Tobacco Documents Online

1 ! | I i i i 1 I I I APMIS SuppL 76: Fol. 10~," 9-47, 1997 Printed in Denmark. All rights reserved Copyright @ APMIS 1997 ISSN 0903-46.~X ISBN 87-16-15626-9 Tobacco smoking L. DREYER,~ J. E WINTHER,~ E. PUKKALA2 and A. ANDERSEN~ ~Institut~ of Cancer Epidemiology, Danish Cancer Society, 2Finnish Cancer R~gistry, ~Cancer R~gistry of Norway Dreyer, L., Winther, J. E, Pukkala, E. & Andersen, A. Tobacco smoking. APMIS Suppl. 76: Vol. 105: 9-47, 1997. Active smoking is causally associated with cancers of the lung, larynx, oral cavity, pharynx, oesoph- agus, pancreas, renal parenchyma, renal p~lvis and urinary bladder, and passive smoking appears to be causally associated with cancer of the lung. Information on smoking habits for the years 1965, 1975 and 1985 shows that more men than women in the Nordic countries were current smokers. The rates of women v~re stable over time and those of men were decreasing, approaching those of women. Lung cancer, in particular, is strongly associated with active smoking: by increasing the number of cigarettes smoked l~r da~ (lifelong) to 5, 10, 20 and 40 or more, the risk increases by five-, eight-, 16- and 30-fold, respectively, over that of people who have never smoked. Thus, with approximately 35% current smokers and 25% former smokers among Nordic men in 1985 and approximately 30% currant smokers and 15% former smokers among Nordic women in that year, by the year 2000 10,000 cases of lung cancer (6,500 in men and 3,500 in women) will be caused by active smoking; this is equivalent to 82% of all cases of lung cancer in thes~ populations. Another 6,000 cancers of other types (4,000 in men and 2,000 in women) are caused annually by active smoking, yielding a total of 16,000 n~w cases each yeax around the turn of the century. This implies that 14% (19°,4 in men and 9% in women) of all incident cancers in the Nordic countries around the year 2000 will b~ caused by active tobacco smoking. In comparison, passive smoking is a minor cause of lung cancer, responsible for approxi- mately 0.6% of all new cases (approximately 70 cases annually) in this area around the turn of the century. K~ words: Tobacco smoking; cancer incidence; prevention. L. Dreyer, Institute of Cancer Epidemiology, Danish Cancer Society. ACTIVE SMOKING Tobacco smoking is the major single cause of human cancer. It has b~en estimated that smok- ing of cigarettes was responsible for 30% of all cancer deaths in the United States in 1978 (1). In the Nordic countries, cigarette smoking has played a key role in the steady increase in cancer incidence observed in people of each sex over the past 30-40 years. In addition, regular to- bacco use is an important cause of non-malig- nant damage to the lung and cardiovascular sys- tem (2). Epidemiological studies initiated during the 1940s and 1950s in response to the dramatic in- crease in mortality from lung cancer observed in Europe and the United States established cigarette smoking as the major causal factor (3-- 10). Subsequently, thousands of scientific in- vestigations have confirmed this conclusion and have provided additional evidence that smoking is a cause of cancers at many other sites as well. In 1986, tobacco smoke was added to the list of agents found by the International Agency for Reseamh on Cancer (IARC) to be carcinogenic to humans (Group 1) (11). Tobacco smoke exerts its predominant card- nogenie effect on those tissues directly exposed, such as the bronchial lining of the lung; how- ever, organs distant from the smoke are also

DREYER et aL affected, since caminogens and pro-carcinogens are absorbed from the lungs into the blood- stream and circulated to all parts of the body. C~rtain carcinogens am found in particularly high concentrations in the urine of smokers (l 1,12). Although the oesophagus and stomach arc not directly exposed to inhaled cigarette smoke, its constituents condense on the mucous membrane of the mouth and pharynx and am swallowed; furthermore, mucus cleared from the lung reaches the oesophagus and stomach. The sites of cancer listed in Table 2.1 are causally linked to tobacco smoking. These are the types included in the overall assessment of the number of cancers avoidable if smoking were eliminated in the Nordic countries. More- over, cancers at the following sites are strongly suspected to be related to tobacco smoking: lip, liver, stomach, uterine cervix and bone marrow (leukaemia), and similar assessment have been made for these cancer types. Throughout the text and tables, however, the numbers of cancers thought to be avoidable have been kept separate from the numbers known to be avoidable. For all cancers causally linked to tobacco smoking, a dose-response relationship is seen, with decreasing risks observed after cessation of smoking. Overall, cancer rates arc modified by the following factors: type of tobacco used (amount of tar in cigarettes, presence or absence of a filter on cigarettes, cigars, cigarillos, pipe), the amount of tobacco smoked per day, the num- ber of years of smoking, the degree of inhalation, and, perhaps, individual susceptibility for cancer development (11, 13). Cohort and case-control studies conducted in the United States and the United Kingdom in the 1950s and 1960s generally showed two-to fourfold lower risk ratios for lung cancer among women than men for comparable levels of to- bacco consumption (13, 14). The difference was, however, partly a consequence of the fact that TABLE 2.1. Cancer types convincingly related to tobacco smoking P~spiratory Digestive Urinary s~ystvrn organs tract Larynx Oral cavity and l~nal paren- tongue chyma Lung Pharynx R~nal pelvis Oesophagus Urinary bladder Pancreas 10 women generally started smoking at a later age than men and more often smoked low-tar, filter- tipped cigarettes. As female smoking habits are approaching those of males, case-control and cohort studies conducted during the two last decades have generally shown little or no differ- ence in the risk for lung cancer by sex, for com- parable levels of tobacco consumption (13-20). Finally, it is important to be aware that to- bacco smoke may interact with other carcino- gens in the environment, e.g. ionizing radiation and asbestos, to produce particularly high rates of lung cancer or, with alcohol, high rates of cancers of the upper respiratory and digestive tract (21). When such synergism exists, it is dif- ficult to separate the effects of each carcinogenic exposure; in this wport, wc attempted to adjust for only the most important, best-described in- teractions. MATERIAL AND METHODS The proportions of cancers in the population that arc attributable to tobacco smoking - the population attributable ri~k percent (PAR°~) - w~re calculated on the basis of information on the sex-specific smok- ing habits in each Nordic country and estimates of the relative risk for each cancer site of relvvance to smoking (s¢~ also "Aims and Background"). Information on smoking Information on the smoking habits of men and women in the five Nordic countries was revi~w¢~l for the years 1965, 1975 and 1985 (Tables 2.2 and 2.3; 22-31). As the latency for the development of most solid tumours is about 15 y~ars, we assume that these cohorts are those most likely to contribute to the can- cot rates in 1980, 1990 and 2000. More detailed infor- mation on smoking habits can be obtained from the authors. Smokers of bbth cigarettes and other typ~s of to- bacco have bwn classified according to cigarette con- sumption, as use of cigarettes usually predominates and is associated with the highest relatiw risks for cancer. The term "pipe and cigar smokers only" re- fers to users of pip~s, cigars or cigarillos or combi- nations thereof. As reliable data on duration of smoking, degree of inhalation and age-specific smok- ing habits were not available for all countries, these variables w~re not included in the calculation of at- tributable risks. Detailed information on u,~ of smokeless tobacco (especially in fashion in Sweden today) was not available. The proportion of nonsmokers in each of the study I I I I I I I I I I I I I ,I I I I I

'.r1 l I ] TOBACCO SMOKING TABLE 2.2. Distribution of the adult male population of each of the five Nordic countries according to smoking habits in 1965, 1975 and 1985 Country Year Never Male smokers (%) smokers Ever Former Current Denmark 1965 13 87 1975 22 78 1985 39 61 Finland 1965 32 68 1975 37 63 1985 42 58 Iceland 1965 17 83 1975 20 80 1985 31 69 Norway 1965 22 78 1975 28 72 1985 34 66 Sweden 1965 26 74 1975 32 68 1985 43 57 Moderate" Heavyb 17 70 60 10 16 62 48 14 10 51 30~ 21d 18 50 18 32 32 31 11 20 24 34 15c 19d 19 64 42 22 25 55 37 18 27 42 31c lld 22 56 50~ 6d 27 45 38c 7a 28 38 30e 8d 23 51 42 9 28 40 22c 18r 30 27 11c 16r Data for 1965, 1975 and 1985 9r a close year • Current smokers of fewer than 15 cigarettes per day and smokers of pipe or cigars only b Current smokers of 15 eiga.rettes per day or more ¢ Current smokers of fewer than 20 cigarettes per day and smokers of pipe or cigars only a Current smokers of 20 cigarettes per day or more c Current smokers of 12 or fewer cigarettes per day and smokers of pipe or cigars only f Current smokers of more than 12 cigarettes per day populations was subdivided into "never smokers" and "former smokers". The,subgroup of former smokers is a heterogeneous group consisting of per- sons previously exposed to very different amounts of tobacco smoke, who quit smoking frbm a few months to several years before the survey. A comparison of Tables 2.2 and Table 2.3 shows that more men than women were "ever smokers" throughout the period. The proportion among men decreased consistently over time, however, due in par- tieular to decreases in the numbers of heavy smokers in Finland and of moderate smokers in Denmark, Norway and Sweden. The proportions of "ever smokers" among men in 1985 ranged from about 57% in Sweden to almost 70% in Iceland. In that year, Sweden and Finland had the lowest proportions of current male smokers (about 30%) and Denmark, Iceland and Norway the highest (about 40--50%). In all countries except Denmark, the proportions of for- mer smokers were relatively large (about 30%). The consistent increase over time in the proportion of "never smokers" indicates that fewer and fewer young men are taking up the smoking habit. Among women (Table 2.3), the proportion of"ever smokers" remained largely unchanged over the period 1965-85, the proportions in 1985 ranging from approximately 35% in Finland to 60% in Ice- land. The highest proportion of current smokers (43%) in 1985 was found for Danish women, ap- proaching that among men (51%). Relative risks for cancer Risk estimates for cancers known or suspected to be causally related to tobacco smoking were reviewed on the basis of the IARC monograph on tobacco smoking (11) and subsequent papers on this issue. Particular emphasis was paid to large cohort and case-control studies from the United Kingdom and the United States, and, when available, the Nordic countries, and the most representative relative risks for each cancer site were assigned to each of a num- ber of predefmed smoking categories. Study-specific relative risks associated with cigarette consumption (dose-response curves) are presented site by site be- low; summary estimates are shown in Table 2.4. The relative risks for "pipe and cigar smokers only" are rough estimates of the average risks associated with the speetlie smoking habit. Rough estimates of the relative risks of current smokers for cancer types only suspected to be related to tobacco smoking are shown in Table 2.5. All of the caleuiations are based on the assumption that the relative risks are equal for male and female smokers of equal amounts of tobacco. 11

DREYER et a~ TABLE 2.3. Distribution of the adult female population of each of the five Nordic countries according to smoking habits in 1965, 1975 and 1985 Country Year Never Female smokers (%) smokers Ever Former Current Moderatea Heavyb Denmark 1965 50 50 8 42 37 5 1975 41 59 12 47 37 10 1985 51 49 6 43 28¢ 15a Finland 1965 67 33 10 23 17 6 1975 71 29 13 16 9 7 1985 67 33 15 18 14c 4d Iceland 1965 42 58 13 45 26 19 1975 40 60 20 40 22 18 1985 41 59 22 37 32c 5a Norway 1965 63 37 14 23 22c 1a 1975 59 41 9 32 29c 3d 1985 50 50 18 32 27c 5d Sweden 1965 57 43 20 23 21 2 1975 49 51 17 34 18c 16f 1985 54 46 20 26 14~ 12r Data for 1965, 1975 and 1985 or a close year I| l ! II 1 Current smokers of fewer than 15 cigarettes per day and smokers of pipe or cigars only Current smokers of 15 cigarettes per day or more Current smokers of fewer than 20 cigarettes per day and smokers of pipe or cigars only Current smokers of 20 cigarettes per day or more Current smokers of 12 or fewer cigarettes per day and smokers of pipe or cigars only Current smokers of more than 12 cigarettes per day TABLE 2.4. Summary estimates of the relative risks for cancers known to be causally related to tobacco smoking, by smoking status and number of cigarettes smoked per day Site of cancer Never Former Current smokers smokers smokers No. of cigarettes smoked per day Pipe and cigar 1-9 10-19 20-39 >40 sm°kersa Lung 1.0 5.0 4.6 11.5 22.4 30.0 7.0 ! i 1 ! Larynx 1.0 2.5 2.5 5.4 9.9 13.0 3.5 Oesophagus 1.0 2.0 1.8 3.4 5.9 7.7 3.0 Pharynx 1.0 2.0 1.6 2.9 4.9 6.3 3.5 Mouth and tongue 1.0 2.0 1.6 2.9 4.9 6.3 3.5 Pancreas 1.0 1.0 1.2 1.6 2.1 2.5 1.0 Lower urinary tractb 1.0 1.5 1.5 2.5 4.0 5.0 1.5 Renal parenehyma 1.0 1.0 1.2 1.5 1.9 2.2 1.0 Comprises smokers of pipes, cigars and cigarillos only Includes cancers of the urinary bladder and renal pelvis RESULTS Lung cancer (ICD-7: 162) Lung cancer is a major global health problem and is de facto the commonest malignancy, having surpassed stomach cancer in the early 1980s (32). Lung cancer is particularly prevalent in industrialized countries; in Europe, it ae- 12 counts for 29% of all cancer deaths among men and 8% among women (33). Large variations are seen across country borders, between the two sexes and with age. In the Nordic countries combined, lung cancer is the second most fre- quent cancer in men, after cancer of the pros- tare, and the third most frequent cancer in women, only exceeded by breast cancer and co- I ! I

