Is Cardiovascular Disease Slowing Improvements in Life Expectancy?

6. Contributors to CVD mortality and policy options for improving CVD health

Professor Martin O’Flaherty

Abstract

The urgency for tackling cardiovascular disease is evident as the long-standing decline in mortality could be at risk. Policies at the population level can deliver rapid, large and equitable health and economic gains, with large returns on investments. The policy areas include improving food policy, reducing alcohol intake, smoking and air pollution. Prevention policy should aspire to achieve three main goals: reduce the cardiovascular disease burden, reduce the equity gap and reduce stress in the health care system to make it economically sustainable. Reducing the unequal burden of cardiovascular disease is likely to require a combination of targeted policies in deprived communities alongside structural policies to improve diets, smoking and alcohol intake.

Western countries experienced a unique epidemiological phenomenon, with massive reductions in ischemic heart disease mortality of more than 60% over four decades in the second half of the 20th century (Moran et al., 2014[1]). Some countries – particularly Central European countries – have shown dramatic declines after increases that lasted until the 1990s. However, many countries are currently experiencing rising trends, including China and Mexico.

Trends are not set in stone. They can vary substantially over relatively short time scales. The recent slowing down in reduction in CVD mortality reported in this report points towards an undesirable change in the direction of the drivers of CVD mortality. Thus, understanding what drives those trends continues to be relevant and urgent.

Cardiovascular disease (CVD) is eminently preventable, and modifiable risk factors can explain 90% of its incidence (Moran et al., 2014[1]). Changes in risk factors at the population level and treatments are the two main drivers of CVD, as was clearly shown for coronary heart disease (CHD) by observational studies like MONICA and modelling studies.

Observational and modelling evidence on what drives CHD mortality

During the 1980s and 1990s, the MONICA study, using carefully designed and detailed protocols and methods, was able to explore more precisely the contribution of risk factors and treatments to mortality trends in over 20 diverse populations (Tunstall-Pedoe et al., 2000[2]; Kuulasmaa et al., 2000[3]). The MONICA project’s goals were to measure trends in CVD mortality, and CHD and cerebral-vascular disease morbidity, and to assess the extent to which these trends are related to changes in known risk factors, health care, and significant socio-economic characteristics measured simultaneously in defined communities in different countries. Its principal findings were that about one‑third of the change in CHD mortality rates could be attributed to health care and two‑thirds to changes in risk factors (Tunstall-Pedoe et al., 2000[2]; Tunstall-Pedoe et al., 1999[4]).

Another approach to answering the question is by using modelling approaches. Epidemiological models are a way to synthesise demographics, risk factors and treatment evidence to provide a quantitative summary of the contributions of these drivers to changes in incidence or case fatality. Two models have been designed to answer this question: the US CHD policy model and the IMPACT model.

The US CHD Policy model is a state-transition model developed in the 1980s. It was initially used to examine trends in CHD mortality (Goldman et al., 2001[5]; Hunink et al., 1997[6]) and expected gains in life expectancy from risk factor modifications (Tsevat et al., 1991[7]). This model was also used to evaluate the cost-effectiveness of specific medical interventions for primary and secondary prevention of CHD (Gaspoz et al., 2002[8]; Phillips et al., 2000[9]; Prosser et al., 2000[10]), salt reduction policies (Bibbins-Domingo et al., 2010[11]) and health promotion activities (Tosteson et al., 1997[12]). The model showed that in the US population between 1980‑1990 risk factor changes contributed 50% to the mortality decline, while treatments contributed 43%.

The IMPACT model has been used in more than 20 countries globally to explore the proportion of the change in deaths in terms of contributions of risk factors and evidence-based treatments. IMPACT is a spreadsheet model initially developed by Capewell and colleagues in 2000 (Capewell et al., 2000[13]). This model combines data sources on patient numbers, treatment uptake, treatment effectiveness, risk factor trends and consequent mortality effects. The deaths prevented or postponed over a specified period are then estimated (Unal, Capewell and Critchley, 2006[14]). The model can be used to estimate the proportion of change in mortality attributable to specific treatments or risk factor changes. It can also estimate the future consequences of altering treatment strategies and changing population risk. The model also estimates life-years gained and cost-effectiveness for specific interventions.

