Social Inequalities, Global Public Health, and Cancer Michael Marmot

PART 1 CHAPTER 2 CHAPTER general considerations 2 CHAPTER

chapter 2. GENERAL CONSIDERATIONS Social inequalities, global public health, and cancer Michael Marmot

Introduction equalities in the occurrence of NCDs low-income countries. Within coun- being the silent guest at the table, tries, the so-called diseases of afflu- The World Health Organization there but not openly acknowledged, ence are no longer; the lower peo- and the Government of Uruguay I was invited to speak at the opening ple are in the social hierarchy, the held the Global Conference on experts’ plenary. I began along the higher their risk of NCDs. We cannot Noncommunicable Diseases (NCDs) following lines. deal with NCDs without dealing with in Montevideo in October 2017, a “NCDs are a global health prob- the social determinants of health most welcome initiative. The aim was lem. One purpose of our meeting inequalities. to prepare for the third United Nations here in Montevideo is to plan for an There is a rumour going around General Assembly high-level meeting NCD summit to be held at the United that poor people are poor because on NCDs, another welcome develop- Nations in New York in September they make poor choices, and that ment. At the Montevideo summit we 2018. If you attend that summit and, poor people are unhealthy because discussed strategies to reduce the while there, go to Central Park for they make unhealthy choices. This global burden of NCDs by both pre- a little exercise in green space – rumour is, very largely, a myth. It has vention and treatment. Inequalities in good for mental as well as physical the causal connection backwards. the burden of NCDs were not a strong health – you may find your life at risk. More accurately, it is not mythical feature of the meeting, however. They Mown down by hordes of high-in- that the rumour exists – I read it in surfaced prominently to the extent come joggers.” the press nearly daily – but the evi- that universal health coverage aims Much as I applaud people taking dence points the other way. It is not to deal with inequalities in access to responsibility for their health, these poor choices that lead to poverty but health care; out-of-pocket payments high-income New Yorkers are atyp- poverty that leads to poor choices. for medical care can, and do, impover- ical. Globally, the burden of NCDs is An Indian villager is more likely to ish people globally. To avoid social in- in middle-income and, increasingly, invest in longer-term strategies if the

Chapter 2. Social inequalities, global public health, and cancer 77 harvest has been good. If it has been social gradients in adult mortality these social and economic inequal- poor, he will focus on how to get (Bassanesi et al., 2008). In high-in- ities, one might speculate that there calories for his family tomorrow, not come countries, the gradient is will always be inequalities in health. on strategies for future prosperity. clear: higher socioeconomic posi- Probably, there will be. However, the A single mother may respond to the tion means lower mortality, whether magnitude varies between countries admonition to read bedtime stories to socioeconomic position is measured and, over time, within countries. If the her children – it’s good for their long- by education level, income, occupa- magnitude of inequalities, that is, the term future – that she would if she tional status, or degree of deprivation slope of the gradient, is not fixed, it could be sure that they would have a of area of residence. A recent pub- suggests that action to reduce these bed, let alone a book. lication from Lifepath, a European inequalities should be possible. So it is with healthy choices. Under collaborative research programme, First, consider the variation in different circumstances, people with showed that in 48 cohorts low occu- health inequalities between coun- lower income would be more likely to pational status was associated with tries: it is marked. In Europe, we cal- adopt the choices that are good for higher all-cause mortality (Stringhini culated life expectancy at the age of health. Having time to think about et al., 2017). 25 years by education level. In each exercise is a luxury that people at If all societies have social and country, men with a university educa- the economic margins may not have, economic inequalities – in education tion had a longer life expectancy than quite apart from a lack of amenities; level, income, occupational status, men with only a primary education healthy food may be beyond a house- and living conditions – and the so- (Fig. 2.1), but the gap varied (Marmot hold budget. The stress of marginal cial gradient in health follows from and UCL Institute of Health Equity, employment would be happily for- gone if better jobs were available. Globally, to take effective action Fig. 2.1. Life expectancy for men aged 25 years, by education level, in different European countries. ISCED, International Standard Classification on NCDs, we need to address in- of Education; levels 0–2, pre-primary, primary, and lower secondary equalities in NCDs, and this entails education, or second stage of basic education; levels 5–8, first stage of action on the social determinants of tertiary and second stage of tertiary education. Source: Eurostat [demo_ health. This chapter focuses on what mlexpecedu] (2016). we can do, but first we look at cancer in the context of inequalities in health.

