Abstract
Background: Heart failure with preserved ejection fraction (HFpEF) is an increasing health problem associated with high morbidity and mortality rates. Several reports have shown an association between hypokalemia and clinical outcomes in patients with heart failure (HF). However, the association of hypokalemia with kidney function and clinical outcomes in patients with HFpEF remains unclear.
Methods and Results: We measured serum potassium levels and kidney function in 454 patients with HFpEF (mean age 76 years; 55% men) at admission. Hypokalemia (K+
<3.5 mmol/L) and hyperkalemia (K+
>5.0 mmol/L) were identified in 58 (12.7%) and 11 (2.4%) patients, respectively. Patients with hypokalemia showed renal tubular damage (RTD), defined as a urinary β2-microglobulin to creatinine ratio ≥300 μg/gCr, preserved estimated glomerular filtration rate (eGFR), and plasma volume expansion. Multivariate logistic analysis demonstrated that RTD, preserved eGFR, and plasma volume expansion were significantly associated with hypokalemia. During the median follow-up period of 1,000 days, 82 HF-related events occurred. Kaplan-Meier analysis showed that patients with hypokalemia had a higher rate of HF-related events than those without hypokalemia. Multivariate Cox proportional hazard regression analysis demonstrated that hypokalemia was significantly associated with HF-related events after adjusting for confounding factors.
Conclusions: Hypokalemia is affected by kidney function, notably RTD, in patients with HFpEF. Hypokalemia is a risk factor for HF-related events in patients with HFpEF.
Despite advances in heart failure (HF) treatment, HF remains a public health problem associated with high morbidity and mortality.1 Among the HF subtypes, HF with preserved ejection fraction (HFpEF) is gaining attention owing to its increasing prevalence and high morbidity and mortality.2 HFpEF is characterized by a high prevalence of comorbidities such as kidney dysfunction, anemia, and electrolyte disorders,3 indicating the involvement of non-cardiac factors in its pathophysiology.4
The 2021 European Society of Cardiology (ESC) guidelines for the diagnosis and treatment of acute and chronic HF noted that serum potassium concentrations have a U-shaped relationship with mortality, with the lowest risk of death within a relatively narrow range of 4–5 mmol/L.5 Dyskalemia is defined as hyperkalemia (serum potassium >5.0 mmol/L) and hypokalemia (serum potassium <3.5 mmol/L). Hypokalemia is often caused by loop diuretics and thiazide administration through renal loss and may occur in up to 50% of patients with HF.6 Extreme hypokalemia in patients with HF leads to lethal ventricular arrhythmias and high cardiovascular mortality.7
The kidney plays a pivotal role in maintaining serum potassium concentrations through reabsorption, mainly in the proximal tubule, and excretion, regulated by aldosterone in the distal tubule.8 Kidney dysfunction is a common comorbidity in patients with HFpEF. We have reported that renal tubular damage (RTD) is more common and a risk factor for poor clinical outcomes in patients with HFpEF.9 The predisposing factors for hypokalemia in patients with HFpEF remain unclear.
Plasma volume expansion underlies systemic congestion, which is a well-known clinically and prognostically relevant complication of HF.10 The 2021 ESC guidelines also recommend that blood volume status should be assessed in patients with HF.5 Due to the difficulty of actual plasma volume measurement, its clinical application has been limited. Conversely, estimated plasma volume status was reportedly easy to measure and useful to predict clinical outcomes in patients with HFpEF.11
The aims of this study were to: (1) identify predisposing factors for hypokalemia in patients with HFpEF; and (2) examine whether hypokalemia is associated with HF-related events in patients with HFpEF.
Methods
Ethics Statement
All procedures were performed per the ethical, institutional, and/or national research committee guidelines of the centers at which the studies were conducted, and all procedures complied with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. The study was approved by the Institutional Ethics Committee of Yamagata University School of Medicine; the first and most recently updated institutional review board numbers are H20-62 and 2020-344, respectively.
Study Subjects
This prospective observational study is part of an ongoing Yamagata cardiovascular and cerebrovascular prognosis survey and aimed to elucidate the association between serum potassium concentrations and the clinical outcomes of patients with HFpEF. After written informed consent was obtained from patients, blood samples were collected within 24 h after admission. From the day after informed consent was obtained, all patients were prospectively followed twice a year using telephone or medical records for a median period of 1,000 days.
We enrolled 454 patients with HFpEF who were admitted to our hospital for the diagnosis or treatment of acute or chronic HF exacerbation between 2009 and 2019. HF was diagnosed according to the standard Framingham criteria, including a history of dyspnea and symptomatic exercise intolerance, accompanied by signs of pulmonary congestion or peripheral edema and radiological or echocardiographic evidence of left ventricular enlargement or dysfunction.12 Transthoracic echocardiography was performed by physicians and sonographers who were blinded to the biochemical data. Left ventricular ejection fraction (LVEF), left ventricular (LV) end-diastolic diameter, and the ratio of the mitral inflow E wave to the tissue Doppler e′ wave (E/e′ ratio) were measured. The diagnosis of HFpEF was made using the following criteria: (1) evidence of HF; (2) normal or mildly abnormal LV systolic function (LVEF ≥50%); and (3) evidence of abnormal LV relaxation, filling, diastolic distensibility, or diastolic stiffness (LV mass index >95 g/m2
in women and 115 g/m2
in men; E/e′ at rest > 9; B-type natriuretic peptide [BNP] at rest >35 pg/mL in sinus rhythm and 105 pg/mL in atrial fibrillation; pulmonary artery systolic pressure >35 mmHg).13,14 Diagnoses of hypertension, diabetes, and dyslipidemia were established based on the medical records or history of medical therapy. The exclusion criteria were acute coronary syndrome within 3 months preceding admission, active hepatic disease, and malignant disease. Demographic and clinical data, including age, sex, New York Heart Association (NYHA) functional class, and medications at admission and discharge, were collected from the medical records and interviews with the patients.
