Should the Cohort study start from just one type of exposure or can it be more exposures? Would you give examples of that study
HI Azam
Not you can assess multiple exposure simultaneously. But the issue here would be the non exposed group should comprise all those with none of the risk factors under investigation. to make things easier it is wise to start with a single exposure, while collecting data on others factors that you will consider as confounding variables.
In a cohort study you can study multiple disease however in a case control study you can study multiple risk factors
Cohort studies start from exposure and end with outcom (disease, relapse, death and ...). Generally we follow a set of risk factors (or one specific risk factor ) and search for our interest outcome. However, as we know like Framingham Cohort Study you can consider a set of risk factors and some diseases.
The above answers are very reasonable. It is important to concentrate on one exposure . however, it might be difficult to exclude other exposures or risk factors in one group. therefore the study design is based on one important exposure but confounders must not be ignored in the stage of analysis.
We can concentrate on one exposure, and during the study you can consider the other exposures and at the end of the study you may have other outcomes. During analysis, analyse each outcome with its tested exposure and consider the confounders.
Hi AZAM
I concur with most of the above answers. I just wanted to add that it really depends on your research question. You may be interested in one particular exposure in which case you can clearly identify exposed and unexposed groups at the beginning of the study. But this is not always the case. Let say you have a set of predictors to investigate, then you will start with information on the set of exposures but the outcome is absent (and that will be the most important) and then end with the outcome.
I think you have very good answers, all of the above. I would just like to add that during the design of your cohort study, you must be clear about what you are interested in (as the definition of your "exposed" and "not exposed" groups will depend on this), you may have more than one factor of interest, but that may increase the complexity of your observation. You should take into account that your outcome (and your exposure as well) might be affected by other factors present in each of your groups and therefore you must consider them in the analysis (confounders, effect modifiers). Moreover, the greater complexity of the cohort study, is that you have to follow your cohort and comparison group for enough time as to be able to see the effect (if it is prospective non of your participants can show any signs of the effect during selection). And this is what makes them very sensitive to changes not anticipated during observation (such as changes from not exposed category to exposed or vice versa). They are not practical for the study of rare events (that is why I told you earlier that you must know your effect enough to predict the different biases that may affect the plausibility of your study). If you need more help on this, please check the attached file.
Thank you very much Marcel Kitenge, Taddese Alemu Zerfu, Nida Zahid, Mostafa Dianatinasab, Omran S Habib, Dalia G Mahran, Tahir Kemal Şahin, Tieba Millogo & Angela Gala Gonzalez for the complete and comprehensive answer
Dear friends
Very good answers above.
The cohort study (except for the historical cohort study) usually considers only one factor at a time because of the precise distribution between exposed and unexposed at the start of the follow-up period.
However, even in a prospective cohort study, it is quite possible to consider several factors at once. The Framingham Heart Study, whose cohort is made up of the population of this city, studied the relationship between several risk factors (cholesterol, smoking, obesity, alcohol consumption ...) and cardiovascular diseases. This study, which began in 1950, continues.
In the case of nosocomial infections, a single cohort of patients hospitalized in a service for a given period of time could be considered, and their members would then be divided into different exposure categories with a view to identifying factors of risk.
It is important to note that, if several factors are to be studied simultaneously, the group defined as unexposed is composed of individuals without any of the risk factors to be evaluated.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis