The Polyscore of autonomic parameters predicts mortality and identifies low-risk individuals among diabetic survivors of acute myocardial infarction.

Survivors of an acute myocardial infarction with diabetes mellitus retain an increased mortality risk. Reliable assessment of individual risk is required for effective and cost-efficient medical care in these patients. The Polyscore is a previously established risk predictor consisting of seven autonomic tests derived from electrocardiogram, blood pressure, and respiration. The Polyscore allows classification of survivors of myocardial infarction in groups at low, intermediate and high mortality risk. The aim of this study was to investigate the prognostic value of the Polyscore in diabetic survivors of acute myocardial infarction, which may be impaired by the presence of diabetic autonomic neuropathy. Survivors of an acute myocardial infarction were included in a prospective cohort study during hospitalisation due to the index event at two university hospitals in Munich, Germany. The Polyscore was determined from simultaneous non-invasive 30-min recordings of electrocardiogram, continuous arterial blood pressure, and respiration which were performed in all participants. Patients were followed for 5 years. The primary and secondary outcomes were all-cause mortality and cardiac mortality. 184 of 941 enrolled patients (19.6%) suffered from diabetes mellitus. 5-year-mortality was higher in diabetic patients (15.2%) compared to non-diabetic patients (5.8%). A multivariable Cox regression model confirmed the Polyscore as a strong predictor of mortality in diabetic post-MI patients (intermediate risk: HR 6.56, 95% CI 1.61-26.78, p = 0.004, mortality 22.8%; high risk: HR 18.76, 95% CI 4.35-80.98, p < 0.001, mortality 68.8%). There was no interaction between diabetes mellitus and the Polyscore regarding mortality prediction (p = 0.775). Interestingly, in contrast to the groups at intermediate and high risk (73 patients, 39.7%), the Polyscore identified a majority of diabetic patients (111, 60.3%) with a low mortality risk, comparable to that of low-risk non-diabetic patients (3.6% and 2.1%, respectively, p = 0.339). Consistent results were observed for cardiac mortality. This analysis shows that the Polyscore predicts all-cause and cardiac mortality in diabetic survivors of acute myocardial infarction. Within these patients it identifies a large population not affected by the excess mortality associated with diabetes in this setting. Thus, the Polyscore may facilitate risk-adapted follow-up strategies in diabetic survivors of myocardial infarction.