Nienke Boderie

Towards A Smoke-Free Generation Novel Strategies To Phase Out Tobacco Use Nienke Boderie

Towards a smoke-free generation: novel strategies to phase out tobacco use Op weg naar een Rookvrije Generatie: nieuwe strategieën om tabaksgebruik uit te faseren Nienke Wilhelmina Boderie

Towards a smoke-free generation: novel strategies to phase out tobacco use Doctoral thesis, Erasmus Univeristy Rotterdam Nienke W. Boderie Copyright © Nienke W. Boderie All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without prior permission of the author or the copyright owning journals for previously published chapters Layout & Printed by Ridderprint | www.ridderprint.nl This thesis was printed with financial support of the Department of Public Health, Erasmus Medical Center and of the Erasmus Univeristy

Towards a smoke-free generation: novel strategies to phase out tobacco use Op weg naar een Rookvrije Generatie: Nieuwe strategieën om tabaksgebruik uit te faseren Proefschrift ter verkrijging van de graad van doctor aan de Erasmus Universiteit Rotterdam op gezag van de rector magnificus Prof.dr. A.L. Bredenoord en volgens besluit van het College voor Promoties De openbare verdediging zal plaatsvinden op vrijdag 30 Augustus 2024 om 13.00 uur door Nienke Wilhelmina Boderie geboren te Zoetermeer.

Promotiecommissie Promotoren Prof.dr. F.J. van Lenthe Prof.dr. J.L.W. van Kippersluis Overige leden Prof.dr. K. Stronks Prof.dr. M.C. Willemsen Dr. J.E. Roeters van Lennep Co-promotor Dr. J.V. Been

Science is made up of so many things that appear obvious after they are explained Frank Herbert, Dune, p443

Contents Chapter 1 General introduction 9 Chapter 2 Socioeconomic inequalities in smoking-attributable mortality in Europe; understanding trends 2000-2020 Submitted to Nicotine & Tobacco Research 19 Part I Chapter 3 Assessing public support for extending smoke-free policies beyond enclosed public places and workplaces: protocol for a systematic review and meta-analysis BMJ Open 2021;11:e040167 57 Chapter 4 Public support for smoke-free policies in outdoor areas and (semi-) private places: a systematic review and metaanalysis eClinicalMedicine 2023;59: 101982 73 Chapter 5 Public support for smoke-free private indoor and public outdoor areas in the Netherlands: a trend analysis from 2018-2022 Tobacco Induced Diseases 2024; 22 211 Chapter 6 Smokers’ responses to being addressed when smoking in an outdoor voluntary smoke-free zone: An observational study Tobacco Prevention and Cessation 2021; 7 229 Chapter 7 The next step for a smoke-free generation: a multidisciplinary study of opportunities for expanding smoke-free environments in the Netherlands (translated from Dutch) TSG-Tijdschrift voor gezondheidswetenschappen 2023; 101(2), 21-28. 241

Part II Chapter 8 A social care programme’s impact on quit-smoking intention in multi-problem households: exploring the scarcity theory 259 Chapter 9 Deposit? Yes, please! The effect of different modes of assigning reward-and deposit-based financial incentives on effort Behavioural Public Policy 2023; 1-29. 277 Chapter 10 PERSonalised Incentives for Supporting Tobacco cessation (PERSIST) among healthcare employees: a randomised controlled trial protocol BMJ open 2020; 10(9), e037799 337 Chapter 11 PERSonalised Incentives for Supporting Tobacco cessation (PERSIST) among healthcare employees: evaluation and lessons learned Invited for resubmission at Critical Public Health 357 Chapter 12 General discussion 391 Summary 410 Samenvatting 416 List of Publications 422 Contributing authors 424 PhD portfolio 430 About the author 433 Dankwoord 434

1 Chapter

GENERAL INTRODUCTION

Chapter 1 10 Tobacco use is one of the biggest public health threats to the world, claiming the lives of 8 million people annually.1 Despite the alarming statistic that one in two tobacco users will meet a premature death, tobacco industries continue to thrive, selling to over 1.3 billion smokers worldwide. The start of the tobacco epidemic can be marked in 1880 with the invention of the first practical cigarette cutting machine.2 Since then, the spread of tobacco use and its associated adverse health impact has been described using the tobacco epidemic model (Figure 1). In Western Europe the first stage of this epidemic began in the early 1900s, predominantly among men. In the second stage, smoking rates increased sharply among men and first signs of tobacco related mortality showed. In this stage women started smoking as well. In the 1950s tobacco use peaked among men and the sharp increase in tobacco-related mortality heightened awareness of its detrimental health effects. In the fourth phase, smoking rates decreased among both men and women. Mortality rates remained high, or started to decline among men. Figure 1: Tobacco epidemic model, used from Lopez et al. Tobacco Control 1994; 3; 242-247 The emergence of stage 3 in Western countries coincided with the release of the 1964 U.S. Surgeon General’s report on lung cancer and tobacco, which was one

