Tobacco Documents Online

2063593255 n ~ n,plwry "'~

2063592254.

1.4: References Becher H, Zatonski W, Jockef KH. Passive smoking in Germany and Poland: comparison of exposure levels, sources of exposure, validity, and perception. Epidemiology 1992;3:509- 514. Boffetta P, Agudo A, Ahrens W, Benhamou E, Benhamou S, Darby SC, Ferro G, Fortes C, Gonzalez CA, Jockel KH, Krauss M, Kreienbrock L, Kreuzer M, Mendes A, Merietti F, Nyberg F, Pershagen G, Pohlabeln H, Riboli E, Schmid G, Simonato L, `l°redaniel J, Whitley E, Wichmann HE, Winck C, Zambon P, Saracci R. Multicenter case-control study of exposure to environmental tobacco smoke and lung cancer in Europe. J Natl Cancer Inst 1998;90 (in press). Bresiow NE, Day NE. Statistical Methods in Cancer Research, Vol. I, The Analysis of Case- Control Studies (IARC Scientific Publications No. 32). Lyon: International Agency for Research on Cancer, 1980. Jockei KH, Pohlabeln H, Ahrens W, Krauss M. Environmental tobacco smoke in Germany. Epidemiology (in press). Nyberg F, Agrenius V, Svartengren K, Svensson C, Pershagen G. Environmental tobacco smoke and lung cancer in nonsmokers - does time since exposure play a role? Epidemiology 1998a;9:301-308. Nyberg F, Agudo A, Boffetta P, Fortes C, Conzalez CA, Pershagen GA. European validation study of smoking status and environmental tobacco smoke exposure in nonsmoking lung cancer cases and controls. Cancer Causes Control 1998b;9:173-182. Riboli E, Preston-Martin S, Saracci R, Haley NJ, Trichopoulos D, Becher H, Burch JD, Fontham ETH, Gao YT, Jindal SK, Koo LC, Le Marchand L, Segnan N, Shimizu H, Stanta G, Wu-Williams AH, Zatonski W. Exposure of nonsmoking women to environmental tobacco smoke: a 10-country collaborative study. Cancer Causes Control 1998; 1:243-252.

INTERNATIONAL AGENC r FOR RESEARCH ON CANCER WORLD HEALTH ORGANfZATION European W-fuiticentre Case-Coritrol Study of Lung Cancer in Non-smokers Detailed Results on Exposure to Environmental Tobacco Smoke Paoio Boffetta, Antonio Agudo, Wolfgang Ahrens, Ellen Benhamou, Simone Benhamou, Sarah C. Darby, Gilles Ferro, Cristina Fortes, Carlos A. Gonzalez, Karl-Heinz Jockei, Martin Krauss, Lothar Kreienbrock, Fviichaela Kreuzer, Anabela Mendes, ~=ranco Merletti, Fredrik Nyberg, Goran Pershagen, Hermann Pohlabein, Elio Riboli, Giovanni Schmid, Lorenzo Simonato, Jean Tredaniei, Elise Whitley, Heinz-Erich Wichmann, Carlos Winck, Paola Zambon, Rodolfo Saracci IARC Technical Report No. 33 Lyon, 1998

