1 Basic Epidemiologic Concepts and Principles
II. ETIOLOGY AND NATURAL HISTORY OF DISEASE
III. ECOLOGICAL ISSUES IN EPIDEMIOLOGY
IV. CONTRIBUTIONS OF EPIDEMIOLOGISTS
I What is Epidemiology?
Epidemiology is usually defined as the study of factors that determine the occurrence and distribution of disease in a population. As a scientific term, epidemiology was introduced in the 19th century, derived from three Greek roots: epi, meaning “upon”; demos, “people” or “population”; and logos, “discussion” or “study.” Epidemiology deals with much more than the study of epidemics, in which a disease spreads quickly or extensively, leading to more cases than normally seen.
Epidemiology can best be understood as the basic science of public health. It provides methods to study disease, injury, and clinical practice. Whereas health care practitioners collect data on a single patient, epidemiologists collect data on an entire population. The scientific methods used to collect such data are described in the Epidemiology section of this text, Chapters 1 to 7, and the methods used to analyze the data are reviewed in the Biostatistics section, Chapters 8 to 13.
The scientific study of disease can be approached at the following four levels:
1. Submolecular or molecular level (e.g., cell biology, genetics, biochemistry, and immunology)
2. Tissue or organ level (e.g., anatomic pathology)
3. Level of individual patients (e.g., clinical medicine)
Perspectives gained from these four levels are related, so the scientific understanding of disease can be maximized by coordinating research among the various disciplines.
Some people distinguish between classical epidemiology and clinical epidemiology. Classical epidemiology, which is population oriented, studies the community origins of health problems, particularly those related to infectious agents; nutrition; the environment; human behavior; and the psychological, social, economic, and spiritual state of a population. Classical epidemiologists are interested in discovering risk factors that might be altered in a population to prevent or delay disease, injury, and death.
Investigators involved in clinical epidemiology often use research designs and statistical tools similar to those used by classical epidemiologists. However, clinical epidemiologists study patients in health care settings rather than in the community at large. Their goal is to improve the prevention, early detection, diagnosis, treatment, prognosis, and care of illness in individual patients who are at risk for, or already affected by, specific diseases.1
Many illustrations from classical epidemiology concern infectious diseases, because these were the original impetus for the development of epidemiology and have often been its focus. Nevertheless, classical methods of surveillance and outbreak investigation remain relevant even for such contemporary concerns as bioterrorism, undergoing modification as they are marshaled against new challenges. One example of such an adapted approach is syndromic epidemiology, in which epidemiologists look for patterns of signs and symptoms that might indicate an origin in bioterrorism.
Epidemiology can also be divided into infectious disease epidemiology and chronic disease epidemiology. Historically, infectious disease epidemiology has depended more heavily on laboratory support (especially microbiology and serology), whereas chronic disease epidemiology has depended on complex sampling and statistical methods. However, this distinction is becoming less significant with the increasing use of molecular laboratory markers (genetic and other) in chronic disease epidemiology and complex statistical analyses in infectious disease epidemiology. Many illnesses, including tuberculosis and acquired immunodeficiency syndrome (AIDS), may be regarded as both infectious and chronic.
The name of a given medical discipline indicates both a method of research into health and disease and the body of knowledge acquired by using that method. Pathology is a field of medical research with its own goals and methods, but investigators and clinicians also speak of the “pathology of lung cancer.” Similarly, epidemiology refers to a field of research that uses particular methods, but it can also be used to denote the resulting body of knowledge about the distribution and natural history of diseases—that is, the nutritional, behavioral, environmental, and genetic sources of disease as identified through epidemiologic studies.
II Etiology and Natural History of Disease
The term etiology is defined as the cause or origin of a disease or abnormal condition. The way a disease progresses in the absence of medical or public health intervention is often called the natural history of the disease. Public health and medical personnel take advantage of available knowledge about the stages, mechanisms, and causes of disease to determine how and when to intervene. The goal of intervention, whether preventive or therapeutic, is to alter the natural history of a disease in a favorable way.
A Stages of Disease
The development and expression of a disease occur over time and can be divided into three stages: predisease, latent, and symptomatic. During the predisease stage, before the disease process begins, early intervention may avert exposure to the agent of disease (e.g., lead, trans-fatty acids, microbes), preventing the disease process from starting; this is called primary prevention. During the latent stage, when the disease process has already begun but is still asymptomatic, screening for the disease and providing appropriate treatment may prevent progression to symptomatic disease; this is called secondary prevention. During the symptomatic stage, when disease manifestations are evident, intervention may slow, arrest, or reverse the progression of disease; this is called tertiary prevention. These concepts are discussed in more detail in Chapters 15 to 17.
