CVD, encompassing disorders that affect the heart and blood vessels, represent a major global public health concern due to their high morbidity and mortality rates. CVD remain the leading cause of death worldwide[Reference Gaidai, Cao and Loginov1]. From 1993 to 2019, the prevalence of CVD nearly doubled, and it is projected to increase further by 2024, underscoring the urgent need for effective public health strategies to address the growing burden of these conditions[Reference Murray, Aravkin and Zheng2]. CVD affect blood circulation and include coronary artery disease, heart failure, vascular abnormalities, dyslipidaemia and hypertension (HTN). These conditions are prevalent in populations following Western dietary patterns, with contributing risk factors such as oxidative stress, diabetes, high blood pressure (BP) and elevated cholesterol levels[Reference Toth3]. Adopting healthy lifestyle practices, including balanced dietary habits, is critical for reducing CVD risk[Reference Laranjo, Lanas and Sun4].
Complementary and alternative therapies, particularly those involving natural products, are gaining attention for their potential roles in the prevention and management of chronic diseases, including cardiovascular and metabolic disorders[Reference Valdés-González, Sánchez and Moratilla-Rivera5–Reference Arabi, Shahraki-Jazinaki and Nayyerabadi7]. Natural compounds such as conjugated linoleic acid, Pycnogenol, royal jelly, nuts and ginseng have demonstrated beneficial effects on inflammation, oxidative stress, lipid metabolism, glycaemic control and vascular function[Reference Park, Chung and Chung8–Reference Moher, Shamseer and Clarke10]. For instance, Pycnogenol has shown anti-inflammatory and antioxidant effects, potentially mitigating vascular dysfunction in chronic disease settings[Reference Jafari, Kordkatuli and Mardani11]. Similarly, royal jelly has demonstrated improvements in glycaemic indices and liver enzyme profiles, which are critical in metabolic syndrome and CVD development[Reference Moher, Shamseer and Clarke10]. Regular consumption of nuts, particularly walnuts, may promote satiety and help modulate body composition and cardiovascular biomarkers, suggesting a role in dietary strategies for cardiometabolic disease prevention[Reference Li, Wang and Xu9]. Among these, ginseng has been the focus of increasing scientific inquiry due to its diverse phytochemical composition and its emerging evidence base in supporting cardiovascular health[Reference Follmann, Elliott and Suh12].
Ginseng, a medicinal plant in the genus Panax, is known for its diverse biological activities. Different species of Panax contain various bioactive compounds, including ginsenosides, polysaccharides, polyacetylene compounds and alkaloids, which vary between species, necessitating careful differentiation among them[Reference Xu and Doi13]. Ginseng and its ginsenosides may confer cardiovascular benefits through antioxidative effects, reduced platelet adhesion, regulation of vasomotor function, improvement in lipid profiles and modulation of ion channels[Reference De Andrade, De Mesquita and de Almeida Claro14,Reference Jovanovski, Komishon and Au-Yeung15] .
Numerous systematic reviews have investigated the effects of ginseng on cardiovascular risk factors, yielding mixed results that highlight the complexity of its effects. A meta-analysis by Arabi et al., which included twenty-nine randomised controlled trials (RCT) through January 2024, reported that ginseng supplementation did not significantly alter lipid parameters – such as TAG, total cholesterol (TC), HDL-cholesterol and LDL-cholesterol – across various health conditions[Reference Jovanovski, Smircic-Duvnjak and Komishon16]. Conversely, another systematic review focusing on metabolic parameters, encompassing twenty-three studies up to 2022, found that ginseng supplementation significantly improved markers of blood glucose, BP and lipid levels, suggesting potential metabolic benefits[Reference Vuksan, Sung and Sievenpiper17]. Additionally, a comprehensive review that analysed nineteen meta-analyses up to July 2022 found potential benefits of ginseng in improving fatigue, physical function and metabolic markers, though the low quality of the included studies limited the reliability of these findings[Reference Vuksan, Xu and Jovanovski18]. These mixed results suggest that while ginseng may hold therapeutic promise, more rigorous studies are needed to clarify its efficacy and safety in specific populations.
Given recent literature developments, an updated review of the evidence is warranted. This study aims to address the limitations of earlier systematic reviews by offering comprehensive evidence on ginseng supplementation’s impact on CVD risk factors. Subgroup analyses were conducted to examine findings across specified outcomes, and the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach was used to evaluate the confidence in the evidence across studies. We believe that a rigorously conducted systematic review and meta-analysis will enhance the knowledge of clinicians, therapists and patients regarding ginseng’s efficacy.
Methods
Protocol and registration
This study followed a predetermined protocol based on the 2015 Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines[Reference Gao, Shi and Ma19]. The systematic review and meta-analysis were registered in the Prospective Register of Systematic Reviews (PROSPERO) under registration number CRD42023465688, to ensure transparency. The protocol was also published in BMJ Open [Reference Ma, Benzie and Chu20].
Search strategy and study selection
A comprehensive search was conducted across the Cochrane Central Register of Controlled Trials (CENTRAL), Web of Science, Medline (http://www.ncbi.nlm.nih.gov/pubmed), Scopus and Embase databases to identify articles published up to August 2024. Additionally, prominent journals and reference lists of relevant articles, meta-analyses and reviews on ginseng were manually reviewed to capture any additional eligible studies. To address publication bias, we expanded our search to include grey literature, such as conference proceedings and trial registries, in addition to published studies, though only RCT with sufficient data were ultimately included to ensure consistency in quality and analysis. Grey literature was systematically considered to broaden the evidence base and reduce the risk of overlooking relevant findings.
The primary focus was on the effects of ginseng supplementation. Outcomes investigated included changes in body fat percentage (BF%), BMI, waist circumference (WC), weight, diastolic BP (DBP), pulse pressure, systolic BP (SBP), 2-h glucose, 2-h insulin, fasting insulin, fasting blood sugar (FBS), HbA1c, Homeostatic Model Assessment of Beta-cell Function (HOMA-B), Homeostatic Model Assessment for Insulin Resistance (HOMA-IR), Quantitative Insulin Sensitivity Check Index (QUICKI), C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), high-sensitivity C-reactive protein (hs-CRP), IL-1 receptor antagonist (IL-1Ra), IL-4, IL-6, IL-8, IL-10, TNF-α, total protein, HDL-cholesterol, LDL-cholesterol, oxidised LDL-cholesterol, TC, TAG, albumin, alkaline phosphatase (ALP), alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (BUN), creatinine, gamma-glutamyl transferase (GGT), catalase (CAT), glutathione peroxidase (GPx), glutathione (GSH), glutathione reductase (GSH-Rd), malondialdehyde (MDA), reactive oxygen species (ROS), superoxide dismutase (SOD), total antioxidant capacity (TAC), adiponectin, leptin and heart rate following ginseng supplementation.
The search strategy targeted RCT involving human subjects and utilised keywords including ‘ginseng’ and ‘clinical trial’. Terms were sourced from Emtree and Medical Subject Headings in PubMed, and search strategies were tailored for each database (PubMed, Web of Science, Embase, Scopus and Central) using their respective subject headings. Full details of the search methodology are available in online Supplementary Table 1.
Inclusion/exclusion criteria and data extraction
We included trials that investigated the effects of ginseng supplementation on outcomes such as anthropometric measures, BP, glycaemic profile, inflammatory markers, lipid profile, liver function, oxidative stress parameters, adipokines and heart rate. Only studies published in full-text and in English were considered. Observational studies (cohort, case–control or cross-sectional), letters to the editor, review articles, animal studies and articles not published in English were excluded.
