IJE Advance Access originally published online on October 26, 2005
International Journal of Epidemiology 2005 34(6):1181-1182; doi:10.1093/ije/dyi242
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Editorial |
Real epidemiologists don't do ecological studies?
Professor of Clinical Epidemiology, Department of Social Medicine, University of Bristol, Canynge Hall, Whiteladies Road, Bristol BS8 2PR, UK
E-mail: y.ben-shlomo{at}bristol.ac.uk
The papers by Tapia Granados1,2 and the accompanying commentaries3–8 provide a most welcome debate for epidemiologists interested in the role of macro-level socioeconomic factors in determining mortality risk. Leaving aside technical arguments around the statistical methods, this work is challenging and should be of interest to many readers. The observation that mortality actually increases during periods of economic expansion, though not new, is counterintuitive and strikes deep at the heart of social epidemiology that believes improving socioeconomic conditions should be associated with better health. It is therefore unsurprising that this observation is strongly challenged. However, as Tapia Granados,1,2 Edwards,4 and Rhum8 discuss, there are reasons (access to health care, health behaviours, working conditions, road traffic accidents, etc.) why economic upturns may be associated with worse health in the US. The fact that this paradoxical observation is in the opposite direction to the long-term trends in economic growth and life expectancy, and that epidemiological studies have consistently shown that unemployed populations have worse health than those in employment, does not necessarily refute these ecological patterns. I won't dwell on the arguments here as the papers by Tapia Grandos and the accompanying commentaries go into these in depth and make for excellent reading.
It is perhaps not surprising that most of the research in this area has been undertaken by economists, demographers, or social scientists rather than epidemiologists. However studying fluctuations in economic growth should be of great interest to social epidemiologists and public health practitioners. The last decade or so has seen a major revival for both theoretical9–13 and empirical14 uses of contextual or macro-level variables in explaining health variations. This runs parallel to the emphasis on individual risk factors that emerged from the 1950s and more recently has flowered into molecular and genetic epidemiology. The integration of both macro- and micro-level processes and systems within a life course temporal framework is one of the theoretical challenges for social epidemiologists in the 21st century.15 This explicitly recognizes the complexity of disease causation and the importance of societal factors in affecting both biological and social processes. An additional advantage is that some of these macro-level factors allow us to get as close as one can to an experimental paradigm given our inability to undertake randomized controlled studies for most contextual or policy-related variables. For example the implementation of state regulation around seat belt use or speed limits can be examined in relation to motor traffic accidents both in a before/after comparison as well as across areas if not implemented universally.16 Often, however, multiple factors are at play. Analysis of Russian mortality rates between 1987 and 1994 demonstrated a sharp increase, predominantly for alcohol-related deaths. This most likely reflected a combination of factors such as the reversal of Gorbachev's anti-alcohol campaign in 1985, the end in state monopoly on alcohol imports and sales, and a reduction in the unit cost of alcohol.17
It is perhaps unsurprising that epidemiologists generally avoid such analyses as those undertaken by Tapia Granados, though this is in my opinion a great shame. Most epidemiology courses or textbooks spend much time and effort elaborating the technical aspects of cohort or case–control studies but give little time or space to ecological studies and natural experiments18 and may or may not include statistical methods such as time series analysis. All epidemiology students are usually familiar with John Snow's classic study of cholera mortality rates amongst residents receiving piped water either from the Southwark and Vauxhall or Lambeth water company.19 However, they will not usually have encountered the equally classic study by Zena Stein et al.20 on the Dutch Winter Hunger, which used the awful experience of pregnant women during the World War II to examine the potential effects of under-nutrition during the fetal period on childhood growth and development and has more recently been applied to chronic disease aetiology.21
Simple cross-sectional ecological associations are usually dismissed as mere hypothesis-generating exercises, confounded by the ecological fallacy and very weak evidence for causality. However, temporal ecological data or, where one is fortunate, a population-based natural experiment provides far more robust evidence. For example, the effect of daily variations in air pollution on mortality, where individual confounding factors remain constant over time, provides strong evidence for a causal effect.22 Applying such environmental approaches to variations in health care and economic or social policies has been relatively under-utilized though is clearly not without problems. Ecological comparisons of different health care systems remain confounded by other societal factors that influence the nature of health care provision (insurance or national health) and other social welfare policies. Similarly policy changes that occur over time may be confounded by other secular trends. However, by combining secular and geographical variations it may be possible to examine how likely associations are to be explained by confounding factors since different patterns of confounding are likely to be at play over time in the same geographical area and at the same time between different geographical areas for some of these factors. In addition, the analysis of temporal rather than cross-sectional data can allow one to examine for lagged effects rather than simple acute responses through either the effects of early life programming on disease risk or long-term accumulation of adverse exposures through a variety of different life course pathways.23,24 The ultimate in any ‘fantasy epidemiology’ is the dataset that combines both individual and contextual variables over the life course for many different areas or populations with heterogeneous ecosocial factors. This would enable one to examine the potential influence of macro-level variables on both individual level intermediary causal factors as well as interactions between macro-level and individual variables as policy changes may not affect all segments of society equally or all countries in the same way. Whilst such a dataset may currently be in the realms of make-believe, we must clearly make greater use of other ‘naturalistic’ opportunities to answer policy-related questions.