Kidney Function and Anemia
Urine and venous blood samples were obtained in the early morning within 24 h after admission. Urine β2-microglobulin concentrations were determined using the latex agglutination method (BML Inc., Tokyo, Japan). Urine β2-microglobulin concentrations were corrected for urine creatinine levels (UBCR). RTD was defined as UBCR ≥300 μg/gCr, as reported previously.15,16 The urine albumin concentration was measured quantitatively by immunoturbidimetry in a single spot urine specimen collected early in the morning. Urine albumin concentrations were corrected for urine creatinine (Cr) in a single manner, according to the urine microalbumin-Cr ratio (UACR). Glomerular filtration rate (GFR) was estimated using the diet modification in renal disease equation with the Japanese coefficient, as reported previously.17 According to Kidney Disease Outcomes Quality Initiative (KDOQI) clinical guidelines, reduced estimated GFR (eGFR) and albuminuria were defined as eGFR <60 mL/min/1.73 m2
and UACR ≥30 mg/gCr, respectively.18 Hemoglobin (Hb) concentrations were measured simultaneously. Anemia was defined as Hb <13 g/dL in men and <12 g/dL in women according to the World Health Organization guidelines.19
Definition of Plasma Volume Status
Actual plasma volume (aPV), ideal plasma volume (iPV), and plasma volume status (PVS) were calculated using the following equations:20–22
aPV = (1 − hematocrit) × (a + [b × weight])
where weight is in kilograms, hematocrit is a fraction, a=1,530 and b=41 in men, and a=864 and b=47.9 in women.
iPV = c × weight
where weight is in kilograms, c=39 in men, and c=40 in women.
PVS (%) = [(aPV − iPV) / iPV] × 100
The normal range of PVS remains to be determined. Because aPV is often below iPV, the value of PVS could become less than 0. In the Treatment of Preserved Cardiac Function Heart Failure with an Aldosterone Antagonist Trial (TOPCAT) study, PVS was reported to be less than 0 in 91% of patients with HF.11 Plasma volume expansion was defined as PVS >0.
Cardiovascular Risk Factors
Hypertension was defined as systolic blood pressure ≥140 mmHg, diastolic blood pressure ≥90 mmHg, or the use of antihypertensive medication. Diabetes was defined as a fasting blood sugar level ≥126 mg/dL, HbA1c ≥6.5% (NGSP), or the use of antidiabetic medication. Dyslipidemia was defined as high-density lipoprotein cholesterol <40 mg/dL, low-density lipoprotein cholesterol ≥140 mg/dL, triglycerides ≥150 mg/dL, or the use of lipid-lowering medication.
Endpoints and Follow-up Period
All patients were prospectively followed up by telephone or by reviewing medical records twice a year for a median period of 1,000 days (interquartile range [IQR] 378–1,000 days). The endpoint was defined as an HF-related event. HF-related events were defined as hospitalizations due to the worsening of HF and HF-related death.
Statistical Analysis
The normality of the distribution of continuous variables was confirmed using the Shapiro-Wilk test. Results are expressed as the mean±SD for continuous variables and as numbers and percentages for categorical variables. Skewed values are presented as the median with IQR. Continuous and variables were compared using t-tests, whereas categorical variables were compared using Chi-squared tests. Because of the skewed distribution of B-type natriuretic peptide (BNP) values, log10-transformed BNP values were used for statistical analysis.
The significance of differences in age, eGFR, sodium, chloride, sodium/chloride ratio, phosphate, log10-transformed BNP, LVEF, LV end-diastolic dimension, LV mass index, and the E/e′ ratio among groups was evaluated using analysis of variance with Tukey’s post hoc test. The significance of differences in UACR, UBCR, and PVS levels among groups was determined using the Kruskal-Wallis test.
Receiver operating characteristic (ROC) curves for hypokalemia were plotted and used to compare the predictive accuracy of UBCR for the presence of hypokalemia. ROC curves for HF-related events were plotted and used to measure the predictive accuracy of hypokalemia for HF-related events. Univariate logistic analysis was performed to identify the predisposing risk for hypokalemia, and selected risk factors for hypokalemia were entered into multivariate logistic analysis. There was a close association between PVS and anemia. Thus, we did not include anemia in the multivariate logistic analysis. Survival curves were plotted using the Kaplan-Meier method and compared using the log-rank test. Cox proportional hazard analysis was used to determine the independent predictors of HF-related events. Multicollinearity was assessed using the variance inflation factor. There were the close associations between hypokalemia and RTD, as well as PVS and anemia. Thus, we did not include anemia and RTD in the multivariate Cox proportional hazards regression analysis. Age, sex, log10-transformed BNP concentration, and significant predictors selected in the univariate Cox proportional hazard regression analysis were entered into the multivariate analysis for HF-related events. We calculated the net reclassification index (NRI) and integrated discrimination improvement (IDI) to measure the incremental value of hypokalemia in classifying patients at high or low risk of HF-related events. An analysis using restricted cubic spline was performed to identify any possible non-linear associations between serum potassium and HF-related events.
Statistical significance was set at P<0.05. All statistical analyses were performed using the standard software packages JMP®
(version 14.2; SAS Institute, Cary NC, USA), EZR (Saitama Medical Center, Jichi Medical University, Shimotsuke, Japan), and Stata software package (version 14.1; StataCorp, College Stata, TX, USA).
Results
Baseline Characteristics of All Patients and Clinical Characteristics in the Hypokalemia, Mild Hypokalemia, and Control Groups
The baseline characteristics of the 454 patients are summarized in Table 1. There were 210 patients with NYHA Class II, 143 with NYHA Class III, and 101 with NYHA Class IV. Hypertension, diabetes, and dyslipidemia were identified in 362 (79.7%), 161 (35.5%), and 272 (59.9%) patients, respectively. Previous ischemic heart disease and atrial fibrillation were identified in 81 (17.8%) and 199 (43.8%) patients, respectively. There were 220 (48%) patients with reduced eGFR, 249 (55%) with albuminuria, 217 (48%) with RTD, and 289 (64%) with plasma volume expansion.
