Leave a comment

2.7 The environment

2.7
The environment

A. J. McMichael

Introduction
History of environmental health in Western society
Environmental health issues in the nineteenth century
Environmental health issues in the twentieth century
Categories of environmental exposure

Natural versus human-made environmental hazards

Local versus global environmental hazards
Contribution of environmental exposures to global burden of disease
Environmental health research: some specific issues

Scope, strategies, and policy interface

Other key issues in environmental epidemiology

Profiles of hazardous ambient environmental exposures

The urban environment

Urban populations: ecological footprints and sustainability

‘New and resurgent’ infectious diseases: a global phenomenon

Environmental health research: the wider dimensions

Intertwined relationships: environment, population, poverty, and health

Globalization, environmental impacts, and health
Challenges to science
Main forms of global environmental change

Global climate change

Stratospheric ozone depletion

Biodiversity loss and invasive species

Impairment of food-producing ecosystems

Other global environmental changes
Health as a ‘sustainable state’
Conclusion
Chapter References

Introduction
The meaning of the word ‘environment’, applied to human health, is elastic. Conventionally it refers to the various external factors that impinge on human health, via exposures that are usually shared between members of communities or whole populations and that are predominantly involuntary (that is, not under the control of individuals). The scope of ‘environmental exposures’ is usually confined to physical, chemical, and microbiological agents that are able to induce pathological effects. Occupational exposures, especially in the wage-paid workplace, are an important subset of ‘environmental exposures’—although the specialized topic of the occupational environment is usually treated separately (see Chapter 8.6).
The roles of housing quality, material circumstances, and socio-economic status in the determination of disease patterns have claimed increasing attention from epidemiologists. These considerations invite a more inclusive definition of ‘environment’, one that embraces the built environment, social and economic relations, and the patterns of living that flow from those circumstances. This comprehensive view of the environment is illustrated in Fig. 1, which also shows examples of types of health problems that arise in relation to the various facets and interactive combinations of environmental influences. While recognizing the fundamental importance of the social environment as a determinant of human health, and as a dimension of humankind’s complex ecology, this chapter will not develop that aspect further.

Fig. 1 The major components of ‘the environment’. Examples of environmental diseases and disorders are shown, emphasizing that most arise from the interplay of several aspects of the environment.

History of environmental health in Western society
Concern over environmental conditions has been a driving force in the historical evolution of public health. Indeed, the origins of the modern discipline of epidemiology are substantially grounded in studies of environmental health problems before the twentieth century, in particular: toxic contaminants in locally brewed alcoholic drinks, the patterns of cholera occurrence in London, mortality gradients between different residential areas and socio-economic groups, and some specific occupational exposures. The history of ‘environmental health’ in Western countries was, throughout those centuries, dominated by infectious diseases. That remains true in low-income countries today, where around two-fifths of all deaths are due to infectious disease.
Over the broad sweep of history, radical changes in human ecology and the pattern of contacts between civilizations have largely determined the turbulent tides of infectious diseases and the changing nutritional profiles of populations. The microbiological environment has been transformed over the past 10 000 years as, first, human societies opted for settled farming and herding, and, later, as emerging civilizations with their own distinctive disease pools made commercial and military contact with one another (McNeill 1976). The gradual ‘domestication’ of epidemic infections, via coevolutionary adaptations of microbe and human host, and the attainment of famine-free food supplies in Europe over the past seven to eight centuries laid the foundations for a healthier living environment.
Historians discern several distinct stages in Western society’s relationships to nature. During much of the seventeenth and eighteenth centuries, long-standing ideas persisted about disease arising from nature as God’s judgement on the human condition. The weight of religious authority and the inertia of folklore fostered a generally passive and fatalistic approach to environmental adversity. The wages of sin were disease. However, during those centuries new philosophical perspectives were emerging as the foundations of modern empirical science was being laid in Western Europe. Francis Bacon argued for scientific enquiry based upon empirical observation and comparison. Descartes, having differentiated the human mind from the body, propounded a reductionist framework for studying the external world—a world which, as a machine-like entity, was amenable to disassembly. Newton elucidated the laws of physical motion, light, and gravity. Other scientists revealed various other basic laws of the natural world. With the rapid growth of this Enlightenment thinking and knowledge, the rise of inductive logic, and the application of utilitarianism to the fruits of scientific enquiry, a more ‘activist’ approach to managing and changing the environment emerged.
One consequence of this more interventionist philosophy was the rise of the ‘social hygiene’ movement in Europe, originating in France in the late seventeenth century (Riley 1987). This movement ‘sought to cleanse the environment, to reduce its pathogenic properties and its capacity to promote epidemics’ (Riley 1987). Major social expenditures were required to undertake these ambitious infrastructural development and clean-up projects, ranging from the draining of marshes, the removal of urban refuse, and the improvement of roadways. Governments were persuaded, sometimes grudgingly, that such investments would lead to gains in the health of workforces and to increases in the amount of arable land—and therefore to higher productivity, more taxes, and fuller treasury coffers (Shahi et al. 1997).
Environmental health issues in the nineteenth century
Following the convulsion of the French Revolution at the end of the eighteenth century, more humane and egalitarian social ideologies emerged early in the nineteenth century. It was increasingly recognized that the well being and health of populations were affected by their social and physical environment—and that often there were resultant infectious disease risks throughout the social ranks. There was much talk of ‘miasmas’ (foul emanations arising from decay and putrefaction), especially within urban environments. Following the crisis of urban industrial blight and increased mortality in the 1830s in Britain, the Sanitary Idea emerged and became, temporarily, linked with ideas of urban sustainability—including ideas of recycling sewage, maintaining fertile adjoining soils, attaining local self-sufficiency in food production, and achieving full employment. The proposal for ‘garden cities’ became popular in England, partly inspired by the prospect of minimizing miasmas that were presumed to arise from crowded polluted dark urban environments. This new belief in the possibilities for enlightened collective action, for the general technocratic management of nature, challenged the inherent selfishness of the laissez-faire ideology of the age.
From around the mid-nineteenth century, however, ‘environmental health’ increasingly became a topic of formalized research, pursued by biomedical scientists. Statistics were collected on exposures, disease, and deaths. Public authorities were urged to ameliorate particular offending places to reduce the risks to health. The spectacular rise of bacteriology in the 1880s caused further divergence from the earlier ecological perspective. Microbes were deemed to be the primary cause of disease. This powerful new germ theory, along with new theories of cell biology and heredity, new concern over micronutrient deficiencies, and the medicalization of child-bearing and child-rearing all refocused the health sciences on the individual. Ideas of shared environmental exposures and their risks to health receded.
There has been protracted debate over how best to explain the declining mortality rates in Britain over the past two centuries, a mortality that was dominated by infectious diseases until the second quarter of this century. McKeown (1976) gives most of the credit to improvements in social and environmental factors, arguing in particular that gains in nutrition strengthened human biological defences against the ever-present infectious diseases of early industrial city life. The improvements in food and nutrition flowed from the modern agricultural revolution, with mechanization, diversification of food species, cross-breeding of plant and animal species to increase yields, and more efficient transport networks. Improved housing quality, safe water supplies, increasing literacy, and better domestic hygiene gave further important protection to infants and children against infectious agents. While many commentators have broadly concurred with McKeown, others have championed the role of deliberate public health interventions (Szreter 1988). In France, for example, substantial gains in life expectancy emerged first in Lyon (in the 1850s), then Paris (1860s and 1970s, albeit more protractedly), and then Marseille (around 1890) in direct association with improved public water supply and sanitation in each of those cities.
These improvements in population nutrition, in attributes of the urban environment, and general social progress fostered, beginning in Europe in the nineteenth century, the first broad-based public health revolution. Various health gains arose incidentally, as consequences of changes in human social organization and economic practices. For example, the increased supplies of cattle fodder (alfalfa and turnips) that followed the mechanization of European agriculture from the mid-eighteenth century stimulated an increase in cattle herds. Not only did this lead to a dietary increase in animal protein but it also caused a reduction in human malaria, since the anopheline mosquitoes prefer their blood-meals from bovids rather than from hominids. This unplanned health benefit was further boosted by the fact that the malaria parasite (Plasmodium) cannot complete its lifecycle in cattle as it does in humans (McNeill 1976). Thus, without the deliberate intent of public health practitioners, the population’s nutrition improved and malaria receded.
As the Industrial Revolution progressed in nineteenth-century Western populations, widespread environmental hazards occurred, including urban air pollution (William Blake’s ‘dark satanic mills’ of northern industrial England), microbiological contamination of drinking water, food contamination, and the various physical hazards of congested low-grade squalid housing. The great increase in the mobilization of lead in Europe and North America during the Industrial Revolution is shown in Fig. 2. (So too is the massive effect of the last Ice Age on lead mobilization, via disturbance of geological structures, and the brief but acute spike in lead usage and dissemination during the Classical Greek and, in particular, Roman eras.) Later in the nineteenth century, miasmatic theories of environmental disease causation were replaced by the specific-causation ideas of the germ theory. This was reinforced around the turn of the century by the identification of particular health disorders due to specific occupational exposures such as pitchblende (uranium oxide ore) mining and working in the dyestuffs industry, and by recognition of specific micronutrient deficiencies. Hence the notion of specific causation of specific diseases became dominant.

Fig. 2 Variations in the amounts of lead mobilized into the environment, as reflected in Greenland ice-core concentrations over the past 20 000 years. Note the effects of glaciation, the extensive use of lead during the classical Roman era, and the rise of environmental lead levels in association with the industrial revolution over the past two centuries. (Data from Delmas and Legrand 1998.)