Figure 6.1. Schematic representation of the IMPACT model

Source: Image courtesy of Professors Capewell and O’Flaherty.

To estimate the contribution of medical and surgical treatments to the reductions in CHD mortality, the model integrates information on the number of patients eligible for a specific treatment, the case fatality rate of that group of patients, the relative risk reduction offered by the treatment, and the uptake of the treatment amongst those patients.

Box 6.1. IMPACT CHD methodology examples

Example 1. Estimating the contribution of evidence-based treatments

Men aged 55‑64 given aspirin for acute myocardial infarction: In the Antithrombotic Trialists’ Collaboration meta-analysis, aspirin reduced relative mortality in men with acute myocardial infarction by 15%. In England and Wales in 2000, 10 699 men aged 55‑64 were eligible with a case fatality rate of 17%, and 95% were given aspirin. One year case fatality in men aged 55‑64 admitted with an acute myocardial infarction was approximately 17%. The deaths prevented or postponed for at least a year were therefore calculated as: Patient numbers x treatment uptake x relative mortality reduction x one‑year case fatality = 10 699 x 95% x 15% X 17% = 259 deaths prevented or postponed.

Example 2. Estimating the contribution of risk factors

In England and Wales, the diabetes prevalence in men aged 75‑84 was 4% in 1981 and 7% in 2000, obtained from the Health Survey for England. Using a population attributable risk fraction approach (PARF), we estimated that about 12% respectively of CHD deaths were attributable to diabetes in 1981 and 18% in 2000. The number of actual deaths attributed to diabetes was then calculated by multiplying the PARF times the deaths observed in 1981 and expected in 2000: 2 865 in 1981 and 3 916 in 2000. The difference between these (1 051) represented the change in the number of deaths attributable to the change in diabetes prevalence in the population between 1981 and 2000.

The IMPACT model consistently found that about 40 to 72% of the fall in deaths was attributable to risk factor changes, and 23 to 55% to treatments (Mensah et al., 2017[15]). Particularly powerful drivers were population-wide declines in smoking, blood pressure and cholesterol levels, and acute care and secondary prevention, including heart failure treatments. In countries where chronic heart failure rates were increasing, adverse population-level trends in smoking, cholesterol and blood pressure drove mortality upwards – as in Beijing or Tunisia (Critchley et al., 2004[16]; Critchley et al., 2016[17]).

The models for OECD countries show the patterns observed in western countries. Interestingly, although different modelling approaches have been used in some countries, these approaches found similar insights, strengthening our confidence in the knowledge of what is driving changes in CVD mortality.

Figure 6.2. The proportion of CHD deaths explained and postponed by changes in risk factors and treatments using IMPACT and other models: Selected OECD countries

Note: All results relate to IMPACT model except Finland 1982‑97 (Vartianineen), New Zealand 1978‑1981 (Beaglehole), United States 1968‑1976 (Goldman), United States 1980‑1990 (Hunink), Holland 1978‑1985 (Bots).

Source: Image courtesy of Professors Capewell and O’Flaherty.

Risk factors, particularly changing at the population level, seem to be an essential driver. Importantly, risk factors can react rapidly to changes in their determinants (Capewell and O’Flaherty, 2011[18]). An interesting illustration is Poland in the 1990s, after the socio-economic transformation. The halving of CHD mortality rates in Poland was driven by risk factors, explaining about 54% of the decline and 37% by the increased use of evidence-based treatments (Bandosz et al., 2012[19]). These massive, population-level changes were possibly linked to widespread changes in Polish diets at the population level (Zatonski and Willett, 2005[20]). Furthermore, in populations with rising trends (e.g. China and Mexico), adverse risk factors trends are likely explanations for most of the increase in mortality.

One of the most constant findings in most populations studied with IMPACT is that the almost universally observed increases in obesity and diabetes offset a significant proportion (10‑14%) of the mortality reductions attributed to favourable changes in blood pressure, tobacco smoking, cholesterol levels and physical activity.