Table 1.
Baseline Characteristics in All Patients and in the Hypokalemia, Mild Hypokalemia, and Control Groups of Patients With HFpEF
| Variables |
All patients
(n=454) |
Patients with HFpEF |
P value |
Hypokalemia
(K <3.5 mmol/L;
n=58) |
Mild hypokalemia
(K 3.5–3.9 mmol/L;
n=137) |
Control
(K ≥4.0 mmol/L;
n=259) |
| Age (years) |
73.6±12.7 |
76.6±11.6 |
72.7±12.5 |
73.4±13.0 |
0.1443 |
| Men/women (n) |
246/208 |
31/27 |
75/62 |
140/119 |
0.9843 |
| NYHA Class II/III/IV (n) |
210/143/101 |
20/24/14 |
62/47/28 |
128/72/59 |
0.2024 |
| Hypertension |
362 (79.7) |
50 (86.2) |
112 (81.8) |
200 (77.2) |
0.2246 |
| Diabetes |
161 (35.5) |
20 (34.5) |
45 (32.9) |
96 (37.1) |
0.6950 |
| Dyslipidemia |
272 (59.9) |
34 (58.6) |
82 (59.9) |
156 (60.2) |
0.9747 |
| Ischemic heart disease |
81 (17.8) |
9 (15.5) |
25 (18.3) |
47 (18.2) |
0.8810 |
| Atrial fibrillation |
199 (43.8) |
32 (55.2) |
60 (43.8) |
107 (41.3) |
0.1596 |
| Biochemical data |
| eGFR (mL/min/1.73 m2) |
63±27 |
72±30* |
67±28* |
57±26 |
0.0004 |
| Reduced/preserved eGFR (n) |
220/234 |
20/38 |
59/78 |
141/118 |
0.0070 |
| UACR (mg/gCr) |
37 [13–128] |
58 [30–152]*,† |
31 [10–119] |
35 [12–121] |
0.0346 |
| Albuminuria |
249 (55) |
42 (72) |
70 (51) |
137 (53) |
0.0126 |
| UBCR (μg/gCr) |
270 [85–1,746] |
1,748 [306–7,034]*,† |
270 [87–1,251] |
194 [60–1,184] |
<0.0001 |
| RTD |
217 (48) |
45 (78) |
66 (48) |
106 (41) |
<0.0001 |
| PVS |
5.2 [−4.1, 16.3] |
10.4 [−6.5, 21.5] |
2.5 [−4.8, 15.4] |
5.1 [−4.9, 15.8] |
0.0029 |
| Plasma volume expansion |
289 (64) |
46 (79) |
80 (58) |
163 (63) |
0.0152 |
| Anemia |
255 (56) |
42 (72) |
66 (48) |
147 (57) |
0.0065 |
| Sodium (mmol/L) |
141±3.0 |
142±3.4* |
141±2.5* |
140±3.2 |
<0.0001 |
| Chloride (mmol/L) |
104±3.9 |
103±5.3 |
104±3.1 |
104±3.8 |
0.2703 |
| Sodium/chloride ratio |
1.35±0.04 |
1.38±0.06*,† |
1.36±0.03 |
1.35±0.04 |
<0.0001 |
| Potassium (mmol/L) |
4.0±0.3 |
3.1±0.3*,† |
3.7±0.2* |
4.3±0.3 |
<0.0001 |
| Phosphate (mmol/L) |
3.4±0.7 |
3.2±0.8* |
3.2±0.6* |
3.5±0.7 |
<0.0001 |
| Log10 [BNP] (pg/mL) |
2.41±0.49 |
2.51±0.39*,† |
2.37±0.53* |
2.42±0.49 |
0.1718 |
| Echocardiogram |
| LVEF (%) |
64±9 |
63±9 |
64±9 |
65±9 |
0.4027 |
| LVEDD (mm) |
49±8 |
47±9 |
49±8 |
49±8 |
0.0926 |
| LVMI (g/m2) |
148±51 |
135±37 |
144±51 |
152±53 |
0.0555 |
| E/e′ ratio |
14.8±8.0 |
15.1±7.2 |
14.5±7.4 |
14.8±8.4 |
0.9227 |
| Medication at admission |
| ACEIs or ARBs |
265 (58) |
38 (66) |
81 (59) |
146 (56) |
0.4273 |
| β-blockers |
199 (44) |
27 (47) |
60 (44) |
112 (43) |
0.9003 |
| MRA |
83 (18) |
15 (26) |
21 (15) |
47 (18) |
0.2377 |
| Diuretics |
129 (28) |
22 (38) |
38 (28) |
69 (27) |
0.2373 |
| Furosemide |
82 (18.1) |
13 (22.4) |
19 (13.9) |
50 (19.3) |
0.2584 |
| Median dose (mg) |
20 |
20 |
20 |
20 |
|
| Azosemide |
40 (8.8) |
9 (15.5) |
17 (12.4) |
14 (5.4) |
0.0113 |
| Median dose (mg) |
30 |
30 |
30 |
30 |
|
| Torsemide |
11 (2.4) |
1 (1.7) |
2 (1.5) |
8 (3.1) |
0.5472 |
| Median dose (mg) |
8 |
8 |
6 |
6 |
|
| Tolvaptan |
9 (2.0) |
2 (3.5) |
1 (0.7) |
6 (2.3) |
0.3459 |
| Median dose (mg) |
7.5 |
7.5 |
3.75 |
7.5 |
|
| Medication at discharge |
| ACEIs or ARBs |
278 (61.2) |
38 (65.5) |
86 (62.8) |
154 (59.5) |
0.6268 |
| β-blockers |
290 (63.9) |
38 (65.5) |
88 (64.2) |
164 (63.3) |
0.9463 |
| MRA |
130 (28.6) |
28 (48.3) |
38 (27.7) |
64 (24.7) |
0.0025 |
| Diuretics |
266 (58.6) |
38 (65.5) |
76 (55.6) |
152 (58.6) |
0.4240 |
| Furosemide |
145 (31.9) |
16 (27.6) |
46 (33.6) |
83 (32.1) |
0.7090 |
| Median dose (mg) |
20 |
30 |
20 |
20 |
|
| Azosemide |
101 (22.2) |
18 (31.0) |
29 (21.2) |
54 (20.9) |
0.2491 |
| Median dose (mg) |
30 |
30 |
30 |
30 |
|
| Torsemide |
22 (4.9) |
5 (8.6) |
3 (2.2) |
14 (5.4) |
0.1193 |
| Median dose (mg) |
8 |
8 |
4 |
8 |
|
| Tolvaptan |
16 (3.5) |
4 (6.9) |
3 (2.2) |
9 (3.5) |
0.3078 |
| Median dose (mg) |
7.5 |
7.5 |
7.5 |
7.5 |
|
