P5363Resistin is an independent prognostic factor in type 2 diabetes

Abstract

Background: Subclinical inflammation has been linked to insulin resistance, as well as to vascular complications of type 2 diabetes (T2D). Although in rodent models resistin has been shown to exert metabolic effects, in humans, it is considered mainly an inflammatory marker. Purpose: To investigate the association of resistin and other inflammatory markers with mortality and long-term complications of T2D. Methods: The analysis included 303 T2D patients (pts) - participants of the AVOCADO study. Apart from basic laboratory parameters, baseline resistin, tumor necrosis factor α (TNF-α) and high-sensitivity C-reactive protein concentrations were measured. Pts were followed for a median of 5.4 years. The primary endpoint was death at follow-up (FU). The secondary endpoints included: MACE 1 - a composite of death, acute coronary syndrome (ACS) and stroke or transient ischemic attack (TIA), and MACE 2 - a composite of death, ACS, stroke/TIA, need for coronary or peripheral revascularization, hospitalization for heart failure and need for amputation. Baseline clinical and laboratory parameters were included in the Cox proportional hazards model - all variables found to be predictors of clinical endpoints in univariate analyses were then included in multivariate analyses. Receiver operating characteristic (ROC) analysis was performed, as well as Youden's J statistic to determine the optimal biomarker cut-off point for the prediction of the primary endpoint. Results: At baseline, median age was 68 years, median duration of T2D - 8 years, and 53% of pts were male. Data on death at FU were obtained for all pts: 34 pts (11%) died during FU. In univariate analyses, predictors of death at FU included: age, previous myocardial infarction, glomerular filtration rate, low-density lipoprotein, hemoglobin, TNF-α and resistin concentrations. In multivariate analysis, only older age and higher resistin were predictive of death at FU. In ROC analysis, area under the curve (AUC) for resistin was 0.7. Resistin concentration of 11.4 ng/mL or more had sensitivity of 41%, specificity of 91%, a negative predictive value of 92%, and a positive predictive value of 36% for prediction of death at FU (Youden's index). Data on secondary endpoints were available for 220 pts, of whom 49 pts (22%) reached MACE 1, and 70 pts (32%) reached MACE 2. In univariate analyses, both resistin and TNF-α (among other variables) were predictive of MACE 1 and MACE 2. In multivariate analyses, resistin, but not TNF-α, was an independent predictor of both secondary endpoints. Conclusions: Higher resistin is associated with reduced survival in T2D, irrespectively of other inflammatory markers. Resistin concentration of 11.4 ng/mL or more indicates T2D pts at the highest risk of unfavorable outcomes. The results of our study suggest that resistin might affect long-term prognosis in T2D not only through its inflammatory properties, but also via other, e.g. metabolic, pathways.