Acute Kidney Injury Impacts on Hypokalemia Associated with Yokukansan Preparation: A Retrospective Observational Study

Abstract

The time course of acute kidney injury and hypokalemia remains unelucidated. We investigated whether altered renal function impacts hypokalemia and clinical predictors for acute kidney injury in patients who used Yokukansan preparation. We performed a secondary analysis of retrospective observational cohort data from adult patients who started Yokukansan preparation. The study was conducted from June 2015 to May 2019 at Tokyo Women’s Medical University, Medical Center East. The effect of acute kidney injury (>1.5-fold increase from baseline serum creatinine level) or renal function recovery on hypokalemia (serum potassium level <3.0 mEq/L) was investigated. The clinical predictors for acute kidney injury were determined using a multivariate Cox proportional hazard analysis. Out of 258 patients, 12 patients had both outcomes, and all but one patient experienced in the order of acute kidney injury and hypokalemia. Excluding one patient, hypokalemia occurred in 11/34 (32%) patients after acute kidney injury and 27/223 (12%) patients without acute kidney injury (p = 0.005). Hypokalemia occurred in 9/25 (36%) of acute kidney injury with recovery, 2/9 (22%) of acute kidney injury without recovery, and 27/223 (12%) of no acute kidney injury (p = 0.014). Patients with acute kidney injury showed a late onset of hypokalemia compared with those without acute kidney injury (p = 0.001). In 258 patients, multivariate Cox proportional hazard analysis showed that high systolic blood pressure and mean arterial pressure increased the risk of acute kidney injury. Clinicians should remember that hypokalemia developed after acute kidney injury while Yokukansan preparation treatment.

INTRODUCTION

Acute kidney injury (AKI) is associated with substantial morbidity and mortality in various populations.^1–3) AKI pathogenesis is categorized as prerenal, renal, and postrenal.^4,5) A thorough understanding of AKI etiology (e.g., septic shock, acute heart failure, electrolyte disorders, and medications) is essential to determine therapeutic strategy.⁶⁾

Clinicians strongly invested in AKI treatment to discharge patients earlier and improve the late consequences of AKI. The introduction of intensive treatment for AKI could serve to increase renal potassium excretion, thereby driving the normalization of the serum potassium level because of the renal recovery process and the management of hyperkalemia.⁷⁾ Thus, the recovery of renal function is a potential risk of hypokalemia. Conversely, hypokalemia itself was associated with the deterioration risk of renal function and all-cause mortality.^8,9) There is a lack of information on the chronological order of the clinical events of AKI and hypokalemia.

Yokukansan and Yokukansankachimpihange (both Tsumura Co., Tokyo, Japan), which include an equivalent amount of Glycyrrhiza. Yokukansan contains seven crude components (Atractylodes lancea rhizome, Poria sclerotium, Cnidium sp. rhizome, Uncaria sp. hook, Angelicae radix, Bupleuri radix, and Glycyrrhiza sp.). Yokukansankachimpihange contains Citrus unshiu peel and Pinellia sp. tuber in addition to Yokukansan. Glycyrrhetinic acid, the active metabolite of Yokukansan preparation (Yokukansan and Yokukansankachimpihange), inhibits 11β-hydroxysteroid dehydrogenase type 2, resulting in hypokalemia.¹⁰⁾ Elderly patients often have a high risk of hypokalemia associated with Yokukansan preparation and AKI in clinical settings.^11–13) Additionally, Yokukansan preparation induces pseudo-aldosteronism (hypokalemia and hypertension) and hypertension could be a potential predictor for AKI.^14,15) Yokukansan preparation is often used in the older population to suppress the symptoms such as those of Alzheimer’s disease by blocking the activity of the glutamate nervous system and stimulating serotonin 5-HT_1A receptor.^16,17) Therefore, we selected Yokukansan preparation as an ideal proven drug to address the clinical question regarding the time–courses of AKI and hypokalemia.

This study aimed to identify the timeline of development/recovery of AKI and hypokalemia, the relationship between development/recovery of AKI and hypokalemia, and clinical predictors for AKI.

MATERIALS AND METHODS

Study Design and Patient Eligibility

We further analyzed a previously described single-center retrospective observational cohort dataset collected between June 2015 and May 2019 at Tokyo Women’s Medical University, Medical Center East (a 450-bed university hospital in Tokyo, Japan).¹³⁾ The study population included inpatients and outpatients aged >20 years who started Yokukansan preparation between June 2015 and May 2019 with monitored their serum potassium level during the treatment. The exclusion criteria included a baseline serum potassium level of <3.0 mEq/L, no multiple data for serum creatinine level, and baseline estimated glomerular filtration rate (eGFR) <30 mL/min/1.73 m². This study was conducted in compliance with the Declaration of Helsinki and approved by the Institutional Review Board at Tokyo Women’s Medical University (#5199).

