Abstract
Background:
Korean red ginseng (KRG) is widely used as a dietary supplement, but its effects on uterine leiomyoma growth and reproductive hormones are unclear. We evaluated whether KRG affects leiomyoma growth, routine laboratory indices, or endocrine profiles in premenopausal women with uterine leiomyoma managed expectantly.
Methods:
In this single-center, double-blind, randomized, placebo-controlled, parallel-group trial, premenopausal women aged 20 to 45 years with a uterine leiomyoma of 1 to 8 cm suitable for expectant management were allocated 1:1 to KRG 3 g/d or matching placebo for 6 months. Participants, investigators, ultrasonography examiners, and analysts were blinded. The primary outcome was the between-group difference in the 6-month change in maximal diameter of the index leiomyoma on ultrasonography, analyzed by analysis of covariance adjusted for baseline diameter. Secondary outcomes were complete blood count, ferritin, liver enzymes, and reproductive hormones.
Results:
Between December 2015 and September 2019, 69 participants were randomized (placebo 34, KRG 35) and 61 completed the 6-month assessment (placebo 30, KRG 31); all 8 noncompletions were losses to follow-up. Maximal diameter increased modestly in both groups, with no significant between-group difference in change (mean difference −0.11 cm; 95% confidence interval: −0.33 to 0.12; P = .34); the upper confidence limit lay below the prespecified target difference of 0.33 cm. There were no significant between-group differences in laboratory changes or in baseline-, age-, and body mass index-adjusted log scale hormone ratios, and the multivariate hormone profile did not differ between groups (Pillai’s trace = 0.056; P = .68). In exploratory post hoc analyses, the internal correlation structure among hormone changes was similar in both arms. No treatment-related serious adverse event and no discontinuation for intolerance was documented, although adverse events were ascertained passively.
Conclusion:
In premenopausal women with uterine leiomyoma managed expectantly, 6 months of KRG supplementation did not meaningfully alter leiomyoma growth, routine laboratory indices, or reproductive hormone homeostasis compared with placebo. Within the constraints of a modest sample size, incomplete characterization of the study product, and passive adverse-event ascertainment, these findings do not establish the safety of KRG, and they support no conclusion regarding nonserious adverse effects.
Keywords: dietary supplements, hormones, Korean red ginseng, randomized controlled trial, uterine leiomyoma
1. Introduction
Uterine leiomyoma is the most common benign tumor of the female reproductive tract and can present with heavy menstrual bleeding, pelvic pain/pressure, infertility, and adverse pregnancy outcomes. Management is individualized and ranges from observation to medical, procedural, and surgical options depending on symptom burden, leiomyoma characteristics, fertility goals, and patient preference.[1] Although the lifetime prevalence of leiomyomas is high, many women are asymptomatic or have mild symptoms and are managed expectantly with periodic monitoring.[1,2] Because leiomyoma growth is influenced by steroid hormones and related signaling pathways, patients frequently ask whether commonly used supplements may affect leiomyoma growth or hormonal profiles.
Korean red ginseng (KRG; steamed and dried Panax ginseng C.A. Meyer) is widely consumed in Korea as a health supplement. Steaming and drying alter the ginsenoside profile and increase low-polarity and rare ginsenosides that may have distinct biologic activities.[3,4] Certain ginsenosides have been reported to interact with estrogen receptor signaling in vitro, raising theoretical concerns in estrogen-responsive conditions such as leiomyoma.[5,6] At the same time, systematic reviews and large placebo-controlled trials in other populations have generally reported favorable safety and tolerability of Panax ginseng/KRG, although most trials are not specific to women with leiomyoma.[7,8]
Epidemiologic studies have suggested associations between traditional herbal products and fibroid-related outcomes. For example, population-based cohort studies from Taiwan reported that the use of Chinese herbal products was associated with a decreased risk of incident uterine leiomyomas or leiomyoma-related surgery, and Panax ginseng was among the frequently used components in some analyses.[9,10] However, these observational findings may be influenced by confounding and do not establish whether KRG affects leiomyoma growth under controlled conditions.
Therefore, we conducted a double-blind, randomized, placebo-controlled trial to evaluate whether daily KRG supplementation for 6 months affects leiomyoma size and selected laboratory and hormonal parameters in premenopausal women with uterine leiomyoma managed expectantly.
2. Methods
2.1. Participants
This trial was conducted between December 2015 and September 2019 at the Department of Obstetrics and Gynecology, Gangnam Severance Hospital, Yonsei University College of Medicine, Seoul, Republic of Korea. Eligible participants were premenopausal women aged 20 to 45 years with uterine leiomyoma measuring 1 to 8 cm who were suitable for expectant management. At enrollment, none had urgent indications for pharmacologic therapy, surgery, or uterine artery embolization, and standard treatment options were discussed and deferred based on symptom severity, fertility goals, and patient preference. Women with adenomyosis or endometrial polyps were excluded. Patients were also excluded if they were pregnant, lactating, required immediate leiomyoma-directed treatment, or had major medical conditions that could affect study outcomes.
2.2. Ethical approval and trial registration
The protocol was approved by the Institutional Review Board of Gangnam Severance Hospital (approval No. 3-2012-0230), and written informed consent was obtained from all participants before any study procedures. This trial was registered retrospectively in the Clinical Research Information Service (CRiS; registration No. KCT0011687), a primary registry of the WHO International Clinical Trials Registry Platform, on 6 March 2026 – that is, after completion of both enrollment (December 2015–September 2019) and follow-up. Because the trial was not publicly registered before enrollment, however, prespecification of the outcomes and analyses cannot be independently verified, and the possibility of outcome-reporting bias cannot be formally excluded. The Consolidated Standards of Reporting Trials (CONSORT) 2025 statement was used as a reporting framework for this manuscript, and its use does not imply prospective registration. The endocrine network analyses reported in Section 3.4 were post hoc, exploratory, and were not part of the approved protocol.