| ] ] ! i I i i i I I 1 TOBACCO SMOKING TABLE 2.5. Estimated relative risks for cancers at sites suspected to be related to tobacco smoking Cancer site Non-smokersI Current smokers Lip 1.0 2.0 Liver 1.0 1.5 Stomach 1.0 1.5 Leukaemia 1.0 1.3 Uterine cervix 1.0 2.0 a Never and former smokers lorectal cancer (34). In these countries nearly 10,000 new eases occur annually, approximately 7000 in men and 3000 in women (34--44). Relationship with tobacco consumption The predominant risk factor for lung cancer is cigarette smoking, but industrial exposures and radiation also play important roles (13). The combined results of the most important co- hort studies (Fig. 2.1; 2,8,17,45-52) give a risk for lung cancer that rises linearly, with an ap- proximate excess relative risk of 0.73 per ciga- rette smoked per day. Accordingly, by increasing the number of cigarettes smoked per day to 5, 10, 20 and 40 or more, the risk increases 5.0-, 8.0-, I6- and 30-fold, respectively, over that of "never-smokers". The risk for lung cancer is highest among cigarette smokers, but is also sig- nificantly increased among.smokers of pipes and cigars (Table 2.4; l l,13J. Former smokers have been allocated an average relative risk for lung cancer of 5 frab.le 2.4; 11). Tables 2.6 (men) and 2.7 (women) show the numbers of cases of lung cancer notified to each of the five Nordic cancer registries during 1980 and 1990 and the numbers estimated to occur in year 2000. The tables also give the corresponding crude incidence rates and the calculated pro- portions of cases caused by tobacco smoking (PAR%). While Finnish men had the highest numbers in 1980, with a crude incidence rate of 86 per 100 000 male inhabitants, Danish men took over this position in the next few decades, and the rate expected in 2000 is dose to 100 new eases per year per 100 000 inhabitants. The rate among Swedish men is about half of that seen in Danish men. The etude incidence rates of lung cancer among women (Table 2.7) are usually less than half those seen in the respective male popu- lations, but are dearly increasing over time. Large variations are seen between countries, however, women in Iceland and Denmark having the highest rates throughout the period. The pro- portion of lung cancers due to tobacco smoking among men in the Nordic countries in 1980 was 86%, and this percentage was unchanged in the estimate for 2000. Thus, 5,500-6,000 of the cases of lung cancer diagnosed annually among Nord- ic men in late 1970s and 1980s were due to smok- ing. In the year 2000, the number will be about 6,800 annually (Table 2.6). The corresponding PAR% were and will be somewhat lower in women (Table 2.7). The pro- portion of lung cancers caused by tobacco smoking was 72% around 1980 and 78% in 1990; around 2000, the proportion of lung can- cer among women due to tobacco smoking is expected to be 79%. These estimates are equiva- lent to about 1,300, 2,200 and 3,500 cases of lung cancer annually, respectively. Laryngeal cancer (ICD-7: 161) Laryngeal cancer predominates among men, with an estimated worldwide male:female inci- dence ratio of 7:1 (32). High incidences are re- ported from southern Brazil, Italy, France and Spain (32), while the rates in the Nordic coun- tries are much lower (34-44): in the combined Nordic populations of about 22.5 million in- habitants, about 600 eases were seen among men in 1990, and fewer than 100 among women. Relationship with tobacco smoking Tobacco smoking and alcohol drinking are important risk factors for laryngeal cancer, par- tieularly when combined (53-55); occupational hazards, including exposure to asbestos, may also be of some importance (32). Fig. 2.2 shows the dose-response relationships between ciga- rette smoking and laryngeal cancer, observed in a number of studies of relevance to the Nordic countries (53-64). As alcohol is also a strong risk factor, only studies in which adjustment was made for the effect of alcohol are included. Most of the studies reported an increase in the relative risk for laryngeal cancer that is approxi- mately proportional to the number of cigarettes smoked per day. In general, by increasing the number of cigarettes smoked per day to 5, 10, 20 and 40 or more, the risk for laryngeal cancer increases 2.5-fold, 4.0-, 7.0- and 13-fold, respec- tively, over that of "never-smokers", which corresponds to an excess relative risk of 0.30 per cigarette smoked per day. 13

DREYER ~t ~/, Relative risk 40 30 10 O; 5 10 15 20 25 30:35 No. of cigarettes/day 40 The overall relative risk for laryngeal cancer among smokers of pipes and cigars only is about 3.5 (53, 59, 62, 65), and the correspond- ing risk of the combined group of former smokers is about 2.5 (Table 2.4; 58, 59, 61, 62). Tables 2.8 (men) and 2.9 (women) give the key figures for laryngeal cancer in the Nordic popu- lations. A slightly increasing trend in the crude incidence rates for both men and women is seen over time, Danish men having the highest esti- mated rate of dose to 10 per 100 000 in year 2000. About 70% of the cases of laryngeal can- car in men and 60% in women that occur today and are foreseen to be diagnosed around the turn of the century in the Nordic counties, are caused by tobacco smoking, implying tobacco- 14 Fig. 2.1. Cohort studies: cigarette smoking and lung "°- Carstensen et al (45) " -I- Doll at al (2) ~."" Engholm et al (46) "~" Garlinkel et al (17) "~" Hammond et al (8) "~" Hammond et al (47) "~ Hakulinen at al (48) -e- Hakullnen et al (48) "~" Losslng et al (49) "~" Lund et al (50) "*" McLsughlin at al (51) • 4- Weir et sl (52) "~" Our estimate related PAR% that are almost as high as those for lung cancer. The findings indicate that around the year 2000, about 500 cases of laryngeal cancer will occur annually among Nordic men and 70 among Nordic wgmen, due to tobacco smoking. Oesophageal cancer (ICD-7: 150) Oesophageal cancer is characterized by an ex- treme diversity of rates throughout the world; there are usually more eases among men than women (32). Oesophageal cancer is rare in the Nordic countries in comparison with other parts of the world, such as Asia (34-44). In 1990, about 700 new cases among men and 300 I I

TOBACCO SMOKING a 1 ! ! n i TABLE 2.6. Proportions and numbers 1980, 1990 of cases of lung cancer avoidable among men in the Nordic and 2000 if tobacco smokin~ were eliminated countries in Country and year Annual no. Incidence PAR%• Avoidable of new cases per 100,000 number Nordic countries, 1980 6906 62.2 86 5925 Denmark 2130 84.2 86 1830 Finland 1983 85.8 89 1765 Iceland .42 36.5 87 35 Norway 962 47.5 85 815 Sweden 1789 43.4 83 1480 Nordic countries, 1990 6679 58.6 84 5635 Denmark 2018 79.7 84 1705 Finland 1629 67.5 86 1395 Iceland 61 47.7 86 50 Norway 1159 55.3 84 980 Sweden 1812 42.9 83 1505 Nordic countries, 2000 8102 68.6 84 6815 Denmark 2520 97.4 85 2150 Finland 1879 76.0 86 1620 Iceland 91 64.7 86 80 Norway 1313 59.9 84 1105 Sweden - 2299 52.1 81 1860 I I I I i I Population attributable risk percent TABLE 2.7. Proportions and numbers of eases of lung cancer avoidable among women in the Nordic countries in 1980, 1990 and 2000 if tobacco smokinf~ were eliminated Country and year Annual no. Incidence PAR.%" Avoidable of new eases per 100,000 number Nordic countries, 1980 1857 16.2 72 1330 Denmark 708 27.3 77 545 Finland • 299 12.1 71 210 Iceland 32 28.3 83 25 Norway 235 11.4 67 155 Sweden 583 13.9 68 395 Nordic countries, 1990 2826 24.0 78 2200 Denmark 1101 42.2 80 880 Finland 380 14.9 71 270 Iceland 38 29.9 83 30 Norway 458 21.4 75 345 Sweden 849 19.6 79 675 Nordic countries, 2000 4437 36.4 79 3490 Denmark 1800 67.6 82 1475 Finland 532 20.3 71 380 Iceland 75 53.8 83 65 Norway 624 27.9 79 495 Sweden 1406 31.1 76 1075 ! I • Population attributable risk percent among women were registered in the Nordic countries. Relationship with tobacco smoking Tobacco smoking and alcohol consumption are the major risk factors for cancer at this site in both men and women, especially when prac- tised in combination (21). In a number of studies in which alcohol consumption was ad- justed for, the magnitude of the increased risk for oesophageal cancer, appeared to be directly proportional to the number of cigarettes 15

DREYER ~ .L 4O 35 30 25 20 15 Relative risk Fig. 2.2. Case-control studies: cigarette smoking and la- ryngeal cancer, ad- justed for alcohol consumption. 5 0L 5 10 15 20 25 30 35 No. of cigarettes/day 4O smoked per day (54, 59, 66-73). Fig. 2.3 shows the dose-response relationships observed in rel- evant studies, with an excess relative risk of about 0.17 per cigarette smoked per day. Ac- cordingly, by increasing the number of cigarettes smoked daily to 5, 10, 20 and 40 or more, the risk increases by 1.8-, 2.%, 4.4- and 7.7- fold, respectively, over that of "never-smokers". Smoking of pipes and cigars only appears to increase the relative risk for oesophageal cancer to approximately 3 (Table 2.4; l l, 66, 67, 73, 74). Former smokers have an excess risk be- tween that of current smokers and "never smokers" (59, 66, 69, 70), estimated to be in the order of 2 (Table 2.4). 16 Tables 2.10 (men) and 2.11 (women) show that the crude rates of oesophageal cancer in the Nordic countries are almost unchanged between 1980 and 2000, with annual incidence rates of about 6 cases per 100,000 for men and 3 per 100,000 for women, respectively. Danish men and Finnish and Icelandic women have the highest rates. About 60% of the cases in men and 40% in women are attributable to smoking habits, corresponding to nearly 500 cases in the Nordic countries per year. Pharyngeal cancer (ICD-7: 14.5-148) Pharyngeal cancer is a relatively rare cancer in the Nordic countries (32), approximately 350 C) I I I ! I I I I I I I I I I I I I I

TOBACCO SMOKINO I I I I I TABLE 2.8. Proportions and numbers of eases of laryngeal cancer avoidable among men in the Nordic countries in 1980, 1990 and 2000 ~ tobacco smoking were eliminated Country and year Annual no. InCidence PAR%• Avoidable of new cases per' I00,000 number Nordic countries, 1980 590 5.3 71 420 Denmark 207 8.2 72 150 Finland 130 5.6 77 100 Iceland 6 5.2 74 4 Norway 85 4.2 70 60 Sweden 162 3.9 66 105 Nordic countries, 1990 591 5.2 69 405 Denmark 211 8.3 69 145 Finland 106 4.4 71 75 Iceland 6 4.7 71 4 Norway 108 5.1 69 75 Sweden 160 3.8 67 105 Nordic countries, 2000 720 6.1 68 490 Denmark 254 9.8 71 180 Finland 106 4.3 72 75 Iceland 6 4.3 72 4 Norway 142 6.5 69 95 Sweden ° 212 4.8 63 135 I I I I I I ! • Population attributable risk percent TABLE 2.9. Proportions and numbers of cases of laryngeal cancer avoidable among women in the tries in 1980, 1990 and 2000 ~ tobacco smoking were eliminated Country and year Annual no. Incidence PAR%• of new cases per I00,000 Nordic court- Avoidable number Nordic countries, 1980 74 0.6 52 35 Denmark 28 1.1 58 15 Finland • 8 0.3 50 4 Iceland 2 1.8 67 1 Norway 14 0.7 45 5 Sweden 22 0.5 46 10 Nordic countries, 1990 88 0.7 60 55 Denmark 48 1.8 62 30 Finland 8 0.3 50 4 Iceland 0 0.0 - - Norway 9 0.4 56 5 Sweden 23 0.5 61 15 Nordic countries, 2000 104 0.9 62 65 Denmark 68 2.6 65 45 Finland 8 0.3 50 4 Iceland 0 0.0 - - Norway 4 0.2 61 2 Sweden 24 0.5 56 15 .i ! 1 • Population attributable risk percent eases having been registered in 1990 in men and 125 in women (34-44). Relationship with tobacco smoking Like other cancers of the upper aerodigestive tract, pharyngeal cancer is associated with both tobacco smoking and alcohol drinking, and the combination of the two habits increases the risk further (I 1, 75). Few studies have addressed the risk for pharyngeal cancer associated with reg- ular cigarette smoking. The main results from the available studies are summarized in Fig. 2.4 17

DREYER et al. Relative risk 40 ,.r----~ ~ 0L 0 10 20 :30 40 No. of cigarettes/day (55-57, 59, 75, 76), which shows marked vari- ation in the magnitude of the reported risks, after adjustment for alcohol drinking. The risk for pharyngeal cancer rises with the number of cigarettes smoked per day with an excess rela- tive risk of approximately 0.13 per cigarette smoked. Thus, by increasing the number of cigarettes smoked daily to 5, 10, 20 and 40 or more, the risk increases by 1.7-, 2.3-, 3.7- and 6.3-fold, respectively, over that of "never- smokers". Smokers of pipes and cigars only seemed to have approximately the same risk as smokers of cigarettes (11), with an estimated av- erage of 3.5 (Table 2.4). A relative risk of 2.0 was allocated to the group of former smokers. Tables 2.12 (men) and 2.13 (women) give the 18 I Fig. 2.3. Case-control i studies: cigarette smok- ing and oesophagcal cancer, adjusted for al- I cohol consumption. ! ~ ~m ,t=l (S41 I I "{" Brown it al (66) "~" Fr~lI~hl it a! ($9) I "J" Need it al (69) ~" Tuyn= ~t a1171 ) "~ Tuy~ it =1 (7=) • ~" oiff il||mltel ! I key figures for pharyngeal cancer. Over the study I period, the crude rate increases in men but not in women. The rates are particularly high for [] Danish and Swedish men and Danish women. The calculated PAR% for laryngeal cancer re- lated to tobacco smoking is about 50% in men [] and 40% in women, corresponding to annual | expected numbers of 250 eases among men and 50 among women in the Nordic countries around the year 2000. • Cancers of the oral cavity and tongue (ICD-7; I 141 and 143-144) Cancers of the oral cavity and tongue are also rare in the Nordic countries, with approximate- I II I

i i i i i i i i i i 1 I i I I I TOBACCO SMOKING TABLE 2.10. Proportions and numbers of cases of oesophageal cancer avoidable among men in countries 1980, 1990 and 2000 if tobacco srnokin~ were eliminated Country and year Annual no. Incidence PAR%• per 100,000 of new eases Nordic countries, I980 541 4.9 59 Denmark 111 4.4 62 . Finland 107 4.6 66 Iceland 4 3.5 64 Norway 83 4. I 58 Sweden 236 5.7 55 Nordic countries, 1990 672 5.9 57 Denmark 191 7.5 58 Finland 128 5.3 59 Iceland 10 7.8 61 Norway 114 5.4 57 Sweden 229 5.4 54 Nordic countries, 2000 693 5.9 55 Denmark 194 7.5 58 Finland 103 4.2 59 Iceland 6 4.3 60 Norway 96 4.4 57 Sweden ° 294 6.7 51 the Nordic Avoidable number 325 70 70 3 50 130 380 110 75 5 65 125 385 115 60 4 55 150 Population attributable risk percent TABLE 2.11. Proportions and numbers of eases of oesophageal cancer avoidable among women countries in 1980, 1990 and 2000 if tobacco srnokin~ were eliminated Country and year Annual no. Incidence PAR%" per 100,000 of new cases Nordic countries, 1980 305 2.7 38 Denmark 68 2.6 46 Finland ~. 121 4.9 37 Iceland 2 1.8 53 Norway 26 1.3 32 Sweden 88 2.1 34 Nordic countries, 1990 301 2.6 44 Denmark 74 2.8 49 Finland 88 3.4 37 Iceland 4 3.2 53 Norway 40 1.9 42 Sweden 95 2.2 47 Nordic countries, 2000 332 2.7 44 Denmark 75 2.8 52 Finland 108 4.1 37 Iceland 6 4.3 54 Norway 44 2.0 48 Sweden 99 2.2 43 in the Nordic Avoidable number 115 30 45 1 10 30 125 35 30 2 15 45 145 40 40 3 20 40 • Population attributable risk percent ly 500 eases among men and 350 eases among women in 1990 (34-44). Relationship with tobacco smoking Tobacco smoking and alcohol consumption are known to be important risk factors for these turnout types, particularly when practised in combination (77, 78). Chewing of tobacco pro- ducts is also a risk factor (11). The risk esti- mates from the available studies (Fig. 2.5; 56, 57, 77-82), show a rise with the number of ciga- rettes smoked with relative risks of 1.7, 2.3, 3.7 19

