Comparative analysis of three nutrition scores in predicting mortality after acute myocardial infarction

We investigated the utility of nutrition scores in predicting mortality and prognostic importance of nutrition status using three different scoring systems in patients with acute myocardial infarction (AMI).In total, 1147 patients with AMI were enrolled in this study (72.5 % men; mean age 65.6 years). Patients were divided into three groups according to the geriatric nutritional risk index (GNRI); prognostic nutritional index (PNI); and triglycerides, total cholesterol, and body weight index(TCBI) scores as tertile: low (GNRI ≤ 103.8, n = 382), intermediate (103.8 < GNRI ≤ 112.3, n = 383), and high (GNRI > 112.3, n = 382) GNRI groups; low (PNI ≤ 50.0, n = 382), intermediate (50.0 < PNI ≤ 56.1, n = 383), and high (PNI > 56.1, n = 382) PNI groups; and low (TCBI ≤ 1086.4, n = 382), intermediate (1086.3 < GNRI ≤ 2139.1, n = 383), and high (TCBI > 2139.1, n = 382) TCBI groups.In the GNRI, TCBI, and PNI groups, the cumulative incidence of all-cause death and major adverse cardiovascular events (MACEs) was significantly higher in the low score group, followed by the intermediate and high score groups. Moreover, both intermediate and low PNI groups had a similar cumulative incidence of all-cause death and MACE. The GNRI score (AUC 0.753, 95% CI 0.608~0.745, P = 0.009) had significantly higher areas under the curve (AUCs) than the TCBI (AUC 0.659, 95% CI 0.600~0.719, reference) and PNI (AUC 0.676, 95% CI 0.608~0.745, P = 0.669) scores.Patients with low nutrition scores were at a higher risk of MACE and all-cause death than patients with high nutrition scores. Additionally, the GNRI had the greatest incremental value in predicting risks among the three different scoring systems used in this study.