Unless indicated otherwise, data are expressed as the mean±SD, n (%), or median [interquartile range]. *P<0.05 compared with control; †P<0.05 compared with mild hypokalemia. ACEIs, angiotensin-converting enzyme inhibitors; ARBs, angiotensin II receptor blockers; BNP, B-type natriuretic peptide; E/e′ ratio, ratio of the mitral inflow E wave to the tissue Doppler e′ wave; eGFR, estimated glomerular filtration rate; HFpEF, heart failure with preserved ejection fraction; LVEDD, left ventricular end-diastolic dimension; LVEF, left ventricular ejection fraction; LVMI, left ventricular mass index; MRA, mineralocorticoid receptor antagonists; NYHA, New York Heart Association; PVS, plasma volume status; RTD, renal tubular damage; UACR, urinary albumin-creatinine ratio; UBCR, urinary β2-microglobulin-creatinine ratio.
As shown in Figure 1, patients were divided into 3 groups according to serum potassium concentrations: a hypokalemia group (K <3.5 mmol/L; n=58), a mild hypokalemia group (K 3.5–3.9 mmol/L; n=137), and a control group (K ≥4.0 mmol/L; n=259). Patients in the hypokalemia group had higher levels of eGFR, UACR, UBCR, PVS, sodium, sodium/chloride ratio, and log10-transformed BNP than patients in the other 2 groups. The prevalence of preserved eGFR, albuminuria, RTD, plasma volume expansion, and anemia was higher in patients with than without hypokalemia. Patients in the hypokalemia group were administered more mineralocorticoid receptor antagonists (MRAs) than those in the other 2 groups at discharge. Patients in the mild hypokalemia group had higher eGFR, sodium, and BNP levels and lower phosphate levels than those in the control group (Table 1).
Association of Hypokalemia With Kidney Function
Serum potassium concentrations decreased in the setting of preserved eGFR and RTD (Supplementary Figure 1). There was no difference in serum potassium concentrations according to the presence or absence of albuminuria. Patients were divided into 4 groups based on the presence or absence of preserved eGFR and RTD. The prevalence of hypokalemia was highest in patients with preserved eGFR and RTD (Figure 2A). To assess the predictive capacity for hypokalemia, we compared C-indices for hypokalemia. The C-index of UBCR for hypokalemia was the highest among the parameters for kidney function (Figure 2B–E). Of note, the abnormal cut-off point of UBCR for hypokalemia was 296 μg/gCr, which was almost equal to the definition of RTD.
Risk Factors for Hypokalemia in Patients With HFpEF
To determine the predisposing risk factors for hypokalemia, we performed univariate and multivariate logistic analyses. Univariate logistic analysis showed that RTD was significantly related to the presence of hypokalemia. In addition, age, preserved eGFR, albuminuria, plasma volume expansion, anemia, and prescription of MRAs were related to the presence of hypokalemia. The multivariate logistic analysis demonstrated that RTD was independently related to the presence of hypokalemia (odds ratio 3.84; 95% confidence interval 1.92–8.10; P<0.0001; Table 2).
Table 2.
Univariate and Multivariate Logistic Analyses for Hypokalemia in Patients With HFpEF
| Variables |
Univariate analysis |
Multivariate analysis |
| OR |
95% CI |
P value |
OR |
95% CI |
P value |
| Age |
1.02 |
1.00–1.05 |
0.0474 |
1.00 |
0.98–1.03 |
0.7267 |
| Sex (men vs. women) |
0.97 |
0.56–1.69 |
0.9041 |
|
|
|
| NYHA Class III/IV vs. II |
1.75 |
0.99–3.17 |
0.0519 |
|
|
|
| Hypertension |
1.68 |
0.81–3.96 |
0.1719 |
|
|
|
| Diabetes |
0.96 |
0.52–1.68 |
0.8671 |
|
|
|
| Dyslipidemia |
0.94 |
0.54–1.66 |
0.8301 |
|
|
|
| Ischemic heart disease |
0.83 |
0.37–1.68 |
0.6152 |
|
|
|
| Atrial fibrillation |
1.68 |
0.97–2.95 |
0.6555 |
|
|
|
| Preserved eGFR |
1.92 |
1.10–3.45 |
0.0215 |
2.88 |
1.53–5.57 |
0.0009 |
| Albuminuria |
2.39 |
1.32–4.52 |
0.0033 |
1.49 |
0.76–3.02 |
0.2484 |
| RTD |
4.51 |
2.42–8.94 |
<0.0001 |
3.84 |
1.92–8.10 |
<0.0001 |
| PV expansion |
2.41 |
1.28–4.90 |
0.0058 |
2.38 |
1.16–5.24 |
0.0181 |
| Anemia |
2.25 |
1.25–4.26 |
0.0006 |
|
|
|
| Log10 [BNP]A |
1.30 |
0.96–1.81 |
0.0946 |
|
|
|
| Diuretics |
1.40 |
0.79–2.53 |
0.2474 |
|
|
|
| MRA |
2.69 |
1.53–4.73 |
0.0007 |
2.21 |
1.21–4.06 |
0.0105 |
APer 1-SD increase. CI, confidence interval; OR, odds ratio; PV, plasma volume. Other abbreviations as in Table 1.
