Clinical randomized controlled trial and network pharmacological analysis of blood glucose regulation by Korean red ginseng in patients with type 2 diabetes and impaired glucose regulation

Abstract

Background

This study aimed to evaluate the clinical efficacy and safety of Korean red ginseng through a randomized controlled trial and explore the potential targets and pathways for Korean red ginseng to exert clinical efficacy by combining network pharmacology and clinical trial results.

Methods

We included randomly allocated participants to a test group that received red ginseng or a placebo control group. Network pharmacology was used to analyze the targets of Korean red ginseng and the genes related to type II diabetes, with the core targets being inferred. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes enrichment analyses were conducted to explore the potential mechanisms of action of red ginseng.

Results

Regarding safety, none of the participants showed adverse effects. The test group exhibited significant post-intervention reductions in fasting blood glucose (from 6.44 ± 1.17 to 6.19 ± 1.14 mmol/L), 2-h blood glucose (from 11.96 ± 4.40 to 10.48 ± 3.72 mmol/L), and HbA1c (from 6.27 ± 0.79 to 6.12 ± 0.81 %; all P < 0.01). Immune markers, including CD4/CD8, IgG, IgA, and IgM, remained within normal limits in both groups. Further hierarchical analysis revealed that red ginseng significantly increased IgA levels and natural killer cells in patients with type 2 diabetes with low immunity, with a significant between-group difference. Network pharmacology identified effective components of Korean red ginseng for treating type II diabetes, including ginsenoside Rb1, ginsenoside-Rg3, and Ginsenoside Rb2, with potential targets such as toll-like receptors 2 and 4, tumor necrosis factor, etc.

Conclusion

According to the Evaluation Method of Auxiliary Blood Glucose Lowering Function of health products (2012 Version), we can conclude that RG capsules can help reduce blood glucose levels in individuals with Type 2 diabetes and Impaired Glucose Regulation.

Graphical abstract

1. Introduction

Type 2 diabetes mellitus is a chronic metabolic disease characterized by persistent hyperglycemia resulting from impaired insulin secretion. 90 % of diabetics have type 2 diabetes. Impaired glucose regulation (IGR), which is termed as prediabetes, is an intermediate high-risk condition between normal blood glucose and diabetes hyperglycemia, with patients having an increased risk of T2DM. IGR comprises two states: impaired fasting glucose (IFG) and impaired glucose tolerance (IGT).According to the International Diabetes Federation, the estimated global prevalence of diabetes among people aged 20–79 years is 10.5 % (536.6 million people) in 2021, which is expected to increase to 12.2 % (783.2 million people) in 2045 [1]. The prevention and control of diabetes have become major public health issues of global concern.

Red ginseng (RG) is a valuable medicinal agent with a history of >1000 years; moreover, it is widely used in China, South Korea, Japan, and other Asian countries. The main components in RG are ginsenosides. Additionally, it contains sugars, volatile oils, and other components. Korean red ginseng (KRG) originates from various parts of Korea and is widely used in East Asia [2]. KRG has been shown to exert anti-inflammatory, antioxidant, anti-tumor, anti-depression, anti-asthma, and other effects; further, it is widely used to treat diabetes, cancer, inflammation, nervous system diseases, cardiovascular diseases, and hyperlipidemia [[3], [4], [5], [6], [7]].

Given the above background, JungKwanJang Red Ginseng capsules (with G1899(P) Korean Ginseng Powder (6); Korea Ginseng Corporation, Daejeon, Korea; a safe natural substance) have been prepared to improve glycemic control. This study was conducted to assess its efficacy and safety on blood glucose levels in patients with T2DM and IGR and explore its potential mechanism by combining network pharmacology and clinical trial results.

2. Materials and methods

2.1. Clinical research

Ethical statement

This trial was approved by the ethics committee of the Tianjin First Central Hospital (Ethical Approval Number: 2021N067KY) and conducted in accordance with the Declaration of Helsinki. This study was registered with Chinese Clinical Trail Registry (ChiCTR2100055018).

2.1.1. Experimental products

We obtained JungKwanJang Red Ginseng capsules from Korea Ginseng Corporation. Regarding the production process, raw ginseng materials were washed, steamed, dried, crushed, and sieved (180 mesh) to obtain G1899(P) Korean Ginseng Powder (6), which was then poured into the capsules. Each capsule weighed 0.465 g, with the total saponin content being 3.5–4.8 g/100 g (Batch number: 20210528; Shelf life: 36 months). Intervention samples were subjected to hygiene and acute oral toxicity testing, and the test results were qualified. Supplementary data S1 shows the composition of components of G1899(P) Korean Ginseng Powder (6) used in this experiment. The placebo capsule comprised silicon dioxide, lactose, magnesium stearate, allure red, brilliant blue, lemon yellow, microcrystalline cellulose, and food flavoring essence. All ingredients added to the placebo comply with the National Food Safety Standard for the Use of Food Additives (GB2760). The placebo capsule lacked functional ingredients and it was like the test capsule in terms of taste, packaging, appearance, and dosage form.