General introduction 11 1 of the first reports to draw attention to the lethal consequences of tobacco use. However, it quickly became evident that in itself a desire to quit smoking often proved insufficient to enable successful cessation. Smoking appeared to be not only highly addictive but also profoundly influenced by social and environmental factors; it made cessation a formidable challenge.3 Nowadays, the negative effects of smoking extend well beyond the adverse health effects among individuals, with large losses in productivity and excess health care costs each year.4 On top of that, tobacco production has been estimated to have contributed up to 20% of annual greenhouse gas increases and cigarette butts and other waste forms contaminate beaches and waterways.5 As a response to the wide range of negative effects of tobacco use, many governments implemented tobacco control policies. Tobacco control In 2003, the World Health Organization (WHO) launched the Framework Convention on Tobacco Control (FCTC) in which 168 countries committed themselves “to protect present and future generations from the devastating health, social, environmental and economic consequences of tobacco consumption and exposure to tobacco smoke”. Within the FCTC framework, the WHO introduced a concise set of essential tobacco control policies under the acronym MPOWER, comprising six key components: (1) Monitor tobacco use and prevention policies, (2) Protect people from tobacco smoke, (3) Offer help to quit smoking, (4) Warn about the dangers of tobacco, (5) Enforce bans on tobacco advertising, promotion and sponsorship, and (6) Raise taxes on tobacco. Nowadays, 5.6 billion people, or 70% of the world population, are protected by at least one MPOWER measure at the highest level, meaning that all recommended measures are implemented. In the past 15 years the number of countries with comprehensive smoke-free legislation (i.e. the ‘P’ in MPOWER) increased from 10 (i.e. 5% of countries worldwide) to 74 (38%).6 Policies such as an increased tobacco taxation, smoke-free air regulations and cessation support have significantly reduced the prevalence of smoking and its associated mortality.7

Chapter 1 12 Future tobacco control To build on these past successes and make further progress in tobacco control, additional steps are needed. Preventing youth from initiating smoking and raising them in a tobacco-free environment is the most effective way to do so. This approach, known as creating ‘a smoke-free generation’, has been adopted by many countries as part of their tobacco endgame strategies. In order to reach a smoke-free generation, tobacco control needs to step up its game, building on the successes of previous measures. An example of such a strategy is to expand smoke-free zones from indoor public places and workplaces to novel areas, such as outdoor public places and indoor private places. Another pathway could be to explore novel ways to further promote quit rates. Ideally, a range of measures focused on protection, prevention and cessation support are implemented simultaneously. Smoke-free zones In terms of smoking cessation, smoke-free zones diminish the visibility of smoking, assisting recent quitters in maintaining abstinence, normalizing smokefree environments, and deterring youth from adopting smoking behaviour.11,12 In a recent survey in the Netherlands 25% of smokers stated that smoke-free places such as terraces, entrances to health care facilities or parks would help them smoke less. Furthermore, smoke-free zones have proved to be effective in protecting non-smokers against second-hand smoke exposure. Smokefree policies consistently reduce exposure to second-hand smoke and do not increase exposure in other places such as homes. Also important, reductions in second-hand smoke are sustainable and do not reverse over time.15,16 Moreover, smoke-free policies have had a substantial effect on child health via reductions in preterm birth rates, hospital attendance for asthma and hospitalisation due to respiratory tract infections. In line with the MPOWER guidelines, smoke-free zones are recommended for indoor public places and workplaces, but expanding these policies to outdoor public and indoor private spaces can further improve public (child) health. Implementing smoke-free policies in such novel settings requires public support, which not only is crucial for convincing policy makers, but also fosters a social norm change required for ensuring compliance with the policy. If public support for a smoke-free zone is high, smoking in that area may no longer be socially accepted. A seemingly effective way to increase support for smoke-free zones is the smoke-free generation framework, in which tobacco

General introduction 13 1 control policies are framed positively, i.e., from the perspective of protecting child health rather than restricting smoking.19 Smoking cessation Achieving a smoke-free generation promises long-term public health benefits, but in the short term, it is also essential to further support current tobacco users in quitting. Although 70% of tobacco users express a desire to quit at some point in their lives, only 30% attempt to do so, and among them less than 10% reports a successful quit attempt in the last year.21,22 Research indicates that nicotine replacement therapy (NRT) and individual or group-based counselling can enhance the success rate of quit attempts, but most individuals attempt to quit without assistance.23 Additionally, many cessation programs offer one-size-fits-all solutions, despite significant personality differences among smokers. As smoking is influenced by social and environmental contexts, integrating contextual and personal factors into cessation programs could be a promising approach. Socioeconomic position and smoking cessation Historically, smoking was introduced into society by men of higher socioeconomic standing. Nowadays however, it is most prevalent among individuals with a lower socioeconomic position (SEP). For instance, in the Netherlands, the average smoking prevalence in 2021 was 20%, while among lower educated men between the age of 25 and 45 it was 40%.24 That smoking is more prevalent among individuals with a lower SEP is not solely due to a lack of knowledge or motivation. Indeed, research has shown no social gradient in the intention to quit, quit attempts, use of medication to aid quitting, or the use of stop smoking services. 25 26 Lower SEP individuals do however, face many additional challenges when trying to quit smoking. Often, their addiction is stronger as a result of younger age of smoking initiation and of larger quantities of tobacco used. And finally, when trying to quit, lower SEP smokers often face detrimental conditions to quit smoking, and often experience less peer support due to higher acceptability of smoking in their social environments,29 or due to the poorer quality social networks.30 31 Stress and smoking cessation In addition to, or stemming from, external factors influencing the number and success of quit attempts, personal characteristics may facilitate or hamper quit

Chapter 1 14 success. Stress due to material conditions, or the lack of those, makes it harder to focus on the future. Immediate pressing issues take up the mental bandwidth to think about long term effects of (unhealthy) decisions. Economists refer to this prioritization of the immediate future over the distant future as delay discounting, or a reduced temporal horizon.32 Providing short-term rewards or incentives can help overcome this bias and make quitting more attractive. Delay discounting often presents in individuals with high stress levels. Another aspect of higher stress levels can be reduced cognitive bandwidth. According to Mullainathan and Safir’s theory of scarcity, stress can limit one’s cognitive bandwidth, limiting the mental capacity need to make deliberate and undeliberate (healthy) choices.34 Tackling the sources of stress in tobacco users’ lives may have potential to increase this cognitive bandwidth and improve the success rate of cessation programmes. Towards a smoke-free generation To further reduce the negative impact of tobacco use in society there is thus a need to explore additional strategies to protect and prevent youth from smoking and to improve success rates of quit attempts among those who currently smoke. This thesis will investigate novel approaches for both these aspects in order to contribute to a smoke-free generation Aims of this thesis: In this thesis, I aim to investigate strategies to phase out tobacco use by focussing on protecting against and preventing smoking, as well as on improving quit attempts among current tobacco users. The two key objectives are: • To evaluate public support for and implementation aspects of smoke-free policies in public outdoor and (semi)private places (part I) • To investigate new approaches to improve smoking cessation programmes using personalisation and financial incentives (part II) Outline In order to inform policies aimed at protecting people against the harms of second-hand smoke exposure and decreasing future trends in tobacco use this thesis will investigate smoke-free policies that go beyond indoor public places and workplaces (Part I) and we investigate strategies to incorporate personal

