DEBATE DEBATE

Introduction

For several decades after World War II, many parts of the world undertook committed efforts to eradicate diseases through applied biomedical sciences and promote health through public programs fostering social and economic equity. Both efforts seemed to be successful. Then, in the closing years of the twentieth century, the efforts to promote health were largely abandoned in favor of increasing economic activity per se, at the cost of rapidly rising income disparities within and between countries (UNDP, 1999). At the same time, infectious diseases once thought to be all but eradicated have begun to reemerge as serious problems throughout the world (Lederberg et al., 1992; Waltner-Toews, 1995). All of this is occurring in a context of global environmental change occurring at speeds and a scale unprecedented in recent human history. This would appear to be a good time to reconsider how we think about human diseases, health, and their ecological and social contexts.

What is health?

Good health, according to the preamble of the World Health Organization, is "... a state of complete physical, mental, and social well-being and not merely the absence of disease or infirmity." This is clearly a utopian definition, a heaven-on-earth which humanity may well strive for, but never - except in rare moments perhaps when good wine and good friends conspire together - achieve. Microbiologist Rene Dubos, as an antidote to this utopianism, suggested that good health was simply a "modus vivendi enabling imperfect [people] to achieve a rewarding and not too painful existence while they cope with an imperfect world" (Last, 1988:57). A Dictionary of Epidemiology (Last, 1988:57) also proposes a more pedestrian, but perhaps more workable, definition of health as: "A state characterized by anatomical, physiological, and psychological integrity, ability to perform personally valued family, work, and community roles; ability to deal with physical, biological, psychological, and social stress; a feeling of well-being; and freedom from the risk of disease and untimely death."

In general, definitions of the health of plants, animals, people, communities, and ecosystems include some notion of current balance and harmony and some notion of reserve (Waltner-Toews & Wall, 1997) or capacity to respond and adapt to a changing environment (Constanza et al., 1992). Furthermore, health is directly related to the achievement of desirable and feasible goals, and disease is but one constraint on achieving those goals. Thus the achievement of health is a social activity within biophysical constraints, rather than a biomedical activity within social constraints. Furthermore, those who possess the skills required to prevent and treat disease - analysis, diagnosis, treatment, professional authority - are ill-equipped to promote health. Disease prevention and control programs, in fact, may actually create ill-health. If nothing else, an ecosystem approach to health makes this transparent.

Why do we need approaches to improving or achieving health?

Despite the disclaimer of health being "not merely the absence of disease" in the WHO definition, nearly all efforts to improve human health in the past century have been aimed at ridding the world of various diseases. Indeed, those accomplishments are so remarkable that one might ask, given the obvious benefits of freedom from smallpox, measles, and polio, why one would even bother to discuss alternatives to disease control. Why not simply throw out our vague and luxurious notions of "complete physical, social, and mental well-being" and simply redefine health as "absence from disease or pathology"? Based on the concept of health as the capacity to achieve culturally defined goals, and on the failures of the biomedical model, I would like to suggest that there are sound practical and theoretical reasons for not doing this.

First, many disease-control programs are no longer effective. The rising tide, globally, of multi-resistant organisms and pesticide-resistant insect vectors is the direct, unintended result of therapies we use to control or eliminate them. One short-term response to these "counter-attacks" is simply more of the same - more vaccines, more drugs, more pesticides - as has been proposed by the Institute of Medicine in the United States. In some ways, this is like responding to successful guerrilla warfare by proposing bigger conventional armies and weapons. I suggest that it is time to ponder the wisdom of our bio-military metaphors and linear causal thinking, to address the flaws in reasoning and tactics we have employed to date, and to use our much vaunted intelligence as a species to find more creative solutions.

Second, success in programs which manifestly promote health in some dimensions - such as improvements in agriculture to address food shortage - have had unintended negative effects on other aspects of health, such as disease. Simple talk of creating "supportive environments for health" is characterized by the optimism of a race horse with blinders. Water management programs have had particularly devastating effects by favoring several tropical diseases. Dams are built to generate electrical power, to control flooding, and to generate wealth (all of which are demonstrably supportive of health). Nevertheless they also expand or create new habitat for flora and fauna which cause disease, and remove sources of natural renewal from farmland (Hunter et al., 1982). In Bangladesh, epidemic Kala azar (leishmaniasis) has occurred in populations living within flood control embankments (Minkin et al., 1996), and malaria epidemics, mad cow disease, and cyclosporiasis have all been associated with aggressive agricultural programs (Waltner-Toews, 1999).

More specifically, this cross-sectoral blindness means that control of some diseases through, say, environmental restructuring, leads to the emergence of other diseases. This is most obvious in foodborne diseases, where the industrialization and centralization, which quite naturally accompanied regulations on canning and pasteurization to control botulism and brucellosis, have resulted in the creation of large ecological niches and associated large-scale epidemics of diseases such as salmonellosis. In North America, policies and practices which encourage a voluminous and cheap supply of food, and which on the one hand serve as a preventive against starvation, tend on the other hand to undercut the economic and ecological sustainability of farmers and to create a whole new array of nutritional and disease problems associated with obesity.

Disease control programs themselves can also undermine health in at least two ways. They can disrupt ecological systems which make health possible. Thus we are faced with the dilemma that DDT is useful in bringing malaria under control, but at the same time endangers the integrity of the interactions among insect pollinators, birds, and food production which make sustainable livelihoods, and health, possible. Secondly, and less obviously, food supplementation, vaccination, and drug treatment programs based on a biomedical model can undermine the ability of people to adapt resourcefully to their own environments. They do this by reinforcing the notion that it is appropriate for outside experts to determine which outcomes - among many possible competing ones - are appropriate, which responses are "correct", and who should carry them out. Physicians and veterinarians who are well equipped to diagnose and treat are in general very poorly trained to promote health, which requires negotiation and adaptation.