B Mechanisms and Causes of Disease
When discussing the etiology of disease, epidemiologists distinguish between the biologic mechanisms and the social, behavioral, and environmental causes of disease. For example, osteomalacia is a bone disease that may have both social and biologic causes. Osteomalacia is a weakening of the bone, often through a deficiency of vitamin D. According to the custom of purdah, which is observed by many Muslims, women who have reached puberty avoid public observation by spending most of their time indoors, or by wearing clothing that covers virtually all of the body when they go outdoors. Because these practices block the action of the sun on bare skin, they prevent the irradiation of ergosterol in the skin. However, irradiated ergosterol is an important source of D vitamins, which are necessary for growth. If a woman’s diet is also deficient in vitamin D during the rapid growth period of puberty, she may develop osteomalacia as a result of insufficient calcium absorption. Osteomalacia can adversely affect future pregnancies by causing the pelvis to become distorted (more pear shaped), making the pelvic opening too small for the fetus to pass through. In this example, the social, nutritional, and environmental causes set in motion the biochemical and other biologic mechanisms of osteomalacia, which may ultimately lead to maternal and infant mortality.
Likewise, excessive fat intake, smoking, and lack of exercise are behavioral factors that contribute to the biologic mechanisms of atherogenesis, such as elevated blood levels of low-density lipoprotein (LDL) cholesterol or reduced blood levels of high-density lipoprotein (HDL) cholesterol. These behavioral risk factors may have different effects, depending on the genetic pattern of each individual and the interaction of genes with the environment and other risk factors.
Epidemiologists attempt to go as far back as possible to discover the social and behavioral causes of disease, which offer clues to methods of prevention. Hypotheses introduced by epidemiologists frequently guide laboratory scientists as they seek biologic mechanisms of disease, which may suggest methods of treatment.
C Host, Agent, Environment, and Vector
The causes of a disease are often considered in terms of a triad of factors: the host, the agent, and the environment. For many diseases, it is also useful to add a fourth factor, the vector (Fig. 1-1). In measles, the host is a human who is susceptible to measles infection, the agent is a highly infectious virus that can produce serious disease in humans, and the environment is a population of unvaccinated individuals, which enables unvaccinated susceptible individuals to be exposed to others who are infectious. The vector in this case is relatively unimportant. In malaria, however, the host, agent, and environment are all significant, but the vector, the Anopheles mosquito, assumes paramount importance in the spread of disease.
Host factors are responsible for the degree to which the individual is able to adapt to the stressors produced by the agent. Host resistance is influenced by a person’s genotype (e.g., dark skin reduces sunburn), nutritional status and body mass index (e.g., obesity increases susceptibility to many diseases), immune system (e.g., compromised immunity reduces resistance to cancer as well as microbial disease), and social behavior (e.g., physical exercise enhances resistance to many diseases, including depression). Several factors can work synergistically, such as nutrition and immune status. Measles is seldom fatal in well-nourished children, even in the absence of measles immunization and modern medical care. By contrast, 25% of children with marasmus (starvation) or kwashiorkor (protein-calorie malnutrition related to weaning) may die from complications of measles.
Agents of disease or illness can be divided into several categories. Biologic agents include allergens, infectious organisms (e.g., bacteria, viruses), biologic toxins (e.g., botulinum toxin), and foods (e.g., high-fat diet). Chemical agents include chemical toxins (e.g., lead) and dusts, which can cause acute or chronic illness. Physical agents include kinetic energy (e.g., involving bullet wounds, blunt trauma, and crash injuries), radiation, heat, cold, and noise. Epidemiologists now are studying the extent to which social and psychological stressors can be considered agents in the development of health problems.
The environment influences the probability and circumstances of contact between the host and the agent. Poor restaurant sanitation increases the probability that patrons will be exposed to Salmonella infections. Poor roads and adverse weather conditions increase the number of automobile collisions and airplane crashes. The environment also includes social, political, and economic factors. Crowded homes and schools make exposure to infectious diseases more likely, and the political structure and economic health of a society influence the nutritional and vaccine status of its members.
Vectors of disease include insects (e.g., mosquitoes associated with spread of malaria), arachnids (e.g., ticks associated with Lyme disease), and mammals (e.g., raccoons associated with rabies in eastern U.S.). The concept of the vector can be applied more widely, however, to include human groups (e.g., vendors of heroin, cocaine, and methamphetamine) and even inanimate objects that serve as vehicles to transmit disease (e.g., contaminated needles associated with hepatitis and AIDS). A vector may be considered part of the environment, or it may be treated separately (see Fig. 1-1). To be an effective transmitter of disease, the vector must have a specific relationship to the agent, the environment, and the host.
