How are epidemiological studies of diseases conducted?

There are different study designs for observational epidemiological studies that differ in effort of realization, duration of investigation, validity, and cost. The most common study types are briefly introduced:

In a cohort study, a group of persons (cohort) is observed over a longer period of time to find out whether a disease due to a exposure occurs among several persons. The incidence can be related to the exposure of the individual to the risk factor. Long observational time, possible errors due to preterm dropout of the participants, the formulation of the hypothesis at the beginning of the study and no possibility to give evidence for rare diseases are the shortcomings of this study design.

In a case-control study, a group of cases is compared to a group of healthy persons (controls) to evaluate possible risk factors using the risk estimate odds ratio. The advantages of this design are relatively short study duration, low costs, and the possibility to investigate different risk factors at the same time. Shortcomings are the problems and possible errors concerning the selection of cases and controls (selection bias) and impaired memory (recall bias).

A cross-sectional study design is used to evaluate the occurrence of a specific disease in the population (prevalence) at a definite point in time. Distinct groups of persons (e.g., different occupations) can be compared for the generation of a hypothesis, however this cannot be tested with this design.

A meta-analysis is performed to analyze and to pool the results of different independent studies that may have inhomogeneous findings using statistical methods. A major concern of meta-analysis is the publication bias, e.g., the tendency of authors more likely to publish a "real effect" such as an increased risk than a publication without a peculiar observation.