Ginseng supplementation and vascular function: a systematic review and meta-analysis of clinical trials

Abstract

Background

Ginseng is a source of pharmacologically active compounds with cardioprotective properties. The aim of the current systematic review and meta-analysis was to investigate the effect of ginseng on vascular function.

Methods

Electronic databases, including PubMed, ISI Web of Science, Scopus, and Google Scholar were searched using relevant keywords from inception until February 2025. The eligibility criteria for a full-text review of papers consisted of clinical trials involving subjects aged ≥ 15 years who consumed ginseng and evaluated at least one of the cardiac function markers: flow-mediated dilation (FMD), pulse wave velocity (PWV), augmentation index (AIx), and circulating biomarkers of vascular function.

Results

From a total of 328 initial search records, 13 randomized controlled trials were included in the study. Pooled analysis showed a significant increase in FMD level (effect size: 5, SMD: 0.571%, 95% CI: 0.198, 0.943, P = 0.003), and endothelium-derived nitric oxide levels (effect size: 4, SMD: 0.30, 95% CI: 0.01, 0.59, P = 0.045) as well as a significant decrease in PWV (effect size: 7, SMD: ‒0.29 cm/s, 95% CI: ‒0.51, ‒0.06, P = 0.014). No significant effect was observed in AIx (effect size: 9, SMD: ‒0.53%, 95% CI: ‒1.08, 0.03, P = 0.063) following ginseng intake.

Conclusions

The overall results revealed beneficial effects of ginseng on endothelial function and arterial stiffness, as measured by PWV, while it did not change AIx. Further studies are recommended to investigate the effects of ginseng on vascular function.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12906-025-04936-5.

Introduction

According to the World Health Organization (WHO), cardiovascular disease (CVD), as the leading cause of mortality worldwide, accounted for 32% of all deaths in 2019 [1]. Vascular dysfunction plays a key role in the pathogenesis of CVD [2]. Vascular hemostasis is regulated by endothelium-derived factors, including endothelium-derived nitric oxide (eNO) and prostacyclin as vasodilator agents, as well as thromboxane A2 and endothelin-1 (ET-1) as vasoconstrictor substances [3]. An imbalance between endothelium-derived vasodilator and vasoconstrictor agents accelerates cardiovascular disorders. Endothelium-derived NO synthase (eNOS) is an enzyme that produces eNO through the arginine-citrulline cycle [4]. Endothelium-derived NO has multiple physiological actions, including the relaxation of vascular smooth muscle cells [5], the inhibition of leukocyte adhesion and platelet aggregation [6–8], and the prevention of vascular inflammation [9].

Nutraceutical compounds have been the focus of recent research due to their high tolerance and safety. Ginseng, a medicinal herb, has been associated with reducing cardiometabolic risk factors such as insulin resistance, hypertension, and dyslipidemia [10–12]. The Asian pharmacopeia considers ginseng to be an “all-healing” tonic due to its exceptional chemical profile [13]. The two major species of ginseng are Panax ginseng C.A. Meyer, referred to as Asian ginseng, and Panax quinquefolius L., known as American ginseng [14]. Among the numerous bioactive compounds in ginseng, triterpenoid saponins, known as ginsenosides, are mainly responsible for its pharmacological effects [15, 16]. Based on experimental studies, ginseng can increase eNO production, prevent endothelial cell apoptosis, and inhibit angiotensin-converting enzyme (ACE), consequently improving endothelial function [17, 18]. Clinical evidence also reveals that ginseng may decrease the risk of cardiovascular events by lowering vascular stiffness in both the central aorta and peripheral arteries [19–22]. Arterial stiffness, which is measured by the augmentation index (AIx) and the pulse wave velocity (PWV), is an independent risk predictor for cardiovascular events [23].

There are some systematic reviews and meta-analyses that evaluated the effect of ginseng on blood pressure [24, 25], while none of these studies assessed vascular function parameters, such as endothelial function and arterial stiffness. The results from a meta-analysis of randomized controlled trials (RCTs) showed a lowering effect of ginseng supplementation on systolic blood pressure (SBP) and diastolic blood pressure (DBP) [25]. Another meta-analysis also showed beneficial effects of Korean ginseng on SBP and DBP [24]. Many studies support the potential protective effects of ginseng on vascular function, although researchers have not yet reached a consensus. In the present study, we aimed to systematically review all available clinical trials assessing the effect of ginseng on endothelial function, arterial stiffness, and selected circulatory biomarkers of vascular function.