Current health and disease control programs often work against each other across organizational scales. Drawing inferences about populations based on studies of individuals is termed the atomistic fallacy, and is widespread and widely tolerated in epidemiological studies. Ironically, the converse fallacy - drawing inferences about individuals from population studies - is vigorously guarded against. What this means is that all efforts are focused on finding individual determinants of disease, and the broad systemic conditions - the very conditions which determine whether or not healthy human communities are sustainable - are largely, by design, ignored. Thus we have the absurdity of governments in some industrialized countries giving away groundwater to private companies, who then wrap it in plastic and sell it back to the original owners of the water (the citizens of the country) under the pretense that this is good for their individual health. Even if the water in the bottle could be demonstrated to be superior to tap water (which it is not) it would still have major negative consequences for population health because of the energy and resources required for manufacture and disposal of the bottles.

Problems are solved at an individual level but become major problems at a regional or global level. Thus, saving children through vaccination without concomitant programs in education, nutrition, agriculture, and sustainable livelihoods undermines the health of whole communities and condemns them to slow and painful death and disintegration (McMichael, 1993). Indeed, the tension between sustainable population health, which requires a certain death and replacement rate, and individual health, for which death is the ultimate negative outcome, has no solution within current biomedical models (Waltner-Toews, 2000). The idea that death and maybe even disease might in some sense be important for sustainable health cannot even be conceptualized in a normal biomedical framework.

Finally, current disciplinary-bound approaches to health, which focus on biomedical and personal behavioral issues, inhibit health researchers and workers from addressing the real causes - which reflect irreducible interactions among economics, politics, and ecosystems. This is addressed more fully in a later section of this paper on emerging infectious diseases.

The practical problems, contradictions, and dilemmas that have arisen in our quest for less disease and better human health do not simply reflect lack of investment in public health and scientific infrastructure: sewers, water supplies, laboratories, vaccines. These are all important. However, this dis-investment in the public good, widespread under conditions of globalization, is itself a product of a lack of appreciation of the nature and complexity of the problems we are addressing.

The theoretical basis for a new approach to health

The world we live in can be described as what Russell Ackoff has called "a mess", that is, "interacting problems or issues that are not easy to appreciate as a whole" (Flood & Carson, 1993: 11). Furthermore, it is clear that the complexity we see in the world around us is not only a property of the world we are studying, but also a property of the interaction between ourselves and the world. Normal, conventional science assumes an objective, external observer. For public health and ecological issues, where people - researchers and subjects alike - are internal to the subject being studied, this assumption is untenable.

Causal models used by epidemiologists and biomedical researchers tend to be linear: specific disease outcomes are defined and studies are designed to identify determinants of those outcomes. In all but a few rare occasions, epidemiological studies focus on determinants of disease in individuals. Models of disease can rarely accommodate the disease outcome itself as a cause of other outcomes which themselves connect back to the disease, or to other diseases or health measures. For instance, we have tended to assume that if scientists identify smoking as a cause of cancer, or tick bites as the vehicle for infection by a parasite, then the public health and behavioral changes necessary to prevent these diseases will be made post haste by an informed public. Cross-scale interactions between individual health and disease outcomes, agricultural and economic policies, social expectations created and reinforced by advertising and entertainment, cultural habits of scientists and non-scientists, evolutionary ecology, and epidemic diseases are rarely studied.

Based on the problems identified above, any new approach to studying and promoting health must at least encompass multiple ecological, social, and health outcomes, and "not just the absence of disease". These multiple outcomes need to be expressed in a context of a set of interactions, with people inside them, rather than looking at the environment as an external "support" or "threat" to health. Secondly, to be adequate to the new tasks facing us, any new theory and plan of action must account for problems of scale and inter-scalar connections.

Systems theory explicitly addresses the connections between various elements, and thus provides a useful starting point. A system is simply a set of elements which interact with each other within a certain boundary. Many attempts have been made to reduce people, farms, households, or ecosystems to well-defined, machine-like systems which can be simulated mathematically. While such models provide some useful information, those which are the least realistic tend to provide the strongest implications for action, while those that are the most realistic provide understanding but suggest no obvious solution to the problems being studied. Puccia & Levins (1985) suggest that there are unavoidable trade-offs in systems modeling between generality, realism, and precision. "No model", they state, "can be general, precise, and realistic"(Puccia & Levins, 1985: 9).

In the face of this intransigence of the mess we are in, some scholars retreat to the details of reductionism, with the assurance that if they are not helping the situation, they may at least be supplying data which might be useful. However, the systems literature itself seems to present us with a kind of post-modern smorgasbord of systems ideas, ranging from chaos and catastrophe to soft, dynamic, complex, and ecosystems. Some authors (Funtowicz & Ravetz, 1994) argue that the interactions between people and ecosystems reflect an "emergent complexity", which cannot be captured by even our most sophisticated complex systems models. This may be true, and should engender humility into all our actions, but does not provide a very useful guide for action itself.

SOHO systems

Among the options available in the new systems smorgasbord is a set of ideas which brings together conventional, soft, complex, chaotic, management, participatory, and ecosystems. These ideas, subsumed under an umbrella of "the ecosystem approach", are based on the view that the mess we live in can be usefully understood as self-organizing, holarchic, open (SOHO) systems (Kay et al., 1999).

