International Journal of Epidemiology
©International Epidemiological Association 1990
Vol. 19, No. 1
Printed in Great Britain
LEADING ARTICLE
Is Molecular Epidemiology a Germ
Theory for the End of the Twentieth
Century?
DANA LOOMIS AND STEVE WING
A recent article in this journal by Vandenbroucke1
drew a parallel between the possible intellectual fate of
some current ideas in epidemiology and the last century's miasma theory of epidemics. Like some nineteenth century investigators, many contemporary
epidemiologists are indeed guided by the principle that
the origins of disease are environmental in a very broad
sense. This is sometimes called the 'black box' view of
epidemiology,2 and it emphasizes the documentation
of exposure-disease associations in conjunction with
the idea that disease can be prevented by altering the
environment without the need of detailed knowledge
of pathogenic mechanisms. Vandenbroucke's thesis is
that epidemiologists who subscribe to this view, like
the miasmists, their philosophical predecessors, are
destined to be the losers in a struggle with a superior
scientific idea. For the miasmists that competing idea
was the germ theory of disease; for the epidemiologists
of today, in his view, it is molecular biology.
Such exhortations that epidemiology must alter its
philosophy and methods to survive as a legitimate
scientific discipline are not unusual3 but they are often
based on a distorted view of the history and goals of
epidemiology. This view is derived from a misinterpretation of the last century's intellectual contest
between theories about the origins of mass disease.
The conventional wisdom of today is that those nineteenth-century epidemiologists who subscribed to the
theory that epidemics are completely explainable by
the presence of specific agents in all affected populations won that contest because they were right, that is,
because such agents are, by definition, the unique
causes of disease. The virtue of their ideas is supposDepartment of Epidemiology, School of Public Health, University of
North Carolina, Chapel Hill, NC, USA.
edly proven by the subsequent conquest of infectious
disease by scientific medicine and the downfall of competing causal theories.1'3
The achievements of scientific medicine in this century obscure other aspects of these historical events
however. The nineteenth century contest between the
contagion and miasma theories of the origin of epidemics is often held up as a singular turning point in the
history of medical thought, but that contest grew from
a tension as old as western medicine between the
heroic capabilities of the healer and the often unglamorous work of the hygienist.4 The same tension continued even after the germ theory of disease was widely
adopted, as the concept of specific aetiology and ideas
emphasizing the role of environment altenatively
gained and lost favour throughout this century.5 The
fact that the debate continues today in the competition
for grant money between high-technology molecular
investigations and classical exposure-disease studies
indicates that approaches philosophically descended
from the germ theory have not dispatched their competitors as successfully as they are often credited with
having done.
Moreover, the triumph of scientific medicine over
epidemic disease is not the clear victory commonly
supposed. The great plagues of Europe disappeared
long before the emergence of bacteriology.6 In the late
nineteenth and early twentieth centuries, impressive
declines in mortality also took place before either
effective medical therapies or prophylactic treatments
had been developed, and occurred largely as a result of
social and environmental change intended to improve
the quality of life and enhance economic production.7-8
What, then, is the importance of this history today?
Both of the last century's leading causal theories are
certainly still with us, but neither has retained its nineteenth century form. It is virtually beyond dispute that
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
certain specific agents are necessary for disease to
occur. Nevertheless, in infectious disease epidemiology, the position accorded to infectious organisms
has changed markedly in the last hundred years. The
intellectual dominance of Koch's postulates declined
by the early twentieth century, and infectious agents by
themselves came to be regarded as incomplete
explanations for the occurrence of mass disease.9 In
spite of their logical power, Koch's causal criteria lead
to an oversimplified view of infectious disease aetiology,10 and because of this realization, the original
theory has been broadened to admit other aspects of
the agent, the environment, and the host"—including
characteristics of the population of which the individual host is a part.12
At the same time, factors like history, geography,
and social status remain exceedingly powerful predictors of the health of human populations, as they
were a century ago, and social change is still an important force behind trends in disease rates.1314 These
observations are usually taken as evidence for the aetiological role of environment.15 For its part, though, the
environmental theory of disease aetiology has changed
by increasingly taking on logic and methods which
mimic the identification of single aetiological agents
and the quantification of their independent effects
through experimentation.16"21 Despite their evolution
along convergent paths since the nineteenth century,
the descendents of the nineteenth century contagion
and miasma theories retain distinct identities which
imply that both theories are right, in the sense that
each provides explanations not satisfactorily made by
the other. However, their growing similarity demonstrates why they are both also wrong.
