1-- 76.5 % (according the exposure rate of disease) The OR for active smoking in males and females are shown in Table 3. Table 3. OR For Active Smoking in Males and Females Cases Smokers Nonsmokers 29 OR (m) _ ---- = 5.8 5 Controls. (female) Smokers.. Nonsmokers Smokers 63 29 53 73 5 2 22 45 73 OR(t)= --- = 3.32 22 95%CI=2.59-12.44 Exposure rate (cases) = 92.9% Exposure rate (controls) = 68.7% .929 (5.8 -1) PAR%= - -- --- .929 (5.8 -1) + 1 .687 (5.8 -1) PAR%= ------------------------ 95 % CI = 1.96 - 4.42 Exposure rate (cases) = 65.3% Exposure rate (cont.) = 38.9% .653 (3.32 -1) = 81.7% PAR%= - .653 (3.32 -1) + 1 .389 (3.32-1) = 76.7% PAR%= ------------------ = 47.4% -3- '

.687 (5.8 -1) + 1 .389 (3.32 -1) + 1 I I I The OR for amount of smoking is shown in Table 4. I Table 4. OR and Amount of Smoking I No. of Cig./day OR (male) OR'(femaIe) I 0 1 1 I l- 1.66 1.47 10- 2.98 2.52 20- 14.78 6.27 I 30- 27.72 The OR for duration of smoking is shown in Table 5. I Table 5. OR for Duration of Smoldng I Year of smoking OR 95 % Cl I 0 1 1- 1.73 1.38 - 2.18 I 20- 3.00 2.17 - 4.16 40- 5.20 3.49-7.75 In analyzing the age at which smoking began, cases were shown to st controls (Table 6). art smoking earlier than I I ' ~ 4- i V N W tn I I

I 1 I I I I I I I I I I I I I I I I Table 6. Age in Which Smolting Began Male Female Cases 22.4 18.9 Controls 24.7 28.2 P < 0.05 < 0.01 OR was found to be higher among those who started to smoke at an earlier age (Table 7). Table 7. OR and Age Smolting Began Age starting to smoke OR Nonsmoker 1 _ 21 1.59 16-20 3.1 < 15 6.30 ~.61 In summary, the OR increased with the amount and with duration of smoking. Moreover, cases were shown to begin smoking at an earlier age than the controls. Summary of Studies in Nonsmoking Females Pertaining to Exnosure to Passive Smoke The OR of nonsmoking female cases being exposed to passive smoke from parents, siblings and colleagues was not significantly higher than the nonexposed group. However, exposure to smoke from husbands resulted in an elevated OR (Table 8). Table 8. OR of Passive Smoking in Females -5- I

I I 34 x 52 OR = ---------- = 41 x 20 2.16 95 % CI = 1.03 - 4.53 Exposure rate(cases) _ Exposure rate (control) = 34 34 + 20 41 63 % 44.1 % 41 + 52 .63(2.16-1) .441(2.16-1) PAR % = ----------------------- = 42.5 % PAR % _ ----------------------- = .63 (2.16 - 1) + 1 .441(2.16-1)+1 42.5 % Table 9 shows that OR of female lung cancer increased with number of cigarettes smoked per day by their husbands and with duration of exposure. Table 9. OR of Duration and Amount of Spousal Smoking Amount OR 95% CI Duratlon. OR of exp. of exp. (Yr) 95% CI- 0 (cig/d) 1 0 1 1- 1.4 1.12 - 1.76 1- 1.49 1.15 - 1.94 10- 1.97 1.42 - 2.72 20- 2.23 1.54 - 3.22 20- 2.76 1.85 - 4.10 40- 3.32 2.11 - 5.22 OR for active and passive smoking in relation to female lung cancer cases is shown in Table 10. Table 10. OR of Active and Passive Smoking Wives . .... --------------- - -- - -- - -- - -- - ---- --------- Nonsmoking Smoking Nonsmoking 1.0 3.32 (1.96 - 4.42) Husband Smoking 2.16 (1.03 - 4.53) 4.90 (1.8 - 9.5) -6- N O O i V 00 Gf N ? ~ ' I I I I I I I I I I ' I ' I I I