Hypokalemia and Clinical Outcomes in Patients With HFpEF
During the follow-up period, 82 HF-related events occurred. Kaplan-Meier analysis revealed that the hypokalemia group had the greatest risk for HF-related events among the 3 groups (Figure 3).
Univariate Cox proportional hazard regression analysis showed that hypokalemia was significantly associated with HF-related events. In addition, age, NYHA functional class, reduced eGFR, albuminuria, RTD, plasma volume expansion, and anemia were associated with HF-related events. Multivariate Cox proportional hazard regression analysis demonstrated that hypokalemia was an independent predictor of HF-related events after adjusting for age, sex, NYHA functional class, reduced eGFR, albuminuria, plasma volume expansion, and log10-transformed BNP concentration (hazard ratio 4.07; 95% confidence interval 2.31–7.05; P<0.0001; Table 3).
Table 3.
Univariate and Multivariate Cox Proportional Hazard Regression Analyses for Heart Failure-Related Events in Patients With HFpEF
| Variables |
Univariate analysis |
Multivariate analysis |
| HR |
95% CI |
P value |
HR |
95% CI |
P value |
| Age |
1.04 |
1.02–1.06 |
0.0006 |
1.02 |
0.99–1.04 |
0.0839 |
| Sex (men vs. women) |
1.25 |
0.81–1.94 |
0.3112 |
1.52 |
0.98–2.41 |
0.0642 |
| NYHA Class III/IV vs. II |
2.04 |
1.33–3.36 |
0.0013 |
1.91 |
1.20–3.12 |
0.0060 |
| Hypertension |
1.24 |
0.72–2.29 |
0.4593 |
|
|
|
| Diabetes |
0.86 |
0.54–1.34 |
0.5107 |
|
|
|
| Dyslipidemia |
1.25 |
0.80–1.98 |
0.3243 |
|
|
|
| Ischemic heart disease |
0.99 |
0.56–1.67 |
0.9854 |
|
|
|
| Atrial fibrillation |
1.10 |
0.72–1.69 |
0.6555 |
|
|
|
| Reduced eGFR |
2.78 |
1.76–4.51 |
<0.0001 |
3.19 |
1.92–5.44 |
<0.0001 |
| Albuminuria |
1.66 |
1.07–2.66 |
0.0238 |
1.04 |
0.64–1.76 |
0.8652 |
| RTD |
2.60 |
1.65–4.22 |
<0.0001 |
|
|
|
| Plasma volume expansion |
3.69 |
2.08–7.15 |
<0.0001 |
2.45 |
1.31–5.03 |
0.0039 |
| Anemia |
3.06 |
1.86–5.30 |
<0.0001 |
|
|
|
| Log10 [BNP] A |
1.27 |
0.99–1.64 |
0.0596 |
0.82 |
0.61–1.11 |
0.1952 |
| Hypokalemia vs. control |
3.29 |
1.94–5.49 |
<0.0001 |
4.07 |
2.31–7.05 |
<0.0001 |
| Mild hypokalemia vs. control |
1.20 |
0.71–1.99 |
0.4931 |
1.26 |
0.73–2.12 |
0.3984 |
| Diuretics |
1.33 |
0.85–2.10 |
0.2165 |
|
|
|
| MRA |
0.99 |
0.61–1.58 |
0.9870 |
|
|
|
APer 1-SD increase. HR, hazard ratio. Other abbreviations as in Tables 1,2.
Improvement of Reclassification by Adding Hypokalemia to Predict HF-Related Events
To examine whether the model fit and discrimination improved when hypokalemia was added to the baseline model, we evaluated improvements in the C-index, NRI, and IDI. The baseline model included age, sex, NYHA functional class, reduced eGFR, albuminuria, plasma volume expansion, and log10-transformed BNP. The addition of hypokalemia to the baseline model improved the C-index for HF-related events, with significant NRI and IDI (Figure 4). Finally, we created restricted cubic spline plots for HF-related events. Unexpectedly, there was no U-shaped relationship between the serum potassium concentration and HF-related events. The restricted cubic spline plots demonstrated that the adjusted hazard ratio increased with decreasing serum potassium concentrations when serum potassium concentrations were <4.0 mmol/L (Figure 5).
Discussion
The novel findings of the present study are as follows: (1) among patients with HFpEF, those with hypokalemia showed a higher prevalence of preserved eGFR, albuminuria, RTD, and plasma volume expansion than those with normal K+
concentrations; (2) multivariate logistic analysis demonstrated that RTD was a risk factor for hypokalemia; (3) Kaplan-Meier analysis demonstrated that hypokalemia imposed the greatest risk for HF-related events; (4) multivariate Cox proportional hazard regression analysis demonstrated that hypokalemia was an independent predictor for HF-related events; and (5) hypokalemia improved the predictive ability for HF-related events.
Kidney Function and Hypokalemia
The prevalence of dyskalemia reportedly differs by HF subtype and phase.23,24 Among HF subtypes, several reports have indicated that the prevalence of hypokalemia is high among patients with HFpEF. The incidence and prevalence of hypokalemia ranged, approximately, from 5% to 20% in patients with HFpEF.23,25,26 A previous study reported that the prevalence of hypokalemia at admission and discharge was 13.2% and 4.3%, respectively, in patients with acute HF, suggesting the correction of hypokalemia during hospitalization.24 Therefore, we focused on serum potassium concentrations at admission in patients with HFpEF to examine the predisposing factors for hypokalemia. Following previous reports, the prevalence of hypokalemia was 12.7% in the present study.
