Abstract Body: Introduction: Previous studies have suggested that increased body fat is associated with elevated cardiovascular disease (CVD) risk. We have previously shown that apart from common vascular risk factor modification, adipose tissue-specific factors, such as adiponectin, may mediate the relationship between body fat and vascular risk. Here, we use large-scale omics data to explore whether pharmacologic targeting of the adiponectin pathway may contribute to lowering CVD risk.

Methods: To proxy the effect of pharmacologically perturbing the adiponectin pathway we used a drug-target mendelian randomization (MR) approach. We selected genome-wide, independent (r²<0.1) variants in the vicinity (300 kb) of the ADIPOQgene from a GWAS of circulating adiponectin levels (n=~38,600). Using these variants, we performed MR on ischemic stroke (n=62,100)/subtypes (large artery=6,399; cardioembolic=10,804; small vessel=6,811), white matter hyperintensities (WMH) (n=42,310), and coronary artery disease (CAD) (n=181,522) using GWAS data of the respective traits. We also leveraged pQTL data from two proteomics studies (deCODE; 35,559 Icelanders; 4,719 proteins, and UK Biobank Pharma Proteomics Project; ~50,000 participants; ~3000 proteins) to perform MR at the proteome-wide level, with the aim of identifying the downstream proteomic effect of adiponectin targeting.

Results: Fourteen genetic variants in the ADIPOQ locus were associated with circulating adiponectin levels. Higher genetically-proxied, ADIPOQ-mediated circulating adiponectin levels were associated with lower WMH (IVW β=-0.131 per μg/ml increase in adiponectin levels; 95% CI: (-0.217) - (-0.044); p=0.003) and CAD (IVW OR 0.93 per μg/ml increase in adiponectin levels; 95% CI: 0.88-0.98; p=0.006) risk, but not ischemic stroke/subtypes (all p-values > 0.05). Results were consistent in sensitivity analyses. After FDR correction per proteomic dataset (p-FDR < 0.05), higher genetically-proxied, ADIPOQ-mediated circulating adiponectin levels were associated with proteins involved in various pathways including the inflammatory (e.g. CXCL17, CRP), thrombogenic (e.g. factor X, plasminogen), cardiometabolic (e.g. IGFBPL1, NOTCH3), and oncologic (e.g IL9) pathways (Figures 1, 2).

Conclusions: Pharmacologic targeting of circulating adiponectin levels may be associated with decreased risk of CAD and covert brain damage. The extent that this effect is driven by pleiotropic mediatory proteomic pathways warrants further research.