2.1.2. Participants

2.1.2.1. Inclusion criteria

• •Patients with T2DM
• •Having stable condition after diet control or oral hypoglycemic treatment, no changes in the drug regimen or dosage
• •Only taking maintenance doses (i.e., fasting blood glucose ≥7 mmol/L (126 mg/dl) and 5.6–7 mmol/L [100–126 mg/dl] or 2-h postprandial blood glucose ≥11.1 mmol/L (200 mg/dl) and 7.8–11.1 mmol/L [140–200 mg/dl] in patients with T2DM and IGR, respectively).

2.1.2.2. Exclusion criteria

• •Patients with type I diabetes
• •Patients aged <18 years or >65 years
• •Pregnant or breastfeeding women
• •Patients allergic to the test sample
• •Patients with major organ complications such as heart, liver, and kidney diseases, or other serious diseases.
• •Patients with mental illness
• •Patients taking glucocorticoids or other drugs that affect blood glucose levels
• •Patients who could not cooperate in terms of the controlled diet
• •Patients with diabetic ketosis, acidosis, or infection within the past 3 months
• •Patients who were immunocompromised due to taking immunosuppressive drugs
• •Short-term use of products that may have affected the results
• •Patients who did not adhere to the study protocol or with incomplete data
• •Patients who were considered unsuitable for trial participation by the researchers for any other reason
• •Patients participating in other clinical trials.

2.1.3. Experimental design and grouping

A randomized-blinded method was used to allocate participants into the test and control groups, which were matched according to several potentially confounding factors, including glycosylated serum protein or glycosylated hemoglobin, blood glucose, sex, age, disease course, and medication type (sulfonylureas and biguanides). Each group required a minimum of 50 participants. Specialist statisticians used SPSS software to generate random numbers that could be used to randomize 107 participants. All investigators, participants, and their caregivers were blinded to group assignment throughout the study period.

2.1.4. Treatment dose and duration

Patients in both groups took three capsules twice a day for 90 consecutive days. Prior to the intervention period, each participant received a controlled diet plan that corresponded with their sex, age, labor intensity, and ideal weight, with reference to their original living habits. During the test, participants were asked to maintain the controlled diet and not to change the type and dose of the hypoglycemic drugs they were taking.

2.1.5. Statistical analyses

Continuous data were tested for homogeneity of variance and normality. The t-test was used for normally distributed data with homogeneity of variance. A nonparametric test was used for nonnormally distributed data or heterogeneous variance. The Wilcoxon signed-rank test was used for intra-group comparisons, and the Mann-Whitney U test was used for intergroup comparisons. Differences between the two groups before and after treatment were compared. If normality and homogeneity of variance were satisfied, the two-sample t-test was used; otherwise, the two-sample Mann-Whitney U test was used. Continuous data is expressed as mean (standard deviation). Categorical variables, such as the effective rate, frequencies, and percentiles, were analyzed using the chi-square test to calculate differences between the groups. Data analysis was conducted using SPSS 27.0 (IBM Corp., Armonk, NY). P < 0.05 was considered statistically significant.

2.2. Network pharmacology study on the improvements in patients with T2DM mediated by red ginseng

2.2.1. Screening active components and corresponding targets

The active ingredients of RG were searched based on the Traditional Chinese Medicine Systems Pharmacology (TCMSP) [8], Bioinformatics Analysis Tool for Molecular Mechanism of TCM (BATMAN-TCM) [9], and SymMap v2 [10] databases. The active ingredients were screened based on whether the medicinal ingredients in the database had corresponding targets.

2.2.2. Determination of T2DM-related and common targets with active components

Information regarding the disease-related genes was obtained using the GENECARDs database [11]. Specifically, the database was queried using "type 2 diabetes" as the key word, followed by screening out of targets with a relevance score <10. Additionally, RG targets and T2DM-related genes were analyzed using Venny 2.1 [12] in order to identify common targets for further analysis.

2.2.3. Network Construction of common targets

Protein-protein interaction (PPI) is crucially involved in diverse aspects of life processes, including biological signal transduction, regulation of gene expression, and energy and material metabolism [13]. The STRING database [14] was used to conduct PPI network analysis based on the identified common targets [15]. The species was confined to “Homo sapiens”, with the parameter of moderate confidence exceeding 0.4. The resulting PPI network diagram was imported into the Cytoscape software. Next, the CytoNCA plugin was used to calculate the centrality size of each gene in the RG-T2DM PPI network with moderate centrality. The size of the gene graph was divided based on the centrality value, with drug- and disease-related genes being distinguished using different colors. The MCODE plugin was used to select the core subgroups of the drug-disease action network, with genes with the highest scores being labelled as "key cluster" genes.