Two reviewers (F.M. and P.B.) conducted the initial search and screened studies based on eligibility criteria. Data management was performed using EndNote X7 to merge results and remove duplicates. The selection process followed PRISMA guidelines, and eligible articles were reviewed using a standardised data extraction form created by the primary reviewers (A.J. and V.M.). Two independent researchers (H.M. and K.K.) extracted the data, with any unresolved disagreements referred to the project supervisor (A.A.) for consensus.
Methodological quality assessment and strength of evidence evaluation
Two reviewers (M.A. and B.P.) conducted a detailed assessment of the methodological quality of the included trials using the Cochrane criteria. Any disagreements were resolved through discussion or, if necessary, with input from a third reviewer (A.H.F.). The criteria used to evaluate each trial included adequacy of sequence generation, allocation concealment, blinding, handling of incomplete outcome data, selective outcome reporting and identification of other potential biases.
The GRADE framework was applied to assess and summarise the quality of evidence for each outcome included in the meta-analysis. This approach categorises evidence into four levels: ‘high’, ‘moderate’, ‘low’ or ‘very low’. Evidence quality could be adjusted based on methodological and statistical factors. Two independent reviewers (A.J. and A.A.) assessed each outcome for factors such as study design, risk of bias, consistency, precision, directness, effect size, dose–response gradient and publication bias, with a third reviewer (V.M.) serving as an adjudicator if needed.
Statistical analysis
Statistical analysis was performed using Stata software, version 17·0 (StataCorp). We extracted pre- and post-treatment means, sd and participant counts from both intervention and placebo groups across clinical outcomes, including lipid profile, glycaemic control, anthropometric indices, inflammation, BP, oxidative stress, liver function, adipokines and heart rate. Standardised mean differences (SMD) between intervention and placebo groups were calculated to assess the impact of ginseng supplementation.
A random effects model using the DerSimonian–Laird method was applied to account for heterogeneity across studies. In cases where sd for changes were not reported, they were estimated using the Follmann method[Reference Xu, Lu and Wu21], assuming a correlation coefficient (R) of 0·5. When only the standard error (se) was provided, we calculated sd using the formula sd = se × √n, where n is the sample size. The sd of the change difference were calculated as sd = √[(sd pre-treatment)2 + (sd post-treatment)² – (2R × sd pre-treatment × sd post-treatment)], with a correlation coefficient of 0·5.
Meta-regression assessed the potential impact of ginseng dosage (grams per day) and duration on CVD risk factors. For the meta-regression, we used a random effects model to explore sources of heterogeneity among studies. Heterogeneity was evaluated using the I² statistic and Cochran’s Q test. Meta-regression was conducted using the metareg command in Stata, and covariates with a P-value < 0·10 were considered potentially significant modifiers of effect size.
To handle heterogeneity and enhance the generalizability of our results, we performed subgroup analyses based on factors such as study population, study quality, intervention age, gender, ginseng dosage, continent and duration of intervention allowing us to explore variations in outcomes across different contexts and improve the applicability of our findings. These approaches were complemented by standard sensitivity analyses to assess the robustness of our conclusions, ensuring a comprehensive evaluation of the evidence. Additionally, a non-linear model was applied to integrate dose–response data from multiple studies, exploring the effects of varying ginseng dosages on CVD risk factors[Reference Yang, Li and Zhou22]. Sensitivity analysis was performed using a leave-one-out approach, in which the meta-analysis was recalculated after excluding one study at a time to assess the robustness of the results.
Results
Study selection
The study selection process is illustrated in Fig. 1. A total of 20 739 studies were identified through database searches: PubMed (n 2325), ISI Web of Science (n 2841), Scopus (n 5362), Embase (n 8761) and the Cochrane Library (n 1450). After removing 9796 duplicates, irrelevant studies and animal studies, 10 943 studies were screened based on titles and abstracts. Of these, 10 802 were excluded, resulting in 141 full-text studies evaluated further. Seventy-four studies were excluded for not reporting the desired outcomes, as detailed in online Supplementary Table 2. Consequently, 70 studies comprising 4506 participants were included in the systematic review and meta-analysis.
Flowchart of study selection for inclusion trials in systematic review.
Study characteristics
The characteristics of the included studies are summarised in Table 1. Online Supplementary Figures 1–9 show the SMD and 95 % CI for changes in BF%, BMI, WC, weight, DBP, pulse pressure, SBP, 2-h glucose, 2-h insulin, fasting insulin, FBS, HbA1c, HOMA-B, HOMA-IR, QUICKI, CRP, ESR, hs-CRP, IL-1 receptor antagonist, IL-4, IL-6, IL-8, IL-10, TNF-α, total protein, HDL-cholesterol, LDL-cholesterol, oxidised LDL-cholesterol, TC, TAG, albumin, ALP, ALT, AST, BUN, creatinine, GGT, CAT, GPx, GSH, GSH-Rd, MDA, ROS, SOD, TAC, adiponectin, leptin and heart rate.
Characteristic of included studies in the meta-analysis
ALP, alkaline phosphatase; ALT, alanine aminotransferase; AMI, acute myocardial infarction; AST, aspartate aminotransferase; B, both sex; BF, body fat; BFM, body fat mass; BW, body weight; CON, control group; DB, double-blinded; DBP, diastolic blood pressure; ED, erectile dysfunction; F, female; FBG, fasting blood glucose; FBS, fasting blood sugar; FG, fasting glucose; FM, fat mass; FPG, fasting plasma glucose; FPI, fasting plasma insulin; HC, hip circumference; HOMA-IR, Homeostatic Model Assessment of Insulin Resistance; hs-CRP, high-sensitivity C-reactive protein; HTN, hypertension; IFG, impaired fasting glucose; IGT, impaired glucose tolerance; IL-1Ra, IL-1 receptor antagonist; INT, intervention group; LDH, lactate dehydrogenase; MDA, malondialdehyde; MetS, metabolic syndrome; NAFLD, non-alcoholic fatty liver disease; NO, nitric oxide; NR, not reported; QUICKI, Quantitative Insulin Sensitivity Check Index; RCT, randomised controlled trial; SBP, systolic blood pressure; SOD, superoxide dismutase; T2D, type 2 diabetes; T2DM, type 2 diabetes mellitus; TB, triple-blinded; TBARS, thiobarbituric acid reactive substances; TC, total cholesterol; WC, waist circumference; WHR, waist-to-hip ratio.
