Contents - Previous - Next

This is the old United Nations University website. Visit the new site at http://unu.edu

Viewing disease as the synergistic interaction of host, agent, and environment

Nevin S. Scrimshaw

Abstract

The agent of disease, whether it is an infectious organism, the deficiency or excess of a nutrient, a toxin, or the expression of a gene, is not sufficient to explain its cause. The impact of the agent depends on host factors and the influence of physical, biological, and social factors in the environment. For nutritional disorders those of the social environment, including the complex of educational, economic, religious, political, and other factors, are likely to be the most important and also the most difficult to understand and modify. If nutritional and other diseases are viewed as problems in human ecology, the primary health-care approach to their prevention follows naturally.

Introduction

The aetiology of disease can only be fully understood in the context of three sets of factors, those pertaining to the host, the agent, and the environment. The latter must be interpreted broadly to include the physical, biological, and social environment. Of these, it is usually the social environment, with its complex of educational, economic, religious, political, and other factors, that determines the environmental circumstances and behaviours that turn out to be most important. Unfortunately, environmental factors are also likely to be the most difficult to understand and to modify.

Long before human societies acquired any scientific knowledge of the specific agents of disease, they evolved complex sets of explanations, associating disease with specific environmental and social factors. Although some of these have no scientific basis that we can discern, many of them appear quite rational to scientists with modern medical knowledge. Clearly, much can be learned about the epidemiology of disease from traditional beliefs.

The belief in ancient Rome that malaria was due to "bad air" from the Pontine marshes indicated an important step in the chain of multiple causation, even if it was Anopheles mosquitoes that made air dangerous. Similarly, the association between rats and plague was recognized long before identification of the bacillus and a rodent flea, its initial carrier. The consequences of eating poisonous mushrooms or fish were understood without any knowledge of the kinds of toxic compounds they contain.

Beriberi was associated with rice diets and pellagra with corn diets long before thiamine and niacin were identified as the nutrient deficiencies responsible. Nor was it necessary for physicians to understand the metabolism of nucleic acid for the nobles of seventeenth-century Europe to recognize the relationship between a diet rich in organ meats and red wine and the symptoms of gout.

Even some forms of cancer were appropriately associated with specific occupations and environments, although the mechanisms were, and to a considerable extent still are, uncertain. In the diagnosis and treatment of the diseases that we classify as psychological or psychosomatic, folk beliefs are particularly instructive, and indigenous treatments based on them may be more successful than the "scientific" therapy of contemporary Western medicine.

Thus environmental factors were often recognized to be part of the chain of causation of many specific diseases much before the responsible agents were known. With this recognition, it was sometimes possible to achieve successful prevention without knowing that a disease was caused by an infectious agent, a missing nutrient, or a toxin. Of course, diseases are also caused by psychological factors, which do not fit as neatly into the scheme of epidemiological analysis, perhaps because we do not yet know enough.

The element missing from this discussion of disease aetiology up to this point is the status of the host. The occurrence of disease cannot be accounted for simply by factors associated with the disease agent and environment alone. The same combination of these may produce serious illness or even death in one person yet have no effect on another. For example, if the presence of the tubercle bacillus were solely responsible for the occurrence of tuberculosis in nineteenth-century Europe, the disease would have been nearly universal. Instead, most individuals experienced the infection without developing the disease. Similarly, the same rich diet mentioned earlier that provoked attacks of gout in men was no threat to pre-menopausal women in the same society. Even under conditions of severe nutritional deficits of thiamine or niacin, not all individuals succumbed to the corresponding disease. Such host factors as age, sex, and physiologic status must also be taken into account.

An additional host factor, often overlooked, is individual metabolic variation. We find it almost impossible to conduct even a small nitrogen balance study concerned with the capacity of human subjects to utilize various food proteins or essential amino acids without encountering an "outlier." The range of human metabolic idiosyncrasy is almost infinite and may turn up as a prominent factor in the occurrence of a disease in unpredictable ways at almost any time.

Limitations to the identification of specific causative agents

Too often physicians think primarily in terms of an agent when identifying the cause of a disease. Unfortunately, this is perpetuated in their teaching of medical students and young house officers. Because modern clinical medicine has developed more or less effective countermeasures against the agents of most diseases, this approach usually works reasonably well in the hospital. However, it may fail miserably in preventing a relapse once patients return to the environments from which they came, and as an approach to prevention of disease in a population, it is usually ineffective.

Both the clinician and the public health physician should take into consideration all available knowledge of the factors related to an illness in an individual or a population in order to attack it at its most vulnerable point in the complex of causation, or indeed at more than one weak point. The use of accumulated, contemporary knowledge of all three categories of causative factors will make possible advances in medical science and world health that will not be achieved with any other approach. Moreover, it is only in this way that the remarkable variation in the frequency and severity of disease with time, place, and person can be accounted for.

Variations In disease with time

The remarkable decline in most of the major infections diseases of the past in Europe before either knowledge of their specific agents or the availability of any effective means of treatment or prevention has been described by McKeown [1]. Deaths from tuberculosis dropped precipitously in country after country before any effective specific therapy was available. Similar drops were seen in England and Wales for bronchitis, pneumonia, and influenza; whooping cough; measles; scarlet fever and diphtheria; smallpox; and infections of the ear, pharynx, and larynx mainly before the responsible agents were known [1]. The end result is seen in table 1. Since there is no evidence for a change in the virulence of the infectious agents, the explanation must lie in environmental and/or host factors. McKeown makes a convincing case that the principal change was improvement in a factor dependent mainly on the social environment namely, quality of the diet, which influenced the nutrition status of the host. This, in turn, increased host resistance.

TABLE 1. Standardized death rates from airborne diseases: England and Wales (per million)

In 1900, the death rate from diarrhoea! disease in infants in New York City was 4,496 per 100,000; by 1961 it was only 45. For children one to four years of age, the drop was from 470 to 2 in the same period [2]. There is still no treatment that can bring about this result. It was undoubtedly due to a number of factors, principally, improved hygiene and better nutrition.

There is much to be learned from disease period)city. Seasonal peaks of diarrhoea! disease reflect, primarily, environmental factors associated with fluctuations in temperature and rainfall. In developing countries, such peaks are often associated with seasonal increases in kwashiorkor a few weeks later. In this case, diarrhoea precipitates the syndrome in children whose nutrition status was already borderline. It exacerbates malnutrition by reducing nutrient absorbtion.. More important, diarrhoea! and other infections divert amino acids for the synthesis of immune globulins and other protein sources, and cause increased losses of amino acid nitrogen in the urine. The effect of infections on food intake through loss of appetite and "therapeutic" withdrawal of solid food are additional factors. Even though kwashiorkor is a form of severe protein-calorie malnutrition, its appearance could never be understood from ape praisal of diet alone.

Effect of place on disease

Some diseases are clearly related to place. Infections such as schistosomiasis and onchocerciasis require quite specific physical environments that are suitable for the carrier, a snail in the first case and a Simulium fly in the second. Hookworm requires warm, moist soil and also social habits that pollute the soil with human faeces and that allow bare feet to come into contact with the infected soil.

Since intestinal parasites are normally spread by contamination of soil with human farces, it was surprising for a survey in Arctic Greenland [3] to find prevalence rates comparable to those of the tropics. By single stool examination of 663 persons, the intestinal parasite infection rate was 72 per 100. The explanation was found in the tropical conditions inside the homes that allowed for both survival of the infective forms of helminths and protozoa and their ingestion by family members.

Malaria can occur only under conditions suitable for the mosquito carrier. Since there are a number of different species capable of transmitting this disease, the habitats are quite varied. As in the case of intestinal parasites, however, social environmental factors will be the ultimate determinant as to whether or not malaria occurs in a given place. Not only was malaria once the scourge of upstate New York under the name of "Genesee Valley fever," but it was also a disease of winter in Holland when the mosquitoes moved indoors with the coming of cold weather. Most significant is the human capability of eliminating malaria in any region through measures to control the mosquito carrier, prophylaxis and treatment to prevent infection of the mosquito, or both.

