Epidemiology: A branch of medical science that deals with the incidence, distribution, and control of disease in a population.
—Merriam-Webster’s Collegiate Dictionary, 10th Edition.
If you read the paper, listen to news radio, or catch the nightly news on the television, you have no doubt heard about dozens of research studies linking food and nutrition to health. It’s a hot topic; after all, everyone eats! Many of the nutrition-health studies we hear about are epidemiological—population—studies. Epidemiology is a valid and valued form of research, and has long been of special importance in the public health arena. Unfortunately, it is a type of research, with some limitations, that is frequently misinterpreted. Media coverage of epidemiological results is frequently inaccurate, or is described in a way that renders the findings less than useful—or even meaningless—to the average person.
Why does this happen? There are a number of reasons: primary among them is that space in the paper or time “on the air” is limited, so the reporter must make a decision about which facts to report. In those cases, attention-grabbing headlines take priority over the supplemental information that people usually need to really understand the research and its implications. Second, many reporters who write about nutrition research may have no formal training in how to interpret the often complicated studies, and therefore may make mistakes in communicating a study’s findings.
Learning a bit about epidemiological research is a good way to help make sure that we, as consumers, better understand research that may affect our health. Once we know what to look for, it’s a lot easier not only to interpret what we read and hear about research, but also to decide if it’s important to us.
Epidemiology 101
Basically, epidemiological studies look at populations to investigate potential associations between aspects of health (such as cancer and heart disease) and diet, lifestyle, genetics, or other factors within populations. Epidemiological studies are observational in nature and the outcomes need further study through other types of research such as intervention or clinical studies. This type of research yields information about the distribution and determinants of disease or other health outcomes for further study, but it does not establish cause and effect. For example, comparing per capita meat consumption between various countries may reveal an association between eating meat and certain types of cancer. However, one could not correctly conclude that eating more meat causes the cancers, because other unrelated factors in the diets of meat eaters may truly be the culprits (such as total fat, calories, or the lower levels of vegetable consumption of meat eaters).
Within the category of epidemiological research, there are various types of research designs. Prospective studies monitor a group of people over a period of time to observe the effects of diet, behavior, and other factors on health or the incidence of disease, whereas retrospective studies look at events and behaviors that have already taken place (University of California Berkeley Wellness Letter, 1996). Epidemiological research may also be either observational or experimental. In experimental research, part of the population receives some sort of treatment (sometimes called an intervention) and the results are compared with the results for those who do not receive the treatment (the control group.) Observational research suggests associations or correlations between characteristics based on observed differences. To determine cause and effect, experimental research must be conducted.
The roots of epidemiology
Maybe the word epidemiology is new to you, but the basic concepts of epidemiological research have been around for centuries. Early Greeks and Romans recognized that symptoms of certain diseases were associated with environmental conditions (for example “marsh fever” occurred more frequently among people who lived in swampy areas). Early epidemiologists considered not only infectious diseases like cholera and plague, but also environmental hazards like lead, chronic diseases, and even occupational risks when conducting their research.
Traditional quantitative epidemiology began in the 17th century because of concern for the public’s health. That early epidemiological research focused on the plague and poor sanitation (and its accompanying diseases). Infectious disease epidemiology and the germ theory (which states that single agents relate one to one to specific diseases) characterized the mid-to-late 19th century and the first half of the 20th century. For example, John Snow’s work on the cause of cholera in the mid-1850s was followed by the study of tuberculosis, anthrax, and later, polio. Beginning in the 1920s, those who started to search for nonmicrobiological causes of disease ran against the tide of current scientific thinking. Joseph Goldberger in the United States was such a pioneer when he established nutritional deficiency as a cause of pellagra. Thus, the study of nutrition and disease was born.
Of course, as is the case today, the conclusions of early epidemiologists were not always correct. Nevertheless, over the years the science of epidemiology matured and became more refined. It has played an enormous role in helping to quantify the magnitude of such diseases as HIV/AIDS, breast cancer, Alzheimer’s disease, and tobacco-related disease—and has even suggested some factors that influence their occurrence. Traditionally, epidemiology has been a guiding force behind many public health policies and programs worldwide. The decline in infectious diseases and the rise in the relative importance of noncommunicable diseases (diseases that are not contagious and that cannot be transmitted from person to person) have led to the development of modern epidemiology. Today, consumers may interpret epidemiological data as relating to individuals when in fact, epidemiological studies are intended for focus on broad population groups.
What do these studies mean to me?
For most people, interpreting epidemiological research isn’t an easy task—especially when the full research study is not available for consideration, as is usually the case for consumers. How, then, can we begin to understand whether a study is relevant to our lives? There are a few ways to quickly ascertain whether a given piece of research is applicable to one’s life. Granted, this is the “quick-and-dirty” approach to interpreting research, and of course, taking the time to read an entire research report critically is always best. In a pinch, however, the following tips can get you started:
Consider the type of risk expressed in the study.
Remember that it’s best to consider both types of risk when evaluating a study (see the sidebar “Risky Business”). Absolute risk will tell you how likely the outcome is to happen overall, whereas relative risk compares the risk between groups. If the absolute risk is negligible, then the relative risk may be irrelevant.
Assess whether the results are generalizable to other groups.
Studies are frequently conducted with rather narrow population groups, making for stronger, more valid results, but sacrificing the ability to generalize. For example, a study conducted with men may not mean that the same results would occur in a population of women. Obviously, animal research cannot be easily generalized to humans; the results of research conducted with adults may not be applicable to children, and so on. Researchers usually give those kinds of admonitions at the end of the research study, but these types of statements often get lost (or are not included) due to media’s time or space constraints in reporting of research.
Remember that an association or correlation doesn’t prove cause and effect.
It cannot be stressed enough that an association or correlation does not prove cause and effect. This is one of the easiest mistakes to make when interpreting research—the media do it frequently and so do consumers. When an association is suggested, more research is always needed to determine whether that association really exists, and why.
View research studies as discussions among scientists.
Almost no one gets to have the final word, as it is rare that a study provides a final, complete answer. Flip-flopping opinions among experts, although frustrating for the public, reflect the fact that science is a dynamic, evolving process.
The bottom line
Regardless of whether the study is epidemiological or another type of research, don’t change your diet on the basis of the findings of just one new study. For the reason mentioned in the last tip in the list of tips given above, it’s unnecessary—and perhaps even unwise—to modify dietary behavior on the basis of the findings of one study especially if you are considering eliminating an entire food category. Wait for confirming studies—and plenty of them—before making a change.
R.i.s.k.y Business
If you don’t fully comprehend the differences between relative and absolute risk, you may completely misunderstand the importance of a research study. Absolute risk refers to the actual risk of an occurrence—the chance that a specific outcome will occur. Relative risk puts risk in comparative terms—for example, the outcome rate for people exposed to the factor in question compared with the outcome rate for those not exposed to the factor. A relative risk of more than 1 indicates an increased risk of the outcome under investigation; a relative risk of less than 1 indicates a decreased risk of the outcome. Relative risks are the most commonly used measure of morbidity or mortality in the medical literature today. In many cases, however, the absolute risk is a far more relevant statistic for the public.
For example, a news report that claims that people who consume Product X are 75 percent more likely to suffer from Problem Y than those who don’t consume it may seem to be a pretty convincing and important study. This type of statement is an expression of relative risk. In the same example, the absolute risk of someone in the study suffering from Problem Y may have been only 1 percent. In this case, the absolute risk is the more meaningful result, as it tells us that only 1 percent of people suffer from Problem Y anyway—it was the expression of the results as a relative risk that made the problem seem more important than it really is. However, relative risk can also make a problem appear to be less important than it actually is. Therefore, it is important to consider both relative risk and absolute risk when evaluating research results.
