Home Statistical Dictionary Cohort Study
Cohort Study
An observational study design that follows a group of individuals sharing a defining characteristic (a cohort), classified by exposure status, forward over time to compare the rate at which an outcome develops in exposed versus unexposed members.
In Plain English
A cohort study follows the natural, forward direction of cause and effect: you identify a group of people, classify them by whether or not they have some exposure of interest (smokers vs. non-smokers, say), and then follow both groups forward through time to see who develops the outcome you're studying and at what rate. This forward-looking structure lets cohort studies directly estimate incidence and relative risk in a way a case-control study's backward-looking design cannot, but it comes at a real cost, following a large group of people over a long period, especially for a rare outcome, requires substantial time, resources, and vulnerability to attrition, which is exactly the trade-off that makes case-control studies the more practical choice for investigating rare outcomes instead.
Definition
A cohort study enrolls a group of individuals (the cohort), classifies them by exposure status to the risk factor under investigation, and follows them prospectively (forward in time) to compare the incidence of the outcome between exposed and unexposed members, allowing direct estimation of relative risk (the ratio of incidence rates between exposed and unexposed groups), a quantity case-control studies cannot directly estimate. Cohort studies can be prospective (the cohort is assembled and followed forward from the present) or retrospective/historical (using existing historical records to reconstruct exposure status and outcomes that have already occurred, but still classifying by exposure and comparing subsequent incidence in the same forward logical structure); cohort studies are well suited to studying multiple outcomes from a single exposure and to establishing a clear temporal sequence between exposure and outcome, but they are resource-intensive, particularly for rare outcomes requiring very large cohorts or long follow-up periods, and are vulnerable to attrition (loss to follow-up) that can bias results if it is differential between exposure groups.
Properties
- Cohort studies can directly estimate incidence rates and relative risk, a fundamental advantage over case-control studies, because the cohort is assembled and followed based on exposure status rather than outcome status, the natural population-level rates of the outcome in each exposure group are preserved and directly observable.
- A single, well-designed cohort study can investigate multiple different outcomes simultaneously from the same underlying exposure classification, a practical efficiency case-control studies (which must be specifically designed around a single outcome of interest, the one used to define cases) cannot easily replicate.
- Cohort studies are substantially more resource-intensive than case-control studies, particularly for rare outcomes, since a large cohort must be followed, often for years, to accumulate enough outcome events for reliable analysis, and are specifically vulnerable to attrition bias if participants who drop out differ systematically, by exposure status or by outcome risk, from those who remain.
At a Glance