DREYER et al. Relative risk 40: 2o! ........................ ~---+ ...... t ,V' ,-" 5; • -{- - - ..--:'- ; -'- 0 10 20 30 40 No. of cigarettes/day and 6.3 associated with daily smoking of 5, 10, 20 and 40 or more cigarettes. This corresponds to an excess relative risk of 0.13 per cigarette smoked per day. Former smokers have a relative risk of 2.0 and smokers of pipes and cigars only one of 3.5 (Table 2.4). The epidemiological characteristics of can- cers of the oral cavity and tongue are given in Tables 2.14 (men) and 2.15 (women). As seen for cancer of the pharynx, the crude rate among men increases while that of women is almost un- changed. Particularly high rates were registered in Danish and Norwegian men and in Danish women. About 50% of cancers of the oral cavity and tongue in men and 40% in women are 20 Fig. 2.4. Case-control studies: cigarette smok- ing and pharyngeal can- cer, adjusted for alcohol consumption. caused by smoking, which corresponds to an ex- pected annual total of about 500 cases around the year 2000. Pancreatic cancer (ICD-7: 157) Pancreatic cancer is moderately frequent on a worldwide basis and in the Nordic countries (32). In 1990, about 1400 cases were seen in men and 1700 in women (34-44). Relationship with tobacco smoking Tobacco smoking is the only environmental factor that has been consistently associated with an increased risk (22), and the risk appears to rise linearly with the number of cigarettes ! i I I I I I I I I I I I I I I I

TOBACCO SMOKING i I TABLE 2.12. Proportions and numbers of cases of pharyngeal cancer avoidable among men in the Nordic coun- tries in 1980, 1990 and 2000 if tobacco smokinf, were eliminated Country and year Annual no. Incidence PAR%• Avoidable of new cases per I00,000 number Nordic countries, 1980 270 2.4 57 150 Denmark 61 2.4 61 35 Finland 39 1".7 61 25 Iceland 3 2.6 63 2 Norway 35 1.7 56 20 Sweden 132 3.2 54 70 Nordic countries, 1990 347 3.0 54 190 Denmark 99 3.9 57 55 Finland 51 2.1 55 30 Iceland 1 0.8 61 1 Norway 53 2.5 55 30 Sweden 143 3.4 51 75 Nordic countries, 2000 456 3.9 52 235 Denmark 129 5.0 55 70 Finland 54 2.2 55 30 Iceland 1 0.7 57 1 Norway 83 3.8 54 45 Sweden ° 189 4.3 47 90 Population attributable risk percent TABLE 2.13. Proportions and numbers of eases of pharyngeal cancer avoidable among women in the Nordic countries in 1980, 1990 and 2000 i/tobacco smokin~ were eliminated Country and year Annual no. Incidence PAR%a of new cases per 100,000 Nordic countries, 1980 132 1.2 34 45 Denmark • 34 1.3 43 15 Finland 30 1.2 32 10 Iceland 2 1.8 49 1 Norway 21 1.0 28 5 Sweden 45 I.I 31 15 Nordic countries, 1990 124 1.1 40 50 Denmark 34 1.3 46 15 Finland 27 1.1 33 10 Iceland 0 0.0 - - Norway 16 0.7 38 5 Sweden 47 1.1 42 20 Nordic countries, 2000 121 1.0 41 50 Denmark 42 1.6 47 20 Finland 17 0.6 33 5 Iceland I 0.7 49 0 Norway 22 1.0 44 I0 Sweden 39 0.9 38 15 Avoidable number I I • Population attributablerisk percent smoked per day (Fig. 2.6; 2, 45, 46, 51, 83--91). Thus, by increasing the number of eighrettes smoked daily to 5, 10, 20 and 40 or more, the risk increases by 1.2-, 1.4-, 1.8- and 2.5-fold, re- speetively, over that of "never-smokers". This corresponds to an excess relative risk of 0.04 per cigarette smoked per day. Former smokers (2, 83) and smokers of pipes and cigars only (91, 92) do not have a detectable increase in risk (Table 2.4). 21

DREYER et al. • Fig. 2.5. Case-con- Relative risk trol studies: ciga- 40 r---'--------- ~ ~ rctte smoking and ~ | cancers of the oral ~ / cavity and tongue, [ | adjusted for alcohol :35 t ................................. ....[ consumption. 0 1 0 20 :30 40 No. of cigarettes/day Tables 2.16 (men) and 2.17 (women) give Cancers of the lower urinary tract (ICD-7: 181, the key figures for pancreatic cancer in the 180 partly) Nordic countries. The crude rate for women The,,. term "cancers of the lower urinary" increases between 1980 to 2000, but no clear tract includes both cancers and bemgn papil- trend is sccn for men. As the association with lomas of the urinary bladder, renal pelvis, cigarette smoking is weaker than that for ureter and urethra. Cancer of the bladder is cancers of the respiratory trac~, the PAR% for pancreatic cancer is about 17% in men and 10-12% in women, corresponding to an annual, expected number of 550 pancreatic cancers in men and women around the y~ar 2000. 22 relatively common, ranking eleventh on a glo- bal basis (32). Cancers of the lower urinary tract arc particularly common in the Nordic countries, with approximately 4150 new cases in men and 1550 in women in 1990 (34- I I

TOBACCO SMOKING TABLE 2.14. Proportions and numbers in the Nordic countries in Country and year Annual no. of new cases of cases of cancer of the oral cavity and tongue avoidable among men 1980, 1990 and 2000 if tobacco smoking were elimi hated Incidence PAR%~ • Avoidable per 100,000 number Nordic countries, 1980 388 3.5 57 220 Denmark 86 3.4 61 55 Finland 52 2.3 61 30 Iceland 3 2.6 63 2 Norway 83 4.1 56 45 Sweden 164 4.0 54 90 Nordic countries, 1990 449 3.9 54 245 Denmark 139 5.5 57 80 Finland 65 2.7 55 35 Iceland 6 4.7 61 4 Norway 86 4.1 55 45 Sweden 153 3.6 51 80 Nordic countries, 2000 611 5.2 52 315 Denmark 191 7.4 55 105 Finland 74 3.0 55 40 Iceland 4 2.8 57 2 Norway 130 5.9 54 70 Sweden ° 212 4.8 47 100 • Population attributable risk percent TABLE 2.15. Proportions and numbers of cases of cancer of the oral cavity and tongue avoidable among women in the Nordic countries in 1980, 1990 and 2000 if tobacco smokin$ were eliminated Country and year Annual no. Incidence PAR%• Avoidable of new cases per 100,000 number Nordic countries, 1980 268 2.3 34 90 Denmark • 70 2.7 43 30 Finland 48 1.9 32 15 Iceland 0 0.0 - - Norway 54 2.6 28 15 Sweden 96 2.3 31 30 Nordic countries, 1990 341 2.9 40 140 Denmark 85 3.3 46 40 Finland 72 2.8 33 25 Iceland 1 0.8 49 0 Norway 51 2.4 38 20 Sweden 132 3.0 42 55 Nordic countries, 2000 363 3.0 41 145 Denmark 108 4.1 47 50 Finland 50 1.9 33 15 Iceland 5 3.6 49 2 Norway 77 3.4 44 35 Sweden 123 2.7 38 45 • Population attributable risk percent Relationship with tobacco smoking The predominant risk factor for these can- cars is cigarette smoking, but industrial ex- posures also play an important role in de- veloped countries. In Africa and the Eastern Mediterranean, schistosomal infection is of particular importance (32). There is some vari- ation in the magnitude of the reported risk for moderate smokers in particular, as illustrated in Fig. 2.7 (2, 45, 46, 50-52, 93-101). AI- 23

DREYER et al. Relative risk 40~ Bouohardy et al (83) Car=lenten et =! (45) C, uzlck at al (~4) Doll et al {2) Engholm et al (46) Falk ~ at (85) F=rrow ~ al (86) MacMzhon ¢ zl (88) McLlughltn mal (51) Sl~e~m~ ¢ ~1 (82) ~Wynder = 11 (~) +Zheng = =L 1~3 (91) ~Our estimate Fig. 2.6. Cohort and case-control studies: cigarette smoking and pan- erratic cancer. 0 10 2O 3O 4O NO. of cigarrets/day though most of the studies cited do not in- clude papillomas of the lower urinary tract, the smoking habits of patients with these neo- plasms appear to be similar to those of pa- tients with invasive bladder tumours (102, 103). For equal levels of tobacco smoking, the relative risks are generally higher for cancers of the renal pelvis and ureter than for those of the urinary bladder (11). Overall, by increasing the number of cigarettes smoked per day to 5, lO, 20 and 40 or more, the risk for cancers in the lower urinary tract increases by 1.5-, two, thr~e and fivefold, respectively, over that of "never-smokers", corresponding to an excess 24 relative risk of 0.10 per cigarette smoked per day. Smoking of pipes and cigars only also in- creases the relative risk for bladder cancer (45, 103-104), but at lower levels, most likely around 1.5 (Table 2.4). The group of former smokers was allocated an average relative risk for cancers of the lower urinary tract of 1.5 (Table 2.4). The rates of cancers of the urinary tract have and will increase during the last decades of this century (Tables 2.18 and 2.19). Denmark has remarkably high incidence rates in both men and women, that in men are pre- dicted to exceed 50 new cases per 100,000 per ( [ I i I ! I I ! 0 0~, ! I I I I I

I l i l l l i I I I TOBACCO SMO~LINO TABLE 2.16. Proportions and numbers of cases of pancreatic cancer avoidable among men in the Nordic countries in 1980, 1990 and 2000 if tobacco smokin8 were eliminated Country and year Annual no. Incidence PAR%= Avoidable of new cases Per lO0,O00 number Nordic countries, 1980 1588 14.3 18 280 Denmark 407 16.1 17 70 Finland • 254 11.0 28 70 Iceland 15 13.0 19 3 Norway 268 13.2 17 45 Sweden 644 15.6 14 90 Nordic countries, 1990 1389 12.2 17 230 Denmark 311 12.3 16 50 Finland 280 11.6 20 55 Iceland 18 14.1 15 3 Norway 246 11.7 16 40 Sweden 534 12.6 15 80 Nordic countries, 2000 1814 15.4 17 315 Denmark 376 14.5 20 75 Finland 423 17.1 22 90 Iceland 18 12.8 20 4 Norway 307 14.0 17 50 Sweden ° 690 15.6 14 95 Population attributable risk percent TABLE 2.17. Proportions and numbers of cases of pancreatic cancer avoidable among women in countries in 1980, 1990 and 2000 if tobacco smoking were eliminated Country and year Annual no. Incidence PAR%= per 100,000 of new eases Nordic countries, 1980 1489 13.0 9 Denmark , 389 15.0 12 Finland 287 11.6 I0 Iceland 14 12.4 18 Norway 220 10.7 8 Sweden 579 13.8 7 Nordic countries, 1990 1670 14.2 13 Denmark 359 13.8 14 Finland 404 15.8 10 Iceland 13 10.2 16 Norway 267 12.5 12 Sweden 627 14.5 15 Nordic countries, 2000 2030 16.7 12 Denmark 417 15.7 18 Finland 464 17.7 9 Iceland 16 11.5 18 Norway 340 15.2 14 Sweden 793 17.5 10 the Nordic Avoidable number 135 45 30 3 15 40 215 50 40 2 30 95 245 75 40 3 45 80 Population" attributable risk p~rcent year before the year 2000. The low incidence rate in Finland may well re, fleet less reporting of benign and borderline papillomas of the bladder. In 1990, about 41% of the male cases and 32% of the female cases were attributed to tobacco smoking, corresponding to more than 2000 cancers of the lower urinary tract in the Nordic countries combined. More than 2,500 25

Relative risk 40! No. of cigarettes/day cases are expected to occur in 2000 due to smoking in these populations. Renal cancer (ICD-7:180 partly) Cancers of the renal parenchyma, mainly re- hal-cell carcinomas, show moderate geographic variation, but the incidence rates in Eskimo populations and in the Scandinavian countries are among the highest in the world (32). In the Nordic countries in 1990, about 1500 cases were recorded in men and 1000 in women (34-44). Relationship with tobacco smoking Tobacco smoking is a risk factor for renal can- cer, but occupational exposures may also play a 26 Fig. 2. 7. Cohort and case-control studies: cigarette smoking and lower urinary tract cancer. role (32). The previously reported dose- response curves for cigarette smoking and cancer of the kidney (2, 45, 46, 51, 52, 105-112; Fig. 2.8) all show a modest increase, with overall relative risks of 1.2, 1.3, 1.6 and 2.2 for smokers of 5, 10, 20 and 40 or more cigarettes per day, respectively, when compared with the rates in neversmokers. This corresponds to an excess relative risk of 0.03 per cigarette smoked per day. Smokers of pipes and cigars only and former smokers have no measurable increase in risk (Table 2.4). The crude rates of renal cancer in the Nordic countrics have increased slightly since 1980 and will apparently continue to increase in the years to come (Tables 2.20 and 2.21). The recorded rates [ ! ! [ 1 ! I [