Data Collection

The patients’ electronic medical records were reviewed to collect data for demographic characteristics (sex, age, height, body weight, and body mass index), primary diagnosis (delirium, dementia, and other psychological illness), clinical laboratory data (serum albumin, blood urea nitrogen, serum creatinine, serum sodium, serum potassium, and serum chloride), vital signs (systolic blood pressure (SBP), diastolic blood pressure (DBP), and heart rate), details of the Yokukansan preparation (dose and follow-up period), and medications of interest (benzodiazepine, ramelteon, suvorexant, cholinesterase inhibitors (donepezil, rivastigmine, and galantamine), N-methyl-D-aspartate antagonist (memantine), typical or atypical antipsychotics, oral corticosteroids, angiotensin-converting enzyme inhibitors, angiotensin II receptor blockers, loop diuretics, thiazide diuretics, and potassium-sparing diuretics). We calculated the ratio of blood urea nitrogen to serum creatinine to assess the volume status.¹⁸⁾ A prediction equation, including sex, age, and serum creatinine level, was used to determine the eGFR.¹⁹⁾ Because we could not obtain blood gas data from the electronic medical records, we considered the gap between serum sodium and serum chloride as an alternative marker for the acid–base disturbance.²⁰⁾ We calculated mean arterial pressure (MAP) using the following equation.

  

Outcome

We assessed two study outcomes during the Yokukansan preparation treatment: AKI and hypokalemia (serum potassium level <3.0 mEq/L). Because data on urine volume were unavailable, we assessed AKI based on Kidney Disease Improving Global Outcomes (KDIGO) criteria, using the definition of >1.5 times increase from the baseline serum creatinine level.²¹⁾ There are concerns that the KDIGO criteria of a fixed increase of serum creatinine gives a false-positive result.²²⁾ Additionally, KDIGO timing criteria (48 h or 7 d) were primarily designed for intensive care settings.²¹⁾

Statistical Analysis

Normally distributed data were reported as mean ± standard deviations and compared using Student’s t-test, whereas non-normally distributed data were reported as medians and interquartile ranges (IQR) and compared using Mann–Whitney’s U-test. Categorical data were presented as numbers and percentages, and heterogeneity was assessed using chi-square test.

Time-to-event outcomes (AKI or hypokalemia associated with Yokukansan preparation) were analyzed based on Kaplan–Meier curves. The follow-up period was defined as the time from the introduction of Yokukansan preparation to the discontinuation of Yokukansan preparation or the development of outcome (AKI or hypokalemia).

We compared the median onset and the incidence of hypokalemia between patients with and without AKI. Because we considered increased potassium excretion as the evidence of renal function recovery, we further examined the development of hypokalemia according to AKI recovery. The recovery of renal function was defined as no less than a 1.0-fold decrease from baseline serum creatinine level within 1 week after AKI.²³⁾ We corrected the p value by Bonferroni’s correction. The corrected p values were set at 0.025 (0.05 divided by 2) assuming that two intergroup analyses were conducted. The clinical features of interest in patients with AKI and hypokalemia were summarized. Furthermore, Kaplan–Meier curves analyzed the cumulative incidence of hypokalemia stratified by AKI above mentioned. This comparison was used by log-rank test. The index dates were the start of Yokukansan preparation for patients without AKI and the development of AKI for patients with AKI.

We preliminary investigated clinical predictors for AKI using a multivariate Cox proportional hazard analysis. We assumed that approximately 12% of the older population developed AKI.^24,25) Thus, our study required at least 250 patients to identify three independent variables in the final model if we assumed that 10 patients were required to detect one independent variable. The dependent variable was AKI development during the Yokukansan preparation treatment. The independent variables were demographic characteristics, primary diagnosis, clinical laboratory data, vital signs, the Yokukansan preparation dose, and medications of interest. eGFR was categorized as a binary variable (< 60 or ≥ 60 mL/min/1.73 m²) because this cut-off value was used for diagnosing chronic kidney disease.²⁶⁾ The daily dose of the Yokukansan preparation was categorized as ≥7.5 or <7.5 g because a daily dose of 7.5 g of Yokukansan preparation has been used in clinical trials.²⁷⁾ We prescreened potential independent variables using univariate Cox proportional hazard analyses, and all independent variables with a p value <0.10 were further evaluated in the multivariate Cox proportional hazard analysis. We confirmed each independent variable for multicollinearity. Additionally, we constructed three final models using three blood pressure components (SBP, DBP, and MAP), because the effect of each blood pressure on renal function differed.²⁸⁾ We used a stepwise backward selection method to construct three final models according to the Akaike information criterion. The robustness of the final models was confirmed using a stepwise forward selection method. Three final models were used to determine the hazard ratio (HR) and 95% confidence interval (95% CI). We attempted to examine the interactions of independent variables by introducing a cross-product term of respective independent variables into each final model as needed.

