BMJ

Confidential: For Review Only

Trends i n colorectal mortality in Europe: an analysis of the WHO mortality database

Journal: BMJ

Manuscript ID: BMJ.2015.026370

Article Type: Research

BMJ Journal: BMJ

Date Submitted by the Author: 10-Apr-2015

Complete List of Authors: Ait Ouakrim, Driss; Centre for and Biostatistics, Melbourne School of Population and Global Health, University of Melbourne, Pizot, Cecile; International Prevention Research Institute (iPRI), Boniol, Magali; International Prevention Research Institute (iPRI), Malvezzi, Matteo; Universitá degli Studi di Milano, Department of Clinical Sciences and Community Health Boniol, Mathieu; International Prevention Research Institute (iPRI), ; University of Strathclyde Institute for Global Public Health at iPRI, Negri, Eva; IRCCS - Istituto di Ricerche Farmacologiche ‘Mario Negri’, Department of Epidemiology Bota, Maria; International Prevention Research Institute (iPRI), ; University of Strathclyde Institute for Global Public Health at iPRI, Jenkins, Mark; Centre for Epidemiology and Biostatistics, Melbourne School of Population and Global Health, University of Melbourne, Bleiberg, Harry; Jules Bordet Institute, Autier, Philippe; University of Strathclyde Institute of Global Public Health at iPRI, ; International Prevention Research Institute (iPRI),

Keywords: colorectal cancer, mortality, screening, epidemiology

https://mc.manuscriptcentral.com/bmj Page 1 of 33 BMJ

1 2 3 4 N = 3361 5 6 7 8 TrendsConfidential: in colorectal cancer mortality For in Review Europe: an analysis Only of the 9 10 WHO mortality database 11 12 13 14 1 2 15 Driss Ait Ouakrim Research Fellow , Cécile Pizot Research Officer , Magali Boniol Assistant 16 Statistician 2, Matteo Malvezzi Post Doctoral Fellow 3, Mathieu Boniol Vice-President 17 Biostatistics 2, 4, Eva Negri Head of laboratory 5, Maria Bota Research Officer 2, 4 , Mark A Jenkins 18 1 6 19 Professor and Director , Harry Bleiberg Professor Emeritus , Philippe Autier Vice-President 2, 4 20 Population Research . 21 22 23 1 Centre for Epidemiology and Biostatistics, Melbourne School of Population and Global Health, 24 25 University of Melbourne, 3010, Australia 26 2 International Prevention Research Institute (iPRI), 69006 , 27 28 3 Department of Clinical Sciences and Community Health, Universitá degli Studi di Milano, 29 30 20156 , 31 4 University of Strathclyde Institute for Global Public Health at iPRI, 69006 Lyon, France 32 33 5 Department of Epidemiology, IRCCS - Istituto di Ricerche Farmacologiche ‘Mario Negri’, 20156 34 Milan, Italy. 35 36 6 Jules Bordet Institute, 1000 , 37 38 39 40 Correspondence: Dr Philippe Autier, International Prevention Research Institute (iPRI), Cours 41 Lafayette 95, 69006 Lyon Cedex 08, France; E-mail: [email protected]; website : www.i- 42 pri.org. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 1 https://mc.manuscriptcentral.com/bmj BMJ Page 2 of 33

1 2 3 Abstract (n=297) 4 5 Objectives: To examine changes in colorectal cancer (CRC) mortality in 34 European countries 6 7 between 1989 and 2011 8 Confidential: For Review Only 9 Design: Retrospective trend analysis. 10 11 12 Data source: World Health Organisation mortality database. 13 14 Population: Female and male colorectal cancer deaths between 1989 and 2011. 15 16 17 Main outcomes measures: Time trends in colorectal cancer mortality rates using join point 18 19 regression analysis. Rates were age adjusted using the European Standard Population. 20 21 Results: From 1989 to 2011, declines in CRC mortality of more than 25% in men and 30% in 22 23 women occurred in Austria, Switzerland, Germany, the United Kingdom, Belgium, the Czech 24 25 Republic, Luxemburg and Ireland. In contrast, CRC mortality rates declined by less than 15% in 26 27 the Netherlands and Sweden for both sexes. Increases of 28 to 45% were observed in men in 28 Spain, Portugal and Greece that were not observed in women. Smaller or no declines occurred 29 30 in most Central European countries. Substantial mortality increases of 20 to 102% occurred in 31 32 Croatia, Macedonia, and Romania for both sexes and in most Eastern European countries for 33 34 men. In countries with declining CRC mortality, declines were generally more important for 35 36 women and people aged less than 65 years. Overall, from 1989 to 2011, CRC mortality in the 27 37 38 European member states has decreased by 13% for men and by 27% for women. As a 39 40 comparison, In the United States, CRC mortality declined by 40% for men and by 39% for 41 42 women. 43 44 Conclusions: There is considerable disparity in the level of CRC mortality between European 45 46 countries, as well as between men and women and between age categories. Countries with the 47 48 greatest declines in CRC mortality are characterised by greater accessibility to screening 49 50 services, especially endoscopic screening, and specialised care. In countries where rates are 51 52 sharply increasing, lifestyle factors such as poor diet and alcohol drinking also play a major role. 53 54 55 56 57 58 59 60 2 https://mc.manuscriptcentral.com/bmj Page 3 of 33 BMJ

1 2 3 Introduction 4 5 6 Colorectal cancer (CRC) is the second most commonly diagnosed cancer in the world and has 7 8 poorConfidential: prognosis when metastasised to lymph For nodes orReview distant organs. In 2012, Only it was estimated 9 1 10 that 241.6 thousand European men were diagnosed with CRC and 113.2 thousand died from it. 11 For European women, 205.2 thousand CRC cases and 101.5 thousand CRC deaths were 12 13 recorded that year.1 Estimations also indicated substantial differences in CRC mortality across 14 15 European countries.2 16 17 18 Over the last two decades in Europe, early detection of CRC has increased through screening 19 20 and easier access to endoscopic removal of adenomatous polyps (the commonest CRC 21 precursor lesion).3 At the same time, new CRC treatments have been developed and their 22 23 availability improved.4 Taken together, these factors would be expected to lead to a decrease in 24 25 CRC mortality over time; however, effectiveness and access to these mortality-reducing factors 26 27 may not be equal across age groups, sex, and nationality. Further, the prevalence of some of 28 29 the main risk factors for CRC has been increasing (e.g. adiposity and sedentary lifestyle) and is 30 5 31 unequally distributed between countries. 32 33 In order to better understand the possible reasons underlying the heterogeneity in CRC 34 35 mortality across European countries, we analysed age and sex-specific trends in CRC mortality 36 37 rates from 1970 to 2011 in 34 European countries. We were most interested in recent CRC 38 39 mortality trends (1989-2011) because the most profound changes in screening and treatment 40 41 efficiency took place after 1988. 42 43 Methods 44 45 46 Mortality data 47 48 Number of CRC deaths was obtained from the World Health Organisation (WHO) mortality 49 6 50 database for 34 European countries for the period 1970–2011. For the sake of comparison, we 51 52 also extracted data for the United States of America (USA). 53 54 Between 1970 and 2011, data on cause of death was classified using three versions of the 55 56 international Classification of Diseases (ICD). We defined CRC mortality as any death with an 57 58 ICD 8 code of A048-A049, an ICD 9 code of B093-B094, or an ICD 10 code of C18-C21. We also 59 60 3 https://mc.manuscriptcentral.com/bmj BMJ Page 4 of 33

1 2 3 included any death with an ICD code for of the anus and/or anal canal because it was 4 5 unclear from the database how some countries used the different ICD coding schemes to 6 7 classify cancers into the various CRC sub-sites—“colon”, “recto-sigmoid junction”, “rectum”, 8 Confidential: For Review Only 9 “anus” and “anal canal”. 10 11 For a majority of Western European countries, data were available for most of or the period of 12 13 interest. Cyprus was not included because data were only available for four years (1999, 2000, 14 15 2004, 2006). For Switzerland, data were available until 2010. Furthermore, a change in coding 16 17 practice in 1994 resulted in an over-reporting of cancer mortality before 1994. We therefore 18 19 applied a correction factor of 0·94 to all mortality rates before 1995, as recommended by Lutz 20 7 21 et al. For most central and eastern European countries, data were available since the early-mid 22 23 1980s until 2009 – 10, with the exception of Bulgaria for which data were available for the 24 25 entire period. For Slovakia and the Republic of Macedonia data were available from 1991 and 26 27 1992 respectively, until 2010. Years for which data were missing are presented in 28 Supplementary Table 1. 29 30 31 Statistical analyses 32 33 We used the direct method and age-specific population estimates from the WHO mortality 34 35 database 6 to compute age-adjusted CRC mortality rates according to the age distribution of the 36 37 standard European population.8 Joinpoint regression analysis over the whole period was 38 39 performed to identify years where statistically significant changes in mortality trends occurred. 40 41 Annual percent changes (APCs) (relative change) were then computed for each country, by 42 43 fitting a regression line to the natural logarithm of the rates on the period 1989-2011 and over 44 45 the last five years of available data (2007-2011). Because of missing data, regression was fitted 46 over the period 1992 – 2010 for Slovakia and 1991 – 2010 for Macedonia. Based on the 47 48 underlying joinpoint model, overall percent changes were then derived for the period 1989- 49 50 2011. We conducted these analyses separately for men and women of all ages, and for the age 51 52 categories <65, 65 to 79 and ≥80 years. The joinpoint regression analysis was conducted using 53 54 the publicly available “joinpoint software” from the Surveillance Research Program of the U.S. 55 9 56 National Cancer Institute. Modelling parameters are presented in Supplementary Table 2. 57 58 59 60 4 https://mc.manuscriptcentral.com/bmj Page 5 of 33 BMJ