Tackling the drivers of CVD mortality in populations

We can only speculate on the causes of the current slowdown in CVD mortality. It is likely that the factors are mostly related to changes in population drivers of incidence including diet, smoking, physical activity and alcohol and increasing trends in obesity and diabetes, rather than worsening of case-fatality rates. The OECD recently reported improvements in acute case-fatality rates in several countries that are experiencing the slowdown (OECD, 2019[21]), although less is known about case-fatality rates for later complications of ischemic heart disease, like heart failure patients living in the community or people living long-term with stroke and other forms of vascular disease.

The urgency to tackle CVD is evident as the long-standing decline in mortality could be at risk. We are now facing a slowdown in mortality improvements in several countries and a reversal of mortality decline in the United States, with CVD mortality slowdowns playing a significant role (Sidney et al., 2016[22]; Public Health England, 2018[23]). Furthermore, this is associated with persistent and continuing disparities and, in the United Kingdom, the slowdown in mortality improvements overall is most marked in more deprived populations (Public Health England, 2019[24]).

Policies at the population level can deliver the gains that are needed. They essentially affect the environment in which we live, can be shaped by policy tools such as regulations, trade, legal and fiscal policies and, crucially, can deliver rapid, large and equitable health and economic gains, with large returns on investments (Masters et al., 2017[25]; Capewell and O’Flaherty, 2011[18]). The policy areas include improving food policy, for example in relation to salt and sugar reduction, and reducing alcohol intake, smoking and air pollution. All of this will have a substantial impact on CVD, and will also favourably affect diseases caused by the same risk factors, resulting in cascade effects in reducing the overall burden of non-communicable diseases (NCDs).

However, not all this policy effect will be instantaneous. While CVD trends are likely to respond quickly, longer lag times for other diseases will mean that it will take longer to realise the overall health gains that we can expect. For example, tackling obesity to reduce diabetes prevalence will result in a reduction of dementia prevalence over the medium and long term, while CVD will be reduced faster (Bandosz et al., 2020[26]).This will then create additional pressures on the health care system and society that will continue to respond while the full impact of prevention across the board unfolds. Crucially, the recent slowdown in CVD mortality will compound the problem, resulting in about a 15% increase in health care costs, with social care costs increasing twice as fast (Collins et al., 2019[27]).

To recover the lost ground, we need a rethink of prevention. The drivers are likely to be the same as in the earliest stages of the CVD epidemic. We need to leverage this knowledge to reduce both the overall burden and the inequalities. Prevention policy goals should aspire to achieve three main goals: reduce the CVD burden, reduce the equity gap and reduce stress in the health care system to make it economically sustainable.

Reducing the unequal burden of CVD is likely to require a combination of targeted policies in deprived communities alongside structural policies to improve diets, smoking and alcohol intake. For example, there is evidence from the United States that the most substantial health gains and disparities reduction can be achieved by making fruit and vegetables more affordable for Supplemental Nutrition Assistance Program (SNAP) participants, alongside a fiscal policy to tax sugar-sweetened beverages for the whole population, resulting in substantial numbers of cases prevented (Pearson-Stuttard et al., 2017[28]). Similar insights were gained in the United Kingdom, combining approaches to identify and manage high-risk individuals alongside population-level strategies on smoking and food; for example, the recently reported reduction in sugar intake after the successful implementation of the sugary drinks tax, resulting in a reduction of about a third of the volume of sugars from soft drinks per capita and per day, equivalent to a reduction of sugar intake of more than 4 grams a day (Kypridemos et al., 2018[29]; Kypridemos et al., 2016[30]; Bandy et al., 2020[31]). Furthermore, this will result in substantial reductions in demands faced by the health care system.

As many population-level prevention strategies do not rely on health care budgets, a move towards structural, population-level prevention might free resources to invest in other pressing areas, like the interface of health and social care, reducing inequalities, and tackling the pressures on health systems exerted by population ageing. Thus, an urgent research priority should be to help policy-makers decide which combination of strategies over time can offer the best overall approach for tackling the challenge of CVD in our populations.

Inaction in addressing CVD prevention will have profound social, economic and equity costs, costs that are avoidable with our current knowledge about the prevention of CVD and NCDs.

Acknowledgements: the IMPACT CHD model, developed by Professor Capewell at the University of Liverpool, is the result of the contributions from researchers from many countries around the globe. My sincere thanks to all our collaborators and colleagues.