The modern incarnations of these theories are
centred on agents of disease that are differently conceptualized, but similarly alienated from the ecological
context in which they exist. Neither approach effectively addresses the interdependence of multiple
agents or how human populations become exposed
and susceptible to them. This failure derives from a
view of populations as mere aggregations of individuals (ie vehicles for quantifying exposure-disease
associations) rather than as organized groups with
relational properties that cannot be deduced by
measurements on individuals. While molecular techniques can directly address some important weaknesses of 'black box' environmental theories22 they do
not possess inherent scientific qualities that make them
superior tools for understanding the determinants and
distribution of disease in populations, despite their
often-proclaimed advantages over other methods.123
No specific method or technique can provide a solu-
tion, because the problems with existing epidemiological theories of cause arise from their fundamental
conceptual, rather than technical, inadequacy.
To avoid having to make a choice between scientifically unsatisfactory alternatives, it is necessary to
replace the meaning of cause underlying both the
'black box' epidemiology that Vandenbroucke called a
modern miasma theory, and its complement, the contemporary 'germ theory' that argues for a mechanistic,
molecularized epidemiology, narrowly focused on the
discovery of a new class of agents. Despite their apparent antagonism, these epidemiological theories share
derivation from a global scientific paradigm that has
been termed Cartesian reductionism.24 In epidemiology this paradigm equates causal inquiry with discovering the inherent dose-response relationships
between agents and diseases through rigorous hypothesis testing. The weakness of causal theories arising
from the Cartesian paradigm is their failure to recognize that the action of interdependent parts (agents,
aspects of the environment, and individuals) is not an
immutable, ahistorical characteristic of those parts,
but is, instead, dependent on the properties of the
whole system in which they operate.24 The reductionist
paradigm does not treat populations as structured, historical systems; this omission is analogous to a physiological paradigm that would ignore the development,
differentiation, and organization of cells in an organism. This fallacy makes it clear that cause in epidemiology is not a property of agents, but one of complex
systems in which the population phenomena of health
and disease occur.25
This concept of cause is not new. It has a long history
connected to a line of thought that emerges clearly in
Rudolf Virchow's exploration, nearly a century and a
half ago, of causal ideas ranging from cellular to social
aspects of the origins of disease.26"28 The same conceptual thread has continued to our own time in writings that give a glimpse of the structure of an
alternative epidemiological paradigm.4'9-25-29"32
The success of epidemiology as a science will ultimately depend upon its ability to predict, explain, and
act through the use of theory. John Snow's work on
cholera was important, not because he was a contagionist,1 but because he had a complex, integrative
theory that was able to explain important observations
and lead to practical intervention.33 The challenge for
epidemiologists facing the 21st century is to develop
similarly sophisticated theories of cause that acknowledge the complex systems in which the health-disease
process is embedded. This entails embracing both new
advances from laboratory science and old concerns
about the role of environment without isolating either
one from the whole systems of which they are essential
parts. Meeting this challenge demands no less than a
PHILOSOPHY OF MODERN EPIDEMIOLOGY
34
paradigm shift in epidemiology, one which is being
stimulated by the vast public health consequences of
global environmental, demographic, and social
changes now taking place.
ACKNOWLEDGEMENTS
We thank Naomar de Almeida-Filho, Carl Shy, and
David Savitz for thoughtful reviews of the manuscript.
REFERENCES
Vandenbroucke J P. Is 'The Causes of Cancer' a miasma theory for
the end of the Twentieth century? ImJ Epidemiol 1988; 17:
708-9.
2
Peto R. The need for ignorance in cancer research. In: Duncan R,
Weston-Smith M, eds. The Encyclopaedia of Medical Ignorance. Oxford, Pergamon Press, 1984,129-33.
3
Feinstein A R. Scientific standards in epidemiologic studies of the
menace of daily life. Science 1988; 242: 1257-63.