2.3. Study design, randomization, and intervention
This study was designed as a double-blind, randomized, placebo-controlled, parallel-group trial and is reported in accordance with the CONSORT 2025 statement, which updates prior CONSORT guidance for parallel-group randomized trials.[11] Participants were randomized in a 1:1 ratio to KRG or placebo using a computer-generated random allocation sequence created before enrollment by a study team member. To ensure allocation concealment, study products were packaged as identical units and labeled with sequential study codes according to the allocation list, and participants were assigned the next available code in order of enrollment. Participants, clinical investigators, ultrasonography assessors, and data analysts remained blinded to treatment allocation until completion of follow-up and database lock.
KRG and placebo capsules were provided by the Korea Ginseng Corporation (Daejeon, Korea). Each KRG capsule contained 500 mg of red ginseng with a stated total ginsenoside content of 10 mg per capsule (manufacturer specification). Because the study products were manufactured as a historical lot, a lot-specific certificate of analysis and detailed HPLC-based ginsenoside profile could not be retrieved; therefore, we report the manufacturer-provided specification for total ginsenosides and describe the product source, dose, and administration schedule in detail. The daily dose corresponded to 3 g of KRG (approximately 60 mg/d of total ginsenosides based on the manufacturer specification).
The KRG group took 2 capsules 3 times daily (total 3 g/d) for 6 months, while the placebo group took identically appearing capsules for 6 months. Medication adherence and adverse events were assessed during follow-up visits, including a mid-study visit at approximately 3 months, and participants were advised to avoid initiating new fibroid-directed therapies during the study period whenever clinically feasible.
2.4. Sample size
The sample size was determined a priori for the prespecified primary endpoint: the between-group difference in the 6-month change in maximal diameter (cm) of the index leiomyoma measured by 2-dimensional ultrasonography.[11] The target difference was selected following the DELTA2 framework, which recommends that the difference used in a sample size calculation be 1 that is both important to patients and clinicians and realistic for the intervention under study.[12] Two considerations were combined. First, in a leiomyoma of 4 to 5 cm – the size range expected in a population eligible for expectant management – a difference of 0.33 cm corresponds to approximately 7% of the maximal diameter and, because volume is proportional to the cube of diameter, to approximately a 20% to 25% difference in estimated volume. A change of 20% in volume is the threshold conventionally used to define meaningful growth or spontaneous regression in imaging-based natural-history studies of leiomyoma.[13,14] Second, a target difference must be distinguishable from measurement variability. Reproducibility data for 2-dimensional ultrasonography indicate a mean intraobserver coefficient of variation of approximately 5.9% for leiomyoma diameter, and increases of up to approximately 20% in diameter may reflect measurement error rather than true growth.[15] A target difference of 0.33 cm therefore represents the smallest between-group difference that we judged to be both clinically interpretable and reliably distinguishable from short-term measurement variability.
Assuming a common standard deviation of 0.45 cm for the 6-month change in maximal diameter – a conservative assumption adopted in the absence of directly applicable prior data – a two-sided α of 0.05, 80% power, and 1:1 allocation, 60 evaluable participants (30 per group) were required. Allowing for up to 15% attrition, we planned to randomize 70 participants during the prespecified recruitment window.
2.5. Measurements
Baseline anthropometrics (height, weight, and body mass index [BMI]) were obtained at the initial visit. Blood tests included complete blood count (CBC), serum ferritin, liver enzymes (serum glutamic oxaloacetic transaminase [SGOT], serum glutamic pyruvic transaminase [SGPT], and gamma-glutamyl transferase [gamma-GT]), and a steroid hormone profile (follicle-stimulating hormone [FSH], luteinizing hormone [LH], estradiol [E2], testosterone, and sex hormone-binding globulin [SHBG]). Blood samples were collected at baseline (0 months) and at 6 months; steroid hormones were measured in the mid-follicular phase when feasible.
The primary efficacy outcome was changes in the maximal diameter of the index leiomyoma from baseline to 6 months. Leiomyoma size was assessed by transvaginal sonography or transrectal sonography and measured as the largest diameter on sagittal and coronal sections; in cases with multiple leiomyomas, the largest leiomyoma at baseline was defined as the index leiomyoma and was followed longitudinally. Maximal diameter, rather than estimated volume, was selected as the primary measure for 3 reasons. First, the largest diameter is the measurement routinely documented in clinical ultrasound reports and is the metric on which counseling and management thresholds are based in everyday practice, so a treatment effect expressed in these terms is directly interpretable by clinicians. Second, ellipsoid volume derived from 2-dimensional ultrasonography propagates measurement error across 3 orthogonal diameters, given the intraobserver coefficient of variation reported for this modality;[15] a volume-based endpoint would have been appreciably noisier than a diameter-based endpoint at the sample size available. Third, diameter permits direct comparison with the imaging-based natural-history literature used to contextualize the present findings. Only the maximal diameter of the index leiomyoma was recorded in this trial; 3 orthogonal diameters were not captured, and a volume estimate could therefore not be derived. Ultrasound examinations were performed by 2 examiners, both of whom remained blinded to treatment allocation throughout the study and who followed a common measurement protocol. Leiomyoma anatomical type was recorded according to the International Federation of Gynecology and Obstetrics (FIGO) leiomyoma subclassification system.[16] For the between-group comparison, FIGO types were collapsed into 3 prespecified categories defined by the extent of intramural involvement: types 0 to 1 (<50% intramural, submucosal aspect), types 2 to 5 (50% or more intramural), and types 6 to 8 (<50% intramural on the subserosal aspect, or other). For descriptive purposes, the index leiomyoma is additionally reported using the conventional FIGO anatomical designations: submucosal (types 0–2), intramural (types 3–4), subserosal (types 5–7), and other (type 8).
