Daniel Kim - New Horizons in Modeling and Simulation for Social Epidemiology and Public Health

Using nationally‐representative data, Kim (2016) estimated the impacts of state and local spending on welfare and education on the risks of dying from major causes. Each additional $250 per capita spent on welfare predicted a 3‐percentage point lower probability of dying from any cause, and each additional $250 per capita spent on welfare and education predicted a 1.6‐percentage point lower probability and a nearly 1‐percentage point lower probability of dying from coronary heart disease (CHD). To put such numbers into context, these changes are on the order of reductions achieved through treating a patient with high blood pressure or cholesterol—representing clinically meaningful changes (Kim 2016).

In a cross‐national study that implemented IV analysis to enhance causal inference, Kim et al. (2011) further estimated the population health impacts of raising social capital across 40 countries. Among those aged 15–74 years in 40 nations with at least 40% of the country trusting of others, raising country percentages of social trust by 20 percentage points in countries with at least 30% of a country’s citizens trusting of others and by 10 percentage points in countries with 30–40% average country trust was predicted to avert nearly 287 000 deaths per year.

Finally, Kondo et al. (2009) conducted a meta‐analysis of cohort studies including roughly 60 million participants in which people living in regions with high‐income inequality had an excess risk for premature mortality independent of their SES, age, and sex. The estimated excess mortality risk was 8% for each 0.05 unit increase in the Gini coefficient (a common measure of income inequality theoretically ranging from 0, representing perfect equality, to 1, corresponding to perfect inequality). While this excess risk appears modest, all of society is exposed to income inequality, such that the aggregate effects can be significant (Kondo et al. 2009). The authors estimated that if the inequality–mortality relation is truly causal, more than 1.5 million deaths (9.6% of total adult mortality in the 15–60 age group) could be averted in 30 OECD countries by reducing the Gini coefficient to below the threshold value of 0.3 (Kondo et al. 2009).

Notably, according to Figure 1.2, there should also be substantial impacts of intervening on the social determinants of health on health inequities across population groups, as defined along social axes such as gender, race/ethnicity, and SES. For example, government spending on public assistance programs (e.g. Aid to Families with Dependent Children) and tax credit programs (e.g. the Earned Income Tax Credit) should reduce income disparities between the rich and the poor and thereby reduce associated gaps in health, since income is a strong determinant of health and disease.

As the U.S. National Academy of Sciences panel concluded in its report, if the United States fails to address its growing health disadvantage in the near future, it will lag even further behind comparable countries in life expectancy and across a wide range of other population health outcomes. By adversely affecting the productivity of the workforce through worse population health, the economy of the United States would also continue to suffer, whereas other countries would continue to reap the economic benefits of having healthier populations. Because of how much is at stake, the panel concluded that it would hence be at the United States' peril that it continue to ignore its growing health disadvantage (National Research Council and Committee on Population 2013). Meanwhile, other countries will still need to maintain their efforts on addressing the social determinants of health if they wish to sustain and/or improve their relative standings in the Health Olympics.

Overall, the findings summarized in this chapter make a strong case for intervening at the policy level on social determinants to improve population health and reduce population health inequities. It is also clear that much more empirical evidence is needed if we wish to establish the population health impacts of the social determinants of health. These evidence gaps include estimates of the effects of social determinants of health on the incidence of diseases and on morbidity outcomes such as DALYs; the estimated population‐wide health impacts of intervening on the social determinants of health through scaled‐up interventions and policies; and economic evaluations (e.g. cost‐effectiveness) of such interventions.

In the next chapter, we move beyond traditional analytic approaches to provide a rationale for the use of systems science methods. In particular, we introduce two major sets of analytical tools for modeling and simulating impacts of the social determinants of health: agent‐based modeling and microsimulation models. These two novel system science tools and their growing applications in social epidemiology and public health form the primary substance of this book.

1 Adema, W., Fron, P., and Ladaique, M. (2011). Is the European welfare state really more expensive?: indicators on social spending, 1980–2012; and a manual to the OECD social expenditure database (SOCX). OECD Social, Employment and Migration working papers, No. 124, OECD Publishing.

2 Bambra, C., Gibson, M., Amanda, S. et al. (2010). Tackling the wider social determinants of health and health inequalities: evidence from systematic reviews. Journal of Epidemiology and Community Health 64: 284–291.

3 Bezo, B., Maggi, S., and Roberts, W.L. (2012). The rights and freedoms gradient of health: evidence from a cross‐national study. Frontiers in Psychology 3: 441.

4 Bostic, R.W., Thornton, R.L.J., Rudd, E.C., and Sternthal, M.J. (2012). Health in all policies: the role of the US Department of Housing and Urban Development and present and future challenges. Health Affairs 31: 2130–2137.

5 Braveman, P., Egerter, S., and Williams, D.R. (2011). The social determinants of health: coming of age. Annual Review of Public Health 32: 381–398.

6 Centers for Disease Control and Prevention (CDC) (2008). Smoking‐attributable mortality, years of potential life lost, and productivity losses‐United States, 2000–2004. Morbidity and Mortality Weekly Report 57: 1226–1228.

7 Cepeda, M.S., Boston, R., Farrar, J.T., and Strom, B.L. (2003). Comparison of logistic regression versus propensity score when the number of events is low and there are multiple confounders. American Journal of Epidemiology 158 (3): 280–287.

8 Charter, O. (1986, November). Ottawa Charter for health promotion. In First International Conference on Health Promotion, pp. 17–21.

9 Chung, H. and Muntaner, C. (2006). Political and welfare state determinants of infant and child health indicators: an analysis of wealthy countries. Social Science and Medicine 63: 829–842.

10 Davey Smith, G., Sterne, J.A.C., Fraser, A. et al. (2009). The associatiodczcn between BMI and mortality using offspring BMI as an indicator of own BMI: large intergenerational mortality study. British Medical Journal 339: b5043.

11 Declaration of Alma‐Ata. (1978). Pan American Health Organization. https://www.paho.org/English/DD/PIN/alma‐ata_declaration.htm(accessed 1 July 2019).

12 Delany, T., Lawless, A., Baum, F. et al. (2015). Health in all policies in South Australia: what has supported early implementation? Health Promotion International 31 (4): 888–898.

13 Egger, M., Smith, G.D., and Sterne, J.A. (2001). Uses and abuses of meta‐analysis. Clinical Medicine 1 (6): 478–484.

14 Fish, J.S., Ettner, S., Ang, A. et al. (2010). Association of perceived neighborhood safety on body mass index. American Journal of Public Health 100: 2296–2303.

15 Fowler, K.A., Dahlberg, L.L., Haileyesus, T., and Annest, J.L. (2015). Firearm injuries in the United States. Preventive Medicine 79: 5–14.