Which statement best differentiates relative risk (RR) and odds ratio (OR) and when each is most appropriate?

• A. RR compares the odds of exposure among cases vs controls; OR compares risk in exposed vs unexposed in cohort studies; RR is intuitive for cohort data; OR approximates RR when the outcome is rare.
• B. RR compares risk in exposed vs unexposed in cohort studies; OR compares the odds of exposure among cases vs controls; RR is intuitive for cohort data; OR approximates RR when the outcome is rare.
• C. RR and OR are the same; They have no difference.
• D. RR is used only for cross-sectional studies; OR is used only for experimental studies.

Answer: (B)

At its heart, this compares how a measure of association is derived from study design and what each statistic is best at capturing. Relative risk directly compares probabilities: in a cohort you start with people who are exposed and unexposed and then see how many in each group develop the outcome. RR = risk in exposed divided by risk in unexposed, so it feels intuitive because you’re looking at actual risk in each group.

Odds ratio comes from case-control designs, where you start by selecting people with the outcome (cases) and without the outcome (controls) and then look back at exposure. It compares the odds of having been exposed among cases versus controls. Because you don’t know the true incidence in a case-control sample, you can’t compute risk, but you can compute odds, and the OR captures the strength of association in that framework. When the outcome is rare in the population, the odds ratio closely approximates the relative risk, which is why OR is frequently used and reported in case-control studies and in logistic regression. As the outcome becomes more common, the OR can overstate the association relative to the RR, which is why interpretation matters.

The other options mix up what each measure compares and are not appropriate descriptions of how RR and OR relate to study designs.