In the case of human malaria, the vector is a mosquito of the genus Anopheles, the agent is a parasitic organism of the genus Plasmodium, the host is a human, and the environment includes standing water that enables the mosquito to breed and to come into contact with the host. Specifically, the plasmodium must complete part of its life cycle within the mosquito; the climate must be relatively warm and provide a wet environment in which the mosquito can breed; the mosquito must have the opportunity to bite humans (usually at night, in houses where sleeping people lack screens and mosquito nets) and thereby spread the disease; the host must be bitten by an infected mosquito; and the host must be susceptible to the disease.
D Risk Factors and Preventable Causes
Risk factors for disease and preventable causes of disease, particularly life-threatening diseases such as cancer, have been the subject of much epidemiologic research. In 1964 a World Health Organization (WHO) expert committee estimated that the majority of cancer cases were potentially preventable and were caused by “extrinsic factors.” Also that year, the U.S. Surgeon General released a report indicating that the risk of death from lung cancer in smokers was almost 11 times that in nonsmokers.2
Advances in knowledge have consolidated the WHO findings to the point where few, if any, researchers now question its main conclusion.3 Indeed, some have gone further, substituting figures of 80% or even 90% as the proportion of potentially preventable cancers, in place of WHO’s more cautious estimate of the “majority.” Unfortunately, the phrase “extrinsic factors” (or its near-synonym, “environmental factors”) has often been misinterpreted to mean only man-made chemicals, which was certainly not the intent of the WHO committee. In addition to man-made or naturally occurring carcinogens, the 1964 report included viral infections, nutritional deficiencies or excesses, reproductive activities, and a variety of other factors determined “wholly or partly by personal behavior.”
The WHO conclusions are based on research using a variety of epidemiologic methods. Given the many different types of cancer cells, and the large number of causal factors to be considered, how do epidemiologists estimate the percentage of deaths caused by preventable risk factors in a country such as the United States?
One method looks at each type of cancer and determines (from epidemiologic studies) the percentage of individuals in the country who have identifiable, preventable causes of that cancer. These percentages are added up in a weighted manner to determine the total percentage of all cancers having identifiable causes.
A second method examines annual age-specific and gender-specific cancer incidence rates in countries that have the lowest rates of a given type of cancer and maintain an effective infrastructure for disease detection. For a particular cancer type, the low rate in such a country presumably results from a low prevalence of the risk factors for that cancer. Researchers calculate the number of cases of each type of cancer that would be expected to occur annually in each age and gender group in the United States, if the lowest observed rates had been true for the U.S. population. Next, they add up the expected numbers for the various cancer types in the U.S. They then compare the total number of expected cases with the total number of cases actually diagnosed in the U.S. population. Using these methods, epidemiologists have estimated that the U.S. has about five times as many total cancer cases as would be expected, based on the lowest rates in the world. Presumably, the excess cancer cases in the U.S. are caused by the prevalence of risk factors for cancer, such as smoking.
1 BEINGS Model
The acronym BEINGS can serve as a mnemonic device for the major categories of risk factors for disease, some of which are easier to change or eliminate than others (Box 1-1). Currently, genetic factors are among the most difficult to change, although this field is rapidly developing and becoming more important to epidemiology and prevention. Immunologic factors are usually the easiest to change, if effective vaccines are available.
Box 1-1 BEINGS Acronym for Categories of Preventable Cause of Disease
“B”—Biologic and Behavioral Factors
The risk for particular diseases may be influenced by gender, age, weight, bone density, and other biologic factors. In addition, human behavior is a central factor in health and disease. Cigarette smoking is an obvious example of a behavioral risk factor. It contributes to a variety of health problems, including myocardial infarction (MI); lung, esophageal, and nasopharyngeal cancer; and chronic obstructive pulmonary disease. Cigarettes seem to be responsible for about 50% of MI cases among smokers and about 90% of lung cancer cases. Because there is a much higher probability of MI than lung cancer, cigarettes actually cause more cases of MI than lung cancer.
Increasing attention has focused on the rapid increase in overweight and obesity in the U.S. population over the past two decades. The number of deaths per year that can be attributed to these factors is controversial. In 2004 the U.S. Centers for Disease Control and Prevention (CDC) estimated that 400,000 deaths annually were caused by obesity and its major risk factors, inactivity and an unhealthy diet.4 In 2005, using newer survey data and controlling for more potential confounders, other CDC investigators estimated that the number of deaths attributable to obesity and its risk factors was only 112,000.5 Regardless, increasing rates of obesity are found worldwide as part of a cultural transition related to the increased availability of calorie-dense foods and a simultaneous decline in physical activity, resulting in part from mechanized transportation and sedentary lifestyles.6–11
Obesity and overweight have negative health effects, particularly by reducing the age at onset of, and increasing the prevalence of, type 2 diabetes. Obesity is established as a major contributor to premature death in the United States,12,13 although the exact magnitude of the association remains controversial, resulting in part from the complexities of the causal pathway involved (i.e., obesity leads to death indirectly, by contributing to the development of chronic disease).
Multiple behavioral factors are associated with the spread of some diseases. In the case of AIDS, the spread of human immunodeficiency virus (HIV) can result from unprotected sexual intercourse between men and from shared syringes among intravenous drug users, which are the two predominant routes of transmission in the United States. HIV infection can also result from unprotected vaginal intercourse, which is the predominant transmission route in Africa and other parts of the world. Other behaviors that can lead to disease, injury, or premature death (before age 65) are excessive intake of alcohol, abuse of both legal and illegal drugs, driving while intoxicated, and homicide and suicide attempts. In each of these cases, as in cigarette smoking and HIV infection, changes in behavior could prevent the untoward outcomes. Many efforts in health promotion depend heavily on modifying human behavior, as discussed in Chapter 15.
Epidemiologists are frequently the first professionals to respond to an apparent outbreak of new health problems, such as legionnaires’ disease and Lyme disease, which involve important environmental factors. In their investigations, epidemiologists describe the patterns of the disease in the affected population, develop and test hypotheses about causal factors, and introduce methods to prevent further cases of disease. Chapter 3 describes the standard approach to investigating an epidemic.
During an outbreak of severe pneumonia among individuals attending a 1976 American Legion conference in Philadelphia, epidemiologists conducted studies suggesting that the epidemic was caused by an infectious agent distributed through the air-conditioning and ventilation systems of the primary conference hotels. Only later, after the identification of Legionella pneumophila, was it discovered that this small bacterium thrives in air-conditioning cooling towers and warm-water systems. It was also shown that respiratory therapy equipment that is merely rinsed with water can become a reservoir for Legionella, causing hospital-acquired legionnaires’ disease.
An illness first reported in 1975 in Old Lyme, Connecticut, was the subject of epidemiologic research suggesting that the arthritis, rash, and other symptoms of the illness were caused by infection with an organism transmitted by a tick. This was enough information to enable preventive measures to begin. By 1977 it was clear that the disease, then known as Lyme disease, was spread by Ixodes ticks, opening the way for more specific prevention and research. Not until 1982, however, was the causative agent, Borrelia burgdorferi, discovered and shown to be spread by the Ixodes tick.
Smallpox is the first infectious disease known to have been eradicated from the globe (although samples of the causative virus remain stored in U.S. and Russian laboratories). Smallpox eradication was possible because vaccination against the disease conferred individual immunity and produced herd immunity. Herd immunity results when a vaccine diminishes an immunized person’s ability to spread a disease, leading to reduced disease transmission.
Most people now think of AIDS when they hear of a deficiency of the immune system, but immunodeficiency also may be caused by genetic abnormalities and other factors. Transient immune deficiency has been noted after some infections (e.g., measles) and after the administration of certain vaccines (e.g., live measles vaccine). This result is potentially serious in malnourished children. The use of cancer chemotherapy and the long-term use of corticosteroids also produce immunodeficiency, which may often be severe.
In the 1950s it was shown that Japanese Americans living in Hawaii had a much higher rate of MI than people of the same age and gender in Japan, while Japanese Americans in California had a still higher rate of MI than similar individuals in Japan.14–16 The investigators believed that dietary variations were the most important factors producing these differences in disease rates, as generally supported by subsequent research. The Japanese eat more fish, vegetables, and fruit in smaller portions.
Denis Burkitt, the physician after whom Burkitt’s lymphoma was named, spent many years doing epidemiologic research on the critical role played by dietary fiber in good health. From his cross-cultural studies, he made some stunning statements, including the following17:
“By world standards, the entire United States is constipated.”
“Don’t diagnose appendicitis in Africa unless the patient speaks English.”
“African medical students go through five years of training without seeing coronary heart disease or appendicitis.”
“Populations with large stools have small hospitals. Those with small stools have large hospitals.”