2.2.4. GO function and KEGG pathway enrichment analyses

Enrichment analysis primarily comprised Gene Ontology (GO) functional enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment. Prior to analysis using the Metascape platform [16], we imported the common targets into the “Multiple Gene List” and subsequently clicked “Custom Analysis.” GO functional enrichment analysis encompassed biological processes (BP), cellular components (CC), and molecular functions (MF) [15]. Additionally, among the outcomes of the KEGG enrichment analysis, we selected the top 20 signaling pathways markedly correlated with the drug-disease interaction [17].

2.2.5. Construction of the “active components-gene target-disease” network

Prior to analysis, we created two Excel files named “Network” and “Type”, respectively [18], and successively imported them into Cytoscape 3.10.1 software. Finally, we successfully the “active components-gene target-disease” network [19] using Cytoscape.

3. Results

3.1. Clinical research results

3.1.1. Baseline characteristics

After recruitment and screening, we randomly assigned 107 participants to the test (n = 53) and control (n = 54) groups. During the trial, six participants (two and four in the test and control groups, respectively) dropped out for reasons unrelated to the product. Table 1 presents the baseline characteristics of the 101 participants. There were no significant between-group differences at baseline in terms of age, sex, fasting blood glucose, HbA1c, or drug use (P > 0.05) (Table 1).

Table 1.

Baseline characteristics ($\bar{x}$ ±SD).

	Test group (n = 51)	Control group (n = 50)
Male/Female	13/38	16/34
Age (years old)	60.61 ± 5.47	58.86 ± 6.43
Course of disease(years)	4.25 ± 1.18	4.26 ± 1.32
FBG(mmol/L)	6.44 ± 1.17	6.43 ± 2.23
2hPG(mmol/L)	11.96 ± 4.40	11.56 ± 5.96
HbA1c(%)	6.27 ± 0.79	6.30 ± 1.47
Unmedicated subjects	4	5
Sulfonylureas	8	7
Biguanides	8	9
α-Glucosidase inhibitor	4	5
Biguanides + α-Glucosidase inhibitor	6	7
Biguanides + Sulfonylureas	11	9
Biguanides + Sulfonylureas + α-Glucosidase inhibitor	10	8

Inter-group comparison P > 0.05.

3.1.2. Observation results of safety indicators

As shown in Table 2, the blood pressure; heart rate; routine blood, urine, and stool test results; and liver and kidney function indicators of the participants remained within the normal range before and after the intervention period. Throughout the trial period, no adverse or allergic reactions, including nausea, flatulence, diarrhea, or abdominal pain, were observed.

Table 2.

Changes of safety indicators before and after the test ($\bar{x}$ ±SD).

