Journal of Epidemiology Statistical Data J Epidemiol 2021;31(2):145-151 Who Smokes in Europe? Data From 12 European Countries in the TackSHS Survey (2017–2018) Silvano Gallus1, Alessandra Lugo1, Xiaoqiu Liu1, Panagiotis Behrakis2, Roberto Boffi3, Cristina Bosetti4, Giulia Carreras5, Liliane Chatenoud6, Luke Clancy7, Xavier Continente8,9,10, Ruaraidh Dobson11, Tobias Effertz12, Filippos T. Filippidis13, Marcela Fu14,15,16,17, Gergana Geshanova18, Giuseppe Gorini5, Sheila Keogan7, Hristo Ivanov18, María J. Lopez8,9,10, Angel Lopez-Nicolas19, José Precioso20, Krzysztof Przewozniak21,22,23, Cornel Radu-Loghin24, Ario Ruprecht3, Sean Semple11, Joan B. Soriano17,25, Polina Starchenko24, Marta Trapero-Bertran26, Olena Tigova14,15,17, Anna S. Tzortzi2, Constantine Vardavas2, Vergina K. Vyzikidou2, Paolo Colombo27, Esteve Fernandez14,15,16,17, and the TackSHS Project Investigators 1 Department of Environmental Health Sciences, Istituto di Ricerche Farmacologiche Mario Negri IRCCS, Milan, Italy 2 Institute of Public Health of the American College of Greece, Athens, Greece 3 Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy 4 Department of Oncology, Istituto di Ricerche Farmacologiche Mario Negri IRCCS, Milan, Italy 5 Istituto per lo Studio, la Prevenzione e la Rete Oncologica, Florence, Italy 6 Department of Public Health, Istituto di Ricerche Farmacologiche Mario Negri IRCCS, Milan, Italy 7 TobaccoFree Research Institute Ireland, Dublin, Ireland 8 Agència de Salut Pública de Barcelona, Barcelona, Spain 9 CIBER en Epidemiología y Salud Pública (CIBERESP) (Biomedical Research Centre Network for Epidemiology and Public Health), Madrid, Spain 10 Institut d’investigació Biomèdica Sant Pau (IIB St. Pau), Barcelona, Spain 11 University of Stirling, Stirling, Scotland 12 University of Hamburg, Hamburg Business School, Institute for Law & Economics, Hamburg, Germany 13 Department of Primary Care and Public Health, Imperial College, London, UK 14 Tobacco Control Unit, Institut Català d’Oncologia, L’Hospitalet de Llobregat, Barcelona, Spain 15 Tobacco Control Research Unit, Institut d’Investigació Biomèdica de Bellvitge, L’Hospitalet de Llobregat, Barcelona, Spain 16 Universitat de Barcelona, L’Hospitalet de Llobregat, Barcelona, Spain 17 Consortium for Biomedical Research in Respiratory Diseases (CIBER en Enfermedades Respiratorias, CIBERES), Madrid, Spain 18 Smoke Free Life Coalition, Sofia, Bulgaria 19 Universidad Politécnica de Cartagena, Cartagena, Spain 20 Instituto de Educação, Universidade do Minho, Braga, Portugal 21 Maria Sklodowska-Curie National Research Institute of Oncology, Warsaw, Poland 22 Foundation “Smart Health - Health in 3D”, Warsaw, Poland 23 Collegium Civitas, Warsaw, Poland 24 European Network for Smoking Prevention, Bruxelles, Belgium 25 Hospital Universitario La Princesa, Madrid, Spain 26 Research Institute for Evaluation and Public Policies (IRAPP), Universitat Internacional de Catalunya (UIC), Barcelona, Spain 27 DOXA Institute, Milan, Italy Received December 19, 2019; accepted January 21, 2020; released online April 4, 2020 ABSTRACT Background: Population data on tobacco use and its determinants require continuous monitoring and careful inter-country comparison. We aimed to provide the most up-to-date estimates on tobacco smoking from a large cross-sectional survey, conducted in selected European countries. Methods: Within the TackSHS Project, a face-to-face survey on smoking was conducted in 2017–2018 in 12 countries: Bulgaria, England, France, Germany, Greece, Ireland, Italy, Latvia, Poland, Portugal, Romania, and Spain, representing around 80% of the 432 million European Union (EU) adult population. In each country, a representative sample of around 1,000 subjects aged 15 years and older was interviewed, for a total of 11,902 participants. Results: Overall, 25.9% of participants were current smokers (31.0% of men and 21.2% of women, P < 0.001), while 16.5% were former smokers. Smoking prevalence ranged from 18.9% in Italy to 37.0% in Bulgaria. It decreased with increasing age (compared to <45, multivariable odds ratio [OR] for ≥65 year, 0.31; 95% confidence interval [CI], 0.27–0.36), level of education (OR for low vs high, 1.32; 95% CI, 1.17–1.48) and self-rated household economic level (OR for low vs high, 2.05; 95% CI, 1.74–2.42). The same patterns were found in both sexes. Address for correspondence. Silvano Gallus, ScD, Department of Environmental Health Sciences, Istituto di Ricerche Farmacologiche Mario Negri IRCCS, Via Mario Negri 2; 20156 Milan, Italy (e-mail: silvano.gallus@marionegri.it). The full list of TackSHS investigators is provided at the end of the manuscript. DOI https://doi.org/10.2188/jea.JE20190344 HOMEPAGE http://jeaweb.jp/english/journal/index.html 145 Smoking Prevalence in Europe: The TackSHS Survey Conclusions: These smoking prevalence estimates represent the most up-to-date evidence in Europe. From them, it can be derived that there are more than 112 million current