General introduction 15 1 characteristics or circumstances into smoking cessation programmes (Part II). First, I investigated the socioeconomic gradient in trends in smoking-attributable mortality in Europe (Chapter 2). Part I starts with a description the methodology (Chapter 3) and findings (Chapter 4) of a systematic review evaluating public support across the globe for smoke-free policies that go beyond indoor public places and workplaces. Chapter 5 investigates recent trends in support for novel smoke-free policies in the Netherlands specifically. In 2019, Rotterdam was one of the first Dutch cities to launch an outdoor smoke-free zone, which encompasses an area where the Erasmus Medical Center Rotterdam and two school are located. In Chapter 6 we explore how people who smoked in this outdoor smoke-free zone responded to being addressed about their smoking behaviour. In Chapter 7, we describe results from a multidisciplinary team effort to explore various aspects relevant to implementing smoke-free policies in outdoor and (semi)private places and how these apply to the Netherlands specifically. Part II starts with Chapter 8, in which we examine how gaining control over external stressors impacts (the intention to change) health behaviour among people living in vulnerable circumstances, including smoking. Further, we explored how incentives may most efficiently be implemented to support potential behaviour change (Chapter 9). In Chapter 10, we used that information to design a randomised controlled trial to investigate the effect of personalised incentives on sustained abstinence from smoking among hospital employees (Chapter 10). Chapter 11 presents the findings of this study, including the lessons we learned during its implementation.

Chapter 1 16 References 1. The Lancet: Global Burden of Disease. https://www.thelancet.com/gbd. 2. Wipfli, H. & Samet, J. M. One Hundred Years in the Making: The Global Tobacco Epidemic. Annu. Rev. Public Health 37, 149–166 (2016). 3. Macintyre, S. & Ellaway, A. Ecological Approaches: Rediscovering the Role of the Physical and Social Environment. in Social Epidemiology vols 332–348 (2000). 4. Reitsma, M. B. et al. Smoking prevalence and attributable disease burden in 195 countries and territories, 1990–2015: a systematic analysis from the Global Burden of Disease Study 2015. The Lancet 389, 1885–1906 (2017). 5. World Health Organisation, Tobacco: poisoning our planet. https://iris.who.int/bitstream/ handle/10665/354579/9789240051287-eng.pdf?sequence=1 (2022). 6. WHO report on the global tobacco epidemic 2023. https://www.who.int/teams/healthpromotion/tobacco-control/global-tobacco-report-2023. 7. Levy, D. T., Huang, A.-T., Havumaki, J. S. & Meza, R. The Role of Public Policies in Reducing Smoking Prevalence: Results from the Michigan SimSmoke Tobacco Policy Simulation Model. Cancer Causes Control CCC 27, 615–625 (2016). 8. The tobacco endgame: a qualitative review and synthesis | Tobacco Control. https:// tobaccocontrol.bmj.com/content/25/5/594. 9. Puljević, C. et al. Closing the gaps in tobacco endgame evidence: a scoping review. Tob. Control 31, 365–375 (2022). 10. McDaniel, P. A., Smith, E. A. & Malone, R. E. The tobacco endgame: a qualitative review and synthesis. Tob. Control 25, 594–604 (2016). 11. Kelly, B. C., Vuolo, M., Frizzell, L. C. & Hernandez, E. M. Denormalization, smoke-free air policy, and tobacco use among young adults. Soc. Sci. Med. 211, 70–77 (2018). 12. Semple, S. et al. Smoke-free spaces: a decade of progress, a need for more? Tob. Control 31, 250–256 (2022). 13. Fernández, E. et al. Changes in Secondhand Smoke Exposure After Smoke-Free Legislation (Spain, 2006–2011). Nicotine Tob. Res. 19, 1390–1394 (2017). 14. Tsai, Y.-W., Chang, L.-C., Sung, H.-Y., Hu, T. & Chiou, S.-T. The impact of smoke-free legislation on reducing exposure to secondhand smoke: differences across gender and socioeconomic groups. Tob. Control 24, 62–69 (2015). 15. Haw, S. J. & Gruer, L. Changes in exposure of adult non-smokers to secondhand smoke after implementation of smoke-free legislation in Scotland: national cross sectional survey. BMJ 335, 549 (2007). 16. Al-Delaimy, W., White, M., Gilmer, T., Zhu, S. & Pierce, J. P. The California Tobacco Control Program: Can We Maintain the Progress? Results from the California Tobacco Survey, 1990-2005. (2023). 17. Faber, T. et al. Effect of tobacco control policies on perinatal and child health: a systematic review and meta-analysis. Lancet Public Health 2, e420–e437 (2017). 18. Radó, M. K. et al. Effect of smoke-free policies in outdoor areas and private places on children’s tobacco smoke exposure and respiratory health: a systematic review and meta-analysis. Lancet Public Health 6, e566–e578 (2021). 19. Willemsen, M. C. & Been, J. V. Accelerating tobacco control at the national level with the Smoke-free Generation movement in the Netherlands. Npj Prim. Care Respir. Med. 32, 58 (2022). 20. Willemsen, M. C. & Been, J. V. Accelerating tobacco control at the national level with the Smoke-free Generation movement in the Netherlands. Npj Prim. Care Respir. Med. 32, 1–6 (2022).