Holarchy

A SOHO system is characterized by patterns of interaction occurring in nested hierarchies (sometimes called holarchies: Checkland & Scholes, 1990; Allen & Hoekstra, 1992). Unlike say, a military hierarchy, in which there is a chain of authority, each unit (holon) in a holarchy is both a whole thing and a part of something larger. For instance, individual people are parts of families, which are parts of neighborhoods or villages, which in turn make up larger communities and so on. This means that an initiative, such as improving public hygiene, taken at any given scale (for instance the neighborhood) has implications for holons of which it is a part (the city) as well as the holons of which it is comprised (individuals and families). The speed of change at one scale (say introduction of genetically engineered crops at a field scale) may disrupt ecological and social systems of which they are a part, and which have co-evolved over much longer periods of time, so that the expected benefits (improved nutrition to individuals) may not accrue in the expected way (Giampietro, 1994).

Feedback loops

The interactions in SOHO systems are can be represented as a mixture of positive and negative feedback loops. For instance, people engage in various economic activities - such as clearing land for agriculture, irrigation, mining, house-building - in order to make money to improve the quality of their lives. Wealth generated by these activities may be used to build better roads, schools, and sewage disposal facilities. People who have more schooling may be better able to solve social and public health problems - at which time they may see that some of the activities which made the schools possible may themselves be identified as problems. Agricultural activities or manufacturing may, for instance, result in greater pollution of the water supply and the environment, heavier stress on energy use, and general deterioration of the ecosystem. Some diseases may be prevented when swamps are drained or damns are built, even as habitats for new ones are created.

Self-organization, attractors, and surprise

In natural SOHO systems, it appears that, as high quality energy (referred to in the thermodynamics literature as exergy) and information are pumped into the system, the feedback loops become organized in such a way as to make more effective use of the entering resources, build more structure, and enhance their own survivability. It is this combination of feedbacks, boundaries, and openness which results in what is called self-organization. Self-organization is necessary for life to occur. All living things - organisms, ecological systems, socio-ecological systems - must remain both bounded, with a set of internally relatively stable interactions, and open to receiving resources and energy, and dumping waste, if they are to remain alive.

Some elements in any ecosystem are more tightly connected than others, and more essential to their mutual well-being and/or the well-being of the system overall. The importance of connections is not determined by sector (agriculture, health, business, social, environmental), but by flows of useful energy, resources, and information related to patterns of self-organization. Thus, activities which enable a community to make more effective and elaborate use of natural resources and information are likely to have a greater impact on the viability of a given population than health care activities.

As already suggested, these feedback loops in SOHO systems tend to organize themselves in certain patterns which are coherent. Systems researchers call this self-organization "attractors". Most ecosystems - because of the energy and resources available to them - seem to have a propensity to fall into a certain limited set of possibilities. Despite advertising claims to the contrary, not everything is possible, and we cannot all become whatever we want. We - and the ecosystems and societies we live in - are comprised of physical elements which constrain our possibilities. Nevertheless, human intentionality and creativity can push or alter these constraints, which may result in new system states - or just in general disintegration. In the latter case, if life is to continue, some new sets of mutually supportive interactions need to arise. C. S. Holling, drawing on studies of forest fires and spruce-bud worm infections, has described a normal pattern of development in many ecosystems which depends on at least one stage of local disintegration. According to his "lazy-eight" model, these ecosystems follow a path from exploitation of disconnected resources into greater conservation and stored energy, creative destruction with release of biomass (small patches of fires or infestation), reorganization, and then exploitation again (Holling, 1986). Farmers follow a very controlled version of this as they plant, husband, harvest, cultivate, and then plant again.

Changes between system states may be quite sudden. Ideas of thresholds and breakpoints are well-known in both the epidemiological and ecological literature (May, 1977; Kay et al., 1999). Disease organisms increase to critical levels at which time the probability of adequate contact increases to the point where the epidemic explodes.

There is good evidence that ecosystems can exist in different steady states, reaching critical points and then suddenly reorganizing, during which time there may be drastic changes in species composition and diseases. Research on ice cores from Greenland has indicated that global temperature changes on the order of 5-16^oC have occurred over mere decades during global climatic changes in the past (Severinghaus & Brook, 1999; Taylor, 1999). Such "flips" are attributed to the crossing of temperature thresholds required to keep global ocean currents moving in particular ways. These kinds of threshold effects followed by catastrophic changes between attractors have been demonstrated for a variety of systems, social as well as ecological (Casti, 1994; Kay et al., 1999).

This means that gradualist views of disease changes in relation to climate, for instance, may be a very poor basis on which to base organizational response plans, and hence undermine that part of health which comprises the ability to adapt and respond to stress. On the other hand, these insights may also be used to create health benefits disproportionately large relative to the apparent effort. Putting speed bumps on all city streets, and narrowing them, could within a few years result in cleaner air, less respiratory disease, and healthier people who walk more. Of course this would also result in the loss of income from motor-related activities and a change in the physical structure of the cities and in the structure of the national economy. Paying farmers for managing landscapes sustainably as well as for producing commodities would change the entire structure of rural communities, migration to cities, international trade, patterns of foodborne disease, and global economic power. While the exact outcome in either case could not be predicted, an informed public could at least see the general shape of the system options.

Multiple perspectives

Because such complex feedback loops have both positive and negative effects, different people will look at the situation (and evaluate it) differently. Where one person sees the excitement of economic activity, another person sees deforestation, where one person sees disease control by draining swamps, another person sees loss of wildlife and clean water provided by the filtering effects of wetlands. Where one person sees disease control through metal roofing, another person sees increased economic and environmental costs and less comfortable houses. This means that, as the scientific description gets better, the problems are not necessarily resolved, only clarified.