Environmental health issues in the twentieth century
The germ theory, which was later qualified by an increasing appreciation of the modulating influences of environmental conditions and endogenous host characteristics, held sway during the early decades of the twentieth century. New knowledge about the vector-based transmission of major diseases such as malaria, schistosomiasis, dengue fever, yellow fever, trypanosomiasis (African sleeping sickness), leishmaniasis (kala-azar), and onchocerciasis (river blindness) led to new environmental management strategies, predominantly in the tropical and subtropical regions. This included the spraying of mosquito breeding sites with dichlorodiphenyltrichloroethane (DDT), the control of surface water, the curtailment of certain types of vegetation, and the control of alternative mammalian host species.
The spread of industry and motorized transport systems in the early part of the twentieth century hugely increased the inventory of human-made environmental chemical exposures. As noted in Fig. 2, it also contributed to the marked rise in environmental lead exposures. Various notorious urban air pollution episodes occurred during the middle decades of the century. These were an important stimulus to the new generation of environmental legislation in Western countries during the 1960s and 1970s. In the latter half of the twentieth century, led particularly by environmental regulatory agencies in the United States and drawing on the predominant model of specific causation, a reductionist and quantitative mode of research and risk management has prevailed. The approach has been essentially of an itemizing kind, with emphasis on identifying specific diseases attributable to specific environmental exposures. In the light of that research and the ensuing quantitative risk assessments, official exposure standards have been set for a long succession of specific environmental agents. The task has proved endless.
Meanwhile, in the latter third of the twentieth century we became aware of an additional dimension of hazard resulting from human-made environmental contaminants. In her influential book Silent Spring, Carson (1962) argued that pervasive forms of environmental contamination by non-biodegradable bioaccumulating pesticides were ecologically damaging. Non-human species and whole ecosystems were being adversely affected; humans, argued Carson, were therefore on notice that such chemicals could have ripple effects throughout nature, eventually impinging on human biology and well being.
In the 1970s we became increasingly aware of acid rain. Environmental health hazards were thus transcending landscapes, crossing boundaries, and growing in scale. In the 1980s, suspicions arose about the health consequences of cumulative exposures to various families of environmental contaminants, especially the long-lived (‘residual’) chlorinated hydrocarbons and several of the heavy metals. Evidence from animal species, if not from humans, indicated that such exposures could disrupt the workings of the immune system, reproductive system, and neurological system. This increasing emphasis on thinking about ecological and biological systems, within the framework of ‘environmental health’, were harbingers of today’s systems-oriented concerns about larger-scale environmental change, ecological disruption, and their impacts on human population health.
Categories of environmental exposure
External environmental exposures can be conveniently classified in a two-by-two table, differentiating exposures according to whether they are natural or human-made phenomena, or whether they are of local or global scale. Our modern preoccupation is with human-made environmental hazards. Historically, however, preindustrial concerns focused more on aspects of the natural environment. This was memorably encapsulated by Hippocrates, writing nearly two and a half millennia ago in Airs, Waters, Places:
[C]onsider the seasons of the year, and what effects each of them produces . . . then the winds, the hot and the cold, especially such as are common to all countries, and then such as are peculiar to each locality . . . concerning the waters which people use, whether they be marshy and soft, or hard and running from elevated and rocky situations, and then if saltish and unfit for cooking . . .
Natural versus human-made environmental hazards
Natural environmental influences remain relevant today. They include extremes of weather, locally circulating infectious agents, physical disasters, and local micronutrient deficiencies that reflect soil composition. For example, almost one-fifth of the world population lives on ancient, leached, and often mountainous soils which are deficient in iodine. This puts many such populations at risk of dietary iodine deficiency disorders, including goitre, reproductive impairment, and congenital disorders including cretinism, deafness, and neuromuscular disorders (Hetzel and Pandav 1994). Likewise, there are pockets of exposure to selenium deficiency in China and elsewhere, causing Keshan disease of the heart muscle in young adults and Kashin–Beck disease of bones and joints in older people (Appleton et al. 1996). In recent years it has become apparent that there is widespread exposure to arsenic from both soil and water, causing skin lesions, cardiovascular disorders, and various cancers in south-west Taiwan, the Obuasi region of Ghana, and parts of South America. The problem of arsenic-containing groundwater has reached extreme proportions in Bangladesh and West Bengal where there is increasing reliance on deep tubewells—as an alternative to faecally contaminated surface waters (Mazumder et al. 1998).
While most of these natural environmental health hazards can be defined and studied on a local scale, some have a larger dimension. For example, the quasi-periodic El Niño events, occurring every 5 to 7 years, entail a worldwide perturbation of climatic patterns that originate in natural oceanic-atmospheric fluctuations in the eastern Pacific region. These events often lead to environmental disasters and disturbances which, conditional on the vulnerability profile of local populations, pose various physical, microbiological, and other types of hazards to human health (Epstein 1999; Kovats et al. 1999).
The distinction between natural and human-made environmental hazards is not always clear. Several experiences in India in recent decades are illustrative, where natural environmental/nutritional health hazards have unexpectedly improved or become worse following human interventions in the wider environment (Gopalan 1999). Such interventions can disturb geochemical processes in soil and water, causing altered human ingestion of certain metals and trace elements. Pellagra (tryptophan deficiency) was alleviated in the 1980s by basic, economically driven changes in India’s production and market price of alternative cereal grains. Wheat and, to a lesser extent, rice substantially displaced pulses and jowar. Meanwhile, goitre and intellectual stunting due to dietary iodine deficiency increased their geographic range because the spread of irrigation and the expansion of sugar-cane production both depleted soil iodine levels.
Local versus global environmental hazards
Today, the topic of ‘environmental health’ must accommodate a further larger-scale dimension of external environmental influence. This type of health hazard arises from the disruption of the Earth’s ecological and geophysical systems and processes. This disruption jeopardizes the flow of nature’s ‘goods and services’: climatic stability, food yields, the supply of clean and fresh water, and the healthy functioning of biotically diverse natural ecosystems that recyle nutrients, cleanse air and water, and produce useful materials (Daily 1997). The disruption of these systems can affect population health in more diverse, and often less direct and less immediate, ways than those of specific conventional environmental hazards that pose direct local hazards via injury, toxicity, nutritional deficiency, or infection (Last 1992).
The distinction between these two categories of environmental influences upon health is perhaps best characterized as a difference in the scale of environmental change and in the immediacy and directness of action. The conventional ‘environmental health’ focus of epidemiologists and public has been on hazards that arise from local human-made environmental indiscretions. In the popular view, prototypical environmental health events include the disasters of Chernobyl, Bhopal, Seveso, Love Canal, Minamata Bay, and the London smog of 1952. In contrast, global environmental changes entail disruptions of complex ecological and geophysical systems.
In industrialized countries attention over the past 50 years has been directed predominantly to the plethora of chemical contaminants entering air, water, soil, and food, along with various physical hazards such as ionizing radiation, non-ionizing radiation, urban noise, and road trauma. As technologies evolve and as levels of consumption rise, the list of candidate hazards seems endless. In the 1990s we began to worry about the cancer hazard of electromagnetic radiation from mobile phones, about the risk to the fetus from chlorinated organic chemicals that form in chlorine-treated water supplies, and about the possible toxicity, allergenic, and other consequences of genetically modified foods. In low-income countries, the major environmental concerns continue to be the microbiological quality of drinking water and food, the physical safety of housing and work, indoor air pollution, and traffic hazards.
Understandably, most environmental epidemiological research continues to be focused upon these specific direct-acting environmental hazards, and usually within a localized setting. The specificity of both exposure and outcome, and the apparent directness of the causal relationship, are amenable to investigation with the existing repertoire of epidemiological study designs, and these enable the relationship between varied levels of individual or group exposure and the probability of some specified health outcome to be determined. Dose–response relationships are described, usually via statistical modelling. Then, if the causal interpretation is convincing, the estimated dose–response relationship can be used to guide directly the setting of standards and regulations. There is now a need to extend the repertoire of epidemiological research and risk assessment methods to accommodate the study of the health impacts of larger-scale environmental change and ecological disruption.
Contribution of environmental exposures to global burden of disease
The relative importance of ‘environmental’ exposures as a cause of human disease and premature death remains contentious. In industrialized countries the main concern in recent decades has been with the succession of chemical contaminants entering the air, water, soil, and food, and with various physical hazards such as ionizing radiation, non-ionizing radiation (electricity transmission), urban noise, and road trauma. Typical examples of quantifiable environmental health relationships include those between benzene exposure and leukaemia, blood lead concentration and IQ, and tropospheric ozone and asthma attacks. In less developed countries, the persisting environmental concerns are with the microbiological quality of drinking water and food, the physical safety of housing and work, indoor air pollution, exposure to the elements (especially during extreme weather events), and the hazards on chaotic local roads.
Assessing the contribution of environmental exposures to the global burden of disease is difficult for several reasons. Firstly, our knowledge about disease aetiology is incomplete. Secondly, the environmental exposure being assessed is often a moving target: many of today’s diseases are primarily the result of yesterday’s exposures. Consider, for example, how the mixture of urban air pollution, ever changing over time as technologies and transport systems evolve, is likely to affect the incidence of chronic respiratory disease over the course of several decades. Thirdly, environmental exposures affect not just the occurrence of disease, but may also affect the subsequent clinical management and eventual health outcome.
Depending on the definitions and assumptions used, estimates of the environmental contribution to the global burden of disease vary. An assessment conducted for the fifth anniversary of the Rio de Janeiro Earth Summit estimated that about 25 per cent of the global burden of disease and premature death, as measured in disability-adjusted life years, was caused by environmental hazards, including the workplace environment (WHO 1997). Smith et al. (1999), in a recent comprehensive analysis that encompassed disease initiation, progression, and case outcome, estimate that 25 to 33 per cent of the global burden of disease and premature death is attributable to direct environmental risk factors. Using published dose–response data for major exposure–disease relationships, Murray and Lopez (1999) have estimated (Table 1) that around one-quarter of the global burden of disease is due to ‘environmental exposures’ (at all ages), including around one-sixth of the total burden in young people (birth to 14 years).

Table 1 Estimated proportion of global burden of illness, injury, and premature mortality attributable to environmental exposures, measured with a common unit (disability-adjusted life year (DALY))

Some caution is needed in the choice of the health risk coefficients that are incorporated in this type of calculation. In particular, there are some environmental exposures that vary on both a short-term and long-term basis. Ambient temperature is one such exposure: it varies on a daily basis, a seasonal basis, and (with incipient climate change) on a decadal scale. Likewise the concentrations of various air pollutants vary over short periods. This day-to-day variation in exposure levels presents a challenge to environmental epidemiologists: How do the biological impacts of acute fluctuations in exposure relate to the induction of chronic disease by prolonged exposure to above-average levels?
The answer is that we do not yet really know. In general, it is easier to estimate the health risks associated with acute fluctuations in the exposure than to study the long-term health consequences of sustained exposure. Hence, daily time-series studies within single populations, particularly studies of mortality, have become prominent within air pollution epidemiology. The regression-based risk coefficients from these studies have then been widely used to estimate the excess annual mortality within a population with a specified average level of air pollution. Yet such calculations are inappropriate (McMichael et al. 1998), since daily time-series data provide no direct information about the extent of life-shortening associated with the excess daily deaths (many of which result from exacerbation of well-advanced disease, especially cardiovascular disease). Therefore, such data cannot contribute to the estimation of the effects of prolonged exposure to air pollution upon chronic disease incidence and death rates—even though that latter category of effect is of much greater public health importance. (Indeed, a comparative risk assessment carried out for The Netherlands (de Hollander et al. 1999) estimates that the health-impairing effect of chronic exposure to air-borne particulates is at least an order of magnitude greater than the effect of short-term exposures.) The long-term effects are best estimated via cohort studies. Time-series studies can identify the acute toxic effects, and can assist in identifying the most noxious pollutants, but they cannot quantify the long-term health impacts of air pollution.
Before reviewing the scope of local, and then global, environmental health hazards, the major research strategies, and some of the distinctive needs of this topic area, will be summarized.
Environmental health research: some specific issues
Scope, strategies, and policy interface
The prime task of environmental epidemiological research is to elucidate causal relations between environmental exposures and impaired states of health. It also seeks to quantify risks to populations, to develop appropriate interventions to reduce environmental risks to health, and to evaluate the effectiveness of such intervention. Epidemiology is the basic quantitative science of environmental health research. The concepts and methods of epidemiology are addressed in detail in Part 6. In essence, epidemiological research describes and explains variations and temporal changes in the pattern of illness and disease between and within populations. Most environmental epidemiology is observational (non-experimental), and this imparts certain well-known challenges to research design and data interpretation. Where health benefit is anticipated from interventions that reduce environmental exposures, experimental studies can be carried out.
Historically, epidemiology has played a crucial and largely self-sufficient role in identifying the environmental health hazards posed by relatively high levels of exposure, such as heavy air pollution (e.g. the London smog of 1952), heavy metals (especially in the occupational setting) in air, water, and food, solar ultraviolet irradiation, and environmental tobacco smoke. Those early studies were mostly done in industrializing countries, where research expertise existed and where there were technical and information resources available. Increasingly, the spotlight of attention is shifting to environmental exposures in low-income countries and in former communist countries undergoing rapid social and economic transition (often with a background of extensive environmental pollution and degradation). In low-income countries there is commonly a compounded range of environmental exposures. There are environmental health hazards arising from traditional practices (such as the use of solid fuels and biomass for domestic heating and cooking, and the consumption of fungal toxins in stored foods in tropical regions); those that continue to arise from microbiological contamination of water, food, and soil; and those that are now emerging via industrialization, the extension of mining and oil extraction, forest clearance and irrigation, urbanization, and the rapid increase in private car ownership.
Many environmental exposures occur at dose rates and total doses that are low compared to occupational exposures and personal habits such as cigarette smoking. This situation presents the epidemiologist with the difficult task of detecting modest increments in risk. Yet the importance of these external environmental exposures is threefold.

1.
The exposures often impinge on a large proportion of people within the population, thereby resulting in a large aggregate health impact (this is a socio-economic criterion), i.e. while the individual ‘risk’ is small, the population ‘effect’ may not be.