TOBACCO SMOKING TABLE 2.18. Proportions and numbers of cases of lower urinary tract cancer avoidable among men countries in 1980, 1990 and 2000 if tobacco smokinf, were eliminated Country and year Annual no. Incidence PAR%• per 100,000 in the Nordic Avoidabl~ of now cases number Nordic countries, 1980 3329 30.0 42 1410 Denmark 1054 41.7 43 455 Finland 364 15.8 53 195 Iceland 27 23.5 46 15 Norway 637 31.5 42 270 Sweden 1247 30.3 38 475 Nordic countries, 1990' 4146 36.4 41 1695 Denmark 1204 47.5 41 490 Finland 576 23.9 45 260 Iceland 36 28.1 43 15 Norway 794 37.9 41 330 Sweden 1536 36.3 39 600 Nordic countries, 2000 5067 42.9 41 2060 Denmark 1409 54.5 43 610 Finland 735 29.7 46 340 Iceland 49 34.9 45 20 Norway 980 44.7 41 405 Sweden " 1894 42.9 36 685 • Population attributable ri/s, k percent TABLE 2.19. Proportions and numbers of cases of lower urinary tract cancer avoidable among women in the Nordic countries in 1980, 1990 and 2000 if tobacco smoking were eliminated Country and year Annual no. Incidence PAR%= Avoidable of nvw eases per 100,000 number Nordic countries, 1980 Denmark Finland Iceland Norway Sweden Nordic countries, 1990 1238 10.8 26 315 389 15.0 31 120 146 5.9 25 35 13 11.5 41 5 211 10.2 22 45 479 11.4 22 110 1534 13.0 32 495 451 17.3 34 155 232 9.1 25 60 10 7.9 39 4 258 12.0 29 75 583 13.5 34 200 1867 15.3 32 600 554 20.8 38 210 264 10.1 25 65 17 12.2 41 5 333 14.9 34 I 15 699 15.5 30 205 Denmark Finland Iceland Norway Sweden Nordic countries, 2000 Denmark Finland Iceland Norway Sweden Population attributable risk p~rc~nt are highest in Sweden. The PAR% for renal cancer and smoking is about 14% in men and 10% in women, which reflects the moderate relationship seen between cigarette smoking and this type of cancer. Owing to the relatively high incidence of renal cancer in the Nordic countries, however, an- nually almost 400 cases per year can be attributed to the current smoking habits of the popula- tions. For all the cancer sites described above the 27

DR.EYER et el, Relative risk 40~ 0 10 20 30 40 No. of cigarettes/day Benhamou et al (I05) Brownson et al (106) Carstensen et =I (45) Doll et ad (2) Engholm et all (48) Hlatt et =I (107) "~'McLaughlin et al McLaughlin et al (109) Musoat et =I (110) Weir et el (52) "'Wynder et el (111) et =1 (112) ~"Yu "I~" °ur Fig. 2.& Cohort and. case-contro] studies: cigarette smoking and renal cancer. ! I I I I I ! i ! I I ! i cause-effect relationship with tobacco smoking has been confirmed; however, for the following sites a causal relationship has only been suspected. Lip cancer (ICD-7: 140) Lip cancers are not common in most popula- tions, although high rates are seen in Canada and in latitudes near the equator. High risks are also observed among persons living in rural areas and those with outdoor occupations (32). In 1990, about 500 cases were registered among men in the Nordic countries and 100 among women (34-44). 28 Relationship with tobacco smoking Data from epidemiological studies are con- sistent in demonstrating a causal relationship between exposure to sunlight and cancers of the lip (32). The effect of tobacco smoking on the risk for lip cancer has been investigated in only one cohort study; no association was found (50). Most case-control studies (ll, ll3), how- ever, have reported an increased risk for lip can- cer among pipe smokers, while the association between cigarette smoking and lip cancer has been less consistent (1 l, 56, 114). In this review, the risk for lip cancer has been I I

I i I I I I I TOBACCO SMOKING TABLE 2.20. Proportions and numbers of cases of renal cancer avoidable among men in the Nordic countries in 19~ tobacco smoking were eliminated Country and year Annual no. Incidence PAR%" Avoidable of new cases j~er 100,000 number Nordic countries, 1980 1346 12.1 14 190 Denmark 248 9.8 13 30 Finland 197 8.5 24 50 Iceland 11 9.6 16 2 Norway 207 10.2 15 30 Sweden 683 16.6 12 80 Nordic countries, 1990 1453 12.7 13 195 Denmark 279 11.0 12 35 Finland 275 11.4 17 45 Iceland 22 17.2 13 3 Norway 275 13.1 14 40 Sweden 602 14.2 11 70 Nordic countries, 2000 1801 15.3 14 260 Denmark 333 12.9 18 60 Finland 411 16.6 19 75 Iceland 19 13.5 16 3 Norway 317 14.5 14 45 Sweden " 721 16.3 10 75 Population attributable ris,k percent TABLE 2.21. Proportions and numbers of cases of renal cancer avoidable among women in the Nordic countries in 1980, 1990 and 2000 if tobacco smoking were eliminated Country and year Annual no. Incidence PAR°/o* of new cases per 100,000 Nordic countries, 1980 983 8.6 7 70 Denmark . 210 8.1 9 20 Finland 168 6.8 8 15 Iceland 5 4.4 16 1 Norway 156 7.6 7 10 Sweden 444 10.6 5 25 Nordic countries, 1990 1045 8.9 10 105 Denmark 215 8.2 10 20 Finland 206 8.1 8 15 Iceland 6 4.7 14 Norway 185 8.6 11 20 Sweden 433 10.0 11 50 Nordic countries, 2000 1292 10.6 10 125 Denmark 283 10.6 16 45 Finland 307 11.7 7 25 Iceland 15 10.8 15 2 Norway 186 8.3 12 20 Sweden 501 I 1.1 7 35 Avoidable number t Population attributable risk percent set at 2.0 for all current smokers compared with "never-smokers" and "former smokers" (Table 2.5). Tables 2.22 (men) and 2.23 (women) show that the crude incidenc~ rates of lip cancer de- crease throughout the study period for men, but no trend is seen for women. About 120 cases in men and 30 in women in the Nordic countries will probably bc attributable to tobacco smoking by the year 2000, corresponding to about 25% of male lip cancer cases and 20% of female cas~s. 29

DREYER et aL TABLE 2.22. Proportions and numbers of cases of lip cancer possibly avoidable among men in the Nordic coun- tries in 1980, 1990 and 2000 if tobacco smoking were eliminated Country and year Annual no. Incidence PAR%a Avoidable of new eases per 100,000 number Nordic countries, 1980 589 5.3 36 210 Denmark 143 5.7 41 60 Finland 149 6.4 33 50 Iceland 5 4.3 39 2 Norway 111 5.5 36 40 Sweden 181 4.4 34 60 Nordic countries, 1990 460 4.0 30 135 Denmark 110 4.3 38 40 Finland 121 5.0 24 30 Iceland 3 2.3 35 1 Norway 86 4.1 31 25 Sweden 140 3.3 28 40 Nordic countries, 2000 451 3.8 26 120 Denmark 89 3.4 34 30 Finland 130 5.3 25 35 Iceland 6 4.3 30 2 Norway 83 3.8 28 25 Sweden 143 3.2 21 30 Population attributable risk percent TABLE 2.23. Proportions and numbers of cases of lip cancer possibly avoidable among women in countries in 1980, 1990 and 2000 if tobacco smokinf were eliminated Country and 3~ear Annual no. PAR°/ot of new cases the Nordic Incidence Avoidable per 100,000 number Nordic countries, 1980 104 0.9 22 20 Denmark 28 1.1 29 10 Finland 34 1.4 19 5 Iceland 0 0.0 - - Norway 13 0.6 19 2 Sweden 29 0.7 19 5 Nordic countries, 1990 115 1.0 23 25 Denmark 19 0.7 32 5 Finland 36 1.4 14 5 Iceland 1 0.8 28 0 Norway 23 1.1 24 5 Sweden 36 0.8 25 10 Nordic countries, 2000 126 1.0 20 30 Denmark 10 0.4 31 3 Finland 38 1.5 15 5 Iceland 2 1.4 27 1 Norway 33 1.5 24 10 Sweden 43 1.0 19 10 Population attributable risk percent Liver cancer (ICD-7: 155.0) Cancer of the liver is rare in the Nordic coun- tries in comparison with many other areas of the world, where the prevalence of chronic ear- 30 riers of hepatitis B surface antigen and/or ex- posure to aflatoxins is common (32). In the Nordic countries in 1990, about 600 eases were registered in men and 400 in women (34-44).

rl "1 TOBACCO SMOKING TABLE 2.24. Proportions and numbers of cases of liver cancer possibly avoidable among men in countries in 1980, 1990 and 2000 if tobacco smoking, were eliminate Country and year Annual no. Incidence PAR%a per 100,000 the Nordic Avoidable of new eases number Nordic countries, 1980 584 5.3 22 130 Denmark 145 5.7 26 40 Finland 98 4.2 20 20 Iceland 3 2.6 24 I Norway 53 2.6 22 10 Sweden 285 6.9 21 60 Nordic countries, 1990 613 5.4 17 100 Denmark 137 5.4 24 30 Finland 127 5.3 14 15 Iceland 5 3.9 21 1 Norway 52 2.5 18 10 Sweden 292 6.9 16 45 Nordic countries, 2000 642 5.4 15 o 95 Denmark 129 5.0 21 25 Finland 156 6.3 15 25 Iceland 7 5.0 17 1 Norway 51 2.3 16 10 Sweden o 299 6.8 12 35 i i I 1 .! Population attributable risk percent TABLE 2.25. Proportions and numbers of eases of liver cancer possibly avoidable among women in the Nordic countries in 1980, 1990 and 2000 if tobacco smokin$ were eliminated Country and year Annual no. Incidence PAR%a Avoidable of new eases per 100,000 number Nordic countries, 1980 407 3.6 12 55 Denmark • 105 4.0 17 20 Finland 77 3.1 10 10 Iceland 7 6.2 18 1 Norway 39 1.9 10 4 Sweden 179 4.3 10 20 Nordic countries, 1990 430 3.7 14 55 Denmark 93 3.6 19 15 Finland 102 4.0 8 10 Iceland 5 3.9 16 1 Norway 43 2.0 14 5 Sweden 187 4.3 15 25 Nordic countries, 2000 453 3.7 11 50 Denmark 81 3.0 18 15 Finland 127 4.9 8 10 Iceland 3 2.2 16 0 Norway 47 2.1 14 5 Sweden 195 4.3 10 20 Population attributable risk tx:rcent Relationship with tobacco smoking Hepatocellular carcinomas have repeatedly been related to smoking, but no definite con- elusion has been reached because of the many confounding factors, such as alcohol, hepatitis B virus and aflatoxins (11, 32). An additional methodological problem of many of the studies has been the potential inclusion of histologically unconfirmed eases, some of which might represent metastatic spread to the 31

DREYER et al. TABLE 2.26. Proportions and numbers of cases of stomach cancer possibly avoidable among men in the Nordic countries in 1980, 1990 and 2000 if tobacco smokin~ were eliminated Country and year Annual no. Incidence PAR%a Avoidable of new cases per 100,000 number Nordic countries, 1980 3018 27.2 22 665 Denmark 574 22.7 26 150 Finland .678 29.3 20 135 Iceland 32 27.8 24 I0 Norway 595 29.4 22 130 Sweden 1139 27.7 21 240 Nordic countries, 1990 2380 20.9 17 405 Denmark 385 15.2 24 90 Finland 570 23.6 14 80 Iceland 34 26.6 21 5 Norway 532 25.4 18 95 Sweden 859 20.3 16 135 Nordic countries, 2000 2322 19.7 15 345 Denmark 438 16.9 21 90 Finland 560 22.6 15 80 Iceland 40 28.4 17 5 Norway 422 19.3 16 65 Sweden 862 19.5 12 105 ! Population attributable risk percent TABLE 2.27. Proportions and numbers of cases of stomach cancer possibly avoidable among women countries in 1980, 1990 and 2000 if tobacco smokin~ were eliminated Country and year Annual no. Incidence PAR°/oI per 100,000 of new cases in the Nordic Avoidable number Nordic countries, 1980 2072 18.1 12 245 Denmark 356 13.7 I7 60 Finland 572 23.2 10 60 Iceland 17 15.0 18 3 Norway 418 20.3 10 45 Sweden 709 16.9 10 75 Nordic countries, 1990 1713 14.6 13 225 Denmark 287 11.0 19 55 Finland 489 19.2 8 35 Iceland 18 14.2 16 3 Norway 341. 15.9 14 45 Sweden 578 13.3 15 85 Nordic countries, 2000 1708 14.0 11 195 Denmark 278 10.4 18 50 Finland 491 18.8 8 40 Iceland 17 12.2 16 3 Norway 319 14.2 14 45 Sweden 603 13.3 10 55 • Population attributable risk percent liver of tobacco-induced cancers in other or- gans. The risk for hepatocellular carcinoma of cur- rent smokers in comparison with "never smokers" and former smokers has been set at 32 1.5 (Table 2.5). The crude rates of liver cancer appears to be stable over time (Tables 2.24 and 2.25). The highest reported rates are those in Sweden. By 2000 about 15% of the liver cancers in Nordic men and 11% of the female cases may

1 1 I 1 I I TOBACCO SMOKING bc attributable to smoking, corresponding to about I00 cases in men a.nd 50 cases in women, annually. Gastric cancer (ICD-7: 151) Gastric cancer is frequent worldwide, with particularly high rates in Japan and other areas of eastern Asia (32). A remarkable feature of the cpidcmiology of gastric cancer in most parts of Europe and North America is the decline in its incidence over the last four to five decades (32). In the Nordic countries in 1990, about 2,400 cases were recorded in men and 1,700 in women (34-44). Relationship with tobacco smoking Cohort studies of the risks for cancer among smokers also addressed the risk for gastric cancer (2, 50, 51, 115). Some (116, 117) but not all (118, 119) oase-control studies showed a dose-response relationship with smoking. It was shown recently that infection with a bacterium, Helicobacter pylori, is the main risk factor for gastric cancer (120). This bacterium is a potential confounder which has not been adjusted for in these studies. In this review, the rate of gastric cancer in current smokers has been set at 1.5 times the rate for non-smokers (Table 2.5). It can be seen in Tables 2.26 (men) and 2.27 (women) that the crude rates of gastric cancer decrease throughout the study period. About 15% of the eases in men and 11% of those in women in the year 2000 may be attributable to smoking. This corresponds to an expected num- ber of about 350 cancers in men and 200 in women, annually. Cancer of the uterine cervix (ICD-7: 171) Worldwide, cancer of the uterine cervix is the second most common cancer among women. The regions of highest risk are sub-Saharan Africa, Central and South America and Sou- theast Asia (32). During 1990, about 1,600 new cases of cervical cancer were registered among women in the Nordic countries (34--44). Relationship with tobacco smoking Many recent studies have found that to- bacco smoking increases the risks for both intraepithelial neoplasia and invasive cervical cancer; in addition many studies have shown a dose-response relationship with tobacco con- sumption, a reduction in the relative risk among former smokers and a correlation be- tween clearly tobacco-related cancers and cer- vical cancer (121-126"). Nevertheless, the avail- able studies do not allow us to conclude TABLE 2.28. Proportions and numbers of cancers of the uterine cervix possibly avoidable in the Nordic countries in 1980, 1990 and 2000 if tobacco smokinf were eliminated Country and year Annual no. Incidence PAR°A• Avoidable of new cases per I00,000 number Nordic countries, 1980 1679 14.7 23 380 Denmark 589 22.7 29 175 Finland 181 7.3 19 35 Iceland 7 6.2 31 2 Norway 386 18.7 19 75 Sweden 516 12.3 19 95 Nordic countries, 1990 1585 13.5 26 415 Denmark 540 20.7 32 I70 Finland 139 5.4 14 20 Iceland 16" 12.6 28 4 Norway 401 18.7 24 95 Sweden 489 11.3 25 125 Nordic countries, 2000 1189 9.8 24 285 Denmark 406 15.2 31 125 Finland 91 3.5 15 15 Iceland 16 11.5 27 4 Norway 256 11.4 24 60 Sweden 420 9.3 19 80 • Ppulation attributable risk percent 33