All statistical analyses were two-sided and performed using JMP^® Pro 14 (SAS Institute Inc., Cary, NC, U.S.A.). A p value <0.05 was considered statistically significant unless otherwise mentioned. Figures were prepared using GraphPad Prism ver. 8.00 (GraphPad Software, San Diego, CA, U.S.A.).

RESULTS

Study Population

A flow chart of the eligible populations is depicted in Fig. 1. We identified 678 patients at Tokyo Women’s Medical University, Medical Center East, who were >20 years old and who had started Yokukansan preparation between June 2015 and May 2019. Of them, we excluded 13 with a baseline serum potassium level <3.0 mEq/L, 307 with no multiple data for serum creatinine, and 100 with a baseline eGFR <30 mL/min/1.73 m². The remaining 258 patients were included in the main analysis.

Fig. 1. The Flow Chart of Eligible Populations

We identified 678 patients between June 2015 and May 2019 at Tokyo Women’s Medical University, Medical Center East, who were >20 years old and used Yokukansan preparation (Yokukansan and Yokukansankachimpihange). We excluded 420 patients because of baseline serum potassium level <3.0 mEq/L, no multiple data for serum creatinine level, and baseline estimated glomerular filtration rate <30 mL/min/1.73 m² (13, 307, and 100 patients, respectively). The final study cohort included 258 patients. Of them, 35 (14%) developed acute kidney injury, and 223 (86%) of patients did not. Hypokalemia developed in 12 (34%) patients in the acute kidney injury group and 27 (12%) patients in the no acute kidney injury group. Among 12 patients with acute kidney injury and hypokalemia, 11 (92%) patients experienced subsequent hypokalemia after acute kidney injury. N, number; K, potassium; Cr, creatinine; eGFR, estimated glomerular filtration rate, AKI; acute kidney injury.

Table 1 shows the baseline clinical characteristics of the study cohort. Approximately 60% of patients were male. The median age was younger in the AKI group than the no AKI group (75 [68–81] vs. 80 [73–85] years old, p < 0.01). The majority of patients had hypoalbuminemia (mean serum albumin <3.0 g/dL), and 83% of the patients in both groups received a daily dose of ≥ 7.5 g Yokukansan preparation.

Table 1. Baseline Characteristics in the Study Population Who Underwent the Yokukansan Preparation

Variables	AKI, N = 35	No AKI, N = 223	p Value
Demographic characteristics
Male, N (%)	22 (63)	126 (57)	0.48
Age, years	75 [68–81]	80 [73–85]	<0.01
Height, cm	159 ± 9	158 ± 10	0.43
Body weight, kg	55 ± 15	54 ± 12	0.69
Body mass index, kg/m2	22 ± 4	22 ± 4	0.74
Main diagnosis			0.06
Delirium, N (%)	27 (77)	134 (60)
Dementia, N (%)	3 (9)	16 (7)
Other psychological illness, N (%)	5 (14)	73 (33)
Laboratory data
Serum Alb, g/dL	2.5 ± 0.7	2.9 ± 0.7	<0.01
BUN, mg/dL	16 [11–35]	16 [12–22]	0.47
Serum Cr, mg/dL	0.75 [0.65–1.03]	0.75 [0.57–1.01]	0.84
BUN/Cra)	24 [15–41]	20 [16–28]	0.11
eGFRb), mL/min/1.73 m2	69 [50–87]	68 [48–87]	0.83
eGFRb), < 60 mL/min/1.73 m2, N (%)	15 (43)	91 (41)	0.82
Serum Na, mEq/L	141 ± 5	139 ± 5	0.12
Serum K, mEq/L	3.9 ± 0.6	3.9 ± 0.6	0.58
Serum Cl, mEq/L	103 ± 5	102 ± 5	0.93
Serum Na–serum Clc), mEq/L	36 ± 4	36 ± 3	0.71
Vital signs
SBP, mmHg	131 ± 22	133 ± 24	0.59
DBP, mmHg	68 ± 16	74 ± 15	0.06
MAPd), mmHg	88 ± 16	94 ± 16	0.10
Heart rate, beats per min	79 ± 16	86 ± 22	0.07
Details of Yokukansan preparatione)
Type of Yokukansan preparatione)			0.35
Yokukansan, N (%)	35 (100)	220 (99)
Yokukansankachimpihange, N (%)	0 (0)	3 (1)
Yokukansan preparatione) dose ≥ 7.5 g/d, N (%)	29 (83)	186 (83)	0.99
Follow-up period, days	14 [6–47]	10 [4–26]	0.29
Medications of interest
Central nervous system medications, N (%)	30 (86)	173 (78)	0.25
Benzodiazepine, N (%)	5 (14)	34 (15)	0.88
Ramelteon, N (%)	22 (63)	112 (50)	0.16
Suvorexant, N (%)	5 (14)	16 (7)	0.18
Cholinesterase inhibitorsf), N (%)	3 (9)	19 (9)	0.99
Memantine, N (%)	0 (0)	14 (6)	0.04
Typical antipsychotics, N (%)	2 (6)	11 (5)	0.85
Atypical antipsychotics, N (%)	6 (17)	44 (20)	0.72
Oral corticosteroids, N (%)	4 (11)	11 (5)	0.16
ACEi or ARB, N (%)	11 (31)	58 (26)	0.51
Loop diuretics, N (%)	8 (23)	27 (12)	0.11
Thiazide diuretics, N (%)	0 (0)	3 (1)	0.35
Potassium-sparing diuretics, N (%)	7 (20)	19 (9)	0.06