smokers in the EU-28. Lower socio-economic status is a major determinant of smoking habit in both sexes. Key words: tobacco; cigarette smoking; smoking prevalence; cross-sectional study; survey; Europe; TackSHS Copyright © 2020 Silvano Gallus et al. This is an open access article distributed under the terms of Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. In this paper we aimed to describe the TackSHS survey, a INTRODUCTION multi-country survey conducted in 2017–2018 in 12 EU MS, In 2008, the World Health Organization (WHO) introduced the providing the main results on smoking prevalence and its MPOWER policy package, a series of technical measures and correlates in Europe. resources to assist country-level implementation of the WHO Framework Convention on Tobacco Control (FCTC), developed in response to the globalization of the tobacco epidemic. The first METHODS of these technical measures involves the monitoring of prevention The TackSHS survey18 has been conducted in 12 strategically policies and tobacco use.1 selected EU countries (Bulgaria, England, France, Germany, Several national surveillance systems are available to monitor Greece, Ireland, Italy, Latvia, Poland, Portugal, Romania, and smoking prevalence and trends in most countries.2 However, Spain), representing geographical, legislative and cultural data from country-specific populations are poorly comparable, variations across the EU. since they are not collected using a standardized study This cross-sectional study was coordinated by Istituto di design—including sampling methods, mode of interviewing Ricerche Farmacologiche Mario Negri IRCCS (Mario Negri and questionnaires—which represents a prerequisite for valid Institute; Milan, Italy). The fieldwork was conducted by DOXA, comparisons of smoking prevalence and smoking patterns across the Italian branch of the Worldwide Independent Network=Gallup different countries.3,4 International Association, and its European partners. To the best of our knowledge, the only standardized and representative investigations that systematically monitor smoking Sample selection prevalence among adults in more than one European country, are In each country, we considered a sample of around 1,000 the WHO Global Adult Tobacco Survey (GATS)5,6; the Survey of individuals aged 15 years and older (participants’ age was ≥16 in Health, Ageing and Retirement in Europe (SHARE)7; and the England, ≥18 in Ireland, 15–64 in Greece, and 15–74 years in European Commission (EC) Special Eurobarometers.8–10 GATS Latvia), representative of the general population in terms of age, covers three European Union (EU) member states (MS; ie, Greece, sex, habitat (ie, geographic area and=or size of municipality) and, Poland, and Romania)5,6; SHARE is a longitudinal study, con- in some countries, socio-economic characteristics. The survey ducted in five waves between 2004 and 2013, collecting infor- included a total of 11,902 subjects. The whole population domain mation on smoking habit among individuals aged 50 years and represented 79.2% of the whole EU population aged 15 years or older from 20 European countries7; Eurobarometer is a represent- over (432 million inhabitants). ative survey conducted in all the 28 EU MS on approximately Sampling methods varied across countries: in Bulgaria, 1,000 face-to-face interviews per country, based on selected Greece, Italy, Latvia and Romania, a multi-stage sampling was thematic studies. The most recent Eurobarometer surveys focusing used and respondents were randomly chosen to be representative on tobacco smoking indicated that the overall smoking prevalence of the population in terms of sex, age, and geographic area (in in the EU had slightly decreased from 29% in 2009 to 26% in 2014, Italy, representativeness by socio-economic characteristics was and remained stable until 2017.8,9,11 Among the other tobacco also ensured); in Germany, Ireland, Poland, Portugal, and Spain, control projects, the International Tobacco Control Policy stratified random sampling was used, combining also quotas on Evaluation Project (the ITC Project) includes longitudinal studies sex and age and social class in Ireland; in England, United in 10 European countries. However, all these studies, which Kingdom, cluster sampling with quotas on age, sex, socio- include samples not representative of the general population or economic status (SES), region, urban=rural dwelling was used to smokers only, cannot provide smoking prevalence estimates.12 The ensure national representativeness; in France, quotas on age, sex, above-mentioned European surveillance systems are important and region, and city size were used for the selection of survey useful tools; however, they are not always able to investigate in- participants. depth key tobacco control measures and smoking determinants. Consequently, a few specific studies, mainly supported by the EC, Questionnaire have been conducted to analyse this issue.3,13–15 These include, the A