Methods

Eligibility criteria

The review protocol has been registered in the PROSPERO database of Systematic Reviews (registration number: CRD42023395145). Human studies investigating the effects of ginseng on functional measures of arterial function, such as endothelium-dependent vasodilation, endothelium-independent vasodilation, arterial stiffness, and vascular tone, as well as selected circulatory biomarkers of vascular function, are included in this systematic review and meta-analysis. The selection criteria were: clinical trials on subjects aged ≥ 15 years, with no limit on health status, in which the intervention group was supplemented with ginseng (as pure powder, extract, or fresh form); clinical trials with a control group in which the only difference between the groups was the consumption of ginseng; and trials with sufficient data for baseline and endpoint values of at least one of flow-mediated dilation (FMD), PWV, AIx, and serum levels of eNO, intercellular adhesion molecule-1 (ICAM-1), vascular adhesion molecule-1 (VCAM-1), ET-1, and E-selectin. There were no restrictions regarding language, publication year, and participants’ conditions. Only those outcomes that had at least three relevant studies were included in this systematic review and meta-analysis. The exclusion criteria were as follows: non-clinical studies; trials in which ginseng was supplemented in combination with other herbs or interventions; trials with no appropriate control group; and trials that lacked sufficient information on the baseline or duration of the intervention.

Search strategy and data extraction

The study selection and construction of the flow diagram were conducted in accordance with the guidelines of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [26]. We used the PICOS approach to design this systematic review, as follows:

• Population: Human populations;

• Intervention: Ginseng supplementation;

• Comparator: A placebo or active control, with the sole distinction between the groups being the consumption of ginseng;

• Outcomes: Primary outcomes were direct measures of vascular function, including FMD, PWV, and AIx. The secondary outcomes were circulatory levels of eNO, ICAM-1, VCAM-1, ET-1, and E-selectin;

• Setting: Clinical trials.

We performed electronic searches in the following databases: PubMed, ISI Web of Science, Scopus, and Google Scholar, using relevant keywords and terms from inception to February 2025. Key search terms were “ginseng” OR “Panax” OR “quinquefolius” OR “Ashwagandha” OR “Withania somnifera” OR “Eleutherococcus senticosus” OR “Acanthopanax Senticosus” OR “Eleutherococcus” OR Oplopanax” AND “arterial stiffness” OR “vascular function” OR “endothelial function” OR “flow-mediated dilation” OR “peak wave velocity” OR “carotid thickness” OR “intima-media thickness” OR “carotid plaques” OR “nitrate-mediated dilation” OR “Nitric oxide” OR “IMT” OR “endothelial” OR “endothelium” OR “endothelium-dependent blood flow” OR “vascular resistance” OR “augmentation index” OR “adhesion molecules” OR “adhesion molecule” OR “intercellular adhesion molecule-1” OR “ICAM-1” OR “vascular adhesion molecule-1” OR “VCAM-1” OR “endothelin-1” OR “ET-1” OR “endothelin 1” OR “E-selectin” AND “trial”. The supplementary Table 1 details the search strategy. Screening of the titles and abstracts for included trials was performed independently by two authors (AE and FHS). Papers deemed eligible for full-text review were appraised for eligibility separately by the two authors. The following data were extracted from the included studies: author/date/origin, characteristics of participants, study design, intervention characteristics including dosage and duration of intervention, and study outcomes. At each stage, any disagreements were resolved by discussion within the research team.

Study quality and risk of bias within the studies

The quality of the included studies was assessed using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions [27]. This 7-item instrument encompasses the assessment of random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, selective outcome reporting, incomplete outcome data, and other sources of bias. Interpretation of the scale was conducted at the study level, whereby each study was classified as either high risk of bias (high risk of bias for one or more key domains), unclear risk of bias (unclear risk of bias for one or more key domains), or low risk of bias (low risk of bias for all key domains).

Statistical analysis

Mean changes in study outcomes were used as primary endpoints to evaluate the differences between the intervention and comparison groups. Where the changes in outcomes were assessed at multiple time points, only data at the longest duration of follow-up were extracted for meta-analysis. In articles where the outcomes were not represented in the main text or tables, the data were extracted from the graphs using the online “WebPlotDigitizer” tool. We computed mean differences (MD) and 95% confidence intervals (95% CI) for trials with sufficient data to perform statistical analyses in STATA software version 17. The random effects model was used for estimating pooled effects. The effect sizes for FMD, PWV, AIx, and eNO in each trial were determined as the standardized mean difference (SMD) and their 95% CI. Pooled SMDs and their 95% CIs are presented in forest plots to evaluate the overall effects of ginseng compared to controls. Sensitivity analyses were used to test the robustness of the analyses using the “one-study-removed” approach. Inter-study heterogeneity was examined using the Cochran Q test (considered significant for P values < 0.10) and I-squared (I²) index, with values of < 25, 25 − 50, and > 50 to exhibit low, moderate, and high heterogeneity, respectively [28]. Subgroup analyses were used to explore heterogeneity based on dosage, intervention duration, participants’ age, and health status. Publication bias was tested by visual inspection of funnel plots, as well as Begg’s and Egger’s regression tests. In addition, the trim-and-fill analysis was used to adjust for any significant publication bias.