Studies were published between 1998 and 2024 and were conducted in Brazil[Reference Mucalo, Jovanovski and Rahelić47], Canada[Reference Delui, Fatehi and Manavifar44,Reference Hosseini, Ehsanpour and Asgari45,Reference Mucalo, Jovanovski and Vuksan50,Reference Hosseini, Alipour and Ghadiry57] , China[Reference Ahn, Hong and Choi31,Reference Cha, Kim and Kim55,Reference Al-Kuraishy and Al-Gareeb61,Reference Hosseini, Ghaedi and Zakerkish62] , Croatia[Reference Bang, Kwak and Ahn48,Reference Baek, Heo and Fava65] , Iran[Reference Cho, Ahn and Lee42,Reference Choi, Park and Jang43,Reference Cho, Son and Kim49,Reference Choi, Kim and Kim64] , Iraq[Reference Chung, Kim and Seol73], Japan[Reference Chung, Lee and Lee74], South Korea[Reference Han, Choe and Kim23–Reference Kim, Kim and Kim30,Reference Kim, Yoo and Park32–Reference Yoon, Kang and Vassy41,Reference Kim, Cho and Yoo46,Reference Oh, Park and Kim51–Reference Seo, Hong and Yun54,Reference Hong, Lee and Kim56,Reference Jung, Lee and Kim58–Reference Park, Shin and Kim60,Reference Kim, Kim and Lee63,Reference Hwang, Park and Choi66–Reference Sung, Kang and Jung72,Reference Hong, Lee and Yoon75–86] , Spain[Reference Hosseini, Alipour and Zakerkish87], Thailand[Reference Mohammadi, Hadi and Kord-Varkaneh88], the UK[Reference Hernandez-Garcia, Granado-Serrano and Martín-Gari89] and the USA[Reference Yokozawa, Satoh and Cho90–Reference Kim, Sung and Chang92]. Participant ages in intervention groups ranged from 34·2 to 64 years, with Ginseng doses varying from 50 to 6000 mg/d over intervention durations of 2–35 weeks. Sample sizes ranged from 6 to 195 participants, with studies including only males[Reference Kim, Kim and Kim30,Reference Rhee, Kim and Bae33,Reference Lee, Yoo and Kim39,Reference Mucalo, Jovanovski and Rahelić47,Reference Jung, Lee and Kim58,Reference Attele, Wu and Yuan84,Reference Hosseini, Alipour and Zakerkish87,Reference Mohammadi, Hadi and Kord-Varkaneh88] , only females[Reference Caron, Hotsko and Robertson25,Reference Oh, Park and Kim51,Reference Park, Oh and Choi52,Reference Lee, Lee and Yoo59,Reference Park, Shin and Kim60,Reference Jung, Hwang and Park67,Reference Park, Ahn and Kim69,Reference Park, Kim and Kim70,Reference Jung, Kang and Oh77,Reference Park, Lee and Lee80,Reference Hernández-García, Granado-Serrano and Martín-Gari82,Reference Lee and Ji83,Reference Oh, Chung and Yong91] , both sexes[Reference Han, Choe and Kim23,Reference Kim and Lee24,Reference En-Yuan, Ya-Jun and Lian-Xing26–Reference Dickman, Koenig and Ji29,Reference Ahn, Hong and Choi31,Reference Kim, Yoo and Park32,Reference Reeds, Patterson and Okunade34–Reference Lee, Lee and Oh38,Reference Park, Lee and Lee40–Reference Kim, Cho and Yoo46,Reference Bang, Kwak and Ahn48,Reference Mucalo, Jovanovski and Vuksan50,Reference Rhee, Cho and Kim53–Reference Hosseini, Alipour and Ghadiry57,Reference Hosseini, Ghaedi and Zakerkish62–Reference Hwang, Park and Choi66,Reference Kwon, Jang and Liu68,Reference Shen, Gwak and Joo71–Reference Hyun, Ahn and Kim76,Reference Kim, Chon and Seon78,Reference Seong, Bose and Han79,Reference Shen, Lee and Joo81,Reference Bostock, Kirkby and Garry85,86,Reference Hernandez-Garcia, Granado-Serrano and Martín-Gari89,Reference Yokozawa, Satoh and Cho90,Reference Kim, Sung and Chang92] or not reporting gender[Reference Cho, Son and Kim49,Reference Al-Kuraishy and Al-Gareeb61] . Participants had cardiovascular or hypertensive disorders[Reference En-Yuan, Ya-Jun and Lian-Xing26–Reference Onanong Kulaputana28,Reference Kim, Yoo and Park32,Reference Cho, Son and Kim49,Reference Park, Oh and Choi52,Reference Chung, Lee and Lee74] , metabolic or glycaemic disorders[Reference Dickman, Koenig and Ji29,Reference Kim, Kim and Kim30,Reference Cho, Ahn and Lee42–Reference Hosseini, Ehsanpour and Asgari45,Reference Bang, Kwak and Ahn48,Reference Mucalo, Jovanovski and Vuksan50,Reference Seo, Hong and Yun54,Reference Hosseini, Alipour and Ghadiry57,Reference Al-Kuraishy and Al-Gareeb61–Reference Baek, Heo and Fava65,Reference Shen, Gwak and Joo71,Reference Sung, Kang and Jung72,Reference Hong, Lee and Yoon75,Reference Hyun, Ahn and Kim76,Reference Kim, Chon and Seon78,Reference Seong, Bose and Han79,Reference Shen, Lee and Joo81,86] or general health or non-specific conditions[Reference Han, Choe and Kim23–Reference Caron, Hotsko and Robertson25,Reference Ahn, Hong and Choi31,Reference Rhee, Kim and Bae33–Reference Yoon, Kang and Vassy41,Reference Kim, Cho and Yoo46,Reference Mucalo, Jovanovski and Rahelić47,Reference Oh, Park and Kim51,Reference Rhee, Cho and Kim53,Reference Cha, Kim and Kim55,Reference Hong, Lee and Kim56,Reference Jung, Lee and Kim58–Reference Park, Shin and Kim60,Reference Hwang, Park and Choi66–Reference Park, Kim and Kim70,Reference Chung, Kim and Seol73,Reference Jung, Kang and Oh77,Reference Park, Lee and Lee80,Reference Hernández-García, Granado-Serrano and Martín-Gari82–Reference Bostock, Kirkby and Garry85,Reference Hosseini, Alipour and Zakerkish87–Reference Kim, Sung and Chang92] . The following are the sample sizes for the intervention and control groups across various outcomes: BF%: n 173 (intervention: 86, control: 87), BMI: n 784 (intervention: 419, control: 365), WC: n 272 (intervention: 152, control: 120), weight: n 220 (intervention: 111, control: 109), DBP: n 1537 (intervention: 826, control: 711), pulse pressure: n 214 (intervention: 105, control: 109), SBP: n 1537 (intervention: 826, control: 711), 2-h glucose: n 379 (intervention: 204, control: 175), 2-h insulin: n 123 (intervention: 61, control: 62), FBS: n 2515 (intervention: 1354, control: 1161), fasting insulin: n 1163 (intervention: 630, control: 533), HbA1c: n 660 (intervention: 376, control: 284), HOMA-B: n 80 (intervention: 46, control: 126), HOMA-IR: n 966 (intervention: 529, control: 437), QUICKI: n 369 (intervention: 200, control: 169), CRP: n 223 (intervention: 112, control: 111), ESR: n 169 (intervention: 111, control: 58), hs-CRP: n 740 (intervention: 428, control: 312), IL-1R: n 102 (intervention: 54, control: 58), IL-4: n 306 (intervention: 153, control: 153), IL-6: n 908 (intervention: 505, control: 403), IL-8: n 102 (intervention: 54, control: 48), IL-10: n 205 (intervention: 106, control: 99), TNF-α: n 933 (intervention: 487, control: 446), total protein: n 840 (intervention: 455, control: 385), HDL-cholesterol: n 2196 (intervention: 1192, control: 1004), LDL-cholesterol: n 1873 (intervention: 1004, control: 869), oxidised LDL-cholesterol: n 153 (intervention: 77, control: 76), TC: n 2740 (intervention: 1490, control: 1250), TAG: n 2403 (intervention: 1318, control: 1085), albumin: n 740 (intervention: 405, control: 335), ALP: n 729 (intervention: 399, control: 330), ALT: n 1758 (intervention: 952, control: 806), AST: n 1586 (intervention: 863, control: 723), BUN: n 786 (intervention: 427, control: 359), creatinine: n 1507 (intervention: 807, control: 700), GGT: n 620 (intervention: 351, control: 269), CAT: n 228 (intervention: 150, control: 78), GPx: n 241 (intervention: 147, control: 94), GSH: n 195 (intervention: 124, control: 71), GSH-Rd: n 170 (intervention: 112, control: 58), MDA: n 425 (intervention: 242, control: 183), ROS: n 170 (intervention: 112, control: 58), SOD: n 331 (intervention: 193, control: 138), TAC: n 254 (intervention: 170, control: 84), adiponectin: n 165 (intervention: 99, control: 66), leptin: n 61 (intervention: 46, control: 15) and heart rate: n 713 (intervention: 383, control: 330).
