In a case-control study of physical activity and coronary heart disease, investigators match cases and controls on body mass index before analyzing the association. After matching, the odds ratio falls close to unity even though stratified analysis suggests a real effect. The most likely explanation is:
- A Overmatching, because body weight lies partly on the causal pathway between activity and heart disease ✓
- B Recall bias introduced by matching
- C Berkson's bias due to hospital-based control selection
- D Effect modification by body weight masking the true association
Correct answer: A. Overmatching, because body weight lies partly on the causal pathway between activity and heart disease
Explanation
Matching on a variable that is itself influenced by the exposure, or otherwise part of the causal chain, removes real differences in exposure between cases and controls and drives the estimate toward the null. This is overmatching. Berkson's bias relates to hospital-based sampling, not matching choices, and effect modification is a finding to be explored, not a mechanism producing spurious nullity here.
Reference: Modern Epidemiology (Rothman, Greenland, Lash), 4th ed.
High-yield for: NEET PGINI-CETNExTFMGEUSMLEPLABMRCP
Written and medically reviewed by the StethoPrep medical team.