Introduction

Potassium plays an essential role in cellular function, particularly in neural and muscular tissue, including cardiomyocytes, where it is crucial for maintaining resting membrane potential and electrical stability.1 However, potassium homeostasis is often disrupted in patients with heart failure (HF) because of neurohormonal activation, medications (e.g. diuretics and mineralocorticoid receptor antagonists), and comorbidities (e.g. kidney dysfunction).1–5 Although there are many studies of the association between potassium concentrations and outcomes in HF, most have examined a single left ventricular ejection fraction (LVEF) phenotype or HF overall, without any stratification by LVEF.3,6 Second, most prior investigations have typically examined risk related to potassium using categories thought to reflect hypokalaemia, normokalaemia, and hyperkalaemia. However, it is not clear whether the potassium cut-points used to define these conventional categories are the thresholds that best reflect the association between potassium levels and outcomes in HF and recently interventions to increase potassium levels have been shown to improve cardiovascular outcomes.7 It is also not known whether the pattern of association between potassium levels and risk is the same in patients with HF and reduced ejection fraction (HFrEF, LVEF ≤40%) and those with HF and preserved ejection fraction (HFpEF, LVEF ≥50%).4,6,8–12 To address these critical knowledge gaps, we conducted a comprehensive analysis using individual participant-level data from 12 major randomized controlled trials, comprising 7 trials in HFrEF and 5 trials in HFpEF.13–24 By stratifying the population according to HF phenotype and examining a wide spectrum of baseline potassium levels, the aims of this study were to determine whether the relationship between serum potassium levels and clinical outcomes differs between HFrEF and HFpEF, and to identify the potassium range associated with the lowest risk in each subgroup.