Safety outcomes included changes in laboratory parameters and clinical adverse events reported during follow-up, including any need for additional fibroid-related treatment. Adverse events were assessed at each follow-up visit through open-ended questioning and review of medical records. Serious adverse events were defined as events resulting in hospitalization, surgery, or significant disability and were recorded throughout the study period. No structured symptom checklist, standardized severity or causality grading, or dedicated adverse-event case report form was used. Documentation was limited to the occurrence or absence of serious adverse events. Nonserious events were not recorded in a form permitting categorical tabulation.
2.6. Statistical analysis
Continuous variables are presented as mean ± standard deviation or median (range), and categorical variables as number (percentage). FIGO categories (0–1, 2–5, and 6–8) were compared between groups using the Fisher–Freeman–Halton exact test due to small expected cell counts. Within-group pre-post shifts were explored by collapsing FIGO categories into 0 to 1 versus 2 to 8 and applying McNemar’s exact test. Analyses were conducted in a modified intention-to-treat population defined as all randomized participants with available baseline and 6-month outcome data (complete-case analysis).
Baseline characteristics were compared using the independent t-test or the Mann–Whitney U test, as appropriate. For the primary outcome, between-group differences were assessed using analysis of covariance (ANCOVA) with 6-month maximal diameter as the dependent variable, treatment group as the fixed factor, and baseline maximal diameter as a covariate; change scores are additionally presented for descriptive purposes. Within-group changes from baseline to 6 months in raw laboratory variables were assessed using paired t tests, and between-group differences in raw change scores were assessed using independent t-tests.
Because hormone distributions were right-skewed, hormone analyses were additionally performed on the log scale. Within-group hormonal change is presented as the geometric mean ratio (GMR; posttreatment/pretreatment) with 95% confidence intervals (CIs). Between-group hormonal effects were estimated using ANCOVA models for log-transformed 6-month values adjusted for log-transformed baseline value, age, and BMI, and are reported as adjusted KRG/placebo ratios with 95% CIs. The overall multivariate treatment effect across the 5 hormones was assessed using multivariate analysis of variance. Exploratory hormone network analyses were performed using Pearson correlations of log-change values within each treatment group and canonical correlation analysis between the gonadotropin block (Δlog FSH, Δlog LH) and the steroid/SHBG block (Δlog E2, Δlog testosterone, Δlog SHBG) after residualizing all variables for age and BMI; significance was assessed with 5000 permutations.
Safety outcomes (laboratory parameters and reported adverse events) were summarized descriptively. Endocrine network analyses – principal component analysis (PCA) of log-change hormone profiles, within-group Pearson correlations, and canonical correlation analysis – were conducted post hoc and were not prespecified.
All tests were 2-sided, and P < .05 was considered statistically significant. All statistical analyses were performed using R version 4.4.2 (R Foundation for Statistical Computing, Vienna, Austria).
3. Results
3.1. Trial flow
A total of 69 patients who met the eligibility criteria were randomly assigned to double-blind treatment with KRG or placebo. Among them, 61 completed the 6-month intervention and had both baseline and 6-month outcome data (placebo n = 30; KRG n = 31); these participants were included in the primary analysis (Fig. 1). Eight participants (11.6%) did not complete the 6-month assessment, balanced between arms (placebo, n = 4; KRG, n = 4). All 8 were lost to follow-up. No participant discontinued because of an adverse event, intolerance of the study product, or initiation of leiomyoma-directed treatment. Because noncompletion was balanced between arms and attributable to a single reason unrelated to the study product or to the outcome, the missing data were judged to be plausibly missing completely at random.
Figure 1.

CONSORT flow diagram (69 randomized; 61 analyzed). CONSORT = Consolidated Standards of Reporting Trials.
3.2. Baseline characteristics and leiomyoma outcomes
Table 1 shows the baseline characteristics of the study participants. Baseline age, BMI, gravidity, parity, hemoglobin level, and maximal leiomyoma diameter were comparable between the placebo and KRG groups (e.g., age 37.93 ± 3.92 vs 38.84 ± 4.06 years, P = .379; maximal diameter 4.21 ± 1.76 vs 4.78 ± 1.60 cm, P = .192).
Table 1.
Baseline characteristics of the study participants.
| Variables* | Placebo group | KRG group | P-value |
|---|---|---|---|
| (n = 30) | (n = 31) | ||
| Age (yr) | 37.93 ± 3.92 | 38.84 ± 4.06 | .379 |
| BMI (kg/m2) | 23.57 ± 2.25 | 23.67 ± 2.51 | .876 |
| Gravidity | 0 (0–3) | 0 (0–3) | .567 |
| Parity | 0 (0–2) | 0 (0–2) | .760 |
| Baseline hemoglobin (g/dL) | 12.63 ± 1.59 | 12.23 ± 1.16 | .270 |
| Anemia (hemoglobin < 12 g/dL), n (%) | 8 (26.7) | 12 (38.7) | .416† |
| Maximal diameter of index leiomyoma (cm) | 4.21 ± 1.76 | 4.78 ± 1.60 | .192 |
| Anatomical type of index leiomyoma, by extent of intramural involvement, n (%) | 1.000‡ | ||
| FIGO type 0–1 (< 50% intramural) | 7 (23.3) | 8 (25.8) | |
| FIGO type 2–5 (≥ 50% intramural) | 22 (73.3) | 22 (71.0) | |
| FIGO type 6–8 (< 50% intramural, subserosal aspect, or other) | 1 (3.3) | 1 (3.2) | |
| Anatomical type of index leiomyoma, conventional grouping, n (%) | .710‡ | ||
| Submucosal (FIGO type 0–2) | 16 (53.3) | 20 (64.5) | |
| Intramural (FIGO type 3–4) | 13 (43.3) | 10 (32.3) | |
| Subserosal (FIGO type 5–7) | 1 (3.3) | 1 (3.2) |
Anatomical type is presented in 2 complementary ways: by the extent of intramural involvement (types 0–1, <50% intramural on the submucosal aspect; types 2–5, ≥50% intramural; types 6–8, <50% intramural on the subserosal aspect, or other), the prespecified grouping used for the between-group comparison; and by the conventional FIGO anatomical designations (submucosal, types 0–2; intramural, types 3–4; subserosal, types 5–7) for clinical interpretability. No participant had a type 8 leiomyoma. Where multiple leiomyomas were present, the type of the index (largest) leiomyoma is reported.
