PMEP Home Page --> Pesticide Active Ingredient Information --> EXTOXNET: The Extension Toxicology Network --> Toxicology Information Briefs --> Epidemiology

E  X  T  O  X  N  E  T
Extension Toxicology Network

A Pesticide Information Project of Cooperative Extension Offices of Cornell University, Michigan State University, Oregon State University, and University of California at Davis. Major support and funding was provided by the USDA/Extension Service/National Agricultural Pesticide Impact Assessment Program.

  Toxicology
Information
Brief
Epidemiology

Publication Date: 9/93

INTRODUCTION

Epidemiology is the study of diseases in populations of humans or other animals, specifically how, when and where they occur.

Epidemiologists attempt to determine what factors are associated with diseases (risk factors), and what factors may protect people or animals against disease (protective factors). The science of epidemiology was first developed to discover and understand possible causes of contagious diseases like smallpox, typhoid and polio among humans. It has expanded to include the study of factors associated with non-transmissible diseases like cancer, and of poisonings caused by environmental agents.

Epidemiological studies can never prove causation; that is, it cannot prove that a specific risk factor actually causes the disease being studied. Epidemiological evidence can only show that this risk factor is associated (correlated) with a higher incidence of disease in the population exposed to that risk factor. The higher the correlation the more certain the association, but it cannot prove the causation.

For example, the discovery of the link between cigarette smoking and lung cancer was based on comparisons of lung cancer rates in smokers and non-smokers. The rates of lung cancer are much higher in smokers than in non-smokers. Does this prove that cigarette smoking causes lung cancer? No. In order to prove that cigarette smoking is the factor causing this increase in lung cancer, it was necessary to expose animals to tobacco smoke and tobacco smoke extracts. This was done under highly controlled conditions where the only difference between the controls (animals not exposed to smoke) and treated animals was the exposure to smoke. These laboratory studies proved the causal association between smoking and increased risk of cancer.

Epidemiological studies can be divided into two basic types depending on (a) whether the events have already happened (retrospective) or (b) whether the events may happen in the future (prospective). The most common studies are the retrospective studies which are also called case-control studies. A case-control study may begin when an outbreak of disease is noted and the causes of the disease are not known, or the disease is unusual within the population studied.

The first step in an epidemiological study is to strictly define exactly what requirements must be met in order to classify someone as a "case." This seems relatively easy, and often is in instances where the outcome is either there or not there (a person is dead or alive). In other instances it can be very difficult, particularly if the experts disagree about the classification of the disease. This happens often with the diagnosis of particular types of cancer. In addition, it is necessary to verify that reported cases actually are cases, particularly when the survey relies on personal reports and recollections about the disease made by a variety of individuals.

The strength of an epidemiological study depends on the number of cases and controls included in the study. The more individual cases that are included in the study, the more likely it is that a significant association will be found between the disease and a risk factor. Just as important is determining what behavioral, environmental, and health factors will actually be studied as possible risk or protective factors. If inappropriate factors are chosen, and the real factors are missed, the study will not provide any useful information. In such an instance, an association may be found between an inappropriate factor, and the disease because this inappropriate factor which we will call factor 1, is associated with another factor, factor 2, which is actually related to the disease, but which was not studied. In such an instance, factor 1 is called a confounding variable, because it confounds the interpretation of the results of the study. Thus, it is very important that epidemiologist choose the proper factors to study at the outset, and not study too many factors at once, since the possibility of finding confounding factors increases with the addition of more variables.

ESTABLISHING THE LINK

Epidemiology relies heavily on statistics for establishing and quantifying the relationships between risk factors and disease, and for establishing whether or not there is an excessive amount of a particular disease occurring in a specific geographic area. Medical records can provide invaluable historical data for establishing trends in the incidence of diseases. There are vast collections of medical record information all over the world, and sorting through the data can be a very expensive and time consuming process. In addition, what can be gained from the records is only as good as the information that they contain, and often the information is scanty or impossible to verify.

One source of information commonly used are death certificate registries, which usually contain information about cause of death. Using information from such a registry, someone once did a study that showed an unusually high incidence of deaths from lung cancer in a large agricultural valley city. There was no question that the rates of deaths from lung cancer in this city were much higher than in other cities of a similar size and location. Many interpretations of this data were presented ranging from pesticide use to agricultural burning. Someone finally noticed that this particular city contained a renowned hospital treatment center for patients with lung cancer. As it turned out, the very high rate of cancer deaths could be explained by the numbers of lung cancer patients who came there from all over the state for treatment which was most often unsuccessful.

Epidemiologist are often called upon to investigate apparent "clusters" of disease in specific geographical areas. For example, a woman in a community may have a miscarriage, and later learn that other people she knows in the neighborhood have also had miscarriages within the last couple of years. It may appear to her and her friends and family that a lot of miscarriages have occurred in their neighborhood. An epidemiologist may be called in to determine if the rate of miscarriages is higher than usual. To do so, the epidemiologist must interview all of the women in the community, and another similar community in a different geographical area, or select appropriate samples of women to be interviewed concerning their reproductive histories. This information will be validated through hospital records, and then analyzed and compared to other similar studies. Even in situations where there is not a higher than normal rate, it may seem higher to the inhabitants, because diseases are not evenly distributed throughout populations.

INTERPRETING THE RESULTS

If the rate of occurrence of a disease in the general population of the USA is 10 per thousand, it does not mean that every group of one thousand people tested will provide 10 cases of disease. Some groups may have 5 or less, others 15 or more. The mean rate is 10 per thousand not the rate in every thousand people tested. This uneven distribution is similar to the way chocolate chips cluster in cookies. Most of the time they are pretty well spread throughout, however some cookies may have all the chips clumped together on one side. The only way to determine whether a cluster is a "real" cluster or just a "chance" cluster is to do a full scale epidemiological study, which is an expensive and time consuming process, and may still give disappointing results.

SUMMARY

Retrospective and prospective studies have been extremely valuable in discovering links between chemical exposure and disease. Perhaps the best example, again, is the association of cigarette smoking and lung cancer and emphysema. Epidemiological studies have also been especially useful in the occupational sector where workers have been exposed to a small number of chemicals, at high dosage rates for periods of time. Epidemiological studies are occasionally relatively easy, and especially significant when they uncover a very high incidence of an unusual disease in a population. For example, the finding of a very small number (about 10) of cases of a very rare liver tumor in workers heavily exposed to vinyl chloride, was a strong signal that vinyl chloride was the causative agent. Animal studies proved this was the case. Epidemiological studies are least powerful in studying very common diseases that occur at high incidence rates in many different populations. In such instances, it is necessary to include huge numbers of subjects in the studies, sometimes thousands. Such large studies have been undertaken in regard to cardiovascular disease, and many factors which influence the development of cardiovascular disease have been found. Such large scale efforts have really paid off because they have provided information about diet and exercise habits that can be used to prevent the development of cardiovascular disease.

The same is true with respect to chemically induced diseases. We all reap the benefits of epidemiological studies of workers exposed to chemicals in occupational settings. We also reap the benefits of studies done on patients who must take certain medications daily, and who sometimes develop side effects. Epidemiological studies of disease related to chemical exposure are very difficult in the general population because of the multitude of chemicals to which we are daily exposed. This is, after all, a world made up of chemicals, just as we ourselves are.