Results

As Fig. 1 shows, our search yielded 328 publications, of which 40 duplicate articles were removed, 261 of the remaining 288 records were excluded based on their titles and abstracts. Finally, 27 full-text articles were collected and further screened, and 13 studies were included in the analysis.

Fig. 1.

Flow diagram of the literature search and study selection process

Characteristics of included studies

The characteristics of the included studies are described in Table 1. These studies were conducted in Canada [21, 29–33], Korea [19, 20, 34–36], Croatia [22], and Taiwan [37]. Many participants were described as healthy volunteers [20, 21, 29, 31], while others included subjects with prehypertension or hypertension [22, 34, 36], metabolic syndrome [35] or T2DM [32, 33], stable coronary artery disease [19], hemodialysis [37], or combined hypertension and T2DM [30]. Seven trials were of crossover design [19, 21, 29, 31–33, 37] and six were parallel studies [20, 22, 30, 34–36]. A total of 774 participants aged 30–64 years were included in the meta-analysis from all 13 RCTs. The daily dosages ranged from 0.21 to 6 g/d, and the duration of supplementation ranged from 180 min to 12 weeks. Korean ginseng (Panax Ginseng), American ginseng (Panax Quinquefolius L.), and Siberian ginseng (Acanthopanax Senticosus or Eleutherococcus Senticosus) were used in the trials as hydroalcoholic extracts, powders, or fresh forms. Of the 13 studies, six were included in the analysis more than once: two compared various dosages of supplements; one assessed the impact of ginseng separately on heart-femoral PWV and brachial-ankle PWV; two used various species of ginseng as interventions, and one evaluated the effect of ginseng in populations with different conditions. Only six [20, 30, 32, 34, 35, 37] out of 13 studies assessed intervention compliance rates. Nine studies evaluated adverse effects during the study period [20, 29, 31–37].

Table 1.

Overview of the characteristics and main findings of the clinical trials included in the systematic review