2.
The exposures are encountered on an essentially involuntary, and often unequal, basis (an ethical criterion).

3.
The exposures are amenable to control at source (a practical criterion).
There are several other characteristic features of environmental epidemiological research. Firstly, many types of environmental exposures (such as drinking-water fluoride levels or urban air pollution levels) impinge fairly evenly over whole communities, which makes the individual-level comparison of health impacts difficult. Secondly, any one particular environmental exposure is likely to be accompanied by other environmental exposures. This results, at least, in the need to take account of confounding effects—and the combined exposures may mean that there are significant biological interactions. Thirdly, modern understanding of the subtle and complex processes of development and functioning of various organ systems—especially the central nervous system, immune system, and reproductive system—have led to an awareness of how environmental exposures may induce ‘subclinical’ impairment of those systems.
In the light of these complexities environmental health research increasingly is undertaken in an interdisciplinary fashion. For example, research into the effect of low-level environmental lead exposure on the cognitive development of young children has required the integrated consideration of the results of epidemiological studies, animal experimental research, and neuropathological and molecular toxicological studies. The development of molecular biology over the past several decades has yielded many new techniques for measuring ‘internal’ exposure, especially in relation to carcinogenesis. Molecular biological markers are also useful for measuring putative biological mechanistic phenomena (which strengthens the basis for causal inference) and, in some situations, as early preclinical biological outcomes.
Other key issues in environmental epidemiology
Units of ‘exposure’: collective versus individual exposure
Environmental exposures, as discussed above, often impinge approximately equally on most or all members of a community. Hence, many epidemiological studies have depended on comparing groups of people defined by area of residence, city, or occupation. Even where it is likely that there is moderate exposure difference between individuals, because of interindividual behavioural variation, it is often difficult in practice to estimate those exposure differences. Notwithstanding the implications of many textbooks, the reflex discounting of population-level (‘ecological’) studies is inappropriate. There is impressive historical precedent for the informativeness of such an approach, such as Snow’s comparison of cholera death rates in adjoining suburbs in mid-nineteenth century London (Snow 1855). Furthermore, various environmental epidemiological questions are essentially of a ‘population’ kind. For example: How does the mortality impact of heatwaves differ between coastal and midcontinental urban populations? Are migrant populations at increased risk of childhood leukaemia because of their encounter with unfamiliar local viruses?
Time lags between exposure and outcome
A recurring difficulty in environmental epidemiology, especially in relation to chronic disease outcomes, is that many exposures change over time. In effect, today’s health events may be primarily the result of exposure in earlier decades. It is well documented, for example, that the composition and level of air pollution has changed in most cities over recent decades. Therefore cross-sectional correlations of levels of community exposure to air pollution and disease rates may be misleading. The phenomenon of temporal trends in exposure variables is not limited to the environmental realm; the same applies to dietary habits, oral contraceptive formulation, and so on. However, there is a particular temptation in environmental epidemiology to examine readily available cross-sectional data.
The time-lag question also arises in relation to the study of acute effects. Is the mortality toll of a heatwave greatest on the day of maximum temperature, or one or two days later? To answer this sort of question it is necessary to test a succession of different short-term lag periods.
Total exposure assessment
To estimate the health risk associated with an environmental exposure agent it is necessary to consider all the ways it might reach people. Some pollutants can readily reach people through several different routes. Therefore, if exposure assessment is based on environmental sampling it may be necessary to sample several media. For example, airborne lead pollution, arising mainly from leaded fuel in motor vehicles, can spread through the environment via air, water, soil, and food. Even though the original emissions of lead may have only been to air, exclusive attention to that route would greatly underestimate the actual total exposure—and it is the latter that determines the risk to health.
This problem can either be tackled by multimedia measurements or by the use of some integrating biological measure. Examples of the latter include blood lead concentration, adipose tissue dioxin concentration, and breast-milk concentrations of polychlorinated biphenyls.
Spatial analysis
Many of the relationships being investigated in environmental epidemiology depend on the analysis of spatial relations. Simple techniques entail, for example, the use of concentric circles around point sources of environmental pollutants or of residential distance from main highways. As the amount and quality of information on the spatial distribution of the exposure hazard, and of its sources, increases, along with the equivalent information on the distribution of health outcome measures, so the possibilities increase for formal spatial analysis (Lawson et al. 1999). Geographic information systems afford an analytic technique for identifying spatial relationships via the integration of ‘layers’ of spatial data about one or more exposures, other confounding or modifying factors, and health outcomes. In situations that entail a mix of input variables of quantitative, ordered, and nominal types, multivariate ordination techniques such as non-metric multidimensional scaling ordination are becoming increasingly widely used. A range of research strategies has evolved over the past decade for the study of ‘small-area’ environmental epidemiological statistical analysis (Elliott et al. 1992). Improved techniques for smoothing data and for dealing with autocorrelation and outlier observations are now available. The use of these various techniques has been facilitated by the rapid increases in computing power.
There is a largely untapped wealth of historical and current spatial environmental data, much of it from satellite remote-sensing. There is the opportunity, indeed need, for the public health sciences to become more integrated with the relevant scientific networks, and to learn about and access these resources. Global ‘observing systems’ were created during the 1990s, with detailed data on the world’s land surfaces (terrestrial systems), oceans, and atmosphere. These are auspiced within the United Nations family of organizations. Other regional or national environmental data sets are also available, particularly from the well-resourced American agencies such as the National Aeronautical and Space Administration and the National Oceanic and Atmospheric Administration.
Specific toxicity versus generic organ system effects
Most toxicological studies and most environmental epidemiological studies aim to characterize effects and risks associated with specific exposure agents. When such exposures yield specific avoidable clinical outcomes (e.g. bladder cancer, renal failure, encephalopathy), then this approach makes good sense. However, there has been increasing recognition that certain organ systems can be cumulatively, adversely, affected by multiple exposures over time. Such exposures can result in immune system suppression, endocrine disruption, and cognitive impairment.
Both the immune system and the endocrine system entail complex interactive networks of organs, cells, and chemical messengers. It is not surprising that many exogenous organic chemicals can cause metabolic perturbation of these systems. Those are not so much ‘toxic’ as ‘pharmacological’ effects. There is accruing evidence that implicates organochlorine pesticides and other organic chemicals such as butyltin in suppression of the human immune system (WRI 1998; Whalen et al. 1999). There is suggestive, but inconclusive, evidence of a decline in sperm count over the past 50 years (Sharpe and Skakkebaek 1993), although the interpretation of the evidence is hampered by the constituent datasets being neither representative nor standardized. Plausibility comes from various observations of impaired fertility and reproduction in other mammals, birds, and fish.
There has been specific concern about the possible involvement of endocrine-disrupting xeno-oestrogens in breast cancer in women. Some organochlorine compounds such as DDT (especially its metabolite dichlorodiphenyldichloroethylene) may have weak oestrogenic effects and are therefore suspected of increasing the risk of hormone-dependent cancers, particularly breast cancer in women (Davis et al. 1998). Breast cancer is the most common cancer among women in many Western countries and most of the major risk factors for breast cancer, such as early menarche, late menopause, nulliparity, late conception of the firstborn, and hormone replacement treatment, suggest that oestrogen is important in the pathogenesis of breast cancer. However, the results from studies of breast cancer in relation to environmental exposures to organochlorine compounds have been inconsistent.
Risk assessment: the research–policy interface
The increasing awareness within educated communities about the potential health risks posed by ambient environmental exposures has brought an increased expectation that governments will assess risks, prescribe standards, and ensure environmental risk management. Epidemiologists have therefore become increasingly engaged at this research–policy interface, seeking to summarize the range of published research findings, to derive dose–response relationships, and to identify critical levels of exposure.
Methods of quantitative risk assessment have evolved over the past two decades, particularly within the United States—and particularly in relation to exposure agents deemed to be causes of cancer (Samet et al. 1998; Corvalan et al. 1999; Nurminen et al. 1999). Proponents proclaim the merits of quantitative risk assessment as a successful social application of otherwise disparate results of a myriad of epidemiological and toxicological studies. Critics indicate the problems in averaging across epidemiological studies, the uncertainty of the form of dose–response functions, the difficulties of extrapolating between dose ranges and species, and the questionable assumption that single factors act independently of one another. The lack, or imprecision, of epidemiological data within parts of the exposure range of interest often necessitates supplementation with, or reliance on, animal toxicological data (McMichael and Woodward 1999).
Quantitative risk assessment, in its fullest form, enables the current or future burden of disease attributable to a particular profile of exposure, within an entire population, to be estimated. Thus, for a specified urban population, we may estimate how many episodes of asthma are attributable to an annual pattern of daily fluctuations in specified air pollutants. Or we may estimate the total sick days resulting from that asthma, or the total loss of productive working days. It is then possible to extend the analysis to estimate the economic costs to the population. Indeed, if the preventability of asthma is sufficiently well understood, cost–benefit analysis can titrate the ‘savings’ from averted asthma episodes against the costs of reducing the pollutant levels. This extension of quantitative risk assessment into the realm of cost–benefit analysis has highlighted the need for a common metric, applicable to disparate health outcomes. The disability-adjusted life year has recently been promoted and widely used as one such common metric that enables ‘comparative risk assessment’ (Murray and Lopez 1999). A recent example of how such an approach can facilitate the comparison of environmental risks and guide the choice of environmental interventions has been published by de Hollander et al. (1999), showing, for example, that the long-term effects of particulate air pollution account for almost 60 per cent of the total environment-attributable health loss in The Netherlands.
There is an important extension to this item. As we recognize the prospect of large-scale environmental declines that are likely to jeopardize the well being and health of future generations, so we encounter an ethical problem. How do we weigh the health needs of future generations against the manifest needs of the present generation? Unavoidably, there must be a trade-off since generalized and globally equitable economic development today, based on current technologies and patterns of consumption, will greatly increase the pressures on the biosphere. Indeed, some would argue that unless it is shown that obligations to future generations differ qualitatively from those to the present generation, our undoubted first priority must be to deal with existing problems—buoyed by the hope that solutions to future sustainability will emerge. The argument is morally and philosophically complex, but it is one that will be eased by improved assessments of the range and magnitude of likely future impacts on human population health. As ever, social policy-making will be assisted and enhanced by fuller information and more extensive risk assessments.
Profiles of hazardous ambient environmental exposures
Overview
The advent of the new environmentalism in Western societies, during the latter third of the twentieth century, has been associated with widespread reductions in various forms of environmental pollution. Sulphate and particulate air pollution has declined, largely in response to legislative and regulatory initiatives; likewise acid rain has been curbed, the use and release of heavy metals has decreased, pesticide use has been constrained (and DDT banned), and the nuclear power industry has become more tightly controlled. These were the sorts of environmental issues that were prominent at the agenda-setting United Nations Conference on the Human Environment in Stockholm in 1972.
Meanwhile, car usage and urban transport systems in general, and hence exhaust emissions, have proliferated. The rise of photochemical-oxidant air pollutants (especially ozone), and the increase in fine particulates (especially from diesel engines) have posed particular risks to cardiovascular and respiratory health. Food production methods have tended to become intensified and the long-distance transport of fresh and processed foods has increased. Both developments have been associated with an apparent rise in the occurrence of episodes of microbiological and chemical contamination of food (McMichael 1999a). The occurrence of bovine spongiform encephalopathy (‘mad cow disease’) in the United Kingdom, and its subsequent transmission to humans as new variant Creutzfeldt–Jakob disease, illustrates well the sort of unexpected public health consequence that can arise from aberrant methods of intensified livestock production.
In low-income countries the profile of environmental exposures is more mixed, reflecting the ‘old’ and the ‘new’. Microbiological hazards, especially in drinking water, remain widespread in rural populations, shanty towns, and urban slums: approximately 40 per cent of the world’s population still lack safe drinking water and 60 per cent lack sanitation. Domestic air pollution levels are often high, especially where biomass fuels or coal are used for heating and cooking. In consequence, high rates of infant and child mortality from diarrhoeal diseases and acute respiratory infections persist (Wang and Smith 1999). Meanwhile, as industries proliferate, as chemical-intensive agriculture spreads, and as cities fill with cars, trucks, and buses, so the ambient environment acquires a range of additional physical, air-borne, water-borne, and food-borne hazardous exposures. The combination of population pressures and economic intensification is placing increasing stresses on local environments and these result in both immediate environmental hazards and in the longer-term depletion of natural resource stocks, alteration of physicogeochemical cycles (e.g. the public health problem of arsenic-contaminated groundwater from deep tubewells sunk throughout Bangladesh and West Bengal), and disruption of ecological systems.
As we enter the twenty-first century, more than half of all people are living in urban environments. This statistic will continue to rise during the coming several decades as Homo sapiens becomes an urbanized species. It is important to think about the urban environment as a ‘habitat’, as a system of interacting conditions, exposures, and processes—physical, chemical, and microbiological; demographic, social, and cultural. It will suffice, in this chapter, to mention just a few of the distinguishing features of the ever-evolving urban environment.
The urban environment
Contemporary sources of hazards
The modern urban configuration—which is both variable and evolving in countries around the world—comprises industrial activities, concentrated transport systems, intensive waste generation, and fluid and often novel patterns of social relations and interactions. This configuration poses various environmental risks to health. Some of the risks may be overt, as with road trauma or the increase in asthma hospital admissions during air pollution crises. Others, however, are non-acute and more insidious. Environmental lead exposure, which blunts young children’s intelligence, is a good example of the latter type of urban exposure. The best estimate, based on cohort studies in Western populations, is that children whose blood lead concentrations during early childhood differ by around 10 µg/dl have a resultant difference of 2 to 3 IQ points, relative to an expected population mean of 100 (Tong and McMichael 1999). Such exposure differentials typically occur between the top and bottom quintiles of children within the urban environment.
The material quality of housing is another important dimension of the urban environment. It appears to be an important determinant of seasonal patterns of morbidity and mortality. For example, the older housing stock in the United Kingdom is a likely contributor to the well-documented above-average excess of winter mortality in that country relative to most other European countries. Housing quality, including dampness, may contribute to early-life exposure to fungal spores and to house dust mites, both of which are likely initiators of asthmatic predisposition in children.
The three contrasting examples of urban environmental hazards discussed below all reflect basic aspects of our urban living style and environment. They are urban air pollution, the various physical and social hazards of transport systems, and the vulnerability of inner-urban populations to heatwaves.
Urban air pollution
Urban air pollution has, in recent decades, become a worldwide public health problem. The earlier industrial/domestic air pollution from coal burning has been replaced by pollutants from motorized transport which form photochemical smog in summer and episodes of heavy haze of particulates and nitrogen oxides in winter.
Studies relating ambient air pollution levels to health risks were, until the 1970s, largely confined to examining particular extreme episodes of very high outdoors air pollution levels (e.g. the famous London smog of 1952). These episodes were associated with a marked increase in total mortality, especially cardiovascular and respiratory deaths, and with various respiratory disorders. More recently, long-term follow-up studies of populations exposed at different levels of air pollution, especially particulates, indicate that the higher the background levels of exposure the greater the mortality risk (Dockery et al. 1993; Pope et al. 1995).
In China, urban air pollution with particulates and sulphur dioxide is increasing sharply. The main source of pollution is the industrial use of coal, with its relatively high sulphur content. However, emissions from domestic cooking and heating fuels are also increasing. Indeed, nationally, the much greater health hazard is from indoor air pollution (Smith et al. 1999). The World Bank has recently assessed that morbidity and mortality in Chinese cities will increase steeply over the coming two decades. On current industrialization trends, it has been estimated that premature deaths will increase from around 200 000 to over half a million per year by 2020, while chronic bronchitis cases and annual bouts of respiratory symptoms are expected to triple (World Bank 1997).
An interesting contemporary policy application of these risk estimates is in assessing the avoidable health impacts attendant upon reduction of fossil fuel combustion undertaken to mitigate greenhouse gas emissions. Worldwide, it has been estimated that 7 million premature deaths could be avoided by 2020 if there were worldwide compliance with the level of carbon dioxide emission reduction recommended in the Kyoto Protocol (WGPHFFC 1997). Related estimations for China indicate that, if that country were to comply with the carbon dioxide emission cutbacks of the Kyoto Protocol, then by 2020 the annual avoidance of premature deaths from ambient (external) air pollution in China would be in the range of 2000 to 16 000 (Wang and Smith 2000). Furthermore, the equivalent number of avoidable deaths for the simultaneous reductions in indoor exposure (where coal is currently the main domestic fuel and exposures are often extreme) would be a vast 50 000 to 500 000. The width of those ranges reflects both the existence of alternative technological approaches to emission reductions and the uncertainties of the dose-specific risks to health.
Epidemiologists have developed a diverse and increasingly sophisticated set of methods for assessing the health impacts of air pollution. Nevertheless, the issue remains constrained by difficulties in exposure assessment, by the uncertain differentiation between acute and chronic effects, by the need to elucidate interactive effects between coexistent air pollutants that are often highly correlated, and by the fact that the profile of air pollution keeps evolving as human activity patterns change.
Transport systems
As modern cities grow in size, transport systems become a prominent feature of the environment. In various of Europe’s main cities, public transport systems were established many decades ago before the rising counterpressure of privately-owned cars. Hence there are undergound train systems in many cities, and extensive tram-car networks in a few (such as Amsterdam). Car ownership and travel has increased spectacularly over the past 50 years, in much of the world. Indeed, in the second quarter of the twentieth century, the oil, tyre, and automobile industries in the United States took deliberate collective action to buy up and dismantle most of the nation’s urban light-rail systems, thus making the future safe for the private car (Newman and Kenworthy 2000). In many of the cities of eastern and southeastern Asia, rapid unplanned growth in private car ownership has created serious environmental health problems (including the escalation of lead pollution). Road networks and highways have become dominant influences on urban topography, and they often subdivide and fracture communities. There are two sides to this ledger: privately owned cars create new opportunities and freedoms, while also creating new social and public health problems (McMichael 1996).
In addition to the well-known environmental problems of exhaust gas emissions—comprising various noxious gases that cause local air pollution, gases that contribute to acid rain, and the release of the greenhouse gas carbon dioxide—the other major public health detriments of car-based transport systems are as follows.