Qualitative data assessment
Using the Cochrane Risk of Bias Assessment tool (ROB2), nine studies were rated as good quality[Reference Rhee, Kim and Bae33,Reference Lee, Lee and Oh38,Reference Park, Lee and Lee40,Reference Oh, Park and Kim51,Reference Hosseini, Ghaedi and Zakerkish62,Reference Park, Ahn and Kim69,Reference Seong, Bose and Han79,Reference Bostock, Kirkby and Garry85,Reference Kim, Sung and Chang92] , twenty two as fair[Reference Kim and Lee24,Reference Ahn, Hong and Choi31,Reference Reeds, Patterson and Okunade34,Reference Delui, Fatehi and Manavifar44,Reference Mucalo, Jovanovski and Rahelić47,Reference Bang, Kwak and Ahn48,Reference Cha, Kim and Kim55,Reference Hosseini, Alipour and Ghadiry57,Reference Jung, Hwang and Park67,Reference Park, Kim and Kim70,Reference Shen, Gwak and Joo71,Reference Jung, Kang and Oh77,Reference Kim, Chon and Seon78,Reference Park, Lee and Lee80–Reference Attele, Wu and Yuan84,86–Reference Mohammadi, Hadi and Kord-Varkaneh88,Reference Yokozawa, Satoh and Cho90] and the remaining studies as poor[Reference Han, Choe and Kim23,Reference Caron, Hotsko and Robertson25–Reference Kim, Kim and Kim30,Reference Kim, Yoo and Park32,Reference Kim, Park and Kang35–Reference Kwon, Bose and Song37,Reference Lee, Yoo and Kim39,Reference Yoon, Kang and Vassy41–Reference Choi, Park and Jang43,Reference Hosseini, Ehsanpour and Asgari45,Reference Kim, Cho and Yoo46,Reference Cho, Son and Kim49,Reference Mucalo, Jovanovski and Vuksan50,Reference Park, Oh and Choi52–Reference Seo, Hong and Yun54,Reference Hong, Lee and Kim56,Reference Jung, Lee and Kim58–Reference Al-Kuraishy and Al-Gareeb61,Reference Kim, Kim and Lee63–Reference Hwang, Park and Choi66,Reference Kwon, Jang and Liu68,Reference Sung, Kang and Jung72–Reference Hyun, Ahn and Kim76,Reference Hernandez-Garcia, Granado-Serrano and Martín-Gari89,Reference Oh, Chung and Yong91] (Table 2).
Quality of included studies in the meta-analysis
H, high risk of bias; L, low risk of bias; U, unclear risk of bias.
Effects of ginseng supplementation on anthropometric indices
Ginseng supplementation did not show significant effects on BF%, BMI, WC or weight (online Supplementary Fig. 1). Egger’s test did not indicate significant publication bias for BF% (P = 0·798), BMI (P = 0·652), WC (P = 0·909) or weight (P = 0·511). Sensitivity analyses, excluding individual studies, revealed no changes in these results.
Effects of ginseng supplementation on blood pressure
Ginseng supplementation did not demonstrate significant effects on DBP (online Supplementary Fig. 2), pulse pressure or SBP. Egger’s test indicated no significant publication bias for DBP (P = 0·552), pulse pressure (P = 0·905) or SBP (P = 0·966). Sensitivity analysis, excluding studies by En-Yuan et al.[Reference Kim, Cho and Yoo46] (SMD: –0·29; 95 % CI: –0·57, –0·02), Hwang et al.[Reference Kim and Lee24] (SMD: –0·28; 95 % CI: –0·57, 0·00) and Dickman et al.[Reference Oh, Chung and Yong91] (SMD: –0·30; 95 % CI: –0·57, –0·02) resulted in changes to the overall effect estimates for DBP. Excluding studies by En-Yuan et al.[Reference Kim, Cho and Yoo46] (SMD: –0·26; 95 % CI: –0·51, –0·01) altered the overall results on SBP.
Effects of ginseng supplementation on glycaemic profile
Our meta-analysis did not show any significant changes in 2-h glucose (online Supplementary Fig. 3), 2-h insulin, fasting insulin, FBS, HbA1c, HOMA-B, HOMA-IR and QUICKI.
Unlike other parameters in this subgroup, significant publication bias was observed in 2-h glucose (Egger’s P = 0·010), and QUICKI (Egger’s P = 0·004). Sensitivity analysis showed no alteration in this group’s parameters.
Effects of ginseng supplementation on inflammatory markers
Studies indicated a significant reduction in hs-CRP upon ginseng administration (SMD = −0·23, 95 % CI: −0·38, −0·08; P = 0·002; I2 = 0·0, P = 0·463) (online Supplementary Fig. 4). However, no significant results were found on CRP, ESR, IL-1 receptor antagonist, IL-4, IL-6, IL-8, IL-10, TNF-α and total protein.
No publication bias was suggested among the parameters in this group. Sensitivity analysis showed significant changes in IL-4 results upon excluding Yang et al.[Reference Ahn, Hong and Choi31] (SMD: −0·36, 95 % CI: −0·64, −0·08).
Effects of ginseng supplementation on lipid profile
According to our results, ginseng supplementation is not beneficial on HDL-cholesterol (online Supplementary Fig. 5), LDL-cholesterol, oxidised LDL-cholesterol, TC and TAG.
Sensitivity analysis shows the exclusion of Beak et al.[Reference Han, Choe and Kim23] leads to changes in overall results in LDL-cholesterol levels and TC levels. Egger’s test indicated no evidence of publication bias in lipid profile parameters.
Effects of ginseng supplementation on liver function
Ginseng supplementation demonstrated a beneficial effect on GGT levels (SMD = −0·20, 95 % CI: −0·37, −0·04; P = 0·015; I2 = 0·0, P = 0·619) (online Supplementary Fig. 6). However, no changes were seen in albumin, ALP, ALT, AST, BUN and creatinine.
No publication bias was suggested among the parameters in this group. Unlike other parameters in this group, sensitivity analysis showed that removing the study by Shen et al.[Reference Attele, Wu and Yuan84] (SMD: −0·16, 95 % CI: −0·34, 0·02) affected the overall results on GGT.
Effects of ginseng supplementation oxidative stress parameters
Trials revealed a significant impact on GSH-Rd (SMD = 0·90, 95 % CI: 0·38, 1·42; P = 0·001; I2 = 58·5, P = 0·065) (online Supplementary Fig. 7), ROS (SMD = −0·94, 95 % CI: −1·27, −0·60; P = <0·001; I2 = 0·0, P = 0·824) and SOD (SMD = 0·49, 95 % CI: 0·10, 0·87; P = 0·014; I2 = 65·2, P = 0·008). However, no changes were seen on CAT, GPx, GSH, MDA and TAC.
A considerable publication bias was observed for CAT (Egger’s P = 0·013), whereas no publication bias was detected for the other oxidative stress parameters.
