Effectiveness of Chinese herbal medicine compared with angiotensin II receptor blockers in patients with diabetic kidney disease: A hospital-based matched cohort study

Abstract

Angiotensin II receptor blockers (ARBs) are one of the standard treatments for diabetic kidney disease (DKD). Some patients may opt for Chinese herbal medicine (CHM) of their own free will. However, there is no real-world evidence regarding the effectiveness and safety of CHM. We aimed to explore the effectiveness of CHM for DKD in comparison to ARBs. We enrolled 732 DKD patients (72 used only CHM and 661 used ARBs) from 2007 to 2016, and all patients were followed until December 2016 at China Medical University Hospital in Taiwan. A total of 355 ARB users and 71 CHM users were analyzed after propensity score matching. The estimated glomerular filtration rate (eGFR) after treatment was 84.9 ± 28.1 ml/min/1.73 m² in CHM users, which was higher than that (67.8 ± 35.4 ml/min/1.73 m²) in ARB users (p < 0.001). The change in the eGFR was −6.0 ± 21.4 ml/min/1.73 m² in CHM users and −12.9 ± 24.8 ml/min/1.73 m² in ARB users (p = 0.029). The blood urea nitrogen (BUN) and creatinine levels of patients taking CHM were 22 ± 16 mg/dl and 0.9 ± 0.4 mg/dl, respectively, and were lower than those (30 ± 28 mg/dl and 1.7 ± 2.0 mg/dl) of patients taking ARBs (p = 0.025 and p = 0.003). Using linear regression with adjustments for age, sex, BMI, baseline eGFR, and HbA1c levels, we found that the declines in the eGFR/baseline eGFR and changes in the urine albumin–creatinine ratio (ACR) were comparable between the two groups (p = 0.86 and 0.73). This study suggests that CHM may have comparable effectiveness to ARBs, which provides insights for further investigations.

Graphical abstract

List of abbreviations

DKD

Diabetic kidney disease

CHM

Chinese herbal medicines

ACEIs

Angiotensin-converting enzyme inhibitors

ARBs

Angiotensin II receptor blockers

CKD

Chronic kidney disease

eGFR

Estimated glomerular filtration rate

ACR

Urine albumin–creatinine ratio

TCM

Traditional Chinese medicine

GMP

Good manufacturing practice

BUN

Blood urea nitrogen

HbA1c

Hemoglobin A1c

CAD

Coronary artery disease

HTN

Hypertension

CVA

Cerebrovascular accident

SLE

Systemic lupus erythematosus

SBP

Systolic blood pressure

DBP

Diastolic blood pressure

CT

Computed tomography

MRI

Magnetic resonance imaging

BHJRST

Bai-Hu-Jia-Ren-Shen-Tang

JSSQW

Ji-Sheng-Shen-Qi-Wan

SNS

Si–Ni-San

MZRW

Ma-Zi-Ren-Wan

ZBDHW

Zhi-Bai-Di-Huang-Wan

1. Introduction

Diabetes affected more than 463 million people in 2019,¹ and its prevalence increased from 12.1 % to 19 % between 2005 and 2014.² Diabetic kidney disease (DKD) is a major cause of end-stage renal disease and is responsible for 45 % of cases requiring dialysis.³ DKD may increase the health-care burden because diabetes trends are increasing among young adults.^2,4,5 Microalbuminuria is the cardinal symptom of DKD⁶ and is linked to renal function decline. Angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin II receptor blockers (ARBs) are the current standard treatments for DKD.^{7, 8, 9, 10, 11} Both ACEIs and ARBs reduce albuminuria and slow DKD progression.