1 2 3 Results 4 5 6 Figure 1 provides an overview of the evolution of CRC mortality by gender and for all ages 7 8 betweenConfidential: 1970 and 2011 (or the last year ofFor availabl e Reviewdata). CRC mortality trends Only for the age 9 10 categories <65, 65 to 79 and ≥80 years are available online as supplementary material to this 11 article (Supplementary Figures 1 - 3). Overall, Important differences can be observed between 12 13 men and women and between the included countries. Over the period 1989 – 2011, in all 34 14 15 countries, CRC mortality has increased by a median of 6% for men and decreased by a median 16 17 of 14.7% for women (Tables 1 and 2). In the 27 European Union (EU) member states, CRC 18 19 mortality has decreased by 13% in men and by 27% in women. The largest drops for both sexes 20 21 occurred in Western and Northern Europe. Important drops also took place in Southern and 22 23 some Eastern-European countries for women. In most countries with decreasing mortality 24 25 rates, the decline initiated between 1970 and up to the mid-1990s. 26 27 For men, CRC mortality rates in 1989-91 across European countries ranged from 10.7 per 28 29 100,000 in Greece to 53 per 100,000 in the Czech Republic (Table 1). Over the period 1989- 30 31 2011, CRC mortality trends ranged from a 44% decline in Austria to a 102% increase in 32 33 Romania. In nine countries, declines started before 1980 (Figure 1). Declines of 10% or more 34 35 were observed in 15 countries, while increases of 10% or more were also observed in 15 36 countries. The average level of CRC mortality for European men has therefore remained 37 38 constant at 27.3 CRC deaths per 100,000 between 1989-91 and 2009-11. In the later period 39 40 however, the difference in CRC mortality between the two countries with the highest and 41 42 lowest CRC mortality rates was reduced from 5 to 3 folds. 43 44 45 For women, CRC mortality rates in 1989-91 ranged from 8.9 per 100,000 in Greece to 28.2 per 46 47 100,000 in Hungary (Table 1). Over the period 1989-2011, CRC mortality trends ranged from a 48 49 50.4% decline in Austria to a 56.1% increase in Macedonia. Declines started before 1980 in 15 50 51 countries (Figure 1). Twenty-five countries experienced a decline in CRC mortality. This decline 52 53 occurred in a large number of Southern, Central and Eastern European countries and in 54 contrast with the male population in these countries, which was still experiencing increases in 55 56 mortality rates. Declines of more than 10% were observed in 20 countries, while increases were 57 58 observed in nine countries. The average CRC mortality rate for European women decreased 59 60 5 https://mc.manuscriptcentral.com/bmj BMJ Page 6 of 33

1 2 3 from 18.1 per 100,000 in 1989-91 to 15.3 per 100,000 in 2009-11. Although mortality rates in 4 5 1989-91 were on average about one half lower for women compared to men, subsequent 6 7 decreases in mortality in Europe were more pronounced in women. In most countries, declines 8 Confidential: For Review Only 9 in CRC mortality usually started earlier for women. In countries experiencing mortality 10 11 increases, the rates for men were substantially higher than for women. 12 13 Overall, for both men and women, a comparison of the annual percent changes (APC) over the 14 15 last 5 years of observations (2007 to 2011) to the APCs over the entire 22-year period 16 17 suggested that the declines in CRC mortality may be accelerating. 18 19 20 Important differences were also noticeable between countries, including amongst those with 21 comparable socio-economic characteristics. For example, in the Netherlands, Germany and 22 23 Austria, CRC mortality rates in 1989-91 were similar. Between 1989 and 2011, CRC mortality 24 25 rates in the Netherlands declined by 11.6% and 16.7% for men and women respectively. In 26 27 comparison, Germany and Austria saw their CRC mortality rates decline by more than 30% for 28 29 both sexes. 30 31 In Central, Eastern and Southern European countries, CRC mortality rose until later years. In 32 33 1989-91, the highest CRC mortality rates were in Hungary, Slovakia and the Czech Republic. In 34 35 other countries, mortality trends stabilized or showed first signs of decline after 2000. 36 37 Sustained mortality increases in both sexes were observed in Latvia, Macedonia, Romania, and 38 39 the Russian Federation (Figure 1). 40 41 A weak inverse relationship between CRC mortality rates in 1989-91 and changes in mortality in 42 43 subsequent years is noticeable (Figure 2). The largest declines were observed in Western 44 45 European countries. Increasing mortality rates were observed mostly in countries situated in 46 47 Southern, Central and Eastern Europe, at the exception of the Czech Republic, which had one of 48 49 the strongest declines in mortality for both sexes. 50 51 For all of Europe combined, the declines or increases in CRC mortality were less marked in men 52 53 and women under age 65 years compared to those in older age groups. However, there were 54 55 major differences between regions. In most Western and Northern Europe, declines were 56 57 generally more pronounced in people aged less than 65 years (Tables 1 and 2). Although some 58 59 heterogeneity existed among these countries as decreases tended to be more important for 60 6 https://mc.manuscriptcentral.com/bmj Page 7 of 33 BMJ

1 2 3 older persons in Austria, Germany, Switzerland, Ireland and the United Kingdom. Age 4 5 differences in trends were most discernible in Central and Eastern European countries, where 6 7 people aged 80 years or older experienced dramatic increases in CRC mortality compared with 8 Confidential: For Review Only 9 younger age groups. 10 11 In the USA, from 1989 to 2011, mortality declined by 40% for men and by 39% for women 12 13 (Tables 1 and 2). Mortality declines were greatest for people aged 80 years or older, and lowest 14 15 for those younger than 65 years. 16 17 18 Discussion 19 20 Our analysis documents the considerable diversity of CRC mortality trends across European 21 22 countries over the last four decades. If rates across Europe show less variability in recent 23 24 periods than in 1989, the difference between the highest and lowest rates is still threefold. On 25 26 average, there were sustained declines in CRC mortality between 1989 and 2011 for most 27 28 northern and western European countries as well as some central European countries. Our 29 results suggest that these declines are likely to continue and that they have been accelerating 30 31 in the recent past. Our findings also confirm the stabilised mortality trends noticed in most 32 33 central and eastern European countries since the early 2000s, particularly for women and 34 35 people younger than age 65 years. Some of these countries, however, continue to have very 36 37 high rates, particularly for the male population. Changes in CRC mortality over time may be due 38 39 to the contribution of a variety of factors, including demographic characteristics, lifestyle, 40 41 disease awareness, screening and access to effective treatment. Of note, declines often started 42 43 one or two decades before the introduction of any screening programme or the diffusion of 44 45 effective CRC treatments seen in the 1990s. 46 47 Age and sex specific declines in CRC mortality 48 49 The considerable differences in age-specific trends in CRC mortality observed in most countries 50 51 could be due to older age being associated with more advanced stage at diagnosis and less 52 53 intensive treatment. 10 11 Moreover, younger CRC patients have a better survival irrespective of 54 55 patient’s characteristics, stage at diagnosis and treatment received. 12 Changes in lifestyle in 56 57 successive generations might also have played a role, with a lower prevalence of risk factors 58 59 associated with CRC death in younger generations. This birth cohort phenomenon is perhaps 60 7 https://mc.manuscriptcentral.com/bmj BMJ Page 8 of 33

1 2 3 more relevant to Central European countries, where dramatic changes in lifestyle have taken 4 5 place after 1989, that should lead to decreases in the CRC burden that are first noticeable in 6 7 subjects less than 65 years of age. 13 14 8 Confidential: For Review Only 9 10 CRC mortality declines started earlier and were usually more important for women than for 11 men. Conversely, in countries where CRC mortality was on the rise, the increases were usually 12 13 more marked for men. Several behavioural and physiological factors have been identified as 14 15 potential explanations for this difference in the burden of CRC between the sexes. For example, 16 17 use of oestrogenic hormone by women has been associated with a reduced risk of CRC.15 It has 18 19 also been shown that compared to women, men are less likely to participate to CRC 20 16 21 screening. Men have on average less contact with the health care system, are less well 22 23 informed about health issues, give less attention to symptoms and are less inclined to seek 24 17 18 25 medical advice. Men have less varied dietary habits and higher levels of smoking and 26 27 alcohol consumption than women which might also be associated with their higher CRC 28 mortality rates.19 29 30 31 In addition, men have a 1.5 to 2-fold greater prevalence of large bowel adenoma than women, 32 20-22 33 irrespective of age. This might be the reason why in randomized trials of flexible 34 35 sigmoidoscopy screening, risk reductions tended to be more pronounced for men than for 36 women.23-25 In contrast, in randomized trials on guaiac faecal occult blood test (FOBT) 37 38 screening, reductions in CRC mortality were similar for both sexes, presumably because 39 40 invasion of surrounding tissues by CRC may cause bleeding while bleeding is a rare event during 41 42 adenoma development.26-28 Thus for men, there might be some rational to support endoscopic 43 44 screening over FOBT. 45 46 Lifestyle 47 48 49 Part of the disparities in CRC mortality trends may be due to contrasting prevalence of lifestyle 50 51 risk factors across Europe. For instance, the acute increase in alcohol consumption observed 52 53 since 1989 in countries of the former Soviet Union is likely to have contributed to the higher 54 CRC incidence and mortality levels in these countries. 29-31 The improvement in CRC mortality 55 56 rates in the younger age groups of some central and eastern European countries may, at least 57 58 59 60 8 https://mc.manuscriptcentral.com/bmj Page 9 of 33 BMJ

