Send to

Choose Destination
Am Heart J. 2017 Jun;188:99-108. doi: 10.1016/j.ahj.2017.03.011. Epub 2017 Mar 23.

True rate of mineralocorticoid receptor antagonists-related hyperkalemia in placebo-controlled trials: A meta-analysis.

Author information

Universität des Saarlandes, Klinik für Innere Medizin III, Homburg/Saar, Germany. Electronic address:
Universität des Saarlandes, Klinik für Innere Medizin III, Homburg/Saar, Germany.
University of Michigan School of Medicine, University of Michigan, University of Michigan Health System, Ann Arbor, MI.
Universität des Saarlandes, Institut für Medizinische Biometrie, Epidemiologie und Medizinische Informatik, Homburg/Saar, Germany.



Mineralocorticoid receptor antagonists (MRA) improve survival in heart failure with reduced ejection fraction but are often underused, mostly due to concerns of hyperkalemia. Because hyperkalemia occurs also on placebo, we aimed to determine the truly MRA-related rate of hyperkalemia.


We performed a meta-analysis including randomized, placebo-controlled trials reporting hyperkalemia on MRAs in patients after myocardial infarction or with chronic heart failure. We evaluated the truly MRA-related rate of hyperkalemia that represents hyperkalemia on MRA, corrected for hyperkalemia on placebo (Pla), according to the equation: True MRA (%)=(MRA (%) - Pla (%))/MRA (%).


A total number of 16,065 patients from 7 trials were analyzed. Hyperkalemia was more frequently observed on MRA (9.3%) vs placebo (4.3%) (risk ratio 2.17, 95% CI 1.92-2.45, P<.0001). Truly MRA-related hyperkalemia was 54%, whereas 46% were non-MRA related. In trials using eplerenone, hyperkalemia was documented in 5.0% on eplerenone and in 2.6% on placebo (P<.0001). In spironolactone trials, hyperkalemia was documented in 17.5% and in 7.5% of patients on placebo (P=.0001). Hypokalemia occurred less frequently in patients on MRA (9.3%) compared with placebo (14.8%) (risk ratio 0.58, CI 0.47-0.72, P<.0001).


This meta-analysis shows that in clinical trials, 54% of hyperkalemia cases were specifically related to the MRA treatment and 46% to other reasons. Therefore, non-MRA-related rises in potassium levels might be underestimated and should be rigorously explored before cessation of the evidence-based therapy with MRAs.

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center