Sensitivity analysis showed that the exclusion of the study by Kim et al.[Reference Kwon, Jang and Liu68] altered the results for CAT (SMD: 0·34; 95 % CI: 0·00, 0·69). The exclusion of the study by Kim et al.[Reference Lee, Lee and Oh38] changed the overall effect on GSH (SMD: 0·58; 95 % CI: 0·11, 1·05). The study by Dickman et al.[Reference Oh, Chung and Yong91] showed changes in sensitivity analysis on MDA (SMD: −0·73; 95 % CI: −1·36, −0·10) Additionally, the exclusion of the study by Kim et al.[Reference Lee and Ji83] affected the results for SOD (SMD = 0·43, 95 % CI: −0·01, 0·87).
Effects of ginseng supplementation on adipokines
No significant changes were observed in adiponectin or leptin levels following ginseng supplementation (online Supplementary Fig. 8).
Sensitivity analysis revealed no significant alterations upon exclusion of any of the included studies. No publication bias was detected for either adiponectin or leptin.
Effects of ginseng supplementation on heart rate
The pooled results from studies examining the effects of ginseng on heart rate showed no significant decrease (online Supplementary Fig. 9).
Sensitivity analysis indicated no significant changes in the overall results after the exclusion of any of the studies. There was no publication bias detected for heart rate, as indicated by Egger’s test.
Subgroup analysis
In the subgroup analysis (Table 3) of ginseng interventions, participants were categorised by health conditions, geographical region, dosage, duration of intervention, gender, age and study quality. The health conditions were divided into cardiovascular or hypertensive disorders, metabolic or glycaemic disorders and general or non-specific conditions. These included patients with coronary artery disease, type 2 diabetes mellitus, HTN, chronic fatigue and more.
Description of the analysis and subgroup results of ginseng supplementation
ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BF%, body fat percentage; BUN, blood urea nitrogen; CAT, catalase; CRP, C-reactive protein; DBP, diastolic blood pressure; ESR, erythrocyte sedimentation rate; FBS, fasting blood sugar; GPx, glutathione peroxidase; GSH, glutathione; GSH-Rd, glutathione reductase; GGT, gamma-glutamyl transferase; HbA1c, glycated Hb; HC, hip circumference; HOMA-B, Homeostatic Model Assessment of Beta-cell Function; HOMA-IR, Homeostatic Model Assessment of Insulin Resistance; hs-CRP, high-sensitivity C-reactive protein; IL-1R, IL-1 receptor; MDA, malondialdehyde; QUICKI, Quantitative Insulin Sensitivity Check Index; ROS, reactive oxygen species; SMD, standardised mean difference; SBP, systolic blood pressure; SOD, superoxide dismutase; TAC, total antioxidant capacity; TC, total cholesterol; WC, waist circumference.
Boldface is used to highlight statistically significant P-values (P < 0·05) to enhance clarity and ensure these key results stand out for better comprehension.
The health conditions were organised into three distinct, non-overlapping categories for analysis: (1) cardiovascular or hypertensive disorders, (2) metabolic or glycaemic disorders and (3) general health or non-specific conditions. The cardiovascular or hypertensive disorders category included patients with heart-related conditions such as coronary artery disease or heart failure, as well as those with BP disorders like HTN or prehypertension, and individuals with elevated cholesterol levels. The metabolic or glycaemic disorders category comprised participants with conditions affecting blood sugar regulation, such as type 2 diabetes, metabolic syndrome or impaired glucose tolerance, alongside related issues like non-alcoholic fatty liver disease. The general health or non-specific conditions category encompassed a diverse group, including healthy adults, smokers, physically active individuals, postmenopausal women and individual with fatigue, among others. Although these categories were designed to be mutually exclusive for analytical clarity, we recognise that some participants may have concurrent conditions (e.g. a patient with both diabetes and HTN). In such instances, participants were assigned to the category corresponding to their primary diagnosis or the condition most relevant to the study’s focus, as reported by the original study authors, ensuring a consistent and interpretable subgroup analysis.
Participants came from Asia, Europe and Canada/USA, with ginseng dosages split into those receiving less than 3000 mg/d, more than 3000 mg/d or unspecified doses. The duration of interventions was categorised as either 8 weeks or less or longer than 8 weeks. Gender groups included male, female, both or not specified, while age was categorised as 48 years or younger or older than 48 years. Finally, the study quality was assessed as poor, fair or good to evaluate the reliability of the outcomes. This structured subgrouping allows for a detailed exploration of ginseng’s effects across various demographics and conditions.
In our subgroup analysis, ginseng supplementation showed significant effects in patients with cardiovascular or hypertensive disorders on DBP, TNF-α, LDL-cholesterol and TC. In those with metabolic or glycaemic complications, significant changes were seen in FBS, total protein, albumin and AST. For the general population, ginseng improved HOMA-IR, GGT and hs-CRP. Positive results for SBP were consistent across all health conditions.
Based on continent, ginseng supplementation showed improvements in hs-CRP, IL-6, GGT and MDA among Asian participants. For MDA, studies conducted in Canada or the USA reported significant findings, while for SBP, only one study from Croatia reported promising outcomes[Reference Baek, Heo and Fava65].
In the gender subgroup analysis, ginseng supplementation showed significant improvements in both males and females for hs-CRP, IL-6, GGT and MDA. In females, significant changes were observed in total protein and albumin, whereas in males, improvements were noted in DBP, SBP, albumin and GGT.
In the dose subgroup analysis, ginseng supplementation at doses lower than 3000 mg showed significant improvements in hs-CRP, while doses higher than 3000 mg resulted in significant changes in MDA, based on one study[Reference Lee and Ji83].
In the duration subgroup analysis, ginseng supplementation for less than 8 weeks showed significant improvements in DBP and total protein, while interventions lasting more than 8 weeks resulted in significant changes in hs-CRP, LDL-cholesterol and TC.
In the age subgroup analysis, ginseng supplementation in participants younger than 48 years showed significant improvements in DBP, hs-CRP, ALT and MDA. In contrast, those older than 48 years exhibited significant changes in HOMA-IR, LDL-cholesterol and TC. Ginseng supplementation showed an increase in fasting insulin in participants younger than 48 years.
Based on the quality of the included studies, significant improvements in DBP, hs-CRP, LDL-cholesterol and albumin were observed in studies rated as poor. In studies with fair quality, notable changes were seen in GGT and MDA. Additionally, studies classified as good quality demonstrated significant effects on MDA and total protein.
Meta-regression and non-linear dose–response analysis
A non-linear dose–response regression model was used to examine the relationship between ginseng dosage and cardiovascular outcomes. The meta-regression analysis revealed a significant association between ginseng dosage and HOMA-IR (P = 0·044), suggesting that higher doses may improve insulin resistance. Additionally, a significant association was found between intervention duration and MDA (P = 0·007), indicating that prolonged supplementation could increase MDA levels, which may be undesirable (online Supplementary Figs. 10–11).
The dose–response analysis also demonstrated significant associations between ginseng dose and several markers, including FBS (P < 0·001), hs-CRP (P = 0·043), IL-6 (P = 0·041), DBP (P = 0·022), IL-10 (P = 0·048), fasting insulin (P = 0·012) and total protein (P = 0·010), suggesting that higher doses reduce these markers. Conversely, a positive association was found between MDA (P = 0·008) and duration of supplementation, indicating that prolonged use may elevate MDA levels (online Supplementary Figs. 12–13).
