Diabetes ginseng therapeutic effects of hypertension ginseng

Research Article Effect of coadministration of enriched Korean Red Ginseng (Panax ginseng) and American ginseng (Panax quinquefolius L) on cardiometabolic outcomes in type-2 diabetes: A randomized controlled trial

Elena Jovanovski a,f, Lea Smircic-Duvnjak b, Allison Komishon a, Fei (Rodney) Au-Yeung a,f, John L. Sievenpiper a,c,d,e,f, Andreea Zurbau a,f, Alexandra L. Jenkins a, Mi-Kyung Sung g, Robert Josse d,f, Dandan Li a,f, Vladimir Vuksan a,c,d,f,*

a r t i c l e i n f o Article history: Received 21 March 2019 Received in Revised form 25 September 2019 Accepted 4 November 2019 Available online 14 November 2019 Keywords: Ginseng Hypertension Randomized controlled trial Type-2 diabetes

^a Clinical Nutrition and Risk Factor Modiﬁcation Centre, St. Michael’s Hospital, Toronto, Canada

^b School of Medicine University of Zagreb, University Clinic for Diabetes, Endocrinology and Metabolic Diseases, Zagreb, Croatia

^c Li KaShing Knowledge Institute, St. Michael’s Hospital, Toronto, Canada

^d Division of Endocrinology & Metabolism, St. Michael’s Hospital, Toronto, Canada

^e Toronto 3D Knowledge Synthesis and Clinical Trials Unit, St. Michael’s Hospital, Toronto, Canada

^f Department of Nutritional Sciences, Faculty of Medicine, University of Toronto, Toronto, Canada

^g Department of Food and Nutrition, Sookmyung Women’s University, Yongsan-gu, Seoul, Republic of Korea

a b s t r a c t

Background: Diabetes mellitus and hypertension often occur together, amplifying cardiovascular disease (CVD) risk and emphasizing the need for a multitargeted treatment approach. American ginseng (AG) and Korean Red Ginseng (KRG) species could improve glycemic control via complementary mechanisms. Additionally, a KRG-inherent component, ginsenoside Rg3, may moderate blood pressure (BP). Our objective was to investigate the therapeutic potential of coadministration of Rg3-enriched Korean Red Ginseng (Rg3-KRG) and AG, added to standard of care therapy, in the management of hypertension and cardiometabolic risk factors in type-2 diabetes.
Methods: Within a randomized controlled, parallel design of 80 participants with type-2 diabetes (HbA1c: 6.5e8%) and hypertension (systolic BP: 140e160 mmHg or treated), supplementation with either 2.25 g/day of combined Rg3-KRG þ AG or wheat-bran control was assessed over a 12-wk inter- vention period. The primary endpoint was ambulatory 24-h systolic BP. Additional endpoints included further hemodynamic assessment, glycemic control, plasma lipids and safety monitoring.

Results: Combined ginseng intervention generated a mean ± SE decrease in primary endpoint of 24-h systolic BP (-3.98 ± 2.0 mmHg, p ¼ 0.04). Additionally, there was a greater reduction in HbA1c (-0.35 ± 0.1% [e3.8 ± 1.1 mmol/mol], p ¼ 0.02), and change in blood lipids: total cholesterol (-0.50 ± 0.2 mmol/l, p ¼ 0.01), non-HDL-C (-0.54 ± 0.2 mmol/l, p ¼ 0.01), triglycerides (-0.40 ± 0.2 mmol/l, p ¼ 0.02) and LDL-C (-0.35 ± 0.2 mmol/l, p ¼ 0.06) at 12 wks, relative to control. No adverse safety outcomes were observed.