Author/date	Country	Study design	Study participants	Nutritional intervention			Main outcomes
				Groups	Dose of supplement	Duration
Chung, et al./ 2010 [19]	Korea	Crossover RCT	20 Patients with stable angina and ≥ 50% stenosis in at least one coronary artery; Mean age: 62.4 ± 3.1 yrs; BMI: 25.5 ± 0.8 kg/m2	Powder of Korean red ginseng Vs placebo	2.7 g/d (900 mg three times daily)	10 weeks	Heart femoral PWV and brachial ankle PWV
Jovanovski, et al./ 2010 [21]	Canada	Crossover RCT	17 Healthy individuals; Mean age: 30 ± 9 yrs; BMI: 25 ± 3 kg/m2.	Dried ground Korean red ginseng rootlet part of root Vs Hydroalcoholic extract of Korean red ginseng rootlet ginsenosides Vs Hydroalcoholic extract of Korean red ginseng rootlet polysaccharides Vs placebo	Korean red ginseng root: 3 g/d (500 mg six times daily) Korean red ginseng ginsenosides extract: 1.22 g/d Korean red ginseng polysaccharides extract: 0.21 g/d	180 min	AIx
Chen, et al./ 2012 [37]	Taiwan	Crossover RCT	18 Intradialytic and HTN; Mean age: 56.2 ± 12.7 yrs; BMI: 24.1 ± 4.2 kg/m2 20 Intradialytic and no HTN; Mean age: 49.3 ± 11.5 yrs; BMI: 23.1 ± 6 kg/m2.	Fresh slices of Korean red ginseng Vs no treatment	3.5 g at each hemodialysis session for twelve consecutive hemodialysis treatments)	4 weeks	eNO
Park, et al./ 2012 [35]	Korea	Parallel RCT	48 Participants with metabolic syndrome; Mean age: 43.1 ± 10.6 yrs in ginseng group and 46.2 ± 11 yrs in controls; BMI: Not reported	Powder of Korean red ginseng root Vs placebo	4.5 g/d (1.5 g three times daily)	12 weeks	PWV
Rhee, et al./ 2012 [36]	Korea	Parallel RCT	64 Participants with hypertension; Mean age: 55 ± 9 yrs in ginseng group and 58 ± 6 yrs in controls; BMI: 24.9 ± 2.4 kg/m2 in ginseng group and 24.7 ± 2.6 kg/m2 in controls.	Korean red ginseng powder Vs placebo	3 g/d	12 weeks	PWV
Mucalo, et al./ 2013 [22]	Croatia	Parallel RCT	64 Participants with hypertension; Mean age: 62.1 ± 8.8 yrs in ginseng group and 63.9 ± 10.9 yrs in controls; BMI: 33.4 ± 5.6 kg/m2 in ginseng group and 29.9 ± 4.9 kg/m2 in controls.	Hydroalcoholic extract of American ginseng root Vs placebo	3 g/d (two 500 mg capsules three times daily)	12 weeks	AIx
Jovanovski, et al./ 2014 [31]	Canada	Crossover RCT	16 Healthy individuals; Mean age: 30 ± 9 yrs; BMI: 25 ± 3 kg/m2.	Dried ground Korean red ginseng rootlet part of root Vs Hydroalcoholic extract of Korean red ginseng rootlet ginsenosides Vs Hydroalcoholic extract of Korean red ginseng rootlet polysaccharides Vs placebo	Korean red ginseng root: 3 g/d (500 mg six times daily) Korean red ginseng ginsenosides extract: 1.22 g/d Korean red ginseng polysaccharides extract: 0.21 g/d	180 min	FMD
Jovanovski, et al./ 2014 [29]	Canada	Crossover RCT	23 Healthy individuals; Mean age: 25 ± 2 yrs; BMI: 22 ± 0.6 kg/m2.	Hydroalcoholic extract of Korean red ginseng root Vs placebo	400 mg/d	180 min	AIx
Shishtar, et al./ 2014 [33]	Canada	Crossover RCT	25 Participants with T2DM; Mean age: 63 ± 9 yrs; BMI: 29.3 ± 4.3 kg/m2.	Powder of Korean white ginseng root Vs placebo	1 g/d 3 g/d 6 g/d	4 h	AIx
Cha, et al./ 2016 [34]	Korea	Parallel RCT	62 Prehypertensive individuals; Mean age: 42.7 ± 12.6 yrs in ginseng group and 41.4 ± 10.7 yrs in controls; BMI: 24.3 ± 2.80 kg/m2 in ginseng group and 24.6 ± 2.87 kg/m2 in controls.	Powder of Korean red ginseng Vs placebo	5 g/d (10 capsules containing 500 mg ginseng)	12 weeks	eNO
Vuksan, et al./ 2018 [32]	Canada	Crossover RCT	24 Participants with T2DM; Mean age: 64 ± 7 yrs; BMI: 27.8 ± 4.6 kg/m2.	Hydroalcoholic extract of American ginseng root	3 g/d (Three times daily)	8 weeks	eNO
Jovanovski, et al./ 2020 [30]	Canada	Parallel RCT	80 Patients with hypertension and T2DM; Mean age: 59.4 ± 7.4 yrs in ginseng group and 60.5 ± 6.9 yrs in controls; BMI: 28.6 ± 3.4 kg/m2 in ginseng group and 29.6 ± 4.3 kg/m2 in controls.	Hydroalcoholic extract of combined American ginseng root and Hydroalcoholic extract of Rg3-enriched Korean ginseng extract Vs placebo	2.25 g/d (Three times daily)	12 weeks	AIx PWV
Oh, et al./ 2020 [20]	Korea	Parallel RCT	76 Healthy individuals; Mean age: 45 ± 1.6 yrs in high-dose ginseng group, 43.6 ± 1.7 yrs in low-dose ginseng group, and 46.8 ± 1.8 yrs in placebo group; BMI: 24.8 ± 0.5 kg/m2 in high-dose ginseng group, 25.4 ± 0.6 kg/m2 in low-dose ginseng group, and 25.4 ± 0.6 kg/m2 in placebo group.	Water extract of dried fruit of Siberian ginseng	500 mg/d 1000 mg/d (Two capsules two times daily)	12 weeks	FMD PWV

Abbreviations: AIx, augmentation index; BMI, body mass index; eNO, endothelial-derived nitric oxide; FMD, flow-mediated dilation; HTN, hypertension; PWV, pulse wave velocity; RCT, randomized controlled trial; T2DM, type 2 diabetes mellitus

