Case 2 (Questions 19 and 20) Suppose a team of researchers c…
Case 2 (Questions 19 and 20) Suppose a team of researchers conducted a prospective cohort study on the effect of physical activity on the 6-month risk of depression. To assess potential effect modification by gender, the researchers estimated the effect of physical activity on depression separately for men and women. In men, the estimated risk difference was -0.11 and the estimated risk ratio was 0.71. In women, the estimated risk difference was -0.07 and the estimated risk ratio was 0.80. Based on these effect estimates the researchers concluded that gender is indeed an effect modifier, so the researchers decided to report the conditional effect estimates in their manuscript. After submitting the manuscript for publication, a reviewer comments that gender may also be a confounder of the effect of physical activity on depression and that it would be important to adjust for gender in the analyses. The team of researchers from case 2 agrees with the reviewer that gender may be a confounder, however they argue that their conditional effect estimates are already adjusted for confounding by gender. Please describe how stratification addresses: Confounding by gender in case 2 (max. 3 points). Effect modification by gender in case 2 (max. 3 points).