Fig. 3: The effects of the IL6 genetic instrument mirror the effects of IL-6 inhibition as captured in clinical trials.

a, IVW MR associations of genetically proxied downregulation of IL-6 signaling through perturbation in IL6 (12-variant instrument, scaled to a 24% decrease in CRP levels) with 8 circulating biomarkers measured in the RESCUE trial⁹. Data are presented as log changes (% change) in levels and error bars correspond to 95% CIs. The IL6 genetic instrument derived was from a CRP GWAS (UK Biobank + CHARGE; n = 575,531). Biomarker associations derived from multiple GWAS datasets as detailed in Supplementary Table 1 (sample sizes vary by biomarkers). b, The effects of 30 mg of ziltivekimab (anti-IL-6 monoclonal antibody) administered subcutaneously every 4 weeks over 12 weeks versus placebo on the 12-week change in 8 circulating biomarkers in the RESCUE trial (led to an 88% placebo-adjusted 12-week decrease in serum CRP)⁹. Data are presented as log changes (% change) in levels; error bars correspond to 95% CIs. RESCUE trial: n = 66 per group (30 mg ziltivekimab versus placebo). c, Correlation of the effects of the IL6 genetic instrument (scaled to 24% decrease in CRP; n = 575,531) and ziltivekimab in RESCUE (30 mg ziltivekimab versus placebo; n = 66 per group) on the 8 biomarkers (Spearman’s ρ = 0.91). Each point represents an effect estimate for a biomarker, as derived by IVW MR with horizontal and vertical error bars indicating the 95% CIs for genetic instrument and ziltivekimab effects, respectively. d, IVW MR associations between genetically downregulated IL-6 signaling through perturbation in IL6 (scaled to a 24% decrease in CRP; n = 575,531) and autoimmune disease outcomes in publicly available GWASs (8,156 cases and 416,495 controls for polymyalgia rheumatica and 35,871 cases and 240,149 controls for rheumatoid arthritis). Data are presented as ORs per 24% decrease in CRP and error bars correspond to 95% CIs. All reported P values are two sided.

Source data