Aldosterone activity, a regulator of potassium excretion, is reportedly augmented in patients with HF independent of treatment with renin-angiotensin-aldosterone inhibitors.27 Although we did not examine serum aldosterone concentrations, patients in the hypokalemia group had higher sodium and lower potassium concentrations than those without hypokalemia, suggesting mineralocorticoid receptor activation.
Only a few reports have indicated that anemia and kidney function are related to hypokalemia.25,26 The common comorbidities of HFpEF include kidney dysfunction and anemia, which exacerbate clinical outcomes in patients with HFpEF.28 Plasma volume expansion has been suggested as the primary pathophysiologic mechanism in patients with HFpEF.29 Thus, we calculated the PVS from the hematocrit as an indicator of plasma volume and demonstrated an association between hypokalemia and plasma volume expansion. We hypothesized that plasma volume expansion, rather than anemia per se, contributed to the development of hypokalemia.
The association between hypokalemia and eGFR remains controversial. Some reports have indicated that reduced eGFR is associated with a high incidence of hypokalemia,24,26 whereas others have indicated that preserved eGFR is related to hypokalemia at discharge in hospitalized patients with HFpEF.23,30 Notably, we demonstrated, for the first time, the close relationship between hypokalemia and UBCR, a marker for renal tubular reabsorption.31,32 Up to 78% of patients with hypokalemia showed RTD. The significant relationship between hypokalemia and RTD was maintained after multivariate analysis, including the known risk for hypokalemia. These results raise the possibility that impaired reabsorption in the renal tubule is the underlying mechanism of hypokalemia. To support this, we compared serum phosphate concentrations between patients with and without RTD and found that serum phosphate concentrations were lower in patients with than without RTD (Supplementary Figure 2), indicating the disturbance of reabsorption. The prevalence of hypokalemia was highest in patients with RTD and preserved eGFR. This result is consistent with previous reports in hospitalized patients with HFpEF.23,30 RTD is a final common pathway to end-stage renal disease, leading to reduced eGFR.33 Thus, RTD may also explain the relationship between reduced eGFR and the incidence of hypokalemia as described in the SwedeHF and TOPCAT trials.24,25 Considering these findings, hypokalemia in patients with HFpEF may result from several factors, such as mineralocorticoid receptor activity, plasma volume expansion, and RTD.
Hypokalemia and HF-Related Events
The impact of hypokalemia on clinical outcomes in patients with HFpEF is debatable. Hypokalemia at discharge in patients with acute HFpEF was associated with neither all-cause mortality nor composite events of mortality and HF-related events.23 The SwedeHF registry demonstrated that new-onset hypokalemia was not associated with HF hospitalization in patients with HF whose baseline serum potassium concentrations were 3.5–5.0 mmol/L, although 20.3% of patients experienced at least one episode of hypokalemia.25 Concurrently, Nishihara et al have reported that the lowest K+
quartile (K+
<4.1 mmol/L) at discharge was associated with cardiovascular events in patients with HFpEF.30 The TOPCAT trial showed that recurrent and/or new-onset hypokalemia was associated with cardiovascular and all-cause mortality.24 The PARAGON-HF trial showed an association between hypokalemia (K+
<4.0 mmol/L) and HF rehospitalization in outpatients with HFpEF.26
We reconfirmed the usefulness of low serum potassium concentrations (K+
<3.5 mmol/L) at admission in predicting HF-related events in patients with HFpEF. Because a significant association between hypokalemia and clinical outcomes was maintained in the multivariate analysis, it is plausible that hypokalemia could independently predict clinical outcomes. The reason why extreme hypokalemia is associated with HF-related events remains unclear. One possible explanation is that serum potassium concentrations may be a marker of HF severity; another possible explanation is an effect of associated comorbidities. Thus, RTD potentially affected the clinical outcomes of patients with HFpEF with hypokalemia in the present study. Because RTD is induced by renal parenchymal hypoxia,34,35 HF-induced RTD may exacerbate hypokalemia at admission in patients with HFpEF. We have reported that RTD plays a multifunctional role in the development of anemia, hypoalbuminuria, and malnutrition, all of which worsened clinical outcomes in patients with HF.36–38 Because the present study was a prospective observational study, we could not determine the mechanism by which hypokalemia exacerbates clinical outcomes.
Importantly, we showed that the prediction models for HF-related events improved significantly after including hypokalemia, as evidenced by an improvement in the C-index, NRI, and IDI. This result confirms the clinical importance of hypokalemia in predicting clinical outcomes and raises the possibility that hypokalemia could be an additional clinical factor in patients with HFpEF.
According to the American College of Cardiology (ACC) expert consensus decision pathway on the management of HFpEF, the prescription of sodium–glucose cotransporter 2 (SGLT2) inhibitors and MRAs are recommended in HFpEF patients with fluid retention.39 It was reported that plasma volume was corrected by SGLT2 inhibitors in patients with diabetes.40 The SGLT2 inhibitor canagliflozin was reported to slow the progression of kidney disease and reduce UBCR in Japanese type 2 diabetes patients with chronic kidney disease.41 In addition, the Eplerenone Post-Acute Myocardial Infarction Heart Failure Efficacy and Survival Study (EPHESUS) data in HF patients after myocardial infarction showed potassium-sparing diuretics and MRAs decreased estimated PVS, leading to improved clinical outcomes.42 Our results from restricted cubic spline analysis suggested that the correction of serum potassium to >4.0 mmol/L may be beneficial to prevent HF-related events. Therefore, it is plausible that guideline-directed medical therapies such as SGLT2 inhibitors and MRAs could prevent hypokalemia and potential subsequent HF-related events through the correction of plasma volume and RTD in patients with HFpEF.
