Problem
In a hypothetical study to examine the association between obesity with coronary heart disease (CHD) 9,400 patients were selected to participate. Participants were classified as obese or not and followed over five years to assess if they were diagnosed with coronary heart disease. The results are shown in the table below:
|
|
CHD
|
No CHD
|
Total
|
|
Obese
|
79
|
745
|
824
|
|
Not Obese
|
286
|
8290
|
8576
|
|
Total
|
365
|
9035
|
9400
|
1. What is the exposure variable? What is the disease?
2. Calculate the relative risk.
3. Can you think of any confounding variables? What are they?
The investigators suspected that people who have high cholesterol are more likely to have CHD.
As part of the design of the study, the investigators obtained the blood cholesterol levels of the participants. The information was classified as high when the blood cholesterol was 240 mg/dL or higher. The normal blood cholesterol is less than 240 mg/dL.
1. Of the 79 people who were obese and had CHD, 55 had high blood cholesterol and 24 had normal cholesterol.
2. Of the 745 people who were obese and did not have CHD, 51 had high bold cholesterol levels and 694 had normal blood cholesterol.
3. Of the 286 people who were not classified as obese and had CHD, 5 had high blood cholesterol.
4. Of the 8,290 people who were not obese and did not have CHD, 5 had high levels of blood cholesterol.
• Is high blood cholesterol a confounding variable? (Show your work to provide evidence).
• Based on this study design, what are possible the sources of bias? Why?