General introduction 17 1 21. Zimmerman, A. M. Minder mensen stoppen met roken. Trimbos-instituut https://www. trimbos.nl/actueel/nieuws/minder-mensen-stoppen-met-roken/ (2022). 22. CDCTobaccoFree. Smoking Cessation: Fast Facts. Centers for Disease Control and Prevention https://www.cdc.gov/tobacco/data_statistics/fact_sheets/cessation/ smoking-cessation-fast-facts/index.html (2022). 23. Edwards, S. A., Bondy, S. J., Callaghan, R. C. & Mann, R. E. Prevalence of unassisted quit attempts in population-based studies: a systematic review of the literature. Addict. Behav. 39, (2014). 24. Roken | Opleiding | Volksgezondheid en Zorg. https://www.vzinfo.nl/roken/opleiding. 25. Droomers, M., Schrijvers, C. T. M. & Mackenbach, J. P. Educational differences in the intention to stop smoking: Explanations based on the Theory of Planned Behaviour. Eur. J. Public Health 14, 194–198 (2004). 26. Droomers, M., Schrijvers, C. T. M. & Mackenbach, J. P. Educational differences in the intention to stop smoking: explanations based on the Theory of Planned Behaviour. Eur. J. Public Health 14, 194–198 (2004). 27. Kotz, D. & West, R. Explaining the social gradient in smoking cessation: it’s not in the trying, but in the succeeding. Tob. Control 18, 43–46 (2009). 28. Schaap, M. M. & Kunst, A. E. Monitoring of socio-economic inequalities in smoking: Learning from the experiences of recent scientific studies. Public Health 123, 103–109 (2009). 29. Hitchman, S. C., Fong, G. T., Zanna, M. P., Thrasher, J. F. & Laux, F. L. The relation between number of smoking friends, and quit intentions, attempts, and success: Findings from the International Tobacco Control (ITC) Four Country Survey. Psychol. Addict. Behav. 28, 1144–1152 (2014). 30. Meijer, E., Gebhardt, W. A., Van Laar, C., Kawous, R. & Beijk, S. C. A. M. Socio-economic status in relation to smoking: The role of (expected and desired) social support and quitter identity. Soc. Sci. Med. 1982 162, 41–49 (2016). 31. Shiffman, S. & Rathbun, S. L. Point process analyses of variations in smoking rate by setting, mood, gender, and dependence. Psychol. Addict. Behav. 25, 501–510 (2011). 32. Matta, A. da, Gonçalves, F. L. & Bizarro, L. Delay discounting: concepts and measures. Psychol. Neurosci. 5, 135–146 (2012). 33. Mani, A., Mullainathan, S., Shafir, E. & Zhao, J. Poverty Impedes Cognitive Function. Science 341, 976–980 (2013).

2 Chapter

SOCIOECONOMIC INEQUALITIES IN SMOKINGATTRIBUTABLE MORTALITY IN EUROPE: UNDERSTANDING TRENDS 2000-2020 Nienke W. Boderie, Alyson van Raalte, Jasper V. Been, Matthias Bopp, Patrick Deboosere, Terje Andreas Eikemo, Ramune Kalediene, Mall Leinsalu, Di Long, Pekka Martikainen, Olof Östergren, Maica Rodríguez-Sanz, Frank J. van Lenthe, Wilma J. Nusselder

Chapter 2 20 Abstract Introduction Smoking is one of the most important behavioural contributors to morbidity and mortality worldwide. However, the effects of smoking are not evenly distributed across society. We investigated trends in educational inequalities in smokingattributable mortality in Europe and changes in its contribution to educational inequalities in partial life expectancy over time. Method Partial life expectancy between age 50-80 was calculated between 2000 and 2020 in ten European countries (Austria, Belgium, Denmark, Estonia, Finland, Italy (Turin), Lithuania, Spain (Barcelona), Sweden and Switzerland). We estimated the smoking-attributable fraction (SAF) using the Preston-Glei-Wilmoth method. Changes in partial life expectancy by education and educational inequalities in partial life expectancy were decomposed into the age-specific contributions attributable and not-attributable to smoking, using the continuous change model. Results Among men, SAF decreased over time for all countries, but remained largest among those with lower education. For women SAF increased over time, but with a less profound educational gradient. For men, the contribution of smoking to educational differences in partial life expectancy ranged between 0.2 and 2.3 years and decreased between 2000 and 2020. Among women, it ranged between -0.1 and 0.9 years and increased or stabilized over time, except for Denmark. Conclusion Although smoking-attributable mortality decreased among men in all educational groups, smoking remains an important factor contributing to educational inequalities in life expectancy. For women the contribution of smoking to educational inequalities in life expectancy is increasing in most countries. The need for tobacco control measures to reduce these disparities remains high, especially for women.