1 2 3 in part, be related to protective lifestyle factors, such as a more diversified diet (e.g., containing 4 5 lower amounts of fat) and the decline in alcohol consumption. 13 6 7 8 TheConfidential: so called Mediterranean diet has been For associate dReview with a moderately protective Only effect 9 32 10 against CRC. However, lower adherence to this dietary pattern over the last thirty years has 11 been reported in Portugal, Spain, Italy and Greece 33-36 , which might be associated with the 12 13 rising or modestly decreasing CRC mortality rates observed in these countries. 14 15 16 Several studies have consistently reported associations between obesity, physical inactivity and 17 37-40 18 diabetes mellitus, and CRC occurrence and poor prognosis. The prevalence of these risk 19 5 20 factors varies substantially between European countries and tends to increase in many 21 communities. These increasing prevalences may also explain the higher CRC mortality rates 22 23 observed in some countries and particularly in those where populations have limited access to 24 25 screening and efficient therapies. 26 27 28 Awareness 29 30 Another potential reason behind the large and early declines in CRC mortality observed in some 31 32 countries might the level of awareness of CRC among physicians and the public. Awareness 33 34 encompasses greater attention to bowel symptoms, to a family history of colorectal cancer or a 35 36 personal history of bowel inflammatory disease. Increasing disease awareness has encouraged 37 38 the spread of opportunistic screening activities and spurred patients to consult more quickly for 39 bowel symptoms and doctors to speed up referrals for bowel examination, e.g., diagnostic 40 41 colonoscopy and barium enema.41 42 43 44 Screening 45 46 In the top twelve countries where CRC mortality has most improved, as well as in the USA, 47 48 increasing screening activities based on FOBT or on endoscopic methods have taken place since 49 50 the 1990s through organised programmes, opportunistic screening or a combination of both 51 3 42 43 52 approaches. 53 54 In several countries, the publication of randomised trials demonstrating the ability of FOBT and 55 56 sigmoidoscopy screening to reduce the risk of CRC death gave the signal for the introduction of 57 58 national or regional CRC screening programmes based on FOBT (e.g., the UK, France) or rather 59 60 9 https://mc.manuscriptcentral.com/bmj BMJ Page 10 of 33

1 2 3 based on colonoscopy (e.g., Austria, Czech Republic, Germany, Poland). 3 42 43 Since then the 4 5 landscape of CRC screening in Europe has been rapidly evolving. However, precise data on CRC 6 7 screening activities in countries are scarce and available statistics rarely distinguishes between 8 Confidential: For Review Only 9 colonoscopies done for screening, diagnosis or surveillance purposes. In other countries, like in 10 11 the United Kingdom, statistics are based on hospital records and do not comprehend most of 12 44 45 13 the endoscopic activities done outside hospital settings. 14 15 The few available estimates, particularly those for countries with comparable high quality 16 17 health care systems but different levels of screening activity, are consistent with the idea that 18 19 screening participation is a key factor in the decrease of CRC mortality. For example, compared 20 22 46 21 21 to France , Austria and Germany , availability of colonoscopy services is more limited in 22 47 23 the Netherlands. Until 2013, only pilot projects based on FOBT screening have been run in the 24 3 25 Netherlands and in Sweden . A survey among people aged 50 years and older using a standard 26 27 questionnaire found that in 2004, the proportion of men and women who had had at least one 28 FOBT over the last ten years were 61% in Austria, 53% in Germany, 24% in France, 15% in 29 30 Sweden and 4% in the Netherlands 48 . For 10-year endoscopic examination of the large bowel, 31 32 proportions were 24% in Austria and Germany, 25% in France, 12% in Sweden and 10% in the 33 34 Netherlands. These observations agrees with recent data from the USA, showing that high 35 36 participation rates to endoscopic screening have had a large impact on mortality and incidence 37 49 38 from CRC . 39 40 In countries where CRC mortality has been declining, sex and age differences in the level of 41 42 screening uptake might explain the differences in sex and age-specific CRC mortality. For 43 44 example, in Austria, the largest decline in mortality was observed for those aged less than 65 45 46 years in our analysis, the age category in which colonoscopic screening has been shown to be 47 50 48 most prevalent. In contrast, the largest decline in CRC mortality in the USA was observed for 49 50 men aged 80 years and older. This is likely due to the fact that endoscopic screening in USA is 51 52 usually undertaken at an older age (65 years or older, as access to Medicare is allowed) with a 53 higher participation for men.16 51 54 55 56 57 58 59 60 10 https://mc.manuscriptcentral.com/bmj Page 11 of 33 BMJ

1 2 3 Treatment 4 5 6 New treatments, constant advances in surgical technics and therapeutic protocols have also 7 52 8 certainlyConfidential: played a key role in reducing of CRCFor mortality. Review There has been an Only increasing 9 10 recognition that specialized care, received in medical centres treating large numbers of patients 11 with the same condition, is associated with higher survival rates.53 54 Over the last two decades, 12 13 the management of CRC patients has substantially improved, with the introduction of 14 15 multidisciplinary teams, more accurate staging, novel surgical approaches, and more effective 16 17 chemotherapy and radiotherapy regimens.4 18 19 20 Potential limitations 21 22 The results presented in this study should be interpreted with caution as the reliability of death 23 55 24 certification may vary between countries. This means that the quality of the WHO mortality 25 26 data might vary between countries, which in turn might have impacted the accuracy of our 27 28 mortality rates. However, colorectal cancer is a major neoplasm, and its diagnosis and 29 certification are consistent in most European countries.56 Therefore, it is unlikely that errors in 30 31 diagnostic or certification may have affected our findings. Missing mortality data may also have 32 33 influenced trends, but after 1988, years with missing data were infrequent. Another potential 34 35 limitation might be that for countries having experienced strongly divergent age-specific CRC 36 37 mortality trends, we were unable to determine to what extent those variation were due to 38 39 changes in the ICD over the study period. Given these changes in coding practices, we opted for 40 41 an inclusive approach in the definition of the anatomic location of CRC, by taking into account 42 43 deaths due cancers of the anus and the anal canal in our analysis. It should be noted that these 44 45 cancers represent only about 1% of all cancers of the large bowel. Increases in life expectancy 46 may have led to a higher mean age of the population at risk in the last open ended age category 47 48 (85 years or more) leading to suboptimal age adjustment in this age group. 49 50 51 Conclusions 52 53 Overall, since 1970, CRC mortality in men has been declining in half of European countries, and 54 55 CRC mortality in women has been declining in two-third of European countries. The underlying 56 57 driver for these declines is likely to be a combination of better public awareness of the disease, 58 59 changes in lifestyle, greater participation to screening, improved treatment and management 60 11 https://mc.manuscriptcentral.com/bmj BMJ Page 12 of 33

1 2 3 protocols. In Southern, Central and Eastern Europe, however, where mortality rates have either 4 5 risen, or modestly decreased, there are clear opportunities for improvement through primary 6 7 and secondary prevention and better access to specialized care. Effective strategies already in 8 Confidential: For Review Only 9 place in several European countries, which have achieved large declines in CRC mortality since 10 11 1989, represent valuable models for the design and implementation of such public health 12 13 policies. 14 15 Declaration of interest 16 17 18 All authors have completed the ICMJE uniform disclosure form at 19 20 www.icmje.org/coi_disclosure.pdf and declare: no support from any organisation for the 21 submitted work; no financial relationships with any organisations that might have an interest in 22 23 the submitted work in the previous three years; no other relationships or activities that could 24 25 appear to have influenced the submitted work. 26 27 28 Funding 29 30 The study was mainly funded by the International Prevention Research Institute (iPRI), Lyon 31 32 (France). DA was supported by the Australian National Health and Medical Research Council 33 34 (NHMRC) under the Centre for Research Excellence scheme. The iPRI and the NHMRC had no 35 36 involvement in the design, conduct and reporting of the study. 37 38 Copyright 39 40 “The Corresponding Author, Philippe Autier, has the right to grant on behalf of all authors and 41 42 does grant on behalf of all authors, a worldwide licence to the Publishers and its licensees in 43 44 perpetuity, in all forms, formats and media (whether known now or created in the future), to i) 45 46 publish, reproduce, distribute, display and store the Contribution, ii) translate the Contribution 47 48 into other languages, create adaptations, reprints, include within collections and create 49 50 summaries, extracts and/or, abstracts of the Contribution, iii) create any other derivative 51 52 work(s) based on the Contribution, iv) to exploit all subsidiary rights in the Contribution, v) the 53 54 inclusion of electronic links from the Contribution to third party material where-ever it may be 55 located; and, vi) licence any third party to do any or all of the above.” 56 57 58 59 60 12 https://mc.manuscriptcentral.com/bmj Page 13 of 33 BMJ

