A cohort study of coffee intake and pancreatic cancer yields a crude relative risk of 2.5. When the data are stratified by sex, the relative risk is 1.0 among men and 4.0 among women, and these stratum-specific estimates differ significantly from each other. The correct conclusion is:
- A Sex is an effect modifier; the crude estimate should not be reported alone ✓
- B Sex is a confounder; the adjusted pooled estimate should be reported
- C The finding is due to chance since stratum estimates differ from the crude value
- D Simpson's paradox has occurred and the crude estimate is valid
Correct answer: A. Sex is an effect modifier; the crude estimate should not be reported alone
Explanation
Effect modification is present when the strength of association itself differs across levels of a third variable, as shown here by RRs of 1.0 versus 4.0. The appropriate response is to report the sex-specific estimates rather than collapse them. In confounding, the stratum-specific estimates are similar to each other but differ from the crude estimate, which is exactly the pattern seen here and rules out option B.
Reference: Gordis Epidemiology, 6th ed.
High-yield for: NEET PGINI-CETNExTFMGEUSMLEPLABMRCP
Written and medically reviewed by the StethoPrep medical team.