References

[26] Bandosz, P. et al. (2020), “Potential impact of diabetes prevention on mortality and future burden of dementia and disability: a modelling study”, Diabetologia, Vol. 63/1, pp. 104-115, http://dx.doi.org/10.1007/s00125-019-05015-4.

[19] Bandosz, P. et al. (2012), “Decline in mortality from coronary heart disease in Poland after socioeconomic transformation: Modelling study”, BMJ (Online), Vol. 344/7842, http://dx.doi.org/10.1136/bmj.d8136.

[31] Bandy, L. et al. (2020), “Reductions in sugar sales from soft drinks in the UK from 2015 to 2018”, BMC Medicine, Vol. 18/1, p. 20, http://dx.doi.org/10.1186/s12916-019-1477-4.

[11] Bibbins-Domingo, K. et al. (2010), “Projected effect of dietary salt reductions on future cardiovascular disease”, New England Journal of Medicine, Vol. 362/7, pp. 590-599, http://dx.doi.org/10.1056/NEJMoa0907355.

[13] Capewell, S. et al. (2000), “Explanation for the decline in coronary heart disease mortality rates in Auckland, New Zealand, between 1982 and 1993.”, Circulation, Vol. 102/13, pp. 1511-6, http://dx.doi.org/10.1161/01.cir.102.13.1511.

[18] Capewell, S. and M. O’Flaherty (2011), “Rapid mortality falls after risk-factor changes in populations”, The Lancet, Vol. 378/9793, pp. 752-753, http://dx.doi.org/10.1016/s0140-6736(10)62302-1.

[27] Collins, B. et al. (2019), “OP19 Will social care need more resources? A modelling study of health and social costs in england and wales for alternative future cardiovascular disease scenarios”, J Epidemiol Community Health, Vol. 73/Suppl 1, pp. A9-A9, http://dx.doi.org/10.1136/jech-2019-ssmabstracts.19.

[17] Critchley, J. et al. (2016), “Contrasting cardiovascular mortality trends in Eastern Mediterranean populations: Contributions from risk factor changes and treatments”, International Journal of Cardiology, Vol. 208, pp. 150-161, http://dx.doi.org/10.1016/j.ijcard.2016.01.031.

[16] Critchley, J. et al. (2004), “Explaining the increase in coronary heart disease mortality in Beijing between 1984 and 1999”, Circulation, Vol. 110/10, pp. 1236-1244, http://dx.doi.org/10.1161/01.CIR.0000140668.91896.AE.

[8] Gaspoz, J. et al. (2002), “Cost effectiveness of aspirin, clopidogrel, or both for secondary prevention of coronary heart disease”, New England Journal of Medicine, Vol. 346/23, pp. 1800-1806, http://dx.doi.org/10.1056/NEJM200206063462309.

[5] Goldman, L. et al. (2001), “The effect of risk factor reductions between 1981 and 1990 on coronary heart disease incidence, prevalence, mortality and cost”, Journal of the American College of Cardiology, Vol. 38/4, pp. 1012-1017, http://dx.doi.org/10.1016/S0735-1097(01)01512-1.

[6] Hunink, M. et al. (1997), “The recent decline in mortality from coronary heart disease, 1980-1990; The effect of secular trends in risk factors and treatment”, Journal of the American Medical Association, Vol. 277/7, pp. 535-542, http://dx.doi.org/10.1001/jama.277.7.535.

[3] Kuulasmaa, K. et al. (2000), “Estimation of contribution of changes in classic risk factors to trends in coronary-event rates across the WHO MONICA Project populations”, Lancet, Vol. 355/9205, pp. 675-687, http://dx.doi.org/10.1016/S0140-6736(99)11180-2.

[30] Kypridemos, C. et al. (2016), “Cardiovascular screening to reduce the burden from cardiovascular disease: Microsimulation study to quantify policy options”, BMJ (Online), Vol. 353, http://dx.doi.org/10.1136/bmj.i2793.

[29] Kypridemos, C. et al. (2018), “Future cost-effectiveness and equity of the NHS Health Check cardiovascular disease prevention programme: Microsimulation modelling using data from”, PLoS Med, Vol. 15/5, https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5973555/ (accessed on 6 February 2020).

