The association of concomitant serum potassium and glucose levels and in-hospital mortality in patients with acute myocardial infarction (AMI). Soroka acute myocardial infarction II (SAMI-II) project.

Background: Acute myocardial infarction (AMI) is associated with significant systemic metabolic changes. These changes include increased plasma concentrations of counter-regulatory hormones and changes in potassium (K, mEq/L) and glucose (mg/dL) levels. The latter are associated with outcomes and investigated as potential focus for intervention; glucose-insulin‑potassium (GIK) solution.

Objectives: To evaluate the associations of concomitant K and glucose (K/glucose) levels with in-hospital mortality in AMI patients.

Methods: AMI patients hospitalized in a tertiary Medical Center through 2002-2012 were studied. K/glucose levels were divided into equally sized categories. The intermediate category (glucose 124-143 mg/dL, K 4-4.9 mEq/L) was the reference group. The associations of these tests with the outcome were assessed using Generalized Estimating Equations model which included the interaction of K and glucose levels, adjusted for the patient's baseline characteristics and other laboratory results.

Results: 17,670 AMI admissions (mean age 67.8 ± 4.0 years, 66.6% males, mortality rate 7.7%) were included; 112,531 results of K/glucose tests were recorded. Univariate and multivariate analyses showed that K/glucose levels were significantly associated with in-hospital mortality, with highest risk being in patients with concomitant low K (<3.7 mEq/L) and high glucose (≥217 mg/dL), adjOR = 2.53. It seems that low-normal glucose levels attenuate the increased risk associated with low K.

Conclusions: The highest independent risk for mortality is found with low K and concomitant high glucose levels. Additional studies evaluating mechanisms and therapeutic interventions in K/glucose levels in this setting are warranted.