The initial observed AR(p) difference was maintained: 36.3% (95% CI: [24.3-50.3]) in study A and 69.7% (95% CI: [51.7-83.2]) in study B. Further investigations highlighted the potential selection bias introduced in both studies, especially among controls.