Effect of diabetes on serum potassium concentrations in acute coronary syndromes

Abstract

Objectives: To compare serum potassium concentrations in diabetic and non-diabetic patients in the early phase of acute coronary syndromes.

Background: Acute phase hypokalaemia occurs in response to adrenergic activation, which stimulates membrane bound sodium-potassium-ATPase and drives potassium into the cells. It is not known whether the hypokalaemia is attenuated in patients with diabetes because of the high prevalence of sympathetic nerve dysfunction.

Methods: Prospective cohort study of 2428 patients presenting with acute coronary syndromes. Patients were stratified by duration of chest pain, diabetic status, and pretreatment with beta blockers.

Results: The mean (SD) serum potassium concentration was significantly higher in diabetic than in non-diabetic patients (4.3 (0.5) v 4.1 (0.5) mmol/l, p < 0.0001). Multivariate analysis identified diabetes as an independent predictor of a serum potassium concentration in the upper half of the distribution (odds ratio 1.66, 95% confidence interval 1.38 to 2.00). In patients presenting within 6 hours of symptom onset, there was a progressive increase in plasma potassium concentrations from 4.08 (0.46) mmol/l in patients presenting within 2 hours, to 4.20 (0.47) mmol/l in patients presenting between 2-4 hours, to 4.24 (0.52) mmol/l in patients presenting between 4-6 hours (p = 0.0007). This pattern of increasing serum potassium concentration with duration of chest pain was attenuated in patients with diabetes, particularly those with unstable angina. Similar attenuation occurred in patients pretreated with beta blockers.

Conclusion: In acute coronary syndromes, patients with diabetes have significantly higher serum potassium concentrations and do not exhibit the early dip seen in non-diabetics. This may reflect sympathetic nerve dysfunction that commonly complicates diabetes.

Figure 1

Serum potassium concentrations by duration of chest pain. Data are mean (SEM). Upper panels show the effects of pretreatment with β blockers in 1118 patients for whom complete data were available. Lower panels show effects of diabetes in 1124 patients for whom complete data were available.

Figure 2

Serum glucose concentrations by duration of chest pain in 932 patients for whom complete data were available.