The classic disease of place is endemic goitre, because it occurs in persons living in regions with low iodine content in the soil who are dependent on food and water from the same region. Massive glaciation over much of the northern temperate regions removed iodine-rich soil, and there has not been sufficient time for it to be replenished by air currents. Mountainous regions also tend to be iodine-poor. Yet the problem can be wholly overcome by iodizing salt, and even without this measure, the increasing exchange of food between iodine-rich and iodine poor regions is causing the disease to disappear in many localities.

It is perhaps a truism of medical science that under similar conditions of exposure to the agent, the clinical outcome of many diseases is very much an individual matter. It is sometimes assumed that even before immunization was available, measles was rarely a cause of death in North America or Europe because of the quality of medical care. Yet, medical care had virtually nothing to do with measles mortality, since it was not medical attention that prevented deaths or decreased the chances of becoming infected; before immunization, nearly everyone developed measles at some time during childhood in both industrialized and developing countries. In 1960, mortality rates were 90 times higher in Mexico and 189 times higher in Guatemala than in the United States, with even higher multiples for some South American and African countries. These high rates were, and for most developing countries still are, due to the poorer nutrition status of children in these countries.

Tuberculosis is an equally compelling example. As an intern in Gorgas Hospital in Panama, I saw tuberculosis as a quite different disease-extensively disseminated, rapidly progressive, and in 1945, often fatal-compared with the generally mild cases I saw as a medical student in Rochester, New York. Once again, a poorly nourished population was much more susceptible. This experience could be duplicated in almost any developing country at the time and lends credence to the previous explanation of the fall in mortality from this and other infectious diseases in Europe during the last century. While modern streptomycin-isoniazid therapy is effective even in male nourished individuals, the fall in tuberculosis mortality considerably antedates the availability of this therapy.

Non-nutrient factors in nutritional disease

Thus far, the examples have emphasized the effect of nutrition status on host resistance to infectious disease. Host factors are also of significance in understanding the occurrence of nutritional disease itself. Individuals are more vulnerable to malnutrition during certain physiologic states, namely at times of growth, pregnancy, and lactation. Frequently, malnutrition can be explained by concurrent pathological states that can range from parasites and acute infections diseases to malignancies. Nevertheless, host factors merely increase or decrease the probability that a given intake of a nutrient will be sufficient for an individual.

An understanding of nutritional disease causation requires recognition of the relevant environmental factors. These operate in three different ways: by affecting the production and physical availability of food, by altering the requirement of the host for nutrients, and by influencing the actual intake of nutrients. For each of these mechanisms it is necessary to look systematically and separately at the effects of the physical, biologic, and social environment.

The multiple ways in which all three kinds of environmental factors combine to determine food production and availability are beyond the scope of this discussion. The environmental factors influencing the nutrient requirements of the host are primarily the acute and chronic infections, as previously mentioned. While these arise out of the biological environment, they are largely determined by physical and social factors. Those environmental factors that determine actual nutrient intake, however, are particularly important for the physician to understand. Failure to do so is likely to lead to advice that is impractical or unacceptable.

The problem is that medical schools have great difficulty conveying to their students an adequate appreciation of environmental factors as causative of disease. The specific agent can usually be identified with the help of the laboratory, and the host is available for interview and evaluation. But the environment in which the pathological condition originated is generally not directly accessible. Moreover, medical students themselves, and the faculty as well, often have little firsthand experience with either the macro or the microenvironments of their patients. In many cases, even an accurate description of the environmental conditions for a given patient may be so removed from the social experience of the physician that it is not comprehended. The situation is far different from the time in which family doctors did possess this kind of knowledge to help them understand disease in their patients.