1 2 3 “I, Philippe Autier, The Corresponding Author of this article contained within the original 4 5 manuscript which includes any diagrams & photographs within and any related or stand alone 6 7 film submitted (the Contribution”) has the right to grant on behalf of all authors and does grant 8 Confidential: For Review Only 9 on behalf of all authors, a licence to the BMJ Publishing Group Ltd and its licencees, to permit 10 11 this Contribution (if accepted) to be published in the BMJ and any other BMJ Group products 12 13 and to exploit all subsidiary rights, as set out in our licence set out at: 14 15 http://www.bmj.com/about-bmj/resources-authors/forms-policies-and-checklists/copyright- 16 17 open-access-and-permission-reuse.” 18 19 X I am one author signing on behalf of all co-owners of the Contribution. 20 21 Contributors statement 22 23 24 Driss Ait Ouakrim (DA), Cécile Pizot (CP), Magali Boniol (MGB), Matteo Malvezzi (MM), Mathieu 25 26 Boniol (MTB), Eva Negri (EN), Maria Bota (MB), Mark A Jenkins (MAJ), Harry Bleiberg (HB), 27 28 Philippe Autier (PA). 29 30 Concept: PA, EN, HB 31 32 Data collection: CP, DA, MB, MM, 33 34 Data analysis : DA, CP, MTB, MB 35 36 Drawing graphics : DA, CP, MGB, MM 37 38 Article writing and editing: DA, MAJ, PA 39 Discussion: all co-authors 40 41 42 All authors, external and internal, had full access to all of the data (including statistical reports 43 44 and tables) in the study and can take responsibility for the integrity of the data and the 45 46 accuracy of the data analysis. 47 48 Transparency declaration 49 50 The lead author, Philippe Autier, affirms that this manuscript is an honest, accurate, and 51 52 transparent account of the study being reported; that no important aspects of the study have 53 54 been omitted; and that any discrepancies from the study as planned (and, if relevant, 55 56 registered) have been explained. 57 58 Ethical approval not required 59 60 13 https://mc.manuscriptcentral.com/bmj BMJ Page 14 of 33

1 2 3 4 What is already known on this subject? 5 6 • Colorectal cancer (CRC) is a major public health issue in most western countries. 7 8 Confidential:• Since 1989, changes in risk factor prevalenceFor (e.g.Review, obesity, alcohol drinking),Only 9 participation in screening and access to specialised care and effective therapies have 10 taken place in many European countries. 11 12 13 14 What this study adds 15 16 • Colorectal cancer mortality is declining in an increasing number of European countries 17 despite persistent differences between men and women and between specific regions 18 in Europe. 19 20 • The largest declines over the period 1989 – 2011 were observed in countries where 21 participation in screening has increased and where access to specialized care has 22 been optimized. 23 24 • In most Central European countries, CRC mortality has been stable or slightly 25 decreasing since the early 2000s. CRC mortality is still increasing in most Eastern 26 27 European countries. 28 29 • Strategies already in place in several European countries could be used as models to 30 design and implement effective health policies to prevent death from colorectal 31 cancer. 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 14 https://mc.manuscriptcentral.com/bmj Page 15 of 33 BMJ

1 2 3 Table 1 – Changes in colorectal cancer mortality for European men between 1989 and 2011 4 5 Mortality change for all ages (%) Mortality change 1989-2011 by age group (%) 6 Annual * Year Mean mortality For 1989-2011 change Annual Overall 7 start of last 5 8 Confidential: Fordecline Review Only 9 Countries 1989-91 2009-11 Annual Overall years <64 65-79 80+ <64 65-79 80+ 10 Austria 36.0 21.6 -2.6 -44.1 1993 -3.6 -3.2 -2.6 -2.0 -51.1 -44.4 -35.6 11 Switzerland 27.1 18.7 -2.2 -38.4 1970 0.5 -1.6 -2.2 -2.6 -30.2 -38.6 -44.2 12 Germany 33.4 22.6 -2.1 -36.7 1973 -2.0 -2.2 -1.9 -2.1 -39.2 -34.8 -37.5 13 United Kingdom 31.4 21.4 -2.0 -35.8 1990 -1.6 -2.6 -2.0 -1.4 -43.5 -35.4 -26.4 14 Luxembourg 31.8 21.6 -1.9 -33.8 1982 -5.2 -2.1 -2.1 -1.1 -37.4 -37.4 -21.1 15 Belgium 30.2 21.7 -1.7 -31.9 1970 -1.8 -1.9 -1.8 -1.5 -34.6 -32.5 -28.7 16 Czech Republic 53.0 37.5 -1.6 -30.1 1991 -3.7 -2.3 -1.6 -0.7 -39.5 -30.0 -14.8 17 18 Ireland 35.3 25.7 -1.5 -28.4 1970 -4.0 -2.9 -1.1 -0.7 -47.2 -21.6 -15.3 19 Denmark 35.5 27.6 -1.4 -26.9 1978 -1.0 -2.1 -1.4 -0.7 -37.8 -26.6 -14.7 20 France 27.5 21.6 -1.2 -23.9 1976 -1.3 -1.1 -1.5 -0.9 -21.4 -28.2 -18.7 21 Norway 30.8 25.4 -1.1 -22.0 1994 -0.6 -2.1 -0.9 -0.5 -36.9 -18.4 -11.3 22 Finland 19.4 16.1 -1.0 -20.2 1992 -2.0 -1.0 -0.8 -1.5 -20.0 -15.3 -27.9 23 Sweden 23.2 19.7 -0.7 -14.9 1976 -0.5 -1.2 -0.5 -0.7 -23.0 -10.3 -14.5 24 EU 27 27.8 24.9 -0.6 -13.0 1995 -0.6 -1.0 -0.6 -0.3 -19.2 -11.9 -7.4 25 Netherlands 28.4 25.7 -0.6 -11.6 1975 -1.1 -0.8 -0.4 -0.6 -16.3 -7.9 -13.0 26 Italy 24.0 22.0 -0.5 -11.2 1995 -0.7 -1.0 -0.6 0.1 -19.5 -13.3 2.8 27 Iceland 22.0 24.9 -0.2 -5.0 1970 17.0 -1.1 -0.4 0.7 -21.0 -7.4 15.7 28 29 Malta 20.7 24.5 0.2 5.2 na -2.5 -1.9 1.2 0.9 -34.8 30.6 21.6 30 Hungary 45.6 50.4 0.3 6.7 na 1.0 0.4 0.4 -0.2 10.3 8.8 -3.4 31 Latvia 26.7 29.2 0.4 8.7 na 1.1 -0.9 0.4 2.3 -17.7 8.7 64.4 32 Slovenia 32.5 38.8 0.5 11.8 1995 -0.9 -0.8 0.7 1.8 -16.1 16.4 46.8 33 Ukraine 24.6 27.7 0.6 13.5 na 0.1 0.4 0.5 1.4 8.6 12.5 35.9 34 Estonia 25.1 30.2 0.7 15.4 na 0.7 -1.6 0.9 3.3 -30.4 21.4 106.4 35 Russian Federation 25.9 29.6 0.7 17.3 2006 -0.8 -0.1 0.8 2.1 -2.4 20.2 58.2 36 Lithuania 24.6 30.0 0.8 18.0 na 1.0 -0.6 0.9 2.6 -12.6 21.7 74.4 37 Slovakia ¶ 36.5 43.5 0.8 18.8 1999 -0.6 -0.7 1.2 2.3 -14.9 29.9 64.9 38 Belarus § 21.5 26.2 0.9 21.9 1997 1.7 0.0 1.0 2.8 -0.3 25.7 83.0 39 Spain 21.9 28.5 1.1 27.8 na 1.4 0.7 1.1 1.6 16.6 27.5 41.6 40 41 Bulgaria 22.7 28.5 1.2 30.0 na -2.1 0.3 1.3 3.1 5.8 33.2 96.7 42 Portugal 23.8 30.4 1.3 31.5 na -0.1 0.8 1.1 2.1 18.8 27.2 58.6 43 Greece 10.7 15.3 1.7 43.5 na 0.4 1.0 1.5 2.5 23.5 40.0 73.3 44 Poland 22.4 30.6 1.7 44.7 na -0.4 0.1 2.0 3.6 1.7 56.0 118.4 45 Croatia 26.4 42.8 2.3 64.0 na 2.6 1.4 2.0 4.0 37.0 55.4 139.0 46 TFYR Macedonia ¶ 14.0 22.7 3.2 98.3 na -2.6 1.7 3.2 7.2 44.6 101.7 365.2 47 Romania 13.9 26.4 3.2 102.0 na 2.5 2.2 3.4 5.6 60.0 109.8 233.1 48 United States 26.0 16.8 -2.3 -39.8 1977 -2.2 -1.6 -2.5 -2.7 -30.3 -43.3 -45.0 49 Summary statistics (EU 27 and United States not included) 50 Mean 27.3 27.3 0.0 4.8 -0.3 -0.8 0.1 1.0 -12.4 7.2 40.4 51 Median 26.2 26.0 0.3 6.0 -0.7 -1.0 0.4 0.8 -18.6 8.8 18.7 52 53 Minimum 10.7 15.3 -2.6 -44.1 -5.2 -3.2 -2.6 -2.6 -51.1 -44.4 -44.2 54 Maximum 53.0 50.4 3.2 102.0 17.0 2.2 3.4 7.2 60.0 109.8 365.2 * 55 Age adjusted (European standard) rate per 100,000 person years. 56 ¶ Mean mortality calculated for the period 91-92 due to missing data. 57 § Similarly, for Belarus, the mean mortality corresponds to the mortality rate in 2011. 58 EU: European Union (27 countries); na: not applicable 59 60 15 https://mc.manuscriptcentral.com/bmj BMJ Page 16 of 33

