LUNG CANCER AND CIGARETTE SMOKING IN WOMEN: A MULTICENTER CASE-CONTROL STUDY IN EUROPE Antonio AGUDO 1 * , Wolfgang AHRENS 2 , Ellen BENHAMOU 3 , Simone BENHAMOU 4 , Paolo BOFFETTA 5 , Sarah C. DARBY 6 , Francesco FORASTIERE 7 , Cristina FORTES 7 , Vale ´rie GABORIEAU 5 , Carlos A. GONZ ´ ALEZ 1 , Karl-Heinz J¨ OCKEL 8 , Michaela KREUZER 9 , Franco MERLETTI 10 , Hermann POHLABELN 2 , Lorenzo RICHIARDI 10 , Elise WHITLEY 11 , Heinz-Erich WICHMANN 9 , Paola ZAMBON 12 and Lorenzo SIMONATO 12 1 Catalan Institute of Oncology, L’Hospitalet, Barcelona, Spain 2 Bremen Institute for Prevention Research and Social Medicine, Bremen, Germany 3 National Institute of Health and Medical Research (INSERM), Paris, France 4 Institut Gustave Roussy, Villejuif, France 5 International Agency for Research on Cancer (IARC), Lyon, France 6 Imperial Cancer Research Fund, Oxford, UK 7 Epidemiology Unit Latium Region, Rome, Italy 8 Institute for Medical Informatics, Biometry and Epidemiology and West-German Cancer Center, Essen, Germany 9 GSF Institute of Epidemiology, Neuherberg, Germany 10 Unit of Cancer Epidemiology, University of Turin and Center for Oncologic Prevention, Turin, Italy 11 Department of Social Medicine, University of Bristol, Bristol, UK 12 Venetian Tumour Registry, University of Padua, Padua, Italy The association between cigarette smoking and lung can- cer risk in women was investigated within the framework of a case-control study in 9 centres from 6 European countries. Cases were 1,556 women up to 75 years of age with histo- logically confirmed primary lung cancer; 2,450 controls with age distribution similar to cases were selected. The predom- inant cell type was adenocarcinoma (33.5%), with similar proportions for squamous-cell type (26.4%) and small-cell carcinoma (22.3%). Overall, smoking cigarettes at any time was associated with a 5-fold increase in lung cancer risk (odds ratio 5.21, 95% confidence interval 4.49 – 6.04); corresponding figures for current smoking habits were 8.94, 7.54 –10.6. The association showed a dose-response relationship with dura- tion of the habit and daily and cumulative lifetime smoking. A significant excess risk of 70% was associated with every 10 pack-years smoked. After 10 years of smoking cessation, the relative risk decreased to 20% compared to current smokers. The following characteristics were associated with a higher relative risk: inhalation of smoke, smoking non-filter ciga- rettes, smoking dark-type cigarettes and starting at young age. The association was observed for all major histological types, being the strongest for small-cell type carcinoma, fol- lowed by squamous-cell type and the lowest for adenocarci- noma. The proportion of lung-cancer cases in the population attributable to cigarette smoking ranged from 14% to 85%. We concluded that women share most features of the asso- ciation between cigarette smoking and lung cancer observed in men. Int. J. Cancer 88:820 – 827, 2000. © 2000 Wiley-Liss, Inc. During the first half of this century, lung cancer was a rare disease in women, but after 1950, incidence and mortality rates have sharply increased in many developed countries. In 1987, lung cancer became the leading cause of cancer death in U.S. women (Ernster, 1994), and today, it accounts for 25% of estimated female cancer deaths compared with 16% for breast cancer (Landis et al., 1999). Within the European Union, age-standardised mortality has levelled off or declined for most cancer sites with the exception of lung cancer in women, which has risen persistently from 8.9% to 9.6  10 –5 between 1985–1989 and 1990 –1994 (Levi et al., 1999). Between 1970 and 1990, there was a sharp rise in female lung-cancer rates in all European Union countries except for Spain and Greece (Lo ´pez-Abente et al., 1995). The epidemic of lung cancer in women followed the widespread introduction of cigarette smoking. Over the past few decades, there has been a profound change in female smoking patterns in Euro- pean countries: the increase began earlier in northern countries (late 1960s or early 1970s), where prevalence has now peaked, while in most southern countries the proportion of women who smoke is still increasing in young birth cohorts (Graham, 1996). While there are wide variations in point prevalence, the patterns and trends are similar, with a curve characterised by an initial rise before the prevalence levels off. In most countries, it is young women in higher socio-economic groups who have led the way into cigarette smoking, and similarly, prevalence rates appear to level off and eventually fall first among women who are privileged in terms of education (Graham, 1996). Although a large body of evidence exists on the carcinogenic effects of tobacco smoke, most studies were based mainly on men, and some aspects of the risk associated with smoking have been limited in women by low statistical power. Furthermore, there may be differences between men and women regarding epidemiology of lung cancer, including different smoking behaviours (Ernster, 1994; Kreuzer et al., 2000), histological patterns (Wynder and Muscat, 1995) and host susceptibility (Risch et al., 1993; Zang and Wynder, 1996). A large series of female cases and controls have been studied within the framework of a collaborative European investigation on lung cancer. They provide a unique opportunity to update the assessment of the magnitude of risk of lung cancer among women in Europe, including a better knowledge of the risk associated with some qualitative and quantitative aspects of the habit. Grant sponsor: European Commission DG-XII; Grant number: EV5V- CT94-0555; Grant sponsor: Association pour la Recherche sur le Cancer, European Commission; Grant number: 90CVV01018; Grant sponsor: Caisse Nationale d’Assurance Maladie des Travailleurs Sociaux; Grant sponsor: Federal Ministry for Education, Science, Research and Technol- ogy (Germany); Grant number: 01 HK 546; Grant sponsor: Federal Min- istry of Work and Social Affairs (Germany); Grant number: IIIb 7-27/13; Grant sponsor: Federal Office of Radiation Protection (Germany); Grant numbers: St Sch 1066; St Sch 4047; St Sch 4074/1; St Sch 4006; St Sch 4112; Grant sponsor: Ministero dell’Universita ` e Ricerca Scientifica e Tecnologica (MURST); Grant sponsor: Italian Association for Cancer Research (AIRC); Grant sponsor: Regione Piemonte-Ricerca Finalizzata; Grant sponsor: National Research Council; Grant number: 91.00327.CT04; Grant sponsor: Spanish Ministry of Health; Grant number: 89002300; Grant sponsors: Imperial Cancer Research Fund; Department of Health (UK); Department of the Environment (UK). *Correspondence to: Department of Epidemiology and Cancer Regis- tration, Catalan Institute of Oncology (ICO), Av. Gran Via s/n, Km 2.7 E-08907 L’Hospitalet, Barcelona, Spain. Fax: +34 93 260 7783. E-mail: a.agudo@ico.scs.es Received 23 March 2000; Revised 20 June 2000; Accepted 21 June 2000 Int. J. Cancer: 88, 820 – 827 (2000) © 2000 Wiley-Liss, Inc. Publication of the International Union Against Cancer