a) A ratio of BUN/Cr was used as an alternative marker of volume status. b) The formula used to calculate the eGFR included sex, age, and serum Cr level. c) The gap between serum Na and serum Cl is an alternative marker of the acid–base disturbance. d) We calculated MAP using the following equation (MAP = 1/3 × SBP + 2/3 × DBP). e) The Yokukansan preparation includes Yokukansan and Yokukansankachimpihange. f) Cholinesterase inhibitors include donepezil, rivastigmine, and galantamine. AKI; acute kidney injury, N; number, Alb; albumin, BUN; blood urea nitrogen, Cr; creatinine, eGFR; estimated glomerular filtration rate, Na; sodium, K; potassium, Cl; chloride, SBP; systolic blood pressure, DBP; diastolic blood pressure, MAP; mean arterial pressure, ACEi; angiotensin-converting enzyme inhibitor, ARB; angiotensin II receptor blocker. Normally, distributed data were reported as the mean ± standard deviation and were compared using Student’s t-test. Non-normally distributed data were reported as the median and interquartile range and compared using Mann–Whitney’s U-test. Categorical data were presented as numbers and percentages and were compared using chi-square test.

Trends of AKI and AKI Recovery on Hypokalemia

The Kaplan–Meier curve revealed that the cumulative incidence for AKI was 35 of 258 (14%) patients with a median follow-up period of 10 (IQR: 5–26) days (Fig. 2). The Kaplan–Meier curve for the cumulative incidence of hypokalemia is depicted in Supplementary Fig. 1.

Fig. 2. Kaplan–Meier Curve for Acute Kidney Injury

X-Axis and Y-axis show follow-up days and the cumulative incidence of acute kidney injury, respectively. Solid line shows the Kaplan–Meier curve for acute kidney injury, and dotted lines represent 95% confidence intervals. The number at risk is represented below the X-axis. AKI, acute kidney injury.

There were 12 of 35 (34%) patients with both AKI and hypokalemia. The median onset of hypokalemia was 35 (IQR: 18–146) days in patients with AKI, which was significantly longer than in patients without AKI (18 d, IQR:10–36) (p = 0.001). Table 2 shows the clinical features of interest in patients with both AKI and hypokalemia. There were 11 of 12 (92%) patients who experienced subsequent hypokalemia after AKI. The median period from AKI to the development of hypokalemia was 30 (IQR: 13–47) days.

Table 2. Summary of Characteristics in Patients with Acute Kidney Injury and Hypokalemia

N	AKI recovery	AKI to hypokalemia, d	Sex	Age, years	BMI, kg/m2	Serum Alb, g/dL	Serum Cr, mg/dL	Serum Cr ratio at AKIc)	K reducing drugsa)
1	Yes	47	M	79	24	2.2	0.32	1.8	Yes
2	Yes	22	M	82	22	2.1	0.69	3.6	Yes
3	Yes	40	F	67	18	2.5	0.29	3.7	No
4	Yes	35	M	68	22	2.5	0.52	1.8	No
5	Yes	138	M	67	23	1.5	0.86	3.1	No
6	Yes	78	M	68	24	1.9	0.83	1.8	Yes
7	Yes	30	M	76	24	1.3	1.03	1.7	No
8	Yes	16	M	74	NA	3.7	0.78	1.5	No
9	Yes	13	F	90	18	1.7	1.11	1.5	No
10	No	4	F	84	18	3.3	0.73	1.7	Yes
11b)	No	NA	M	80	16	2.4	1.68	1.9	No
12	No	13	M	68	23	3.2	0.72	1.6	No

a) K reducing drugs include loop and thiazide diuretics. b) The patient developed hypokalemia before AKI. c) Serum creatinine ratio at AKI indicates serum creatinine level at AKI dividing by baseline serum Cr level. N; number, AKI: acute kidney injury, BMI; body mass index, Alb; albumin, Cr; creatinine, K; potassium, M: male, F: female, NA: not available.