a I I I I I I I I I I I I I I I I I If a smoking female has a husband who also smoked, the OR was 4.9. The exposure rate was 64.9% (137/211). PAR %=[0.649(4.9-1)] [0.649(4.9-1)1/ [0= 71.7%. For the 126 smoking female lung cancer cases 90.34 (126 x 0.717), and for the 67 nonsmoking female lung cancer cases 22.64 (67 x 0.338) were due to passive smoking. Therefore in Tianjin 58.5% (113/193) of female lung cancer may be attributed to smoking or passive smoking. Other Risk Factors for Lung Cancer in Females: Occupational Exposure OR of occupational exposure, e.g., textile workers, workers exposed to asbestos, benzene, etc., was 3.1, 95% CI = 1.58 - 6.02. Since the exposure rate was low in the Tianjin population (about 5%), the PAR% attributed to the occupational exposure was only about 9.5% and is much less than that due to smoking. Exposure to Cooking With Coal These results are shown in Table 11. Table 11. OR of Female Lung Cancer Due to Cooking With Coal Duration of Exposurc (hr) OR ° 95% CI 1 x 1 (1.5 r x 20 yr) 1. - 1. 2 x 10° (1.5 hr/d x 40 yr) 2.36 (1.66 - 3.34) 3 x 104 (2 hr/d x 42 yr) 3.62 (2.34 - 5.55) 4 x 10° (3 hr/d x 37 yr) 5.56 (3.40 - 9.10) The OR for exposure to coal fume for 2 x 104 hr, i.e., 1.5 hr per day for 40 years, was 2.36 which was nearly equal to the OR of passive smoking for 20 years. It is, however, much less than the OR for active smoking in females (3.32). Interactive Effects of Risk Factors for Lung Cancer in Females Multifactor analysis by conditional logistic regression method showed that the combination of active smoking, passive smoking, occupational exposure, history of lung disease and 4 x 10^ hr cooking with coal resulted in an OR of 50, 95 % CI = 13.7-185.3, in comparison with those without the above risk factors and cooking with coal for less than 3 x 10^ hrs. -7- I

I Discussion Due to the low rate of smoking among females, it seems likely that risk factors other than cigarette smoking exists for lung cancer. Because smoking rate is very high in men in all areas of China, it is conceivable that there is high exposure to passive smoke among females. A number of studies have addressed the effects of exposure to passive smoke, especially in relation to the incidence of lung cancer in nonsmoking females; the results, however, have been inconsistent and at times, contradictory, between studies. Hirayama showed data that suggest that the incidence of some cancers may be elevated in nonsmoking wives with smoking husbands than those with nonsmoking husbands. Similar results were obtained in our studies. Although cooking with coal (exposure to coal fumes) was one of the risk factors for lung cancer, exposure is expected to gradually decrease with the increased use of gas. The risk associated with occupational exposure is also expected to be small. These considerations, therefore, raise the possibility that the primary risk factors for female lung cancer are active and passive smoking (58.5 % female lung cancer was attributed to smoking). However, in the absence of actual exposure data in other areas, such conclusions must await further investigation in the future in other places in China. The PAR% of male lung cancer attributable to smoking was 76.7- 81.7, consistent with the notion that smoking was the main cause of male lung cancer. In conclusion, both active and passive smoking are important risk factors of lung cancer in Tianjin. The interactive effect among some risk factors increased the OR substantially. Therefore stopping smoking might significantly decrease the OR and conceivably also the incidence of lung cancer. While most other risk factors has a lower OR compared to smoking, the combined OR can also increase quite significantly if other risk factors are allowed to be combined with smoking. -8- I I I I I I I I I I I I I I I