A cohort study of a dietary factor and myocardial infarction shows a crude relative risk of 1.8. When stratified by sex, the relative risk is 4.2 in men and 1.0 in women. The Mantel-Haenszel adjusted estimate differs little from the crude value. The correct conclusion is that sex acts as:

• A A confounder, and the adjusted pooled estimate should be reported
• B An effect modifier, and stratum-specific estimates should be reported separately ✓
• C A source of information bias, and misclassification correction is needed
• D An intermediate variable, and stratification should be abandoned

Explanation

When the stratum-specific estimates differ substantially from each other, the single summary measure misrepresents reality regardless of whether the adjusted value resembles the crude one; this is effect modification, and the honest presentation is separate estimates for men and women. Confounding requires strata to be homogeneous but shifted from the crude value, in which case adjustment is appropriate. Sex precedes the exposure and cannot lie on its causal pathway, ruling out an intermediate variable.

Reference: Rothman, Greenland and Lash, Modern Epidemiology, 3rd ed.

High-yield for: NEET PGINI-CETNExTFMGEUSMLEPLABMRCP

Written and medically reviewed by the StethoPrep medical team.