An epidemiological study to answer the following research

Imagine you have been asked to help design an epidemiological study to answer the following research question: is consumption of coconut oil associated with lower lifetime risk of developing Alzheimer’s disease among people living in Canada? You contemplate three different study designs that could be used to answer this question: a cross-sectional survey, a case-control study, and a prospective cohort study.

1. First, consider the cross-sectional design. You create an online survey that gets posted to Facebook groups for seniors who are interested in information about healthy ageing. Comment on how selection bias may be a concern with this study design.

2.  Next, consider the case-control design. You sample cases from an Alzheimer’s and brain health clinic at a hospital. Controls are sampled from the physiotherapy (rehabilitation) clinic at the same hospital. Comment on how selection bias may be a concern with this study design.

3. Suppose you give all selected cases and controls and questionnaire to complete, asking about their lifetime dietary habits, including a range of items including coconut oil consumption. Comment on how information bias may be a concern with this study design, and indicate whether you expect misclassification of exposure to be differential or non-differential.