BMI = body mass index, FIGO = International Federation of Gynecology and Obstetrics, KRG = Korean red ginseng.
Values are presented as mean ± standard deviation, median (range), or n (%). Between-group comparisons by independent t test or Mann–Whitney U test for continuous variables.
Fisher exact test.
Fisher–Freeman–Halton exact test.
Baseline FIGO category distribution did not differ between groups (0–1/2–5/6–8: 23.3%/73.3%/3.3% in placebo vs 25.8%/71.0%/3.2% in KRG; Fisher–Freeman–Halton exact P = 1.000). At 6 months, FIGO category distribution remained similar between groups (20.0%/76.7%/3.3% vs 19.4%/77.4%/3.2%; exact P = 1.000), and within-group shifts were infrequent (McNemar exact P = 1.000 for placebo and P = .500 for KRG). By the conventional anatomical grouping, the index leiomyoma was submucosal (FIGO types 0–2) in 16 of 30 participants (53.3%) in the placebo group and 20 of 31 (64.5%) in the KRG group, intramural (types 3–4) in 13 (43.3%) and 10 (32.3%), and subserosal (types 5–7) in 1 (3.3%) and 1 (3.2%), respectively, with no between-group difference (Fisher–Freeman–Halton exact P = .71). No participant had a type 8 leiomyoma. Anemia (hemoglobin below 12 g/dL) was present at baseline in 8 participants (26.7%) in the placebo group and 12 (38.7%) in the KRG group (P = .42).
Figure 2 shows the maximal diameter of the index leiomyoma at baseline and 6 months. In the placebo group, maximal diameter increased from 4.21 ± 1.76 cm to 4.59 ± 1.95 cm (paired P < .001), and in the KRG group it increased from 4.78 ± 1.60 cm to 5.05 ± 1.83 cm (paired P < .001). The between-group difference in change was not statistically significant (mean difference −0.11 cm; 95% CI: −0.33 to 0.12; P = .337).
Figure 2.

Maximal diameter of the index leiomyoma before and after 6 months of supplementation. (A) Placebo group: 4.21 ± 1.76 cm versus 4.59 ± 1.95 cm (paired P < .001). (B) KRG group: 4.78 ± 1.60 cm versus 5.05 ± 1.83 cm (paired P < .001). The between-group difference in change was not significant (P = .34). KRG = Korean red ginseng.
3.3. Laboratory and hormone outcomes
Table 2 summarizes raw changes in CBC, liver enzymes, ferritin, and steroid hormone concentrations from baseline to 6 months. There were no statistically significant between-group differences in changes in hemoglobin/hematocrit, ferritin, liver enzymes, or sex hormone parameters (FSH, LH, E2, testosterone, and SHBG) over 6 months (all P > .05). Liver enzyme values remained within clinically acceptable ranges in both groups throughout follow-up, supporting the absence of laboratory signals of hepatotoxicity in this cohort.
Table 2.
Changes in complete blood count, ferritin, liver enzymes, and raw steroid hormone concentrations before and after treatment.
| Variables | Placebo group (n = 30) | Within-group P | KRG group (n = 31) | Within-group P | Between-group P* |
|---|---|---|---|---|---|
| Hemoglobin (g/dL) | Baseline: 12.63 ± 1.59 Endpoint: 12.44 ± 1.70 |
.20 | Baseline: 12.23 ± 1.16 Endpoint: 11.96 ± 1.43 |
.21 | .76 |
| Hematocrit (%) | Baseline: 37.64 ± 4.02 Endpoint: 37.39 ± 4.16 |
.50 | Baseline: 36.83 ± 3.26 Endpoint: 36.28 ± 3.79 |
.32 | .66 |
| Ferritin (ng/mL) | Baseline: 49.42 ± 38.09 Endpoint: 51.35 ± 36.06 |
.53 | Baseline: 49.43 ± 38.42 Endpoint: 49.73 ± 38.21 |
.84 | .63 |
| AST (IU/L) | Baseline: 22.07 ± 6.14 Endpoint: 23.97 ± 6.83 |
.054 | Baseline: 21.00 ± 6.57 Endpoint: 21.06 ± 7.80 |
.94 | .15 |
| ALT (IU/L) | Baseline: 17.70 ± 7.47 Endpoint: 18.60 ± 6.61 |
.17 | Baseline: 17.52 ± 7.69 Endpoint: 17.58 ± 6.77 |
.91 | .32 |
| GGT (IU/L) | Baseline: 14.50 ± 5.84 Endpoint: 14.47 ± 5.03 |
.96 | Baseline: 15.77 ± 11.47 Endpoint: 14.39 ± 5.41 |
.38 | .43 |
| FSH (mIU/mL) | Baseline: 8.97 ± 9.36 Endpoint: 7.03 ± 3.03 |
.25 | Baseline: 10.14 ± 13.27 Endpoint: 10.00 ± 9.33 |
.95 | .53 |
| LH (mIU/mL) | Baseline: 8.70 ± 5.63 Endpoint: 7.45 ± 4.46 |
.12 | Baseline: 10.53 ± 7.19 Endpoint: 9.27 ± 3.82 |
.38 | >.99 |
| Estradiol (pg/mL) | Baseline: 141.06 ± 215.91 Endpoint: 113.15 ± 82.79 |
.48 | Baseline: 119.64 ± 115.63 Endpoint: 94.92 ± 79.83 |
.27 | .94 |
| Testosterone (ng/mL) | Baseline: 0.55 ± 0.46 Endpoint: 0.53 ± 0.42 |
.68 | Baseline: 0.52 ± 0.52 Endpoint: 0.47 ± 0.39 |
.30 | .72 |
| SHBG (nmol/L) | Baseline: 99.55 ± 50.92 Endpoint: 103.61 ± 45.98 |
.35 | Baseline: 96.30 ± 41.98 Endpoint: 94.26 ± 39.15 |
.52 | .25 |
Values are mean ± standard deviation. Within-group comparison between baseline and 6 months by paired t test.