Health inequalities: the gradient

All countries from which we have good data show inequalities in mortality. Such inequalities are not confined to poor health for the poor, but follow a social gradient (Marmot, 2015). In low-income countries, where the systematic collection of data on inequalities is uncommon, we have data on child mortality from Demographic and Health Surveys 2 T 2 (Gwatkin, 2007). In country after 2 A ISCED 2 2 country, the lower the wealth quintile, 2 L the higher the mortality rate of chil- 2 2 dren under the age of 5 years. Specific analyses from middle-income countries such as Brazil show E C D EE EL HR IT H PL PT RO SI S I SE NO TR

nnual dierence E in onths 222 2 D 2 22 2 A

2 T 2 2 A ISCED 2 2 2 L 2 2

E C D EE EL HR IT H PL PT RO SI S I SE NO TR

2013). Sweden has the second long- Fig. 2.2. Trends in difference in life expectancy between areas of greatest est life expectancy at the age of 25 deprivation and the average. Source: compiled from Barr et al. (2017). years, and a narrow gap between those with a primary education and nnual dierence E in onths

those with a tertiary education. Men 2 CHAPTER living in the countries located to the east of Sweden have a lower average life expectancy, and there is a wider GENERAL CONSIDERATIONS 222 2 gap in life expectancy between those D with a primary education and those 2 22 with a tertiary education. Looking at 2 A it a different way, the health disadvantage of living in these countries is greater for those with a primary education than for those with a tertiary education. They do know how to get good health in Bulgaria, Estonia, Hungary, and Romania … it happens for men with university education. It is those lower in the hierarchy who The data are, at the least, consistent of cancers in the United Kingdom are suffer most. with the notion that having an explic- preventable, and that globally the es- The magnitude of health inequali- it policy to do something about the timate is similar (30–50%). According ties also varies within countries over problem may help. to Cancer Research UK, the routes time. Part of that variation may result to lower cancer risk are, in order of from conscious policy decisions. In Cancer and health inequalities importance, to be a non-smoker, England, Barr et al. (2017) looked at maintain a healthy weight, eat fruits Cancer has perhaps featured less the gap in life expectancy between and vegetables, consume less al- than some other diseases in the con- the poorest 20% of local authorities cohol, be “SunSmart”, eat less pro- sideration of health inequalities. This and the remainder; their results are cessed meat and red meat, eat a volume is a timely reminder that that summarized in Fig. 2.2. The New high-fibre diet, be physically active, should no longer be the case, as il- Labour government, elected in 1997, and eat less meat. lustrated by Fig. 2.3. Although the developed a national strategy to re- Each of these is socially deter- contribution of cancer to absolute in- duce health inequalities. In the period mined, and many show a social gra- equalities in all-cause mortality var- before the strategy was put in place, dient; more unhealthy behaviours ies among countries, it is substantial the health gap between the poorest are observed in those lower in the in each case. Unexpectedly, cancer 20% and the remainder was widen- hierarchy. It used to be thought that looms large in middle-income coun- ing. During the period of the strategy, in low- and middle-income countries tries. It is, however, more difficult to the gap narrowed. In the period after cigarette smoking was more com- obtain data on social inequalities. the strategy, when a Conservative- mon in groups with higher income or There are exceptions, such as the led coalition government changed higher education level, unlike the gra- demonstration of inequalities by edu- policy direction, health inequalities dient seen in high-income countries. cation level in Colombia, particularly increased again. This is no longer the case. In many for cancers of the stomach and cer- This simple correlation in time lower- and upper-middle-income vix, both of which are linked to infec- does not prove causation, nor does countries, smoking is more com- tion (de Vries et al., 2015). it tell us what feature, if any, of gov- mon lower in the hierarchy (Global ernment policy might have made the Cancer inequalities: the causes Tobacco Economics Consortium, most difference. What the data do 2018). Similarly with obesity, the so- show is that the magnitude of health Cancer Research UK (Gordon- cial gradient is clear in North America inequalities can vary quite quickly. Dseagu, 2006) concluded that 40% and Europe, as illustrated in Fig. 2.4