1 2 3 Table 2 – Changes in colorectal cancer mortality for European women between 1989 and 2011 4 5 Mortality change for all ages (%) Mortality change 1989-2011 by age group (%) 6 Annual * Year Mean mortality For 1989-2011 change Annual Overall 7 start of last 5 8 Confidential: Fordecline Review Only 9 Countries 1989-91 2009-11 Annual Overall years <64 65-79 80+ <64 65-79 80+ 10 Austria 22.4 11.9 -3.1 -50.4 1975 -2.1 -3.4 -3.7 -2.2 -52.8 -56.3 -38.4 11 Germany 24.0 13.9 -2.9 -47.3 1973 -2.8 -2.9 -3.3 -2.2 -47.8 -52.2 -39.0 12 United Kingdom 21.3 13.3 -2.4 -41.7 1970 -2.7 -2.9 -2.6 -1.7 -47.2 -43.5 -31.4 13 Belgium 20.2 13.5 -2.2 -38.9 1970 -3.3 -2.3 -2.4 -1.9 -39.9 -41.4 -34.4 14 Switzerland 16.2 10.4 -2.2 -38.2 1970 -5.3 -1.7 -2.1 -2.7 -30.8 -37.5 -45.7 15 Czech Republic 27.9 17.9 -2.1 -37.8 1986 -2.5 -2.7 -2.3 -1.2 -45.5 -39.6 -23.1 16 Ireland 21.9 14.4 -2.0 -36.1 1970 -2.6 -2.9 -1.5 -1.8 -48.2 -28.3 -32.6 17 18 Luxembourg 19.9 15.6 -2.0 -35.4 1970 -1.4 -1.3 -3.1 -1.3 -24.7 -49.4 -24.5 19 Denmark 27.1 20.0 -1.5 -29.0 1970 0.3 -2.3 -1.6 -0.6 -40.0 -29.4 -13.2 20 EU 27 18.9 14.4 -1.4 -27.0 1976 -1.2 -1.5 -1.6 -1.0 -28.2 -30.1 -20.6 21 Finland 13.7 11.0 -1.3 -24.9 1970 0.9 -0.5 -1.7 -1.4 -10.7 -31.6 -27.3 22 France 16.2 12.5 -1.3 -24.8 1975 -1.0 -1.2 -1.5 -1.1 -22.9 -29.0 -20.8 23 Iceland 16.4 10.9 -1.1 -22.3 1970 -9.7 -1.6 -0.9 1.5 -29.5 -18.0 38.2 24 Italy 15.7 13.2 -1.0 -19.3 1993 -1.4 -1.0 -1.2 -0.6 -20.4 -22.5 -13.0 25 Hungary 28.2 24.4 -0.9 -18.5 2000 -0.4 -0.3 -1.1 -1.4 -6.4 -21.8 -26.8 26 Estonia 18.3 15.3 -0.9 -18.1 1994 -1.2 -2.4 -1.4 3.0 -41.7 -27.4 91.0 27 Netherlands 20.7 17.6 -0.8 -16.7 1970 -0.1 -0.7 -0.7 -1.2 -14.3 -13.5 -23.5 28 29 Malta 17.1 14.0 -0.8 -15.9 1979 -6.6 -2.4 1.0 -0.9 -40.8 25.0 -17.5 30 Sweden 17.0 14.6 -0.7 -13.4 1975 -2.3 -1.4 -0.4 -0.3 -26.1 -8.1 -6.0 31 Norway 21.3 19.0 -0.6 -13.3 1980 -2.7 -1.4 -0.4 -0.1 -27.0 -9.3 -2.2 32 Slovenia 18.9 17.5 -0.5 -9.6 1995 -4.1 -1.5 -0.3 0.5 -28.1 -6.5 11.6 33 Slovakia 20.5 18.6 -0.3 -6.9 1999 -4.8 -1.1 -0.3 0.9 -20.9 -7.3 21.4 34 Latvia 17.7 16.5 -0.3 -6.5 1997 -1.6 -1.2 -0.4 1.9 -23.4 -9.1 51.0 35 Lithuania 16.3 15.6 -0.3 -5.9 1993 -1.7 -1.2 -0.4 1.8 -23.0 -8.7 47.7 36 Spain 14.3 14.2 -0.3 -5.7 1994 0.7 -0.6 -0.5 0.5 -13.1 -9.8 10.8 37 Ukraine 16.7 15.9 -0.3 -5.4 na -0.1 -0.2 -0.3 0.0 -4.5 -7.2 -0.8 38 Portugal 15.6 15.3 0.0 0.1 1990 0.1 -0.1 -0.4 0.7 -2.0 -7.9 17.2 39 Bulgaria 15.8 15.8 0.1 1.2 1991 -4.8 -0.8 0.1 2.1 -16.2 1.9 58.7 40 41 Belarus 15.0 15.1 0.1 1.3 1999 1.1 -0.3 0.1 1.2 -6.6 1.7 31.3 42 Poland 15.5 16.0 0.1 2.9 1996 -0.4 -0.7 -0.1 2.2 -14.5 -2.5 61.7 43 Russian Federation 17.9 18.9 0.3 6.8 na -0.6 -0.4 0.3 2.0 -8.3 7.9 54.3 44 Greece 8.9 9.8 0.7 15.6 1999 -1.8 -0.2 -0.3 2.8 -5.3 -5.7 83.1 45 Croatia 17.6 20.1 0.8 19.0 na 1.0 0.2 0.5 2.3 3.7 11.8 63.5 46 Romania 10.5 14.5 1.7 45.8 na 0.3 0.7 1.9 4.0 16.4 50.0 135.3 47 TFYR Macedonia 9.6 14.2 2.0 56.1 na -1.7 1.1 1.8 6.1 27.3 46.5 269.6 48 United States of America 17.7 11.4 -2.2 -38.8 1970 -3.1 -1.7 -2.6 -2.3 -30.9 -43.6 -40.4 49 Summary statistics (EU 27 and United States not included) 50 Mean 18.1 15.3 -0.8 -12.7 -1.9 -1.2 -0.9 0.3 -21.6 -14.1 18.4 51 Median 17.4 15.2 -0.8 -14.7 -1.7 -1.2 -0.5 -0.1 -23.0 -9.6 -1.5 52 53 Minimum 8.9 9.8 -3.1 -50.4 -9.7 -3.4 -3.7 -2.7 -52.8 -56.3 -45.7 54 Maximum 28.2 24.4 2.0 56.1 1.1 1.1 1.9 6.1 27.3 50.0 269.6 * 55 Age adjusted (European standard) rate per 100,000 person years. 56 ¶ Mean mortality calculated for the period 91-92 due to missing data. 57 § Similarly, for Belarus, the mean mortality corresponds to the mortality rate in 2011. EU: European Union (27 countries); na: not applicable 58 59 60 16 https://mc.manuscriptcentral.com/bmj Page 17 of 33 BMJ

1 2 3 4 Figure captions 5 6 Figure 1 - Evolution of colorectal cancer mortality in Europe between 1970 and 2011. 7 Figure 2 - Percentage changes in CRC mortality between 1989-2011 in European countries and 8 Confidential: For Review Only 9 in the USA according to the mean CRC mortality in 1989-91 (AU: Austria; BEL: Belgium; BlR: 10 Belarus; BUL: Bulgaria; CRO: Croatia; CZE: Czech Republic; DEN: Denmark; GRE: Greece; EST: 11 Estonia; FIN: Finland; FR: France; GER: Germany; HUN: Hungary; ICE: Iceland; IRE: Ireland; IT: 12 13 Italy; LAT: Latvia; LIT: Lithuania; LUX: Luxembourg; MAC: Macedonia; MAL: Malta; NET: 14 Netherlands; NOR: Norway; POL: Poland; POR: Portugal; RO: Romania; RUS: Russian Federation; 15 SLK: Slovakia; SLV: Slovenia; SPA: Spain; SWE: Sweden; SWI: Switzerland; UKR: Ukraine; UK: 16 United Kingdom). 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 17 https://mc.manuscriptcentral.com/bmj BMJ Page 18 of 33 Figure 1 - Evolution of colorectal cancer mortality in Europe between 1970 and 2011.