ALT = alanine aminotransferase, AST = aspartate aminotransferase, FSH = follicle-stimulating hormone, GGT = gamma-glutamyl transferase, KRG = Korean red ginseng, LH = luteinizing hormone, SHBG = sex hormone-binding globulin.
Between-group comparison of raw change values by independent t test.
Because hormone distributions were right-skewed, hormone changes were additionally analyzed on the log scale and are summarized in Table 3. Within-group GMRs (posttreatment/pretreatment) were 0.90 for FSH, 0.88 for LH, 1.10 for E2, 1.03 for testosterone, and 1.07 for SHBG in the placebo group, and 1.03, 0.96, 0.83, 0.99, and 0.99, respectively, in the KRG group. Among the within-group analyses, only SHBG in the placebo group showed a marginal increase (GMR 1.07; 95% CI: 1.00–1.14; P = .047).
Table 3.
Hormone changes after intervention.
| Hormone | Placebo, GMR (95% CI) | Within-group P | KRG, GMR (95% CI) | Within-group P | Adjusted KRG/placebo ratio (95% CI) | Between-group P |
|---|---|---|---|---|---|---|
| FSH | 0.90 (0.72–1.11) | .33 | 1.03 (0.84–1.25) | .80 | 1.18 (0.92–1.50) | .18 |
| LH | 0.88 (0.73–1.06) | .19 | 0.96 (0.77–1.19) | .72 | 1.21 (0.96–1.54) | .11 |
| Estradiol | 1.10 (0.79–1.52) | .59 | 0.83 (0.63–1.09) | .19 | 0.81 (0.57–1.13) | .21 |
| Testosterone | 1.03 (0.84–1.27) | .74 | 0.99 (0.81–1.21) | .90 | 0.95 (0.73–1.22) | .67 |
| SHBG | 1.07 (1.00–1.14) | .047 | 0.99 (0.93–1.05) | .69 | 0.93 (0.86–1.01) | .079 |
GMR denotes the posttreatment/pretreatment ratio. The between-group ratio was estimated from the analysis of covariance model log(posttreatment value) ~ group + log(baseline value) + age + body mass index and is reported as the adjusted KRG/placebo ratio. The overall multivariate treatment effect across the 5 hormones was assessed using multivariate analysis of variance (Pillai’s trace = 0.056, P = .68).
CI = confidence interval, FSH = follicle-stimulating hormone, GMR = geometric mean ratio, KRG = Korean red ginseng, LH = luteinizing hormone, SHBG = sex hormone-binding globulin.
In ANCOVA models adjusted for baseline value, age, and BMI, adjusted KRG/placebo ratios were 1.18 for FSH (95% CI: 0.92–1.50; P = .182), 1.21 for LH (95% CI: 0.96–1.54; P = .110), 0.81 for E2 (95% CI: 0.57–1.13; P = .209), 0.95 for testosterone (95% CI: 0.73–1.22; P = .671), and 0.93 for SHBG (95% CI: 0.86–1.01; P = .079). Multivariate analysis of variance showed no significant overall treatment effect on the combined hormone profile (Pillai’s trace = 0.056, P = .680).
3.4. Exploratory post hoc endocrine network analyses
The following analyses were post hoc and are reported as hypothesis-generating. PCA of the log-change hormone profile yielded a first component explaining 51.3% and a second explaining 16.2% of the variance, with no separation between treatment groups (see Fig. S1, Supplemental Digital Content 1, which shows the PCA and canonical coupling of hormone change scores). Within-group correlations among hormone change scores were similar in the 2 arms, with Δlog FSH positively correlated with Δlog LH (placebo r = 0.74; KRG r = 0.64) and inversely correlated with Δlog E2 (placebo r = −0.49; KRG r = −0.70; see Fig. S2, Supplemental Digital Content 2, which shows group-specific correlation heatmaps of hormone change scores). Canonical correlation between the gonadotropin block (Δlog FSH, Δlog LH) and the steroid/SHBG block (Δlog E2, Δlog testosterone, Δlog SHBG), adjusted for age and BMI, was significant in each arm (placebo r = 0.66, permutation P = .02; KRG r = 0.77, permutation P = .002; see Table S1, Supplemental Digital Content 3, which summarizes network relationships among hormone changes). These findings are consistent with an intact reproductive endocrine network under both conditions but do not constitute evidence of a treatment effect.