One challenge to creating systems descriptions is that of selecting what to put into them and what to leave out. We cannot describe everything about everything! Nor does it seem appropriate for expert scientists (which scientists?) to determine what is important and desirable for everyone else. The scientific, ecological information is important, but not sufficient. For this reason, some researchers have developed the idea of extended peer groups and "post-normal science". In normal science, one's disciplinary peers determine the "success" and "quality" of one's work. If we are talking about sustainable, healthy communities, then clearly there are others who will have something important to contribute. This is especially important given the uncertainty of scientific predictions with regard to complex systems.

Given this complex theoretical base and the uncertainty of our knowledge, how can we do scientific research and science-based development in support of health?

What is the ecosystem approach? The practical implications of thinking in terms of complex systems

The concept of an "ecosystem" dates back at least to British ecologist Arthur Tansley in 1935, and has undergone various permutations in the decades since then (Bocking, 1994). The ecosystem approach, as described in this paper, is a conceptual and management approach developed and applied by ecologists working with the International Joint Commission of the Great Lakes (Allen et al., 1991). These huge freshwater lakes, which straddle the Canada-US border, are ringed with some of the biggest industrial cities in North America, such as Chicago, Cleveland, Hamilton, and Toronto. In this situation, standard approaches to environmental management, which were designed for smaller areas or parks not usually inhabited by people, did not seem appropriate.

Combining insights from ecosystem ecology, complex systems theory, catastrophe theory, and hierarchy theory, ecosystem researchers and managers developed a process that anticipates change (without predicting its exact timing and nature) and creates adaptive solutions in a participatory manner. The process itself grew largely out of the work of Peter Checkland, a systems scientist and specialist in business management (Checkland & Scholes, 1990). Arguing that human activity systems could not be understood or managed like machines, he developed a seven-step "Soft Systems Methodology" which focused on how to understand the goals, perceptions, and transformative actions of people in organizations. This methodology is best worked out for organizations which have definable goals. Allan & Hoekstra (1992), among others, have explored how to combine such a methodology with the insights gained from natural sciences and ecology. Still others, such as James Kay and members of the "Dirk Gently Group" (The Dirk Gently Group is an informal network of researchers into the problems of decision-making under conditions of complexity and uncertainty. Membership varies, but includes Silvio Funtowicz, Gilberto Gallopin, Mario Giampietro, James Kay, Bruna De Marchi, Tamsyn Murray, Martin O'Connor, Jerry Ravetz, and David Waltner-Toews) have struggled with the problems of how to accommodate human activity systems, multiple perspectives, and the constraints of ecological and energy networks.

According to the Ecological Committee of the International Joint Commission, in the ecosystem approach "there is not one material ecosystem to which our definitions must conform. Rather, the human actor must accept responsibility for erecting definitions and be prepared to change them when the purpose of the description changes" (Allen et al., 1991:5). Generalizing this understanding to recognize that there are many human actors, with many legitimate perspectives, Kay & Schneider (1994:38) have argued that using an ecosystem approach means "changing in a fundamental way how we govern ourselves, how we design and operate our decision-making processes and institutions, and how we approach the business of environmental science and management."

Central to an ecosystem approach rooted in complex systems theories, then, is the notion that achieving sustainability requires bringing together a variety of legitimate stakeholders, drawing on a variety of accepted bodies of knowledge, to negotiate a learning path based on a series of conflict resolutions within ecological constraints. Continual learning based on free flow of information and mutual respect, and investment in democratic local governance, are keys to success (Funtowicz & Ravetz, 1994; Roling & Wagemakers, 1998).

This means that, in the ecosystem approach, the practical, social, and institutional dimensions are of as much concern to researchers as the scientific and scholarly concerns. This can get very complicated very quickly, if we remember that these systems are holarchies. Thus, even if the people in one region can agree on a common plan of action, they will need to pay attention to the larger system of which they are a part. On the one hand, urban neighborhoods may not be able to solve all their water problems without help from city or even national authorities in charge of water distribution systems; on the other hand, if local neighborhoods clean up their streets simply by shipping garbage elsewhere, this may create a much bigger, more concentrated problem than the one we started with. Thus local participatory action, while essential, is always conditional, just as individual freedom in society is conditional on interactions with other "layers" in the holarchy (Waltner-Toews & Wall, 1997).

Both scientific studies and participatory action research (PAR) are necessary ingredients of an ecosystem approach, and both must be reoriented to a systems understanding of reality. Participatory development without science is just politics; science without participatory development is an academic exercise, and without a systems perspective, neither can be used effectively to promote sustainable development or ecosystem health. There is no single way to visually represent this kind of multi-level, multiple perspective research (nor is it the intent of this paper to describe the details of such an approach). Several models have been proposed. A cube with axes representing scale, perspective, and system goals may be used to classify research and management activities (Figure 1, as adapted from VanLeeuwen et al., 1998). Kay et al. (1999) have proposed a two-armed schematic which emphasizes the interactions between natural sciences and socio-cultural visions (Figure 2). Murray and colleagues (Murray et al., 1999) have proposed a flow diagram, the Adaptive Methodology for Ecosystem Sustainability and Health. This puts the emphasis on process, and may give greater guidance for planning and action (Figure 3 - Murray et al., 1999).