Socioeconomic inequalities in smoking-attributable mortality in Europe; understanding trends 2000-2020 21 2 Introduction Ever since the Black-report in 1980, many strategies have been proposed to combat socioeconomic inequalities in health.1 Health inequalities however remained persistent throughout Europe, with socioeconomically more advantageous groups living longer and in better health.2 One of the most important behavioural factors contributing to socioeconomic inequalities in life expectancy is smoking.3 Patterns in smoking and smoking related mortality have often been described using the tobacco epidemic model. This model consists of four stages in which smoking rates among men initially increased and then began to decline following the increase in smoking-related mortality. Among women, a similar pattern is observed, but lagged in time and with a less extreme peak in smoking prevalence.4, 5 Prevalence rates not only differ by sex, but also between countries and socioeconomic groups,6, 7 and individual smoking behaviour is in part determined by social factors that change over time. For example, smoking is influenced by gender norms,8 and while the male-female smoking ratio has decreased, smoking prevalence has a negative educational gradient for both genders.9 Whereas smoking was initially primarily adopted among higher educated individuals, it has become more prevalent among lower educated individuals.10 In the past 20 years, smoking patterns in Europe have changed tremendously and are declining overall, 11, 12 which can in part be attributed to tobacco control measures.13 In Europe, the Tobacco Control Scale (TCS) quantifies the stringency of the implementation of tobacco control policies and has been shown at the ecological level to be associated with a lower prevalence of smoking and higher quit rates over the last decade.13 However, tobacco control measures are unlikely to decrease socioeconomic inequalities, except for price increases.14 Most tobacco control policies are aimed at smoking reduction in the general population and not focussed on decreasing inequalities. In an evaluation of 27 European countries, tobacco control policies were associated with smoking cessation among individuals with a higher socioeconomic position but not among those with a lower socioeconomic position.15 Given the developments in smoking behaviour and tobacco control policies, it is likely that educational inequalities in smoking-attributable mortality are changing as well, although with a delay. A wide range of studies investigated the role of smoking in educational differences in different European countries.16-18 Comparing several countries with different social, economic and political structures will allow countries to learn from each other, which can help to inform health policy and recommendations.

Chapter 2 22 Gregoraci et al. 19 described the changes in smoking-attributable mortality comparing 14 European countries between 1990 and 2004, revealing that inequalities in smoking remain one of the most important entry points for reducing inequalities in mortality. In this current paper, we extend the research on trends in educational inequalities in smoking-attributable mortality in Europe to the period 2000 to 2020. To gain better insight into how smoking contributed to inequalities in partial life expectancy between age 50 and 80, we investigate the contribution of smoking attributable mortality to changes in partial life expectancy for each educational group as well as to educational inequalities in partial life expectancy. Methods Data sources Total and cause-specific mortality by five-year age group, sex and educational level for the period 2000-2020 were collected and harmonized for countries with at least four data points within this period: Austria, Belgium, Denmark, Estonia, Finland, Italy (Turin), Lithuania, Spain (Barcelona), Sweden and Switzerland. Data were part of a post-census longitudinal mortality follow-up, and for most countries, data was available in five-year periods. Barcelona data is from cross-sectional datasets, where the register of mortality is linked to the municipal population registry each year. Educational level was assessed as the highest level of completed education according to the International Standard Classification of Education (ISCED),20 and categorized into three groups: lower (ISCED 0-2, up to lower secondary education), middle (ISCED 3-4, upper secondary education) and high (ISCED 5-8, tertiary education). Following the Preston-Glei-Wilmoth method, we analysed age 50 and over. Educational information was not available for all country-time points at older ages, therefore we excluded ages 80 and onwards. Analysis First, the time periods for which data were available differed between the countries, hence all-cause and lung cancer mortality rates were smoothed and interpolated using penalized splines.21 As a result, yearly mortality estimates for each country between 2000 and 2020 were obtained by age, sex and education, allowing for comparisons between countries. For Lithuania smoothed results did not fit the observed results at the beginning and end of the observation period, hence the results for 2000-2003 and 2020 were omitted. See Appendix I for a comparison between observed and smoothed results. Partial life expectancies (ages 50-79),

Socioeconomic inequalities in smoking-attributable mortality in Europe; understanding trends 2000-2020 23 2 also known as temporary life expectancies or restricted mean survival times,22, 23 were estimated from all-cause mortality life tables for each year. Second, to estimate the smoking-attributable fraction (SAF) we used the PrestonGlei-Wilmoth (PGW) indirect estimation method.24 In short, the PGW method uses a regression model, run on 21 high-income countries over 1950-2003, to estimate the statistical relationship between lung cancer mortality and mortality from other causes of death, accounting for age, calendar year, country, and interactions between mortality and calendar year and mortality and age. The coefficients from this model,24 when combined with the assumption that mortality from lung cancer among non-smokers resembles age and sex estimates from the Cancer Prevention Study II (CPS-II),18 are used to estimate the overall smoking-attributable fraction (SAF) of deaths. SAF was estimated for each year, sex, age and educational group in each country. Age-specific SAF was then combined with all-cause mortality to estimate its impact on life expectancy. Finally, SAF was standardized using the European Standard Population (2013) to allow for comparison between countries.25 Third, cause-deleted life tables that use death rates where the risk of dying from a specified cause is eliminated, in this case smoking-related causes, were calculated to estimate the partial life expectancy loss attributable to smoking between age 50 and 80.26, 27 Finally, for each country and sex, the changes in partial life expectancy (age 50 to 80) by education between 2000 and 2020, and the absolute educational differences in life expectancy between low and high educated in each year, were decomposed into the age-specific contributions in years attributable and nonattributable to smoking, using the continuous change model developed by Horiuchi et al.28 All analyses were done using R (version 4.2.2) and the DemoDecomp package for the decompositions.29 An extensive guide on how to use the Horiuchi model in R is available.30 Given the differences in mortality patterns between men and women, results are presented separately by sex. Results Over time large differences between countries and educational groups were observed in smoking attributable mortality, as seen in Figure 1. For women, SAF was higher for lower educated in all countries, and stable or increasing in many countries regardless of educational level. The exceptions were Denmark and Sweden, who have both experienced a decrease in SAF since mid-2010. Among men SAF declined in all countries and among all educational groups, and SAF was highest in the lower educational group for all countries (Appendix II).

