Incidence records how often new outcomes occur during observation, while risk expresses the probability of an outcome in a defined group over that period. Relative risk compares the risk among exposed participants with the risk among unexposed participants. Together, these measures show both the occurrence of an outcome and its association with the exposure.
Statistical adjustment helps account for confounding variables, which are factors related to both the exposure and the outcome that can distort an observed association. By considering these variables during analysis, researchers can produce comparisons that better reflect the exposure-outcome relationship. This is especially important when exposed and unexposed groups differ in relevant characteristics.
A prospective design collects exposure information and then follows participants forward as outcomes occur. A retrospective design uses existing records to reconstruct exposure and outcome information from an earlier period. The distinction concerns how the data are assembled, while both designs compare outcomes between defined exposure groups and can examine later health effects.
Follow-up time determines how long researchers can observe outcomes and calculate measures such as incidence or risk. Loss to follow-up reduces the available outcome information and can weaken reliability, particularly if participants who leave differ from those who remain. Careful attention to observation duration and participant retention is therefore essential when interpreting findings.
Researchers first classify participants according to the relevant exposure or characteristic, then collect or identify outcome information over the observation period. They compare outcome occurrence between groups using measures such as incidence, risk, or relative risk, and may apply statistical adjustment for confounding variables. The resulting estimates describe the observed relationship between exposure and outcome.
A cohort study is useful when assigning an exposure would be impractical or unethical. Instead of deliberately changing participants’ circumstances, researchers observe existing exposed and unexposed groups and track their outcomes. This approach supports investigation of potential causes and estimation of disease risk while avoiding the ethical or practical barriers associated with randomized experiments.
Cohort studies can examine whether a prior exposure or characteristic is associated with later disease or other health outcomes. Their longitudinal structure allows researchers to estimate outcome risk and compare its occurrence across groups. Consequently, they are suited to questions about potential causes, disease development, and health effects that emerge over time.