1 Austria Belarus Belgium 2 60 60 60 3 50 50 50 4 40 40 40 5 30 30 30 6 20 20 20 7 10 10 10 8 0 Confidential:0 For Review0 Only 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 9 10 11 Bulgaria Croatia Czech Republic 12 60 60 60 13 50 50 50 14 40 40 40 15 30 30 30 16 20 20 20 17 10 10 10 18 0 0 0 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 19 20 21 Denmark Estonia Finland 22 60 60 60 23 50 50 50 24 40 40 40 25 30 30 30 26 20 20 20 27 10 10 10 28 0 0 0 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 29 30 France Germany Greece 31 32 60 60 60 33 50 50 50 34 40 40 40 35 30 30 30 36 20 20 20 37 10 10 10 0 0 0 38 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 39 40 Hungary Iceland Ireland 41 42 60 60 60 43 50 50 50 44 40 40 40 45 30 30 30 46 20 20 20 47 10 10 10 0 0 0 48 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 49 50 Italy Latvia Lithuania 51 52 60 60 60 53 50 50 50 54 40 40 40 55 30 30 30 56 20 20 20 57 10 10 10 0 0 0 58 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 59 60 18 https://mc.manuscriptcentral.com/bmj Page 19 of 33 BMJ

Luxembourg TFYR Macedonia Malta

1 60 60 60 2 50 50 50 3 40 40 40 4 30 30 30 5 20 20 20 6 10 10 10 0 0 0 7 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 8 Confidential: For Review Only 9 Netherlands Norway Poland 10 11 60 60 60 12 50 50 50 13 40 40 40 14 30 30 30 15 20 20 20 16 10 10 10 0 0 0 17 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 18 19 Portugal Romania Russian Federation 20 21 60 60 60 22 50 50 50 23 40 40 40 24 30 30 30 25 20 20 20 26 10 10 10 0 0 0 27 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 28 29 Slovakia Slovenia Spain 30 31 60 60 60 32 50 50 50 33 40 40 40 34 30 30 30 35 20 20 20 36 10 10 10 0 0 0 37 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 38 39 Sweden Switzerland Ukraine 40 41 60 60 60 42 50 50 50 43 40 40 40 44 30 30 30 45 20 20 20 46 10 10 10 0 0 0 47 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 48 49 United Kingdom EU-27 United States 50 51 60 60 60 52 50 50 50 53 40 40 40 54 30 30 30 55 20 20 20 56 10 10 10 0 0 0 57 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 58 59 60 19 https://mc.manuscriptcentral.com/bmj BMJ Page 20 of 33 Figure 2 - Percentage changes in CRC mortality between 1989-2011 in European countries and in the USA according to the mean CRC mortality in 1989-91 (AU: Austria; BEL: Belgium; BlR: Belarus; BUL: Bulgaria; CRO: 1 2 Croatia; CZE: Czech Republic; DEN: Denmark; GRE: Greece; EST: Estonia; FIN: Finland; FR: France; GER: 3 Germany; HUN: Hungary; ICE: Iceland; IRE: Ireland; IT: Italy; LAT: Latvia; LIT: Lithuania; LUX: Luxembourg; MAC: 4 Macedonia; MAL: Malta; NET: Netherlands; NOR: Norway; POL: Poland; POR: Portugal; RO: Romania; RUS: 5 Russian Federation; SLK: Slovakia; SLV: Slovenia; SPA: Spain; SWE: Sweden; SWI: Switzerland; UKR: Ukraine; 6 7 UK: United Kingdom). 8 Confidential: For Review Only 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 20 https://mc.manuscriptcentral.com/bmj Page 21 of 33 BMJ

1 2 3 References 4 5 1. Ferlay J, Steliarova-Foucher E, Lortet-Tieulent J, et al. Cancer incidence and mortality patterns in Europe: 6 estimates for 40 countries in 2012. Eur J Cancer 2013; 49 (6):1374-403. 7 2. Bosetti C, Levi F, Rosato V, et al. Recent trends in colorectal cancer mortality in Europe. Int J Cancer 8 Confidential:2011; 129 (1):180-91. For Review Only 9 3. Altobelli E, Lattanzi A, Paduano R, et al. Colorectal cancer prevention in Europe: Burden of disease and 10 status of screening programs. Prev Med 2014; 62C :132-41. 11 4. Brenner H, Kloor M, Pox CP. Colorectal cancer. The Lancet 2014; 383 (9927):1490-502. 12 5. Boniol M, Autier P. Prevalence of main cancer lifestyle risk factors in Europe in 2000. Eur J Cancer 13 14 2010; 46 (14):2534-44. 15 6. Organization WH. Mortality database: World Health Organization, 2013. 16 7. Lutz JM, Pury P, Fioretta G, et al. The impact of coding process on observed cancer mortality trends in 17 Switzerland. Eur J Cancer Prev 2004; 13 (1):77-81. 18 8. Waterhouse J, Muir C, Correa P, et al. Cancer incidence in five continents. Vol III . Geneva, Switzerland: 19 World Health Organization, 1976. 20 9. Jointpoint regression program [program]. Version 3.4.2. version, 2009. 21 10. Wallace D, Walker K, Kuryba A, et al. Identifying patients at risk of emergency admission for colorectal 22 cancer. Br J Cancer 2014; 111 (3):577-80. 23 24 11. Bojer AS, Roikjaer O. Elderly patients with colorectal cancer are oncologically undertreated. Eur J Surg 25 Oncol 2015; 41 (3):421-5. 26 12. McKay A, Donaleshen J, Helewa RM, et al. Does young age influence the prognosis of colorectal 27 cancer: a population-based analysis. World J Surg Oncol 2014; 12 :370. 28 13. Zatonski WA, Bhala N. Changing trends of diseases in Eastern Europe: closing the gap. Public Health 29 2012; 126 (3):248-52. 30 14. Ondrusova M, Muzik J, Hrcka R, et al. Do we know the cause of the highest colorectal cancer incidence, 31 the changes in the mortality trends and the clinical stages in the Slovak and Czech Republic, the 32 representatives of the Central European region? Neoplasma 2011; 58 (4):283-90. 33 34 15. Roy HK, Bianchi LK. Differences in colon adenomas and carcinomas among women and men: potential 35 clinical implications. JAMA 2009; 302 (15):1696-7. 36 16. Stock C, Haug U, Brenner H. Population-based prevalence estimates of history of colonoscopy or 37 sigmoidoscopy: review and analysis of recent trends. Gastrointest Endosc 2010; 71 (2):366-81 e2. 38 17. Robb K, Stubbings S, Ramirez A, et al. Public awareness of cancer in Britain: a population-based survey 39 of adults. Br J Cancer 2009; 101 Suppl 2 :S18-23. 40 18. Quaife SL, Forbes LJ, Ramirez AJ, et al. Recognition of cancer warning signs and anticipated delay in 41 help-seeking in a population sample of adults in the UK. Br J Cancer 2014; 110 (1):12-8. 42 19. Svensson E, Grotmol T, Hoff G, et al. Trends in colorectal cancer incidence in Norway by gender and 43 44 anatomic site: an age-period-cohort analysis. Eur J Cancer Prev 2002; 11 (5):489-95. 45 20. Murakami R, Tsukuma H, Kanamori S, et al. Natural history of colorectal polyps and the effect of 46 polypectomy on occurrence of subsequent cancer. Int J Cancer 1990; 46 (2):159-64. 47 21. Pox CP, Altenhofen L, Brenner H, et al. Efficacy of a nationwide screening colonoscopy program for 48 colorectal cancer. Gastroenterology 2012; 142 (7):1460-7.e2. 49 22. Jezewski-Serra D, Salines E. Évaluation épidémiologique du programme de dépistage organisé du cancer 50 colorectal en France. France: Institut de Veille Sanitaire, 2013. 51 23. Atkin WS, Edwards R, Kralj-Hans I, et al. Once-only flexible sigmoidoscopy screening in prevention of 52 colorectal cancer: a multicentre randomised controlled trial. Lancet 2010; 375 (9726):1624-33. 53 54 24. Holme O, Loberg M, Kalager M, et al. Effect of flexible sigmoidoscopy screening on colorectal cancer 55 incidence and mortality: a randomized clinical trial. JAMA 2014; 312 (6):606-15. 56 25. Schoen RE, Pinsky PF, Weissfeld JL, et al. Colorectal-cancer incidence and mortality with screening 57 flexible sigmoidoscopy. N Engl J Med 2012; 366 (25):2345-57. 58 59 60 https://mc.manuscriptcentral.com/bmj 21 BMJ Page 22 of 33