1.
Fatal and non-fatal injuries of car occupants, pedestrians, and cyclists. The global annual total of deaths caused by traffic now approaches 1 million, and the death rates are growing most rapidly in many of the world’s poorer countries as urban traffic proliferates.

2.
Physical disruption of neighbourhoods. This contributes to social fragmentation and isolation. In conjunction with (1) above, this physical dominance of roads and traffic diminishes levels of physical activity, particularly in young schoolchildren who are constrained from walking to school and from exploring and playing in their local residential environments. This is an important contributor to the emerging problem of obesity in urban populations everywhere.

3.
Chronic increases in noise levels, with disturbance of sleep patterns and exacerbations of social tensions.
Thermal stress: urban vulnerability and mortality
Severe heatwaves adversely affect health. This is particularly so in the centre of large cities, where temperatures may be higher than in the suburbs and the surrounding countryside, and where the relief of night-time cooling may be reduced. These manifestations of the ‘heat island’ effect are due to the heat-retaining concrete, asphalt, and masonry structures of most inner cities, the physical obstruction of cooling breezes, and the lack of parks and trees. Note that much of the problem is a micro-ecological one, reflecting the form and materials of urban design.
The 1990s was the warmest decade on record (that is, since the mid-nineteenth century), and, from proxy measures of temperature, for at least the last eight centuries. In July 1995, more than 460 extra deaths were certified as due to the effects of the extreme heatwave in Chicago in July, when temperatures reached 40°C (Semenza et al. 1996). Studies of such episodes elsewhere in North America and Europe have shown that those most vulnerable to heat-related illness and death are the elderly, the sick, and the urban poor. In the Chicago heatwave, the rate of heat-related death was much greater in those living in poorly ventilated apartment-block housing (Semenza et al. 1996). That same high pressure system subsequently affected the United Kingdom. An estimated 768 extra deaths (an approximately 10 per cent excess) occurred during the ensuing 5-day heatwave in England and Wales, relative to the equivalent period in 1993–1994 (Rooney et al. 1998). In Greater London, where daytime temperatures were higher (and where there was lesser cooling at night), mortality increased by around 15 per cent during the heatwave.
These studies have underscored the mix of determinants of mortality excess during heatwaves: the urban environment itself as part of modern human ecology, the cultural and technological adaptation to coping with heat (cities in northern Europe and northern United States cope less well than do southern cities), and the circumstances or characteristics of certain individuals that render them particularly vulnerable.
Urban populations: ecological footprints and sustainability
There is another qualitatively different dimension to the topic of urban populations and environmental health. Urban populations play the dominant role in the mounting pressures that currently jeopardize the sustainability of current human ecology. Cities have increasingly large ‘ecological footprints’ (Rees 1996). There are undoubted ecological benefits of urbanism, including economies of scale, shared use of resources, and opportunities for reuse and recycling. Equally, though, there are great ‘externalities’. Urban populations depend on food grown elsewhere, on raw materials (timber, metals, fibre, and so on) and energy sources (especially fossil fuels) extracted from elsewhere, and on disposing their voluminous wastes elsewhere.
Urban populations thus depend on the natural resources of ecosystems that, in aggregate, are vastly larger in area than the city itself. The highly urbanized Netherlands consumes resources from a total surface area 15 times larger than itself. Folke et al. (1996) have studied the renewable resource appropriations by the cities of the Baltic Sea region. The estimated consumption of resources (wood, paper, fibres, and food (including seafood)) by 29 cities depends upon a total area many hundred times greater than their combined area. Similarly, Rees (1996) has estimated that the almost half-million residents of Vancouver, Canada, occupying just 11 400 hectares, actually use the ecological output and services of 2.3 million hectares (Table 2)—a ratio of 207:1.