In the international bestseller ‘Freakonomics’, Steven Levitt and Stephen Dubner describe the natural experiment imposed by the Chicago Public School system on school choice, whereby parents who wished to send their child to a high performing school outside their area were randomly selected to get their choice or not, through a lottery-based system, thus allowing researchers to determine whether school choice made a difference on academic performance.25 Sadly for us, policy-makers rarely allocate interventions on a random basis and often do not even pilot major initiatives. However, different macro-level policies within populations over time and between populations should remain fertile ground to try and test aetiological hypotheses both around aetiology as well as evaluating the public health impact of such policies. Perhaps we need more freaky epidemiology?
Acknowledgments
The author would like to thank Shah Ebrahim and Debbie Lawlor for their helpful comments.
References
1 Tapia Granados JA. Increasing mortality during the expansions of the US economy, 1900–1996. Int J Epidemiol 2005;34:1194–202.
2 Tapia Granados JA. Response: On economic growth, business fluctuations, and health progress. Int J Epidemiol 2005;34:1226–33.
3 Catalano R, Bellows B. Commentary: If economic expansion threatens public health, should epidemiologists recommend recession? Int J Epidemiol 2005;34:1212–13.
4 Edwards RD. Commentary: Work, well-being, and a new calling for countercyclical policy. Int J Epidemiol 2005;34:1222–25.
5 Brenner MH. Commentary: Economic growth is the basis of mortality rate decline in the 20th century—experience of the United States 1901–2000. Int J Epidemiol 2005;34:1214–21.
6 McKee M, Suhrcke M. Commentary: Health and economic transition. Int J Epidemiol 2005;34:1203.
7 Neumayer E. Commentary: The economic business cycle and mortality. Int J Epidemiol 2005;34:1221–22.
8 Ruhm CJ. Commentary: Mortality increases during economic upturns. Int J Epidemiol 2005;34:1206–11.
9 Krieger N. Epidemiology and the web of causation: Has anyone seen the spider? Soc Sci Med 1994;39:887–903.[CrossRef][Web of Science][Medline]
10 Krieger N. Theories for social epidemiology in the 21st century: an ecosocial perspective. Int J Epidemiol 2001;30:668–77.
11 Susser M, Susser E. Choosing a future for epidemiology: II. From black box to Chinese Boxes and eco-epidemiology. Am J Public Health 1996;86:674–77.
12 Susser M, Susser E. Choosing a future for epidemiology: 1. Eras and paradigms. Am J Public Health 1996;86:668–73.
13 Diez-Roux AV. Bringing context back into epidemiology: variables and fallacies in multilevel analysis. Am J Public Health 1998;88:216–22.
14 Diez-Roux AV, Nieto FJ, Muntaner C et al. Neighborhood environments and coronary heart disease: a multilevel analysis. Am J Epidemiol 1997;146:48–63.
15 Ben-Shlomo Y, Kuh D. A life course approach to chronic disease epidemiology: conceptual models, empirical challenges and interdisciplinary perspectives. Int J Epidemiol 2002;31:285–93.
16 Grabowski DC, Morrisey MA. The effect of state regulations on motor vehicle fatalities for younger and older drivers: A review and analysis. Milbank Q 2001;79:517–45.[CrossRef][Web of Science][Medline]
17 Leon DA, Chenet L, Shkolnikov VM et al. Huge variation in Russian mortality rates 1984–1994: Artefact, alcohol, or what? Lancet 1997;350:383–88.[CrossRef][Web of Science][Medline]
18 Rothman KJ, Greenland S. Modern Epidemiology. Philadelphia: Lippincott-Raven, 1998.
19 Davey Smith G. Commentary: Behind the broad street pump: aetiology, epidemiology and prevention of cholera in mid-19th century Britain. Int J Epidemiol 2002;31:920–32.
20 Stein Z, Susser M, Saenger G, Marolla F. Famine and Human Development. The Dutch Hunger Winter of 1944–1945. London: Oxford University Press, 1975.
21 Ravelli ACJ, van der Meulen JHP, Michels RPJ et al. Glucose tolerance in adults after prenatal exposure to famine. Lancet 1998;351:173–77.[CrossRef][Web of Science][Medline]
22 Schwartz J. Air pollution and daily mortality: a review and meta analysis. Environ Res 1994;64:36–52.[Medline]
23 Ben-Shlomo Y, Davey Smith G. Deprivation in infancy or in adult life: which is more important for mortality risk? Lancet 1991;337:530–35.[CrossRef][Web of Science][Medline]
24 Kuh D, Ben-Shlomo Y, Lynch J, Hallqvist J, Power C. Life course epidemiology. J Epidemiol Community Health 2003;57:778–83.
25 Levitt SD, Dubner SJ. Freakonomics—A Rogue Economist Explores the Hidden Side of Everything. Penguin Group, 2005.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  R. Aquino, N. F. de Oliveira, and M. L. Barreto Impact of the Family Health Program on Infant Mortality in Brazilian Municipalities Am J Public Health, January 1, 2009; 99(1): 87 - 93. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I A Jacobs, M T Podobny, and D Bilusich One mechanism underlying contrasting health-economy findings Int. J. Epidemiol., August 1, 2007; 36(4): 929 - 931. [Full Text] [PDF]
|
|
|
|
|
|
D. March and E. Susser The eco- in eco-epidemiology Int. J. Epidemiol., December 1, 2006; 35(6): 1379 - 1383. [Full Text] [PDF]
|
|
|
|
|
|
G. DAVEY SMITH Cultural climate, physical climate, life, and death Int. J. Epidemiol., April 1, 2006; 35(2): 211 - 212. [Full Text] [PDF]
|
|
|
|
|
|
G. D. Smith Equal, but different? Ecological, individual and instrumental approaches to understanding determinants of health Int. J. Epidemiol., December 1, 2005; 34(6): 1179 - 1180. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||