	Test group (n = 51)			Control group (n = 50)
	Before test	After test	decrease	Before test	After test	decrease
safety indicators
Systolic pressure (mmHg)	132.37 ± 17.39	131.43 ± 17.07		129.30 ± 15.21	127.88 ± 14.99
Diastolic pressure (mmHg)	73.82 ± 9.55	72.92 ± 9.26		72.36 ± 12.46	72.14 ± 11.30
Heart rate (times/minutes)	75.63 ± 9.58	75.80 ± 9.46		73.04 ± 11.85	72.96 ± 11.48
RBC( × 1012/L)	4.66 ± 0.51	4.66 ± 0.52		4.63 ± 0.43	4.70 ± 0.40
WBC( × 109/L)	5.69 ± 1.31	5.64 ± 1.32		6.08 ± 1.59	5.80 ± 1.21
PLT( × 109/L)	241.96 ± 55.10	243.04 ± 57.59		241.42 ± 50.62	246.62 ± 48.50
HGB(g/L)	141.31 ± 13.84	140.00 ± 14.24		137.66 ± 12.38	138.66 ± 11.73
TP(g/L)	72.50 ± 4.68	72.15 ± 4.68		71.15 ± 3.42	71.87 ± 3.07
Alb(g/L)	47.49 ± 2.13	47.35 ± 2.93		46.51 ± 1.93	46.96 ± 2.67
ALT(U/L)	24.41 ± 21.32	21.66 ± 15.79		16.38 ± 6.76	17.92 ± 10.42
AST(U/L)	21.77 ± 11.76	19.75 ± 12.09		17.11 ± 3.66	18.69 ± 6.41
Urea(mmol/L)	5.32 ± 1.41	5.49 ± 1.35		5.11 ± 1.20	5.26 ± 1.30
Cre(μmol/L)	64.14 ± 15.79	66.92 ± 15.16		62.04 ± 11.88	64.60 ± 10.84
Urine routine(except for urine sugar)	Normal	Normal		Normal	Normal
Stool routine	Normal	Normal		Normal	Normal
efficacy indicators
FBG (mmol/L)	6.44 ± 1.17	6.19 ± 1.14∗	0.25 ± 0.74	6.43 ± 2.23	6.49 ± 2.37	−0.06 ± 0.67
0.5hPBG (mmol/L)	10.86 ± 2.19	10.64 ± 2.12		10.53 ± 3.69	10.64 ± 3.29
1hPBG (mmol/L)	12.68 ± 3.62	12.28 ± 3.21		11.31 ± 5.35	11.90 ± 4.54
2hPBG (mmol/L)	11.96 ± 4.40	10.48 ± 3.72∗∗	1.48 ± 2.21##	11.56 ± 5.96	11.31 ± 6.11	0.25 ± 1.45
AUC-OGTT (mmol/L × h)	22.53 ± 5.97	21.32 ± 5.17∗∗	–	21.14 ± 9.10	21.52 ± 8.25	–
HbA1c (%)	6.27 ± 0.79	6.12 ± 0.81∗∗	0.16 ± 0.36#	6.30 ± 1.47	6.45 ± 1.52	−0.09 ± 0.78
Fasting insulin (uIU/mL)	12.29 ± 10.30	13.39 ± 11.21∗	–	9.59 ± 4.79	10.09 ± 4.46	–
DPP-4(U/L)	114.46 ± 22.52	107.29 ± 24.72	–	107.19 ± 24.40	107.31 ± 28.78	–
TC (mmol/L)	5.33 ± 1.27	5.09 ± 1.09∗	–	5.21 ± 0.83	5.35 ± 0.96	–
TG (mmol/L)	1.76 ± 1.22	1.92 ± 1.20	–	1.54 ± 0.80	1.78 ± 1.14	–
CD4/CD8	1.99 ± 0.84	1.96 ± 0.90	–	1.94 ± 1.06	1.97 ± 1.02	–
IgA (g/L)	2.69 ± 1.04	2.69 ± 1.16	–	2.86 ± 1.03	2.96 ± 0.96	–
IgG (g/L)	13.21 ± 2.60	13.12 ± 2.68	–	13.34 ± 2.49	13.39 ± 2.60	–
IgM (g/L)	0.99 ± 0.37	1.03 ± 0.44	–	0.98 ± 0.44	1.03 ± 0.52	–
NK Cell activity(%)	24.88 ± 4.09	23.06 ± 9.31	–	24.65 ± 4.25	22.48 ± 9.20	–
Neutrophils ( × 109/L)	3.10 ± 0.87	3.08 ± 0.87	–	3.49 ± 1.19	3.28 ± 0.97	–
Macrophage (monocyte) ( × 109/L)	0.39 ± 0.12	0.39 ± 0.12	–	0.44 ± 0.16	0.42 ± 0.12	–
Comprehensive score	68.25 ± 6.78	71.80 ± 7.07∗∗##		68.28 ± 8.64	67.70 ± 8.39
Individual perception	22.20 ± 2.76	23.25 ± 2.54∗∗#		22.30 ± 3.59	22.02 ± 3.32
Psychological feelings	22.55 ± 3.33	23.69 ± 3.32∗∗#		22.00 ± 3.80	21.98 ± 3.64
Physiological sensation	23.51 ± 3.17	24.86 ± 3.37∗∗		23.98 ± 4.00	23.70 ± 3.89

In safety indicators,self-comparison and inter-group comparison P > 0.05; in efficacy indicators,Self-comparison ∗P < 0.05 ∗∗P < 0.01; inter-group comparison #P < 0 0.05 ##P < 0.01.

3.1.3. Observation results of efficacy indicators

As shown in Table 2, there was a post-intervention decrease in fasting plasma glucose levels in the test group (P < 0.05). Moreover, 0.5-h and 1-h postprandial blood glucose levels showed a post-intervention non-significant decreasing trend in the test group. The 2-h postprandial blood glucose levels in the test group showed a significant post-intervention decrease (1.48 ± 2.21, P < 0.01), with a significant between-group difference in the percentage decrease (P < 0.05). Additionally, there was a significant between-group difference in the percentage decrease in 1-h postprandial blood glucose levels (P < 0.01). The area under the curve of the oral glucose tolerance test (AUC-OGTT) showed a significant decrease in the test group (P < 0.01) but not in the control group. The HbA1c level showed a significant post-intervention decrease (0.16 ± 0.36) in the test group (P < 0.01), with a significant between-group difference in the percentage decrease (P < 0.01). There was a significant post-intervention decrease in fasting insulin levels in the test group (P < 0.05), with no significant between-group difference (P > 0.05). There was no significant post-intervention change in DPP-4 levels in the test group and no significant between-group difference (both P > 0.05). Furthermore, the total cholesterol level showed a significant post-intervention decrease in the test group (P < 0.05), with no significant between-group difference (P > 0.05). There was no significant post-intervention change in triglyceride levels in the test group (P > 0.05). As shown in Table 2, there was no significant post-intervention change in humoral immunity (IgG, IgA, and IgM), cellular immunity (CD4/CD8), and innate immunity (natural killer [NK] cells, neutrophils, and monocytes) in the test group (P > 0.05).