For example, it may be obvious to the physician that a child with oedema caused by severe protein calorie malnutrition would benefit from being given cow's milk, but a recommendation that the child be given milk may be culturally unacceptable or, more likely, economically impossible. Yet, advice to give the child milk that the mother cannot or does not follow does no good. Far too often we find that the mother does the best she can to follow such advice, but for economic reasons must dilute the milk. The result is that the condition becomes worse and may end fatally.

The tendency to withdraw solid food when a child develops diarrhoea is a particularly complex and important example. Any infectious disease affects appetite and results in decreased food intake, but there is an apparent logic to "resting" the gastrointestinal tract under these conditions, a practice still followed by most physicians. The problem is that if the diarrhoea persists, the child who is already in borderline nutrition status rapidly becomes depleted further. In some cases, this leads to clinical kwashiorkor with the diarrhoea persisting because of protein deficiency rather than from the original infection. Moreover, it was shown conclusively in 1948 [4, 5] and has been repeatedly confirmed [6] that feeding children with diarrhoea to the extent that their appetite permits improves recovery regardless of any increase in the stool volume. The physician may have only intended that food be withheld for 24 or 48 hours, but the practice, once initiated, is likely to be continued by the mother as long as diarrhoea persists. It is still responsible for many thousands of deaths each year in developing countries.

The cultural factors that influence food consumption are commonly the decisive ones. Malnutrition often results from dietary habits based on ignorance or food prejudices. One example is the unwillingness in many different cultures to give eggs to young children because of the conviction that they will not be digested or will make the child ill [7]. Many other beliefs deeply rooted in a culture for reasons that may or may not be discernible are counterproductive from a purely nutritional point of view. Regardless of their origins, the proscription or prejudice against the eating of pork by Muslims, beef by Hindus, and poultry or fish by some other groups must be regarded as other examples [8]. Nutritionally negative beliefs that have a valid basis in a society must also be recognized. Adulterated, contaminated, unrefrigerated cow's milk may indeed be a cause of diarrhoea in technically underdeveloped areas, as it was in the United States and Europe until well into this century [2].

Harmful nutritional beliefs and practices are by no means confined to primitive societies or to developing countries. In the United States and Europe, food faddism is one of the most common causes of malnutrition. At best, the so-called health foods and diets are needlessly expensive sources of nutrients and at worst, nutritionally unbalanced and dangerous. Unfortunately, even physicians have been responsible for reducing diets that are nutritionally inadequate and have resulted in serious disease states. Quackery and faddism also result in cases of toxicity from excessive doses of vitamins. Vitamins A, D, and E are particularly dangerous. Recently, toxic effects from "megadoses" of even the water-soluble B vitamins and vitamin C have begun to appear [9].

By far the most significant environmental factor influencing food consumption in underprivileged groups is economic. The relatively high cost of protective foods and limited purchasing power are major factors in the appearance of malnutrition in developing countries and among the poor everywhere. In developing countries, there is little hope of improving the nutrition status of vulnerable groups unless persons can acquire the food they need. If they are landless and cannot produce it, they must be able to buy it. However, funds to do so require employment opportunities and adequate wages. Entitlement programmes such as food stamps or other means of subsidizing food acquisitions can be of help, but are not a substitute for economic opportunity.

Economic factors cannot be separated from political ones. Government land tenure, extension, credit, price subsidies, and price controls have a great deal to do with the adequacy of food production in a country. Subsidies and price controls may make it easier for the poor to buy food. Legislation requiring the fortification of salt with iodine can eliminate endemic goitre from a country at no additional cost to the consumer.

Cereal products in the United States and Canada have long been fortified with thiamine, riboflavin, niacin, and iron. A large-scale trial in India has shown the feasibility of preventing the multiple adverse consequences of iron deficiency by adding iron to salt for populations not consuming processed cereal products. A nationwide trial in Guatemala demonstrated that the addition of vitamin A to sugar is feasible and that it restores blood levels to normal. The point is that there are many different ways of preventing disease and improving health. The health sector and its personnel represent only one approach to the goal of "Health for All." Other sectors also have much to contribute, and ultimately it is government policy that is most influential in determining the level of nutritional and general health in a country.