3.5. Safety and tolerability
No treatment-related serious adverse event was documented in either group during the 6-month study period. No participant discontinued the study product. All 8 noncompletions were losses to follow-up rather than withdrawals for intolerance or adverse events, and no participant required leiomyoma-directed treatment during the study period. Objective laboratory indices remained within clinically acceptable ranges in both groups, with no significant between-group differences in the change in hemoglobin, hematocrit, ferritin, aspartate aminotransferase, alanine aminotransferase, or gamma-glutamyl transferase (Table 2). Because adverse events were ascertained passively and nonserious events were not recorded in a structured format, these observations should be interpreted as the absence of documented serious harm rather than as evidence of overall tolerability.
4. Discussion
In this double-blind, randomized, placebo-controlled trial of premenopausal women with uterine leiomyoma managed without immediate intervention, 6 months of KRG supplementation did not significantly change the course of leiomyoma growth compared with placebo. Although maximal diameter increased modestly within both groups over 6 months, the between-group difference in change was small and not statistically significant. Importantly, the observed pattern of variable growth over time is consistent with the natural history of leiomyomas reported in imaging-based longitudinal studies, in which many tumors increase in size over months while others remain stable or regress.[13,14]
The magnitude of growth observed here can be placed against published natural history. In longitudinal imaging studies of untreated premenopausal women, the median leiomyoma growth rate is approximately 9% in volume per 6 months, with a wide distribution ranging from approximately −89% to +138%; roughly one third of tumors increase by more than 20% in volume, and approximately 7% regress by more than 20% over the same interval.[13,14,17] In the present trial, mean maximal diameter increased by 0.38 cm in the placebo group and by 0.27 cm in the KRG group over 6 months, corresponding to relative increases of approximately 9% and 6% in diameter. Growth under placebo therefore lies within the range expected from natural history, which supports the internal validity of the comparison and argues against systematic measurement drift. Of note, the upper limit of the 95% CI for the between-group difference (0.12 cm) lies well below the prespecified target difference of 0.33 cm; an acceleration of leiomyoma growth attributable to KRG of a magnitude defined in advance as clinically meaningful is therefore unlikely. The lower limit (−0.33 cm) coincides with the target difference, so a modest KRG-associated reduction in growth cannot be excluded, and establishing or refuting such an effect would require a substantially larger trial.
From a mechanistic perspective, ginseng contains multiple bioactive ginsenosides whose relative composition changes with steaming and drying (red ginseng processing).[3,4] Some ginsenosides have been reported to interact with estrogen receptor signaling in vitro, suggesting a theoretical pathway by which ginseng could influence hormone-responsive tissues.[5,6] However, in the present trial, KRG supplementation did not produce clinically meaningful between-group differences in circulating gonadotropins, E2, testosterone, or SHBG, supporting the interpretation that this regimen does not substantially perturb systemic sex-steroid homeostasis in premenopausal women with leiomyoma. Human pharmacokinetic studies also suggest that systemic exposure to individual ginsenosides after oral KRG is limited and variable, which may partially explain the absence of detectable endocrine effects at commonly used doses.[18]
In post hoc exploratory analyses, the correlation structure among hormone change scores – particularly the positive coupling of Δlog FSH with Δlog LH and its inverse coupling with Δlog E2 – was similar in both arms, and age- and BMI-adjusted canonical correlations between the gonadotropin and steroid/SHBG blocks were significant in each arm. These observations are consistent with an intact reproductive endocrine network under both conditions, but they were not prespecified, were not corrected for multiplicity, and are secondary to the primary clinical outcome. They should be regarded as hypothesis-generating only. This interpretation is broadly consistent with a 2022 systematic review of randomized placebo-controlled trials in menopausal women, which reported no consistent direct effects of ginseng on sex hormones, biomarkers, or endometrial thickness.[19]
Our findings should also be interpreted in the context of prior observational evidence. Large population-based cohort studies from Taiwan reported that the use of Chinese herbal products was associated with a decreased risk of incident leiomyomas and/or subsequent fibroid-related surgery, and Panax ginseng was among the commonly prescribed components in some analyses.[9,10] These data are hypothesis-generating, but the endpoints (incident diagnosis or surgery) differ from the present trial’s focus on short-term growth of existing fibroids, and residual confounding is difficult to exclude in real-world prescription datasets. Taken together, the observational literature and the current randomized trial suggest that ginseng exposure is unlikely to be strongly pro-growth for uterine leiomyomas, but further studies are needed to reconcile differences in study design, populations, and outcomes.
Symptoms such as heavy menstrual bleeding, pain, and quality of life are the most clinically meaningful outcomes in uterine leiomyomas, and guideline-based management is driven primarily by symptom burden and patient goals.[1] In this study, women with severe bleeding or an immediate need for leiomyoma-directed treatment were excluded, and the trial focused on imaging, laboratory, and hormonal outcomes. Consequently, we cannot determine whether KRG affects symptom trajectories, and future trials should incorporate validated symptom and quality-of-life instruments as co-primary or key secondary endpoints, consistent with current guidance on outcome reporting in randomized trials.[1,20] The cohort comprised premenopausal women with leiomyomas of moderate size (mean maximal diameter 4.2 and 4.8 cm) who were eligible for expectant management and had no urgent indication for intervention. Validated symptom and quality-of-life instruments were not administered, and structured data on menstrual bleeding severity, pain, and bulk symptoms are not available. Baseline hemoglobin and the prevalence of anemia are therefore reported as objective surrogates. Our findings are consequently applicable to the counseling of asymptomatic or mildly symptomatic women considering KRG under expectant management and should not be extrapolated to women with a substantial symptom burden.
Several limitations warrant discussion. First, the trial was registered retrospectively, more than 6 years after the end of enrollment. Although the study proceeded under institutional review board approval and an a priori protocol that specified the primary outcome and the sample size assumption, retrospective registration means that readers cannot independently confirm that the reported outcomes and analyses were those originally planned. This is a recognized threat to the credibility of a randomized trial, and our findings should be weighed accordingly.
