Imagine that you had to design a prospective cohort study to test the following
ID: 200093 • Letter: I
Question
Imagine that you had to design a prospective cohort study to test the following hypotheses: A diet high in cholesterol affects the risk (or rate) of cardiovascular disease among men over 50 years old. To help you with your design (especially question 9a and 9b), some of the other known causes of cardiovascular disease are: family history of cardiovascular disease (certain genes), obesity (body mass index > 30), and smoking. Assume for this assignment that these other factors have fairly strong evidence showing that they cause diabetes.
9a. Complete the following table to summarize the data you plan to collect. Be specific about your definitions of the variable for a 2x2 table (for example – what is your case definition for cardiovascular disease, how would you define your diet as low/high in cholesterol?)
Data to Collect
Specify variable
(e.g. diabetes, physical activity, etc.)
Definition of variable
(How will you categorize this variable for a 2x2 table?)
Exposure
Outcome
Confounders
9b. Describe one of the plausible confounders in your study. Describe why it is a confounder and why it might bias the interpretation of your results. Be specific about which direction this confounder is likely to bias your results – i.e. how does your measured association compare to what the true underlying cause-effect relationship would be.
9c. Describe a plausible 3rd variable that might “interact with” your exposure (serve as an effect modifier). How would your results look because of the presence of this interacting variable/effect modifier? Feel free to use fictitious numbers/rates if it facilitates your explanation.
Data to Collect
Specify variable
(e.g. diabetes, physical activity, etc.)
Definition of variable
(How will you categorize this variable for a 2x2 table?)
Exposure
Outcome
Confounders
Explanation / Answer
because of presence of interacting variable effect is feel to free that it faciliated numner