Define confounding in a cohort study and provide an example.

• A. Confounding occurs when a study fails to recruit enough participants.
• B. Confounding occurs when a third variable is associated with both exposure and outcome, distorting the exposure-outcome association.
• C. Confounding arises from misclassification of exposure.
• D. Confounding arises from random variation.

Answer: (B)

Confounding in a cohort study occurs when a third variable is related to both the exposure and the outcome, causing the observed association between exposure and outcome to be distorted. In other words, the apparent effect of the exposure might actually be partly or wholly due to this other variable.

For example, imagine a study looking at coffee consumption and heart disease. If people who drink more coffee are also more likely to smoke, and smoking increases the risk of heart disease, the study might (falsely) attribute the higher heart disease risk to coffee when it’s really due to smoking. Adjusting for smoking (through statistical methods or study design) can reveal the coffee-heart disease relationship more accurately.

The other options involve different kinds of bias or error that aren’t confounding: recruiting too few participants affects study power; misclassifying exposure causes information bias; random variation is sampling error.