GRADE assessment
The GRADE profile of ginseng supplementation, indicating the certainty of outcomes is presented in Table 4. Quality of evidence for FBS, total protein, oxidised LDL-cholesterol, ALT, AST, BUN, GGT and heart rate were moderate. ROS, creatinine, albumin, ALP, TC, TAG, TNF-α, IL-6, HOMA-B, weight and BMI were graded as low and the rest of the outcomes were very low in quality (Fig. 2).
GRADE profile of ginseng supplementation on CVD outcomes
ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BF, body fat; BUN, blood urea nitrogen; CAT, catalase; CRP, C-reactive protein; DBP, diastolic blood pressure; ESR, erythrocyte sedimentation rate; FBS, fasting blood sugar; GPx, glutathione peroxidase; GSH, glutathione; GSH-Rd, glutathione reductase; GGT, gamma-glutamyl transferase; HbA1c, glycated Hb; HC, hip circumference; HOMA-B, Homeostatic Model Assessment of Beta-cell Function; HOMA-IR, Homeostatic Model Assessment of Insulin Resistance; hs-CRP, high-sensitivity C-reactive protein; IL-1R, IL-1 receptor; MDA, malondialdehyde; QUICKI, Quantitative Insulin Sensitivity Check Index; ROS, reactive oxygen species; SBP, systolic blood pressure; SMD, standardised mean difference; SOD, superoxide dismutase; TAC, total antioxidant capacity; TC, total cholesterol; WC, waist circumference.
* Downgraded since more than 50 % of the participants were from high-risk bias studies.
† Downgraded since the participants included were less than 400 people.
‡ Downgraded since the 95 % CI crosses the threshold of interest.
§ Downgraded for indirectness in the country.
|| Publication Bias was detected through Egger and Begg’s test. (P-value < 0·05).
¶ The I 2 value was > 50 % (or heterogeneity among the studies was high).
Credibility assessment for included outcomes. hs-CRP, high-sensitivity C-reactive protein; GGT, gamma-glutamyl transferase; FBS, fasting blood sugar; BUN, blood urea nitrogen; AST, aspartate aminotransferase; ALT, alanine aminotransferase; TC, total cholesterol; ROS, reactive oxygen species; HOMA-B, Homeostatic Model Assessment of Beta-cell Function; ALP, alkaline phosphatase; TAC, total antioxidant capacity; SOD, superoxide dismutase; MDA, malondialdehyde; IL-1Ra, IL-1 receptor antagonist; GSH-Rd, glutathione reductase; GSH, glutathione; GPx, glutathione peroxidase; ESR, erythrocyte sedimentation rate; CRP, C-reactive protein; CAT, catalase; WC, waist circumference; SBP, systolic blood pressure; QUICKI, Quantitative Insulin Sensitivity Check Index; HOMA-IR, Homeostatic Model Assessment of Insulin Resistance; DBP, diastolic blood pressure; BF, body fat.
Discussion
Summary of findings
This systematic review and meta-analysis synthesised data from 70 RCT (4506 participants) to assess ginseng’s effects on various health outcomes, particularly highlighting its potential for metabolic and cardiovascular health. Key findings indicated significant reductions in hs-CRP and ROS, along with increases in antioxidant markers (SOD, GSH-Rd), which suggest ginseng’s anti-inflammatory and antioxidant properties. However, other oxidative stress (CAT, GSH, GPx, MDA and TAC) and inflammatory biomarkers (TNF-α, IL and ESR) were not altered significantly following ginseng supplementation. Linear and non-linear associations between these biomarkers and the dose suggested that the differences in the administered dosage can be one of the possible reasons for this controversy in these findings. Dose–response analysis revealed that lower doses were associated with more decrease in FBS (300 mg/d), hs-CRP (300 mg/d), IL-6 (300 mg/d), DBP (100 mg/d) and insulin (50 mg/d). Ginsenosides, the primary active compounds in ginseng, interact with specific receptors, including nuclear receptors like PPAR, which play a role in anti-inflammatory and antioxidant activities. Excessive doses may cause receptor downregulation or desensitisation, diminishing the response[Reference Jang, Hwang and Cho93]. The recommended dose of ginseng is 200–600 mg of extract[Reference Arafa, Refaey and Abd El-Ghafar94]. High doses of ginseng can lead to mild adverse effects, including slight elevations in BP or gastrointestinal disturbances, which may reduce its net therapeutic benefit[Reference Yang, Li and Lu95]. Lower doses are associated with fewer side effects, allowing the beneficial effects of ginseng to predominate.
According to the results of subgroup analysis, ginseng supplementation was associated with reductions in SBP, DBP, TNF-α, LDL-cholesterol and TC in participants with cardiovascular or hypertensive disorders. Additionally, in the metabolic or glycaemic disorders subgroup, ginseng supplementation resulted in significant decrease in FBS. These findings indicated that high baseline values are one of the factors that can influence the effectiveness of ginseng on cardiometabolic risk factors. Moreover, the observed changes in GGT levels, highlighted ginseng’s potential role in liver health. However, without a significant effect on other liver biomarkers, this finding must be interpreted with caution.
Comparison with previous studies
Although hs-CRP was significantly reduced in our study, previous findings have been mixed, with some studies, such as those in acute myocardial infarction[Reference En-Yuan, Ya-Jun and Lian-Xing26] and menopausal populations[Reference Park, Ahn and Kim69], reporting no significant changes in hs-CRP with ginseng supplementation for 8 and 3 months, respectively. However, Hosseini et al. conducted a study on type 2 diabetes patients that demonstrated a significant decrease in hs-CRP levels after 8 weeks of using a standardised extract of ginseng[Reference Liu, Wang and Qian96]. Our subgroup analysis demonstrated variability by geography, suggesting population-specific responses to ginseng. Similarly, a pooled analysis of six trials involving 308 participants showed no significant change in hs-CRP overall; however, significant reductions were observed in studies conducted in Iran, indicating possible geographical variability in ginseng’s effectiveness[Reference Zhang, Ye and Lu97].
While a meta-analysis of ten studies reported reductions in TC and LDL-cholesterol in metabolic syndrome patients[Reference Ghavami, Ziaei and Foshati98], our findings highlighted these reductions specifically in the cardiovascular group, indicating potential differences based on underlying health conditions.
Similarly, we observed significant antioxidant effects, consistent with reports that ginsenosides enhance GSH levels and antioxidant enzymes such as GPx, SOD and CAT[Reference Mohanan, Subramaniyam and Mathiyalagan99,Reference Kim, Kim and Kang100] . Ginsenosides have been linked to increased SOD and CAT activities in various studies, which parallels our findings on reduced ROS and increased SOD, highlighting ginseng’s role in oxidative stress mitigation[Reference Zhao, Zhang and Chen101]. However, while previous studies focus on general oxidative markers, our analysis suggests specific effects on ROS, adding a targeted dimension to ginseng’s antioxidant profile.
Mechanisms of action
Ginseng’s health benefits are attributed to a range of bioactive compounds, primarily ginsenosides, which have been shown to exert anti-inflammatory, antioxidant and metabolic regulatory effects at the molecular level. Ginseng, especially through its active ginsenosides, regulates inflammatory responses by inhibiting key signalling pathways like NF-κB and mitogen-activated protein kinases[Reference Han, Sung and Lee102,Reference Reay, Scholey and Milne103] . The NF-κB pathway, crucial in the transcription of pro-inflammatory cytokines, is suppressed by ginsenosides, which reduce NF-κB activation and the phosphorylation of its inhibitor, IκBα. This inhibition leads to decreased production of pro-inflammatory cytokines, such as TNF-α, thereby reducing systemic inflammation. Ginseng also inhibits the MAPK pathway, specifically the extracellular signal-regulated kinase 1/2, c-Jun N-terminal kinase and p38 branches, which further suppresses pro-inflammatory cytokine synthesis and cellular stress responses[Reference Shishtar, Sievenpiper and Djedovic104]. This anti-inflammatory effect is coupled with a reduced production of inflammatory mediators such as PG E2 while also suppressing the expression of enzymes like cyclooxygenase-2. These molecular pathways play a crucial role in reducing oxidative stress and inflammation, further supporting the increase in antioxidant enzymes like SOD, CAT and GPx[Reference Wee, Park, Chung, Benzie and Wachtel-Galor105].