DREYER et aL TABLE 2.29. Proportions and numbers of eases of leukaemia possibly avoidable among men in the Nordic coun- tries in 1980, 1990 and 2000 if tobacco smokinf, were eliminated Country and year Annual no. Incidence PAR%~ Avoidable of new eases per 100,000 number Nordic countries, 1980 1278 11.5 15 185 Denmark 329 13.0 17 .55 Finland 208 9.0 13 25 Iceland 11 9.6 16 2 Norway 193 9.5 14 30 Sweden 537 13.0 14 75 Nordic countries, 1990 1180 10.4 12 130 Denmark 330 13.0 16 50 Finland 198 8.2 9 15 Iceland 10 7.8 14 I Norway 180 8.6 12 20 Sweden 462 10.9 10 45 Nordic countries, 2000 1082 9.2 10 105 Denmark 331 12.8 14. 45 Finland 188 7.6 9 15 Iceland 9 6.4 11 1 Norway 167 7.6 10 15 Sweden 387 8.8 7 30 [ [ [ [ ( [i ( • Population attributable risk percent TABLE 2.30. Proportions and numbers of cases of leukaemia possibly avoidable among women in countries in 1980, 1990 and 2000 if tobacco smokin~ were eliminated Country and year Annual no. Incidence PAR%• per 100,000 of new cases the Nordic Avoidable number Nordic countries, 1980 954 8.3 8 70 Denmark 248 9.6 11 25 Finland 155 6.3 6 10 Iceland 5 4.4 12 1 Norway 164 8.0 6 10 Sweden 382 9.1 6 25 Nordic countries, 1990 953 8.1 9 85 Denmark 238 9.1 12 30 Finland 153 6.0 5 5 Iceland 8 6.3 10 1 Norway 157 7.3 9 15 Sweden 397 9.2 9 35 Nordic countries, 2000 952 7.8 7 65 Denmark 228 8.6 12 25 Finland 151 5.8 5 5 Iceland 11 7.9 10 1 Norway 150 6.7 9 15 Sweden 412 9.1 5 20 I • Population attributable risk percent whether the relationship is causal or due to confounding by factors such as those associ- ated with sexual habits, including persistent in- fections with certain types of human papil- lomavirus (127-128). The relative risk for e, er- 34 vical cancer of current smokers versus non- smokers was set at 2.0 (Table 2.5). As seen in Table 2.28, the evade rate of can- cer of the uterine cervix in the Nordic coun- tries decreases over time. About 25% of the I I I I

TOBACCO SMOKING TABLE 2.31. Summary of proportions and numbers of cancers avoidable in the Nordic countries combined in 2000 ~ smokin~ were eliminated Tobacco-associated Both sexes Men Women cancer sites PAR Avoidable PAR Avoidable PAR Avoidable % number % number % number i I I All confirmed sites 53 15.740 56 10.875 46 4.865 Lung 82 10.305 84 6.815 79 3.490 Larynx 67 555 68 490 62 65 Pharynx 50 285 52 235 41 50 Oesophagus 51 530 55 385 44 145 Oral cavity and tongue 48 460 52 315 41 145 Lower urinary tract 39 2.660 41 2.060 32 600 Pancreas 15 560 17 315 12 245 Kidney 13 385 14 260 10 125 All suspected sites I4 1.290 15 665 14 625 Lip 24 150 26 120 20 30 Uterine cervix 24 285 - - 24 285 Liver 13 145 15 95 11 50 Stomach 13 540 15 345 11 195 Leukaernia 9 170 10 105 7 65 I I 1 I i ! i i i i l PAR%, population attributab.le risk p~rcent cases are possibly attributable to smoking, which will correspond around 2000 to about 300 eases, annually. Leukaemia (ICD-7: 204) Leukaernia is a rare malignancy with rela- tively little international variation in incidence (32). In 1990, about 1,200 ~aew eases were regis- tered in the Nordic countries among men and 950 among women (34-44). Relationship with tobacco smoking Ionizing radiation, certain drugs and a few industrial chemicals are among the well-docu- mented risk factors for leukaemia (32). Recent up-dating of large prospective studies showed weak positive association in males between ciga- rette smoking and leukaemia especially of the myeloid sub-types (2, 129-I32); however, in a follow- up of two American Cancer Society pro- spective studies, no association was found in women (130). Furthermore, case-control studies have shown contlieting results (133--134). In this review, the risk for leukaemia among current smokers in eomparisun with nonsmokers was set at 1.3 (Table 2.5). The crude rates of leukaemia for men and women have been decreasing over time. About 10% of cases in men and 5% in women are pos- sibly related to smoking habits, corresponding to an expected number of about 200 cases in the two sexes combined in the year2000. All sites combined Table 2.31 gives an overview of the estimates for cancer morbidity in the Nordic countries in 2000 due specifically to the prevailing smoking habits of the respective national populations. It can be seen that almost 16,000 cancers could be avoided annually if tobacco smoking were given up totally; this corresponds to slightly more than 50% of all newly diagnosed tobacco-re- lated cancers. Further, 1,300 cancers could probably be avoided (Table 2.31, bottom). Figs. 2.9 (men) and 2.10 (women) give the same type of information in graphic form. Some 11,000 cancers in men and 5000 in women could be avoided annually by stopping tobacco smoking, cancers of the lung and urinary blad- der being the most significant. Table 2.32 gives the numbers avoidable independently for each of the Nordic countries and for the two sexes combined. The PAR% indicate the proportion of to- bacco-related cancers actually caused by the to- bacco smoking habits of We population. A supplementary figure of interest is the number of tobacco-induced cancers as a proportion of all cancers diagnosed. Table 2.33 gives both measures for the Nordic countries for 1980, 35

DREYER et al. No. of cases 8,000 Fig. 2.9. Cancers avoidable in the year 2000 if cigarette smok- ing w~re eliminated, men. 6,000 4,000 2,000 Site 1990 and 2000. Overall, 19% of all cancers that will be notified among Nordic men in the year 2000 are expected to be due to tobacco smok- ing, ranging from 14% in Sweden to 26% in Denmark. The equivalent figure for Nordic women will be 9%, ranging from 5% in Finland to 15% in Iceland. For all five countries and the two sexes com- bined, smoking-induced cancers account for 14% of all cancers notified and this figure in- creases to 15% if possible tobacco-related can- cers are included. Denmark and Iceland will have the highest percentage, with an estimate of about 20% for the two sexes combined. The lowest proportion of smoking-induced cancers will be seen in Sweden (11%). 36 DISCUSSION The calculations presented in this chapter on smoking-induced cancers in the Nordic coun- tries were made on the basis of the most detailed information on smoking habits available for the Nordic countries around 1965, 1975 and 1985. The aim was to give a detailed, accurate esti- mate of the burden of tobacco-related cancers in the five countries, separately and combined. Information on the proportions of smokers, levels of smoking and types of tobacco used was combined with the appropriate relative risk esti- mates for each of a number of tobacco-related cancers. It would have been a further advantage for the study if data on the duration of regular

I I I i ! Fig. 2.10. Cancers avoidable in the year 2000 if cigarette smok- ing wcr¢ eliminated, wornc'n. TOBACCO SMOKING No. of cases 8,000 Tobacco-related Other causes 6,000 ........................................... 4,000 1 1 2,000 Site I I I I l l, smoking had been available in a standardized format, as we]] as data on inhalation habits, which wa.s not the case. By selective use of risk estimates obtained from studies in the United States, Europe and particularly in the Nordic countries partial indirect adjustment v~re made for these variations in smoking habits. Over the past f~w dew~ades, cigarett~ design and manufacture have changed considerably, af- f~ting the smoke composition and yields of tar (32). A number of studies suggest that people who smoke lilter-tipp~d or low-tar cigarettes have a lower risk of lung eancer than people who smoked the types of cigarettes that wcm available until the late 1950s (135-140). Because of the latency of smoking-induced cancers, it is unlikely that the full effects of these changes in tobacco products are reflected in our figures for tobacco-related cancers in the year 2000. The tar content of cigarettes in the Nordic countries is still relatively high compared with other places, although it has clearly decreased since the 1970s. If the tar content of cigarettes in ge- neral was lower in some of the studies that served as sources for risk estimates, this might again have resulted in a possible underestima- tion of the proportions of cancers in the Nordic countries caused by tobacco smoking. It cannot tm excluded,, however, that the estimates for 2000 are a little too high because of the g~neral decrease in the tar content of cigarettes. We have good mason to believe that our ¢sti- 37

DREYER ¢t al. TABLE 2.32. Summary of proportions and numbers of cancers avoidable in men and women in each of the Nordic countries in 2000 if tobacco smoking were eliminated Tobacco-associated Denmark Finland Iceland Norway Sweden cancer sites PAR Avoidable PAR Avoidable PAR Avoidable PAR Avoidable PAR Avoidable % number % number % number % number % number All confirmed sites 61 5.325 52 2.905 60 195 52 2.610 46 4.700 Lung 84 3.625 83 2.000 85 145 82 1.600 79 2.935 Larynx 70 225 70 80 72 4 69 97 62 150 Oesophagus 56 155 48 100 57 5 54 75 49 190 Pharynx 52 90 45 35 53 1 50 55 44 105 Mouth and 52 155 45 55 53 4 50 105 44 145 tongue Pancreas 19 150 15 130 19 5 15 95 12 175 Lower urinary 42 820 40 405 44 25 39 520 34 890 tract Kidney 17 105 14 100 15 5 13 65 9 110 All suspected sites 23 410 13 230 18 20 17 250 12 385 Lip 33 35 23 40 29 3 27 35 21 40 Liver 20 40 11 35 17 1 15 15 11 55 Stomach 20 140 11 120 17 10 15 110 11 160 Uterine cervix 31 125 15 15 27 4 24 60 19 80 Leukaemia 13 70 7 20 10 2 10 30 6 50 tl l t I PAR%, population attributable risk percent TABLE 2.33. Proportions of tobacco-related cancers and all cancer types caused by tobacco smoking in the Nordic populations diagnosed in 1980, 1990 and 2000 (estimated) Aetiological Denmark Finland Iceland Norway Sweden All countries proportions M/F All M/F All M/F All M/F All M/F All M/F All [ [ PAR% 1980 63/43 57 74/33 63 58/55 57 57/28 48 50/28 43 60/34 52 1990 .60/52 57 63/32 54 55/54 55 57/40 51 51/41 48 57/43 52 2000 62/59 61 62/33 52 61/60 60 56/46 .52 49/41 46 56/46 53 of all cancers 1980 27/8 17 33/5 19 19/12 16 18/4 11 15/4 9 21/5 13 1990 25/10 17 25/5 14 20/10 15 18/6 12 14/6 10 19/7 13 2000 26/14 20 23/5 14 20/15 18 18/8 13 14/7 11 19/9 14 PAR%, population attributable risk percent 1 1 l mates of the numbers of tobacco-related can-. eers in the Nordic populations are valid. After subtracting the estimated number of tobacco- related lung cancers from the total incidence of lung cancer in the Nordic countries, we foun~ the resulting incidence rate (per 100,000 inhabi- tants) to be very dose to the rates of lung cancer previously reported in large cohort studies in which the rates of lung cancer among "never smokers" were estimated. The burden of cancer due to tobacco smoking is usually estimated as the proportion of cancer deaths in a population that can be attributed to 38 smoking, simply because mortality statistics are available from all developed countries; this is not the case for cancer incidence data. The pro- portion of cancer deaths in the United States has been estimated to be about 30% (1), which is 0 higher than the 15'A incidence data found in our study• A recent study has estimated that the pro- portions of mortality from cancer in the Nordic countries in 1995 were about 30% in Denmark, 21% in Finland, 18% in Norway, and 15% in Sweden; no data were available for Iceland (141). The difference due partly to the fact that inci- dence data include people who have survived

i I I 1 i ! 1 i i I I TOBACCO SMOKING TABLE 2.34. Numbers of men and women who have never smoked in each of the five Nordic countries, who were exposed to tobacco "smoke from their spouse, 1965, 1975 and 1985 Country Year Numbers married to current smokers Men Women Nordic countries 1965 301,000 1,64.7,000 1975 493,000 1,304,000 1985 674,000 1,147,000 Denmark 1965 62,000 400,000 1975 127,000 314,000 1985 224,000 331,000 Finland 1965 75,000 368,000 1975 66,000 263,000 1985 92,000 292,000 Iceland 1965 3,000 11,000 1975 4,000 10,000 1985 6,000 9,000 Norway 1965 45,000 312,000 1975 o 86,000 317,000 1985 113,000 200,000 Sweden 1965, 116,000 556,000 1975 210,000 400,000 1985 239,000 315,000 Smoking habits in 1965, 1975 and 1985 or a close year their cancer and that tobacco-related cancers in general and lung cancer in jgartieular belong to the group of cancers with poor survival rates. To our knowledge, cancer incidence rates have been used for calculating the numbers and pro- portions of smoking-induced cancer in whole populations in only a few previous studies. In an Australian study conducted in the 1980s (142), it was estimated that about one-third of all incident cancers in Australia would be preventable on the basis of current knowledge, and that about half of these, i.e. 18%, could be prevented by eliminating tobacco smoking. The latter estimate seems reasonable and is within the range of the estimates given in this report for the Nordic countries. In conclusion, tobacco smoking is a major cause of cancer in the Nordic countries; it will be responsible for about 15% of all cancers (10% in women and 20% in men) by the year 2000. These are also the proportions of cancer that could be avoided if present smoking habits were given up totally. Even if that occurred today, however, it would take a long time, possibly more than 40 years, before the full effect on cancer rates was achieved, mainly because the risk for tobacco-re- lated cancers among former smokers will be mod- erately increased for decades. PASSIVE SMOKING Tobacco smoking is a source of widespread con- tamination of the indoor air of homes, public places and workplaces with agents that clearly cause cancer at higher doses. The first indi- cation that environmental tobacco smoke was potentially carcinogenic appeared in the early 1980s, with the report of an increased risk for lung cancer among nonsmoking women mar- ried to smokers (143-144). Since then, more than 25 epidemiological studies of this relation- ship have been published; in general, a modestly increased risk for lung cancer has been seen among passive smokers, with higher risks in persons married to heavy smokers than in those married to light smokers (145). Studies of passive smoking in e.g. workplaces have given less consistent findings (146-152). Exposures during childhood seem to result in no increased risk for lung cancer later in life (149, 152-159). TABLE 2.35. Sex-specific incidence rates of lung cancer per 100,000 in neversmokers in each of the five Nordic countries in 1980, 1990 and 2000 Country Lung cancer incidence in neversmokers Men Women 1980 1990 2000 1980 1990 2000 Nordic countries 10.0 Denmark 14.7 Finland 9.6 Iceland 5.5 Norway 8.0 Sweden 8.6 9.9 11.4 4.7 5.5 8.0 14.4 15.5 6.9 9.2 12.7 9.9 10.5 3.5 4.3 5.9 8.1 9.5 4.9 5.3 9.0 9.4 10.1 3.8 5.5 5.9 7.6 10.2 4.4 4.1 7.4 39