Conclusion: Coadministration of Rg3-KRG þ AG is an effective addon for improving BP along with attaining favorable cardiometabolic outcomes in individuals with type 2 diabetes. Ginseng derivatives may offer clinical utility when included in the polypharmacy and lifestyle treatment of diabetes.
Clinical trial registration: Clinicaltrials.gov identiﬁer, NCT01578837;

* Corresponding author. Clinical Nutrition and Risk Factor Modiﬁcation Centre, St. Michael’s Hospital, 30 Bond Street, Toronto, Ontario, M5B 1M8, Canada.
E-mail address: v.vuksan@utoronto.ca (V. Vuksan).

https://doi.org/10.1016/j.jgr.2019.11.005
p1226-8453 e2093-4947/$ e see front matter © 2020 The Korean Society of Ginseng, Published by Elsevier Korea LLC. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Introduction

Diabetes mellitus has emerged as a global health epidemic, often coexisting with hypertension which presents a multifold increased risk for cardiovascular disease (CVD) [1]. Within the context of comprehensive cardiovascular risk management, achieving blood pressure (BP) targets remains a clinical challenge, emphasized recently by stricter guidelines reafﬁrming the beneﬁts of tight BP control [2]. Thus, novel combined treatment adjuncts, as part of a multifaceted approach, are required to attain adequate cardiometabolic control in this population.

With a diversiﬁed pursuit for effective management strategies, nonpharmacological approaches are receiving growing attention for meeting public health needs. Ginseng species have long been used for their therapeutic potential and consistent rank among the top ﬁve most commonly consumed herbs globally [3]. Two of the most common species, the American ginseng (Panax quinquefolius L.) and Korean Red Ginseng (Panax ginseng Meyer) are deﬁned by a distinctive proﬁle of dammarane-type glycosides called ginseno- sides, purported to be the principal pharmacologically active con- stituents of ginseng. Although the phytochemical proﬁle of ginseng has drawn signiﬁcant research traction in the area of metabolic function, beneﬁcial clinical correlation is limited.

Recent data suggests that AG and KRG may improve glycemic markers with two distinct but complementary mechanisms of ac- tion: AG appears to promote glucose-dependent insulin secretion, while KRG may increase insulin sensitivity [4,5]. Acutely, both AG and KRG were reported to ﬂatten postprandial glucose concentra- tions without increased risk of hypoglycemia [6e8]. Furthermore, experimental evidence points to a vasodilatory effect of KRG- derived ginsenoside fractions [9e11]. The observed vascular relaxation stems from several proposed mechanisms largely mediated by nitric oxide stimulating pathways [10]. The most potent vasodilatory response in animal and in vitro models appears to be elicited by ginsenoside Rg3, present in KRG [11]. More recently, we have shown that oral administration of KRG enriched with ginsenoside Rg3 acutely improved BP and arterial stiffness in normotensive individuals [12].

Thus, we hypothesized that combining the glucose-lowering

effects of both KRG and AG with the potential hemodynamic ben- eﬁts of Rg3-enriched KRG (Rg3-KRG) may offer a multifaceted treatment advantage to diabetes patients with elevated BP. The objective of the present study, therefore, was to evaluate the efﬁ- cacy and safety of a 12-wk coadministration of Rg3-KRG þ AG ex- tracts on systolic BP and in addition, explore associated cardiometabolic risk factors in individuals with type-2 diabetes managed by standard of care.

Research design and methods

Participants

Prospective participants were recruited using electronic and print advertisements, or through internal participant recruitment system of the patients within two participating universitye afﬁliated hospital centers (St. Michael’s Hospital Toronto, Canada, and KB Merkur Hospital, Zagreb, Croatia). Individuals were eligible to participate if they had a diagnosis of type-2 diabetes for 2:1 year, HbA1c values ::: 8.5%, clinically diagnosed hypertension deﬁned by the use of antihypertensive agents or seated systolic BP of 2:140 mmHg or diastolic BP 2: 90 mmHg on two occasions and BMI<35 kg/m². Main exclusion criteria included: insulin therapy; systolic BP of 2:160 mmHg or diastolic BP 2: 100 mmHg; heart failure; liver dysfunction; serum triglyceride >4.5 mmol/l; a history of coronary, arrhythmic or stroke events; inﬂammatory bowel disease; bleeding

disorder; use of anticoagulants, antiplatelet, sympathomimetic or antidepressant drugs. All participants provided written informed consent. The study protocol met the regulatory framework of Health Canada and the institutional review boards of the partici- pating sites. The trial was undertaken in compliance with Good Clinical Practice guidelines. The Clinicaltrials.gov identiﬁer is NCT01578837.