Moreover, ginseng regulates the expression of pro-apoptotic and anti-apoptotic proteins, such as by decreasing caspase-3 activity and increasing B-cell lymphoma 2 levels, enhancing cell survival and stability in liver tissue[Reference Kim and Lee24,Reference NP106] . This broadened biochemical response could explain the significant change in GGT levels, reflecting potentially increased enzymatic activity in hepatocytes and highlighting ginseng’s multifaceted contribution to liver health.
Ginseng has shown potential as a complementary treatment for managing BP in individuals with HTN. Ginsenosides, particularly Rg1, Rb1 and Rg3, have been found to stimulate nitric oxide production in endothelial cells by activating endothelial nitric oxide synthase. Nitric oxide is a key molecule in promoting vasodilation, leading to a reduction in vascular resistance and BP[Reference Song107]. Some ginsenosides, such as Rg3, have been shown to block calcium channels in vascular smooth muscle cells, leading to muscle relaxation and a decrease in vascular tone[Reference Onanong Kulaputana28]. Ginsenosides may influence the renin-angiotensin-aldosterone system, which plays a central role in BP regulation by controlling fluid balance and vasoconstriction. Studies suggest that ginseng inhibits angiotensin-converting enzyme, leading to decreased levels of angiotensin II, a potent vasoconstrictor. Reduced angiotensin II activity can lower BP by preventing blood vessel narrowing and promoting sodium excretion[Reference Ahn, Hong and Choi31].
Additionally, our results indicate that ginseng supplementation leads to a significant reduction in TC and LDL-cholesterol, among individuals with cardiovascular or hypertensive disease. Ginsenosides also play a role in lipid regulation, which may contribute to their cardiovascular benefits. Studies have shown that ginsenosides influence lipid synthesis and transport by modulating enzymes involved in cholesterol metabolism, such as 3-hydroxy-3-methylglutaryl CoA reductase[Reference Sung, Kang and Jung72]. This modulation leads to reduced levels of TC and LDL-cholesterol.
In our meta-analysis, we did not observe a significant effect of ginseng supplementation on FBS, which aligns with the findings of several previous studies that also reported no significant impact on glycaemic control[Reference Reay, Scholey and Milne103]. Despite these consistent results, some research suggests that ginseng may have a potential role in managing blood glucose levels[Reference Shishtar, Sievenpiper and Djedovic104], indicating that further investigation is warranted to explore the conditions and mechanisms under which ginseng might demonstrate beneficial effects on glucose regulation. Subgroup analysis showed that in the metabolic or glycaemic disorders subgroup, ginseng supplementation resulted in significant decrease in FBS. The mechanisms by which ginseng exerts its hypoglycaemic effects are not entirely understood, but several hypotheses have been proposed. Several studies suggest that ginseng exerts its hypoglycaemic effects through multiple mechanisms, including the stimulation of insulin secretion and improved glucose uptake[Reference Wee, Park, Chung, Benzie and Wachtel-Galor105]. Ginsenosides have been shown to promote glucose-stimulated insulin release from pancreatic β-cells, enhance insulin sensitivity and activate AMP-activated protein kinase, a key regulator of glucose and lipid metabolism[Reference Wee, Park, Chung, Benzie and Wachtel-Galor105,Reference NP106] .
The observed improvements in oxidative stress and inflammatory markers (e.g. hs-CRP, ROS, SOD) align with the broader evidence on natural compounds modulating cardiometabolic pathways. For instance, conjugated linoleic acids have demonstrated anti-inflammatory and antioxidant properties, though their effects on lipid metabolism and body composition remain inconsistent, mirroring the variability seen in ginseng’s impact on lipid profiles[Reference Park, Chung and Chung8]. Similarly, Pycnogenol, a pine bark extract, reduces inflammation via NF-κb inhibition and enhances antioxidant defences – mechanisms analogous to those of ginsenosides – but its efficacy in chronic diseases depends on dosing and population characteristics[Reference Jafari, Kordkatuli and Mardani11]. The null effects of ginseng on glycaemic indices in our analysis parallel findings from studies on royal jelly, where significant reductions in fasting blood glucose were only observed in specific subgroups (e.g. longer interventions or unhealthy populations)[Reference Moher, Shamseer and Clarke10]. This underscores the importance of tailoring supplementation protocols to patient subgroups, as seen in walnut studies, where satiety and lipid benefits were linked to their PUFA and fibre content rather than direct metabolic modulation[Reference Li, Wang and Xu9]. Notably, Zhang et al.[Reference Follmann, Elliott and Suh12] reported significant reductions in SBP and LDL-cholesterol with ginseng, conflicting with our findings, potentially due to differences in study populations (e.g. baseline health status) or ginseng isomer specificity. Collectively, these comparisons highlight that natural products like ginseng exert context-dependent effects, influenced by bioactive compound composition, intervention duration and target populations, necessitating precision in clinical applications.
The observed heterogeneity in ginseng’s efficacy across different population subgroups, particularly the more pronounced cardiovascular benefits in Asian cohorts, might be attributed to several factors. Genetic polymorphisms in drug-metabolising enzymes, especially cytochrome P450 variants like CYP3A4 and CYP2C9, which are known to differ in frequency between Asian and non-Asian populations, may influence the pharmacokinetics of ginsenosides[Reference Munir108]. These genetic variations can affect the biotransformation of ginsenosides to their active metabolites, potentially explaining the enhanced responsiveness observed in Asian populations[Reference Munir108]. Additionally, dietary patterns characteristic of Asian populations, which typically include higher consumption of plant-based foods rich in complementary phytochemicals, may synergistically enhance ginseng’s pharmacological effects through modulation of gut microbiota composition[Reference Lim109]. The gut microbiome plays a crucial role in ginsenoside metabolism, converting primary ginsenosides to more bioavailable and pharmacologically active secondary forms through intestinal bacterial deglycosylation[Reference Chen110]. Differences in intestinal microbiota profiles between ethnic groups may therefore contribute to varying levels of ginsenoside bioactivation and subsequent therapeutic efficacy.
Furthermore, methodological variations and baseline health status differences across studies might explain the discrepancies observed between population groups and health conditions. The significant improvements in lipid profiles (LDL-cholesterol, TC) and BP (DBP) among participants with pre-existing cardiovascular or hypertensive disorders suggest that ginseng’s effects may be more pronounced in individuals with dysregulated cardiometabolic parameters at baseline. This therapeutic threshold effect is consistent with findings for other natural compounds, where clinical benefits are often more evident in populations with established pathophysiological conditions rather than healthy individuals[Reference Wierzejska111]. The duration-dependent effects noted in our subgroup analysis, with certain parameters like hs-CRP, LDL-cholesterol and TC showing significant improvements only after 8 weeks of supplementation, indicate that ginseng’s cardiovascular benefits may require sustained administration to achieve meaningful physiological changes[Reference Follmann, Elliott and Suh12]. This time-dependent response likely reflects the gradual accumulation of bioactive metabolites and the progressive modulation of cellular signalling pathways, particularly those involved in inflammatory cascades and lipid metabolism[Reference Aminifard112]. Additionally, age-related differences in treatment response, evidenced by distinct parameter improvements in younger v. older participants, may be attributed to age-associated variations in pharmacokinetics, receptor sensitivity and baseline inflammatory status[Reference Yuan113].