Excluding one patient who developed hypokalemia prior to AKI, the clinical courses of AKI and hypokalemia associated with Yokukansan preparation are summarized in Table 3. The risk of hypokalemia was higher in 11 of 34 (32%) patients with AKI than in 27 of 223 (12%) patients without AKI (p = 0.005). Of the 34 patients who experienced AKI, 25 (74%) recovered. Hypokalemia was observed in 9 of 25 (36%) patients with AKI recovery, 2 of 9 (22%) patients without AKI recovery, and 27 of 223 (12%) patients without AKI (p = 0.014). The median onset of hypokalemia in patients with and without AKI recovery was significantly longer than that in patients without AKI (56 [23–149] vs. 25 [10–615] vs. 18 [10–36], p < 0.001). However, the Kaplan–Meier curve revealed no difference in the cumulative incidence of hypokalemia (Supplementary Fig. 2, log-rank test, p = 0.303).

Table 3. Summary of Hypokalemic Events with or without Acute Kidney Injury

Variables	AKI	AKI recovery	No AKI recovery	No AKI	p Valuea)	p Valueb)
N	34	25	9	223
Hypokalemia, N (%)	11 (32)	9 (36)	2 (22)	27 (12)	0.005	0.014
Onset of hypokalemia, d	36 [18–148]	56 [23–149]	25 [10–615]	18 [10–36]	0.001	<0.001
Onset of AKI, d	12 [6–49]	14 [6–51]	8 [5–412]	NA	NA	NA

a) p Value focused on AKI and no AKI. b) p Value focused on AKI recovery, no AKI recovery, and no AKI. AKI; acute kidney injury, N; number, NA: not available. Out of 35 patients with AKI, we excluded one patient who developed hypokalemia prior to AKI. Thus, this summary includes patients with AKI (N = 34). We further divided them into AKI recovery (N = 25) and no AKI recovery (N = 9). Non-normally distributed data were reported as median and interquartile range and compared using Mann–Whitney’s U-test. Categorical data were presented as numbers and percentages and compared using chi-square test. p Values were corrected using Bonferroni’s correction. The corrected p value was 0.025 because of 0.05 divided by 2.

Clinical Predictors of AKI

Focusing on included 258 patients, univariate Cox proportional hazard analyses identified SBP, DBP, MAP, oral corticosteroids, angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers, potassium-sparing diuretics, and a daily dose of Yokukansan preparation of ≥ 7.5 g as potential independent variables for AKI. There was no multicollinearity except for the blood pressure component. A stepwise backward selection method determined three final models (Table 4), all of which remained unchanged regardless of a stepwise forward selection method. The final model demonstrated that SBP was the clinical predictor for AKI during Yokukansan preparation treatment (Model 1, HR = 1.007, 95% CI; 1.000–1.013, p = 0.036). However, DBP was an insignificant clinical predictor for AKI (Model 2, HR = 1.008, 95% CI; 0.999–1.018, p = 0.093). Conversely, MAP was a significant predictor for AKI (Model 3, HR = 1.010, 95% CI; 1.001–1.019, p = 0.035). A daily dose of ≥7.5 g/d Yokukansan preparation was significantly associated with AKI in Model 1 (HR = 1.520, 95% CI; 1.026–2.325, p = 0.036). As shown in Table 1, the limited number of patients who used Yokukansan preparation <7.5 g/d might contribute to an alpha error (false-positive). We did not perform further analysis of the interaction between blood pressure components and the daily dose of Yokukansan preparation.

Table 4. Summary of Multivariate Cox Proportional Hazard Analysis for Acute Kidney Injury during the Yokukansan Preparation Treatment

Independent variables	Model 1: SBP			Model 2: DBP			Model 3: MAP
	HR	95% CI	p Value	HR	95% CI	p Value	HR	95% CI	p Value
SBP (per mmHg)	1.007	1.000–1.013	0.036
DBP (per mmHg)				1.008	0.999–1.018	0.093
MAP (per mmHg)							1.010	1.001–1.019	0.035
Yokukansan preparationa) ≥ 7.5 g/d (yes)	1.520	1.026–2.325	0.036	1.385	0.923–2.147	0.119	1.378	0.919–2.134	0.123

a) The Yokukansan preparation includes Yokukansan and Yokukansankachimpihange. SBP; systolic blood pressure, DBP; diastolic blood pressure, MAP; mean arterial pressure, HR; hazard ratio, 95% CI; 95% confidence interval.

DISCUSSION

This exploratory study affirms the development of many hypokalemic events associated with the Yokukansan preparation following AKI. Additionally, hypokalemia associated with Yokukansan preparation is common in AKI. Furthermore, high blood pressure components (SBP and MAP) are associated with AKI during the Yokukansan preparation treatment.