The values for age and BMI were presented as mean ± standard deviation

Meta-analysis

Effect of ginseng on FMD and serum level of eNO

A total of five effect sizes, including 140 subjects, provided data on the effect of ginseng on FMD. The quantitative meta-analysis showed a significant increase in FMD in subjects receiving ginseng compared with controls (Fig. 2; SMD: 0.571%, 95% CI: 0.198, 0.943, P = 0.003). Considering the I² index (38.9%) and the Cochrane Q test (P = 0.162), there was non-significant moderate heterogeneity among the trials. Subgroup analyses based on intervention type, intervention duration, and country could not change the overall effect of ginseng on FMD (Table 2). A sensitivity analysis indicated that removing each of the studies had no significant effect on the pooled effect size. A visual inspection of the funnel plot as well as Egger’s (P = 0.033) test revealed evidence of asymmetry, while Begg’s test showed no evidence of publication bias among trials (P = 0.086). The trim-and-fill method that imputed the effect size by one study did not change the pooled effect size (SMD: 0.442%, 95% CI: 0.018, 0.866). Four effect sizes from three studies were included in the meta-analysis of the effect of ginseng on serum level of eNO, which showed a significant increment in eNO level (Fig. 3; SMD: 0.30, 95% CI: 0.01, 0.59, P = 0.045). There was no heterogeneity between the studies (I² index = 0.0% and P for Cochrane Q test = 0.624). Nevertheless, the result was unstable following removal of each study from the analysis, as excluding each of the studies conducted by Cha et al. (SMD: 0.25%, 95% CI: ‒0.10, 0.60) or Vuksan et al. (SMD: 0.21%, 95% CI: ‒0.12, 0.54) shifted the pooled effect size to an insignificant level. Although visual inspection of the funnel plot of the study precision indicated an asymmetry, the results from Egger’s (P = 0.363) and Begg’s (P = 0.308) tests did not confirm this finding.

Fig. 2.

Forest plot of the effect of ginseng on flow-mediated dilation (FMD)

Table 2.

Subgroup analyses for the effects of ginseng on FMD, PWV, and AIx in the participants of included studies

	Effect sizes, n	SMD (95% CI)	P-within	P-between	I2	P-heterogeneity
Subgroup analysis for FMD
Overall	5	0.57 (0.19, 0.94)	0.003		38.9	0.162
Intervention duration
Acute	3	0.83 (0.31, 1.36)	0.002	0.102	35.5	0.212
Chronic	2	0.29 (–0.11, 0.68)	0.151		0.0	0.762
Intervention type
Powder	1	1.24 (0.48, 2.00)	0.001	0.052	-	-
extract	4	0.42 (0.11, 0.73)	0.008		0.0	0.431
Country
Korea	2	0.29 (–0.11, 0.68)	0.151	0.102	0.0	0.762
Canada	3	0.83 (0.31, 1.36)	0.002		35.5	0.212
Subgroup analysis for PWV
Overall	7	–0.29 (–0.51, ‒0.06)	0.014		17.9	0.293
Intervention type
Powder	4	–0.27 (–0.59, 0.04)	0.091	0.871	16.6	0.308
extract	3	–0.31 (–0.72, 0.09)	0.130		45.7	0.159
Country
Korea	6	–0.33 (–0.60, ‒0.05)	0.020	0.507	27.5	0.228
Canada	1	–0.15 (–0.59, 0.29)	0.503		-	-
Age
< 60 y	4	–0.25 (–0.63, 0.12)	0.190	0.738	47.4	0.127
≥ 60 y	3	–0.34 (–0.65, ‒0.02)	0.036		0.0	0.500
Health status
Patient	5	–0.23 (–0.47, 0.01)	0.062	0.597	0.0	0.435
Healthy	2	–0.42 (–1.10, 0.26)	0.223		65.7	0.088
Subgroup analysis for AIx
Overall	9	–0.53 (–1.08, 0.03 )	0.063		87.2	< 0.001
Intervention duration
Acute	7	–0.79 (–1.43, − 0.15)	0.015	0.003	85.4	< 0.001
Chronic	2	0.31 (–0.02, 0.64)	0.062		0.0	0.471
Intervention type
Powder	4	–0.43 (–0.91, 0.05)	0.078	0.728	61.9	0.049
extract	5	–0.63 (–1.62, 0.36)	0.213		92.6	< 0.001
Country
Croatia	1	0.45 (–0.05, 0.95)	0.075	0.005	-	-
Canada	8	–0.66 (–1.25, − 0.07)	0.029		86.6	< 0.001
Age				0.073
< 60 y	4	–1.16 (–2.25, − 0.06)	0.039		89.1	< 0.001
≥ 60 y	5	–0.07(–0.53, 0.39)	0.770		74.2	0.004
Health status
Patient	5	–0.07(–0.53, 0.39)	0.770	0.073	74.2	0.004
Healthy	4	–1.16 (–2.25, − 0.06)	0.039		89.1	< 0.001

Abbreviations: AIx, augmentation index; FMD, flow-mediated dilation; PWV, pulse wave velocity; SMD, standardized mean difference

Data are pooled weighted mean differences (95% CIs) by a random-effects model

Fig. 3.