first draft of the survey questionnaire was developed in English Pricing Policies and Control of Tobacco in Europe (PPACTE) language by the researchers from Mario Negri Institute, based on survey, conducted in 2010 to investigate the effectiveness of previous tobacco use and secondhand smoke exposure ques- pricing policies on tobacco control in Europe3,16,17; and the present tionnaires.3,19–21 A specific commission with six experts among survey, conducted within the TackSHS Project (http:==tackshs. project partners was created to review the questionnaire and to eu),18 aimed to improve understanding of exposure to second-hand produce a second version. The final version of the English tobacco smoke (SHS) and e-cigarette aerosol in Europe. questionnaire was then developed through the collaboration with 146 j J Epidemiol 2021;31(2):145-151 Gallus S, et al. Table 1. Information about survey methods and sample characteristics in 12 European countries. TackSHS Project, 2017–2018 Sample Age range Survey Country Fieldwork dates Sampling method Representativeness sizea (years) mode Multi-stage stratified random Bulgaria Oct 2017 1,050 ≥15 CAPI Age, sex, habitat sampling with quota support England Jan–Feb 2018 1,013 ≥16 CAPI Cluster sampling with quota support Age, sex, habitat, SES France Nov–Dec 2017 1,018 ≥15 CAPI Quota method Age, sex, habitat Germany Jun 2018 1,031 ≥15 CAPI Stratified random sampling Age, sex, habitat Greece Jun–Jul 2018 1,000 15–64 CAPI Multi-stage random sampling Age, sex, habitat Stratified random sampling with Ireland Nov 2017 941 ≥18 CAPI Age, sex, habitat, SES quota support Italy Nov 2016 1,059 ≥15 CAPI Multi-stage random sampling Age, sex, habitat, SES Latvia Oct 2018 1,022 15–74 CAPI Multi-stage random sampling Age, sex, habitat Multi-stage stratified random Poland Sep 2018 724 ≥16 CAPI Age, sex, habitat sampling with cluster support Portugal Nov–Dec 2017 1,000 ≥15 CAPI Stratified sampling Age, sex, habitat Multi-stage stratified random Romania Jun–Jul 2017 1,018 ≥15 CAPI Age, sex, habitat, SES sampling with quota support Spain Oct 2017 1,026 ≥15 CAPI Stratified random sampling Age, sex, habitat Total Jun 2017–Oct 2018b 11,902 ≥15 CAPI — CAPI, computer assisted personal interview; SES, socio-economic status. a Raw sample size. b This period does not include the time of the survey conducted in Italy (ie, Nov–Dec 2016) since Italy was the pilot country. The participants from Italy are included in all the analyses. all the partners of the TackSHS consortium (eMaterial 1). cigarettes) during their lifetime. Current smokers were smokers The questionnaire was translated into Italian language by Mario who reported smoking at the time they participated in this survey, Negri Institute’s researchers, and into Bulgarian, French, German, while ex-smokers were smokers who stopped smoking by the Greek, Latvian, Russian, Polish, Portuguese, Romanian, and time they participated in this survey.3,21 Smokers also provided Spanish by DOXA partners. The translated versions of the information on age of starting smoking and the number of questionnaire were revised and validated by bilingual (ie, local cigarettes smoked per day. For each country, we calculated language and English language) tobacco control experts. the male-to-female smoking prevalence ratio as the current The questionnaire contains four sections: socio-economic smoking prevalence in men divided by that in women. The 12 and demographic characteristics; cigarette smoking habit and countries were classified into Northern (England, Ireland, and e-cigarette use; exposure to secondhand smoke (SHS) and Latvia), Western (France and Germany), Southern (Italy, Greece, e-cigarette aerosol, plus cigarette smoking and e-cigarette use, Portugal, and Spain), and Eastern (Bulgaria, Poland, and in different indoor and outdoor places; and attitudes and Romania), following United Nation M49 standard,22 and halved perceptions on smoke-free regulations and awareness of SHS by their gross domestic product (GDP) per capita23 into harmful effects. <25.000 € (Latvia, Romania, Poland, Portugal, Greece, and Bulgaria) and ≥25.000 € (England, France, Germany, Ireland, Fieldwork Italy, and Spain). Ad hoc trained interviewers conducted a face-to-face, computer- assisted personal interviewing (CAPI) survey in all the 12 Ethical issues European countries. A test fieldwork was conducted by DOXA in We obtained study approval from a local Ethics Committee Italy in November 2016 among 1,059 participants. The fieldwork in each of the 12 countries (eTable 1). Details on the survey in the other 11 countries was conducted between June 2017 characteristics were provided to all participants by suitably (in Romania) and October 2018 (in Latvia). Table 1 provides qualified professionals through a structured information sheet, information on survey characteristics for each country, including and all the participants provided their consents by ticking the fieldwork period, sample size, age range