From the Department of Epidemiology, Mailman School of Public Health, New York, New York.
Correspondence: Ezra Susser, Department of Epidemiology, Mailman School of Public Health, 722 W. 168th St., New York, NY 10032. E-mail: firstname.lastname@example.org
In the ongoing debate over the value of risk-factor (or “black-box”) epidemiology, proponents have put forward some compelling arguments. The identification of risk factors is vital for public health research; epidemiologists should strive to achieve ever more valid and precise studies of risk factors; and, as Sander Greenland1 shows, an accumulation of initially inexplicable risk factor findings can ultimately provide the basis for a deeper understanding of causes, as well as point the way toward improvements in public health. Thus, the continuing refinement of methods for identifying risk factors is of undoubted value.
At the same time, limiting epidemiology to the identification of risk factors needlessly restricts the scope of our discipline. Current developments call for epidemiologists to adopt a wider and richer framework—a framework that subsumes but does not discard risk-factor epidemiology. One can see epidemiology changing along four lines in parallel.
First, epidemiologists increasingly consider multiple levels of causation. The causes of health outcomes in populations can be sought at macrolevels, like in the distribution of wealth across and within societies; at individual levels, like in individual behavior, and at microlevels, like in the expression of genes in cells. Although a disease ultimately manifests in an afflicted person, its prevention could be directed toward any of these levels. Different approaches are needed to identify the causal processes at each level,2,3 and, equally important, to examine how a process at one level (eg, societal cohesion or cellular changes) can manifest at another level (eg, disease in an individual). Across historical periods, epidemiologists have shifted their focus among these levels of causation.4 However, thinking about multiple levels of causation also fits squarely within epidemiologic traditions. Intimations of this can be seen even in the classic works on cholera in 19th century England featured in this issue of the journal. John Snow and William Budd both entertained ideas about causes operating at many levels.5,6 They recognized that a favorable context for epidemic spread of cholera followed from the social changes wrought by industrialization and urbanization; identified the privately owned and profitable water supply as the primary mode of cholera transmission within this context; noted the influence of household and individual habits on disease transmission; sought to understand its pathophysiology; and speculated on microorganisms as a primary source.
Second, as a central science of public health, epidemiologists are strategically well situated now to investigate the interplay between genetic and environmental factors. Epidemiologists need to comprehend all the causes of population health that could have implications for preventive action. To investigate genetic and nongenetic factors together, our epidemiologic research designs must adapt to incorporate and exploit gene biology. In a simple scenario, a measurable genetic factor could increase the risk of a disorder only in the presence of a specified environmental exposure. The authors of a recent report,7 for example, proposed that the allelic status of the gene encoding the serotonin transporter influences the risk of depression only after exposure to stressful life events.
Third, epidemiologists increasingly recognize the enrichment that follows from examining the trajectory of health and illness over the life course.8,9 The onset of disease in an adult could reflect earlier life experiences in several ways. These include long-deferred effects of an in utero insult; a chain of social and biologic experiences initiated by an early childhood exposure; and cumulative exposure to toxic biologic or social factors over a long period. These alternative trajectories each require different approaches for their investigation.
Fourth, at least some epidemiologists are engaged in expanding their reach and reversing the longstanding trend toward fragmentation into subspecialties. A broader and more unified framework could both sustain the coherence of the discipline and empower each domain by affording it new designs crafted in other domains. For at least half a century, for example, infectious disease and chronic disease epidemiology proceeded largely in parallel. Investigators are now demonstrating that methods developed for infectious diseases—such as modeling transmission of epidemics across and within populations—are relevant for noninfectious diseases and behaviors.
The task before us is to draw together these different domains and then extract the implications for public health. An integrated approach to investigating disease and its prevention will necessarily subsume levels of causation, life course trajectories, kinds of causes, and types of diseases. To describe this emergent way of thinking, we have proposed the term “eco-epidemiology.”10 Calls for integrating risk factor epidemiology within a broader framework have also been articulated by others (eg, Krieger11 and McMichael12). Whether or not the label eco-epidemiology proves durable, the endeavor toward integration is essential.
Finally, we propose that epidemiology will thrive from active and dynamic engagement with many branches of public health and clinical medicine. Probably desirable in any epoch, this collaboration is an absolute necessity in the present time as we face the immediate global health crisis of the HIV/AIDS pandemic.
ABOUT THE AUTHOR
EZRA SUSSER began his career in psychiatry focused on the homeless mentally ill outside the purview of traditional care. Over the years, this led him into psychiatric epidemiology, urban epidemiology, and international health. Not coincidentally, he is the son of epidemiologists Zena Stein and Mervyn Susser (both featured in the Journal's VOICES series).
I would like to acknowledge the contributions of Mervyn Susser, Sharon Schwartz, and Michaeline Bresnahan to the ideas expressed in this piece, and also thank them for helpful comments.
1. Greenland S, Gago-Dominguez M, Castelao JE. The value of risk-factor (“black-box”) epidemiology. Epidemiology. 2004;15:529–535.
2. Rose G. The Strategy of Preventive Medicine. New York: Oxford University Press; 1992.
3. Schwartz S, Diez-Roux AV. Commentary: causes of incidence and causes of cases—a Durkheimian perspective on Rose. Int J Epidemiol. 2001;30:435–439.
4. Susser M, Susser E. Choosing a future for epidemiology: I. Eras and paradigms. Am J Public Health. 1996;86:668–673.
5. Smith GD. Commentary: behind the Broad Street pump: aetiology, epidemiology and prevention of cholera in mid-19th century Britain. Int J Epidemiol. 2002;31:920–932.
6. Vinten-Johansen P, Brody H, Paneth N, et al. Cholera, Chloroform and the Science of Medicine: A Life of John Snow. New York: Oxford University Press; 2003.
7. Caspi A, Sugden K, Moffitt TE, et al. Influence of life stress on depression: moderation by a polymorphism in the 5-HTT gene. Science. 2003;301:386–389.
8. Kuh D, Ben-Schlomo Y, eds. A Life Course Approach to Chronic Disease Epidemiology. New York: Oxford University Press; 2004.
9. Susser E, Terry MB. A conception-to-death cohort. Lancet. 2003;361:797–798.
10. Susser M, Susser E. Choosing a future for epidemiology: II. From black box to Chinese boxes and eco-epidemiology. Am J Public Health. 1996;86:674–677.
11. Krieger N. Epidemiology and the web of causation: has anyone seen the spider? Soc Sci Med. 1994;39:887–903.
12. McMichael AJ. Prisoners of the proximate: loosening the constraints on epidemiology in an age of change. Am J Epidemiol. 1999;149:887–897.
Next to Godliness
“Nothing assists the communication of disease more than the want of personal cleanliness.” - JOHN SNOW
© 2004 Lippincott Williams & Wilkins, Inc.