Study Limitations
This study has some limitations. First, we measured serum potassium concentrations at one point. Second, the rate of HF medication prescriptions was relatively low in the present study because of HFpEF and chronic kidney disease. Third, because the study population was enrolled before the angiotensin receptor-neprilysin inhibitor and SGLT2 inhibitors were clinically applied for HF in Japan, we could not examine the impact of these medicines on serum potassium concentrations and clinical outcomes. Finally, because there were few arrhythmia-related events in this study, we could not determine the impact of dyskalemia on arrhythmia-related events.
Conclusions
Our study found a close association between RTD and hypokalemia, indicating its contribution to the pathophysiology of hypokalemia at admission in patients with HFpEF. Hypokalemia improved the prediction capacity of HF-related events. Therefore, hypokalemia could be a feasible marker of adverse clinical outcomes in patients with HFpEF.
Sources of Funding
This study did not receive any specific funding.
Disclosures
The authors declare that there are no conflicts of interest.
IRB Information
This study was approved by the Ethical Review Committee of Yamagata University Faculty of Medicine (No. H20-62 and 2020-344).
Supplementary Files
Please find supplementary file(s);
https://doi.org/10.1253/circj.CJ-23-0562
References
- 1.
Bui AL, Horwich TB, Fonarow GC. Epidemiology and risk profile of heart failure. Nat Rev Cardiol 2011; 8: 30–41.
- 2.
Dunlay SM, Roger VL, Redfield MM. Epidemiology of heart failure with preserved ejection fraction. Nat Rev Cardiol 2017; 14: 591–602.
- 3.
Streng KW, Nauta JF, Hillege HL, Anker SD, Cleland JG, Dickstein K, et al. Non-cardiac comorbidities in heart failure with reduced, mid-range and preserved ejection fraction. Int J Cardiol 2018; 271: 132–139.
- 4.
Lam CS, Lyass A, Kraigher-Krainer E, Massaro JM, Lee DS, Ho JE, et al. Cardiac dysfunction and noncardiac dysfunction as precursors of heart failure with reduced and preserved ejection fraction in the community. Circulation 2011; 124: 24–30.
- 5.
McDonagh TA, Metra M, Adamo M, Gardner RS, Baumbach A, Bohm M, et al. 2021 ESC guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J 2021; 42: 3599–3726.
- 6.
Ferreira JP, Butler J, Rossignol P, Pitt B, Anker SD, Kosiborod M, et al. Abnormalities of potassium in heart failure: JACC state-of-the-art review. J Am Coll Cardiol 2020; 75: 2836–2850.
- 7.
Skogestad J, Aronsen JM. Hypokalemia-induced arrhythmias and heart failure: New insights and implications for therapy. Front Physiol 2018; 9: 1500.
- 8.
Nakhoul N, Batuman V. Role of proximal tubules in the pathogenesis of kidney disease. Contrib Nephrol 2011; 169: 37–50.
- 9.
Otaki Y, Watanabe T, Shimizu M, Tachibana S, Sato J, Kobayashi Y, et al. Renal tubular damage and clinical outcome in heart failure with preserved ejection fraction and chronic kidney disease. ESC Heart Fail 2023; 10: 2458–2468.
- 10.
Miller WL. Fluid volume overload and congestion in heart failure: Time to reconsider pathophysiology and how volume is assessed. Circ Heart Fail 2016; 9: e002922.
- 11.
Grodin JL, Philips S, Mullens W, Nijst P, Martens P, Fang JC, et al. Prognostic implications of plasma volume status estimates in heart failure with preserved ejection fraction: Insights from TOPCAT. Eur J Heart Fail 2019; 21: 634–642.
- 12.
McKee PA, Castelli WP, McNamara PM, Kannel WB. The natural history of congestive heart failure: The Framingham study. N Engl J Med 1971; 285: 1441–1446.
- 13.
Vasan RS, Levy D. Defining diastolic heart failure: A call for standardized diagnostic criteria. Circulation 2000; 101: 2118–2121.
- 14.
Authors/Task Force Members; McDonagh TA, Metra M, Adamo M, Gardner RS, Baumbach A, Böhm M, et al. 2021 ESC guidelines for the diagnosis and treatment of acute and chronic heart failure: Developed by the Task Force for the Diagnosis and Treatment of Acute and Chronic Heart Failure of the European Society of Cardiology (ESC). With the special contribution of the Heart Failure Association (HFA) of the ESC. Eur J Heart Fail 2022; 24: 4–131.
- 15.
Otaki Y, Watanabe T, Shishido T, Takahashi H, Funayama A, Narumi T, et al. The impact of renal tubular damage, as assessed by urinary beta2-microglobulin-creatinine ratio, on cardiac prognosis in patients with chronic heart failure. Circ Heart Fail 2013; 6: 662–668.
- 16.
Dotare T, Ishiwata S, Matsue Y, Nakamura Y, Sunayama T, Maeda D, et al. Prevalence and prognostic relevance of isolated tubular dysfunction in patients with acute heart failure. Circ J 2022; 86: 709–714.
- 17.
Matsuo S, Imai E, Horio M, Yasuda Y, Tomita K, Nitta K, et al. Revised equations for estimated GFR from serum creatinine in Japan. Am J Kidney Dis 2009; 53: 982–992.
- 18.
Kidney Disease: Improving Global Outcomes Diabetes Work Group. KDIGO 2020 clinical practice guideline for diabetes management in chronic kidney disease. Kidney Int 2020; 98(4S): S1–S115.
- 19.
Nutritional anaemias. Report of a WHO scientific group. World Health Organ Tech Rep Ser 1968; 405: 5–37.
- 20.
Ling HZ, Flint J, Damgaard M, Bonfils PK, Cheng AS, Aggarwal S, et al. Calculated plasma volume status and prognosis in chronic heart failure. Eur J Heart Fail 2015; 17: 35–43.