However this is represented, biomedical researchers, sociologists, anthropologists, natural scientists, and ecologists - both professional and lay - must work together to describe the important patterns of natural resource use and dynamics of the ecosystems in which the communities live. At best, what emerges from this collaboration is an evolutionary understanding of how the current system came to be, what the key interactions are which determine its current state, and, based on historical evidence, what kinds of options (attractors) are available. They also describe the system both in terms of the functioning of the diverse flora and fauna that are present, and in terms of their functions for capturing and disseminating useful energy (exergy) (Kay et al., 1999). Finally, the research process itself involves negotiation of goals and implementation of desirable and feasible actions. Within the ecosystem approach, policies, programs, and actions serve the same function as hypotheses and experiments in laboratory science.

Where does disease fit into this? One might begin by re-thinking disease classification in terms of ecological contexts, rather than simply using Linnean classifications. For example, Mara & Alabaster (1995) have devised an environmental classification of housing-related diseases in developing countries. It is another step in this process to identify how changing housing types might modify the evolution of virulence in malarial parasites. Ewald (1994), among others, has discussed how enclosed housing, which removes prostrate victims of malaria from the pool on which mosquitoes feed, will likely lead to an evolution of malaria to more benign forms. It is also known that metal roofing provides less habitat for Chagas disease vectors; nevertheless, such roofing carries with it ecological costs (use of materials and energy) as well as economic costs which may outweigh their benefits. These pieces of information must be integrated - bringing disease into a health context - as we work intensively with communities in specific places to create healthy social and ecological systems (Gitau et al., 1996; Murray et al., 1999; Yassi et al., 1999).

Because the ecosystem approach is based on a holarchic view of reality, part of the research is to find ways of creating regional and global economic, health, and environmental policies which facilitate, rather than hinder, this work. Health issues need to be put into a holarchic understanding, so we do not run into the situation in which problems are solved at an individual level but become major problems at a regional or global level, or vice versa. For instance, programs which focus on increasing global food production often undermine local food security and health because they fail to account for cross-scale issues related to communication and control. Conversely, control of malaria or the plague in local communities means that controls must also be put in place at regional levels to prevent reemergence locally. At the same time the cost of control, the necessity for multi-sectoral cooperation, and the consequences of failure are considerably higher at the regional level. The only way out of this bind is through global eradication, or the creation of adaptive strategies to live with the disease through better nutrition, public health infrastructure, social support networks, and economic equity, all of have been demonstrated to mitigate the impacts of disease without eradication. Ultimately, local actions will not be sustainable if there are not broader supportive policy environments; conversely, if local communities and ecosystems fail, there will be no global economy or healthy global bio-village.

How will the ecosystem approach aid in achieving health ?

Health is related to the ability to achieve desirable and feasible goals. Therefore, working with people to identify goals, to understand the socioeconomic and ecological constraints and opportunities facing them, and to negotiate resolutions rather than "find" solutions, helps those people to become healthy by definition. There is no biomedical solution which can be imposed from the outside to promote health. Thus, an ecosystem approach to health promotes health in the very means it uses to understand and promote health. Secondly, since the ecosystem approach embeds the social goals of health within our best understanding of the ecological context, it is more likely to be sustainable than an approach to health issues which is based on the fantasy of a social reality disconnected from the biosphere.

Implication of the ecosystem approach for research on tropical and emerging diseases

Emerging, reemerging, and newly recognized diseases in both the tropics and in temperate zones, represent failures - failures to understand the socio-ecological systems we live in, and failures to respond to new understandings as they are uncovered. What we learn from these failures will largely determine how successful we are in creating sustainable and healthy human communities on this planet.

Descriptions of emerging diseases and causes of emergence have been reviewed in several major reports and conferences (Lederberg et al., 1992; Levins et al., 1994; Waltner-Toews, 1995; Meslin, 1997). The Institute of Medicine in the United States, in a 1992 report, identified half a dozen forces which were resulting in the emergence of new diseases and the resurgence of old ones (Table 1 - Lederberg et al., 1992). Peter Drotman of the Centers for Disease Control has pointed out that 150 years earlier, in a report to the Prussian government regarding a typhus epidemic in Upper Silesia, Rudolf Virchow identified causes and made recommendations which are eerily similar to those of the Institute of Medicine (Table 2 - Drotman, 1998). What is clear from both reports - and remarkable given that both men were basic biomedical scientists - is that the causes identified are social, environmental and political. This would come as no surprise to anyone who has done serious research in the area (Evans et al., 1994; Farmer, 1996).

Nevertheless, while Virchow's recommendations are overtly social and political - and hence based on the evidence - those of the IOM seem unconnected with the causes (Table 3). They are at best technical, and at worst merely self-serving (at least for biomedical scientists). Where, one might ask, are the recommendations related to some of the root causes they have identified - technology and industry, international trade, investment in public health infrastructure? Based on the evidence presented, should health researchers not be making strong health representations to organizations like the World Trade Organization and the World Bank, not on how to clean up the disease mess after the fact, but on how to prevent the mess in the first place? It would almost appear that the ideological lenses through which emerging diseases - indeed diseases in general - are being studied, preclude acting on the evidence. This, if nothing else, should raise a warning flag that those who study disease are not necessarily well-equipped to promote health, and that new modes of thought are required to promote health at the beginning of the new millennium.

