Cohort Study

Cohort studies are a fundamental nonrandomized study design used in epidemiology; they are designed to identify whether a particular exposure is related to an outcome of interest. In its simplest form, this observational study design compares one group (“cohort”) of individuals with a particular exposure to a group of individuals who do not have that exposure (“referent group”) over a period of time (longitudinally). Analysis yields a relative risk (RR) that reflects the relative probabilities of exposure in the two populations. Cohort studies can be classified as retrospective or prospective, depending on whether the exposure and covariates that define the cohort are measured before or after the occurrence of disease.

Cohort studies are often considered the best alternative for examining potentially hazardous exposures when random assignment is not feasible or is unethical (as would be the case with harmful exposures such as smoking). Prospective cohorts are considered a robust method for studying common exposures because the timing of outcomes versus exposures is better established, lending greater validity to a causal relationship.

Cohort studies potentially require substantive investment of cost and time, as the outcome of interest might manifest only after a long exposure time. Furthermore, statistical efficiency may not be achieved if the outcome of interest is rare.

Cohort studies are subject to several types of bias. Selection bias might occur when loss to follow-up through mortality or simply dropping out of the study disproportionately affects exposed or unexposed subjects. Also, quality of outcome information might differ between exposed and unexposed subjects. Cohort studies are subject to the problem of confounding: That is, the distribution of nonexposure factors that are associated with the outcome might differ coincidentally in cohort and referent groups. Exposed and unexposed subjects might experience confounding, or differential distribution of nonexposure factors that are associated with the outcome.

A well-designed cohort study begins with a well-defined study population, with particular attention paid to who is included in the study, whether or not new eligible subjects are allowed to enter the cohort as time passes, and how individuals are reclassified if their exposure status changes. The Framingham Heart Study is a well-known cohort study that has been following participants from a small town near Boston, Massachusetts, since 1948. The Framingham Study has been used to identify major cardiovascular disease risk factors, such as high blood pressure, cholesterol, smoking, and diabetes.