[25] Masters, R. et al. (2017), Return on investment of public health interventions: A systematic review, BMJ Publishing Group, http://dx.doi.org/10.1136/jech-2016-208141.

[15] Mensah, G. et al. (2017), “Decline in Cardiovascular Mortality: Possible Causes and Implications”, Circulation Research, Vol. 120/2, pp. 366-380, http://dx.doi.org/10.1161/CIRCRESAHA.116.309115.

[1] Moran, A. et al. (2014), Temporal trends in ischemic heart disease mortality in 21 world regions, 1980 to 2010: The global burden of disease 2010 study, Lippincott Williams and Wilkins, http://dx.doi.org/10.1161/CIRCULATIONAHA.113.004042.

[21] OECD (2019), Health at a Glance 2019: OECD Indicators, OECD Publishing, Paris, https://dx.doi.org/10.1787/4dd50c09-en.

[28] Pearson-Stuttard, J. et al. (2017), “Reducing US cardiovascular disease burden and disparities through national and targeted dietary policies: A modelling study”, PLoS Medicine, Vol. 14/6, http://dx.doi.org/10.1371/journal.pmed.1002311.

[9] Phillips, K. et al. (2000), “Health and economic benefits of increased β-blocker use following myocardial infarction”, Journal of the American Medical Association, Vol. 284/21, pp. 2748-2754, http://dx.doi.org/10.1001/jama.284.21.2748.

[10] Prosser, L. et al. (2000), “Cost-effectiveness of cholesterol-lowering therapies according to selected patient characteristics”, Annals of Internal Medicine, Vol. 132/10, pp. 769-779, http://dx.doi.org/10.7326/0003-4819-132-10-200005160-00002.

[24] Public Health England (2019), “Public health matters”, Health Matters: Ambitions to tackle persisting inequalities in cardiovascular disease - Public health matters, https://publichealthmatters.blog.gov.uk/2019/03/04/health-matters-ambitions-to-tackle-persisting-inequalities-in-cardiovascular-disease/ (accessed on 6 February 2020).

[23] Public Health England (2018), A review of recent trends in mortality in England, http://www.facebook.com/PublicHealthEngland (accessed on 6 February 2020).

[22] Sidney, S. et al. (2016), “Recent trends in cardiovascular mortality in the United States and public health goals”, JAMA Cardiology, Vol. 1/5, pp. 594-599, http://dx.doi.org/10.1001/jamacardio.2016.1326.

[12] Tosteson, A. et al. (1997), “Cost-effectiveness of populationwide educational approaches to reduce serum cholesterol levels”, Circulation, Vol. 95/1, pp. 24-30, http://dx.doi.org/10.1161/01.CIR.95.1.24.

[7] Tsevat, J. et al. (1991), “Expected gains in life expectancy from various coronary heart disease risk factor modifications.”, Circulation, Vol. 83/4, pp. 1194-201, http://dx.doi.org/10.1161/01.cir.83.4.1194.

[4] Tunstall-Pedoe, H. et al. (1999), “Contribution of trends in survival and coronary-event rates to changes in coronary heart disease mortality: 10-year results from 37 WHO MONICA Project populations”, Lancet, Vol. 353/9164, pp. 1547-1557, http://dx.doi.org/10.1016/S0140-6736(99)04021-0.

[2] Tunstall-Pedoe, H. et al. (2000), “Estimation of contribution of changes in coronary care to improving survival, event rates, and coronary heart disease mortality across the WHO MONICA Project populations”, Lancet, Vol. 355/9205, pp. 688-700, http://dx.doi.org/10.1016/S0140-6736(99)11181-4.

[14] Unal, B., S. Capewell and J. Critchley (2006), “Coronary heart disease policy models: a systematic review”, BMC Public Health, Vol. 6/1, p. 213, http://dx.doi.org/10.1186/1471-2458-6-213.

[20] Zatonski, W. and W. Willett (2005), “Changes in dietary fat and declining coronary heart disease in Poland: Population based study”, British Medical Journal, Vol. 331/7510, pp. 187-188, http://dx.doi.org/10.1136/bmj.331.7510.187.