Tropical diseases obviously are related specifically to tropical climate, flora, and fauna. However, if we consider the diseases selected by the World Health Organization's TDR Program (leishmaniasis, onchocerciasis, Chagas disease, leprosy, tuberculosis, African trypanosomiasis, schistosomiasis, dengue, lymphatic filariasis, and malaria), it is also clear that many of these do not simply occur in the tropics because of climate and landscape. Indeed, even the tropical climate is subject to strong influences by the climate-changing behavior of industrialized countries, and landscapes are being transformed by industrialization processes within tropical countries. Diseases such as leprosy and tuberculosis have occurred - and continue to occur - in colder climates where poverty creates homes for them. Even those diseases which are likely constrained by the ecology and climate of the tropics, such as Chagas disease, can clearly be strongly influenced by social and economic practices. Many tropical diseases are thus not only diseases related to climate and environment, but are diseases of poverty. In epidemiological terms, conditions of poverty increase the probability of adequate contact and hence increase the likelihood of epidemics. The disappearance of many infectious diseases - such as cholera, leprosy, tuberculosis, and malaria - has had much more to do with housing, nutrition, and water management than with any advances in biomedical science (Dubos, 1965). This is well established and, I would hope, non-controversial. These public health interventions require public investments. Current drives to evade these requirements by arguing for more research on drugs and vaccines are largely driven by private enterprise economics and a post-Cold War fear of seeming to be interested in the public good, rather than on the evidence.

Even if the financial powers could be persuaded to look at the evidence, however, we need to admit that we are working in a different context than that which the Europeans faced at the beginning of the 20th century. The planet is considerably more crowded, northerners are a great deal more insatiable in our consumption of resources, and the whole human enterprise now threatens the ecological integrity of the planet which makes our lives possible (UNDP, 1998). So, on the other end of the political spectrum, we must say that massive public works programs and environmental restructuring of the kind which freed Europeans from infectious disease need to be more carefully considered with regard to their effects on ecosystems (which Virchow did not consider).

The ecosystem approach is a way to bring together the socioeconomic and biophysical dimensions of health. Nevertheless, what it brings to our understanding of emerging diseases is not so much a way of identifying new variables - though it certainly forces us to cast our net much broader than in conventional research - but a new way of organizing how we think about them, and how we respond to them. Emerging diseases demonstrate to us how our understanding of nature, and our management responses, have been limited. In particular, our understanding of how social and ecological variables appear to behave as complex adaptive systems changes how we think about social, economic, or environmental management programs. Some of the lessons to be learned have been previously reviewed for foodborne diseases (Waltner-Toews, 1996, 1999) and infectious diseases in general (Levins et al., 1994). This paper will emphasize three particular failures in our understanding and response which have characterized recent discussions on emerging diseases. These failures and insights based on them lead directly into a series of recommendations (Table 4).

1) The failure to understand non-linear interactions, chaotic attractors, and catastrophic thresholds. This failure is apparent in the way in which economic development, agricultural trade policies, and the like are treated, as if they were somehow issues separate from social and mental, as well as physical, well-being. It is also apparent in current gradualist expectations of the consequences of global warming. The promotion of health in an ecosystem context must account for multiple inter-related health outcomes, of which disease reduction is but one. Furthermore, it will focus not so much on linear predictions of health improvement through fixed programs (which appears to be impossible), but on creating adaptive organizational responses.

This leads to my recommendations 1a and 1b related to system dynamics: 1a) Research should be expanded on the changes in ecosystems which will most likely push them into the domains of new attractors. What, researchers should be asking, are the likely health, environmental, and social effects of such reorganizations? And 1b) That all economic, agriculture, and business policies, approvals of new technologies, and so on be justified in terms of our best understanding of their likely systemic effects, and that they not be undertaken unless those who live in the system agree that those effects are acceptable and/or that we have in place adaptive mechanisms to deal with those changes.

2) The failure to account for holarchy. Current efforts aimed at preventing diseases in individuals are currently undertaken with little thought as to the consequences for population or ecosystem health. Similarly, population-based programs tend to have a strong autocratic flavor to them which undermines their success.

Failure to account for holarchy has resulted in global trade policies which generate wealth for the global economy, leading to the destruction of local ecosystems and the maintenance of local poverty and infectious diseases (Waltner-Toews, 2000). This is because global and regional policies often destroy local socio-ecological boundaries; individuals whose only loyalty is to the global market undercut the integrity of local social and ecological systems. Of course, healthy individuals, communities, and ecosystems interact with their surroundings and change in response to them. However, globalized commercial trade is just another form of introducing alien and often invasive species into ecosystems; totally free trade is comparable, at the local community and ecosystem level, to pulling the skin off a live animal. Yes, there is more immediate flow of nutrients - but this inevitably leads to local death. Some protectionism is simply an act of survival. Indeed, the very idea of interaction with an environment presupposes the integrity of the boundaries across which the interaction takes place. I must underline, however, that holarchical thinking drives us not to isolationism, nor to globalism, but to interactions with a profound respect for boundaries and scales. Controlling the emergence of new diseases and creating healthy human communities depend on maintaining the tension within holarchies.

Recommendations 2a and 2b are therefore that 2a) Research be expanded on the holarchic boundaries which are important for relative socio-ecological stability and 2b) Policies be developed which respect holarchic boundaries, make explicit environment-economy-health feedback loops within those boundaries, and which encourage trade across those boundaries.

3) The third failure is one of long-term vision, and hence perhaps the most forgivable. On the one hand, this is the failure to take into account the ecological, evolutionary effects of social and economic programs. The obvious examples are those, like Lyme disease, where reforestation and expanding deer populations, coupled with encouragement of people to enjoy outdoor life, have also had negative disease-emergence effects. But the failure is more widespread than that.

This is also evident when apparently narrowly focused changes are proposed - such as the introduction of new genetic varieties of plants on a wide scale. For instance, the introduction of "green revolution" varieties appears now to have played an important part in the systemic changes which have resulted in more than two billion people worldwide with serious micro-nutrient deficiencies, which may be considered a major non-infectious emerging disease (Welch et al., 1997). Farmers chose new varieties because they grew faster and made them more money. This displaced a variety of other crops and in the process restricted the diet for billions of people. They also became considerably more dependent on fossil fuels (for fertilizers and pesticides), contributing to the mining of non-renewable resources, global warming, and the emergence of new diseases, both infectious and non-infectious. Hence a caloric shortage problem was solved in such a way that the system was reorganized in fundamentally unsustainable ways.