Chinese herbal medicine (CHM) is popular in Taiwan. Finished herbal products such as concentrated Chinese herbal granules are reimbursed by the National Health Insurance Program in Taiwan; 45.3 % of chronic kidney disease (CKD) patients and 77.9 % of diabetes patients take CHM.^12,13 CHM may slow DKD progression and prevent cardiovascular complications.¹⁴ Astragalus membranaceus (Astragalus, Huangqi) and Rheum officinale (Rhubarb) were shown to reduce proteinuria in clinical trials.^{15, 16, 17} All previous studies have explored the effect of CHM combined with ARBs/ACEIs,^12,18,19 but no study has investigated the effect of CHM compared with that of ARBs/ACEIs. Such studies are not possible in randomized control trials because ACEIs/ARBs are current standard treatments for patients with DKD. Not using ACEIs or ARBs in one group in a randomized clinical trial would be unethical. We reported a comparison of CHM and ARBs in a real-world setting. These patients opted to use CHM and did not use ARBs. ARBs were selected for this study because less than 10 % of the DKD patients in our database used ACEIs. A propensity score-matched cohort study was performed to analyze patients with DKD. The primary outcome measurements were the decline in the estimated glomerular filtration rate (eGFR) and the urine albumin–creatinine ratio (ACR). During the 10-year follow-up period, declines in the eGFR/baseline eGFR and changes in the urine ACR were comparable between CHM and ARB users. This information may provide valuable information for future pharmacological investigations and clinical trials.

2. Materials and methods

2.1. Study population and variables

We included all patients with DKD from 2007 to 2016 treated at China Medical University Hospital in Taichung, Taiwan, and all patients were followed up to December 2016. China Medical University Hospital provides both Chinese medicine and Western medicine services in Central Taiwan. Licensed traditional Chinese medicine (TCM) doctors are also trained in Western medical departments to ensure a basic understanding of knowledge to communicate with Western medical doctors. Both TCM doctors and Western medical doctors used the same diagnostic coding system (the International Classification of Diseases, Ninth Revision, Clinical Modification; ICD-9-CM coding system). DKD was identified using the ICD-9-CM codes 250.4X or 250.X with an ACR greater than 30 mg/g for three months. CHM users were defined as patients with at least two visits who took CHM for at least 14 days without taking ARBs. All single herbs and herbal formulas made by good manufacturing practice (GMP) pharmaceutical factories were included for further prescription analysis. ARB users were defined as patients with ARBs prescribed for more than 50 % of the follow-up period. Patients who used both CHM and ARBs, those with acute kidney injury according to the 2007 Acute Kidney Injury Network (AKIN) definition, and dialysis, kidney transplant, or cancer patients were excluded.

All patients were followed up at least 2 times every three months to collect data on creatinine levels, the eGFR and other laboratory data. We collected data before treatment and after treatment with CHM or ARBs every three months. For example, when the patients returned to the clinic for regular visits, a blood test was performed before and after treatments at the 3rd, 6th, 9th and 12th months. We recorded the serum creatinine and urine ACR data before treatment and every 3 months after prescribing CHM or ARBs and at the end of follow-up or December 2016, whichever came first. Patients with no creatinine or ACR data were excluded. We used propensity score matching to make the follow-up period similar in both groups.

The primary outcomes were the change in the eGFR calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) formula. The primary outcomes were the changes in the eGFR and ACR during follow-up. eGFR changes were calculated as follows: [eGFR (end of follow-up or December 2016) - the eGFR (baseline)].²⁰ We also calibrated the changes by performing the following calculation: [the eGFR (end of follow-up or December 2016) - the eGFR (baseline)] divided by the eGFR (baseline). ACR changes were calculated as follows: [the ACR (end of follow-up or December 2016) – the ACR (baseline)]. Blood urea nitrogen (BUN) and potassium levels before and after taking ARBs or CHM were recorded.

Demographic characteristics, including age, sex, body mass index, hemoglobin A1c (HbA1c), and comorbidities, were considered covariates in the analysis. Comorbidities, including coronary artery disease (CAD), hypertension (HTN), dyslipidemia, cerebrovascular accident (CVA), and systemic lupus erythematosus (SLE), were recorded. CAD was defined as the presence of ≥1 vessel with ≥50 % luminal diameter narrowing.²¹ HTN was defined as an average systolic blood pressure (SBP) ≥ 140 mmHg, an average diastolic blood pressure (DBP) ≥ 90 mmHg or the current use of blood pressure-lowering medication.²² Dyslipidemia was defined as the presence of one or more of the following conditions: a triglyceride level ≥200 mg/dL, total cholesterol level ≥240 mg/dL, high-density lipoprotein cholesterol level <40 mg/dL, low-density lipoprotein cholesterol level ≥100 mg/dL, or the current use of antilipemic agents.²⁰ CVA was defined as acute global or focal neurological deficits caused by occlusion of an artery or hemorrhage of cerebral circulation for more than 24 h. The diagnosis of CVA was confirmed by a neurologist and computed tomography (CT) or magnetic resonance imaging (MRI).²³ SLE was diagnosed by the attending rheumatologists based on at least 4 of the 11 diagnostic criteria of the American College of Rheumatology classification criteria for SLE.²⁴