Second, the intervention was incompletely characterized. The capsules were supplied by the Korea Ginseng Corporation and manufactured under Korean Good Manufacturing Practice standards, but the manufacturer’s specification of 10 mg total ginsenosides per 500 mg capsule (approximately 60 mg/d at the study dose of 3 g/d) is the only compositional information available to us. Because the products were drawn from a lot manufactured more than a decade before this analysis, neither a lot-specific certificate of analysis nor an HPLC-based individual ginsenoside profile could be retrieved. Ginsenoside composition varies with cultivar, root age, steaming temperature, and extraction method, and the biological activities attributed to red ginseng – including the estrogen-receptor interactions that motivated this trial – are ascribed to specific ginsenosides rather than to total ginsenoside content. In the absence of a lot-specific profile, our null finding cannot be confidently generalized to KRG preparations with a different composition, and the trial cannot be exactly replicated. We recommend that future supplement trials archive a retention sample and report a lot-specific HPLC profile as a minimum standard for the intervention description.
With respect to safety, we monitored laboratory parameters relevant to bleeding and organ function (CBC, ferritin, and liver enzymes) as well as sex steroid-related measures. No significant between-group differences in laboratory changes were observed over 6 months. These findings are consistent with prior systematic reviews and large placebo-controlled trials in other populations that reported good tolerability of Panax ginseng/KRG.[7,8] Nevertheless, laboratory monitoring alone cannot fully characterize clinical safety. Nonserious events were not recorded in a form permitting categorical tabulation. Mild, transient, and subjective events – precisely those most relevant to the tolerability of a dietary supplement – were therefore very likely under-ascertained, and the absence of documented events cannot be read as evidence that none occurred. In addition, with 61 evaluable participants, the trial could only have detected adverse events occurring at a frequency of approximately 5% or higher. What this trial can support is that no serious adverse event, treatment discontinuation, or clinically significant laboratory abnormality was documented over 6 months. Systematic adverse-event collection and longer follow-up are required to exclude rare harms.
This was a single-center trial with a modest sample size and randomized enrollment slightly below the planned target, which may have limited power to detect small between-group differences, uncommon adverse events, or differences in covariance structure. Baseline data for the 8 participants who did not complete follow-up were not retained in the analysis dataset, and a multiple-imputation sensitivity analysis was therefore not possible. Attrition was nevertheless low, balanced between arms, and attributable in every case to loss to follow-up rather than to the study product or to the outcome, which makes a substantial bias from missing data unlikely though not impossible.
Although hormone sampling was scheduled in the mid-follicular phase when feasible, residual confounding related to cycle timing, sampling time, smoking, thyroid status, insulin resistance, and concomitant medication use could not be fully addressed. The source dataset did not contain detailed assay information or measures such as free testosterone or albumin, limiting interpretation of SHBG-related changes.
In addition, intraobserver and interobserver reproducibility of the ultrasound measurements were not assessed locally. Two blinded examiners performed the examinations, and although a common measurement protocol was followed, between-examiner variability may have contributed to the observed variance. In setting the target difference, we therefore relied on published reproducibility estimates for 2-dimensional ultrasound measurement of leiomyoma diameter[15] rather than on locally derived data.
Finally, leiomyoma burden was assessed using maximal diameter rather than volume. Although this measure was selected deliberately for the reasons set out in Section 2.5, it does not capture asymmetric growth, and because 3 orthogonal diameters were not recorded, a volumetric sensitivity analysis was not possible. Volumetric or magnetic resonance imaging-based assessment would better characterize changes in irregularly shaped tumors.[13,17] In addition, although the target difference of 0.33 cm was anchored both to the natural-history literature and to the reproducibility of the measurement, it reflects clinical and investigator judgment rather than a patient-derived estimate of what constitutes an important difference.
5. Conclusion
In premenopausal women with uterine leiomyoma managed expectantly, 6 months of KRG supplementation did not meaningfully alter leiomyoma growth, average concentrations of FSH, LH, E2, testosterone, or SHBG, or key laboratory parameters compared with placebo. Exploratory post hoc network analyses likewise did not suggest disruption of the reproductive endocrine network. These results may inform provisional counseling of women with leiomyoma who are considering KRG, but they do not constitute definitive evidence of safety. Larger, prospectively registered trials with a fully characterized study product, structured adverse-event collection, validated symptoms and quality-of-life outcomes, and longer follow-up are warranted.
Author contributions
Conceptualization: Byung Seok Lee.
Data curation: Yunjeong Park, Heesoo Chung, Jinok Jun, Joo Hyun Park.
Formal analysis: Hyewon Hur, Yunjeong Park, Joo Hyun Park.
Funding acquisition: Byung Seok Lee.
Investigation: SiHyun Cho, Heesoo Chung, Jinok Jun.
Project administration: Byung Seok Lee.
Writing – original draft: Hyewon Hur, Yunjeong Park, Joo Hyun Park.
Writing – review & editing: Hyewon Hur, Heesoo Chung, Jinok Jun, SiHyun Cho, Young Sik Choi, Byung Seok Lee.
Abbreviations:
- ANCOVA
- analysis of covariance
- BMI
- body mass index
- CBC
- complete blood count
- CI
- confidence interval
- CONSORT
- Consolidated Standards of Reporting Trials
- E2
- estradiol
- FIGO
- International Federation of Gynecology and Obstetrics
- FSH
- follicle-stimulating hormone
- GMR
- geometric mean ratio
- KRG
- Korean red ginseng
- LH
- luteinizing hormone
- PCA
- principal component analysis
- SHBG
- sex hormone-binding globulin
This work was supported by the Korean Society of Ginseng through a research grant funded by the Korea Ginseng Corporation, which also supplied the study product and matching placebo. The funder had no role in the design of the study; in the collection, analysis, or interpretation of data; in the writing of the manuscript; or in the decision to submit the work for publication.
