Abstract Background Ventricular arrhythmia (VA) is a frequent cause of sudden cardiac death (SCD). Treatment to prevent SCD includes device therapy with implantable cardioverter defibrillator (ICD). Although the electrophysiology underlying VA have been described, the molecular understanding of VA risk remains unclear. Purpose To identify proteins that are associated with higher and lower risk of VA. Methods SMASH 1 Study was a prospective observational study of 490 patients treated with ICD. The plasma proteome was analyzed with the Olink Explore 384 Cardiometabolic platform at study inclusion, and patients were followed for mean 3.1± 0.7 years. VA events were defined as adjudicated ventricular fibrillation or sustained ventricular tachycardia registered from ICD recordings and clinical events were recorded from electronic health records. Protein concentrations were quantified relatively as normalized protein expression on a log2-transformed concentrations where a large number represents higher protein level in the sample. Results The mean age was 66±12 years, 83% were male and 51% had a primary prevention ICD-indication. Episode(s) of VA occurred in 137 patients (28%) during follow-up. In total, 353 distinct proteins were successfully analyzed in all patients. Among these, high B-type natriuretic peptide (BNP) and N-terminal pro-BNP (NT-proBNP) levels were associated with a higher risk of VA, while Apolipoprotein M (APOM) was the only protein significantly associated with a lower risk of VA after multiple testing using Benjamini-Hochberg correction (Figure, Panel A). After adjusting for age, sex, body mass index, coronary artery disease, heart failure, renal function, left ventricular ejection fraction and antiarrhythmic medication, only low APOM levels remained significantly associated with VA: HR 0.51 (95%CI 0.32-0.79] per log unit increase, p=0.003. Patients in the lowest quartile of APOM had a particularly high risk of incident VA: HR 2.26 (95%CI 1.60-3.19), p=3.61x10-6, compared to quartile 2-4 (Figure, Panel B). High APOM levels were also associated with a lower risk of HF hospitalizations (HR 0.25 [95%CI 0.16-0.40], p<0.001) and all-cause mortality (HR 0.36 [95%CI 0.22-0.60], p<0.001), and these associations remained significant in adjusted models. Conclusions Low levels of APOM are associated with a greater risk of incident VA, independently of established risk factors. Low APOM also predicted risk of HF hospitalizations and mortality in our cohort, which suggest protective cardiovascular effects of APOM. Figure: Association of individual proteomic measures with risk of ventricular arrhythmias (VA).A:Volcano plots of the associations of individual plasma proteins with incident VA. The dotted line represents the significance level at false discovery rate <0.05 and the red dots are significant proteins after multiple testing using Benjamini-Hochberg correction.B: Risk of VA by quartile of Apolipoprotein M.