The test group showed a significant post-intervention improvement in the individual overall feeling, physical feeling, and psychological feeling (P < 0.01); however, this was not observed in the control group (P > 0.05) (Table 2).

3.2. Results of the network pharmacology study

3.2.1. Common targets between “active ingredients” and “type 2 diabetes”

We obtained 21 and 140 active compounds of RG from TCMSP and BATMAN-TCM databases, respectively. After removing the duplicate components and components without corresponding targets, we identified 117 active ingredients. Additionally, we identified 128 corresponding gene symbols. Furthermore, 7413 gene symbols related to the prevention of chronic nephritis were acquired from the GENECARDs database. Venn analysis was used to identify the common gene targets of “red ginseng” and “type 2 diabetes” (Fig. 1[A]). The results showed that RG could regulate 93 related targets in its treatment of T2DM.

Fig. 1.

(A)Venn diagram of drug targets and disease common targets.(B)Protein-protein interaction (PPI) network.(C)the keycluster genes selected by MCODE method.

3.2.2. PPI network and key cluster genes

We uploaded 93 common targets to the STRING website in order to determine mutual interactions. As shown in Fig. 1(B), a PPI network diagram was generated by adjusting these parameters. In this diagram, the nodes serve as symbols for proteins, and the edges act as indications for protein-protein connections. It can be seen that there are 93 nodes and 964 edges, and the mean node degree is 20.7. Using the MCODE plug-in of Cytoscape, the highest-scoring gene network subgroup was identified, which contained 21 key cluster genes (Fig. 1(C)). Further analysis of the 21 key cluster genes found that the degree value, closeness centrality, and betweenness centrality of the eight protein targets toll-like receptor (TLR) 2, Fos, caspase-1, CD80, TLR4, tumor necrosis factor (TNF), interleukin (IL) 6, and IL1-β were higher than the median values. Accordingly, these eight protein targets were considered potential targets for RG intervention in T2DM.

3.2.3. GO and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analyses

We imported 21 key cluster genes into Metascape, and the results of KEGG pathway analysis as well as GO-MF, GO-BP, and GO-CC analyses were obtained individually. The findings indicated that RG could regulate T2DM through several signaling pathways, including lipid and atherosclerosis; tuberculosis, leishmaniasis, TLR signaling pathway; cancer-related pathways; PI3K-Akt signaling pathway; hepatitis C; allograft rejection; small-cell lung cancer; and human T-cell leukemia virus 1 infection (Fig. 2[A]). Combined with the results of the PPI network analysis, our findings suggested that the Toll-like receptor signaling pathway may be among the main pathways of action of RG intervention in T2DM. Notably, GO-BP analysis showed that RG regulates biological processes to treat T2DM, including response to lipopolysaccharides, inflammatory response, regulation of neuroinflammatory response, regulation of epithelial cell apoptotic process, positive regulation of acute inflammatory response, and lymphocyte activation. GO-MF analysis revealed that RG regulated the molecular functions of cytokine receptor binding, cysteine-type endopeptidase activity, apoptotic signaling pathway, protein domain-specific binding, phosphatase binding, and G protein-coupled receptor binding (Fig. 2). GO-CC analysis indicated that the membrane, membrane raft, plasma membrane protein complex, cell body, and organelle outer membrane were involved in the therapeutic effects of RG (Fig. 2).

Fig. 2.

Enrichment analysis of the keycluster genes.(A)KEGG enrichment analysis for signal pathways.(B) biological process, molecular function and cellular component of enrichment analysis.

3.2.4. “Chemical components-target genes-signaling pathways” network

To elucidate the inherent relationship among the chemical components, key targets, and signaling pathways of RG, we used the Cytoscape 3.10.1 software to construct the “chemical components - target genes - signaling pathways” network (Fig. 3).

Fig. 3.

The diagram of Ingredients-Target-Pathway.