1.2: Methods 1.2.1 Study Setting The study was conducted in twelve centres from seven European countries: three each from Germany and Italy, two from Portugal, and one each from Sweden, the United Kingdom, France and Spain. Details of the study design varied among centres; Table 2.1.1 presents selected features of the design. The period of enrolment of cases and controls lasted from 11988 to 1994. The most important difference in the study design among centres was the selection of controls: clinic-based in the centres from France, Portugal, Spain and in one Italian centre; both clinic- and community-based in the centre from the United Kingdom; and community-based in the other centres, The diagnoses of clinic-based controls varied among centres: patients with smoking-related diseases were excluded from the control series in all centres. i here were minor differences among centres regarding age restriction and diagnostic criteria for case eligibility. In some centres there was no age restriction, while in others subjects aged 76 or older were excluded. This combined analysis is therefore restricted to cases and controls up to age 75. Apart from a section on ETS (see below), the common questionnaire included sections on demographic variables, residential history including history of cooking and heating habits, and exposure to known and suspected occupational lung carcinogens. In addition, the centres from Germany, Sweden, Spain, the United Kingdom and France and one Italian centre collected information on dietary habits, from which indicators of intake of vegetables were derived based on country-specific tables of food composition. 1.2.2 The ETS Questionnaire Occasional smoking I he questionnaire included a section aimed at identifying regular smokers, defined as smokers of at least 400 cigarettes or the equivalent amount of cigars or tobacco for pipe. This amount corresponds to about one cigarette per day during one year. Smokers were to be excluded from the study and the questionnaire on ETS exposure was intended to be administered to never smokers. However, the identification of very weak smokers was problematic: therefore, the questionnaire on ETS exposure was introduced by a detailed questionnaire on occasional smoking, aiming to identify low level or discontinuous smoking habits.

(I) age (5 categories), sex, centre and sex-centre interaction terms; (ii) as (i), plus proportion of residence in urban setting during the last 35 years (3 categories); (iii) as (ii), plus education level (3 categories); (iv) as (i), plus duration of exposure to occupational lung carcinogens (4 categories); (v) as (i), plus frequency of consumption of vegetables (2 categories). Educational level was categorized in 3 levels, based on country specific cutpoints aimed to identify low (usually equivalent to primary school), intermediate (equivalent to iunior high school or secondary school) and high educational level (Table 2.1.8). Exposure to occupational carcinogens was categorized in 4 levels (unexposed, 1 st-75th percentile, ?5th-90th percentile and >90th percentile). Consumption of vegetables was categorized based on the median of the consumption of controls. Information on some of the confounders was not collected in selected centres. These cases and controls were excluded from the corresponding analyses (Table 2.1.3). All analyses were repeated after restriction to centres with information on education and residential history and those with information on vegetable consumption. The interaction terms between sex and centre were retained in all regression models since in preliminary analyses their inclusion improved significantly the fit of the models (Table 2.4.1). Their inclusion however affected the results of the ETS exposure indicators only to a minor extent. For selected exposure variables we compared the risk estimates derived from conditional and unconditional logistic regression for the four centres with individually matched data (France, Portugal 1, Spain and United Kingdom) (Table 2.4.2). The unconditional regression model was adjusted for sex, age and centre, as well as for the interaction term sex*centre as in model (i) above. In order to evaluate the differences between the centre-specific results a test for heterogeneity was calculated. This was done by comparing the deviances of regression models excluding and including the interaction terms between the exposure variables and the centre (Table 2.4.3). The difference in the deviances of these two models follows aX2 distribution with degrees of freedom equal to the number of interaction terms. A test for linear trend was calculated by including a trend variable in the logistic model and by testing the significance of the corresponding regression parameter. This was done using a categorical variable with values taken at the median of each exposure category of the variable (Table 2.4.4). We also produced scatter plots of the centre-specific results for selected variables by plotting the ORs against the inverse of their variance (section 2.4.5).

wo ETS from WORKPLACE WOEV Everexposed WOUD Duration (years) n _ jAage, non overlapping WOH Duration (hours) n _~(Aage x 52 x(hours / day) x (days / week)), non overlapping ~ WOWD Weighted duration (hours x level) » _~(Aage x 52 x TrVT3 x(hours / day) x(days / week)) , WOTS Time since last ETS exposure (years) ionnaire and descriotion of exoosure variables

