Updated September 11, 2013.

Question: What are Observational Studies?

There are many different types of scientific studies that give researchers information about the way the body ages. Since aging is a long-term process, longitudinal research is often employed to track a group of subjects for a defined period, usually years.

These studies can involve observation, or intervention. Scientists might use longitudinal research to answer questions about the effect of certain behaviors, like regular exercise or meditation, or foods -- like chocolate or a Mediterranean diet, for example -- on the long-term health of the participants.

Answer:

In an observational study, no intervention takes place. While participants answer detailed questions about the lifestyle habit being investigated, or measurements are taken, no adjustment of the habit itself is suggested by the researchers. During the study period, participants are revisited and surveyed again to chart the habits being studied, and their effects.
The US National Cancer Institute, for example, defines observational studies as those in which “biomedical and/or health outcomes are assessed in pre-defined groups of individuals.” Groups may be defined (or chosen) by age, gender, occupation, where they live, or perhaps grouped according to a disease or condition (for example, heart patients or cancer survivors).
Observational research is valuable because it allows information to be gathered in a large population sample, over a long period of time. There are drawbacks, however. Surveys of lifestyle factors depend on the participant remembering, and accurately reporting, their own behavior.

Eliminating confounding factors – that is, other elements that may influence the outcome being analyzed – is also a challenge for researchers conducting observational studies.

For these reasons, observational studies are most valuable in finding out whether factors are correlated, rather than determining with certainty which behavior caused a certain outcome.

For example, many studies have shown that people who eat chocolate regularly have a lower incidence of cardiovascular disease, but research has not yet determined conclusively that it is the chocolate itself that is responsible for the better heart health.

An interventional study, by contrast, would take two groups made up of similar people, give chocolate in predetermined amounts to the members of one group, but not the other. Over time, measurements of blood pressure, blood lipids etc. would be taken and the two groups compared in order to draw conclusions about causation – that is, the effects caused by the chocolate.

Observational studies are also more appropriate for investigating the effects of negative lifestyle factors like smoking or alcohol consumption, in which interventional research (for example, asking subjects to smoke or drink) would be unethical.

Sorting out causation and correlation: an example

<sub>Sources:

Glossary of Clinical Trials Reporting Program (CTRP) Terms. NIH National Cancer Institute Public Information Sheet. Accessed May 11, 2012.
http://www.cancer.gov/clinicaltrials/conducting/ncictrp/resources/glossary

Is My Study Observational or Interventional? NIH National Heart Lung and Blood Institute. Clinical Research Guidelines. Accessed May 11, 2012.
http://www.nhlbi.nih.gov/crg/app_interventional.php</sub>