Hypokalemia itself is recognized as one of the causes of kidney damage.^8,29) The present study clarified that the majority of patients suffered from hypokalemia after AKI. A comparison of our results with other studies was difficult because this topic has not attracted considerable attention. Although we did not assess the overall information of the patients in our study, the clinical features of hypokalemic nephropathy were long-term hypokalemia, and the causes of serum potassium depletion include malnutrition and the overuse of medications such as diuretics and laxatives.³⁰⁾ However, many patients have hypoalbuminemia and delirium, it is thought that sensitivity to diuretics (e.g., furosemide) may be scarce³¹⁾ and patients receive generous support from medical staff under admission. Moreover, drug-induced hypokalemia was reported to contribute to life-threatening cardiotoxicity such as serious arrhythmia.³²⁾ Therefore, we suggest that medical staff be aware of the occurrence of secondary hypokalemia with Yokukansan preparation after AKI.

We consider that the improvement of renal function is physiologically associated with the normalization of serum potassium levels. In fact, falling serum potassium levels after AKI reflect the recovery process of prerenal and renal AKI in various diagnoses.³³⁾ In addition, we consider that an increase in urine volume is also associated with better potassium renal excretion. Mild AKI was a favorable indicator of the recovery of renal function.³⁴⁾ Therefore, it appears that the conditions of our study population might be easily reversible by optimal treatments because of the AKI definition (>1.5-fold increase from the baseline serum creatinine level). Although the restoration of renal function was desired, we should avert the excessive decline of serum potassium levels. Further analysis was required to clarify the relationship between AKI and other Kampo medicines such as Shakuyakukanzoto.

On the contrary, even AKI without subsequent recovery decreased serum potassium level, which is largely similar to AKI with subsequent recovery. This may be attributed to the administration of intravenous fluids, including a low potassium content, instead of food. A majority of patients with AKI had hypoalbuminemia (Table 2), which supports the presence of a poor nutritional status. We speculate that hypoalbuminemia itself increases the risk of hypokalemia associated with Yokukansan preparation because hypoalbuminemia lowers the glucuronidation capacity.³⁵⁾ Furthermore, the fractional excretion of potassium was reported to be significantly higher in patients with persistent AKI compared to patients with transient AKI.³⁶⁾ Because we were unable to collect a sufficiently large sample size to analyze the effect of renal function recovery on hypokalemia, further analysis is warranted to clarify this topic.

All blood pressure components, but not DBP, are significant risk factors for AKI as we are concerned. Previously, studies demonstrated that any blood pressure components play an essential role in the deterioration of renal function.^37–39) In particular, both SBP and MAP strongly affect AKI development.²⁸⁾ Indeed, few study subjects had severe hypertension. This suggests that even mild hypertension could contribute toward AKI, and our results are supported by Peralta et al.⁴⁰⁾ Furthermore, Yokukansan preparation induces not only hypokalemia but also hypertension.⁴¹⁾ Thus, it is possible that the Yokukansan preparation itself poses an indirect risk of AKI, as shown in Table 4. Close monitoring of serum potassium level and blood pressure is essential for patients who use Yokukansan preparations.

This study has several limitations for interpreting its results. First, because the study is a retrospective design, we could not avert bias and unknown confounding factors. Second, the type and causes of AKI were unknown. Additionally, we were unable to collect data of variables that affect serum potassium level (e.g., food intake, urine volume, and diarrhea) because this was not entirely included in the medical charts. Third, we could not avoid the possibility of over/underestimation of renal function and the incidence of AKI because most study patients have heterogeneous backgrounds such as the elderly, hypoalbuminemia, and intensive care unit (data not shown). Fourth, the timing of the measurement of clinical laboratory data varied. Thus, the index date of outcomes might be inaccurate. Fifth, we could not assess the acid–base imbalance, urine electrolyte clearance, or endocrinological function of the studied patients. Finally, the modest sample size limited detailed analysis of AKI severity.

In conclusion, this exploratory study indicated Yokukansan preparation associated with hypokalemia that might develop after AKI. Further refinement of this timeframe is necessary. Additionally, blood pressure management will enable patients to continue the treatment with the Yokukansan preparation.

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

The online version of this article contains supplementary materials.