Chapter 2. Social inequalities, global public health, and cancer 99 Fig. 2.3. Contribution of selected noncommunicable diseases to absolute inequality (all-cause) death rates between lowest and highest quintiles of community socioeconomic status for those aged 30–64 years. Source: reprinted from Di Cesare et al. (2013), copyright 2013, with permission from Elsevier.

304 years 304 years E E

N N

S S

2 2 2 2 C C Cause o death C C C D A Cause o death C C C D A

for 19 Member States of the European Cancer inequalities: what can dividuals without taking into account Union. be done? the social determinants that lead to Worryingly, in Britain there is a inequalities in these behaviours. social gradient in childhood obesity, It can be argued that the mind is an A review of health inequalities which has been steepening over time important gateway by which social in England, Fair society, healthy determinants affect health equity. (NHS Digital, 2017). Given that child- lives: the Marmot review (Marmot, The above-mentioned risks iden- hood obesity tracks into adulthood, 2010), identified six domains of rec- tified by Cancer Research UK are ommendations to reduce avoidable we are facing increasing inequalities behavioural, that is, controlled by the health inequalities and promote in obesity in the future. mind. As I stated at the beginning health equity: (i) giving every child I have placed emphasis on so- of this chapter, however, it is inad- the best start in life; (ii) education cioeconomic differences in health equate to see these behaviours as and lifelong learning; (iii) employ- and disease. In the Commission on simply being under the control of in- ment and working conditions; (iv) a Equity and Health Inequalities in the Americas, established by the Pan American Health Organization, we Fig. 2.4. Obesity prevalence according to educational attainment, averaged over 19 European Union Member States. Source: reproduced from report are also concerned with ethnic and by Robertson et al. (2007). gender inequalities. We see that throughout the Americas, Indigenous besity prealence groups are disadvantaged in terms 2 of social determinants of health com- pared with non-Indigenous groups. In many countries of the Americas, including the USA, people of African 2 descent are commonly subject to structural racism, which compounds socioeconomic disadvantage. These socioeconomic and ethnic disadvan- tages can combine with differences between sexes in social determi- L S S H nants of health. E

10 minimum income for healthy living The argument I have just stated an environmental component; ex- (everyone should have at least the is not confined to cancer. In recent amples are banning smoking in pub- minimum income that would ena- months, I have been invited to talk lic places and encouraging active ble them to live a healthy life); (v) about social determinants and health transport. Environmental, rather than

healthy and sustainable places equity to groups concerned with simply behavioural, interventions will 2 CHAPTER and environments in which to live medical education, internal medi- also be important in relation to oth- and work; and (vi) taking a social cine, cardiology, surgery, paediatrics, er medical conditions: reducing air determinants approach to preven- thoracic medicine, mental illness, pollution, improving housing quality, GENERAL CONSIDERATIONS tion, that is, not just looking at re- obstetrics and gynaecology, cancer and designing neighbourhoods to ducing smoking and unhealthy diet, control, primary care, pharmacy, promote health and well-being. for example, but also looking at the early child development, violence, We need to make common cause, causes of the social distribution of inclusion health, health psychology, not only to reduce inequalities in can- these behaviours (the causes of the psychosomatic medicine, vegetable cer and specific medical conditions causes). To prevent inequalities in summit, and concepts of honour, as but also to promote health equity cancer, attention must be paid to well as public health and health pro- more generally. A commitment to ac- these six domains through the en- motion, urban renewal, local govern- tion on social determinants of health tire life-course. In other words, pro- ment, and community development. is urgently required. motion of health equity with respect The causes of cancer identified to cancer should be part of a gen- by Cancer Research UK are large- eral approach to reducing health ly behavioural, but interventions to inequalities. address these causes will also have

Key points • About 40% of cancers are judged to be preventable by simple lifestyle changes; these causes of cancer, smoking and obesity principal among them, are socially determined. • These causes follow a social gradient, with lower socioeconomic position corresponding to higher risk. They contribute to inequalities in the occurrence of cancer. • To make progress in reducing inequalities in cancer we must address the causes of the causes of cancer. • Evidence shows that inequalities in mortality vary between and within countries. The implication is that inequalities are not fixed but can be improved.