Study design

We conducted a multicenter, randomized, placebo-controlled parallel design trial. Willing and eligible participants were ran- domized to either the treatment or placebo group with allocation created for each site. All participants and individuals involved in the analysis of the trial were blinded to treatment assignments for the duration of the study. Participants were scheduled to enter a pla- cebo run-in phase for two wks to ensure that BP remained stable and allowed for protocol accustomization. Following the run-in phase, randomized participants entered either the test or control intervention for duration of 12 wks. Individuals were instructed to take study supplements orally in the morning, afternoon and eve- ning each day prior to a meal, except on study visit mornings. Study visits occurred at baseline (week 0) and subsequently at four-wk intervals for the study duration. At each visit, participants attended the clinic after a 10e12 h fast and had anthropometrics taken, symptoms diaries and questionnaires completed, and outcome assessments performed. A >four-wk supply of study capsules was dispensed at every visit in two bottles (95 capsules each) to be started after completion of the baseline procedures. Changes to background antihypertensive and antihyperglycemic therapy and any adverse effects were recorded.

Interventions

Participants were randomized to either combined American ginseng (P. quinquefolius) and Rg3-enriched Korean Red Ginseng (P. ginseng) extract or Wheat-bran control. The combined ginseng supplement consisted of 1.5 g/day AG and 0.75 g/day Rg3-KRG standardized extracts (extracted to contain 75 mg of ginsenoside Rg3 and 375 mg total ginsenosides) or 2.25 g/day of control, con- tained in six gelatin capsules (3 x 500mg AG or control, 3 x 250 mg Rg3-KRG or control). There was no visual difference between the test and control capsules. The participants were instructed to take 2capsules t.i.d. Before each main meal. The AG was provided by the Ontario Ginseng Growers Association and extracted to 10% ginse-

nosides (Canadian Phytopharmaceuticals Corporation, BC, Canada). Dried AG roots were treated in 70% ethanol solution at 80oC for three h. The crude ginsenoside extract was ﬁltered and combined ﬁltrate is vacuum evaporated to appropriate volume. The ﬁnal AG product had the following concentrations of major ginsenosides:

Rb1 (36.76 mg/g), Rb2 (1.73 mg/g), Rb3 (4.44 mg/g), Rg1 (2.52 mg/ g), Rc (10.87 mg/g), Rd (10.28 mg/g), Re (22.03 mg/g) and Rg3 (undetectable). The Rg3-KRG was prepared utilizing a proprietary technology by BTGin Corp. (Daejeon, Korea). Brieﬂy, KRG rootlets underwent 50% and 85% ethanol extraction in consecutive steps. The extract was treated by enzyme and acid hydrolysis to amplify ginsenoside Rg3 content. Beta-glycosidase, produced from Asper-

gillus niger, which has cellulase, hemicellulase, glucosidase activity was used in acidic (pH 2.5~3.5) and thermophilic (65~80o C) con- ditions. To remove acid solution and concentrate Rg3, the reactant was passed through DIAION HP20 resin (Mitsubishi Chemical In- dustries, Tokyo, Japan) packed column. This was followed by

evaporation to powder form under vacuum conditions. The ﬁnal product containing 30% total ginsenosides had the following con- centrations of major ginsenosides: Rb1 (3.77 mg/g), Rg1 (0.57 mg/