4. Discussion

Our clinical results demonstrate that supplementation with KRG over 12 weeks can improve glucose control among individuals with IFG, IGT, or T2DM. Subjects consuming 2.79 g of KRG capsules per day attained significant decreases in serum glucose at 2 h during a 75-g OGTT, fasting blood glucose over the intervention period relative to the placebo group. In addition, there was a significant post-intervention increase in fasting insulin levels in the test group but not in the placebo group. Vuksan et al. [20] reported glucose and insulin regulation effects of KRG in well-controlled T2DM. In this clinical trial, 12 weeks of supplementation with the selected KRG treatment improved plasma glucose levels (decreased 75-g OGTT PG indices by 8–11 %) and plasma insulin (decreased fasting and 75-g OGTT PI indices by 33–38 % and increased insulin sensitivity index by 33 %) in subjects with well-controlled T2DM. Bang et al. [21] reported that supplementation with KRG over 12 weeks can improve glucose control among individuals with IFG, IGT, or newly diagnosed T2DM. Subjects consuming 5 g of KRG capsules per day attained significant decreases in fasting blood glucose, serum glucose at 2 h during a 75-g OGTT over the intervention period relative to the placebo group. There was a significant post-intervention increase in fasting insulin levels in the test group but not in the placebo group. The results of these indicators are consistent with our experimental findings. The observed absolute reduction in fasting blood glucose, fasting insulin levels are considered clinically meaningful in the pre-diabetic and diabetic population, as they are significantly associated with type 2 diabetes. But in our study,the glycosylated hemoglobin (HbA1c) levels in the test group showed a significant post-intervention decrease by 2.43 % (P < 0.01), with a significant between-group difference (P < 0.01). Contrastingly, Bang et al. [21]. and Vuksan et al. [20] both observed no significant change in glycosylated hemoglobin following consumption of RG daily for 12 weeks (P > 0.05). This discrepancy may be attributed to between-study differences in the inclusion/exclusion criteria, dosages, Subject size and ingredient in RG capsules. Some of our subjects had HbA1c levels above 6.5 % before participating in the trial. In addition, due to different processes, our red ginseng capsules may contain rarer ginsenosides such as Rg3, Rk1, Rg5, CK, Rk3, and Rh4. Compared with Ginsenosides Re, Rg1, Rd, and Rb1, these rare ginsenosides have relatively few glycosyl groups and increased hydrophobicity, exhibit relatively high pharmacological activity [22].

Lipid metabolism disorder is a primary pathological change in diabetes and its complications and contributes to glucose metabolism disorder in T2DM. Therefore, regulation of lipid metabolism is crucial for improving blood glucose levels. In the present study, the total cholesterol levels showed a significant post-intervention decrease in the test group (P < 0.05) but not in the placebo group. There was no significant post-intervention change in triglyceride levels in both groups (P > 0.05). Shin et al. [23] found that RG extract (200 mg/kg) reduced the expression of lipid metabolism-related genes (sterol regulatory element binding protein 1c, peroxisome proliferator-activated receptor γ [PPAR γ], fatty acid synthase, stearoyl-CoA desaturase 1, acetyl CoA carboxylase 1), decreased triacylglycerol and total cholesterol levels, and regulated lipid metabolism in HFD-induced T2DM mice over 6 weeks.

Diabetes, which is often accompanied by overeating, hunger, emaciation, and high blood sugar levels, increases the risk of infection and may reduce immunity. RG enhances immunity [[24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35]], with improvements in NK cell activity, cellular immune function, and thymus and spleen indices in mice [26]. Alam et al. found that KRG boosts T and NK cell functions in pigs [31]. We observed no significant changes in humoral immunity (IgG, IgA, and IgM), cellular immunity (CD4/CD8), or innate immunity (NK cells activity, neutrophils, and monocytes) in either group. Subjective health scores showed significant improvements in comprehensive, personal, psychological, and physiological perceptions in the test, but not control, group. Supervisor health scores indicated that RG improved health perception and mental state by lowering blood sugar levels and enhancing the quality of life (Table 2). The lack of improvement in the objective immune indicators suggests that most participants had normal immune function at baseline. We performed further analysis including participants with subjective health scores below 68, which indicates low immunity. Among 33 patients who were identified as immunodeficient (18 and 15 in the test and control groups, respectively), the test group showed significant increases in IgA levels and NK cell activity (P < 0.01), whereas the placebo group did not show significant changes (P > 0.05) (Table 3). These findings may suggest that the immunomodulatory effects of KRG in a pre-diabetic, non-immunocompromised population are subtle and require further investigation in larger, targeted studies.

Table 3.

Changes in blood glucose related indicators, insulin sensitive factors, and immune indicators in immunocompromised people before and after the test($\bar{x}$ ±SD).