DREYER et al. Sidestream smoke, the smoke formed in be- tween puffs and emitted into the air, contains even higher concentrations of several carcino- gens than mainstream smoke, the smoke taken in by smokers (32). The finding of metabolites of lung caminogens in the urine of passive smokers provides supporting evidence that environmental tobacco smoke can cause lung cancer (160). MATERIAL AND METHODS The numbers of lung cancers among people who have never smoked which are attributable to tobacco smoking by their spouses were calculated on the basis of information on sex-specific smoking habits in each of the Nordic countries and the relative risk for lung cancer associated with passive smoking. It was as- sumed that the effect of passive smoking cannot be assessed in active smokers owing to the profound ef- fect of active smoking on the risk for lung cancer. Neither was the risk of former smokers estimated. The proportion of lung cancers/n the general popu- lation that is attributable to passive smoking (PAR%) was calculated by dividing the number of lung can- cers induced by passive smelting by the total number of lung cancers, and multiplying by 100. Smoking habits Data on the smoking habits of men and women in the five Nordic countries were reviewed for the years 1965, 1975 and 1985 (Tables 2.2 and 2.3). Table 2.34 gives the estimated number of "neversmokers" mar- ried to current smokers in each country for each of the three calendar years. "Current smokers" were considered to be smokers of any kind of tobacco (mainly cigarettes); the category "non-smoking spouses" includes the subgroups of "never-smokers" and ex-smokers. The number of women who had never smoked who were exposed passively to tobacco from their spouses decreased throughout the period, while the number of men who had never smoked and were exposed in- creased. Thus, more than I million "n.eversmoking" women and aimost 700,000 "neversmoking" men in the Nordic countries were married to a current smoker in 1985. Relative risk for lung cancer The relative risk of 1.47 for lung cancer used in this review was derived from a European meta-analysis of six studies of the effects of passive smoking (145-147, 153, 156, 161, 162). Most of the studies involved only non-smoking women; however, the carcinogenic ef- fect of passive smoking is assumed to be similar in the two sexes. We estimated the incidence rates of lung cancer among "neversmokers" for 1980, 1990 and 2000 (Table 2.35) on the basis of the estimated numbers of tobacco-induced lung cancers in the Nordic countries in those years, derived from Tables 2.6 for men and 2.7 for women. These rates were combined with in- formation on the numbers of "never-smears" ex- posed to environmental tobacco smoke at home and Table 2.36. Proportions and numbers of cases of lung cancer avoidable among men in the Nordic 1980, 1990 and 2000 if passive smoking were eliminated countries in Country Annual no. Incidence PAR%" Avoidable of n~v cases per 100,000 number Nordic countries, 1980 6906 62.2 0.2 15 Denmark 2130 84.2 0.2 4 Finland 1983 85.8 0.2 3 Iceland 42 36.5 0.2 0 Norway 962 47.5 0.2 2 Sweden 1789 43.4 0.2 4 Nordic countries, 1990 6679 58.6 0.3 15 Denmark 2018 79.7 0.3 5 Finland 1629 67.5 0.2 3 Iceland 61 47.7 0.2 0 Norway 1159 55.3 0.3 3 Sweden 1812 42.9 0.4 5 Nordic countries, 2000 8102 68.6 0.4 35 Denmark 2520 97.4 0.5 15 Finland 1879 76.0 0.2 4 Iceland 91 64.7 0.2 0 Norway 1313 59.9 0.4 5 Sweden 2299 52.1 0.4 I0 • Population attributable risk percent [| i [ [ l 1 ! [ I I I I

] ] ] ] ] | TO]~.CCO SMOI~NG TABLE 2.37. Proportions and numbers of cases of lung cancer avoidable among women in the Nordic countries in 1980, 1990 and 2000 if t~assive smoking were eliminated Country Annual no. Incidence PAR%a Avoidable of new eases per 100,000 number Nordic countries, 1980 1857 16.2 1.5 30 Denmark 708 27.3 1.4 10 Finland 299 12.1 1.7 5 Iceland 32 28.3 0.6 0 Norway 235 11.4 1.9 4 Sweden 583 13.9 1.6 10 Nordic countries, 1990 2826 24.0 1.0 25 Denmark 1101 42.2 1.0 10 Finland 380 14.9 1.2 5 Iceland 38 29.9 0.5 0 Norway 458 ' 21.4 1.2 5 Sweden 849 19.6 0.8 5 Nordic countries, 2000 4437 36.4 0.9 35 Denmark 1800 67.6 0.9 15 Finland 532 20.3 1.3 5 • Iceland 75 53.8 0.4 0 Norway 624 27.9 0.8 5 Sweden o 1406 31.1 0.7 10 Population attributable risk percent the associated relative risk for lung cancer to caleu= late the numbers of cancers attributable to passive smoking. RESULTS Tables 3.36 (men) and 3.37° (women) show the total annual numbers of eases of lung cancer in the Nordic countries, the crude incidence rates and the estimated numbers and proportions (PAR%) of lung cancers caused by passive smoking. While the PAR% for lung cancer is increasing in men, from 0.2% in 1980 to 0.4% in 2000, that in women is decreasing from 1.5 to 0.9% during the same period. In the two sexes combined, the PAR% in 2000 will be about 0.6%, which is equivalent to approximately 70 new lung cancer eases per year in the Nordic countries. DISCUSSION The estimates given in this chapter for the num- bers of lung cancers caused by passive smoking are based on the frequency of exposure of neversmokers to the tobaec.~ smoke of their spouses. Other sources of passive exposure to smoking and the potential effects of passive smoking on active smokers and cx=smokers were not considered; thus estimates may be underestimated. In contrast, some of the in- crease in the relative risk for lung cancer seen in studies of passive smokers may be due to under- reporting of smoking habits in this subgroup of the population. Although 10-15% of lung cancers arising in "neversmokers" may be due to exposures to en- vironmental tobacco smoke, the lung cancer in- eidence in such persons is so low (6-10 per 100,000 inhabitants) that the numbers induced by passive smoking comprise only about 0.6% of all incident lung. cancers diagnosed in the Nordic populations. Consequently, passive smoking is a minor cause of lung cancer in the Nordic countries. REFERENCES 1. Doll, R. & Pete, R.: The causes of cancer: Quantitative estimates of avoidable risks of cancer in the United States today. J. Natl. Can- cerInst. 66: 1193-308, 1981. 2. Doll, R., Pete, R., ItZheatley, IL, Gray, R. & Sutherland, I.: Mortality in relation to smoking: 40 years' observations on male British doctors. Br. Med J. 309: 901-11, 1994. 3. Levin, M. L., Goldstein, H. & Gerhardt, P. 41

DREYER et al. Cancer and tobacco smoking. A preliminary re- port. J. Am. Med. Assoc. 143: 336-8, 1950. 4. Wynder, E. L. & Graham, E. A.: Tobacco smok- ing as a possible etiologic factor in bronchio- genie carcinoma. A study of six hundred and eighty-four proved cases. J. Am. M~l. Assoc. 143: 329-46, 1950. 5. Doll, R. & Hill, A. B.: Smoking and carcinoma of the lung. Preliminary report. Br. Meal. J. ii: 739-48, 1950. 6. Best, E. W. R., Josie, G. H. & Walker, C. B.: A Canadian study of mortality in relation to smoking habits. A preliminary report. Can. J. Public Health .,¢2: 99-I06, 1961. 7. Kahn, H. A.: The Dorn study of smoking and mortality among US veterans: Report on eight and one-half years of observation. Natl. Cancer Inst. Monogr. 19: 1-;125, 1966. $. Hammond, E. C. & Horn, D.: Smoking and death rates report on forty-four months of fol- low-up of 187,783 men. II. Death rates by causes. J. Am. M~cl. Assoc. 166: 1294-308, 1958. 9. Doll, R. & Hill, A. B.: Mortality in relation to smoking: Ten years' observations of British doctors. Br. Med. J. 1: 1460-7, 1964. 10. Doll, R. & Hill, A. B.: Mortality in relation to smoking: Ten years" observations of british doc- tors. Br. Med. J. 1: 1399-410, 1964. 11. International ,4geney for Research on Cancer: Tobacco smoking. Lyon: International Agency for R~-search on Cancer, 1986 (IARC Mono- graphs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans. Vol. 38). 12. Hoffmann, D. & Hart/s, C. C.: M~,hanisms in tobacco carcinogenesis. Cold spring Harbor, NY: Cold Spring Harbor Laboratory, 1966 (Banbury Report 23). 13. Williams, A. & Samet, J.: Lung cancer and cigarette smoking. In: Samet, ~ (ed.) Epidemi- ology of Lung Cancer. New York: Marcel Dekker, Inc~, 1994, pp 71-108. 14. US Department of Health and Human Services The Health Consequences of Smoking: Cancer. A Report of the Surgeon General. Washington DC, US Government Printing Office, 1982. DHHS Publication No.(PHS)82-50179. 15. Gao, Y. T., Blot, W. Z, Zheng, W., Fraumeni, J. F. & Hsu C. IV.: Lung cancer and smoking in Shanghai. Int. J. Epidemiol. 17: 277-80, 1988. 16. Floderus, B., Cederl~f, R. & Friberg, L.: Smok- ing and mortality: a 21-y~ar follow-up based on the Swedish twin registry. Int. J. Epidemiol. 17: 332-40, 1988. 17. Garfinkel, I~ & Stellman, S. D.: Smoking and lung cancer in women: Findings in a prospec- tive study. Cancer Res. 48: 6951-5, 1988. 18. Schoenberg, J. B., Wilcox, H. B., Mason, T.Z, Bill, J. & Stemhagen, A.: Variation in smoking- related lung cancer risk among New Jersey women. Am. J. Epidemiol. 130: 688-95, 1989. 19. Osann, K. E., ,4nton-Culver, H., Kurosaki, T. & Taylor, T.: Sex differences in lung-cancer risk associated with cigarette smoking. Int. J. Can- cer 54: 44-8, 1993. 20. Harris, R. E., Zang, E. ,4., ,4nderson, J. I. & Wynder, E. L.: Race and sex differences in lung cancer risk associated with cigarette smoking. Int. J. Epidemiol. 22: 592-9, 1993. 21. Saracci, R.: The interactions of tobacco smok. ing and other agents in cancer etiology. Epide- miol. Rev. 9: 175-93, 1987. 22. Nielsen, P. E., Krarup N. B., ,4strup, J., Gyntel- berg, F. & Lisse, 0.: Alterations in Danish smoking habits 1958-1976. Ugeskr. Laeg. 140: 2528-32, 1978 (Original in Danish). 23. Roed-Petersen, B.: Smoking habits in East Denmark at October 1979. Scand. J. Soc. Med. 12: 39-47, 1984. 24. Hagerup, L. & Larsen M.: Tobacco smoking and respiratory symptoms in a Danish popula- tion. Ugeskr. Laeg. 133: 1302-6, 1971 (Original in Danish). 25. Gallup Market Research. Denmark, 1975 and 1985. 26. Martelin, T.: The development of smoking habits according to survey data in Finland. Publication of the National Board of Health, Finland, Health Education. Series original re- ports 1/1984. Helsinki 1984 (Original in Fin- nish). 27. Study on health habits 1985. Statistics Finland (unpublish~). 28. Icelandic Heart Association and Committee for Tobacco Use Prevention. 29. Henriksen, T. Z.: R.vkevaner i den norske folkning. Den Norske L~g~forening, 11: 617- 20, 1976. 30. Statens Tobakksskaderdd. Tobakksforbruk i Norge g~ennom 20 ~r. Oslo, 1994. 31. National Smoking and Health Association, Sweden, and The Survey Research C.~ater of the Central Bureau of Statistics. 32. Tomati~, L., Aitio, ,4., Day, N. E., Heseltine, E., Kaldor, J., Miller, A. B., Parkin, D. M. & Ri- boll, E.: Cancer: Causes, Occurrence and Con- trol. Lyon: International Agency for Research on Cancer, 1990. (IARC Scientific Publications No. 100). 33. Moiler Jensen, 0., Est~ve, J., Moiler, H. & Re- nard, H.: Cancer in the European Commnunity and its member states. Eur. J. Cancer 26:1167- 256, 1990. 34. Engeland, A., Haldorsen, T., Tretli, S., Hakulin- en, T., H~rte, L., Luostarinen, T., Magnus, K., Schou, G., Sigvaldason, H., Storm, H. H., Tulin- ius, H. & Vaittinen, P.: Prediction ofcancer inci- dence in the Nordic countries up to the years 15 [ [ 1. [ I 'I