REFERENCES

• 1) Ricci Z, Cruz D, Ronco C. The RIFLE criteria and mortality in acute kidney injury: a systematic review. Kidney Int., 73, 538–546 (2008).
• 2) Susantitaphong P, Cruz DN, Cerda J, Abulfaraj M, Alqahtani F, Koulouridis I, Jaber BL. World incidence of AKI: a meta-analysis. Clin. J. Am. Soc. Nephrol., 8, 1482–1493 (2013).
• 3) See EJ, Jayasinghe K, Glassford N, Bailey M, Johnson DW, Polkinghorne KR, Toussaint ND, Bellomo R. Long-term risk of adverse outcomes after acute kidney injury: a systematic review and meta-analysis of cohort studies using consensus definitions of exposure. Kidney Int., 95, 160–172 (2019).
• 4) Moore PK, Hsu RK, Liu KD. Management of acute kidney injury: core curriculum 2018. Am. J. Kidney Dis., 72, 136–148 (2018).
• 5) Singri N, Ahya SN, Levin ML. Acute renal failure. JAMA, 289, 747–751 (2003).
• 6) Gameiro J, Fonseca JA, Outerelo C, Lopes JA. Acute kidney injury: from diagnosis to prevention and treatment strategies. J Clin Med, 9, 1704 (2020).
• 7) Fry AC, Farrington K. Management of acute renal failure. Postgrad. Med. J., 82, 106–116 (2006).
• 8) Yalamanchili HB, Calp-Inal S, Zhou XJ, Choudhury D. Hypokalemic nephropathy. Kidney Int. Rep., 3, 1482–1488 (2018).
• 9) Kovesdy CP, Matsushita K, Sang Y, et al. Serum potassium and adverse outcomes across the range of kidney function: a CKD Prognosis Consortium meta-analysis. Eur. Heart J., 39, 1535–1542 (2018).
• 10) Tanahashi T, Mune T, Morita H, Tanahashi H, Isomura Y, Suwa T, Daido H, Gomez-Sanchez CE, Yasuda K. Glycyrrhizic acid suppresses type 2 11 beta-hydroxysteroid dehydrogenase expression in vivo. J. Steroid Biochem. Mol. Biol., 80, 441–447 (2002).
• 11) Kamei S, Kaneto H, Irie S, Kinoshita T, Tanabe A, Hirukawa H, Tatsumi F, Shimoda M, Kohara K, Mune T, Kaku K. Pseudoaldosteronism induced by Yokukansan in an elderly Japanese type 2 diabetic patient with Alzheimer’s disease. J. Diabetes Investig., 6, 487–488 (2015).
• 12) Chronopoulos A, Cruz DN, Ronco C. Hospital-acquired acute kidney injury in the elderly. Nat. Rev. Nephrol., 6, 141–149 (2010).
• 13) Hirai T, Yamaga R, Kei M, Hosohata K, Itoh T. Geriatric patients are at a high risk of hypokalemia associated with Yokukansan preparation: a retrospective cohort study. Biol. Pharm. Bull., 43, 1742–1748 (2020).
• 14) Bidani AK, Griffin KA. Pathophysiology of hypertensive renal damage: implications for therapy. Hypertension, 44, 595–601 (2004).
• 15) Conn JW, Rovner DR, Cohen EL. Licorice-induced pseudoaldosteronism: hypertension, hypokalemia, aldosteronopenia, and suppressed plasma renin activity. JAMA, 205, 492–496 (1968).
• 16) Ikarashi Y, Mizoguchi K. Neuropharmacological efficacy of the traditional Japanese Kampo medicine yokukansan and its active ingredients. Pharmacol. Ther., 166, 84–95 (2016).
• 17) Furukawa K, Tomita N, Uematsu D, et al. Randomized double-blind placebo-controlled multicenter trial of Yokukansan for neuropsychiatric symptoms in Alzheimer’s disease. Geriatr. Gerontol. Int., 17, 211–218 (2017).
• 18) Manoeuvrier G, Bach-Ngohou K, Batard E, Masson D, Trewick D. Diagnostic performance of serum blood urea nitrogen to creatinine ratio for distinguishing prerenal from intrinsic acute kidney injury in the emergency department. BMC Nephrol., 18, 173 (2017).
• 19) Matsuo S, Imai E, Horio M, Yasuda Y, Tomita K, Nitta K, Yamagata K, Tomino Y, Yokoyama H, Hishida A. Revised equations for estimated GFR from serum creatinine in Japan. Am. J. Kidney Dis., 53, 982–992 (2009).
• 20) Nagaoka D, Nassar Junior AP, Maciel AT, Taniguchi LU, Noritomi DT, Azevedo LC, Neto LM, Park M. The use of sodium-chloride difference and chloride-sodium ratio as strong ion difference surrogates in the evaluation of metabolic acidosis in critically ill patients. J. Crit. Care, 25, 525–531 (2010).
• 21) Eknoyan G, Lameire N, Eckardt K, Kasiske B. Kidney disease: improving global outcomes (KDIGO) acute kidney injury work group. KDIGO clinical practice guideline for acute kidney injury. Kidney Int. Suppl., 2, 1–138 (2012).