and sampling methods. electronic field in the CAPI questionnaire. The study protocol has been registered in ClinicalTrials.gov (ID: NCT02928536). Variables and definitions The procedures for recruitment of subjects, data collection, Demographic characteristics included age (categorized as <25, storage, and protection (based on anonymous identification code) 25–44, 45–64, ≥65 years) and sex (men and women). Level of are in accordance with the current country-specific legislation. education was categorized as country-specific tertiles of school- This was ratified and signed by DOXA and each of its European ing years. Self-assessment of household family economic status partners. relative to the country-specific population was classified into three levels (higher than average, average, and lower than Statistical analysis average). Never smokers were defined as participants who had Statistical weights were used to generate representative estimates never smoked or had smoked less than 100 cigarettes in their of the general population of each country (individual weight). lifetime. Smokers were defined as participants who reported To calculate results for the entire sample, we applied “country smoking at least 100 cigarettes (including roll-your-own weights”, which combined individual weights with an additional J Epidemiol 2021;31(2):145-151 j 147 Smoking Prevalence in Europe: The TackSHS Survey Table 2. Country-specific prevalence of never, ex- and current smokers, and smoking intensity in the European population aged ;15 years. TackSHS Project, 2017–2018 Prevalence (%) Number of cigarettes per day,a mean (SD) Country Never Current smokers Ex-smokers smokers Total Men Women M=F ratio Total Men Women Bulgaria 46.4 16.6 37.0 39.8 34.4 1.16 14.9 (7.9) 17.1 (8.2) 12.5 (6.8) England 59.9 20.2 19.8 20.3 19.5 1.04 15.5 (9.4) 15.0 (9.4) 15.8 (9.4) France 48.8 20.1 31.0 36.7 25.9 1.42 13.6 (8.7) 15.4 (9.0) 11.3 (7.7) Germany 63.4 12.9 23.7 32.4 15.3 2.12 15.9 (7.0) 17.1 (7.0) 13.4 (6.4) Greece 40.8 25.4 33.8 35.0 32.6 1.07 17.1 (15.8) 21.1 (18.9) 12.9 (10.1) Ireland 65.4 15.0 19.6 22.1 17.2 1.28 NA Italy 69.8 11.3 18.9 22.6 15.5 1.46 12.4 (7.5) 12.9 (7.9) 11.7 (7.1) Latvia 55.7 16.3 28.1 41.3 16.0 2.58 12.9 (7.0) 13.9 (6.8) 10.6 (7.0) Poland 62.9 13.5 23.6 28.7 19.4 1.48 16.4 (11.2) 18.7 (12.7) 13.6 (8.5) Portugal 48.5 14.7 36.8 42.0 32.1 1.31 14.1 (9.4) 16.2 (10.9) 11.7 (6.4) Romania 47.4 18.7 34.0 38.7 29.5 1.31 16.2 (10.0) 18.4 (10.4) 13.3 (8.9) Spain 47.8 20.4 31.8 37.3 26.6 1.40 14.0 (8.0) 14.5 (8.7) 13.3 (7.0) Totalb 57.6 16.5 25.9 31.0 21.2 1.46 14.7 (8.8) 16.0 (9.4) 13.0 (7.7) M=F, male to female; SD, standard deviation. a Smoking intensity, including both manufactured and roll-your-own cigarettes among current smokers (men and women combined). Questions on cigarette smoking intensity was not asked in Ireland. In the other 11 countries, 2,938 out of 3,050 current smokers provided information on the number of cigarettes smoked per day. b Country weights were applied, which combined individual weights with an additional weighting factor, with each country contributing in proportion to its population aged 15 years or over, obtained by Eurostat.24 weighting factor, with each country contributing in proportion Table 3 provides the ORs for current versus non-smokers to its population aged 15 years or over, obtained by Eurostat.24 according to selected individual-level and country-specific We considered descriptive statistics including relative fre- characteristics. Smoking prevalence was higher in men than in quency (%) and its corresponding 95% confidence intervals (CIs) women (OR 1.67; 95% CI, 1.53–1.82). Compared with subjects for categorical variables, and mean and standard deviation (SD) aged <45 years, ORs were 0.97 (95% CI, 0.89–1.07) for 45–64 for continuous variables. years, and 0.31 (95% CI, 0.27–0.36) for ≥65 years. Compared ORs and their corresponding 95% CIs were estimated using with high level of education, the ORs were 1.39 (95% CI, multilevel logistic random-effects models, to take into account 1.25–1.55) for medium and 1.32 (95% CI, 1.17–1.48) for low the heterogeneity between the 12 European countries. The level of education (P for trend <0.001). Compared to those with study country effects were considered as random intercepts, a high household economic status, the ORs were 1.33 (95% CI, and sex, age, and level of education as adjusting variables. 