The new wave of genetically modified organisms is being introduced based on the same naïve view of nature that brought in the first wave of introductions, with the probability they will further destabilize ecological systems (Giampietro, 1994). This will no doubt result in changes to micro-flora which will lead to the emergence of new epidemiological patterns of disease. Certainly this is what has already happened with regard to many foodborne illnesses, such as bovine spongiform encephalopathy, antibiotic-resistant Salmonella DT 104, and verotoxin-producing E. coli, all of which have emerged in the agrifood system as unintended side effects - based on naïve and narrow understandings of social and ecological systems - of economic policies intended to promote cheap food. Agricultural and economic development policies can also be held responsible for the (re)emergence of malaria in several parts of the world.

An understanding of possible long-term effects may not lead us to abandon social projects and policies. Indeed, as Ewald and other evolutionary ecologists have pointed out for malaria and cholera, these may be used to good advantage to devise programs to control disease. But these long-term scenarios should at least cause us to build into their structure and implementation mechanisms to detect probably long-term impacts and to compensatory mechanisms for those who are likely to suffer from those changes.

This is not only a failure to see the long-term effects of human social projects, however, but a failure to understand that ecosystems have co-evolved over a long period of time and that when we invade or disrupt those systems, we do so at some peril, not just to the ecosystem (rain forest destruction) but to ourselves, as we become entangled in life cycles that evolved without us, resulting in the emergence of diseases like Ebola virus and Kyasanur Forest disease.

Hence, recommendations 3a, 3b, and 3c are: 3a) Research be expanded on the co-evolution of ecosystems - including the microbial populations within them - over time. This will require the establishment of long-term sites in various parts of the world and incorporating various types of human activities. Such long-term sites have been set up for "pure" ecological studies; the bases for such teaching-learning sites have already been established in several on-going projects in various parts of the world. What is needed now is to rationalize these, give them some permanent long-term funding, and expand the range of researchers involved; 3b) That new health, economic, and environmental initiatives take into account possible long-term effects on the system dynamics; and 3c) That relatively uninhabited ecosystems be protected from further human invasions by economic and political policies until such a time as we have a fuller understanding of the nature of the infectious agents in the system as currently evolved.

Achieving health for all: Beyond complexity, beyond 2000

Because components of ecosystems interact in ways that undermine our ability to make predictions, in particular because of the kinds of phenomena associated with attractors and threshold changes, it is essential that anyone interested in sustainable health take a broad, ecosystemic approach. Furthermore, because of our limited ability to predict outcomes, it is important that we maintain our flexibility and our options.

In practical terms, this means that local bio- and social diversity must be maintained as a way of maintaining some reserve to respond to the certain changes ahead. An ecosystem approach to emerging diseases, then, involves, on the one hand, monitoring changes in the structures of ecosystems which are leading to the creation of different and often larger niches for potential pathogens. On the other hand, it means working to stop the most dangerous of these changes - working to influence national and international economic and political policies which are creating the disease-promoting conditions. It means doing our research in such a way that the goals such as food security, health, and a convivial life can be achieved without threatening the ecosystems which provide the services that make human communities possible.

Because systemic descriptions and resolutions are context-specific, and because we are talking about bringing together multiple systems descriptions, some argue that the amount of time, energy, money, and work required to do this research is impossible in practice to achieve. This criticism is based on outmoded ideas of both research and development. Although the ecosystem approach to health requires a major commitment of people's activities, these activities are those which should already be going on in any society: planting crops, building bridges, educating children. Within an ecosystem approach, every policy decision is a hypothesis, and every management plan is a test of that hypothesis. Research becomes an integral part of the daily decision-making and evaluating activities of households, communities, regions, and global institutions. This is precisely what makes this an adaptive approach; it does not assume that Truth - relative to sustainability and health - can be established once-and-for-all with a few definitive experiments.

We define health at our boundaries - as individuals, communities, nations and globally - and we achieve health by communicating and negotiating across our boundaries. Think ecosystemically, act holarchically, should be the rallying cry of health workers around the world. There is good evidence that the ecosystem approach can help us achieve a global vision of health, but the uncertainty and contradictions which it is attempting to accommodate must ultimately be applied back to itself. If nothing else, the ecosystem approach brings home the deep understanding that we cannot "manage" the planet for health, but we can look for opportunities to better adapt and feel at home - to be healthy - in an uncertain and contradictory world. Even if our hopes are misplaced, however, and the specific disease control and health outcomes we seek are not achieved, fundamental requirements of the ecosystem approach for open and democratic communication, tolerance, negotiation, and ecological awareness will surely have made the effort worthwhile.

Acknowledgments

I would like to thank the International Development Research Centre (Montevideo) for their financial support. I would also like to thank James Kay, Tamsyn Murray, and other members of the Dirk Gently Group (Jerry Ravetz, Silvio Funtowicz, Mario Giampietro, Bruna de Marchi, Martin O'Connor, Gilberto Gallopin, Silvia Tognetti, Henry Regier, George Francis, Nina-Marie Lister et al.) for keeping these debates alive.

References

ALLEN, T.; BANDURKSI, B. & KING, A., 1991. The Ecosystem Approach: Theory and Ecosystem Integrity. Report to the Great Lakes Advisory Board, International Joint Commission, USA and Canada.         