Chapter 2. Social inequalities, global public health, and cancer 1111 References

Barr B, Higgerson J, Whitehead M (2017). Global Tobacco Economics Consortium (2018). Marmot M (2015). The health gap. London, UK: Investigating the impact of the English health The health, poverty, and financial consequences Bloomsbury. inequalities strategy: time trend analysis. BMJ. of a cigarette price increase among 500 358:j3310. https://doi.org/10.1136/bmj.j3310 million male smokers in 13 middle income NHS Digital (2017). National child measurement PMID:28747304 countries: compartmental model study. BMJ. programme – England, 2016-17. National Health 361:k1162. https://doi.org/10.1136/bmj.k1162 Service. Available from: https://digital.nhs.uk/ Bassanesi SL, Azambuja MI, Achutti A (2008). PMID:29643096 data-and-information/publications/statistical/ Premature mortality due to cardiovascular national-child-measurement-programme/2016- disease and social inequalities in Porto Alegre: Gordon-Dseagu V (2006). Cancer and health 17-school-year. from evidence to action. Arq Bras Cardiol. inequalities: an introduction to current evidence. 90(6):370–9. PMID:18592089 Cancer Research UK. Available from: http:// Robertson A, Lobstein T, Knai C (2007). www.cancerresearchuk.org/prod_consump/ Obesity and socio-economic groups in Europe: de Vries E, Arroyave I, Pardo C, Wiesner C, groups/cr_common/@nre/@pol/documents/ evidence review and implications for action. Murillo R, Forman D, et al. (2015). Trends in generalcontent/crukmig_1000ast-3344.pdf. Available from: http://ec.europa.eu/health/ph_ inequalities in premature cancer mortality by determinants/life_style/nutrition/documents/ educational level in Colombia, 1998-2007. J Gwatkin DR (2007). Socio-economic ev20081028_rep_en.pdf. Epidemiol Community Health. 69(5):408–15. differences in health, nutrition and population https://doi.org/10.1136/jech-2014-204650 within developing countries: an overview. Stringhini S, Carmeli C, Jokela M, Avendaño PMID:25492898 Washington (DC), USA: World Bank. M, Muennig P, Guida F, et al.; LIFEPATH consortium (2017). Socioeconomic status Di Cesare M, Khang YH, Asaria P, Blakely Marmot M (2010). Fair society, healthy and the 25 × 25 risk factors as determinants T, Cowan MJ, Farzadfar F, et al.; Lancet lives: the Marmot review. Strategic review of premature mortality: a multicohort study NCD Action Group (2013). Inequalities in of health inequalities in England post-2010. and meta-analysis of 1.7 million men and non-communicable diseases and effective London, UK. Available from: http://www. women. Lancet. 389(10075):1229–37. https:// responses. Lancet. 381(9866):585–97. instituteofhealthequity.org/resources-reports/ doi.org/10.1016/S0140-6736(16)32380-7 https://10.1016/S0140-6736(12)61851-0 fair-society-healthy-lives-the-marmot-review/ PMID:28159391 PMID:23410608 fair-society-healthy-lives-full-report-pdf.pdf.

EC (2013). Health inequalities in the EU – Marmot M; UCL Institute of Health Equity Final report of a consortium. Consortium lead: (2013). Review of social determinants and Sir Michael Marmot. European Commission. the health divide in the WHO European Available from: https://ec.europa.eu/health/ Region. Copenhagen, Denmark: World Health sites/health/files/social_determinants/docs/ Organization Regional Office for Europe. healthinequalitiesineu_2013_en.pdf. Available from: http://www.euro.who.int/ en/publications/abstracts/review-of-social- determinants-and-the-health-divide-in-the- who-european-region.-final-report.