1.15–1.54) and 2.05 (95% CI, 1.74–2.42) for those with an Country weights were used in all logistic regression models. average and low economic status, respectively. The patterns Analyses were performed with SAS 9.4 (SAS Institute; Cary, NC, observed for all the individual-level characteristics were con- USA). sistent in men and women. Compared with Northern Europe, the ORs of smoking were 1.58 (95% CI, 0.98–2.56) in Western, 1.52 (95% CI, 0.93–2.49) in Southern and 1.57 (95% CI, 0.92–2.70) in RESULTS Eastern European countries. Compared to Northern Europe, the Among 11,902 participants, 57.6% (95% CI, 56.8–58.5%) OR in men was higher in Western Europe (OR 2.17; 95% CI, described themselves as never smokers (49.5% in men and 1.35–3.48), Southern Europe (OR 1.77; 95% CI, 1.09–2.87) and 65.1% in women), 16.5% (95% CI, 15.8–17.2%) as ex-smokers Eastern Europe (OR 1.87; 95% CI, 1.10–3.18). Smoking (19.5% in men and 13.7% in women) and 25.9% (95% CI, prevalence showed no significant difference between countries 25.1–26.7%) as current smokers (31.0% in men and 21.2% in with lower and higher GDP per capita, the ORs for less wealthy women; Table 2). The smoking prevalence was below 20% in countries being 1.22 (95% CI, 0.82–1.81). Italy (18.9%), Ireland (19.6%), and England (19.8%), and over 35% in Bulgaria (37.0%) and Portugal (36.8%). Smoking prevalence in men ranged from 20.3% in England to 42.0% in DISCUSSION Portugal, and in women from 15.3% in Germany to 34.4% in Approximately one out of four adults is a current smoker in Bulgaria. Average smoking intensity was 14.7 cigarettes per day Europe, including one out of three men and one out of five (SD, 8.8), ranging from 12.4 (SD, 7.5) in Italy to 17.1 (SD, 15.8) women. Smoking prevalence showed large differences across in Greece. European countries, being highest in Eastern Europe (28%) and Among current smokers, 76.1% started smoking before age 18 lowest in Northern Europe (20%). years and 96.3% before 25 years. Mean age at starting regular In both men and women, smoking prevalence is highest among smoking was 17.4 (SD, 4.5) overall, 17.1 (SD, 4.1) in males, and young adults aged 25–44 years. Almost all current smokers in 17.8 (SD, 4.8) years in female smokers, and varied between 15.2 Europe started smoking before aged 25 years. Therefore, this (SD, 4.2) in England and 19.1 years (SD, 4.2) in Romania (data study confirms that teenagers and young adults are the most not shown in tables). vulnerable group to start smoking3 and thus are the most 148 j J Epidemiol 2021;31(2):145-151 Gallus S, et al. Table 3. OR for current smokers vs non-smokers (never and ex-smokers combined) and corresponding 95% CI in the European population aged ;15 years, overall and by sex, according to selected individual-level and country-specific characteristics.a TackSHS Project, 2017–2018 Current smoking b Characteristics N Total Men Women % OR (95% CI)c % OR (95% CI)c % OR (95% CI)c Individual level characteristics Sex Women 6,270 21.2 1d Men 5,632 31.0 1.67 (1.53–1.82) Age group, years <25 1,446 24.5 29.4 19.1 1d 1d 1d 25–44 4,079 31.1 36.6 26.0 45–64 4,330 28.9 0.97 (0.89–1.07) 33.3 0.93 (0.82–1.05) 24.9 1.03 (0.90–1.17) ≥65 2,047 11.8 0.31 (0.27–0.36) 16.8 0.36 (0.29–0.43) 7.8 0.26 (0.21–0.33) P for trend <0.001 <0.001 <0.001 Level of educatione High 3,241 24.4 1d 27.4 1d 21.5 1d Medium 4,172 29.4 1.39 (1.25–1.55) 36.0 1.56 (1.35–1.82) 23.2 1.23 (1.05–1.45) Low 4,486 23.6 1.32 (1.17–1.48) 28.7 1.38 (1.18–1.61) 19.2 1.25 (1.06–1.48) P for trend <0.001 <0.001 0.012 Household economic statusf Higher than average 1,588 20.9 1d 23.3 1d 18.4 1d Average 6,264 25.9 1.33 (1.15–1.54) 29.6 1.38 (1.13–1.69) 22.4 1.29 (1.05–1.60) Lower than average 2,905 31.5 2.05 (1.74–2.42) 38.8 2.27 (1.80–2.85) 26.0 1.90 (1.50–2.41) P for trend <0.001 <0.001 <0.001 Country-specific characteristics Geographic area Northern Europe 2,976 20.1 1d 21.1 1d 19.2 1d Western Europe 2,049 26.9 1.58 (0.98–2.56) 34.3 2.17 (1.35–3.48) 20.0 1.13 (0.66–1.94) Southern Europe 4,085 26.2 1.52 (0.93–2.49) 30.5 1.77 (1.09–2.87) 22.2 1.35 (0.78–2.34) Eastern Europe 2,792 28.2 1.57 (0.92–2.70) 33.1 1.87 (1.10–3.18) 24.0 1.37 (0.75–2.50) GDP per capita >25.000 € 6,088 24.7 1d 30.0 1d 19.9 1d ≤25.000 € 5,814 29.9 1.22 (0.82–1.81) 34.6 1.16 (0.73–1.85) 25.8 1.30 (0.87–1.95) CI, confidence interval; GDP, gross domestic product; OR, odds ratio. a Country weights were applied, which combined individual weights with an additional weighting factor, with each country contributing in proportion to its population aged 15 years or over, obtained by Eurostat.24 b Raw sample size. c ORs and their corresponding 95% CIs were estimated using multilevel logistic random-effects models, to take into account the heterogeneity between the 12 European countries. The study country effects were considered as random intercepts, and sex, age and level of education as adjusting variables. Estimates in bold are statistically significant at 0.05 level. d Reference category. e Level of education was categorized as country-specific tertiles of schooling years. The sum does not add to the total because of a few missing values. f Self-assessment of household (family) economic status relative to the country-specific population. This variable is missing for all German subjects, 79 Latvian and, 35 Romanian subjects. important target population for tobacco control interventions measures specifically focused on low socio-economic groups to aimed to decrease nicotine addiction initiation. reduce health inequalities from tobacco. Smoking