Biological senescence risk score. A practical tool to predict biological senescence status

Background Ageing and biological senescence, both related to cardiovascular disease, are mediated by oxidative stress and inflammation. We aim to develop a predictive tool to evaluate the degree of biological senescence in coronary patients. Methods Relative telomere length (RTL) of 1002 coronary patients from the CORDIOPREV study (NCT00924937) was determined at baseline in addition to markers of inflammatory response (hs‐C‐Reactive Protein, monocyte chemoattractant protein‐1, IL‐6, IL‐1β, TNF‐α, adiponectin, resistin and leptin) and oxidative stress (nitric oxide, lipid peroxidation products, carbonylated proteins, catalase, total glutathione, reduced glutathione, oxidized glutathione, superoxide dismutase and peroxidated glutathione). Biological senescence was defined using the cut‐off value defined by the lower quintile of relative telomere length in our population (RTL = 0.7629). We generated and tested different predictive models based on logistic regression analysis to identify biological senescence. Three models were designed to be used with different sets of information. Results We selected those patients with all the variables proposed to develop the predictive models (n = 353). Statistically significant differences between both groups (Biological senescence vs. Nonbiological senescence) were found for total cholesterol, catalase, superoxide dismutase, IL‐1β, resistin and leptin. The area under the curve of receiver‐operating characteristic to predict biological senescence for our models was 0.65, 0.75 and 0.72. Conclusions These predictive models allow us to calculate the degree of biological senescence in coronary patients, identifying a subgroup of patients at higher risk and who may require more intensive treatment.