Licensed TCM doctors prescribed CHM products based on the syndrome differentiation of classical TCM theory. We analyzed the top five most common prescriptions. GMP pharmaceutical factories assure the safety and quality of CHM products without aristolochic acid.

2.2. Ethical considerations

The investigation was carried out following the latest version of the Declaration of Helsinki. The research ethics committee of China Medical University Hospital in Taiwan approved this study (CMUH106-REC2-152). All of the hospital-based datasets were deidentified, and the review board waived the need for informed consent.

2.3. Statistical analysis

Frequencies are shown for all descriptive statistics, percentages are shown for categorical variables, and standard deviations are shown for continuous variables. eGFR and ACR changes between CHM and ARB users were analyzed using the Mann-Whitney U test. The association of eGFR and ACR changes during follow-up was analyzed using linear regression with adjustments for confounding variables, including age, sex, HbA1c level (baseline and after treatment), and glucose level (baseline and after treatment). All analyses were performed using R version 3.6.2 (R Foundation for Statistical Computing, Vienna, Austria). A p < 0.05 was considered statistically significant.

3. Results

We screened 2113 DKD patients treated at China Medical University Hospital from 2007 to 2016, including 1000 women and 1113 men (Fig. 1). A total of 1492 (70.6 %) patients took ARBs, 349 (16.5 %) patients took ARBs and CHM alternately or simultaneously, and 272 (12.9 %) patients took only CHM. After excluding patients who did not meet the inclusion criteria, a total of 732 DKD patients, including 71 CHM users and 661 ARB users, were included for further analysis. The baseline characteristics of the CHM and ARB users were different (Table 1). CHM users were younger, had a shorter follow-up duration, lower BMI, better eGFR values, and lower rates of albuminuria and were less likely to have HTN, dyslipidemia, and CVA than ARB users.

Fig. 1.

Study population flowchart. A total of 2113 newly diagnosed patients with diabetic kidney disease (DKD) were screened. After excluding patients who did not meet the inclusion criteria, a total of 732 DKD patients, including 661 ARB users and 71 CHM users, were included for further analysis. Propensity score matching by age, sex, BMI, follow-up duration, and HbA1c level was applied to ultimately include ARB users (n = 355) and CHM-only users (n = 71).

Table 1.

Baseline characteristics of the patients with diabetic kidney disease.

Characteristics	ARB users (N = 661)		CHM users (N = 71)		p
Age (years)	72.2	12.8	68.2	11.5	0.012
Female, No. (%)	317	48	34	47.9	1.000
Follow-up (years)	3.6	1.8–5.2	2.2	1.5–4.2	0.001
BMI kg/m2	26.4	4.6	24.5	3.4	0.001
SBP	132	21	133	25	0.37
DBP	82	12	81	10	0.68
HbA1c	8.2	2.0	8.0	1.9	0.365
BUN (mg/dl)	18	12	16	9	0.162
Creatinine (mg/dl)	1.2	0.9	0.9	0.5	0.009
eGFR (ml/min/1.73 m2)	71.5	30.6	90.9	29.0	< 0.001
ACR (μg/mg)	649	1706	182	622	0.022
Potassium (meq/L)	4.0	0.5	4.1	0.7	0.082
Comorbidity
CAD	163	24.7	27	38.0	0.021
HTN	584	88.4	32	45.1	< 0.001
Dyslipidemia	345	52.2	3	4.2	< 0.001
CVA	92	13.9	4	5.6	0.075
SLE	1	0.2	13	18.3	< 0.001

Data are expressed as the mean (SD) unless the number (%) is indicated. Duration is expressed as the median and interquartile range.