prevalence among women, exceeding 15% in all the In general, our findings are consistent with those reported in 12 countries, is in an advanced stage of the tobacco epidemic.3,25 the Global Health Observatory data repository of the WHO29 and However, notable differences were observed in the male-to- those found in the 2017 Eurobarometer survey.8 The latter study female smoking prevalence ratio, ranging from 1.04 in England to shows a smoking prevalence for the EU of 26% overall, 30% in 2.58 in Latvia. This suggests that some countries (likely those men, and 20% in women, thus very similar to our estimates. In with a higher male-to-female smoking prevalence ratio) are in an Italy, the smoking prevalence found in TackSHS survey (19% earlier phase compared with others.3,25 overall, 23% in men, and 16% in women) appears slightly lower Multilevel logistic analyses indicate that the level of education than that found in the 2016 national survey conducted on around was inversely related to smoking prevalence in both men and 3,000 subjects (ie, 22% overall, 27% in men, and 17% in women. Analysing other individual level and country specific women).30 Some major discrepancies with current evidence from indicators of SES, we also found that smoking prevalence was the scientific literature were noted particularly in Portugal, where systematically higher in the sub-population with low SES. This we found a smoking prevalence greater than 30% for both men finding was independent of sex, age, and education, as reported in and women. Our estimates are in fact substantially higher than a few other investigations.26 Our results are in broad agreement those reported before 2015 for daily smoking (lower than with current evidence,27,28 and underline the need for specific 20%).2,31,32 The different definition of current smoking and the J Epidemiol 2021;31(2):145-151 j 149 Smoking Prevalence in Europe: The TackSHS Survey different sampling methodology used are not sufficient to explain Funding: This project has received funding from the European the substantial gap found in smoking prevalence. Moreover, the Union Horizon 2020 research and innovation programme under distribution of our Portuguese sample by sex and age group is grant agreement No 681040. The Tobacco Control Unit (ICO) very similar to the official distribution provided by Eurostat.24 is partially supported by the Ministry of Universities and Other adult smoking prevalence estimates for Portugal, including Research, Government of Catalonia (2017SGR139). EFernandez those from the 2010 PPACTE study (32% overall, 36% in men, was partially supported by the Instituto de Salud Carlos III, and 29% in women),3 or the 2015 WHO estimates (32% in men Government of Spain (INT16=00211 and INT17=00103), co- and 14% in women),29 or the 2017 Eurobarometer survey (26% funded by the European Regional Development Fund (FEDER). overall)8 are more similar—but still substantially lower—than our ALugo was supported by an AIRC fellowship for Italy. The work estimates. of SGallus was partially funded by the Italian League Against Overall, the smoking prevalence remained almost the same Cancer (LILT, Milan). (from 27.2% to 26.4%) in the 11 countries that were included in Disclaimer: This paper reflects only the authors’ views and the both the TackSHS and PPACTE (2010) surveys, using similar European Commission is not responsible for any use that may be methodologies and definitions. For France, Portugal, Romania, made of the information it contains. and Spain, we observed increases in smoking prevalence over the Authors’ contributions: SGallus and EFernandez had the last decade but a decrease from 36% to 20% in Ireland. original study idea; EFernandez, GGorini, ALopez-Nicolas The trends in smoking prevalence found in our data match contributed to the finalization of the survey questionnaire; those from the Eurobarometer surveys: the comparison between PColombo provided data from the survey; XLiu and ALugo the 20178 and 200911 surveys indicated a decrease in smoking carried out the statistical analysis; SGallus drafted the article prevalence for all countries except for France and Portugal. in collaboration with XLiu and ALugo; all other authors made Smoking prevalence estimates in Europe remain substantially substantial contributions to conception, design and interpretation higher as compared to those from other high-income countries, of data; all the authors approved the final version of the such as the United States, Canada, and Australia, where less than manuscript. 