Table 2 Ecological footprints of Vancouver and the Lower Fraser Basin

Viewed prospectively, the sustainability of the world’s urban populations and their health thus depends on the continued productivity of, and other ‘services’ provided by, those distant ecosystems. Yet, even so, the magnitude of the environmental externalities of urbanism is growing. The externalities include massive urban contributions to the global accumulation of greenhouse gas, stratospheric ozone depletion, land degradation, local aquifer depletion, and coastal zone destruction. The urbanized developed world, with approximately one-fifth of the world population, currently contributes around three-quarters of all greenhouse gas emissions (IPCC 1996).
Simple arithmetic reveals that the Earth is not large enough to support, sustainably, a future population of 8 to 10 billion living at the level of energy and material consumption of today’s Western middle-class populations. A radical ‘greening’ of social structures, urban design, and technology, with an emphasis on energy efficiency and on reuse and recycling, is required (McMichael and Powles 1999). Shifts in literate consumer preferences and behaviours, via a market place that is socially regulated to ensure that pricing internalizes the full environmental costs, will be a necessary and important input to this process.
‘New and resurgent’ infectious diseases: a global phenomenon
Infectious diseases receded in Western countries throughout the latter nineteenth and most of the twentieth century. Initially, this was largely the result of urban sanitation, improved housing design, personal hygiene, antisepsis in clinical medicine, and the advent of vaccination. The antibiotic era consolidated this increasing suppression of infectious disease as a source of serious morbidity and mortality. However, the receding tide apparently turned during the last quarter of the twentieth century. As the turn of the century approached there was much talk of ‘new and resurgent’ infectious diseases (de Cock and Greenwood 1998).
In 1996, the annual WHO Health Report stated that: ‘Until relatively recently, the long struggle for control over infectious diseases seemed almost over. . . Far from being over, the struggle to control infectious diseases has become increasingly difficult. . . The result amounts to a global crisis’. This hyperbolic language aside, something unexpected has recently happened to patterns of infectious disease. An unusually large number of new or newly discovered infectious diseases have been recorded in the past 25 years, including rotavirus, cryptosporidiosis, legionellosis, the Ebola virus, Lyme disease, hepatitis C, HIV/AIDS, hantavirus pulmonary syndrome, Escherichia coli 0157, cholera 0139, toxic shock syndrome (staphyloccal), and others (Morse 1995; Heymann and Rodier 1997).
There are unusually large-scale influences on infectious disease patterns (Wilson 1995). Populations are becoming interconnected economically, culturally, and physically, enhancing the mixing of people, animals, and microbes from all geographical areas. Human mobility has escalated dramatically, in volume and speed, between and within countries. Long-distance trade facilitates the geographical redistribution of pests and pathogens. This has been well illustrated in recent years by the HIV pandemic, the worldwide dispersal of rodent-borne hantaviruses, and the rapid dissemination of a new epidemic strain of bacterial meningitis along routes of travel and trade (Morse 1995).
Rapid urbanization is expanding the traditional role of cities as gateways for infectious diseases. Population movement from rural areas into cities, and the amplified urban–rural, interurban, and intraurban contacts, is opening new vistas of opportunity to otherwise marginal microbes. This probably assisted the launch of the otherwise poorly-transmissible HIV/AIDS virus in the 1980s (Morse 1995). Likewise, the modern global spread of dengue and dengue haemorrhagic fever (the latter reflecting the increasing geographic overlap of the four viral serotypes) has been aided by the urban expansion of breeding sites for the Aedes mosquito vector. In addition to the ongoing microbial genetic evolution in response to antibiotic use, changes in contemporary human ecology have intensified much of the ‘microbial traffic’ and thus increased the probability that potential human pathogens will find an ecologically supportive pathway or niche (Morse 1993).
Urbanism and the relaxation of traditional cultural norms is yielding newer and freer patterns of human behaviour, including sexual activities and illicit drug use. Modern medical manoeuvres create new ecological opportunities for viruses and prions; hospital admission practices have done similarly for various bacteria (e.g. the Proteus and Pseudomonas genera). Over the past 50 years antibiotics have been used widely and often unwisely (including for livestock and agricultural purposes), thus helping to nurture a new generation of drug-resistant organisms. Likewise, we have inadvertently bred pesticide-resistant mosquitoes, thereby facilitating the dissemination of malaria, yellow fever, dengue fever, and many other vector-borne diseases (Chapin and Wasserstrom 1981).
Infectious disease patterns are also affected by the intensification of food production and processing methods. The notorious bovine spongiform encephalopathy/Creutzfeldt–Jakob disease (‘mad cow disease’) episode in the United Kingdom is a particular example of this problem. On a more familiar front, the reported rates of food poisoning have increased in Western countries during the past two decades, and have almost doubled in the United Kingdom between the mid-1980s and mid-1990s (WHO 1997). The spread of the potentially fatal toxin-producing E. coli 0157 in North America and Europe in the mid-1990s appears to have accompanied beef imported from infected cattle in Argentina (where the rates of human infection with E. coli 0157 are reportedly higher than in Western countries).
The environmentally disruptive impact of technological innovation—whether industrial, agricultural, or medical—in the course of several decades of linear Western-style national economic development in the developing countries made clear that large-scale human interventions in the natural environment often adversely affect infectious disease patterns (Heyneman 1984). Large dams, irrigation schemes, land reclamation, road construction, and population resettlement programmes (as currently occurring in Indonesia) have often potentiated the spread of malaria, dengue fever, schistosomiasis, and trypanosomiasis (Kloos and Thompson 1979; Inhorn and Brown 1990).
Patterns of infectious diseases are widely influenced by land clearance activities in populous regions of the developing world and by the extension of irrigation. In the Sudan, for example, schistosomiasis appeared within several years of the start of the Gezira scheme, a large irrigated cotton project (Fenwick et al. 1981). Various viral haemorrhagic fevers have emerged over the past several decades as intensified land clearance and habitat disruption, especially in South America, have exposed human populations to new viruses that previously circulated exclusively within wilderness ecosystems. For example, the Junin virus, causing Argentine haemorrhagic fever, naturally infects wild rodents (the mouse Callomys callosus). However, the extensive conversion of grassland to maize cultivation in recent decades, stimulating a proliferation of the virus-bearing mice, has exposed farm-workers to this ‘new’ virus.
Meanwhile, casting a longer shadow over the future prospects for human health is the possibility that changes in the world’s climate, the continued loss of biodiversity, and the persistence of large pockets of urban and rural poverty in a market-driven global economy will sustain, even increase, the occurrence of infectious diseases. By the late twentieth century, the ecological complexion of life on Earth had begun rapidly changing. This globalization of our economic activities and culture, worldwide urbanization, increasingly rapid and long-distance mobility, growth in refugee movements, and the intensification of food-producing systems, are creating a world of increasing opportunity for established and potential human pathogens.
Environmental health research: the wider dimensions
Humankind’s unprecedented disruption of various of the Earth’s natural systems at the global level (Vitousek et al. 1997; Lubchenco 1998) reflects the combined pressure of rapidly increasing population size and a high-consumption, energy-intensive, and waste-generating economy. During the last quarter of the twentieth century we have begun to see evidence of a general weakening of the world’s life-supporting systems and processes (Loh et al. 1998; Watson et al. 1998). The resultant risks to population health pose a special research challenge.
These environmental changes, including depletion of stratospheric ozone and long-term changes in global climatic patterns, entail unusually large spatial scales. They also entail temporal scales that extend decades, or further, into the future. Some entail irreversible changes. While some direct impacts on health would result, such as the health consequences of increased floods and heatwaves due to global climate change, or increases in skin cancer due to ozone depletion, many of the impacts would result from the disruption of the ecological processes that are central to food-producing ecosystems or to the ecology of infectious-disease pathogens. Therefore a central research task is to develop mathematical models for carrying out scenario-based health risk assessments that refer to future anticipated scenarios of environmental change.
Over the past three decades, a growing appreciation of ecological relationships and the new insights into the complex interdependencies intrinsic to the biosphere have begun to reshape our understanding of the environmental and ecological influences on the well being, indeed the sustainability, of human societies. The modern agendas for environmental health research and policy now extend beyond the identification, study, and management of specific localized physicochemical and microbiological hazards. We must include the systems-based study, and the sustainable management, of the environment as a life-supporting habitat.
Intertwined relationships: environment, population, poverty, and health
The relationships between ambient environmental conditions, socio-economic circumstances, demographic change and human health are complex and multidirectional. Some of the relationships are immediate: for example, poverty today causes malnutrition tomorrow. Other relationships involve long time lags; for example, current poverty contributes to the need to clear local forests for fuel and to farm marginal lands, and these actions often lead to ecological attrition and hunger in the future. Time lags aside, there is no simple linear causal chain connecting these variables. Population pressure and poverty among rural populations often lead to land degradation and deforestation, with consequences for supplies of food and materials, and within the urban environment they predispose to many other adverse environmental exposures (physical hazards, high exposures to polluted air, and unsafe drinking water), especially in urban slum-dwellers. Poverty influences fertility rates, and vice versa. Environmental degradation often causes further impoverishment—for example, by reducing food yields, depleting fuel-wood, or potentiating destructive floods—and it can also impair health via increased exposure to vector-borne infectious disease (especially following habitat disruption), nutritional deficiencies, and toxic environmental pollutants.
In many African, Asian, and Latin American countries, the average life expectancy is 20 to 30 years less than for rich Western countries (WHO 1999). Infectious diseases remain the main killer, particularly in children below the age of 5 years. Much of this health deficit in poor developing countries reflects the widespread poverty, adverse social consequences of export-oriented economic development, and environmental adversity caused by exploitation of natural resources. The plight of sub-Saharan Africa, with its entrenched poverty and marginalization from the global economy, illustrates well these complex relationships. With two-thirds of the world’s poorest countries, trends in the region’s health, education, and material living standards have reversed in the past two decades and are continuing to fall (Logie and Benatar 1997). Meanwhile, environmental pressures have increased widely, with deforestation, desertification, and the erosion of Africa’s relatively vulnerable soils. More than half the population still lacks safe water and 70 per cent of people lack proper sanitation. Infant mortality rates are over 50 per cent higher than in the world’s other low-income, developing countries. Malaria and tuberculosis are widespread and increasing, while in parts of central, southern, and eastern Africa one in three pregnant women are HIV positive. Logie and Benatar (1997) assess that the two-way relationship between poverty and ill health erodes African economic productivity by at least one-sixth.
These statistics aside, it remains intrinsically difficult to confirm or refute the widely-assumed ‘vicious spiral’ link between poverty, environment, and health. It is certain that both poverty and environmental degradation, via independent pathways, increase the risks to health. It is also clear that there is a strong, albeit complex, relationship of income level to environmental quality. For many important environmental pollutants, as populations undergo an increase in average income there is an inverted U-shaped curve. This is known as the ‘environmental Kuznets curve’ (Grossman 1995). Initially, the pollutant loads increase. Then as wealth, literacy, and political liberalism increase, negative feedback processes ensue and societies take action to reduce the release of those environmental pollutants. However, the indices of several of the larger-scale forms of environmental degradation (such as carbon dioxide emissions) display a clear tendency to a continuing increase. These are the ‘global common’ problems, such as greenhouse gas emissions, for which there is no immediate negative feedback in terms of adverse social impact, health consequences, or market signals. The difference is illustrated in Fig. 3.

Fig. 3 The rise and fall of levels of local environmental pollutants in association with increasing population wealth (the Kuznets curve). As societies become richer and acquire higher expectations, so negative feedback via evolving policies and regulatory action leads to reductions in pollutant levels. However, this type of feedback operates less immediately and less strongly in relation to regional and global environmental changes, the resolution of which requires collective action across diverse populations and jurisdictions.