	Test group (n = 18)		Control group (n = 15)
	Before test	After test	Before test	After test
CD4/CD8	2.04 ± 0.71	1.97 ± 0.83	2.11 ± 0.56	2.03 ± 0.35
IgA(g/L)	2.89 ± 1.24	3.12 ± 1.33∗	3.28 ± 1.22	3.22 ± 0.91
IgG(g/L)	12.65 ± 2.72	13.12 ± 2.82	12.89 ± 2.83	13.07 ± 2.39
IgM(g/L)	1.01 ± 0.35	1.03 ± 0.37	0.95 ± 0.42	0.99 ± 0.44
NK Cell activity (%)	26.13 ± 4.94	30.58 ± 8.08∗∗##	25.45 ± 4.40	21.19 ± 9.02
Neutrophils ( × 109/L)	2.74 ± 0.64	2.83 ± 0.56	3.46 ± 1.49	3.01 ± 1.05
Macrophage (monocyte) ( × 109/L)	0.34 ± 0.10	0.34 ± 0.09	0.43 ± 0.15	0.40 ± 0.10

Comparison within groups∗P < 0.05 ∗∗P < 0.01, comparison between groups##P < 0.01.

To explore the potential mechanism through which RG improves immune deficiency in patients with T2DM, we performed network pharmacology analysis using the TCMSP, BATMAN-TCM, and SYMMAP databases. We identified 117 active ingredients in RG, with the main active ingredient being ginsenosides, including ginsenosides Rb1, Rg3, and Rb2. These constituents target multiple pathways and exert therapeutic effects in T2DM.

Ginsenoside Rb1 is a protopanaxadiol-type saponin predominantly found in Panax ginseng, Panax quinquefolius, and Panax notoginseng [28]. Although its intestinal absorption is relatively low, Rb1 undergoes deglycosylation by gut microbiota and is sequentially hydrolyzed into the secondary metabolite Compound K, thereby enhancing its bioactivity [28]. Studies have demonstrated that Rb1 significantly improves insulin sensitivity by activating the AMPK signaling pathway, thereby suppressing hepatic lipid accumulation and inflammatory responses [29]. Additionally, Rb1 exhibited similar anti-diabetic effects in db/db obese mice, which were attributed to reduced hepatic lipid accumulation and inhibition of adipocyte lipolysis [30]. Given the close relationship between insulin resistance and inflammation, Rb1 ameliorates metabolic disorders associated with obesity and type 2 diabetes mellitus (T2DM) by reducing the production of inflammatory cytokines in the liver and adipose tissue, thereby maintaining glucose and lipid homeostasis. Further research revealed that Rb1 alleviates insulin resistance by suppressing endoplasmic reticulum (ER) stress-mediated inflammasome activation, leading to decreased secretion of inflammatory cytokines [31]. Moreover, Rb1 has demonstrated significant antioxidant and anti-inflammatory effects in various T2DM-related disease models, particularly in skeletal muscle [[32], [33], [34]]. Since both type 1 and type 2 diabetes are characterized by progressive pancreatic β-cell dysfunction, Rb1 mitigates high glucose-induced β-cell apoptosis by inhibiting nitric oxide (NO) production and caspase-3 expression, offering a potential therapeutic strategy for T2DM [35,36].

Ginsenoside Rg3 is primarily found in Panax ginseng, with higher concentrations in Korean red ginseng (KRG) but lower levels in naturally occurring ginseng plants [37,38]. The anti-tumor efficacy of Rg3 has been extensively studied in various cancer types, while its pharmacological benefits in anti-inflammation, antioxidation, anti-aging, and neuroprotection have also been increasingly recognized. In recent years, the therapeutic potential of Rg3 in metabolic syndrome-related diseases, such as metabolic dysfunction-associated steatotic liver disease, obesity, and diabetes, has garnered significant attention. Research indicates that Rg3 significantly enhances glucose uptake in mature 3T3-L1 adipocytes by upregulating the transcription of GLUT4 and IRS-1, as well as increasing PI3K-110α protein levels [38]. Furthermore, Rg3 directly binds to PPARγ in adipocytes, promoting adiponectin secretion and activating adiponectin signaling, thereby alleviating hyperglycemia, hyperlipidemia, and abnormal lipid accumulation [39]. In C2C12 myotubes, Rg3 stimulates insulin signaling by enhancing IRS-1 phosphorylation under both basal and insulin-stimulated conditions, independent of AMPK activation [40]. Additionally, Rg3 significantly enhances insulin secretion in hamster pancreatic HIT-T15b cells and streptozotocin-induced diabetic mice [41].