Questionnaire on occasionai smoking and environmental tobacco smoke k'E TS) Tobacco smoking e Did you ever smoke for longer than one year? Yes..1; No..2 I or more cigarettes/day or half a pack of cigarettes/week or 2 or more packs/month or 4 or more cigarillos/week or 3 or more cigars/week or 3 or more pipes/week if YES to at least one of the questions, go to "Active smoking" section if NO, continue with the next question • Did you ever try to smoke? Yes..".; No ..2 1--i • Have you ever smoked cigarettes, cigar or pipe, even very few occasionally during a social occasion andlor at a particular period of your life? Yes ..1,; No ..2- 17 If yes, from what age did you smoke occasionally? What did you smoke? Cigarettes Filter ....................... I Non filter ................. 2 Cigar ........................ 3 Pipe .......................... 4 Cigarette & pipe ....... 5 ~ ~ i-j U • Did you inhale tobacco smoke? 1:1 Not at all, only in the mouth ....... I A little, just in the throat 2 ............. ~ Deeply into the lung ................... 3 If during this period you smoked mainly the same brand, indicate the brand (otherwise skip) How many? Frequency Per day ......... I Per week ...... 2 Per month ..... 3 Per year........ 4 Per lifetime ... 5 D o` ETS auestionnaire and description of exposure variables

For most quantitative variables, the categorization was done based on the distribution of controls, with cut-points chosen at the 75th and the 90th percentile. Such variables therefore have four categories(unexposed, ist-75th percentile, 76th-90th percentile and >90th percentile); their names start with the letter C. For selected quantitative variables an additional approach was chosen, after combining the unexposed and the lowest exposed groups: the names of these variables begin with the letter D (Table 2.1.2). The choice of the cut-point at the 75th percentile was based on a European study indicating that misclassification of ETS exposure is higher below the 75th percentile of the distribution [Becher et al., 1992]. In general, subjects with missing information on some items of the questionnaire, resulting in a missing exposure variable, were excluded from the analysis for that variable. However, if information for less than 25 percent of the duration of ETS exposure from cohabitants or at the workplace was missing, the average value of the rest of the period of exposure was assigned to the missing period. 1.2.5 Analysis of Subgroups of Cases and Controls Analyses were performed for all centres and for each centre separately. In addition, analyses were conducted on men and women and on centres with popufation-based controls (the three German centres, Italy 1, Italy 2, Sweden and United Kingdom) and centres with clinic- based controls (Itafy 3, France, Spain and the two centres from Portugal). An additional analysis excluded occasional smokers (164 cases and 438 controls who smoked between 1 and 400 cigarettes during their life). We conducted analyses stratified on age (<55, 55-64, and 65-75) and on type of residence (>75 percent of life in urban areas and >75 percent in non-urban areas). Finally, we restricted the analysis to histologically confirmed cases (627/650) and we classified them as squamous cell carcinoma (109 cases), small cell carcinoma (70 cases), adenocarcinoma (333 cases), and other and unknown histologies (115 cases) (Table 2.1.9). In the analysis by histological type, each group of cases was compared with the whole series of controls. 't .2.6 Validation Study We carried out a validation study by interviewing relatives of cases and controls on the non- smoking status of the study subjects. in one of the centres, spouses have also been interviewed regarding their own smoking habits, to validate the report of exposure to ETS ~ from the spouse given by the study subjects. The results of the validation study have been ~ reported separately [Nyberg et ai., 1998b] and are not discussed in detail here. ~ ~ ~ ~ ~ ~

C1 ETS during CHILDHOOD Childhood 0-18 years C'i EV C1 EM C1 EF CITP C'I UN Everexposed Ever exposed to mother Ever exposed to father Exposure to different tobacco products Number of smokers rt _ ~TTSE I C1WN Weighted number of smokers n WTSPE ~ E C1 US Smoking parent years n = 7 (AagexWTSE) I C1 WS Weighted smoking parent years ~ m ~ (dage x WTSPE ) *711 C1 UD Duration (years) n 7:age, non overlapping C1 WD Weighted duration (years) n = 7 (dagexWT,20 I €T_S auestionnair_e_andslescdaYioo_of-exaosure-variahias