Abbreviations: SBP: systolic blood pressure, DBP: diastolic blood pressure, ARB: angiotensin receptor blocker; CHM: Chinese herbal medicine; BMI: body mass index; BUN: blood urea nitrogen; eGFR: estimated glomerular filtration rate; ACR: urine albumin–creatinine ratio; CAD: coronary artery disease; HTN: hypertension; CVA: cerebral vascular accident; SLE: systemic lupus erythematosus; CHF: congestive heart failure.

A propensity score matching approach was considered because the baseline demographic characteristics of the CHM group were quite different from those of the ARB group. Propensity score matching at a 1:5 ratio by age, sex, BMI, follow-up duration, and HbA1c level was applied. After matching, the age, BMI, follow-up duration, and HbA1c level of the CHM group (n = 71) were not different from those of the ARB group (n = 355) (Table 2). The mean age of the ARB group was 70.5 ± 12.6 y/o and that of the CHM group was 68.2 ± 11.5 y/o. The mean BMI of the ARB group was 24.8 ± 3.7 kg/m² and that of the CHM group was 24.5 ± 3.4 kg/m² (p = 0.478). There was no difference in HbA1c levels between the ARB and CHM group (8.4 ± 2.1 % and 8.0 ± 1.9 %, p = 0.183).

Table 2.

Baseline characteristics of the patients with diabetic kidney disease matched for age, follow-up, body mass index, and HbA1c level.

Characteristics	ARB users (N = 355)		CHM users (N = 71)		P
Age (years)	70.5	±12.6	68.2	±11.5	0.165
Female, No. (%)	171	48.2	34	47.9	1.000
Follow-up (years)	2.4	1.6–4.3	2.2	1.5–4.2	0.082
BMI kg/m2	24.8	±3.7	24.5	±3.4	0.478
SBP	132	25	133	25	0.31
DBP	80	9	81	10	0.84
HbA1c	8.4	±2.1	8.0	±1.9	0.183
BUN (mg/dl)	18	±13	16	±9	0.369
Creatinine (mg/dl)	1.1	±0.9	0.9	±0.5	0.112
eGFR (ml/min/1.73 m2)	80.8	±31.4	90.9	±29.0	0.013
ACR (μg/mg)	607	±1635	182	±622	0.031
Potassium (meq/L)	4.0	±0.6	4.1	±0.7	0.586
Comorbidity
CAD	113	31.8	27	38.0	0.381
HTN	281	79.2	32	45.1	<0.001
Dyslipidemia	194	54.6	3	4.2	<0.001
CVA	48	13.5	4	5.6	0.098
SLE	1	0.3	13	18.3	<0.001

Data are expressed as the mean (SD) or number (%), with the median and interquartile range expressed for duration.

Abbreviations: SBP: systolic blood pressure, DBP: diastolic blood pressure, ARB: angiotensin receptor blocker; CHM: Chinese herbal medicine; BMI: body mass index; BUN: blood urea nitrogen; eGFR: estimated glomerular filtration rate; ACR: urine albumin creatinine ratio; CAD: coronary artery disease; HTN: hypertension; CVA: cerebral vascular accident; SLE: systemic lupus erythematosus.

After treatment, the mean BUN level of patients taking CHM was 22 ± 16 mg/dl (Table 3), which was lower than that (30 ± 28 mg/dl) of patients taking ARBs (p = 0.025). The mean creatinine level of patients taking CHM was 0.9 ± 0.4 mg/dl, which was lower than that (1.7 ± 2.0 mg/dl) of patients taking ARBs (p = 0.003). The mean eGFR after treatment was 84.9 ± 28.1 ml/min/1.73 m² in patients taking CHMs, which was higher than that (67.8 ± 35.4 ml/min/1.73 m²) of patients taking ARBs. The mean change in the eGFR was −6.0 ± 21.4 ml/min/1.73 m² in patients taking CHM and −12.9 ± 24.8 ml/min/1.73 m² in those taking ARBs (p = 0.029). The changes in the eGFR divided by the baseline eGFR were not different between the two groups of patients (p = 0.08). The ACR values and changes in the ACR were also comparable (p = −0.130 and p = 0.862, respectively).

Table 3.

Changes in the laboratory data of the patients with diabetic kidney disease after treatment.