Strengths and limitations
While our analysis provides valuable insights, it is important to recognise that we did not observe significant effects of ginseng supplementation on several health outcomes. This lack of significant findings may stem from variations in dosing regimens, participant characteristics and underlying health conditions among the studies included in our meta-analysis. However, when we conducted subgroup analyses, we found meaningful associations for certain outcomes, such as SBP and DBP in individuals with CVD. These insights underscore the need for further studies to explore the impact of ginseng supplementation across diverse populations and health conditions. Future investigations should focus on standardising dosing protocols and examining the mechanisms underlying ginseng’s beneficial effects. Long-term studies are also necessary to evaluate the sustained impact of ginseng supplementation on various health outcomes. A key limitation of this meta-analysis was the inability to conduct subgroup analyses based on ginseng species. While this dimension is clinically relevant due to interspecies variability in ginsenoside composition and pharmacological activity, our effort was hindered by insufficient and inconsistent reporting in the primary studies. Notably, approximately half of the included trials did not specify the ginseng species used, and among those that did, variation in extraction methods and lack of standardised ginsenoside profiling further complicated classification. This lack of consistent botanical specification limited the statistical power and interpretability of species-specific comparisons, precluding robust subgroup analysis along this axis. Therefore, although the therapeutic potential of different ginseng species may differ, the current body of evidence does not yet allow for reliable interspecies meta-analytic evaluation.
Despite these limitations, our meta-analysis highlights the promising health benefits of ginseng, advocating for its consideration as a complementary approach in managing diverse health conditions, while emphasising the need for additional research to strengthen the evidence base.
Future research directions
As ginseng supplementation continues to demonstrate potential health benefits, future research should focus on several key areas to enhance our understanding and application of this natural remedy. First, standardisation of dosage and formulations is essential; research should prioritise the establishment of standardised dosages to address variability in dosing across studies, ensuring optimal doses for various health outcomes. Longitudinal studies are also needed to assess the sustained effects of ginseng supplementation on health outcomes, helping to elucidate the duration necessary for ginseng to exert its beneficial effects and identify any potential long-term impacts on health. Additionally, investigating the specific molecular mechanisms through which ginseng exerts its anti-inflammatory and antioxidant effects is crucial, as understanding these pathways could lead to the development of targeted therapies for metabolic and CVD. Future research should also include diverse populations to evaluate the effectiveness of ginseng across different demographic groups, identifying variations in response to supplementation based on genetic, environmental and lifestyle factors.
Exploring the synergistic effects of ginseng with other herbal supplements or conventional medications may provide valuable insights into its therapeutic potential; thus, research should investigate whether combining ginseng with other agents enhances its health benefits or mitigates any adverse effects. Investigating ginseng’s role in specific clinical populations, such as those with chronic diseases or elderly patients, could further support its use as a complementary therapy, while identifying health conditions that may benefit most from supplementation will enhance its applicability in clinical settings.
To enable future evidence syntheses that assess species-specific effects of ginseng on cardiovascular outcomes, researchers should prioritise rigorous reporting of intervention details. Future RCT should systematically document the botanical species used, provide standardised ginsenoside content and describe extraction and preparation methods in line with established guidelines. Clear and consistent taxonomic reporting will not only enhance reproducibility and transparency but also facilitate more granular meta-analyses, including dose–response and network meta-analyses across ginseng types. Such improvements in methodological reporting are essential to optimise the clinical utility and mechanistic understanding of ginseng supplementation in cardiovascular health.
Lastly, understanding how dietary factors and lifestyle choices interact with ginseng supplementation could shed light on its effectiveness, informing comprehensive health strategies that incorporate ginseng within a broader context of lifestyle management. By addressing these research directions, future studies can build on the current body of evidence, ultimately supporting the integration of ginseng as a beneficial supplement in health management and disease prevention.
Conclusions
In conclusion, this meta-analysis suggests that ginseng supplementation may have potential benefits for some inflammatory markers, liver function and oxidative stress parameters. However, results on some studied biomarkers do not support this finding. In addition, the certainty of evidence for all biomarkers is very low-to-moderate. Therefore, our results must be interpreted with caution. Moreover, observed effect sizes are not clinically significant. Consequently, ginseng can only be considered as a complementary approach in managing various health conditions. Further high-quality research is warranted to confirm these findings and to provide more robust evidence regarding the therapeutic applications of ginseng.
Acknowledgements
The authors acknowledge the use of ChatGPT (GPT-4, OpenAI’s large-scale language-generation model) for assistance in editing and improving the writing style of this article. The artificial intelligence (AI) was used solely for language enhancement and not for creating the content or references of the manuscript. The authors carefully reviewed, edited and revised all AI-assisted texts and take full responsibility for the content and accuracy of this article.
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.
Conceptualization: Ali Jafari. Data curation: Ali Jafari, Helia Mardani, Maryam Abbastabar, Fatemeh Mehdipoor, Bahare Parsi Nezhad, Kosar Kordkatuli, Pardis Bakhtiari, Amir Hossein Faghfourid, Vali Musazadeh, Alireza Alaghi. Formal analysis: Ali Jafari, Vali Musazadeh. Investigation: Ali Jafari, Helia Mardani, Maryam Abbastabar, Fatemeh Mehdipoor, Bahare Parsi Nezhad, Kosar Kordkatuli, Pardis Bakhtiari, Amir Hossein Faghfourid, Vali Musazadeh, Alireza Alaghi. Methodology: Ali Jafari, Vali Musazadeh. Project administration: Ali Jafari, Vali Musazadeh. Software: Ali Jafari, Alireza Alaghi, Vali Musazadeh. Visualization: Ali Jafari, Bahare Parsi Nezhad. Supervision: Ali Jafari, Alireza Alaghi, Vali Musazadeh. Validation: Helia Mardani, Maryam Abbastabar, Fatemeh Mehdipoor, Kosar Kordkatuli, Pardis Bakhtiari, Amir Hossein Faghfouri. Writing – original draft: Ali Jafari, Helia Mardani, Maryam Abbastabar, Bahare Parsi Nezhad, Alireza Alaghi. Writing – review & editing: Ali Jafari, Helia Mardani, Fatemeh Mehdipoor, Kosar Kordkatuli, Pardis Bakhtiari, Amir Hossein Faghfourid, Vali Musazadeh.
Ali Jafari holds an editorial position at Systematic Reviews. All other authors of this publication declare that they have no affiliations with, or involvement in, any organization or entity with a financial interest (including honoraria, educational grants, participation in speakers’ bureaus, memberships, employment, consultancies, stock ownership or other equity interests, expert testimony, or patent-licensing arrangements) or a non-financial interest (such as personal or professional relationships, affiliations, knowledge, or beliefs) relevant to the subject matter or materials discussed in this manuscript.
All relevant data are provided within the manuscript and supplementary file. Additionally, data analysed for this study are available upon request from the corresponding author.
As a systematic review and meta-analysis, our manuscript did not necessitate any referral for review to our institutional clinical ethics committee.
Consent to participate: Not applicable.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S0007114525103607.