In most low-income countries, a ‘dual profile’ of health and disease is now emerging. Rapid and poorly regulated increases in extractive and manufacturing industries typically result in environmental degradation, including air and water pollution, while also increasing the rates of occupational injury and disease (Shahi et al. 1997; Pearce 1996). Meanwhile, the persistence of widespread poverty, lack of safe water and sanitation, and urban crowding ensures the continuation of infectious diseases, especially as a source of childhood mortality. Indeed, recent analyses that have sought to compare the world’s poorest people (as opposed to the conventional aggregation of the world’s poorest countries) with the richest reveal the continued predominance of infectious disease as a cause of disease burden and premature death (Farmer 1999; Gwatkin et al. 1999). Meanwhile, in the urbanizing portions of developing countries, changes in demography (increased life expectancy, decreased fertility) and environmental conditions are transforming the profile of health and disease—illustrated by the increasing prevalence of obesity, hypertension, cigarette smoking, sedentary lifestyle, exposure to ambient air pollution, and the hazards of urban traffic. Chronic non-infectious diseases are increasing alongside the persistence of infectious diseases.
At the beginning of the twenty-first century, various sub-Saharan African countries and parts of South Asia continue to face recurring or worsening ‘subsistence crises’. This situation reflects the excessive weight of ever-increasing population numbers and environmental pressures on the nation’s local resource base, unbuffered by the accrued wealth, trading connections, and political power that high-income countries have (King et al. 1995). In the likely absence of greatly accelerated economic and social development some of these disadvantaged countries face a prospect of further environmental deterioration, resource depletion, uncontrolled urbanization, persistent poverty, and continuing great public health deficits. This, in effect, is a classic Malthusian problem wherein local population needs exceed local environmental carrying capacity (King and Elliott 1996). Such a population can survive, in the short term, by ‘ecological deficit budgeting’—that is, by depleting ‘stocks’ of material, energy, and biotic resources to subsidize deficient existing ‘flows’. Furthermore, in today’s increasingly connected world, these local pressures can be partly offset with international aid and refugee flows, albeit often too late and too little. This has happened recently with Rwanda, Somalia, and parts of West Africa—and may yet occur in India, Bangladesh, and Pakistan where populations are still increasing, land pressures are mounting and environmental conditions are declining (Cassen and Visario 1999).
Meanwhile, more remarkably, the process of ecological deficit budgetting has recently assumed a global dimension (Rees 2000). On any reasonably comprehensive accounting basis it appears that meeting the needs of the current total world population, with its high levels of consumption and waste generation, depends to a substantial extent on the depletion of global stocks of resources and on the overloading of environmental ‘sinks’ (e.g. greenhouse gas accumulation in the lower atmosphere) (Loh et al. 1998; UNEP 1999). This is an extraordinarily important and unprecedented crossroads for humankind to have reached, and it has great implications for the future levels and sustainability of human population health (McMichael and Powles 1999).
We are only slowly perceiving the extent of this deficit-budget dilemma. Its resolution will require more than just a few clever technical fixes (such as the rehabilitation of nuclear power generation and the worldwide diffusion of genetically modified higher-yielding crops). While such technological advances may relieve pressures and extend the biosphere’s effective carrying capacity, a generalized and more radical ‘greening’ of technologies is needed such as, for example, methods of reducing global carbon dioxide emissions by around two-thirds (McMichael and Powles 1999). The degree of required change will not easily be achieved by the accretion of marginal technical advances. Furthermore, reliance on technical fixes is a risky approach that can lead to mishaps—or to unexpected outcomes. (For example, the chlorofluorocarbons, discovered in the 1920s as inert gases useful for refrigeration, undergo a surprising character change in the extreme cold of the stratosphere where they react with ozone molecules.) Therefore, in addition to the undoubted environmental ‘lightening’ gains that technical improvements can make, the wholesale solution of our global environmental problems level will require great collective, and equitable, action.
Social scientists regard that type of co-operation as intrinsically very difficult for human societies to achieve (Caldwell 1990). In particular, human societies generally find it difficult to achieve co-operation whenever substantive changes in social priorities, cultural values, and technological modes are required. This difficulty has been well illustrated by the complex forging and the aftermath of the Kyoto Protocol, adopted under the United Nations Framework Convention on Climate Control in 1997, and which committed developed countries to making small cuts (5 per cent on average) in their national emissions of carbon dioxide. In responding to the threat of global environmental changes, supranational collective action must overcome the self-interested rigidities of national sovereignty, strong vested corporate interests, cultural diversity, the grievances of the very poor against the minority of the very rich in an increasingly unequal world, and today’s dominant philosophies of neoliberalism, individual rights, and the superiority of the marketplace as rational arbiter of social choices. Caldwell, a historian, has observed of this particular contemporary challenge: ‘The co-operative task would require behavior that humans find most difficult: collective self-discipline in a common effort’ (Caldwell 1990).
Globalization, environmental impacts, and health
The world is undergoing a rapid and seemingly inexorable ‘connecting up’. This connectedness is occurring in the economic, political, cultural, physical, and electronic realms. The resultant changing face of economic structures, production, trade, and financial investment has various important consequences for the social and natural environments within which human populations live—and therefore for health (McMichael and Beaglehole 2000).
This contemporary process, now massive in scale, is the culmination of a longer historical process. Economic globalization has been underway for the past 500 years, predominantly under the influence of European civilization, as trade and colonialism have reshaped much of the world, and created the basis for industrialization in Western countries—and now elsewhere. Economic globalization is now much more comprehensive in its embrace of the world’s populations, and it is distinctively a creature of liberalized market forces. Governments of high-income countries, custodians of mixed-economy social democracies a short few decades ago, are now wedded via increasingly powerful international greements and regulations to the primacy of deregulated markets, to the importance of ‘free trade’, and to the need to compete in an open global economy.
One important consequence is that new pressures are being put upon various aspects of the world’s environment as the scale and reach of economic activity increases. The spread of large-scale agribusiness and monoculture crop production is exacerbating soil erosion and land degradation. Freer access to the world’s forests by logging companies is creating problems of erosion, flooding, and silting of reservoirs. The continued building of large dams not only displaces many millions of people, but creates breeding sites for various infectious disease vectors, and carries the risk of occasional massive dam-break disasters. The expansion of long-distance trade creates new opportunities for the spread of pests and pathogens, as evidenced by the introduction of the mosquito species Aedes aegypti, via cargoes of used tyres from Asia to South America, the United States, and parts of Africa (Morse 1995). More generally, the dissemination of invasive species of plants and animals around the world is disrupting ecosystems and, in many cases, affecting the prospects for local health via effects on infectious disease transmission and local food production. The spread of the water hyacinth (introduced from Brazil) over much of Lake Victoria, in eastern Africa, has amplified the breeding sites for schistosomiasis-spreading water snails and diarrhoeal bacteria (Epstein 1998).
Economic globalization, and the segmentation of the world manufacturing workforce, has created a number of local environments with high levels of hazardous occupational and residential exposures. The creation, via taxation incentives to large corporations, of special ‘export industrial zones’ in China, Brazil, and various other low-income countries has spawned chemically and physically hazardous living environments for many poorly paid urban-fringe workers and their families (La Dou 1992).
An associated consequence of this increasingly intensive globalization of economic activity is the escalation in energy use for industrial production, car-based urban transport, long-distance trade, and increased long-distance human mobility. The resultant growth in combustion of fossil fuels is occurring in a world in which the prevailing harsh competitive economic realities do not allow longer-sighted investment in the development of alternative low-carbon energy systems. Meanwhile, as human populations continue to grow in size, and the demand for both timber and pastoral land increases, so forests are logged and cleared respectively.
These expanding economic activities have led to the modern human predicament wherein, for the first time, we have begun to exceed the capacity of the biosphere’s sources and sinks. This is what underlies the various large-scale environmental changes that are now affecting the biophysical systems—the life-support systems—upon which human well being and good health depend in the long term. The following section reviews the main components of this potentially large, important, but as yet unfamiliar category of environmental health hazard.
Global environmental change and health
Human communities have for a long time continued to deplete natural resources and degrade local ecosystems (Diamond 1998). Often, the local consequences have been the restriction or recession in human numbers, and impairment of nutritional status, health, and social viability. This process is now beginning to be played out on a much larger scale. Some of the longer-term consequences for the health of human populations could be commensurately more serious. The scale of human impact on the environment has increased rapidly over the past two centuries, as human numbers have expanded and as the material intensity and energy intensity of economic activity has increased. Global economic activity increased 20-fold in the twentieth century. Meanwhile, in absolute terms, the human population has been growing faster than ever in the last 25 years, capping a remarkable fourfold increase from 1.6 to 6 billion during the twentieth century (Raleigh 1999). The last 3 billion have been added in 14, 13 and, most recently, 12 years respectively. While we remain uncertain of the Earth’s human ‘carrying capacity’ (Cohen 1995), it is expected that the world population will approximate to 9 billion by around 2050, and will probably stabilize at around 10 to 11 billion by the end of the twenty-first century.
Over the past century or so, the typical localized symptoms of human environmental impact have been urban industrial air pollution, chemical pollution of waterways, and the various manifestations of urban squalor in rich and poor countries. These local health hazards are now being supplemented with the health risks posed by changes to some of the planet’s great biophysical systems. Humankind is beginning, unintentionally and at a global level, to alter the conditions of life on Earth—even though we remain largely uncertain, even ignorant, of the long-term consequences (McMichael 1993; Vitousek et al. 1997; Watson et al. 1998). Several ambitious global assessments have estimated that we are now in significant and increasing ‘ecological deficit’, with a manifest decline in natural environmental and ecological resource stocks (Wackernagel and Rees 1996; Loh et al. 1998).
The central issue here is that the underpinnings of human health are being erturbed or depleted. The sustained good health of any population, over time, requires a stable and productive natural environment that yields assured supplies of food and fresh water, that has a relatively constant climate in which climate-sensitive physical and biological systems do not change for the worse, that retains its richness of biodiversity (a source of both present and future value), and that promotes secure livelihoods in agriculture, pastoralism, and fishing along with those in urban professions, trades, and crafts. For the human species, as a ‘social animal’ in extremis, the richness, texture, and stability of the social environment is also important to population health.
Some of these environmental stresses are likely to cause tensions between human communities, leading to conflict and hence to damage to the population’s health (Homer-Dixon 1994). For example, Ethiopia and the Sudan, upstream of Nile-dependent Egypt, increasingly need the Nile’s water for their own crop irrigation. Around the world, many other river systems are shared uneasily between neighbours in unstable regions; these include the Ganges, the Mekong, the Jordan, and the Tigris and Euphrates (Gleick 1998). Approximately 40 per cent of the world’s population, living in 80 countries, now face some level of water shortage. Thus, as we leave the most war-scarred and arms-profiteering century on record, the prospect of international conflict because of the tensions caused by environmental decline, dwindling resources, and ecological disruption continues to cast a long shadow over the prospects for human health.
Challenges to science
The assessment of the risks to population health from global environmental change require several complementary research strategies. Recent research experience in relation to both stratospheric ozone depletion and tropospheric climate change, as sources of risks to human health, is illustrative. Research into the health impacts of these environmental changes can be conducted within three domains.

1.
By reference to analogue situations which, as manifestations of existing natural climatic variability, are deemed likely to foreshadow future aspects of climate change.

2.
By seeking early evidence of changes in health risk indicators or health status occurring in response to actual climate change. Special attention should be paid to climate-sensitive early-responding phenomena.

3.
By using existing empirical knowledge and theory to conduct integrated mathematical modelling (or other forms of integrated assessment) of likely future health outcomes. This is referred to as scenario-based health risk assessment.
Most of the formal assessment of the health impacts of global environmental changes has, to date, focused on the issue of climate change. These have predominantly been assessments of potential health impacts, referring to processes anticipated to occur over future decades. However, for some types of health impact it may soon be possible to observe early changes. For example, if the trend of increasing world temperature since the 1970s continues, then statistically detectable trends in annual heatwave-attributable deaths may soon emerge. There is, thus, an important role for empirical research.
Conversely, scenario-based health risk assessment requires us to apply, via mathematical modelling, our current knowledge and theory to future environmental scenarios. Therein lies a central problem for scientists, policy-makers, and the general public. Science classically operates empirically, via observation, interpretation, replication, prediction, and, as necessary, hypothesis modification. However, having initiated an unintentional global experiment entailing large-scale environmental and ecological changes we cannot sensibly plan to wait decades for sufficient empirical evidence to enable us to describe the health consequences. That would be too great a gamble with an uncertain future. Therefore, to guide society’s transition to the future, we must carry out scenario-based health risk assessment. This must be tempered by the realization that we cannot expect to anticipate all the resultant feedbacks, threshold-based phenomena, and surprise outcomes (Levins 1995)—including under conditions in which climate change interacts with other complex biophysical and ecological systems. In these circumstances, as mentioned earlier, it is essential that policy-making comply with the Precautionary Principle, as enunciated at the 1992 United Nations Conference on Environment and Development in Rio de Janeiro. That policy states that, where the consequences of environmental change are uncertain but potentially serious, perhaps irreversible, then that scientific uncertainty does not justify delaying precautionary preventive action.
The Precautionary Principle moves our thinking towards the realm of uncertainty-based decision-making. In this realm, action is prudently taken in light of what we do not know, and often cannot expect to know, before it may be too late to take corrective action. This mode of decision-making sits in contradistinction to the hard-won edifice of evidence-based quantitative risk assessment as the familiar scientific driver of social policy. The particular and important need to understand and apply uncertainty-based decision-making arises from the increasingly large, but uncertain, risks posed by human-induced changes in large, complex, dynamic environmental systems.
Main forms of global environmental change
The two best-defined ‘global environmental changes’—the depletion of stratospheric ozone by the emission of ozone-destroying gaseous emissions (especially chlorofluorocarbons) and the accumulation of heat-trapping greenhouse gases in the lower atmosphere—each entail changes in ‘global commons’. That is, although the gaseous emissions arise from diverse localized sources, in all continents, their environmental impact is of a diffuse globalized kind. Thus these local emissions result in integrated global changes. These changes entail a range of hazards to human population health, some of which are beyond direct assessment from existing scientific knowledge (McMichael 1993); this necessarily extends the environmental health research agenda.
Global climate change
The Second Assessment Report of the United Nations Intergovernmental Panel on Climate Change (IPCC 1996) concluded that: ‘The balance of evidence suggests a discernible human influence on global climate.’ The Third Assessment Report (IPCC 2001) concluded more firmly that most of the warming (approximately 0.4 °C) since 1975 has been due to humankind’s emissions of greenhouse gases, and that this incipient climate change has begun to alter many physical systems (glaciers, sea ice, permafrost, and rainfall patterns) and many simple biotic processes (flowering time, bird nesting, insect hatching, polewards movement of animal and plant species, and crop-growing seasons). Trends in greenhouse gas emissions will, in the Intergovernmental Panel on Climate Change’s assessment, cause an increase in average world temperature of approximately 2 to 3 °C over this century.
The anticipated health risks from climate change are of different complexion from the more familiar localized environmental health risks due to toxic chemical pollutants. The health effects of climate change would encompass direct and indirect, immediate and delayed effects (McMichael and Haines 1997). While some health outcomes in some populations would be beneficial—for example, some tropical regions may become too hot for mosquitoes or other disease vector organisms, or winter cold-snaps may become milder in temperate zones—most health effects would be adverse. This conclusion follows from the fact that long-term changes in background climatic conditions would alter the functioning of various biophysical and ecological systems that underpin human population health. Thus, the main health consequences would arise from systemic disturbances to the infrastructure of the biosphere.
The anticipated direct health effects include changes in mortality and morbidity from thermal extremes, the respiratory health consequences of altered exposures to photochemical pollutants and aeroallergens (spores, moulds, and so on), and the physical hazards of any resultant increases in storms, floods, and droughts.
Indirect health effects would include alterations in the range and activity of vector-borne infectious diseases (e.g. malaria, dengue fever, and leishmaniasis—the last of these is already present in southern Europe). Predictive mathematical modelling has suggested (Fig. 4) that the geographic zone and seasonality of potential transmission of malaria, and of dengue fever, might increase in many parts of the world (IPCC 1996; McMichael and Haines 1997). In temperate Europe, climate-sensitive vector-borne infections include tick-borne encephalitis and Lyme disease.