Parameters	ARB (n = 355)		CHM (n = 71)		p
BUN (mg/dl)	30	±28	22	±16	0.025
Creatinine (mg/dl)	1.7	±2.0	0.9	±0.4	0.003
eGFR (ml/min/1.73 m2)	67.8	±35.4	84.9	±28.1	<0.001
Changes in eGFR (ml/min/1.73 m2)	−12.9	±24.8	−6.0	±21.4	0.029
Changes in eGFR/baseline eGFR (%)	−0.14	±0.53	−0.03	±0.27	0.080
Potassium (meq/L)	4.3	±0.7	4.0	±0.7	<0.001
ACR (μg/mg)	808	±2636	326	±1005	0.130
Changes in ACR	201	±2665	145	±1209	0.862

Abbreviations: ACR: urine albumin–creatinine ratio; PCR: urine protein–creatinine ratio; BUN: blood urea nitrogen; eGFR: estimated glomerular filtration rate.

Data are expressed as the mean (SD) unless % is indicated.

Linear regression with adjustments for age, sex, BMI, baseline eGFR, and HbA1c level was used to explore the association of eGFR with time. The decline in the eGFR of the CHM and ARB users was not different during the 10-year follow-up (Fig. 2A, p = 0.62). The decline in eGFR/baseline eGFR was comparable between CHM and ARB groups (Fig. 2B, p = 0.86). The urine ACR was not different between the CHM and ARB groups (Fig. 3, p = 0.73), and it increased with time.

Fig. 2.

Estimated glomerular filtration rate (eGFR) during follow-up. (A) Linear regression plot with the 95 % confidence interval for changes in the eGFR during 10 years of follow-up with adjustments for age, sex, BMI, baseline eGFR, blood pressure, and HbA1c level (p = 0.62). (B) Linear regression plot with the 95 % confidence interval for changes in the eGFR/baseline eGFR during ten years of follow-up with adjustments for age, sex, BMI, baseline eGFR, and HbA1c level (p = 0.86).

Fig. 3.

Changes in the urine albumin–creatinine ratio (ACR) during follow-up. Linear regression plot with the 95 % confidence interval for changes in the ACR during ten years of follow-up with adjustments for age, sex, BMI, blood pressure, baseline eGFR, and HbA1c level (p = 0.73).

The top 5 frequently prescribed CHM by TCM doctors for treating DKD outpatients are shown in Table 4, Table 5. San-Qi (Radix notoginseng) was the most frequently prescribed single herb, followed by Dan-Shen (Radix Salviae), Suan-Zao-Ren (Semen Zizyphi Spinosae), Shan-Yao (Rhizoma Dioscoreae Oppositae), and Tian-Hua-Fen (Radix Trichosanthis) (Table 4). Bai-Hu-Jia-Ren-Shen-Tang (BHJRST) was the most frequently prescribed herbal formula, followed by Ji-Sheng-Shen-Qi-Wan (JSSQW), Si–Ni-San (SNS), Ma-Zi-Ren-Wan (MZRW), and Zhi-Bai-Di-Huang-Wan (ZBDHW).

Table 4.

Commonly prescribed single herbs for patients with diabetic kidney disease.

Single herbs		Functional classification according to TCM theory	Frequency	Average daily dose (gm/day)
English name (Chinese Materia Medica name)	Botanical name
San-Qi (Radix Notoginseng)	Panax notoginseng (Burkill) F.H. Chen	dispelling stasis, quickening the blood, and stanching bleeding	439	1.57
Dan-Shen (Radix Salviae)	Miltiorrhizae Salvia miltiorrhiza Bunge	dispelling stasis and quickening the blood (mild action)	425	1.11
Suan-Zao-Ren (Semen Zizyphi Spinosae)	Ziziphus jujuba var. spinosa (Bunge) Hu ex H.F. Chow	nourishing the heart and quieting the spirit	242	1.87
Shan-Yao (Rhizoma Dioscoreae Oppositae)	Dioscorea oppoita THUNB.	supplementing qi and enriching yin	238	1.69
Tian-Hua-Fen (Radix Trichosanthis)	Trichosanthes kirilowii Maxim	clearing heat and enriching yin	238	1.33

Abbreviations: CHM: Chinese herbal medicine; TCM: traditional Chinese medicine.

Table 5.

