A case-control study yields a crude odds ratio of 2.0 for alcohol consumption and esophageal cancer. After stratification by smoking status, the odds ratio is 1.3 among smokers and 1.3 among non-smokers. The most appropriate conclusion is:
- A Smoking modifies the effect of alcohol on esophageal cancer
- B The crude estimate is valid and should be reported unchanged
- C Smoking confounds the association, and the adjusted estimate of 1.3 should be reported ✓
- D Alcohol is causally unrelated, so stratification introduced bias
Correct answer: C. Smoking confounds the association, and the adjusted estimate of 1.3 should be reported
Explanation
When stratum-specific estimates are identical to each other but differ from the crude estimate, a confounder is distorting the crude value, and the common adjusted estimate should be reported. Effect modification would show stratum-specific odds ratios that differ appreciably from one another, requiring separate reporting for each stratum. Here the uniform 1.3 across strata rules out modification and identifies smoking as a confounder of the crude association of 2.0.
Reference: Gordis Epidemiology, 6th ed.
High-yield for: NEET PGINI-CETNExTFMGEUSMLEPLABMRCP
Written and medically reviewed by the StethoPrep medical team.