Ginsenoside Rb2 is an active protopanaxadiol-type saponin widely distributed in various parts of ginseng, with higher concentrations in Panax ginseng and Panax quinquefolius but minimal levels in Panax japonicus and Panax notoginseng [42]. Rb2 is predominantly concentrated in the stems, leaves, and berries of ginseng. Studies have shown that Rb2 significantly ameliorates hyperglycemia, obesity, and insulin resistance by regulating glucose metabolism, body weight, insulin sensitivity, and lipid accumulation, thereby offering a potential therapeutic strategy for diabetes and its complications [43]. Rb2 has demonstrated notable effects on weight reduction and glucose tolerance improvement in animal models [[44], [45], [46]]. In vitro human islet models, 0.5 μg/mL Rb2 significantly increased insulin secretion and promoted β-cell migration [47]. In HFD-induced obese mice, Rb2 treatment markedly decreased body fat weight, improved glucose metabolism, and increased energy expenditure [45,47]. In H4IIE cells, Rb2 attenuates palmitate-stimulated glucose production by upregulating AMPK and its downstream small heterodimer partner (SHP) signaling pathways, thereby inhibiting the expression of gluconeogenic enzymes [48].

Inflammatory factors affect energy intake, storage, and metabolism; interfere with the physiological effects of insulin; and lead to insulin antagonism. TNF-α stimulates hyperglycemic hormone secretion, promotes lipolysis, inhibits hepatocyte-insulin binding, and disrupts insulin signal transduction. IL-6 inhibits insulin-induced lipid and protein synthesis as well as glucose transport and therefore reduces glucose transporter expression. MCP-1 inhibits insulin-mediated glucose uptake and insulin receptor tyrosine phosphorylation. Pathological conditions activate upstream pathways, releasing large quantities of pro-inflammatory factors, such as TNF and IL-6, and resulting in insulin antagonism. This reduces glucose utilization, increases blood glucose levels, and exacerbates inflammatory reactions, which creates a vicious cycle. Combining the results of network pharmacology research and clinical trials, we speculate that the TLR signaling pathway is one of the important signaling pathways of Red Ginseng Capsule to improve the abnormal blood glucose in type 2 diabetes patients and people with glucose injury. Molecular docking suggested that ginsenoside Rb1 interacts with SER120 and VAL93 on TLR4, disrupting its dimerization and blocking TLR activation [49]. TLR2, which is a core target in the TLR pathway, regulates energy substrate utilization, tissue inflammation, dietary lipid intake, and glucose homeostasis. High glucose levels activate TLR expression and cytokine release through NF-κB in vitro and in vivo [50]. Kuo et al. [51] found that TLR2 deficiency reduced local inflammatory cytokine expression and enhanced liver insulin action; moreover, they identified TLR2 as a mediator of liver inflammation and insulin resistance. Ginsenosides inhibits TLR2 activity and activates downstream pathways and therefore reduces the release of pro-inflammatory cytokines. The ginsenosides in red ginseng capsules can competitively inhibit the binding of endotoxins and TLR4, thereby reducing the chronic low-grade inflammatory state of the body, decreasing the production of pro-inflammatory factors such as IL-1 β, blocking the activation of JNK, and inhibiting IRS-1 phosphorylation, alleviating insulin antagonistic symptoms; in addition, inhibition of pro-inflammatory factors can alleviate pancreatic beta cell apoptosis. Inhibiting pro-inflammatory cytokines can also reactivate the function of suppressed NK cells and enhance their vitality.

There are some limitations in this study. Firstly, the sample size of immunocompromised individuals is relatively small, and the results of NK cell activity have certain limitations. A further limitation is the absence of a long-term follow-up period, which means the sustainability of the observed metabolic benefits beyond the 90-day intervention remains unknown. In addition, there is a lack of experimental validation for the analysis results and speculated potential mechanisms of network pharmacology. We will further explore these in future research.

5. Conclusion

In conclusion, this randomized controlled trial demonstrates that 90-day supplementation with G1899(P) Korean Ginseng Powder (6) is safe and effective in improving glycemic control, as evidenced by significant reductions in fasting blood glucose, 2-h postprandial glucose, and HbA1c in patients with T2DM and IGR. An immunomodulatory effect was observed in a subset of patients with low baseline immunity.

Based on the results of network pharmacology analysis and the changes of IgA levels and NK cell activity in people with low immunity, we speculate that RG capsule mediated treatment of type 2 diabetes involves multiple potential targets, which may be related to TLR2, TLR4, IL-6 and TNF. Ginsenosides and other active ingredients in RG capsules may improve the state of chronic inflammation of the body by inhibiting the activation of TLR signaling pathway and the release of inflammatory factors, thereby reducing insulin antagonism and alleviating the symptoms of type 2 diabetes. However, further large-scale, multicenter, prospective, randomized, and controlled studies are needed to validate the inferences of network pharmacology.

Declaration of competing interest

All authors have no conflicts of interest to declare.

Appendix A. Supplementary data

The following is the Supplementary data to this article.