Commonly prescribed Chinese herbal formulas for patients with diabetic kidney disease.

Chinese herbal formula		Functional classification according to TCM theory	Frequency	Average daily dose (gm/day)
English name	Ingredient herbs (Chinese Materia Medica name)
Bai-Hu-Jia-Ren-Shen-Tang (BHJRST)	Rhizoma Anemarrhenae, Gypsum Fibrosum, Radix Glycyrrhizae praeparata cum melle, Semen Oryzae, Radix Ginseng	clearing qi-aspect heat	306	4.08
Ji-Sheng-Shen-Qi-Wan (JSSQW)	Radix Rehmanniae Praeparatae, Fructus Cornus, Rhizoma Dioscorea, Poria, Coetex Moutan, Rhizoma Alismatis, Radix Aconiti Lateralis preparata, Cortex Cinnamomi, Radix Achyranthis bidentatae, Semen Plantagines	supplementing yang and disinhibiting water	279	3.37
Si–Ni-San (SNS)	Bupleuri Radix, Glycyrrhizae Radix, Paeoniae Radix Alba, Fructus Aurantii immaturus	coursing depression and rectifying qi	203	3.24
Ma-Zi-Ren-Wan (MZRW)	Bupleuri Radix, Glycyrrhizae Radix, Paeoniae Radix Alba, Fructus Aurantii immaturus	moist precipitation	188	1.76
Zhi-Bai-Di-Huang-Wan (ZBDHW)	Radix Rehmanniae Praeparatae, Fructus Cornus, Rhizoma Dioscorea, Poria, Coetex Moutan, Rhizoma Alismatis, Rhizoma Anemarrhenae, Cortex Phellodendri	enriching yin and clearing heat	161	3.75

Abbreviations: TCM: traditional Chinese medicine.

4. Discussion

CHM and ARBs resulted in a similar decline in the eGFR and urine ACR in a 10-year follow-up of a hospital-based cohort of DKD patients. As no randomized controlled trial has directly compared ARBs and CHM (all trials compared CHM + ARBs and ARBs), this real-world evidence reveals the effectiveness of CHM for treating renal dysfunction. We used propensity score matching because the characteristics of the patients taking CHM were different from those of patients taking ARBs. We found that approximately 30 % of the diabetes mellitus (DM) patients took CHMs. A total of 16.5 % of patients took ARBs and CHM alternately or simultaneously, and 12.9 % took only CHMs. Due to the high density of TCM physicians in central Taiwan,²⁵ the utilization rate of CHM in this study was higher than that in previous studies.^13,26,27

After propensity score matching, the percentage of patients with hypertension and dyslipidemia was higher in the ARB group (Table 2). Because TCM physicians have limited privileges to conduct laboratory investigations, we suspected that these comorbidities might be underdiagnosed in patients taking CHM. Hyperkalemia is a common side effect of ARBs and is one of the significant concerns of CHM treatment.^{28, 29, 30} There was no significant hyperkalemia found in our cohort. CHM may provide an alternative treatment for DKD without increasing the risk of hyperkalemia.

Aristolochic acid (AA)-containing herbs are associated with urothelial malignancies and renal interstitial fibrosis.^{31, 32, 33} Moreover, there is a dose-dependent relationship between the consumption of AA-containing herbs and the risk of developing end-stage renal disease.³⁴ Most AA-containing herbal products were banned by the Ministry of Health and Welfare, Taiwan, in 2003.³⁵ We used a cohort from 2007, which enabled us to ensure that all of the CHMs were free of AA-containing herbs. According to a previous study, type 2 DM patients taking adjunctive CHM had a decreased risk of kidney failure.³⁶ After adjusting for confounding covariates, the end-stage renal disease risk of the adjunctive Liu-Wei-Di-Huang-Wan group was up to 30 % lower than that of the non-TCM group. Another hospital-based study showed that the eGFR of chronic glomerulonephritis patients might not worsen within 12 months while receiving CHM treatments.³⁷ We observed eGFR changes during the 10-year follow-up in the CHM group, and the declines in the eGFR were not different from those of the ARB group.

