Adiponectin in chronic heart failure: influence of diabetes and genetic variants.

BACKGROUND: We hypothesized that, besides type 2 diabetes (T2D) and body mass index (BMI), circulating adiponectin concentration would be associated with variants of the ADIPOQ gene in patients with chronic heart failure (CHF). We also assessed the influence of these confounders on the prognostic value of adiponectin. METHODS: Plasma adiponectin was measured at entry and after 3 months in approximately 1200 patients with CHF enrolled in the GISSI-HF trial. Four common single-nucleotide polymorphisms (SNPs) spanning the ADIPOQ gene were studied: rs17300539 (-11391G→A), rs266729 (-11377C→G), rs2241766 (+45T→G) and rs1501299 (+276G→T). Associations with clinical characteristics and mortality were evaluated in patients with or without T2D. RESULTS: Adiponectin concentrations were negatively related to BMI, higher in women and older persons, but lower in patients with diabetes. T-allele carriers for rs1501299 and A-allele carriers for rs17300539 had significantly elevated adiponectin concentrations. Irrespective of diabetes, baseline plasma adiponectin was independently associated with mortality (adjusted HR [95%CI] per 1 SD increase in adiponectin concentration = 1·24[1·12-1·37], P < 0·0001) and improved prognostic discrimination beyond clinical risk factors (integrated discrimination improvement, P = 0·005). Patients with increasing adiponectin concentration over 3 months had worse outcome than those with stable levels (unadjusted HR = 1·46[1·09-1·96], P = 0·01); this relation was attenuated by the genetic variants examined and by robust confounders like age, diabetes, BMI or NT-proBNP (adjusted HR = 1·37[0·97-1·94], P = 0·075). CONCLUSIONS: Although diabetes and genetic variants at the ADIPOQ gene influence the circulating levels of adiponectin in CHF, higher plasma adipokine levels, but not genetic variants, are consistently associated with a poor prognosis.