Cohort studies are a type of medical research used to investigate the causes of disease, establishing links between risk factors and health outcomes.
Cohort studies are usually forward-looking - that is, they are "prospective" studies, or planned in advance and carried out over a future period of time.
Retrospective cohort studies look at data that already exists and tries to identify risk factors for particular conditions; researchers are inherently limited in their interpretations because they cannot retrospectively gather missing data.
In a prospective cohort study, researchers first raise a research question, forming a hypothesis about the potential causes of a disease. The researchers then observe a group of people, the cohort, over a period of time (often several years), collecting data that may be relevant to the disease. This allows the researchers to detect any changes in health in relation to the potential risk factors they have identified.
For example, scientists may ask participants to record specific lifestyle details over the course of a study, before going on to analyze any possible correlations between lifestyle factors and disease.
- Cohort studies are used by epidemiologists looking into the factors that affect the health and illness of populations.
- Other terms used to describe cohort studies include: incidence, longitudinal, forward-looking, follow-up, concurrent, retrospective, historic and prospective.
Prospective cohort studies offer one way for researchers to investigate the etiology of a disease, i.e. the cause(s) or root of a disease. When doctors talk about the etiology of a condition they can mean both the cause and the mechanism by which a cause leads to a disease effect.
Finding causes of disease - the best available method
The cohort study design is the best available scientific method for measuring the effects of a suspected risk factor. While randomized controlled trials (RCT) are considered the best, most rigorous way of investigating interventional medicine, they are unethical for testing the causes of disease.
A birth cohort study is a long-term follow-up of people born in the same year. One has followed 17,000 people all born in the same week in 1958.
Cohort studies are observational - the researchers simply observe what happens, without applying any intervention themselves. Conversely, experimental studies, such as RCTs, involve an intervention by the scientists - the introduction of a drug, for example. MNT has more information about randomized controlled trials.
When looking for the causes of disease, it would be unethical to deliberately expose participants to a suspected risk factor (such as would be the case in an RCT). Instead, the design of a prospective cohort study is observational rather than interventional.
Randomized controlled trials in humans are used to test the safety and potential benefit of a treatment. While the harms of a treatment sometimes prove to outweigh the benefits, this form of testing is acceptable to the participants because the investigators are aiming at the outset to develop a new treatment and usually have a reasonable expectation of safety at least, if not a positive effect of treatment.
MNT has produced an explanation of how medical research evidence is graded, offering an overview of the best available study methods for answering different types of medical research questions.
Here are a few key points about cohort studies. More detail and supporting information is in the body of this article.
- Cohort studies typically observe large groups of individuals, recording their exposure to certain risk factors to find clues as to the possible causes of disease.
- They are usually prospective studies that gather data going forward, rather than retrospective cohort studies that look at data already collected.
- The Nurses' Health Study is one example of a large cohort study, and it has produced many important links between lifestyle choices and health by following hundreds of thousands of women across North America.
- Such research can also find social factors that influence health. Thousands of babies all born in the same week in the UK have been followed by a number of large birth cohort studies investigating medical and social factors.
Examples of cohort studies
One famous example of a cohort study is the Nurses' Health Study - a massive, long-running analysis of women's health, originally set up in 1976 to investigate the potential long term consequences of the use of oral contraceptives.
This study recruited its second generation cohort for the Nurses' Health Study II in 1989, and its third-generation cohort of nurses from across the US and Canada in 2010. The nurses in the first NHS were married women aged 30-55; the NHS II and III aimed to look at a more diverse cohort including women aged between 20 and 46.
Numerous and important insights into health and wellbeing have already been gained by researchers using data from the Nurses' Health Study, which is run by the Harvard School of Public Health, and the Brigham and Women's Hospital, both based in Boston, MA.
The following headlines from news stories published by MNT report some of the findings from this huge study of hundreds of thousands of women:
- Can breast cancer risk be predicted by skin moles?
- Multiple sunburns as an adolescent increases melanoma risk by 80%
- Tinnitus less common in women who drink more coffee
- Eating nuts every day may prolong life
- Mediterranean diet linked to longer lifespan and better health.
Because the Nurses' Health Study asks participants about their lifestyle choices, it has yielded much information about the harms and benefits of various factors, including specific types of food in the diet.
Cohort studies are also good at finding relationships between health and environmental factors such as chemicals in the air, water and food - issues that the World Health Organization (WHO) helps researchers to investigate with large-scale cohort studies.
WHO is helping to coordinate separate cohort studies so that their results can be pooled. The larger amounts of data produced in this way would enable the analysis of rare conditions such as some types of cancer.
Big cohorts of babies
In 1958, researchers in the UK launched what has become a highly-regarded large-scale cohort study began that has followed 17,000 people all born in the same week in different regions of Great Britain.
Since then, researchers from the UK's Centre for Longitudinal Studies have launched more studies with new large groups of babies.
The latest, the Millennium Cohort Study, is following 19,000 millennium babies, children born in the UK in 2000-2001. In addition to data on the health of these children and their parents, the study is also looking into child behavior and cognitive development, plus a host of social factors.
How good are cohort studies at finding causes of disease?
Epidemiology and causes of disease are investigated by two types of observational study - cohort and case-control studies.
Cohort studies are considered to be better than case-control studies because they are usually "prospective" whereas the latter - being 'retrospective' - have a number of limitations. MNT has more information about case-control studies.
While cohort studies are graded as the most robust form of medical research after experiments such as randomized controlled trials, they are not always the best form of observational work.
Cohort studies do have some limitations:
Cohort studies are usually forward-looking and can only give clues about disease causes and not definitive proof of links between risk factors and health.
- They are less suited to finding clues about rare diseases - a case-control study identifies cases of disease first and then analyzes exposure to risk factors, whereas cohort studies follow exposure data and watch for any emerging cases of disease.
- They are typically unsuitable for identifying the cause(s) of a sudden outbreak of disease - case-control studies can give quicker results.
- They are expensive to run and usually take many years - often decades - to produce results.
- They can only offer clues about the causes of disease, rather than definitive proof of links between risk factors and health (as with any observational medical research).