1 2 3 26. Kronborg O, Jorgensen OD, Fenger C, et al. Randomized study of biennial screening with a faecal occult 4 blood test: results after nine screening rounds. Scand J Gastroenterol 2004; 39 (9):846-51. 5 27. Scholefield JH, Moss SM, Mangham CM, et al. Nottingham trial of faecal occult blood testing for 6 colorectal cancer: a 20-year follow-up. Gut 2012; 61 (7):1036-40. 7 28. Shaukat A, Mongin SJ, Geisser MS, et al. Long-term mortality after screening for colorectal cancer. N 8 Confidential: For Review Only 9 Engl J Med 2013; 369 (12):1106-14. 10 29. Rehm J, Gmel G. Alcohol consumption and public health in Russia. Lancet 2007; 369 (9578):1975-6. 11 30. Rehm J, Sulkowska U, Manczuk M, et al. Alcohol accounts for a high proportion of premature mortality 12 in central and eastern Europe. Int J Epidemiol 2007; 36 (2):458-67. 13 31. Zaridze D, Brennan P, Boreham J, et al. Alcohol and cause-specific mortality in Russia: a retrospective 14 case-control study of 48,557 adult deaths. Lancet 2009; 373 (9682):2201-14. 15 32. Schwingshackl L, Hoffmann G. Adherence to Mediterranean diet and risk of cancer: a systematic review 16 and meta-analysis of observational studies. Int J Cancer 2014; 135 (8):1884-97. 17 33. Rodrigues SS, Caraher M, Trichopoulou A, et al. Portuguese households' diet quality (adherence to 18 19 Mediterranean food pattern and compliance with WHO population dietary goals): trends, regional 20 disparities and socioeconomic determinants. Eur J Clin Nutr 2008; 62 (11):1263-72. 21 34. Bach-Faig A, Fuentes-Bol C, Ramos D, et al. The Mediterranean diet in Spain: adherence trends during 22 the past two decades using the Mediterranean Adequacy Index. Public Health Nutr 2011; 14 (4):622-8. 23 35. Tourlouki E, Matalas AL, Bountziouka V, et al. Are current dietary habits in Mediterranean islands a 24 reflection of the past? Results from the MEDIS study. Ecol Food Nutr 2013; 52 (5):371-86. 25 36. Bonaccio M, Di Castelnuovo A, Bonanni A, et al. Decline of the Mediterranean diet at a time of 26 economic crisis. Results from the Moli-sani study. Nutr Metab Cardiovasc Dis 2014; 24 (8):853-60. 27 37. Boyle P, Leon ME. Epidemiology of colorectal cancer. Br Med Bull 2002; 64 :1-25. 28 29 38. Kirkegaard H, Johnsen NF, Christensen J, et al. Association of adherence to lifestyle recommendations 30 and risk of colorectal cancer: a prospective Danish cohort study. BMJ 2010; 341 :c5504. 31 39. Mills KT, Bellows CF, Hoffman AE, et al. Diabetes mellitus and colorectal cancer prognosis: a meta- 32 analysis. Dis Colon Rectum 2013; 56 (11):1304-19. 33 40. Peeters PJ, Bazelier MT, Leufkens HG, et al. The Risk of Colorectal Cancer in Patients With Type 2 34 Diabetes: Associations With Treatment Stage and Obesity. Diabetes Care 2014. 35 41. MacArthur C, Smith A. Delay in the diagnosis of colorectal cancer. J R Coll Gen Pract 36 1983; 33 (248):159-61. 37 42. Binefa G, Rodriguez-Moranta F, Teule A, et al. Colorectal cancer: from prevention to personalized 38 39 medicine. World J Gastroenterol 2014; 20 (22):6786-808. 40 43. Kanavos P, Schurer W. The dynamics of colorectal cancer management in 17 countries. Eur J Health 41 Econ 2010; 10 Suppl 1 :S115-29. 42 44. DH Knowledge & Intelligence Team. Further Analysis of HES Endoscopy Data. UK: Department of 43 Health, 2010. 44 45. Galloway JM, Gibson J, Dalrymple J. Endoscopy in primary care--a survey of current practice. Br J Gen 45 Pract 2002; 52 (480):536-8. 46 46. Weiss DG, Homoncik M. Colon cancer screening in Austria - Update 2008. J Gastroenterol Hepatol Erkr 47 2008; 6(2):19-22. 48 49 47. Wilschut JA, Habbema JD, van Leerdam ME, et al. Fecal occult blood testing when colonoscopy 50 capacity is limited. J Natl Cancer Inst 2011; 103 (23):1741-51. 51 48. Stock C, Brenner H. Utilization of lower gastrointestinal endoscopy and fecal occult blood test in 11 52 European countries: evidence from the Survey of Health, Aging and Retirement in Europe (SHARE). 53 Endoscopy 2010; 42 (7):546-56. 54 49. Edwards BK, Ward E, Kohler BA, et al. Annual report to the nation on the status of cancer, 1975-2006, 55 featuring colorectal cancer trends and impact of interventions (risk factors, screening, and treatment) 56 to reduce future rates. Cancer 2010; 116 (3):544-73. 57 58 59 60 https://mc.manuscriptcentral.com/bmj 22 Page 23 of 33 BMJ

1 2 3 50. Haidinger G, Waldhoer T, Vutuc C. Self-reported colonoscopy screening in Austria. Eur J Cancer Prev 4 2008; 17 (4):354-7. 5 51. Gross CP, Andersen MS, Krumholz HM, et al. Relation Between Medicare Screening Reimbursement 6 and Stage at Diagnosis for Older Patients With Colon Cancer. JAMA 2006; 296 :2815-22. 7 52. Cunningham D, Atkin W, Lenz HJ, et al. Colorectal cancer. Lancet 2010; 375 (9719):1030-47. 8 Confidential: For Review Only 9 53. Oliphant R, Nicholson GA, Horgan PG, et al. Contribution of surgical specialization to improved 10 colorectal cancer survival. Br J Surg 2013; 100 (10):1388-95. 11 54. van der Pool AE, Damhuis RA, Ijzermans JN, et al. Trends in incidence, treatment and survival of 12 patients with stage IV colorectal cancer: a population-based series. Colorectal Dis 2012; 14 (1):56-61. 13 55. Boyle P. Relative value of incidence and mortality data in cancer research. Recent Results Cancer Res 14 1989; 114 :41-63. 15 56. Harteloh P, de Bruin K, Kardaun J. The reliability of cause-of-death coding in The Netherlands. Eur J 16 Epidemiol 2010; 25 (8):531-8. 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmj 23 BMJ Page 24 of 33

1 2 3 4 Appendix to the article:”Trends in colorectal cancer mortality in Europe: an analysis of the 5 WHO mortality database” 6 7 8 Confidential: For Review Only 9 by Driss Ait Ouakrim Research Fellow 1, Cécile Pizot Research Officer 2, Magali Boniol 10 Assistant Statistician 2, Matteo Malvezzi Post Doctoral Fellow 3, Mathieu Boniol Vice- 11 President Biostatistics 2, 4 , Eva Negri Head of laboratory 5, Maria Bota Research Officer 2, 4 , 12 Mark A Jenkins Professor and Director 1, Harry Bleiberg Professor Emeritus 6, Philippe Autier 13 2, 4 14 Vice-President Population Research . 15 1 Centre for Epidemiology and Biostatistics, Melbourne School of Population and Global 16 17 Health, University of Melbourne, 3010, Australia 18 2 International Prevention Research Institute (iPRI), 69006 Lyon, France 19 20 3 Department of Clinical Sciences and Community Health, Universitá degli Studi di Milano, 21 20156 Milan, Italy 22 23 4 University of Strathclyde Institute for Global Public Health at iPRI, 69006 Lyon, France 24 5 Department of Epidemiology, IRCCS - Istituto di Ricerche Farmacologiche ‘Mario Negri’, 25 26 20156 Milan, Italy. 27 6 Jules Bordet Institute, 1000 Brussels, Belgium 28 29 30 31 Correspondence: Dr Philippe Autier, International Prevention Research Institute (iPRI), 32 Cours Lafayette 95, 69006 Lyon Cedex 08, France; E-mail: philippe.autier@i-pri. org; 33 website : www.i-pri.org . 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmj 1 Page 25 of 33 BMJ

1 2 3 Supplementary Table 1 – Years with missing colorectal cancer mortality data per country 4 5 Supplementary Table 2 – Model parameters used for the joinpoint regression analysis 6 Supplementary Figure 1 - Evolution of colorectal cancer mortality between 1970 and 2010 7 for men (blue dots) and women (red dots) aged less than 65 years. 8 Confidential: For Review Only 9 10 Supplementary Figure 2 - Evolution of colorectal cancer mortality between 1970 and 2010 11 for men (blue dots) and women (red dots) aged 65-79 years. 12 13 Supplementary Figure 3 - Evolution of colorectal cancer mortality between 1970 and 2010 14 for men (blue dots) and women (red dots) aged 80 years and over. 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmj 2 BMJ Page 26 of 33

1 2 3 Supplementary Table 1 – Years with missing colorectal cancer mortality data per country 4 Country Missing year(s) 5 Austria None 6 1970; 1971; 1972; 1973; 1974; 1975; 1976; 1977; 1978; 1979; 1980; 1983; Belarus 7 1984; 1996; 2004; 2005; 2006; 2010 8 Confidential:Belgium 2000; 2001; 2002; 2011 For Review Only 9 Bulgaria None 1970; 1971; 1972; 1973; 1974; 1975; 1976; 1977; 1978; 1979; 1980; 1981; 10 Croatia 11 1982; 1983; 1984 1970; 1971; 1972; 1973; 1974; 1975; 1976; 1977; 1978; 1979; 1980; 1981; 12 Czech Republic 13 1982; 1983; 1984; 1985 14 Denmark None Estonia 1970; 1971; 1972; 1973; 1974; 1975; 1976; 1977; 1978; 1979; 1980; 1983; 1984 15 Finland None 16 France None 17 Germany 1970; 1971; 1972; 1979 18 Greece None 19 Hungary None 20 Iceland 2010; 2011 21 Ireland 2011 22 Italy 2004; 2005 23 Latvia 1970; 1971; 1972; 1973; 1974; 1975; 1976; 1977; 1978; 1979 24 Lithuania 1970; 1971; 1972; 1973; 1974; 1975; 1976; 1977; 1978; 1979; 1980; 1983; 1984 25 Luxembourg None 26 Malta None 27 Netherlands None 28 Norway None 29 Poland 1997; 1998 30 Portugal 2004; 2005; 2006 31 Romania 1979 32 Russian Federation 1970; 1971; 1972; 1973; 1974; 1975; 1976; 1977; 1978; 1979 1970; 1971; 1972; 1973; 1974; 1975; 1976; 1977; 1978; 1979; 1980; 1981; 33 Slovakia 34 1982; 1983; 1984; 1985; 1986; 1987; 1988; 1989; 1990; 1991; 2011 1970; 1971; 1972; 1973; 1974; 1975; 1976; 1977; 1978; 1979; 1980; 1981; 35 Slovenia 36 1982; 1983; 1984; 2011 37 Spain None 38 Sweden None 39 Switzerland 2011 1970; 1971; 1972; 1973; 1974; 1975; 1976; 1977; 1978; 1979; 1980; 1981; 40 TFYR Macedonia 1982; 1983; 1984; 1985; 1986; 1987; 1988; 1989; 1990; 2011 41 Ukraine 1970; 1971; 1972; 1973; 1974; 1975; 1976; 1977; 1978; 1979; 1980; 1983; 1984 42 United Kingdom 2000; 2011 43 United States 2011 44