15% of adults are current smokers.2,33 Conflicts of interest: None declared. The limitations of the present study include some differences of sampling in the study countries. Notwithstanding, all samples The TackSHS Project Investigators: selected were representative of their population in terms of age, Catalan Institute of Oncology (ICO); Bellvitge Biomedical sex, and geographic area and=or size of municipality. The age Research Institute (IDIBELL), Spain: Esteve Fernández, range of the participants was slightly different in some countries. Yolanda Castellano, Marcela Fu, Montse Ballbè, Beladenta In particular, in Greece the sample was limited to subjects aged Amalia, Olena Tigova 15–64 years, thus likely providing an over-estimation of the adult Public Health Agency of Barcelona (ASPB), Spain: Maria José smoking prevalence. However, even assuming that the smoking López, Xavier Continente, Teresa Arechavala, Elisabet Henderson prevalence in the elderly (representing in Greece 25% of the adult Istituto di Ricerche Farmacologiche Mario Negri IRCCS population24) was 50% that in younger adults—as observed in the (IRFMN), Italy: Silvano Gallus, Alessandra Lugo, Xiaoqiu Liu, other 11 countries—we would estimate a smoking prevalence Cristina Bosetti, Enrico Davoli, Elisa Borroni; around 30%, still similar to the provided estimate. Istituto DOXA, Worldwide Independent Network=Gallup The strengths of our survey include the representativeness International Association, Italy: Paolo Colombo of the adult population of the 12 selected European countries, University of Stirling (UNISTIR), the UK: Sean Semple, the use of the same questionnaire in the 12 countries sampled, Rachel O’Donnell, Ruaraidh Dobson developed by a group of experts on tobacco control, the TobaccoFree Research Institute Ireland (TFRI), Ireland: standardized use of a single definition of current smokers, and Luke Clancy, Sheila Keogan, Hannah Byrne the use of face-to-face interviews. Hellenic Cancer Society - George D. Behrakis Research Lab Our data indicate that there are 112 million current smokers (HCS), Greece: Panagiotis Behrakis, Anna Tzortzi, Constantine in the EU, including 65 million men and 47 million women. We Vardavas, Vergina Konstantina Vyzikidou, Gerasimos Bakelas, confirm that the large majority of smokers started smoking when George Mattiampa they were adolescents, and that lower SES is a major determinant Fondazione IRCCS Istituto Nazionale dei Tumori (INT), of smoking habit in Europe. Tobacco control measures to Italy: Roberto Boffi, Ario Ruprecht, Cinzia De Marco, decrease smoking initiation and to promote smoking cessation Alessandro Borgini, Chiara Veronese, Martina Bertoldi, Andrea should, therefore, be targeted at young people and the sub-group Tittarelli with lower SES, respectively. Increasing tobacco prices, through Istituto per lo Studio, la Prevenzione, e la Rete Oncologica the adoption of additional excise taxes, has been shown to be (ISPRO), Italy: Giuseppe Gorini, Giulia Carreras, Barbara more effective among younger generations and in economically Cortini, Simona Verdi, Alessio Lachi, Elisabetta Chellini deprived populations.28,34 Therefore, price increase is still likely Polytechnic University of Cartagena (UPCT), Spain: Ángel to be a particularly effective tobacco control strategy, particularly López Nicolás, Marta Trapero-Bertran, Daniel Celdrán Guerrero in those European countries where tobacco price is still relatively European Network on Smoking and Tobacco Prevention low and resources for tobacco control are scarce.35 (ENSP), Belgium: Cornel Radu-Loghin, Dominick Nguyen, Polina Starchenko Hospital Universitario La Princesa (IISP), Spain: Joan B ACKNOWLEDGEMENTS Soriano, Julio Ancochea, Tamara Alonso, María Teresa Pastor, IDIBELL group thanks CERCA Programme=Generalitat de Marta Erro, Ana Roca, Patricia Pérez, Elena García Castillo Catalunya for institutional support. 150 j J Epidemiol 2021;31(2):145-151 Gallus S, et al. a cross-sectional survey. Tob Control. 2014;23(e1):e17–e23. APPENDIX A. SUPPLEMENTARY DATA 18. Fernández E, López MJ, Gallus S, et al; TackSHS Project Investigators; TackSHS Project Investigators. Tackling second-hand Supplementary data related to this article can be found at https:== exposure to tobacco smoke and aerosols of electronic cigarettes: the doi.org=10.2188=jea.JE20190344. TackSHS project protocol. Gac Sanit. 2020;34(1):77–82. 19. Pérez-Ríos M, Schiaffino A, López MJ, et al. Questionnaire-based second-hand smoke assessment in adults. Eur J Public Health. 2013; REFERENCES 23(5):763–767. 1. WHO. WHO Report on the Global Tobacco Epidemic, 2008: The 20. Sureda X, Martínez-Sánchez JM, Fu M, et al. Impact of the Spanish MPOWER package. Geneva, World Health Organization. Available smoke-free legislation on adult, non-smoker exposure to secondhand online at https:==www.who.int=tobacco=mpower=mpower_report_ smoke: cross-sectional surveys before (2004) and after (2012) full_2008.pdf (last access: 10 July 2019). 2008. legislation. PLoS One. 2014;9(2):e89430. 2. GBD 2015 Tobacco Collaborators. Smoking prevalence and attrib- 21. IARC. IARC Handbooks of Cancer Prevention Volume 12. Methods utable disease burden in 195 countries and territories, 1990–2015: a for Evaluating Tobacco Control Policies. Available online at https:== systematic analysis from the Global Burden of Disease Study 2015. www.iarc.fr=wp-content=uploads=2018=07=Tobacco_vol12.pdf (last Lancet. 