The authors have no conflicts of interest to declare.
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
Supplemental Digital Content is available in the online version of this article (http://dx.doi.org/10.1097/MD.0000000000051173).
How to cite this article: Hur H, Park Y, Chung H, Jun J, Cho S, Choi YS, Lee BS, Park JH. Effects and safety of Korean red ginseng on the clinical course of uterine leiomyoma: A double-blind randomized placebo-controlled trial. Medicine 2026;105:45(e51173).
Contributor Information
Hyewon Hur, Email: hwhur201@yuhs.ac.
Yunjeong Park, Email: pyj11@yuhs.ac.
Heesoo Chung, Email: chungheesu@yuhs.ac.
Jinok Jun, Email: jinokjun@yuhs.ac.
SiHyun Cho, Email: sihyuncho@yuhs.ac.
Young Sik Choi, Email: yschoi08@yuhs.ac.
Byung Seok Lee, Email: dr222@yuhs.ac.
References
- [1].American College of Obstetricians and Gynecologists’ Committee on Practice Bulletins—Gynecology. Management of symptomatic uterine leiomyomas: ACOG practice bulletin, number 228. Obstet Gynecol. 2021;137:e100–15. [DOI] [PubMed] [Google Scholar]
- [2].Stewart EA, Cookson CL, Gandolfo RA, Schulze-Rath R. Epidemiology of uterine fibroids: a systematic review. BJOG. 2017;124:1501–12. [DOI] [PubMed] [Google Scholar]
- [3].Kim WY, Kim JM, Han SB, et al. Steaming of ginseng at high temperature enhances biological activity. J Nat Prod. 2000;63:1702–4. [DOI] [PubMed] [Google Scholar]
- [4].Fan W, Fan L, Wang Z, et al. Rare ginsenosides: a unique perspective of ginseng research. J Adv Res. 2024;66:303–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Cho J, Park W, Lee S, Ahn W, Lee Y. Ginsenoside-Rb1 from Panax ginseng C.A. Meyer activates estrogen receptor-alpha and -beta, independent of ligand binding. J Clin Endocrinol Metab. 2004;89:3510–5. [DOI] [PubMed] [Google Scholar]
- [6].Lee YJ, Jin YR, Lim WC, et al. Ginsenoside-Rb1 acts as a weak phytoestrogen in MCF-7 human breast cancer cells. Arch Pharm Res. 2003;26:58–63. [DOI] [PubMed] [Google Scholar]
- [7].Kim YS, Woo JY, Han CK, Chang IM. Safety analysis of Panax ginseng in randomized clinical trials: a systematic review. Medicines (Basel). 2015;2:106–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Song SW, Kim HN, Shim JY, et al. Safety and tolerability of Korean Red Ginseng in healthy adults: a multicenter, double-blind, randomized, placebo-controlled trial. J Ginseng Res. 2018;42:571–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Hsu WC, Tsai YT, Hou YC, Lai JN. Prescription of Chinese herbal products is associated with a decreased risk of uterine fibroids: a population-based cohort study. Medicine (Baltimore). 2019;98:e18195. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Wu CT, Tsai YT, Hsieh YT, Fu SL, Lai JN. Traditional herbal medicine and the decreased risk of uterine surgery among women with uterine fibroid in Taiwan. J Herb Med. 2023;42:100759. [Google Scholar]
- [11].Hopewell S, Chan AW, Collins GS, et al. CONSORT 2025 statement: updated guideline for reporting randomised trials. BMJ. 2025;389:e081123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Cook JA, Julious SA, Sones W, et al. DELTA2 guidance on choosing the target difference and undertaking and reporting the sample size calculation for a randomised controlled trial. BMJ. 2018;363:k3750. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Ghosh S, Naftalin J, Imrie R, Hoo WL. Natural history of uterine fibroids: a radiological perspective. Curr Obstet Gynecol Rep. 2018;7:117–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Peddada SD, Laughlin SK, Miner K, et al. Growth of uterine leiomyomata among premenopausal black and white women. Proc Natl Acad Sci U S A. 2008;105:19887–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Moshesh M, Peddada SD, Cooper T, Baird D. Intraobserver variability in fibroid size measurements. J Ultrasound Med. 2014;33:1217–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Munro MG, Critchley HO, Broder MS, Fraser IS; FIGO Working Group on Menstrual Disorders. FIGO classification system (PALM-COEIN) for causes of abnormal uterine bleeding in nongravid women of reproductive age. Int J Gynaecol Obstet. 2011;113:3–13. [DOI] [PubMed] [Google Scholar]
- [17].Baird DD, Patchel SA, Saldana TM, et al. Uterine fibroid incidence and growth in an ultrasound-based, prospective study of young African Americans. Am J Obstet Gynecol. 2020;223:402.e1–402.e18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Choi MK, Jin S, Jeon JH, et al. Tolerability and pharmacokinetics of ginsenosides Rb1, Rb2, Rc, Rd, and compound K after single or multiple administration of red ginseng extract in human beings. J Ginseng Res. 2020;44:229–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Lee HW, Ang L, Lee MS. Using ginseng for menopausal women’s health care: a systematic review of randomized placebo-controlled trials. Complement Ther Clin Pract. 2022;48:101615. [DOI] [PubMed] [Google Scholar]
- [20].Butcher NJ, Monsour A, Mew EJ, et al. Guidelines for reporting outcomes in trial reports: the CONSORT-outcomes 2022 extension. JAMA. 2022;328:2252–64. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
