The excess mortality was predominantly related to congestive heart failure. The difference in outcome between ATHENA and ANDROMEDA could be a consequence of the fact that ANDROMEDA enrolled only patients with advanced heart failure and recent decompensation resulting in hospitalization. In contrast, ATHENA specifically excluded sufferers who had either hemodynamic instability or serious heart failure. It is therefore possible that dronedarone increases cardiovascular mortality among sufferers with advanced and lately decompensated congestive heart failing but reduces cardiovascular mortality in individuals with less severe heart failure.As referred to in the Supplementary Appendix, we used proportional-hazards regression versions to examine the partnership of plasma APOC3 levels to the chance of incident cardiovascular system disease. Secondary Analyses In patients with angiographic coronary heart disease, we evaluated the association of plasma APOC3 levels with total and cardiovascular mortality. We also examined the association of APOC3 mutations with the current presence of hepatic unwanted fat on computed tomography . Information on the methods are offered in the Supplementary Appendix. In the analyses of associations with cardiovascular system disease, we tested four variants and thus considered a P worth of significantly less than 0.0125 to be significant. Genetic association analyses were performed with use of PLINK .