- 21.
Martens P, Nijst P, Dupont M, Mullens W. The optimal plasma volume status in heart failure in relation to clinical outcome. J Card Fail 2019; 25: 240–248.
- 22.
Maznyczka AM, Barakat MF, Ussen B, Kaura A, Abu-Own H, Jouhra F, et al. Calculated plasma volume status and outcomes in patients undergoing coronary bypass graft surgery. Heart 2019; 105: 1020–1026.
- 23.
Lombardi CM, Carubelli V, Peveri G, Inciardi RM, Pagnesi M, Ravera A, et al. Prognostic significance of serum potassium in patients hospitalized for acute heart failure. ESC Heart Fail 2022; 9: 2357–2366.
- 24.
Desai AS, Liu J, Pfeffer MA, Claggett B, Fleg J, Lewis EF, et al. Incident hyperkalemia, hypokalemia, and clinical outcomes during spironolactone treatment of heart failure with preserved ejection fraction: Analysis of the TOPCAT trial. J Card Fail 2018; 24: 313–320.
- 25.
Savarese G, Xu H, Trevisan M, Dahlstrom U, Rossignol P, Pitt B, et al. Incidence, predictors, and outcome associations of dyskalemia in heart failure with preserved, mid-range, and reduced ejection fraction. JACC Heart Fail 2019; 7: 65–76.
- 26.
Ferreira JP, Claggett BL, Liu J, Desai AS, Pfeffer MA, Anand IS, et al. Serum potassium and outcomes in heart failure with preserved ejection fraction: A post-hoc analysis of the PARAGON-HF trial. Eur J Heart Fail 2021; 23: 776–784.
- 27.
Yamashita K, Ito K, Endo J, Matsuhashi T, Katsumata Y, Yamamoto T, et al. Adrenal cortex hypoxia modulates aldosterone production in heart failure. Biochem Biophys Res Commun 2020; 524: 184–189.
- 28.
Mentz RJ, Kelly JP, von Lueder TG, Voors AA, Lam CS, Cowie MR, et al. Noncardiac comorbidities in heart failure with reduced versus preserved ejection fraction. J Am Coll Cardiol 2014; 64: 2281–2293.
- 29.
Lam CSP, Voors AA, de Boer RA, Solomon SD, van Veldhuisen DJ. Heart failure with preserved ejection fraction: From mechanisms to therapies. Eur Heart J 2018; 39: 2780–2792.
- 30.
Nishihara T, Tokitsu T, Sueta D, Takae M, Oike F, Fujisue K, et al. Serum potassium and cardiovascular events in heart failure with preserved left ventricular ejection fraction patients. Am J Hypertens 2018; 31: 1098–1105.
- 31.
Bernier GM, Conrad ME. Catabolsm of human beta-2-microglobulin by the rat kidney. Am J Physiol 1969; 217: 1359–1362.
- 32.
Miyata T, Jadoul M, Kurokawa K, Van Ypersele de Strihou C. Beta-2 microglobulin in renal disease. J Am Soc Nephrol 1998; 9: 1723–1735.
- 33.
Nangaku M. Chronic hypoxia and tubulointerstitial injury: A final common pathway to end-stage renal failure. J Am Soc Nephrol 2006; 17: 17–25.
- 34.
Ronco C, Haapio M, House AA, Anavekar N, Bellomo R. Cardiorenal syndrome. J Am Coll Cardiol 2008; 52: 1527–1539.
- 35.
Legrand M, Mik EG, Johannes T, Payen D, Ince C. Renal hypoxia and dysoxia after reperfusion of the ischemic kidney. Mol Med 2008; 14: 502–516.
- 36.
Otaki Y, Watanabe T, Takahashi H, Narumi T, Kadowaki S, Honda Y, et al. Association of renal tubular damage with cardio-renal anemia syndrome in patients with heart failure. Int J Cardiol 2014; 173: 222–228.
- 37.
Otaki Y, Watanabe T, Takahashi H, Funayama A, Kinoshita D, Yokoyama M, et al. Comorbid renal tubular damage and hypoalbuminemia exacerbate cardiac prognosis in patients with chronic heart failure. Clin Res Cardiol 2016; 105: 162–171.
- 38.
Otaki Y, Watanabe T, Shimizu M, Tachibana S, Sato J, Kobayashi Y, et al. Association of malnutrition with renal dysfunction and clinical outcome in patients with heart failure. Sci Rep 2022; 12: 16673.
- 39.
Kittleson MM, Panjrath GS, Amancherla K, Davis LL, Deswal A, Dixon DL, et al. 2023 ACC expert consensus decision pathway on management of heart failure with preserved ejection fraction: A report of the American College of Cardiology Solution Set Oversight Committee. J Am Coll Cardiol 2023; 81: 1835–1878.
- 40.
Dekkers CCJ, Sjöström CD, Greasley PJ, Cain V, Boulton DW, Heerspink HJL. Effects of the sodium-glucose co-transporter-2 inhibitor dapagliflozin on estimated plasma volume in patients with type 2 diabetes. Diabetes Obes Metab 2019; 21: 2667–2673.
- 41.
Takashima H, Yoshida Y, Nagura C, Furukawa T, Tei R, Maruyama T, et al. Renoprotective effects of canagliflozin, a sodium glucose cotransporter 2 inhibitor, in type 2 diabetes patients with chronic kidney disease: A randomized open-label prospective trial. Diab Vasc Dis Res 2018; 15: 469–472.
- 42.
Rossignol P, Menard J, Fay R, Gustafsson F, Pitt B, Zannad F. Eplerenone survival benefits in heart failure patients post-myocardial infarction are independent from its diuretic and potassium-sparing effects. Insights from an EPHESUS (Eplerenone Post-Acute Myocardial Infarction Heart Failure Efficacy and Survival Study) substudy. J Am Coll Cardiol 2011; 58: 1958–1966.