g), Re (1.86 mg/g), Rf (12.3 mg/g), Rb2 (1.24 mg/g), Rh1 (4.15 mg/g), Rc (0.99 mg/g) and Rg3 (100 mg/g). Individual ginsenosides were analyzed using high-performance liquid chromatography (HPLC). HPLC was carried out on a Liquid Chromatography (LC) system equipped with a quaternary gradient pump (Spectra 4000) and UV detector (Spectra 2000). A reversed-phase column (Hypersil gold C18, 100 mm 4.6 mm, internal 5 mm; Thermo Scientiﬁc) was used for quantitative determination of ginsenosides Rg3. The mobile phase consisted of acetonitrile and water with a ﬂow rate at 1.6e2.5 mL/min and the detection wavelength was set at 203 nm. Control capsules contained standardized wheat bran provided by Rogers Foods Limited (Armstrong, BC, Canada). Quality control of all sup- plements was performed to comply with Health Canada regulations.

Outcome measurements

The prespeciﬁed primary endpoint was the 12-wk difference in 24-h systolic BP. Ambulatory BP monitoring was undertaken at baseline and study-end using a standardized cuff system (Spacelabs ABP Monitor 90207, Spacelabs Healthcare, WA, USA) placed on the nondominant arm. At baseline and week-12, automated readings were obtained at 20-min and 60-min intervals during daytime and night-time, respectively. Mean 24-h measurements were calculated by averaging BP readings collected within an hour and then aver- aging across the 24-h monitoring period for each patient. Ambu- latory records were evaluated for quality according to prespeciﬁed criteria of >50% obtained recordings. Clinic BP was measured at each visit in the seated position by standard sphygmomanometry (Omron HEM-907XL, Omron Healthcare Inc., IL, USA). Three consecutive measurements were obtained one minute apart, and the values averaged for each visit. Exploratory hemodynamic endpoints included end differences in ofﬁce BP, ambulatory dia- stolic, daytime and night-time BPs. Data on supplementary vascular measures were also obtained in this trial and will be reported separately.

Secondary glycemic endpoint was HbA1c at 12 wks, supported

by exploratory fasting plasma glucose and insulin levels. Further exploratory endpoints included plasma lipid measures, in addition to safety outcomes. Laboratory measures were performed at baseline and at four-wk intervals thereafter. HbA1c analysis was performed using HPLC with the Tosh-HLC-723 analyzer (site 1) and Tina Quant turbidimetric inhibition immunoassay with the Cobas Integra 400 Plus analyzer (site 2), fasting glucose by a reaction rate method using the Beckman Synchron LX system (site 1) and Beckman Coulter AU680 (site 2), and insulin was analyzed using the Beckman Ultrasensitive Insulin Assay (Beckman Coulter, Brea, USA). Assays for safety measures [alanine aminotransferase (AST), creatinine, prothrombin time (PT), activated partial thromboplastin time (APTT) and International Normalized Ratio (INR)] and plasma lipid measurements [total cholesterol, HDL-C, non-HDL-C and tri- glycerides (TG)] were performed using standard methodology at the local hospital’s clinic laboratory at each site. LDL-C was calcu- lated using the Friedewald formula. Twelve-lead electrocardio- grams, to assess the QT interval, were collected at weeks 4 and 12. Intervention compliance was assessed by using a returned pill count at each visit.

Statistical analysis

Statistical analyses were performed using the SAS statistical package version 9.2 (SAS institute Inc., Cary, NC, USA). Analyses were performed using the intent-to-treat (ITT) approach, which included all participants that were randomly assigned into the study who took at least one dose of the study intervention. PROC