Chapter 2 24 Figure 1: Age-standardized SAF over time by country, educational level and sex

Socioeconomic inequalities in smoking-attributable mortality in Europe; understanding trends 2000-2020 25 2 Life expectancy Between 2000 and 2020, life expectancy between age 50 and 80 ranged between 17 years (Lithuania, 2007, Lower educated men) and 24.8 years (Switzerland, 2020, Higher educated women). Figure 2 shows the difference in observed life expectancy and a hypothetical scenario in which no individuals had ever smoked. In this hypothetical scenario up to 1.2 years could have been gained is gained in life expectancy for women (Denmark, Lower educated, appendix II), and up to 2.4 years for men (Lithuania, Lower educated men, Appendix III). Lower educated men could have gained more years than higher educated men, while among women differences between educational groups were smaller. Figure 2 also shows changes in hypothetical gained life expectancy if smoking had not occurred; among men the hypothetical gain from never having smoked decreased, while for women it was stable or increasing with Denmark as the exception. Figure 2: Hypothetical gain in partial life expectancy between age 50 and 80 in the absence of smoking histories, from a cause deleted life-table Decomposition analysis Life expectancy increased between 2000 and 2020 among those with middle and higher educational attainment for both men and women. Among the lower educated, large differences were observed. In some cases, e.g. Austrian, Estonian and Swiss men, life expectancy increased more among lower educated than higher educated, while for Lithuanian and Finnish women, life expectancy decreased among the lower educated, by 0.6 and 0.2 years respectively (Appendix III). Life expectancy can change due to changes in smoking-attributable mortality and changes in mortality due to other causes. The contribution of both groups

Chapter 2 26 of causes to the total change in partial life expectancy by educational level is seen in Figure 3. These contributions differed strongly by education. Among men, reductions in smoking-attributable mortality contributed to increases in life expectancy. Among women, especially among lower educated women, increases in smoking-attributable mortality contributed negatively to the life expectancy change. Among higher educated women little to no impact of smoking-attributable mortality on changes in life expectancy was observed for Belgium, Estonia, Finland, Lithuania and Spain (Barcelona). Between 2000 and 2020 reductions in mortality due to other causes contributed to increases in life expectancy (among both sexes) except among the lower educated in Lithuania. In many countries, these increases were larger for higher educated individuals. Figure 3: Decomposition of the educational-group specific change in partial life expectancy between age 50 and 80 into smoking-attributable mortality and other-cause mortality between 2000 and 2020. For Lithuania, this is the change in life expectancy between 2004 and 2019.

Socioeconomic inequalities in smoking-attributable mortality in Europe; understanding trends 2000-2020 27 2 In Figure 4 the gap in partial life expectancy between higher and lower educated groups is decomposed into contributions from smoking-related mortality and other causes, for each year between 2000 and 2020. Lower educated women had up to 2.8 years lower life expectancy compared to higher educated women (Lithuania, 2019). This gap increased over time for all countries, except for Switzerland where the gap seemed to stabilize over time. Lithuania showed a more irregular pattern over time. The contribution of smoking-attributable mortality to the gap in life expectancy between lower and higher educated women differs. In about half of the countries, Austria, Estonia, Finland, Italy (Turin), Lithuania and Spain (Barcelona), the contribution of smoking-attributable mortality to educational inequalities is increasing. In Belgium, Sweden and Switzerland, it is stable since 2010. In Denmark, the contribution of smoking to educational inequalities in life expectancy decreases over time. The contribution of mortality due to other causes on the life expectancy gap was either stable or slowly increasing among women, except for Estonia where it decreased over time. Among men, up to 5 years difference in partial life expectancy between lower and higher educated was observed (Lithuania, 2019, Figure 4), with mixed trends in these inequalities across countries. Increases (Denmark, Lithuania, Spain and Italy) and decreases (Austria, Switzerland, and Estonia) were observed, and in Sweden, Belgium and Finland patterns in educational differences changed around 2010 into slightly increasing, decreasing or stable respectively. A more homogeneous pattern however was observed for the contribution of smoking to the difference in life expectancy between lower and higher educated. Over time, the contribution of smoking to educational inequalities in life expectancy decreased for all countries, except for Lithuania where after an initial increase the gap between lower and higher educated stabilized after 2015. The contribution of mortality from other causes to these inequalities increased in Belgium, Denmark, Finland, Italy (Turin), Lithuania, Spain and to some extent Sweden.

Chapter 2 28 Figure 4: Decomposition of the difference in partial life expectancy between age 50 and 80 between lower and higher educated people into smoking-attributable mortality and other cause mortality for each year between 2000 and 2020

Socioeconomic inequalities in smoking-attributable mortality in Europe; understanding trends 2000-2020 29 2 Discussion Using mortality data from ten European countries we investigated trends in partial life expectancy and hypothetical life expectancy in the absence of smoking histories, and decomposed the changing difference in life expectancy between educational groups into smoking-attributable mortality and other causes. Between 2000 and 2020 the contribution of smoking to the educational inequalities in partial life expectancy increased in most countries for women, while it decreased among men. At the same time, smoking histories continue to have a larger impact on the life expectancies of men compared to women. Previous studies have shown that higher educated men were the first to experience the gains to life expectancy by reductions in smoking-attributable mortality.19, 31 If these gains worked their way down to the lower educated groups, who stopped smoking later, it is expected that mortality inequalities would narrow. Instead, we found that inequalities were still increasing or persisting in many countries, despite reductions in smoking-related mortality leading to the largest life expectancy gains for lower educated men. This is because in most countries, life expectancy gains from non-smoking-related causes were even larger among the higher educated. Women, who generally picked up smoking later than men, are experiencing increases in smoking-attributable mortality at large, but in Sweden and Denmark SAF starts to decrease, which is in line with historic smoking patterns; women in Denmark and Sweden were among the first to start smoking, hence they were also the first to start quitting.32, 33 As with men, the peak of smoking attributable mortality was reached first among women in these countries, which is in line with the described variations in the smoking epidemic between gender and socioeconomic status.4 Within Europe large differences in smoking prevalence and implementation of tobacco control policies are observed,34 and country-level tobacco control scores are associated with lower smoking prevalence.35 Current tobacco control is not the cause of current changes in mortality due to the long period between smoking and onset of diseases. However, past tobacco control measures have had an influence on mortality due to smoking, but have not decreased educational inequalities due to smoking yet. Hence, the need for tobacco control policies that target lower educational groups remains high. We used an indirect method to estimate smoking-attributable mortality. Both a direct method, using smoking prevalence and relative mortality risk, and the