Fig. 4 Changes, relative to the present (a), in the annual transmissibility (number of months) of falciparum malaria under two scenarios of climate change: unmitigated emissions (b) and constrained emissions leading to carbon dioxide stabilization at 550 ppm. (c). These maps have been produced by integrated mathematical modelling, which links modelled future climate scenarios to a model of how the malaria system (mosquito and Plasmodium) responds to changes in temperature and rainfall (based on McMichael et al. (1999b)). The next step in this linked modelling (not shown here) is to estimate the additional numbers of people at risk of malaria under these future climate scenarios. In future, more complex modelling will take account of estimated trends in population vulnerability as a function of social, economic, demographic, and technological changes.

Other indirect effects would include altered transmission of person-to-person infections (especially summer season food-poisoning and water-borne pathogens), the nutritional health consequences of regional changes in agricultural productivity in poorly-resourced populations, and the various physical and psychological health consequences of rising sea levels and population displacement. Diffuse public health consequences would be likely to result from migration and the loss of employment caused by the disruptive effects of climate change upon various economic sectors and vulnerable populations.
Stratospheric ozone depletion
Higher in the atmosphere, depletion of stratospheric ozone by human-made gases such as chlorofluorocarbons is already occurring. Ambient levels on the ground of ultraviolet irradiation are estimated to have increased consequently by up to 10 per cent at mid-to-high latitudes over the past two decades (Basher et al. 1994). Via the Montreal Protocol of 1987, updated in the 1990s, the release of many of these gases has been curtailed. However, a problem remains with illegitimate sales and with the escalating production of halons by China and other low-income countries temporarily exempted from the production ban.
Scenario-based modelling, integrating the processes of emissions accrual in the stratosphere, consequent ozone destruction, increased ultraviolet ray flux, and cancer induction, indicates that European and American populations will experience a 5 to 10 per cent excess in skin cancer incidence during the middle decades of this century (Slaper et al. 1996). Similar modelled projections have been made for other fair-skinned populations living in Australia and New Zealand (Martens 1998).
Biodiversity loss and invasive species
As human demand for space, materials, and food increases, so populations and species of plants and animals are being extinguished increasingly rapidly. An important consequence for humans is the disruption of ecosystems such that ‘natural goods and services’ decline (Daily 1997). Biodiversity loss also means that we are losing, prior to their discovery, many of nature’s chemicals and genes—of the kind that have already conferred enormous medical and health improvement benefits. Myers (1997) estimates that five-sixths of nature’s medicinal tropical vegetative goods have yet to be recruited for human benefit.
Meanwhile, ‘invasive’ species are spreading into new non-natural environments via intensified human food production, commerce, and mobility. These changes in regional species composition have many consequences for human health. For example, as mentioned above, the choking spread of water hyacinth in eastern Africa’s Lake Victoria, introduced from Brazil as a decorative plant, has provided a microenvironment for the proliferation of diarrhoeal disease bacteria and water snails that transmit schistosomiasis (Epstein 1998).
Impairment of food-producing ecosystems
The increasing pressures from intensified agricultural and livestock production are stressing the world’s arable lands and pastures. At the beginning of the twenty-first century an estimated one-third of the world’s previously productive land has been seriously damaged by erosion, compaction, salination, waterlogging, and chemical pollution that destroys organic content (WRI 1998). Similar pressures on the world’s ocean fisheries have left most of them severely depleted or stressed. Unless an environmentally benign and socially acceptable way of using genetic engineering to increase food yields is found, we face a future struggle to produce sufficient food for humankind.
Modelling studies that allow for future trends in trade and economic development have been used to estimate the impacts of climate change upon cereal grain yields (representing two-thirds of world food energy). Globally, a slight downturn of around 2 to 4 per cent appears likely, but this would be substantially greater in food-insecure regions in South Asia, the Middle East, North Africa, and Central America (Parry et al. 1999). Such downturns would increase the number of malnourished people in the world, which already approximates to 800 million people.
Other global environmental changes
Freshwater aquifers in all continents are being depleted of their ancient ‘fossil water’ supplies. Agricultural and industrial demand, amplified by population growth, often greatly exceeds the rate of natural recharge. Water-related political and public health crises loom in several regions within decades, including the Middle East, northern Africa, and parts of south Asia. India, which had a supply of 5500 m3 per person per year in 1950 currently has around 1800 m3 per person (close to the recognized minimum requirement), and this will fall by a further quarter over the coming 25 years (Cassen and Visario 1999). Climate change may reduce supplies further in much of India.
Various semi-volatile organic chemicals (such as polychlorinated biphenyls) are now known to be disseminated worldwide via a sequential ‘distillation’ process in the cells of the lower atmosphere, thereby transferring chemicals from their usual origins in low to mid latitudes to high, indeed, polar latitudes (Watson et al. 1998). In consequence, increasing levels of these chemicals are present in polar mammals and fish, and in traditional human groups who depend upon those food sources. Environment-polluting chemicals are no longer just an issue of local toxicity.
Finally, humankind is rapidly changing the cycling of major elements through the biosphere. Since the 1940s there has been a sixfold increase in the global annual human ‘fixation’ of nitrogen (that is, nitrogen that has been converted from inert form to biologically active nitrate and ammonium ions). This amount now exceeds the annual natural yield of ‘fixed’ nitrogen. Most of this increase derives from the use of nitrogenous fertilizers. Likewise, we have markedly altered the geochemical cycling of sulphur. By affecting the acidity and nutrient balances of the world’s soils and waterways, these two chemical changes are likely to impair the capacity of terrestrial and aquatic food-producing systems. The nitrogen overloading of the biosphere is beginning to affect patterns of vegetation growth and various crop-plant pests and diseases.
Health as a ‘sustainable state’
Population health is conventionally measured cross-sectionally, in tally-card fashion, as an ‘achieved’ entity. However, such a measure is an integration of past activities and consumption patterns; it provides no information about the possibility of sustaining those achieved levels in the future.
These generalized gains in human life expectancy reveal that, recently, the life-supporting capacity of the man-modulated environment has been increasing. But at what future cost? At what stage might depletion of the world’s ecological and biophysical capital rebound against the health of human populations? This is a difficult question for scientists to answer. In principle, the answer will reflect the extent to which the ongoing health gains reflect increases in stocks of human and social capital, as opposed to the depletion of stocks of natural capital (McMichael and Powles 1999). Since many of the Earth’s vital life-supporting systems appear now to be coming under unprecedented stress, the question assumes greater weight. If current trends continue, will long-term risks to human health emerge? Indeed these risks will tend to increase so long as human societies persist with an economy that entails a linear, waste-generating, ‘industrial’ metabolism. Such a metabolism is at odds with the circular metabolism of the rest of nature (wherein every output becomes an input.
The development of indicators of sustainability will be difficult. The uncertainties surrounding the prognostic criteria of sustainability and which are inherent in estimations of the potential health effects of ecological disruption are such that a pluralist, and sceptical, approach is desirable. As we acquire fuller understanding of the determinants of ecological sustainability we will become better placed to monitor our changing circumstances. Some types of indicators—such as the proportion of settled populations living in flood plains—would provide information about future environmental risks. However, many ‘indicators’ would only become informative after the sustainabilty threshold had been passed. Life expectancy, for example, may continue to increase if we persist in subsidizing our material needs by depleting natural capital (soil fertility, groundwater stores, biodiversity, atmospheric constancy, and so on). The challenge to public health scientists, social scientists, and environmental scientists in assisting society to foresee population health risks, and to make the transition to sustainable living successfully, is great.
Conclusion
In the Western world, in the eighteenth century, the dominant and age-old environmental health hazard facing the mass of people was malnutrition and famine, along with endemic childhood infectious diseases and recurrent epidemics. After the 1740s, in Europe, malnutrition and famines receded as the modern agricultural revolution began. The extreme urban crowding, insanitary conditions, and working-class poverty due to early industrialization in the nineteenth century resulted in infectious diseases becoming the pre-eminent environmental health hazard. Subsequently, with the rise of modern large-scale industry and of synthetic organic chemistry in the twentieth century, pollution of local air, water, soil, and food became the major focus of environmental health concern, first in developed countries and more recently in developing countries.
Thus, for the past two centuries, environmental health concerns in the Western world have focused very largely on toxicological or microbiological risks to health from specific factors within the local environment. However, the scale of environmental health hazards is now increasing in range. The escalating impact of human economic activity has begun to alter global ecological and biophysical systems (such as the climate system) which underpin the sustainability of the health of humans—and all other species. It now seems that humankind is attempting to live beyond the planet’s overall carrying capacity. This has great implications for our research agenda and methods. Local environmental health hazards remain an important focus of concern for public health research and for risk management. Meanwhile, the tools of environmental health risk assessment must be further developed, including the use of scenario-based mathematical modelling to foresee the likely health impacts from anticipated environmental changes.
In simpler times, not long past, it was possible to rely on continuing life support from the biosphere, while focusing our environmental health concerns on the hazards arising from human-made contamination or degradation of the local external environment. However, that life-supporting capacity of the global natural environment is now showing signs of declining. The scale of our environmental health agenda must therefore be extended from risks to particular local groups or communities, to risks to whole populations and, indeed, to future generations. This thinking has extra relevance as we enter the age of the genome and must decide how best to apply this new molecular genetic knowledge. As Rudolf Virchow, the famous German nineteenth-century pathologist and public health advocate, might well have said of these emerging issues—this type of environmental health will be politics writ very large indeed.
Chapter References
Appleton, J.D., Fuge, R., and McCall, G.J.H. (1996). Environmental geochemistry and health. Special publication No. 113. Geological Society, London.
Basher, R.E., Zheng, X., and Nichol, S. (1994). Ozone-related trends in solar UV-B series. Geophysical Research Letters, 21, 2713–16.
Caldwell, L.K. (1990). Between two worlds: science, the environmental movement and policy choice. Cambridge University Press.
Carson, R. (1962). Silent spring. Houghton Mifflin, New York.
Cassen, R. and Visario, P. (1999). India: looking ahead to one a half billion people. British Medical Journal, 319, 995–7.
Chapin, G. and Wasserstrom, R. (1981). Agricultural production and malaria resurgence in Central America and India. Nature, 293, 181–9.
Cohen, J. (1995). How many people can the Earth support? Norton, New York.
Corvalan, C.F., Kjellstrom, T., and Smith, K.R. (1999). Health, environment and sustainable development: identifying links and indicators to promote action. Epidemiology, 10, 656–60.
