|Topic||Cohort and Case-Control Studies|
|Activity||Use the following information to answer Questions 1–3 below. An article from the New York Times with the headline “Prevention: Fish Helps Reduce Risk of Polyps in Women” stated in the first sentence that “A new study has found that consumption of omega-3 fatty acids from fish is associated with a reduced risk for one kind of colon polyp, but only in women.”|
1. How would you design a cohort study to evaluate the association between consumption of omega-3 fatty acids and colon polyps? What are some of the strengths and limitations of your cohort study? Please limit your answer to no more than one paragraph.|
2. How would you design a case-control study to evaluate the association between consumption of omega-3 fatty acids and colon polyps? What are some of the strengths and limitations of your case control study? Please limit your answer to no more than one paragraph.
3. Which study design is better for answering this question?
4. Epidemiologic case-control studies often report increased risk of an event given exposure, but we know that we can only calculate the odds ratio in a case-control study as opposed to a risk ratio. Is it important to distinguish between a risk ratio and an odds ratio? When does the odds ratio approximate the risk ratio? When does it approximate the rate ratio?
5. Is there anything that remains unclear from the lecture or readings on cohort and case-control studies that you would like to discuss further in discussion section this week?
|Topic||Random Error and Confounding|
|Activity||Students are asked to read a New York Times article, “Barnyard Dust Offers a Clue to Stopping Asthma in Children,” that describes a recent study comparing asthma prevalence among Amish and Hutterite children (https://www.nytimes.com/2016/08/04/health/dust-asthma-children.html?_r=0).|
1. How does the design of this study deal with the issue of noncomparability that has plagued prior studies that have tried to contribute evidence to support the Hygiene Hypothesis? What are the potential sources of noncomparability that the authors have circumvented? Are there others that you think they might still be concerned about? Casting your mind back to the module on causal inference, which of Hill’s guidelines are met with the publication of this new study? Do you feel the weight of evidence is now sufficient to prove the Hygiene Hypothesis? Please limit your response to between one paragraph and one page, double spaced.|
2. You are interested in calculating the population average BMI at the Mailman School of Public Health. Would you expect to have a wider or narrower confidence interval if you sampled 25 people vs. if you sampled 200 people?
c. There is not enough information to answer this question.
3. In a study of the possible effect of smoking on weight loss, you are concerned about potential confounding by alcohol consumption, which is positively associated with smoking and negatively associated with weight loss. What is one method you CANNOT use to control for confounding in either the design or analysis phase of this study?
4. In 2–3 sentences, please explain why you chose this response.
5. Is there anything that remains unclear from the lecture or readings on random error and confounding that you would like to discuss further in discussion section this week?