TOBACCO SMOKING 2000 and 2010. APMIS 101: Suppl. 38, 1-124, 1993. 35. Danish Cancer Registry, Institute of Cancer Epi- demiology. Cancer Incidence in Denmark 1978, 1979 and 1980. Copenhagen: Danish Cancer Society, I983. 36. Danish Cancer Society, Division for Cancer Epi- demiology, Department of Cancer Registration. Cancer Incidence in Denmark 1990. Copen- hagen: Danish Cancer Society, 1994. 37. Finnish Cancer Registry, The Institute for Stat- istical and Epidemiological Cancer Research: Cancer Incidence in Finland 1980. Helsinki: Cancer Society of Finland, 1983. 38. Finnish Cancer Registry, The Institute for Statistical and Epidemiological Cancer Re- search: Cancer Incidence in Finland 1989 and 1990. Helsinki, Cancer Society of Finland, 1992. 39. Tulinius, H. & Ragnarsson, J.: Cancer Incidence in Iceland 1955-1984. R~kjavik: Icelandic Cancer Society and Director General of Health, 1987. - 40. Icelandic Cancer Registry: Cancer Incidence in Iceland 1990. (p~rsonal,commanication). 41. The Cancer Registry of Norway. Incidence of Cancer in Norway 1980. Oslo: The Cancer R~gistry of Norway, 1982. 42. The Cancer Registry of Norway, Institute for Epidemiological Cancer Research: Incidence of Cancer in Norway 1990. Oslo: The Cancer Registry of Norway, 1992. 43. National Board of Health and Welfare, The Cancer Registry: Ca~lcer Incidence m Sweden 1980. Solna, Socialstyrelsen, 1983. 44. Centre For Epidemiology, The Swedish Cancer Registry: Cancer Incidence in Sweden 1990. Stockholm; Socialstyrelsen, The National Board of Health and Welfare,, 1993. 45. Carstensen, J. M., Pershagen, G. & Eklund, G.: Mortality in relation to cigarette and pipe smoking: 16 years' observation of 25,000 Swed- ish men. J. Epidvmiol. Community Health 41: 166-72, 1987. 46. Engholm, G., Englund, ,~., Gerhardsson de Ver- diet, M., Holm, L.-E. & Stjernberg, N.: R6kn- ing och Cancer. En 20-~.rsuppf61jning .av 135.000 svenska byggnadsarb~tarv. Solna: Ar- betslivsinstitutet. Arbcte och H~dsa (in Swedish) (in press). 47, Hammond, E. C.: Smoking in relation to the death rates of one million men and women. Natl. Cancer Inst. Monogr. 19: 127-204, 1966 48. Hakulinen, T., Pakkala, E., Puska, P., Tuomi- lehto, J. & Vartiainen, E~: Various measures of smoking as predictors of cancer of different types in two Finnish cohorts. In: Colditz, G.A. (Ed.): Proceedings of the Consensus Conference on Smoking and Prostate Cancer, Brisbane, February 1996. Repatriation Medical Auth- ority, Brisbane 1997. 49. Lossing, E. H., Best, E. W. P~, McGregor, J. T., Josie, G. H., Walker, C. B., Delaquis, F. H., Baker, P. M. & McKenzie, A. C.: A Canadian Study of Smoking and Health. Ottawa; Depart- ment of National Health and Welfare, 1966. 50. Lund, E. & Zeiner Henriksen, T.: [Smoking as a risk factor for different cancer forms among 26,000 Norwegian men and women. A com- parison of a material of smokers from the Nor- wegian Cancer Rvgistry] R~king sore risikofak- tot for ulike kreftformer blant 26,000 norske menn og kvinner. En kobling av et rokermateri- ale reed Kreftregisteret over en 12 ars periode. Tidsskr. Nor. Laegeforen. ]01: 1937--40, 1981 51. McLaughlin, J. IC, Hrubec, Z., Blot, W, J. & Fraumeni, J. F., Jr: Smoking and cancer mor- tality among US veterans: A 26-year follow-uIx Int. J. Cancer 60: 190-3, 1995. 52. Weir, J. M. & Dunn, J. E. Jr.: Smoking and mortality: A prospective study. Cancer 25: I05- 12, 1970. 53. Burch, J. D., Howe, G. R., Miller, A. B. & Se- menciw, R.: Tobacco, alcohol, asbestos, and nickel in the etiology of cancer of the larynx: A case-control study. J. Natl. Cancer Inst. 67: 1219-24, 1981. 54. Bardn, :1. E., Franceschi, S., Barra, S., Talami- ni, R. &'La Vecchia, C.: A comparison of the joint effects of alcohol and smoking on the risk of cancer across sites in the upper aerodigestive tract. Cancer Epidemiol. Biomarkers Prey. 2: 519-23, 1993. 55. Tuyns, ~4. J., Estdve, J., Raymond, I.,, Berrino, F., Benhamou, E., Blanchet, F., Boffetta, P., Crosig- nani, P., del Moral, A., Lehmann. W., Merletti, F., P~quignot, G., Riboli, E., Sancho-Gamier, H., Terracini, B., Zubiri, ,4. & Zubiri, L.: Cancer of the larynx/hypopharynx, tobacco and alcohol: IARC international case-control study in Turin and Varese (Italy), Zaragoza and Navarra (Spain), Gvneva (Swit~rland) and Calvados (France). Int. J. Cancer 41: 483-91, 1988. 56. Brug~re, J., Gunnel, P., Leclerc, :1. & Rodriguez, J.: Differential effects of tobacco and alcohol in cancer of the larynx, pharynx, and mouth. Cancer 57: 391-5, 1986 57. Elwood, J. M., Pearson, J. C. G., Skippen, D. H. & Jackson, S. M.: Alcohol, smoking, social and occupational factors in the aetiology of cancer of the oral cavity, pharynx and larynx. Int. J. Cancer 34: 603-12, 1984. 58. Falk, R. T., Pickle, L. W., Brown, L. M., Ma- son, T. J., Buffler, P. A. & Fraumeni, J. F., Jr: Effect of smoking and alcohol consumption on laryngeal cancer risk in coastal Texas. Cancer Res. 49: 4024-9, 1989. 59. Franceschi, S., Talamini, R., Barra, S., Bar~n, 43

DREYER at a~ A. E., Negri, E., Bidoli, E., Serraino, D. & La lZecchia, C.: Smoking and drinking in relation to cancers of the oral cavity, pharynx, larynx, and oesophagus in northern Italy. Cancer Res. 50: 6502-7, 1990. 60. Gugnel, P., Chastang, J. F., Luce, D., Leclerc, ,4. & Brug~re, J.: A study of the interaction of alcohol drinking and tobacco smoking among French cases of laryngeal cancer. J. EpidemioL Community Health 42: 350-4, 1988. 61. Hedberg, K., Vaughan, T. L., White, E., Davis, S. & Thomas, D. B.: Alcoholism and cancer of the larynx: A case-control study in western Washington (United States). Cancer Causes Control 5: 3-8, 1994 62. Muscat, J. E. & Wynder, E. T~: Tobacco, alco- hol, asbestos, and occupational risk factors for laryngeal cancer. Cancer 69: 2244-51, 1992. 63. Olsen, J., Sabro, S. & Fasting, U.: Interaction of alcohol and tobacco as risk factors in cancer of the laryngeal region. J. EpidemioL Com- munity Health 39: 165-8, 1985. 64. Wynder, E. L., Covey, L. S., Mabuchi, K. & Mushinski, M.: Environmental factors in cancer of the larynx: A second look. Cancer 38: 1591- 601, 1976. 65. Wynder, E. L. & Stellman, S. D.: Comparative epidemiology of tobacco-related cancers. Can- cer Res. 37: 4608-22, 1977. 66. Brown, L. M., Blot, W. J., Schuman, S. H., Smith, V. M., Ershow, .4. G., Marl¢~, 1~ D. & Fraumeni, J. F., Jr: Environmental factors and high risk of esophageal cancer among men in coastal South Carolina. 3. Natl. Cancer Inst. 80: 1620-5, 1988. 67. Brown, L M., Hoover, R. N., Greenberg, R. S., Schoenberg, J. B., Schwartz, .4. G., Swanson, G. M., Liff, J. M., Silverman, D. T., Hayes, R~ B. & Pottern, L. M.: Are racial differences in squamous cell esophageal cancer explained by alcohol and tobacco use? J. Natl. Cancer Inst. 86: 1340-5, 1994. 68. La Vecchia, C. & Negri, E.: The role of alcohol in oesophageal cancer in non-smokers, and of tobacco in non-drinkers. Int. J. Cancer 43: 784- 5, 1989. 69. Negri, E., La Vecchia, C., Franceschi, S., Decar- li, .4. & Bruzzi, P.: Attributable risks for oeso- phageal cancer in northern Italy. Eur. J. Cancer 28.4: 1167-71, 1992. 70. Tavani, .4., Negri, E., Franceschi, S. & La Vec- chia, C.: Risk factors for esophageal cancer in women in northern Italy. Cancer 72: 2531-6, 1993. 71. Tuyns, .4. J., Pdquignot, G. & Jensen, O. M.: [Esophageal cancer in Ille-et-Vilaine in relation to levels of alcohol and tobacco consumption. Risks thatare multiply] La cancer de l'oeso- phage en Ille-et-Vilaine en fonction des niwaux de consommation d'alcool et de tabac. Des ris- ques qui se multiplient. Bull. Cancer 64: 45~0, 1977. 72. Tuyns, ,4. J.: Oesophageal cancer in non-smok- ing drinkers and in Non-Drinking smokers. Int. 3. Cancer 32: 443-4, 1983. 73. Tuyns, .4. J. & Esteve, J.: Pipe, commercial and hand-rolled cigarette smoking in oesophageal cancer. Int. 3. EpidemioL 12:110-3, 1983. 74. La Vecchia, C., Liati, P., Decarli, .4., Negrello, L & Franceschi, S.: Tar yields of cigarettes and the risk of oesophageal cancer. Int. 3. Cancer 38: 381-5, 1986. 75. Choi, S. Y. & Kahyo, H.: Effect of cigarette smoking and alcohol consumption in the aeti- ology of cancer of the oral cavity, pharynx and larynx. Int. 3. Epidemiol. 20: 878-85, 1991. 76. Olsen, J., Sabroe, S. & Ipsen, J. Effect of com- bined alcohol and tobacco exposure on risk of cancer of the hypopharynx. J. Epidemiol. Com- munity Health 39: 304-7, 1985. 77. Bundgaard, T., Wildt, J., Frydenberg, M., Elbr~nd, O. & Nielsen, J. E.: Case-control study of squamous cell cancer of the oral cavity in Denmark. Cancer Causes Control 6: 57-67, 1995. 78. Rothman, K. & Keller, A.: The effect of joint exposure to alcohol and tobacco on risk of can- cer of the mouth and pharynx. J. Chronic. Dis. 25: 711-6, 1972. 79. Boffetta, P., Mashberg, A., Winkelmann, 1~. & Garfinkel, L.: Carcinogenic effect of tobacco smoking and alcohol drinking on anatomic sites of the oral cavity and oropharynx. Int. J. Cancer 52: 530-3, 1992. 80. Franceschi, S., Barra, S., La Vecchia, C., Bidoli, E., Negri, E. & Talamini, P,.: Risk factors for cancer of the tongue and the mouth. A case- control study from northern Italy. Cancer 70: 2227-33, 1992. 81. Kabat, G. C., Hebert, J. R. & Wynder, E. L~: Risk factors for oral cancer in women. Cancer Res. 49: 2803-6, 1989. 82. Negri, E., La Vecchia, C., Franceschi, S. & Tav- ani, A.: Attributable risk for oral cancer in northern Italy. Cancer Epidemiol. 2: 189-93, 1993. 83. Bouchardy, C., Clavel, F., La Vecchia, C., l~ay- mond, ~ & Boyle, P.: Alcohol, be~" and cancer of the pancreas. Int. J. Cancer 45: 842-6, 1990. 84. Cuzick, J. & Babiker, A. G.: Pancreatic cancer, alcohol, diabetes mellitus and gall-bladder dis- ease. Int. J. Cancer 43: 415-21, 1989. 85. Falk, R. T., Pickle, L. W., Fontham, E. T., Cot- tea, P. & Fraumeni, J. F. Jr.: Life-style risk fac- tors for pancreatic cancer in Louisiana: A case- control study. Am. J. Epidemiol. 128: 324-36, 1988. 86. Farrow, D. C. & Davis, S.: Risk of pancreatic I !. ! ! I I [

I ! TOBACCO SMOKING cancer in relation to medical history and the use of tobacco, alcohol and coffee. Int. J. Cancer 45: 816-20, 1990. 87. Heuch, L, KvMe, G., Jacobsen, B. K. & Bjelke, E.: Use of alcohol, tobacco and coffee, and risk of pancreatic cancer. Br. J. Cancer 48: 637-43, 1983 88. MacMahon, B., Yen, S., Trichopoulos, D., War- ren, I~ & Nardi, G.: Coffee and cancer of the pancreas. N. Engl. J. Med. 304: 630-3, 1981. 89. Silverman, D. T., Dann, J. A., Hoover, K N., Schiffman, M., Lillemoe, K. D., Schoenberg, J. B., Brown, L. M., Greenberg, R. S., Hayes, R. B., Swanson, G. M., Wacholder, S., Schwartz, A. G., Lift, J. M. & Pottern, L. M.: Cigarette smoking and panere,as cancer: A case-control study based on direct interviews. J, Natl. Can- cer Inst. 86: 1510-6, 1994. 90. Wynder, E. I~, Hail, N. E. & Polansky, M.: Epi- demiology of coffee and pancreatic cancer. Cancer Res. 43: 3900-6, 1983. 91. Zheng, W,, McLaughlin, J. K., Gridley, G., Bjel- ke, E., Schuman, L. M., Silverman, D. T., Wac- holder, S., Co Chien, H. T., Blot, W. J. & Frau- meni, J. F., Jr.: A cohot:t study of smoking, al- cohol consumption, and dietary factors for pancreatic cancer (United States). Cancer Causes Control 4: 477-82, 1993. 92. Howe, G. R., Jain, M., Butch, J. D. & Miller, A. B.: Cigarette smoking and cancer of the pan- creas: Evidence from a population-based case- control study in Toronto, Canada. Int. J. Can- cer 47: 323-8, 1991 93. Augustine, A., Hebert, J." R., Kabat, G. C. & Wynder, E. L: Bladder cancer in relation to cigarette smoking. Cancer R~s. 48: 4405-8, 1988. 94. Burns, P. B. & Swanson, G. M.: Risk of urinary bladder cancer among blacks and whites: The role of cigarette use and occupation. Cancer Causes Control 2: 371-9, 1991. 95. Kunze, E., Chang-Claude, J. & Frentzel-Beyme, R.: Life style and occupational risk factors for bladder cancer in Gvrmany. Cancer 69: 1776- 90, 1992. 96. Lopez-Abente, G., GonzMez, C. A., Errezola, M., Escolar, A., Izarzugaza, I., Nebot, M., & Riboli, E.: Tobacco smoke inhalation pattern, tobacco type, and bladder cancer in Spain. Am. J. Epidemiol. 134: 830-9, 1991. 97. Meller Jensen, 0., K, nudsen, J.B., Serensen, B. L. & Clemmesen, J.: Artificial sweeteners and absence of bladder cancer risk in Copenhagen. Int. J. Cancer 32: 577-82, 1983. 98. Sorahan, T., Lancashire, R. J. & Sole, .G.: Uro- thdial cancer and cigarette smoking: Findings from a regional case-controlled study. Br. J. Urol. 74: 753-6, 1994. 99. Steineck, G., Norell, S. E. & Feychting, M.: Diet, tobacco and urothclial cancer. A 14-year follow-up of 16, 477 subjects. Acta Oncol. 27: 323-7, 1988. I00. Vineis, P., Estdve, J. & Terracini, B.: Bladder cancer and smoking in males: Types of ciga- rettes, age at start, effect of stopping and inter- action with occupation. Int. J. Cancer 34:165- 70, 1984. 101. Wynder, E. L. & Gol~mith, R.: The epidcmi- elegy of bladder cancer. A second look. Cancer 40: 1246-68, 1977. 102. Lockwood, K.: On the etiology of bladder tu- rnouts in K'benhavn-Fredriksberg. An inquiry of 369 patients and 369 controls. Acta Pathol. Microbiol. Scan& 51: 1961. 103. MMler Jensen, 0., Wahrendorf, J., Blettner, M., Knudsen, J. B. & Sorensen, B. L.: The Copen- hagen case-control study of bladder cancer: role of smoking in invasive and non-invasive bladder mmours. J. Epidcrniol. Community Health 41: 30-6, 1987. 104. Mommsen, S., Aagaard, J. & Sell, A.: An epide- miological study of bladder cancer in a predom- inantly rural district. Stand. J. Urol. Ncphrol. 17: 30%12, 1983. 105. Benhamou, S., Lenfant, M. H., Ory-Paoletti, C. & Flamant, P~: Risk factors for rtmal-cell carcinoma in a French case-control study. Int. J. Cancer 55: 32-6, 1993. 106. Brownaon, R. C.: A case-control study of renal cell carcinoma in relation to occupation, smok- ing, and alcohol consumption. Arch. Environ. Health 43: 238-41, 1988. 107. Hiatt, K A., Tolan, K. & Quesenberry, C. P. Jr: R~nal cell carcinoma and thiazide use: A his- torical case-control study (California, USA). Cancer Causes Control 5: 319-25, 1994. 108. La Vecchia, C., Negri, E., D'Avanzo, B. & Franceschi, S.: Smoking and renal cell carci- noma. Cancer Res. 50: 5231-3, 1990. 109. McLaughlin, J. K., Lindblad, P., Mellemgaard, A., McCredie, M., Mandel, J. S., Schlehofer, B., Pommer, W. & Adami, H. 0.: International re- nal-cell cancer study. I. Tobacco use. Int. J. Cancer 60: 194-8, 1995. 110. Muscat, J. E., Hoffmann, D. & Wynder, E. L.: The epidemiology of renal ceil carcinoma. A second look. Cancer 75: 2552-7, 1995. 11 I. Wynder, E. L, Mabuchi, IC & Whitmore, W. F., Jr.: Epidemiology of adenocarcinoma of the kidnsy. J. Natl. Cancer Inst. 53: 1619-34, 1974. 112. Yu, M. C., Mack, T. M., Hanisch, R., Cicioni, C. & Henderson, B. E.: Cigarette smoking, obesity, diuretic use, and coffee consumption as risk factors for renal cell carcinoma. J. Nat. Cancer Inst. 77: 351-6, 1986. 113. Surgeon General: Smoking and Health. Wash- ington DC: Public Health Services, US Govern- ment Printing Office, 1979. 45

DREYER et aL |l 114. Lindqvist, C: Risk factors in lip cancer: A questionnaire survey. Am. J. Epidemiol. 109: 521-30, 1979. 115. Kneller, ~ IE., McLaughlin, J. K., Bjelke, E., Schuman, L. M., Blot, W. J., Wacholder, S., Gridley, G., Co Chien, H. T. & Fraumeni, J. F., Jr.: A cohort study of stomach cancer in a high-risk American population. Cancer 68: 672-8, 1991. 116. You, W. C, Blot, W. J., Chang, Y. S., Ershow, A. G., Yang, Z. T., An, Q., Henderson, B., Xu, G. W., Fraurneni, J. F., Jr. & Wang, T. G.: Diet and high risk of stomach cancer in Shandong, China. Cancer Ros. 48: 3518-23, 1988. 117. Wu-William~, A. H., Yu, M. C. & Mack, T. M.: Life-style, workplace, and stomach cancer by subsite in young men of Los Angeles County. Cancer Res. 50: 2569-76, 1990. 118. Jedrychowski, Ire'., Boeing, H., Wahrendorf, J., Popiela, T., Tobiasz-Adamczyk, B. & Kulig, J.: Vodka ccnsumption, tobacco smoking and risk of gastric cancer in Poland. Int. J. Epidemiol. 22: 606-13, 1993. 119. Agudo, A., Gonz~lez, C. A., Marcos, G., Sanz, M., Saigi, E., Verge, J., Boleda, M. & Ortego, J.: Consumption of alcohol, coifs, and to- bacco, and gastric cancer in Spain. Cancer Causes Control 3: 137-43, 1992. 120. International Agency for Research on Cancer: Schistosomes, liver flukes and Helicobacter pyl- ori. Lyon: International Agency for R~search on Cancer (IARC Monographs on Evaluation of Carcinogenic Risks to Humans, Vol. 61), 1994. 121. Brisson, J., Morin, C., Fortier, M., Roy, M., Bouchard, C., Leclerc, J., Christen, A., Guimont, C., Penault, F. & Meisels, A.: Risk factors for cervical intraepithelial neoplasia: Differences betw~n low-and high-grade lesions. Am. J. El> id~niol. 140: 700-10, 1994. 122. Becker, T. M., Wheeler, C. M., McGough, N. S., Parmenter, C. A., Stidley, C. A., Jamison, S. F. & Jordan, S. W.: Cigarette smoking and other risk factors for cervical dysplasia in southwestern Hispanic and non-Hispanic white women. Cancer Epidemiol. Biomar~rs Prey. 3: 113-9, 1994. 123. Gram, I. T., Austin, H. & Stalsberg, H.: Ciga- rette smoking and the incidence of cervical intraepithelial neoplasia, grade III, and cancer of the cervix uteri. Am. J. Epidemiol. 135: 341- 6, 1992. 124. Jones, C. J., Brinton, L. A., Hamman, R. F., Stolley, P. D., Lehman, H. F., Levine, R. S. & Mallin, 12: Risk factors t'or in situ cervical can- cer: Results from a case-control study. Cancer P~s. 50: 3657-62, 1990. 125. Daling, J. R., Sherman, 12 J., Hislop, T. G., Maden, C., Mandelson, M. T., Beckmann, A. M. & Weiss, IV. S.: Cigarette smoking and the 46 risk of anogenital cancer. Am. J. Epidemiol. 135: 180-9, 1992. 126. Winkelstein, W.: Smoking and cervical cancer- Currant status: A review. Am. J. Epidemiol. 131: 945-57, 1990. 127. Phillips, A. N. & Smith, G. D.: Cigarette smok- ing as a potential cause of cervical cancer: Has confounding been controlled? Int. J. Epidemiol. 23: 42-9, 1994. 128. International Agency for Research on Cancer: Human papillomaviruses. Lyon. International Agency for Research on Cancer, (IARC Mono- graphs on the Evaluation of Carcinogenic Risks to Humans, Vol. 64), 1995. 129. Linet, M. S., McLaughlin, Z 12, Hsing, A. W., Wacholder, S., Co-Chien, H. T., Schuman, M., Bjelke, E. & Blot, Vv'. J.: Cigamtt~ smoking and leukemia: Results from the Lutheran Brotherhood cohort study. Cancer Causes Con- trol 2: 413-7, 1991. 130. Garfinkel, L, & Boffetta, P.: Associations twcen smokingnd leukemia in two American Cancer Society prospective studies. Cancer 65: 2356-60, 1990. 131. Mills, P. 12, Newell, G. R., Beeson, W. L., Fras- er, G. E. & Phillips, R. L.: History of ciga~tte smoking and risk of leukemia and my~loma: Results from the Adventist health study. J. Natl. Cancer Inst. 82: 1832-6, 1990. 132. Doll, R.: Cancers weakly related to smoking. Br. Med. Bull 52: 35-49, 1996. 133. Brownson, ~ C., Chang, J. C. & Davis, J. R.: Cigarette smoking and risk of adult leukemia. Am. J. Epidemiol. 134: 938-41, 1991. 134. Spitz, M. R., Fueger, J. J., Newell, G. R. & Kea- ring, M.: Leukemia and cigarette smoking. Cancer Causes Control 1: 195-6, 1990. 135. Lubin, J. H., Blot, W. Z, Berrino, F., Flamant, R., Gillis, C. R., Kunze, M., Schrn~lhl, D. & Vis- co, G.: Patterns of lung cancer risk according to type of cigarette smoked. Int. J. Cancer 33, 569-76, 1984. 136. Hammond, E. C., Garfinkel, L., Seidman, H. & Lew, E. A.: "Tar" and nicotine conteat of ciga- rette smoke in relation to death rates. Environ. Res. 12: 263-74, 1976. 137. Benhamou, S., Benhamou, E., Auquier, A. & Flamant, R.: Differential effects of tar content, typ~ of tobacco and use of a filter on lung can- cer risk in male cigarette smokers. Int. J. Epide- miol. 23: 437-43, 1994. 138. Tang, J. L., Morris, J. 12, Wald, N. J., Hole, D., Shit~ley, M. & Tunstall-Pedoe, H.: Mortality in r~lation to tar yield of cigarettes: A prospec- tive study of four cohorts. Br. M~l. J. 311: 1530-3, 1995. 139. Hartge, P., Silverman, D., Hoover, R., Schairer, C., Altman, R., Austin, D., Cantor, K., Child, M., Key, C. & Marrett, L~: Changing cigarette I I I I I I I I I I I o o~ o~ o~ o~ o~ o

TOBACCO SMOKING habits and bladder cancer risk: A case-control study. J. Natl. Cancer Inst. 78, 1119-25, 1987. 140. Wynder, E. L~ & Stetlman, S. D.: Impact of long-term filter cigarette usage on lung and lar- ynx cancer risk: A case-control study. J. Natl. Cancer Inst. 62: 471-7, 1979. 141. Peto, B., Lopez, A.D., Boreham, J., Thun, M. & Heath, C. Jr.: Mortality from Smoking in D~- veloped Countries 1950-2000. Oxford: Oxford University Press 1994. 142. Armstrong, B. I~: The epidemiology and pre- vention of cancer in Australia. Aust. N. Z. J. Surg. 58: 179-87, 1988. 143. Hirayama, T.: Non-smoking wives of heavy smokers have a higher risk of lung cancer: A study from Japan. Br. Meal. J..282: 183-5, 1981. 144. Trichopoulos, D., Kalandidi, A., Sparros, L. & MacMahan, B.: Lung cancer and passive smok- ing. Int. J. Cancer. 27: 1-4, 1981. 145. Pershagen, G.: Passive smoking and lung can- cer. In: Samet, J. M. (~d.) Epidemiology of Lung Cancer. New York: Marcel Dekker Inc., pp. 207-18, 1994. 146. Lee, P. N., Chamberlain,'J. & Alderson, M. IL: Relationship of passive smoking to risk of lung cancer and other smoking- associated diseases. Br. J. Cancer..W: 97-105, 1986. 147. Kalandidi, A., Katsouyanni, I~, Voropoulou, N., Bastas, G., Saracci, R. & Trichopoulos, D.: Passive smoking and diet in the etiology of luncer cancer among non-smokers. Cancer Causes Control 1: 15--21, 1990. 148. Kabat, G. C. & Wynder, E. L.: Lung cancer in nonsmokers. Cancer 53: ~214-21, 1984. 149. Wu, `4. H., Henderson, B. E., Pike, M. C. & Yu, M. C.: Smoking and other risk factors for lung cancer in women. J. Natl. Cancer Inst. 74: 747- 51, 1985. 150. Garfinkel, L~, Auerbach, O. & Joubert, L.: Invol- untary smoking and lung cancer: A case-con- trol study. J. Natl. Cancer Inst. 75: 463-9, 1985. 151. Shimizu, H., Morishita, M., Mizuno, IL, Masu- da, T., Ogura, Y., Santo, M., Nishimura, M., Kunishima, K., Karasawa, ~L, Nishiwaki, K., Yanamoto, M., Hisamichi, S. & Tominaga, S.: Case-control study of lung cancer in nonsmok- ing women. Tohoku J. Exp. Med. 154: 389-97, 1988. 152. Wu-Williams, A. H., Dai, X. D., Blot, Ire'., Xu, Z. Y., Sun, X. W., Xiao, H. P., Stone, B. J., Yu, S. F.., Feng, Y. P., Ershow, .4. G., Sun, J., Fraumeni, J. F. Jr. & Henderson, B. E.: Lung cancer among .women in north-east China. Br. ft. Cancer. 62: 982-7, 1990. 153. Pershagen, G., Hrubec,'Z. & Svensson, C.: Pass- ive smoking and lung cancer in Swedish women. Am. J. Epiderniol. 125: 17-24, 1987. 154. Correa, P., Pickle, L. IV., Fontham, E., Lin, Y. & Haenzel, W.: Passive smoking and lung cancer. Lancet 10: 595-7, 1983. 155. Koo, L. C., Ho, J. H. C., Saw, D. & Ho, C.: Measurements of passive smoking and esti- mates of lung cancer risk among nonsmoking Chinese females. Int. J. Cancer 39: 162-9, 1987. 156. Svensson, C., Pershagen, G. & Klominek., 3".: Smoking and passive smoking in relation to lung cancer in women. Aeta Oneol. 28: 623--9, 1989. 157. Janerich, D. T., Thompson, W. D., Varela, L. R., Greenwald, P., Chorost, S., Tucci, C., Za- man, M. B., Melamed, M. R., Kiely, M. & McKneally, M. F.: Lung cancer and exposure to tobacco smoke in the household. N. Engl. J. Med. 323: 632--6, 1990. 158. Sobue, 7"., Suzuki, JL, Nakayama, N., Inubuse, C., Matsuda, M., Doi, 0., Mori, T., l~uruse, K., Fukuoka, H., Yasumitsu, T., Kuwabara, 0., Ichigaya, M., Kurata, M., Kuwabara, M., Naka- hara, N., Endo, S. & Hattori, S.: Passive smok- ing among nonsmoking women and the re- lationship between indoor air pollution and lung cancer incidence - R~sults of a multicenter case-control study. Gan. to O. Rinsho. 36: 329- 33, 1990. 159. Stockwell, H. G., Goldman, A. L., Lyman, G, H., Noss, C. L, Armstrong, A. W., Pinkham, P. A., Candelora, E. C. & Brusa, M. 1~: Environ- mental tobacco smoke and lung cancer risk in nonsmoking women. J. Natl. Cancer Inst. 84: 1417-22, 1992. 160. Hbcht, S. S., Carmella, S. G., Murphy, S. E., Akerkar, S., Brunnemann, K. D. & Hoffmann, D.: A tobacco-specific lung carcinogen in the urine of men exposed to cigarette smoke. N. Engl. J. Med. 329: 1543--6, 1993. 161. Trichopoulos, D., Kalandidi, A. & Sparros, L: Lung cancer and passive smoking: Conclusion of G-reek study. Lancet, Sept 17, 2: 677--8, 1983. 162. Hole, 1). J., Gillia, C. JL, Chopra, C. & Haw- thorne, V. M.: Passive smoking and cardiorespi- ratory health in a general population in the west of Scotland. Br. Med. J. 299: 423-7, 1989. 47