Hyper- and hypokalemia associated with increased mortality in chronic heart failure
Associations of serum potassium levels with mortality in chronic heart failure patientsAldahl M, Jensen ASC, Davidsen L, et al. - Eur Heart J 2017;38(38):2890–2896
A U-shaped relation between potassium and mortality has been reported in acute HF and CVD patients and even low and high potassium levels within the normal range were associated with increased mortality. However, it is not known whether this finding applies to patients with chronic HF, who have often an impaired potassium homeostasis, due to concomitant treatments with loop diuretics, MRAs, and ACE-Is or ARBs [1-4].
In this analysis of Danish registries, the association between different potassium levels and mortality was evaluated in chronic HF patients taking loop diuretics and ACE-Is or ARBs. All patients diagnosed with HF between 1994 and 2012, either during hospital admission or in an ambulatory outpatient setting, were included in the analysis, when they had a prescription for both loop diuretics and ACE-Is or ARBs, as well as at least one serum potassium measurement 90 – 455 days later. The time of the serum potassium measurement defined the baseline of the analysis. Outcome of the study was all-cause mortality within 90 days after the baseline serum potassium measurement.
The following serum potassium intervals were defined (in mmol/L): Interval 1 (hypokalaemia): 2.8-3.4; Interval 2: 3.5-3.8; Interval 3: 3.9-4.1; Interval 4 (reference interval): 4.2-4.4; Interval 5: 4.5-4.7; Interval 6: 4.8-5.0; Interval 7 (mild hyperkalaemia): 5.1-5.5; Interval 8 (severe hyperkalaemia): 5.6-7.4
The study population of 19 549 patients was of advanced age, consisted of more men than women and ~35% had a prior MI.
- Within 90 days from the qualifying serum potassium measurement, 7.1% patients died. The 90-day mortality in the eight potassium intervals from the lowest to the highest were 14.4%, 8.0%, 6.3%, 5.0%, 5.8%, 7.9%, 10.3%, and 21.1% respectively.
- There was a significant increased risk of death in hypokalaemia patients (P≤0.001), as well as in hyperkalaemia patients (P≤0.001 for mild hyperkalaemia and P≤0.001 for severe hyperkalaemia) compared with the reference interval.
- Potassium intervals within the normal range showed significant increased risk for patients between intervals 2 to 4 (P≤0.001 for interval 2; P=0.0076 for interval 3; P ≤0.001 for interval 4).
- In the adjusted multivariable analysis, mortality remained significantly increased in patients with hypokalaemia (HR: 3.16; P≤0.01), mild hyperkalaemia (HR: 1.60; P≤0.01), and severe hyperkalaemia (HR: 3.31; P≤0.001).
- The risk of all-cause mortality was also increased in the following potassium intervals within the normal range: HR for interval 2: 1.62; P≤0.01; HR for interval 3: 1.29; P≤0.01; and HR for interval 6: 1.34; P≤0.01.
- The covariates with a significant impact on mortality were age, gender, AMI, COPD, DM, a high serum creatinine level, and prescriptions for digoxin and potassium supplements.
- The U-shaped restricted cubic spline curve showed an increase in mortality for low and high potassium levels and within the normal range, the interval from 4.0 to 4.8 mmol/L is associated with the lowest risk of death.
In chronic HF patients taking loop diuretics and ACE-Is or ARBs, the short-term all-cause mortality was increased in those with hypokalaemia and mild or severe hyperkalemia, as well as in those in the lower and upper levels of the normal serum potassium range (3.5 – <4.1 mmol/L and 4.8–5.0 mmol/L, respectively) compared with the reference values of 4.2–4.4 mmol/L.
In their editorial article , Pitt and Rossignol acknowledge the U-shaped relationship between serum potassium and mortality, and note that clinicians should be concerned even if potassium values are in the upper or lower normal range, when it comes to chronic HF patients. They criticize the study published by Aldahl et al, because no information is given about reduced or preserved left ventricular EF, which might be important for the interpretation of the data. Furthermore, they raise the question and discuss several possibilities as to how these level ranges can be avoided, what should be done when one encounters these values, and how blunt the U-shaped relationship is. They advise to reconsider the current recommendations regarding the frequency for monitoring serum potassium in chronic HF patients, although the cost effectiveness and efficacy of monitoring strategy need further evaluation. And they conclude: ‘Thus, while we are indebted to the Danish National registry investigators for emphasizing the U-shaped relationship between serum K+ and death in patients with chronic HF, and especially the increased risk of death at levels of serum K+ <4.2 and >4.4mmol/L, levels currently considered by some clinicians as not warranting concern, the study, as with most good studies, raises many more questions. Once we cross the nadir of serum K+ in a patient with chronic HF and make a U-turn we will need further direction as to where we go from here.’