The symptoms of diabetes include thirst and frequent urination. Chinese herbal formulas such as BHJRST are most commonly used to treat thirst in type 2 DM patients.¹³ BHJRST consists of Radix Ginseng, which is known to improve glucose tolerance.^38,39 BHJRST not only works for hyperglycemia but also has a lipid-lowering effect.^40,41 The essential constituents in JSSQW and ZBDHW are the same as those in LWDHW. LWDHW may improve renal function by decreasing fasting blood glucose levels and ACR values in mice.^42,43 JSSQW removes excessive fluid from the body and may improve long-term survival rates in CKD patients.¹⁸ BHJRST, the derivative formulas of LWDHW, and LWDHW itself are the three crucial herbal formulas prescribed by TCM doctors to treat DKD.

San-Qi (Radix notoginseng) and Dan-Shen (Radix Salviae) were the most frequently prescribed single herbs in this study. Previous studies have shown that San-Qi (Radix notoginseng) inhibits the Wnt/β-catenin signaling pathway and prevents epithelial-mesenchymal transformation.⁴⁴ Therefore, San-Qi (Radix notoginseng) protects podocytes against apoptosis.⁴⁵ The combination of San-Qi (Radix notoginseng) with standard medications is associated with decreased albuminuria and proteinuria levels.⁴⁶ Dan-Shen (Radix Salviae) is well known for its antiplatelet aggregation, antifibrotic, and antioxidative stress effects in patients with cardiovascular diseases and stroke. It has also been shown that the active component of Dan-Shen (Radix Salviae) may inhibit urinary protein excretion⁴⁷ and reduce renal interstitial fibrosis.⁴⁸ The clinical findings are illustrated in the graphical abstract. A cartoon summarizing the potential effects and mechanisms of how the ARB/ACEI and CHM work to protect the kidney in DKD is shown in Fig. 4.

Fig. 4.

A cartoon summarizing the potential effects and mechanisms of how the ARB/ACEI and CHM work to protect the kidney in DKD.

There were some potential limitations in this study. First, the data collected in this study were limited to a single medical center in Taiwan. We cannot ensure that CHM-only users did not receive ARBs/ACEIs at other hospitals. We manually checked these 71 CHM users' electronic medical records; no such documentation of ARB/ACEI use was found in the records. In the future, linking in-house electronic medical records and National Insurance Research Database records may minimize bias. Second, due to the nature of this retrospective study, we performed propensity score matching. The matching reduced the sample size for statistical analysis. Third, all patients enrolled in this study were Taiwanese. Therefore, the application of CHM to other populations needs further study. Fourth, although data on many clinical indicators, such as BMI, HbA1c levels, and comorbidities, were collected for analysis, detailed information about the patients’ lifestyles was not available. Finally, the exact mechanisms of the CHMs used in this study require further pharmacological investigations. Based on the findings with these real-world data, we are planning to conduct a clinical trial and investigate the pharmacological mechanisms of these CHMs.

5. Conclusion

This propensity score-matched cohort study suggests that CHM might have a similar effect as ARBs on albuminuria and renal function in early DKD patients. Further pharmacological investigations and clinical trials are warranted.

Ethics approval and consent to participate

The data collection and analysis complied with the Declaration of Helsinki. The Research Ethics Committee of China Medical University Hospital in Taiwan approved this study (CMUH106-REC2-152) and waived the need for informed consent.

Consent for publication

Not applicable.

Availability of data and material

The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.

Funding

This work was financially supported by China Medical University Hospital, Taichung, Taiwan (DMR-107-149). This study was also supported in part by the “Chinese Medicine Research Center, China Medical University” from the Featured Areas Research Center Program within the framework of the Higher Education Sprout Project by Ministry of Education (MOE) in Taiwan (CMRC-CENTER-2) and the “Construction of Precision Medicine in Chinese Medicine” project by the Ministry of Health and Welfare, Taiwan (M09G1047 and M1007013). None of the funders or institutions listed had a role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.

Authors' contributions

YCL participated in the design of the study and drafted the manuscript. YCL, MYW, YCH, CYC and HRY analyzed and interpreted the data. CYC and HRY supervised the project, contributed to the conception and design of the study, and finalized the manuscript. CYC and HRY contributed equally as co-corresponding authors. All the authors have read and approved the final manuscript.

Declaration of competing interest

The authors declare that they have no competing interests.