Fig. 1

Explanation of study design and workflow. A: The total effect of IL- 6 on AD, c, was derived using Two-sample Mendelian randomization (i.e., genetically predicted IL- 6 as exposure and AD as outcome). Assumption 1: genetic variation is strongly correlated with the exposure of interest. Assumption 2: genetic variation is not associated with potential risk factors for the outcome (P < 10^–5). Assumption 3: genetic variation influences only the outcome through exposure. B: The total effect was divided into two components: (i) an indirect effect determined through a two-step process (where'a'represents the overall influence of IL- 6 on peripheral immune cells, and'b'represents the impact of peripheral immune cells on AD, adjusted for IL- 6), calculated using the product method (a*b), and (ii) a direct effect (c'—a*b). C: For the mediation of specific peripheral immune cells combined, the indirect effect was derived using the difference method (c—c'). The proportion mediated was then calculated by dividing the indirect effect by the total effect