• 22) Lin J, Fernandez H, Shashaty MG, Negoianu D, Testani JM, Berns JS, Parikh CR, Wilson FP. False-positive rate of AKI using consensus creatinine-based criteria. Clin. J. Am. Soc. Nephrol., 10, 1723–1731 (2015).
• 23) Duff S, Murray PT. Defining early recovery of acute kidney injury. Clin. J. Am. Soc. Nephrol., 15, 1358–1360 (2020).
• 24) Li Q, Zhao M, Wang X. AKI in the very elderly patients without preexisting chronic kidney disease: a comparison of 48-hour window and 7-day window for diagnosing AKI using the KDIGO criteria. Clin. Interv. Aging, 13, 1151–1160 (2018).
• 25) Turgutalp K, Bardak S, Horoz M, Helvacı I, Demir S, Kiykim AA. Clinical outcomes of acute kidney injury developing outside the hospital in elderly. Int. Urol. Nephrol., 49, 113–121 (2017).
• 26) Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF 3rd, Feldman HI, Kusek JW, Eggers P, Van Lente F, Greene T, Coresh J. A new equation to estimate glomerular filtration rate. Ann. Intern. Med., 150, 604–612 (2009).
• 27) Okamoto H, Iyo M, Ueda K, Han C, Hirasaki Y, Namiki T. Yokukan-san: a review of the evidence for use of this Kampo herbal formula in dementia and psychiatric conditions. Neuropsychiatr. Dis. Treat., 10, 1727–1742 (2014).
• 28) Young JH, Klag MJ, Muntner P, Whyte JL, Pahor M, Coresh J. Blood pressure and decline in kidney function: findings from the Systolic Hypertension in the Elderly Program (SHEP). J. Am. Soc. Nephrol., 13, 2776–2782 (2002).
• 29) Wang HH, Hung CC, Hwang DY, Kuo MC, Chiu YW, Chang JM, Tsai JC, Hwang SJ, Seifter JL, Chen HC. Hypokalemia, its contributing factors and renal outcomes in patients with chronic kidney disease. PLOS ONE, 8, e67140 (2013).
• 30) Bock KD, Cremer W, Werner U. Chronic hypokalemic nephropathy: a clinical study. Klin. Wochenschr., 56 (Suppl. 1), 91–96 (1978).
• 31) Gupta R, Testani J, Collins S. Diuretic Resistance in Heart Failure. Curr. Heart Fail. Rep., 16, 57–66 (2019).
• 32) Khan S, Khan SU. Adverse drug event of hypokalaemia-induced cardiotoxicity secondary to the use of laxatives: a systematic review of case reports. J. Clin. Pharm. Ther., 45, 927–936 (2020).
• 33) Magden K, Yildirim I, Kutu M, Ozdemir M, Peynir S, Altas A, Yildiz G, Hur E. Recovery process in patients followed-up due to acute kidney injury. Hippokratia, 17, 239–242 (2013).
• 34) Mallhi TH, Khan AH, Adnan AS, Sarriff A, Khan YH, Gan SH. Short-term renal outcomes following acute kidney injury among dengue patients: a follow-up analysis from large prospective cohort. PLOS ONE, 13, e0192510 (2018).
• 35) Lu Y, Zhu J, Chen X, Li N, Fu F, He J, Wang G, Zhang L, Zheng Y, Qiu Z, Yu X, Han D, Wu L. Identification of human UDP-glucuronosyltransferase isoforms responsible for the glucuronidation of glycyrrhetinic acid. Drug Metab. Pharmacokinet., 24, 523–528 (2009).
• 36) Maciel AT, Park M, Macedo E. Fractional excretion of potassium in the course of acute kidney injury in critically ill patients: potential monitoring tool? Rev. Bras. Ter. Intensiva, 26, 143–147 (2014).
• 37) Baek SH, Chin HJ, Na KY, Chae D-W, Kim S. Optimal systolic blood pressure in noncritically ill patients with acute kidney injury: a retrospective cohort study. Kidney Res. Clin. Pract., 38, 356–364 (2019).
• 38) Micozkadioglu H. Higher diastolic blood pressure at admission and antiedema therapy is associated with acute kidney injury in acute ischemic stroke patients. Int. J. Nephrol. Renovasc. Dis., 7, 101–105 (2014).
• 39) Yang H, Guo X, Zhang X, Li Z, Yu S, Zheng L, Li W, Zhou Y, Sun Y. The relationship between mean arterial pressure and decreased glomerular filtration rate in rural areas of Northeast China. BMC Nephrol., 16, 137 (2015).
• 40) Peralta CA, Norris KC, Li S, Chang TI, Tamura MK, Jolly SE, Bakris G, McCullough PA, Shlipak M. Blood pressure components and end-stage renal disease in persons with chronic kidney disease: the Kidney Early Evaluation Program (KEEP). Arch. Intern. Med., 172, 41–47 (2012).
• 41) Sontia B, Mooney J, Gaudet L, Touyz RM. Pseudohyperaldosteronism, liquorice, and hypertension. J. Clin. Hypertens. (Greenwich), 10, 153–157 (2008).