Chapter 2 30 indirect method based on lung cancer mortality as used in this study assume a simplified time pattern of smoking effects on diseases, and reviews comparing direct and indirect methods did not identify a best-practice method.36, 37 The PGW method assumes that the lag between smoking initiation and mortality is the same for lung cancer as it is for other smoking-attributable causes. For lung related diseases this indeed seems to be the case, while for cardiovascular diseases, excess mortality occurs earlier in life as it has a shorter lag period then lung cancer.5 The risk of lung cancer also declines faster upon smoking cessation compared to other lung related diseases.38 Oza et al. 39 showed minimal differences between methods incorporating specific lag times and those that did not. Furthermore, the PGW method assumes that the distribution of lung cancer related deaths among non-smokers is stable across countries and over time, which is not always the case.40 Despite these concerns, several studies have made comparisons of different indirect estimation techniques yielding similar results, showing the robustness of the PGW method.17, 24, 41, 42 This study comes with some limitations. First, not only do smoking rates change over time, educational stratification does as well. Therefore we risk lagged selection bias, i.e. there is a risk that a lower or higher educated person dying in 2000 represents another social structure than someone in 2020.43 This means that we should interpret increasing inequalities in partial life expectancy as in part actual increases, and in part a reflection of changing educational structures.44 On the other hand, the results in our study are similar to those using income quintiles as proxy of socioeconomic status, and quintiles have the advantage of being free of distributional change.17 Furthermore, at older ages information on mortality by educational level was unavailable for some countries, therefore we used partial life expectancy up to age 80. The impact of this choice however is expected to be limited, as previous studies have shown sharp declines in the percentage of smoking-related deaths after age 80.45, 46 In other words, smoking mainly contributes to premature mortality before age 80. Finally, the data coverage differed between countries, especially for Italy and Spain which did not have national-level datasets for long time series, and were instead represented by Turin and Barcelona respectively. However, previous literature has shown that smoking-related mortality in Spain nationally between 2016 and 2019 showed similar educational patterns as those shown for Barcelona in the present study.47-49

Socioeconomic inequalities in smoking-attributable mortality in Europe; understanding trends 2000-2020 31 2 Comparing a wide range of European countries for a prolonged period of time is one of the strengths of this paper. Furthermore, the analyses presented in this paper rely on a unique set of data on mortality rates by country, age group, sex and educational group and include data on cause of death. Education has the benefit of being stable throughout the life course and has a strong link to behavioural factors such as smoking.50 However, as socioeconomic position is a complex construct, future research could investigate composite or multiple indicators of socioeconomic status simultaneously. Conclusion Overall, we can conclude that smoking-attributable mortality remains an important factor contributing to educational differences in life expectancy in Europe for both sexes. Among women, the peak in smoking-attributable mortality has not been reached yet in most countries, due to the time lag between smoking exposure and outcome. The good news is that low educated men in most countries are experiencing the greatest reductions in smoking-related mortality– the bad news is that educational gaps in mortality decline are emerging in mortality due to other causes of death. The need for tobacco control measures that decrease educational gaps remains high, especially for women. Acknowledgements Data were collected as part of; the CHAIN project, which has received financial support from the Nordic Research Council (grant number 288638), NTNU and Erasmus MC; the LIFEPATH project, which has received financial support from the European Commission (Horizon 2020 grant number 633666), the project “Longer life, longer in good health, working longer? Implications of educational differences for the pension system”, which has received financial support from Network for Studies on Pensions, Aging and Retirement, and the DEMETRIQ project, which received support from the European Commission (grant numbers FP7-CP-FP and 278511). The mortality data for Switzerland were obtained from the Swiss National Cohort, which is based on mortality and census data provided by the Federal Statistical Office and supported by the Swiss National Science Foundation (grant nos. 3347CO-108806, 33CS30_134273 and 33CS30_148415). We acknowledge, Giuseppe Costa, Nicolas Zengirini, Henrik Brønnum-Hansen and Christoph Waldner, for providing and preparing the data.