Forest plot of the effect of ginseng on serum level of endothelium-derived nitric oxide (eNO)

Effect of ginseng on PWV

The pooled analysis of seven effect sizes from five clinical trials, with 328 subjects, showed a significant decrease in PWV level following ginseng supplementation (Fig. 4; SMD: ‒0.29 cm/s, 95% CI: ‒0.51, ‒0.06, P = 0.014). The I² index (17.9%) and Cochrane Q test (P = 0.293) indicated low inter-trial heterogeneity. In stratified analyses, there was no evidence of differences between subgroups of studies based on type of intervention, country, participants’ age, and health status (Table 2). Further, a sensitivity analysis showed that the exclusion of one effect size reported by Oh et al. turned the pooled SMD into a nonsignificant result (SMD: ‒0.20 cm/s, 95% CI: ‒0.42, 0.01). The results from Egger’s (P = 0.078) and Begg’s (P = 0.133) tests did not confirm the evidence of publication bias.

Fig. 4.

Forest plot of the effect of ginseng on pulse wave velocity (PWV)

Effect of ginseng on AIx

A total of nine effect sizes from five clinical trials, including 358 subjects, evaluated the effect of ginseng on AIx. The pooled analysis of these studies showed no significant change in AIx level following ginseng supplementation (Fig. 5; SMD: ‒0.53%, 95% CI: ‒1.08, 0.03, P = 0.063). A high inter-trial heterogeneity was detected based on the I² index of 87.2% and P < 0.001. The findings from subgroup analyses based on intervention duration and country revealed significant differences between subgroups. Pooling changes from the trials with acute follow-up and trials using Korean ginseng (conducted in Canada) resulted in a significant decrease in AIx when compared with their counterparts (Table 2). A sensitivity analysis showed that omitting each of the studies conducted by Jovanovski et al. (SMD: ‒0.63%, 95% CI: ‒1.25, ‒0.006) or Mucalo et al. (SMD: ‒0.65%, 95% CI: ‒1.24, ‒0.06) shifted the pooled effect size to the significance level. Furthermore, after the exclusion of one effect size from a clinical trial conducted by Shishtar et al., the reduction in AIx levels remained significant as compared to the main analysis (SMD: ‒0.63%, 95% CI: ‒1.24, ‒0.01). A visual inspection of the funnel plot of the study precision did not show an asymmetry. The results from Egger’s (P = 0.301) and Begg’s (P = 0.175) tests also supported this finding.

Fig. 5.

Forest plot of the effect of ginseng on augmentation index (AIx)

Assessment of the quality and risk of bias

Table 3 indicates the risk of bias in the included studies. Except for one study, the remaining trials provided sufficient data on random sequence generation. A total of 11 studies implemented blinding for the participants, researchers, and outcome assessors to the study protocol. Sufficient data on allocation concealment were reported in six clinical trials. None of the studies exhibited bias regarding selective outcome reporting. Nine clinical trials assessed the completeness of the outcomes, and only three clinical trials appeared to have other sources of bias (between-group analysis without adjustment for baseline values of study outcomes). Based on judgments of the risk of bias for each item, four studies were categorized as high quality, seven studies were classified as “high risk” and two studies as “unclear risk” of bias.

Table 3.