45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmj 3 Page 27 of 33 BMJ

1 2 3 Supplementary Table 2 – Model parameters used for the joinpoint regression analysis 4 Parameter Value 5 Model Ln(y) = bx 6 Dependent variable Age Standardised Rate 7 Independent variable Year 8 Confidential:By variable ForAdd Age Reviewand Country Only 9 For EU27 data: add Age and Sex 10 Heteroscedastic errors option Constant variance (homoscedasticity) 11 Number of joinpoint Min=0 and Max=3 12 Method Grid Search 13 Minimum number observations from a joinpoint 6 14 to either end of the data 15 Minimum number of observations between two 6 joinpoints 16 Model selection method Permutation test 17 Significance level 0.05 18 Number of permutations 4499 19

20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmj 4 BMJ Page 28 of 33

Supplementary Figure 1. Evolution of colorectal cancer mortality between 1970 and 2010 for men (blue 1 dots) and women (red dots) aged less than 65 years. 2 3 4 Austria Belarus Belgium

5 25 25 25

6 20 20 20 7 15 15 15 8 Confidential: For Review Only 10 10 10 9 10 5 5 5 0 0 0 11 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 12 13 Bulgaria Croatia Czech Republic 14 25 25 25 15 16 20 20 20 17 15 15 15 18 10 10 10 19 5 5 5 20 0 0 0 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 21 22 Denmark Estonia Finland 23 24 25 25 25 25 20 20 20 26 15 15 15

27 10 10 10

28 5 5 5 29 0 0 0 30 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 31 32 France Germany Greece

33 25 25 25

34 20 20 20 35 15 15 15 36 37 10 10 10 38 5 5 5 0 0 0 39 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 40 41 Hungary Iceland Ireland 42 43 25 25 25 44 20 20 20 45 15 15 15 46 10 10 10 47 5 5 5

48 0 0 0 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 49 50 51 Italy Latvia Lithuania 52 25 25 25 53 20 20 20

54 15 15 15

55 10 10 10

56 5 5 5 57 0 0 0 58 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 59

60 https://mc.manuscriptcentral.com/bmj 5 Page 29 of 33 BMJ

1 2 Luxembourg Macedonia Malta 3 25 25 25 4 20 20 20

5 15 15 15

6 10 10 10 7 5 5 5 8 Confidential: For Review Only 0 0 0 9 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 10 11 Netherlands Norway Poland

12 25 25 25 13 20 20 20 14 15 15 15 15 16 10 10 10 17 5 5 5 0 0 0 18 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 19 20 Portugal Romania Russian Federation 21 22 25 25 25 23 20 20 20 24 15 15 15 25 10 10 10

26 5 5 5

27 0 0 0 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 28 29 30 Slovakia Slovenia Spain 31 25 25 25

32 20 20 20

33 15 15 15

34 10 10 10 35 5 5 5 36 0 0 0 37 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 38 39 Sweden Switzerland Ukraine

40 25 25 25 41 20 20 20 42 43 15 15 15 44 10 10 10 45 5 5 5 0 0 0 46 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 47 48 United Kingdom USA EU-27 49 50 25 25 25 51 20 20 20 52 15 15 15 53 10 10 10

54 5 5 5

55 0 0 0 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 56 57 58 59 60 https://mc.manuscriptcentral.com/bmj 6 BMJ Page 30 of 33

Supplementary Figure 2. Evolution of colorectal cancer mortality between 1970 and 2010 for men (blue 1 dots) and women (red dots) aged 65 to 79 years. 2 3 Austria Belarus Belgium 4 350 350 350 5 300 300 300 6 250 250 250 200 200 200 7 150 150 150 8 100 Confidential:100 For Review100 Only 9 50 50 50 0 0 0 10 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 11 12 Bulgaria Croatia Czech Republic 13 350 350 350 14 300 300 300 15 250 250 250 16 200 200 200 150 150 150 17 100 100 100 18 50 50 50 19 0 0 0 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 20 21 Denmark Estonia Finland 22 23 350 350 350 300 300 300 24 250 250 250 25 200 200 200 26 150 150 150 27 100 100 100 50 50 50 28 0 0 0 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 29 30 31 France Germany Greece

32 350 350 350 33 300 300 300 34 250 250 250 200 200 200 35 150 150 150 36 100 100 100 37 50 50 50 0 0 0 38 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 39 40 Hungary Iceland Ireland 41 350 350 350 42 300 300 300 43 250 250 250 44 200 200 200 45 150 150 150 100 100 100 46 50 50 50 47 0 0 0 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 48 49 Italy Latvia Lithuania 50 51 350 350 350 300 300 300 52 250 250 250 53 200 200 200 54 150 150 150 55 100 100 100 50 50 50 56 0 0 0 57 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010

58 59

60 https://mc.manuscriptcentral.com/bmj 7 Page 31 of 33 BMJ

1 2 Luxembourg Macedonia Malta 3 350 350 350 4 300 300 300 250 250 250 5 200 200 200 6 150 150 150 7 100 100 100 8 50 Confidential:50 For Review50 Only 0 0 0 9 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 10 11 Netherlands Norway Poland

12 350 350 350 13 300 300 300 14 250 250 250 15 200 200 200 150 150 150 16 100 100 100 17 50 50 50 0 0 0 18 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 19 20 Portugal Romania Russian Federation 21 22 350 350 350 300 300 300 23 250 250 250 24 200 200 200 25 150 150 150 26 100 100 100 50 50 50 27 0 0 0 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 28 29 30 Slovakia Slovenia Spain 31 350 350 350 32 300 300 300 250 250 250 33 200 200 200 34 150 150 150 35 100 100 100 36 50 50 50 0 0 0 37 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 38 39 Sweden Switzerland Ukraine

40 350 350 350 41 300 300 300 42 250 250 250 43 200 200 200 150 150 150 44 100 100 100 45 50 50 50 0 0 0 46 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 47 48 United kingdom USA EU-27 49 50 350 350 350 300 300 300 51 250 250 250 52 200 200 200 53 150 150 150 54 100 100 100 50 50 50 55 0 0 0 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 56 57 58 59 60 https://mc.manuscriptcentral.com/bmj 8 BMJ Page 32 of 33

Supplementary Figure 3. Evolution of colorectal cancer mortality between 1970 and 2010 for men (blue 1 dots) and women (red dots) aged 80 years and above. 2 3 Austria Belarus Belgium 4 600 600 600 5 500 500 500 6 400 400 400 7 300 300 300 8 200 Confidential:200 For Review200 Only 9 100 100 100 0 0 0 10 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 11 12 Bulgaria Croatia Czech Republic 13 600 600 600

14 500 500 500

15 400 400 400

16 300 300 300

17 200 200 200 18 100 100 100 19 0 0 0 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 20 21 Denmark Estonia Finland 22 23 600 600 600 24 500 500 500 25 400 400 400 300 300 300 26 200 200 200 27 100 100 100

28 0 0 0 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 29 30 31 France Germany Greece

32 600 600 600 33 500 500 500 34 400 400 400 35 300 300 300 36 200 200 200 37 100 100 100 0 0 0 38 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 39 40 Hungary Iceland Ireland 41 600 600 600

42 500 500 500

43 400 400 400

44 300 300 300

45 200 200 200 46 100 100 100 47 0 0 0 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 48 49 Italy Latvia Lithuania 50 51 600 600 600 52 500 500 500 53 400 400 400 54 300 300 300 200 200 200 55 100 100 100

56 0 0 0 57 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010

58 59 60 https://mc.manuscriptcentral.com/bmj 9 Page 33 of 33 BMJ

Luxembourg Macedonia Malta 1 2 600 600 600 3 500 500 500 4 400 400 400 5 300 300 300 6 200 200 200 100 100 100 7 0 0 0 8 1970Confidential: 1980 1990 2000 2010 1970 1980For 1990 2000Review 2010 1970 1980 Only 1990 2000 2010 9 10 Netherland Norway Poland

11 600 600 600

12 500 500 500 13 400 400 400 14 300 300 300 15 200 200 200 16 100 100 100 0 0 0 17 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 18 19 Portugal Romania Russian Federation 20 600 600 600 21 500 500 500

22 400 400 400

23 300 300 300

24 200 200 200

25 100 100 100

26 0 0 0 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 27 28 29 Slovakia Slovenia Spain 30 600 600 600 31 500 500 500 32 400 400 400 33 300 300 300 34 200 200 200 100 100 100 35 0 0 0 36 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 37 38 Sweden Switzerland Ukraine

39 600 600 600

40 500 500 500 41 400 400 400 42 300 300 300 43 200 200 200 44 100 100 100 0 0 0 45 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 46 47 United Kingdom USA EU-27 48 49 600 600 600 500 500 500 50 400 400 400

51 300 300 300

52 200 200 200

53 100 100 100

54 0 0 0 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmj 10