Chapter 2 32 References 1. Bambra C, Smith KE, Garthwaite K, Joyce KE, Hunter DJ. A labour of Sisyphus? Public policy and health inequalities research from the Black and Acheson Reports to the Marmot Review. Journal of epidemiology & community health. 2011;65(5):399-406. 2. Mackenbach JP. Health inequalities: Europe in profile: Produced by COI for the Department of Health; 2006. 3. World health Organization. WHO report on the global tobacco epidemic 2021: addressing new and emerging products. 2021 978 92 4 003209 5. 4. Thun M, Peto R, Boreham J, Lopez AD. Stages of the cigarette epidemic on entering its second century. Tob Control. 2012;21(2):96-101. 5. Lopez AD, Collishaw NE, Piha T. A descriptive model of the cigarette epidemic in developed countries. Tobacco control. 1994;3(3):242. 6. Janssen F. Changing contribution of smoking to the sex differences in life expectancy in Europe, 1950-2014. Eur J Epidemiol. 2020;35(9):835-41. 7. Janssen F, El Gewily S, Bardoutsos A. Smoking epidemic in Europe in the 21st century. Tobacco control. 2021;30(5):523-9. 8. Bilal U, Beltrán P, Fernández E, Navas-Acien A, Bolumar F, Franco M. Gender equality and smoking: a theory-driven approach to smoking gender differences in Spain. Tobacco Control. 2016;25(3):295-300. 9. Legleye S, Khlat M, Beck F, Peretti-Watel P. Widening inequalities in smoking initiation and cessation patterns: A cohort and gender analysis in France. Drug and Alcohol Dependence. 2011;117(2):233-41. 10. Mackenbach JP. Health inequalities: Persistence and change in modern welfare states: Oxford University Press, USA; 2019. 11. Pesce G, Marcon A, Calciano L, Perret JL, Abramson MJ, Bono R, et al. Time and age trends in smoking cessation in Europe. PLOS ONE. 2019;14(2):e0211976. 12. Marcon A, Pesce G, Calciano L, Bellisario V, Dharmage SC, Garcia-Aymerich J, et al. Trends in smoking initiation in Europe over 40 years: A retrospective cohort study. PLOS ONE. 2018;13(8):e0201881. 13. Feliu A, Filippidis FT, Joossens L, Fong GT, Vardavas CI, Baena A, et al. Impact of tobacco control policies on smoking prevalence and quit ratios in 27 European Union countries from 2006 to 2014. Tobacco Control. 2019;28(1):101-9. 14. Smith CE, Hill SE, Amos A. Impact of population tobacco control interventions on socioeconomic inequalities in smoking: a systematic review and appraisal of future research directions. Tob Control. 2020;30(e2):e87-95. 15. Bosdriesz JR, Willemsen MC, Stronks K, Kunst AE. Tobacco control policy and socioeconomic inequalities in smoking in 27 European countries. Drug and Alcohol Dependence. 2016;165:79-86. 16. Martikainen P, Ho JY, Preston S, Elo IT. The changing contribution of smoking to educational differences in life expectancy: indirect estimates for Finnish men and women from 1971 to 2010. J Epidemiol Community Health. 2013;67(3):219-24. 17. Martikainen P, Mäkelä P, Peltonen R, Myrskylä M. Income differences in life expectancy: the changing contribution of harmful consumption of alcohol and smoking. Epidemiology. 2014;25(2):182-90. 18. Östergren O, Martikainen P, Tarkiainen L, Elstad JI, Brønnum-Hansen H. Contribution of smoking and alcohol consumption to income differences in life expectancy: evidence using Danish, Finnish, Norwegian and Swedish register data. J Epidemiol Community Health. 2019;73(4):334-9.

Socioeconomic inequalities in smoking-attributable mortality in Europe; understanding trends 2000-2020 33 2 19. Gregoraci G, Van Lenthe FJ, Artnik B, Bopp M, Deboosere P, Kovacs K, et al. Contribution of smoking to socioeconomic inequalities in mortality: a study of 14 European countries, 1990–2004. Tobacco control. 2017;26(3):260-8. 20. Statistics UIf. International standard classification of education: ISCED 2011. Comparative Social Research. 2012;30. 21. Camarda CG. MortalitySmooth: An R package for smoothing Poisson counts with P-splines. Journal of Statistical Software. 2012;50:1-24. 22. Arriaga EE. Measuring and explaining the change in life expectancies. Demography. 1984;21(1):83-96. 23. Royston P, Parmar MKB. Restricted mean survival time: an alternative to the hazard ratio for the design and analysis of randomized trials with a time-to-event outcome. BMC Medical Research Methodology. 2013;13(1):152. 24. Preston SH, Glei DA, Wilmoth JR. Contribution of smoking to international differences in life expectancy. International differences in mortality at older ages: Dimensions and sources. 2011:105-31. 25. Eurostat. Revision of the European Standard Population 2013 Edition. European Union, , 2013 ISSN 1977-0375. 26. Beltrán-Sánchez H, Preston SH, Canudas-Romo V. An integrated approach to causeof-death analysis: cause-deleted life tables and decompositions of life expectancy. Demographic research. 2008;19:1323. 27. Preston SH. The contours of demography: Estimates and projections. Demography. 1993;30(4):593-606. 28. Horiuchi S, Wilmoth JR, Pletcher SD. A decomposition method based on a model of continuous change. Demography. 2008;45(4):785-801. 29. Riffe T. DemoDecomp: decompose demographic functions. R Package version; 2018. 30. Raalte AAv, Nepomuceno MR. Decomposing gaps in healthy life expectancy. International handbook of health expectancies: Springer; 2020. p. 107-22. 31. Long D, Mackenbach J, Martikainen P, Lundberg O, Brønnum-Hansen H, Bopp M, et al. Smoking and inequalities in mortality in 11 European countries: a birth cohort analysis. Population Health Metrics. 2021;19(1):3. 32. Godtfredsen NS, Holst C, Prescott E, Vestbo J, Osler M. Smoking Reduction, Smoking Cessation, and Mortality: A 16-year Follow-up of 19,732 Men and Women from the Copenhagen Centre for Prospective Population Studies. American Journal of Epidemiology. 2002;156(11):994-1001. 33. Ahacic K, Kennison R, Thorslund M. Trends in smoking in Sweden from 1968 to 2002: Age, period, and cohort patterns. Preventive Medicine. 2008;46(6):558-64. 34. Willemsen MC, Mons U, Fernández E. Tobacco control in Europe: progress and key challenges. Tobacco Control. 2022;31(2):160-3. 35. Teshima A, Laverty AA, Filippidis FT. Burden of current and past smoking across 28 European countries in 2017: A cross-sectional analysis. Tobacco induced diseases. 2022;20. 36. Pérez-Ríos M, Montes A. Methodologies used to estimate tobacco-attributable mortality: a review. BMC Public Health. 2008;8(1):22. 37. Tachfouti N, Raherison C, Obtel M, Nejjari C. Mortality attributable to tobacco: review of different methods. Archives of public health. 2014;72:1-7. 38. Dresler CM, León ME, Straif K, Baan R, Secretan B. Reversal of risk upon quitting smoking. The Lancet. 2006;368(9533):348-9. 39. Oza S, Thun MJ, Henley SJ, Lopez AD, Ezzati M. How many deaths are attributable to smoking in the United States? Comparison of methods for estimating smoking-attributable mortality when smoking prevalence changes. Preventive medicine. 2011;52(6):428-33.

RkJQdWJsaXNoZXIy MTk4NDMw