Assessment of risk of bias in the studies included in the meta-analysis

Study	Random sequence generation	Allocation concealment	Blinding of participants and personnel	Blinding of outcome assessment	Incomplete outcome data	Selective outcome reporting	Other sources of bias
Chung, et al. 2010	L	U	L	L	L	L	H
Jovanovski, et al./ 2010	L	U	L	L	L	L	L
Chen, et al./ 2012	H	H	H	H	L	L	H
Park, et al. 2012	L	L	L	L	L	L	L
Rhee, et al. 2012	L	U	L	L	H	L	L
Mucalo, et al. 2013	L	L	L	L	H	L	H
Jovanovski, et al./2014	L	L	L	L	L	L	L
Jovanovski, et al.2014	L	L	L	L	L	L	L
Shishtar, et al. 2014	L	L	L	L	L	L	L
Cha, et al./ 2016	L	H	L	L	L	L	L
Vuksan, et al./ 2018	L	U	L	L	H	L	L
Jovanovski, et al./ 2020	L	U	U	U	L	L	L
Oh, et al. 2020	L	L	L	L	H	L	L

Discussion

In the present meta-analysis, ginseng supplementation significantly improved FMD, PWV, and serum levels of eNO, with no remarkable changes in AIx. Flow-mediated dilation measures eNO-mediated responses in the arteries of the peripheral circulation to assess endothelial function. The increase in FMD following ginseng supplementation was highly robust to sensitivity analysis. Flow-mediated dilation, the main index of endothelial function, reflects the ability of the vascular endothelium to produce eNO. It is worth noting that each 1% increase in FMD can reduce cardiovascular events by more than 10% [38]. Inter-trial heterogeneity for the effect of ginseng supplementation on FMD was small. The results also indicated that the effectiveness of ginseng did not differ in diverse subgroups based on duration and type of ginseng; however, further trials should be conducted to elucidate the effect of ginseng on vascular function in subgroups based on age, duration, comorbidity, and other independent cardiovascular risk factors.

Among the more than 300 ginsenosides found in ginseng, the ginsenosides Rg1, Rg3, Rb1, and Re are the main compounds responsible for the different pharmacological effects of ginseng [39–41]. In preclinical studies, ginsenosides, particularly Rg3, stimulated endothelial eNOS and increased the production of eNO [42]. This increase in eNO level plays a critical role in regulating endothelial function through vasorelaxant effects, inhibition of platelet aggregation and leukocyte adhesion, and prevention of hyperplasia in vascular smooth muscle cells [43–45]. The primary mechanism by which ginsenosides increase eNO levels and enhance endothelium-dependent vasodilation involves the stimulation of the phosphoinositide 3-kinase/protein kinase B-endothelial nitric oxide pathway, as an eNOS inhibitor diminished the vasodilatory activity of ginsenosides [46]. In line with improvements in FMD, our meta-analysis also revealed a significant increase in plasma eNO following ginseng consumption. Furthermore, experimental studies revealed that protopanaxatriol, protopanaxadiol, and ginsenoside Rh2, which are the most potent components of ginseng extract, can inhibit ACE [47], thereby reducing angiotensin II levels, preventing bradykinin degradation, and inducing vasodilatory effects.

In addition to preclinical studies, several meta-analyses of clinical trials indicated the improving effects of ginseng on vascular-related disorders, mainly by stimulating eNO production. The findings from a meta-analysis of 23 RCTs performed in 2022 revealed a significant reduction in SBP (-3.23 mmHg) and DBP (-1.48 mmHg) following ginseng supplementation. This effect was greater in individuals with prehypertension, hypertension, and metabolic syndrome than in individuals with diabetes mellitus, overweight/obesity, and hyperlipidemia, or postmenopausal subjects [25]. The authors attributed the antihypertensive effects of ginseng to its stimulating effects on eNO synthesis. Another meta-analysis of 18 RCTs reported that ginseng-based medicines were more effective in treating angina pectoris compared to nitrates [48]. A meta-analysis of 28 RCTs also revealed a remarkable improvement in left ventricular end-diastolic diameter and volume and cardiac function in patients with acute decompensated heart failure following ginseng consumption [49]. The findings of Ren et al. in their meta-analysis also showed that ginseng consumption decreased nitroglycerin dosage in patients with stable angina pectoris [50].