2017;389(10082):1885–1906. accessed on: 1 August 2019). 2008. 3. Gallus S, Lugo A, La Vecchia C, et al. Pricing Policies And Control 22. Statistics Division. Department of Economic and Social Affairs. of Tobacco in Europe (PPACTE) project: cross-national comparison United Nations Standard country or area codes for statistical use of smoking prevalence in 18 European countries. Eur J Cancer Prev. (M49). 1999. 2014;23(3):177–185. 23. World Bank. GDP per capita (current US$), World Bank national 4. Bogdanovica I, Godfrey F, McNeill A, Britton J. Smoking accounts data and OECD National Accounts data files. Available prevalence in the European Union: a comparison of national and online at https:==data.worldbank.org=indicator=NY.GDP.PCAP.CD? transnational prevalence survey methods and results. Tob Control. locations=EU (last accessed on: 1 August 2019). 2018. 2011;20(1):e4. 24. European Commission. Eurostat. Database. Available online at 5. Centers for Disease Control and Prevention. Global Adult Tobacco https:==ec.europa.eu=eurostat=data=database (last access: 10 July Survey Collaborative Group. Global Adult Tobacco Survey (GATS): 2019). Core Questionnaire with Optional Questions, Version 2.0. Atlanta, 25. Lopez AD, Collishaw NE, Piha T. A descriptive model of the GA. cigarette epidemic in developed countries. Tob Control. 1994;3(3): 6. Mbulo L, Palipudi KM, Nelson-Blutcher G, Murty KS, Asma S; 242–247. Global Adult Tobacco Survey Collaborative Group. The process of 26. Huisman M, Kunst AE, Mackenbach JP. Inequalities in the cessation among current tobacco smokers: a cross-sectional data prevalence of smoking in the European Union: comparing education analysis from 21 countries, global adult tobacco survey, 2009–2013. and income. Prev Med. 2005;40(6):756–764. Prev Chronic Dis. 2015;12:E151. 27. Bosdriesz JR, Willemsen MC, Stronks K, Kunst AE. Tobacco 7. Serrano-Alarcón M, Kunst AE, Bosdriesz JR, Perelman J. Tobacco control policy and socio-economic inequalities in smoking in 27 control policies and smoking among older adults: a longitudinal European countries. Drug Alcohol Depend. 2016;165:79–86. analysis of 10 European countries. Addiction. 2019;114(6):1076– 28. Hiscock R, Bauld L, Amos A, Fidler JA, Munafò M. Socioeconomic 1085. status and smoking: a review. Ann N Y Acad Sci. 2012;1248:107– 8. Special Eurobarometer 458 report. Attitudes of Europeans towards 123. tobacco and electronic cigarettes. 2017. 29. WHO. Tobacco use. Data by country. Global Health Observatory 9. Special Eurobarometer 429 report. Attitudes of Europeans towards data repository. Available online at http:==apps.who.int=gho=data= tobacco and electronic cigarettes. 2015. node.main.65 (last accessed on: 10 December 2019). 2015. 10. Special Eurobarometer 485 report. Attitudes of Europeans towards 30. Lugo A, Zuccaro P, Pacifici R, et al. Smoking in Italy in 2015–2016: tobacco. 2012. prevalence, trends, roll-your-own cigarettes, and attitudes towards 11. Special Eurobarometer 332 report. Tobacco. 2010. incoming regulations. Tumori. 2017;103(4):353–359. 12. ITC Project. International Tobacco Control Policy Evaluation Project. 31. Carreira H, Pereira M, Azevedo A, Lunet N. Trends in the Available online at https:==itcproject.org= (last accessed on: 08 prevalence of smoking in Portugal: a systematic review. BMC January 2020). Public Health. 2012;12:958. 13. Balia S. Survival expectations, subjective health and smoking: 32. Drope J, Schluger N, Cahn Z, et al. The Tobacco Atlas. Atlanta: evidence from SHARE. Empir Econ. 2014;47(2):753–780. American Cancer Society and Vital Strategies. 2018. 14. Trias-Llimós S, Muszyńska MM, Cámara AD, Janssen F. Smoking 33. CDC. Centers for Disease Control and Prevention. Current Cigarette cessation among European older adults: the contributions of marital Smoking Among Adults in the United States. Available online at and employment transitions by gender. Eur J Ageing. 2017;14(2): https:==www.cdc.gov=tobacco=data_statistics=fact_sheets=adult_data= 189–198. cig_smoking=index.htm (last access: 10 July 2019). 2019. 15. ECRHS. European Community Respiratory Health Survey. Avail- 34. IARC. IARC Handbooks of Cancer Prevention. Vol. 14. Effective- able online at http:==www.ecrhs.org=Default.htm (last access: 10 ness of Tax and Price Policies for Tobacco Control. Lyon, France; July 2019). 2011. 16. Gallus S, Lugo A, Ghislandi S, La Vecchia C, Gilmore AB. Roll- 35. Joossens L, Feliu A, Fernandez E. The Tobacco Control Scale 2019 your-own cigarettes in Europe: use, weight and implications for in Europe. Brussels: Association of European Cancer Leagues, fiscal policies. Eur J Cancer Prev. 2014;23(3):186–192. Catalan Institute of Oncology. Available from: http:==www. 17. Joossens L, Lugo A, La Vecchia C, Gilmore AB, Clancy L, Gallus tobaccocontrolscale.org=TCS2019.pdf (last access: 30 June 2020). S. Illicit cigarettes and hand-rolled tobacco in 18 European countries: 2020. J Epidemiol 2021;31(2):145-151 j 151