4
Dubos R. Mirage of Health: Utopias, Progress and Biological
Change. New York, Harper and Row. 1959.
5
Kunitz S J. Explanations and ideologies of mortality patterns. Pop
Dev Rev 1987; 13: 379-408.
6
McNeill W H. Plagues and Peoples. Garden City, NJ, Anchor Press/
Doubleday, 1976.
'McKeown T. The Role of Medicine. London, Nuffield Provincial
Hospitals Trust, 1976.
8
Tesh S. Political ideology and public health in the nineteenth century. ImJ Health Sen 1982; 12: 321-42.
9
Gordon J E. The twentieth century, yesterday, today and tomorrow
(1920-). In: Top F H, (ed), The History of American Epidemiology. St Louis, Mosby, 1953, 114-67.
10
Stewart G T. Limitations of the germ theory. Lancet 1968; 1:
1077-81.
" Evans A S. Causation and diseases. A chronological journey. Am J
Epidemiol 1978; 108: 249-58.
"Greenwood M. Epidemiology, Historical and Experimental Baltimore, Johns Hopkins Press, 1932.
13
Wing S. The role of medicine in the decline of hypertension-related
mortality. Int J Health Serv 1984; 14: 649-66.
14
Wing S. Social inequalities in the decline of coronary mortality. Am J
PublHlth 1988; 78: 1415-16.
" Doll R. An epidemiological perspective of the biology of cancer.
Cancer Res 1978; 38: 3573-83.
1
16
Yerushalmy J, Palmer C E. On the methodology of investigations of
etiologic factors in chronic diseases. J Chron Dis 1959; 10:
27^0.
17
Hill A B. The environment and disease: Association or causation?
Proc Roy Soc Med 1965; 58: 295-300.
18
Rothman K J. Causes. Am J Epidemiol 1976; 104: 587-92.
" Wang J D, Miettinen O S. Occupational mortality studies; principles
of validity. Scand J Work Environ Hlth 1982; 8: 153-8.
20
Robins J. A graphical approach to the identification and estimation
of causal parameters in mortality studied with sustained exposure periods. J Chron Dis 1987; 40 (Suppl 2): 139-61S.
21
Khoury M J, Flanders W D, Greenland S, Adams M J. On the
measurement of susceptibility in epidemiologic studies. Am J
Epidemiol 1989; 129: 183-90.
22
Gordis L. Challenges to epidemiology in the next decade. Am J Epidemiol 1988; 128: 1-9.
23
Perera F P, Weinstein B I. Molecular epidemiology and carcinogenDNA adduct detection: New approaches to studies of human
cancer causation. J Chron Dis 1982; 35: 581-600.
24
Levins R, Lewontin R. The Dialectical Biologist. Cambridge, MA,
Harvard University Press, 1985, 267-88.
25
Selva-Sutter E A. La continuada interaction sociobiologica en la
causalidad estructuraly en la articulation social de los procesos
de saludenfermedad. San Salvador, Facultad de Medicina,
Universidad de El Salvador, 1987.
26
Waitzken H. The social origins of illness: A neglected history. Int J
Hlth Serv 1981; 11:77-103.
27
Rosen G. What is social medicine. A genetic analysis of the concept.
Bull Hist Med 1947; 21: 674-733.
28
Ackerknecht E H. Rudolf Virchow: Doctor, Statesman, Anthropologist. Madison WI. University of Wisconsin Press, 1953.
29
Greenwood M. Epidemics and Crowd Diseases. An Introduction to
the Study of Epidemiology. New York, MacMillan, 1935.
30
Sydenstricker E. Health and Environment. New York, McGrawHill, 1933.
31
Stallones R. Environment, Ecology and Epidemiology. Pan
American Health Organization, Scientific Publication No.
231. 1971.
32
Laurell A C. Epidemiologia: proceso colectivo salud-enfermedad.
Information Cientifica y Tecnoldgica 1985; 7: 16-9.
33
Cameron D, Jones IG. John Snow, the Broad Street pump and modern epidemiology. Int J Epidemiol 1983; 12: 393-6.
34
Kuhn T S. The Structure of Scientific Revolutions. Chicago, University of Chicago Press, 1970.