Daily, G.C. (ed.) (1997). Nature’s services. Societal dependence on natural ecosystems. Island Press, Washington, DC.
Davis, D.L., Axelrod, D., Bailey, L., Gaynor, M., and Sasco, A. (1998). Rethinking breast cancer risk and the environment: the case for the Precautionary Principle. Environmental Health Perspective, 6, 523–9.
De Cock, K. and Greenwood, B. (ed.) (1998). New and resurgent infections. Detection and management of tomorrow’s epidemics. Wiley, Chichester.
De Hollander, A.E.M., Melse, J.M., Lebret, E., and Kramers, P.G.N. (1999). An aggregate public health indicator to represent the impact of multiple environmental exposures. Epidemiology, 10, 606–17.
Delmas, R.J. and Legrand, M. (1998). Trends recorded in Greenland in relation with Northern Hemispheric anthropogenic pollution. IGBP Global Change Newsletter, 36, 14–17.
Diamond, J. (1998). Guns, germs and steel. The fate of human civilizations. Jonathan Cape, London.
Dockery, D.W., Pope, C.A., and Xu, X. (1993% environmental epidemiology: methods for small area studies. Oxf29. An association between air pollution and mortality in six US cities. New England Journal of Medicine, 329, 1753–9.
Elliott, P., Cuzick, J., English, D., and Stern, R. (ed.) (1992). Geographical andord University Press.
Epstein, P.R. (1998). Weeds bring disease to the east African waterways. Lancet, 351, 577.
Epstein, P.R. (1999). Climate and health. Science, 285, 347–8.
Farmer, P. (1999). Infections and inequalities. The modern plagues. University of California Press, Berkeley, CA.
Fenwick, A., Cheesmond, A.K., and Amin, M.A. (1981). The role of field irrigation canals in the transmission of Schistosoma mansoni in the Gezira Scheme, Sudan. Bulletin of the World Health Organization, 59, 777–86.
Folke, C., Larsson, J., and Sweitzer J. (1996). Renewable resource appropriation. In Getting down to Earth (ed. R. Costanza and O. Segura). Island Press, Washington, DC.
Gleick, P.H. (1998). The world’s water. The biennial report on freshwater resources 1998–1999. Island Press, Washington, DC.
Gopalan, C. (1999). The changing epidemiology of malnutrition in a developing society: the effect of unforeseen factors. Bulletin of the Nutrition Foundation of India, 20, 1–5.
Grossman, G. (1995). Pollution and growth: what do we know? In The economics of sustainable development (ed. I. Goldin and L.A. Winters). Cambridge University Press.
Gwatkin, D.R., Guillot, M., and Henneline, P. (1999). The burden of disease among the poor. Lancet, 354, 586–9.
Hetzel, B.S. and Pandav, C.S. (1994). SOS for a billion. The conquest of iodine deficiency disorders. Oxford University Press, Bombay.
Heymann, D.L. and Rodier, G. (1997). Reemerging pathogens and diseases out of control. Lancet, 349, 8–9.
Heyneman, D. (1984). Development and disease: a dual dilemma. Journal of Parasitology, 70, 3–17.
Homer-Dixon, T.F. (1994) Environmental scarcities and violent conflict: evidence from cases. International Security, 19, 5–40.
Inhorn, M.C. and Brown, P.J. (1990). The anthropology of infectious disease. Annual Reviews of Anthropology, 19, 89–117.
IPCC (Intergovernmental Panel on Climate Change) (1996). Climate change. Report of Working Group I, 1995 (ed. J.T. Houghton, L.G. Meira Filho, B.A. Callander, et al.). Cambridge University Press.
IPCC (Intergovernmental Panel on Climate Change) (2001). Climate change 2001: the scientific basis. Climate change 2001: impacts, adaptations and vulnerability. Cambridge University Press.
King, M. and Elliott, C. (1996). Averting a world food shortage: tighten your belts for CAIRO II. British Medical Journal, 313, 995–7.
King, M., Elliott, C., and Hellberg, H. (1995). Does demographic entrapment challenge the two-child paradigm? Health Policy and Planning, 10, 376–83.
Kloos, H. and Thompson, K. (1979). Schistosomiasis in Africa: an ecological perspective. Journal of Tropical Geography, 48, 31–46.
Kovats, R.S., Bouma, M., and Haines, A. (1999). El Niño and health. WHO/SDE/PHE/99.4, WHO, Geneva.
La Dou, J. (1992) The export of industrial hazards to developing countries. In Occupational health in developing countries (ed. J. Jeyaratnam). Oxford University Press.
Last, J.M. (1992). Global environment, health and health services. In Maxcy Rosenau Last. Public health and preventive medicine (ed. J.M. Last and R.B. Wallace), pp. 677–86. Appleton Lange, Norwalk, CT.
Lawson, A., Biggeri, A., Boehning, D., Lesaffre, E., Viel, J-F., and Bertollini, R. (1999). Disease mapping and risk assessment for public health. Wiley, Chichester.
Levins, R. (1995). Preparing for uncertainty. Ecosystem Health, 1, 47–57.
Logie, D.E. and Benatar, S.R. (1997). Africa in the 21st century: can despair be turned to hope? British Medical Journal, 315, 1444–6.
Loh, J., Randers, J., MacGillivray, A., et al. (1998). Living planet report, 1998. WWF International, Switzerland.
Lubchenco, J. (1998). Entering the century of the environment: a new social contract for science. Science, 279, 491–7.
McKeown, T. (1976). The modern rise of population. Academic Press, New York.
McMichael, A.J. (1993). Planetary overload: global environmental change and the health of the human species. Cambridge University Press.
McMichael, A.J. (1996). Transport and health: assessing the risks. In Health at the crossroads: transport policy and urban health (ed. T. Fletcher and A.J. McMichael), pp. 9–26. Wiley, Chichester.
McMichael, A.J. (1999a). Dioxins in the Belgian food chain: chickens and eggs. Journal of Epidemiology and Community Health, in press.
McMichael, A.J. (1999b). Urbanisation and urbanism in industrialised nations, 1850–present: implications for human health. In Urbanism, health and human biology in industrialised countries (ed. L. Schell and S. Ulijasek), pp. 21–45. Cambridge University Press.
McMichael, A.J. and Beaglehole, R. (2000). The changing global context of public health. Lancet, 356, 495–9.
McMichael, A.J. and Haines, A. (1997). Global climate change: the potential effects on health. British Medical Journal, 315, 805–9.
McMichael, A.J. and Powles, J.W. (1999). Human numbers, environment, sustainability and health. British Medical Journal, 319, 977–80.
McMichael, A.J. and Woodward, A.J. (1999). Quantitative estimation and prediction of human cancer risk: its history and role in cancer prevention. In Quantitative estimation and prediction of human cancer risks (ed. S. Moolgavkar, D. Krewski, and L. Zeise), pp. 1–10. IARC Scientific Publications No. 131, Oxford University Press.
McMichael, A.J., Anderson, H.R., Brunekreef, B., and Cohen, A. (1998). Inappropriate use of daily mortality analyses for estimating the longer-term mortality effects of air pollution. International Journal of Epidemiology, 27, 450–3.
McMichael, A.J., Bolin, B., Costanza, R., et al. (1999a). Globalization and the sustainability of health: an ecological perspective. BioScience, 49, 205–10.
McMichael, A.J., Kovats, R.S., Martens, W.J.M., Nijhof, S., de Vries, P., and Livermore, M. (1999b). Comparative scenarios of climate change and health: modelling future patterns of malaria and thermal stress. In Global climate change and its impacts (ed. G. Jenkins), pp. 24–7. Department of the Environment, London.
McNeill, W. (1976). Plagues and people. Doubleday, New York.
Martens, W.J.M. (1998). Health and climate change: modelling the impacts of global warming and ozone depletion. Earthscan, London.
Martens, W.J.M., Kovats, R.S., Nijhof, S., et al. (1999). Climate change and future populations at risk of malaria. Global Environmental Change, 9 (Supplement), S89–107.
Mazumder, D.N.G., Haque, R., Ghosh, N., et al. (1998). Arsenic levels in drinking water and the prevalence of skin lesions in West Bengal, India. International Journal of Epidemiology, 27, 871–7.
Morse, S.S. (1993). Examining the origins of emerging viruses. In Emerging viruses (ed. S. Morse). Oxford University Press, New York.
Morse, S.S. (1995). Factors in the emergence of infectious diseases. Emerging Infectious Diseases, 1, 7–15.
Murray, C.J. and Lopez, A. (1996). Global burden of disease. Harvard School of Public Health, Cambridge, MA.
Murray, C.J. and Lopez, A. (1999). On the comparable quantification of health risks: lessons from the Global Burden of Disease Study. Epidemiology, 10, 594–605.
Myers, N. (1997). Biodiversity’s genetic library. In Nature’s services. Societal dependence on natural ecosystems (ed. G. Daily), pp. 255–73. Island Press, Washington, DC.
Newman, P. and Kenworthy, G. (1999). Cities and sustainability: overcoming automobile dependence. Island Press, Washington, DC.
Nurminen, M., Nurminen, T., and Corvalan, C.F. (1999). Methodologic issues in epidemiologic risk assessment. Epidemiology, 10, 585–93.
Parry, M., Rosenzweig C., Iglesias, A., Fischer, G., and Livermore, M.T.J. (1999). Climate change and global food security: a new assessment. Global Environmental Change, 9 (Supplement), S51–68.
Pearce, N. (1996). Traditional epidemiology, modern epidemiology, and public health. American Journal of Public Health, 86, 678–83.
Pope, C.A., Thun, M.J., Mohan, M., et al. (1995). Particulate air pollution as a predictor of mortality in a prospective study of US adults. American Journal of Respiratory and Critical Care Medicine, 151, 669–74.
Raleigh, V.S. (1999). World population and health in transition. British Medical Journal, 319, 981–4.
Rees, W. (1996). Revisiting carrying capacity: area-based indicators of sustainability. Population and Environment, 17, 195–215.
Rees, W. (2000). Patch disturbance, ecofootprints, and biological integrity: revisiting the limits to growth (or why industrial society is inherently unsustainable). In Ecological integrity in the world’s environment and health (ed. D. Pimentel, L. Westra, and R. Noss), pp. 139–56. Island Press, Washington, DC.
Riley, J.C. (1987). The eighteenth-century campaign to avoid disease. St Martin’s Press, New York.
Rooney, C., McMichael, A.J., Kovats, R.S., and Coleman, M. (1998). Excess mortality in England and Wales during the 1995 heatwave. Journal of Epidemiology and Community Health, 52, 482–6.
Samet, J.M., Schnatter, R., and Gibb, H. (1998). Invited commentary: epidemiology and risk assessment. American Journal of Epidemiology, 148, 929–36.
Semenza, J.C., Rubin, C.H., Falter, K.H., et al. (1996). Heat-related deaths during the July 1995 heatwave in Chicago. New England Journal of Medicine, 335, 84–90.
Shahi, G.S., Chen, L., Levy, B.S., Binger, A., Kjellstrom, T., and Lawrence, R.S. (1997). A historical perspective. In International perspectives on environment, development, and health. toward a sustainable world (ed. G.S. Shahi, B.S. Levy, A. Binger, T. Kjellstrom, and R.S. Lawrence), pp. 21–47. Springer, Basel.
Sharpe, R.M. and Skakkebaek, N. (1993). Are oestrogens involved in falling sperm counts and disorders of the male reproductive tract? Lancet, 341, 1392–5.
Slaper, H., Velders, G.J.M., Daniel, J.S., de Gruijl, F.R., and van der Leun, J.C. (1996). Estimates of ozone depletion and skin cancer incidence to examine the Vienna Convention achievements. Nature, 384, 256–8.
Smith, K., Corvalan, C., and Kjellstrom, T. (1999). How much global ill health is attributable to environmental factors? Epidemiology, 10, 573–84.
Snow, J. (1855). On the mode of communication of cholera. Churchill, London.
Szreter, S. (1988). The importance of social intervention in Britain’s mortality decline c. 1850–1914: a re-interpretation of the role of public health. Social History of Medicine, 1, 1–37.
Tong, S. and McMichael, A.J. (1999). The magnitude, persistence and public health significance of cognitive effects of environmental lead exposure in childhood. Journal of Environmental Medicine, 1, 103–10.
UNEP (United Nations Environment Programme) (1999). Global environment outlook 2000. Earthscan, London.
Vitousek, P.M., Mooney, H.A., Lubchenco, J., and Melillo, J.M. (1997). Human domination of Earth’s ecosystems. Science, 277, 494–9.
Wackernagel, M. and Rees, W. (1996). Our ecological footprint. Reducing human impact on the Earth. New Society Publishers, Canada.
Wang, X. and Smith, K.R. (2000). Secondary benefits of greenhouse gas control: health impact in China. Environmental Science and Technology, 33, 3056–61.
Watson, R.T., Dixon, J.A., Hamburg, S.P., Janetos, A.C., and Moss, R.H. (1998). Protecting our planet. Securing our future. Linkages among global environmental issues and human needs. UNEP, USNASA, World Bank, Washington, DC.
WGPHFFC (Working Group on Public Health and Fossil-Fuel Combustion) (1997). Short term improvements in public health and global-climate policies on fossil-fuel combustion: an interim report. Lancet, 350, 1341–9.
Whalen, M.M., Loganahtan, B.G., and Kannan, K. (1999). Immunotoxicity of environmentally relevant concentrations of butyltin on human natural killer cells in vitro. Environmental Research, 81, 108–16.
WHO (World Health Organization) (1997). Health and environment in sustainable development. Five years after the Earth Summit. WHO, Geneva.
WHO (World Health Organization) (1999). World health report 1999. Making a difference. WHO, Geneva.
Wilson, M.E. (1995). Infectious diseases: an ecological perspective. British Medical Journal, 311, 1681–4.
World Bank (1997). Clear water, blue skies: China’s environment in the new century. World Bank, Washington, DC.
WRI (World Resources Institute) (1998). 1998–99 world resources. A guide to the global environment. environmental change and human health. Oxford University Press.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: