Research Refresher: Epidemiology

Understanding the relationship between diet, lifestyle, genetics and health is challenging. The scientific community uses various research study types, all with their own strengths and limitations, to shed light on these relationships.

Knowing the strength and limitations of various types of research allows health professionals to make better judgments on the accuracy and value of nutritional data.

Evaluating Diet’s Role in Health

In terms of evaluating the role of diet in various diseases, research is either observational (epidemiological) or experimental (clinical trials or laboratory research).

Randomized control clinical trials are considered the gold standard in nutritional research because they can determine cause and effect; however, they are often not practical for a variety of reasons such as cost, ethics, complexities of the diet.

The vast majority of data linking diet to various health issues such as heart disease or cancer is the result of epidemiological studies. Data from such studies can be used to determine associations and possible risks in a population, but not to establish cause and effect.

Considerations of Epidemiological Research

When reviewing data from epidemiological studies, it is important to consider the limitations before making conclusions.

Assessing Diet
Assessing dietary intake is difficult in populations. To find either a protective or risk effect of a specific food, food group or nutrient is complicated by the variability and nature of eating. Diets are rarely static – people eat different foods, in varying amounts, in different combinations. As a result ‘exposure’ to specific food risk factors is less clear.

Assessing Risk
Epidemiological studies do not prove causality; they express possible risk. Study results are typically reported as relative risk (RR) or odds ratio (OR). The RR expresses the risk for a disease in persons exposed to the factor as a ratio or multiple of the risk in persons not exposed.

The strength of an association can be evaluated by looking at the RR. An RR less than 2.0 is not considered strong. Further, expressing results as an RR can make a problem appear worse than it really is. It is important to consider both the RR and the absolute risk (how likely the outcome is to happen overall) when evaluating research results.

Confounding Variables
It’s also important to consider the confounding variable – another factor that has not been controlled for in the study, but may affect outcomes. For example, an observed association between diet and cancer might be confounded by other factors such as smoking or physical activity. Although these are often statistically adjusted, there is still some degree of error.

Bottom Line

Diet-related epidemiological studies contain many interconnected, complex diet-related variables and limitations to research methods. In addition, this common type of nutrition research does not determine cause and effect. It is important to consider these limitations when making conclusions on nutrition data and translating that data into consumer messages.

Types of Epidemiological Studies

Nutritional epidemiology studies typical or usual dietary patterns to assess long term diets and measure exposure to certain foods or nutrients.

• Ecologic studies – look at large populations and compare the frequency of disease rates and dietary intakes.
• Case control studies – compare two similar groups of people (cases and controls) and collect data on past dietary intake.
• Prospective Cohort Studies – assess the intake of a large group of people over a long period of time and then compare those who develop disease and those who don’t.

Each of these types of studies has limitations. For systematic reviews and evidenced-based recommendations, the results from prospective cohort studies are regarded as strongest.