Causality in the Environment and Health: The Utility of the Multiplex Variable

CAUSALITY IN THE ENVIRONMENT AND HEALTH: THE UTILITY OF THE MULTIPLEX VARIABLE ERICJ. CASSELL, M.D* and MICHAEL D. LEBOWITZ, PhD.i The model of cause and effect that has been so useful in infectious disease and classic epidemiology has failed in the search for understanding the relationship between the environment and health. That is to say, despite intense effort, no single or even simple combination of common environmental constituents (pollutants or weather factors) has stood up as a cause of human disease or illness in the concentrations found in ambient atmosphere [1-15]. Further, but equally important, methods for air pollution control, that is, establishment of ambient-air or emission standards, pollutant by pollutant, based on the infectious disease model, have not been particularly successful in the management of environmental problems. The weakness of that approach has been especially apparent as a result of the current "energy crisis." The failure seems to us to arise from a durable but nontheless simplistic view of reality common to and now under attack in many areas of science. We believe the basic problem is in our understanding of causality . The concepts of causality as expressed in many of the medical and behavioral sciences have been broadened and made diffuse past utility for further research or control. Another way of saying this is that we have become too wary of making causal statements, replacing them instead with statements about correlation, association, or probability of association. This has come about for good reasons as the phenomena under study have become more complex, making simple or unguarded causal statements increasingly difficult. Indeed, most simplistic statements of "cause" in environmental effects have turned out to be incorrect . Because our world demands action, the scientist so frequently constrained on paper about cause-and-effect statements often behaves as if, in fact, he believed that specific environmental variables (i.e., SO2) were a cause of human disease. "Clinical professor of public health, Cornell University Medical College, New York, New York 10021. tAssociate professor of internal medicine, University of Arizona, College of Medicine, Tucson, Arizona 85724. 338 I Eric J. Cassell and Michael D. Lebowitz · Multiplex Variable We have, then, a curious paradox. On the one hand the situations of our world, such as the effect of the environment on health, seem to demand control, and control seems to require some concept of cause. On the other hand our increasing appreciation of the complexity of the problems increasingly warns us off statements of cause. The resolution of this paradox would seem to demand a reassessment of the concept of cause. The Concept of Cause We are compelled to return to Hume. Hume said that all that can be proved about the phenomenon called cause and effect has been proved when the two events considered "cause" and "effect" are shown to constantly occur together [16]. He argued that causal connections are contingent , not logically necessary. In other words, there is no logical manner by which one can arrive at cause and effect before the fact. As Pap [17] has pointed out, Hume's insight is often ignored because of two common confusions. The first confusion is that the observation comes to stand for its concept. For example, the tubercle bacillus has come to be equated with tuberculosis. That is, one tends to forget the intervening steps or processes involved in the coincidence. The original utility that followed the demonstration of the tubercle baccillus in tuberculosis—guiding the search for cause in other diseases—becomes diluted as the confusion that tuberculosis is the tubercle bacillus directs us away from the power and importance of the intervening processes. When we study health in terms of environmental observations this confusion of the observation for a concept in the infectious disease model persuades us to look for a single cause (i.e., pollutant), regardless of the intervening processes or the discontinuity of the chain of events that is in violation of Hume's demand for constant conjunction. We explain away our difficulties and failures by invoking doctrines of complexity, multifactoriality , contributory variables, etc. The second confusion pointed out by Pap is that once having explained (as distinct from having observed) why an event A...