Conclusion

The treatment and prevention of most disease will be more effective if it includes consideration of all relevant, available data that pertain to the host, and in the broadest sense, to the environment. It is usually these multiple influences that will determine whether a potential disease agent will or will not cause disease in a specific individual. In short, the practice of clinical medicine will benefit from an epidemiological approach to causation, and for preventive medicine, it is essential.

This method of viewing disease is of value whether the agent of disease in an individual is perceived by modern medical science to be due to a microbial organism, a nutrient deficiency or excess, a toxin, a carcinogen, trauma, or whether it is perceived to be of psychologic or metabolic origin. In every case, it is a complex of other factors that determines where and when the agent will be present, its qualitative and quantitative nature, and the influence it will have on the host.

It should be evident that improvements are required in medical education to reflect this more ecological consideration of disease. Epidemiology, nutrition, and preventive medicine are still neglected subjects in the medical curriculum. Moreover, the whole system of medical teaching places major emphasis on arriving at the proper diagnosis. Epidemiological factors are brought in as an aid to diagnosis, but they rarely receive appropriate emphasis for their contribution to treatment and prevention, both primary and secondary.

There should be far more emphasis on prevention in medical practice and it should be based as much as possible on sound knowledge of the multiple factors involved in each case. Fortunately, this is beginning to happen to some degree. As the relationship between cigarette smoking and lung cancer and coronary heart disease has become more evident, physicians are beginning to emphasize the hazards of smoking. Obesity and excessive alcohol consumption receive some attention as risk factors, but few physicians urge the amount of exercise required for cardiovascular fitness.

The United States along with some 153 other countries has subscribed to the World Health Organization's stated goal of "Health for all by the year 2000." However, an approach to this goal by the finest system of secondary and tertiary medical care to the neglect of the primary prevention of disease is a travesty of the concept. In the context of the declaration made in Alma-Ata in which the goal was formulated, primary health care includes "education concerning prevailing health problems and methods of identifying, preventing, and controlling them; promotion of food supply and proper nutrition; an adequate supply of safe water and basic sanitation; maternal and child health care, including family planning; immunization against the major infectious diseases; prevention and control of locally endemic diseases; appropriate treatment of common diseases and injuries; and provision of essential drugs." Note that appropriate treatment and provision of essential drugs are by no means at the top of the list.

We must train physicians who view health care in this broader context and endeavour to prevent, as well as treat, disease. If diseases are viewed as problems in human ecology of multi-factorial origin, the Alma-Ata approach to primary health care follows naturally. The most important advance that medicine can now make in terms of the well-being of the greatest number of persons is to develop a more holistic approach to the treatment and prevention of human disease.

References

1. McKeown T. The role of medicine: dream, mirage, or nemesis? Princeton, NJ, USA: Princeton University Press, 1979.

2. Scrimshaw NS. Ecological factors in nutritional disease. Am J Clin Nutr 1964;14:112-22.

3. Gordon JE, Freundt EA, Brown EW Jr, Babbott FL Jr. Endemic and epidemic diarrhea! disease in arctic Greenland. Am J Med So 1961;242:158-74.

4. Chung A. The effect of oral feeding at different levels on the absorption of foodstuffs in infantile diarrhea. J Pediatr 1948;33:1-13

5. Chung A, Viscorová B. The effect of early oral feeding versus early oral starvation on the course of infantile diarrhea. J Pediatr 1948;33:14-22.

6. Chen LC, Scrimshaw NS, eds. Diarrhea and malnutrition. Interactions, mechanisms, and interventions. New York and London: Plenum Press, 1983.

7. Jelliffe DB. Nutrition in the subtropics and tropics. Monograph Series no. 29. Geneva: WHO, 1968.

8. Simoons FJ. Eat not this flesh. Madison, Wis, USA: University of Wisconsin Press, 1961.

9. Alhadeff L, Gualtieri CT, Lipton M. Toxic effects of water-soluble vitamins. Nutr Rev 1984;42:33-40.

Contents - Previous - Next