Arterial stiffness affects the elasticity of the blood vessels and arterial compliance. The rigidity of the arterial walls diminishes arterial compliance, impairing their ability to absorb pulsatile wave propagation. This leads to the early return of the reflected pressure wave back to the left ventricle, resulting in systolic hypertension and CVD [51]. Assessment of arterial stiffness is conducted using PWV and AIx [52, 53]. Using PWV, time intervals between pulse waves were measured to quantify arterial stiffness. The AIx also uses the difference between the second and first systolic peaks to measure arterial stiffness. Several studies have demonstrated beneficial effects of polyphenols/polyphenol-rich foods on arterial stiffness [54–57]. Nevertheless, our results showed significant changes in PWV but not in AIx following ginseng consumption; both of which were less stable to sensitivity analysis. We found substantial heterogeneity between trials assessing the effects of ginseng on AIx, but not for PWV, indicating that intervention status may be the major source of heterogeneity. The stratified analysis showed a lowering effect of ginseng supplementation on AIx in the studies using Korean ginseng and those with an acute duration of intervention. It appears that the studies conducted by Jovanovski et al. [30], Mucalo et al. [22], and Shishtar et al. [33] had the greatest impacts on the results of AIx; after excluding each of them, the results became statistically significant. Both studies conducted by Jovanovski et al. [30] and Mucalo et al. [22] were the only studies in which American ginseng was used. Previous studies have shown that the amount of ginsenoside content in Korean ginseng is about twice that of the ginsenoside content in American ginseng [58]. Moreover, the amount of other pharmacologically active substances, such as non-saponin compounds, phenol compounds, acid polysaccharides, and polyethylene compounds in Korean ginseng is higher than that in American ginseng [58]. In another study by Jovanovski et al., consumption of ginsenoside fraction, but not the polysaccharide fraction, led to a significant reduction in AIx levels [21]. In the study conducted by Shishtar et al., the acute effects of various dosages of Korean white ginseng powder were examined in patients with T2DM [33]. The results revealed a significant reduction in AIx following the consumption of 3 g ginseng, while 6 g of ginseng worsened AIx response [59]. Therefore, ginseng species and dosage may be determining factors in the response of vascular function to ginseng consumption.

Knowledge gaps and future directions

The bioactive compounds of the different species of ginseng used in the included trials varied. In particular, high variability in the ginsenoside content of ginseng varieties, as well as different extraction methods, may lead to varying vascular responses to ginseng. Furthermore, half of the included studies did not evaluate compliance with the intervention. Compliance rates affect estimates of the intervention’s effect and should be measured in future studies. None of the included studies considered confounding factors such as dietary habits, physical activity, body mass index (BMI), and concomitant medication use in the statistical analyses, while all of these factors may influence vascular function and interfere with the effects of ginseng supplementation. There were not enough studies on the other markers of arterial function, including ET-1, a potent vasoconstrictor; ICAM-1; VCAM-1; and E-selectin, which are vascular inflammatory markers. Therefore, further studies are recommended to investigate the effects of ginseng on these biomarkers. Considering the relatively high inter-trial heterogeneity in studies related to AIx, caution should be exercised when interpreting the results. In addition, most studies were conducted in Canada and Korea and this could affect the generalizability of the results to other populations due to the dietary and genetic variations among the populations. Further studies are recommended to evaluate the effectiveness of ginseng on vascular function, especially in populations with CVD, to draw more definitive conclusions.

Conclusion

The results of the current meta-analysis, the first of its kind, demonstrated improving effects of ginseng consumption on FMD, PWV, and serum levels of eNO, suggesting that ginseng can be recommended as an adjuvant in the management of vascular disorders in patients with CVD. Future studies should consider the best methodological designs, such as standardization of the intervention to ensure that the percentage of phytochemicals and bioactive compounds remains sufficient for the therapeutic effects of ginseng.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

ACE

Angiotensin converting enzyme

AIx

Augmentation index

BMI

Body mass index

95% CI

95% Confidence intervals

CVD

Cardiovascular diseases

DBP

Diastolic blood pressure

eNO

Endothelial-derived nitric oxide

eNOS

Endothelial-derived NO synthase

ET-1

Endothelin-1

FMD

Flow-mediated dilation

HTN

Hypertension

ICAM-1

Intercellular adhesion molecule-1

MD

Mean differences

PWV

Pulse wave velocity

RCT

Randomized controlled trial

SBP

Systolic blood pressure

SMD

Standardized mean difference

T2DM

Type 2 diabetes mellitus

VCAM-1

Vascular adhesion molecule-1

WHO

World health organization

Author contributions

AE: Contributed to the study conception, design and data collection, and interpretation and drafting the manuscript; screened articles and extracted data. NK: Contributed to the interpretation and drafting the manuscript. NN: Contributed to data collection; screened articles and extracted data. FHS: Participated in revising the paper critically and approving the version of the manuscript being submitted; and all authors: contributed to the manuscript, and read and approved the final manuscript.

Funding

This study was funded by Urmia University of Medical Sciences, Urmia, Iran., Grant agreement no. 12079.

Data availability

Data is provided within the manuscript or supplementary information files.

Declarations

Ethics approval and consent to participate

The study protocol was approved by the ethics committee of Urmia University of Medical Sciences, Urmia, Iran (Ethics code: IR.UMSU.HIMAM.REC.1402.004).

Consent for publication

All authors declared their consent for publication.

Competing interests

The authors declare no competing interests.