ALLEN, T. & HOEKSTRA, T. W., 1992. Toward a Unified Ecology. New York: Columbia University Press.         

BOCKING, S., 1994. Visions of nature and society: A history of the ecosystem concept. Alternatives, 20:12-18.         

CAST, J. L., 1994. Complexification. New York: HarperCollins Publishers.         

CHECKLAND, P. & SCHOLES, P., 1990. Soft Systems Methodology in Action. John Chichester: Wiley & Sons.         

CONSTANZA, R.; NORTON, B. & HASKELL, B., 1992. Ecosystem Health: New Goals for Environmental Management. Washington, D.C.: Island Press.         

DUBOS, R., 1965. Man Adapting. New Haven: Yale University Press.         

EVANS, R. G.; BARER, M. L. & MARMOR, T. R., 1994. Why are Some People Healthy and Others Not? The Determinants of Health in Populations. New York: Aldine de Gruyter.         

EWALD, P., 1994. Evolution of Infectious Disease. Oxford: Oxford University Press.         

FARMER, P., 1996. Social inequalities and emerging infectious diseases. Emerging Infectious Diseases, 2:259-269.         

FUNTOWICZ, S. & RAVETZ, J., 1994. Emergent complex systems. Futures, 26:568-582.         

GIAMPIETRO, M., 1994. Sustainability and technological development in agriculture: A critical appraisal of genetic engineering. Bioscience, 44:677-689.         

GITAU, T.; McDERMOTT, J. & WALTNER-TOEWS, D., 1997. Design and implementation of a program to assess the health of an intensively farmed highlands agroecosystem in Kenya. Epidemiologie et Santé Animale, 31-32:2-7.         

HOLLING, C. S., 1986. The resilience of terrestrial ecosystems: Local surprise and global change. In: Sustainable Development of the Biosphere (W. M. Clark & R. E. Munn, eds.), pp. 292-320, Cambridge: Cambridge University Press.         

KAY, J.; REGIER, H.; BOYLE, M. & FRANCIS, G., 1999. An ecosystem approach for sustainability: Addressing the challenge of complexity. Futures, 31:721-742.         

LAST, J. M., 1988. A Dictionary of Epidemiology. Oxford: Oxford University Press.         

LEDERBERG, J.; SHOPE, R. & OAKS, S., 1992. Emerging Infections: Microbial Threats to Health in the United States. Washington, D.C.: National Academy Press.         

MARA, D. D. & ALABASTER, G. P., 1995. An environmental classification of housing-related diseases in developing countries. Journal of Tropical Medicine and Hygiene, 98:41-51.         

MAY, R., 1977. Thresholds and breakpoints in ecosystems with a multiplicity of stable states. Nature, 269:471-477.         

MESLIN, F. X., 1997. Global aspects of emerging and potential zoonoses: A WHO perspective. Emerging Infectious Diseases, 3:223-228.         

MINKIN, S. F.; RAHMAN, R. & ISLAM, M. A., 1996. Flood control embankments and epidemic kala-azar in Bangladesh. Ecosystem Health, 2:215-226.         

MURRAY, T.; KAY, J.; WALTNER-TOEWS, D. & RAEZ-LUNA, E., 1999. Adaptive Methodology for Ecosystem Sustainability and Health (AMESH): An Introduction. Conference on Conservation Medicine, White Oak Conservation Center, Florida, April 29-May 2.         

PUCCIA, C. & LEVINS, R., 1985. Qualitative Modeling of Complex Systems. Cambridge: Harvard University Press.         

ROLING, N. G. & WAGEMAKERS, M. A., 1998. Facilitating Sustainable Agriculture. Cambridge: Cambridge University Press.         

SEVERINGHAUS, J. P. & BROOK, E. J., 1999. Abrupt climate change at the end of the last glacial period inferred from trapped air in polar ice. Science, 286:930-933.         

TAYLOR, K., 1999. Rapid climate change. American Scientist on the Web, 87 <http://www.amsci.org/ articles/99articles/taylor/html>.         

UNDP (United Nations' Development Program), 1998. Overview of Human Development Report 1998. <http://www.undp.org/wdro/e98over.htm>.         

UNDP (United Nations' Development Program), 1999. Human Development Report. Oxford: Oxford University Press.         

VanLEEUWEN, J.; NIELSEN, N. O. & WALTNER-TOEWS, D., 1998. Ecosystem health: An essential field for veterinary medicine. Journal of the American Veterinary Association, 212:53-57.         

WALTNER-TOEWS, D., 1995. Changing patterns of communicable disease: Who is turning the kaleidoscope? Perspectives in Biology and Medicine, 39:43-55.         

WALTNER-TOEWS, D., 1996. An agroecosystem perspective on foodborne illnesses. Ecosystem Health, 2:177-185.         

WALTNER-TOEWS, D., 1999. Mad Cows and Bad Berries. Alternatives Journal: Environmental Thought, Policy and Action, 25:38-44.         

WALTNER-TOEWS, D., 2000. The end of medicine: The beginning of health. Futures, 312:655-667.         

WALTNER-TOEWS, D. & WALL, E., 1997. Emergent perplexity: In search of post-normal questions for community and agroecosystem health. Social Science and Medicine, 45:1741-1749.         

WELCH, R.; COMBS, G. & DUXBURY, J., 1997. Toward a "greener" revolution. Issues in Science and Technology, Fall:50-58.         

YASSI, A.; MAS, P.; BONET, M.; TATE, R.; FERNANDEZ, N.; SPEIGEL, J. & PEREZ, M. L., 1999. Applying an ecosystem approach to determinants of health in Centro Habana. Ecosystem Health, 5:3-19.