MIXED was used to determine between-treatment and within- treatment end-change in efﬁcacy, safety and compliance end- points at week-12 as the response variable, with baseline and center measures as covariates. Missing efﬁcacy data were imputed using the PROC MIXED and MIANALYZE. Multiple imputation method assumes that the imputed values are randomly selected from the distribution of true missing values. Predictors of miss- ingness at week-12 were used in the imputation. Chi² test was used to determine the differences in participants’ categorical character- istics. We also conducted additional exploratory analyses for the effect of treatment by sex, antihypertensive medication type (NO stimulating vs. non-NO stimulating), antihyperglycemic medica- tion type and baseline BP and HbA1c values. We tested for inter- action of treatment BP effect with age (<60 vs. >60yrs), sex and BMI (<30 and >30 kg/m²). Finally, in prespeciﬁed subgroup anal- ysis, we evaluated the association of the 12-wk change in ambu- latory BP values with corresponding baseline BP measures using linear regression analysis. Assuming a mean 24-h systolic BP dif- ference from control of 4.7 mmHg [13] and a standard deviation of

6.8 mmHg, at a signiﬁcance level of a ¼ 0.05 and 1-b ¼ 0.8, a total of

68 participants were required. With an estimated 20% attrition rate, a sample size of 82 participants or 41 per treatment group were to be recruited as per entry criteria.

Results

Participants

A total of 1,360 individuals were prescreened by phone out of which 361 individuals were assessed for eligibility (Fig. 1). Of these,

85 participants were randomly assigned to receive combined AG&Rg3-KRG intervention or control. Five participants dis- continued the study in the run-in phase before receiving the study intervention (4 participants failed to attend the baseline mea- surement; 1 participant discontinued for medical reasons), result- ing in 80 participants for which data was available at baseline to undertake analysis (modiﬁed ITT). Seventy-eight completed the entire study protocol following run-in, with the attrition rate of 3%. Baseline demographic characteristics are summarized in Table 1 and were similar across the treatment groups. There were no sig- niﬁcant differences in baseline covariates between the two centers (data not shown). Most of the participants were receiving an oral antihyperglycemic agent (98%) and antihypertensive agent (96%) at baseline (http://hyper.ahajournals.org.myaccess.library.utoronto. ca/content/56/5/824.long, Table 1). Metformin was taken by 96% of participants where 62% were also on a DPP-4 inhibitor. In addi- tion, 32 participants (39%) were receiving 2:2 antihypertensive drugs. Sixty-six percent of patients were on lipid lowering medi- cation. Linear regression analysis revealed a signiﬁcant negative association between intervention differences with ginseng and baseline systolic BP (R² ¼ 0.24, p < 0.001). However, there were no signiﬁcant interactions between treatment groups and sex, age, BMI, baseline BP or HbA1c or antihypertensive medication with respect to BP.

3.2. Office and 24-h ambulatory BP

Administration of AG&Rg3-KRG for 12 wks achieved a mean reduction in 24-h systolic BP of 3.98 2.0 mmHg (p . 0.03) relative to control (Fig. 2). No within-treatment change from baseline to the end of 12 weeks in mean 24-h systolic BP was observed for either the control group (1.08 2.4 mmHg, p . 0.65)

or combined ginseng group ( 1.70 2.2 mmHg, p . 0.44). Participants receiving AG&Rg3-KRG also had greater reduction in daytime systolic BP ( 4.47 2.1 mmHg, p . 0.03), but not in night-time systolic BP ( 2.52 2.8 mmHg, p . 0.37). Meanwhile, no differences were observed in mean 24-h-, daytime- or night-timediastolic BP ( 0.96 1.6 mmHg, p . 0.54; 1.54 1.7 mmHg, p . 0.34; 0.23 2.1 mmHg, p . 0.91, respectively) when compared with control. Changes in 24-h ambulatory BP parameters

for the two groups are displayed in Table 2.

Based on mean sitting office BPs recorded at week-12, systolic BP decreased in the ginseng group by 5.22 2.6 mmHg (p . 0.02)

compared with the baseline values; no change was evident in the control group ( 0.91 2.9 mmHg, p . 0.76), with a between treatment effect of 4.46 2.5 mmHg (p . 0.07). A similar patternwas observed with office diastolic BP, with a reduction from baseline to the end of 12 weeks in the ginseng treatment arm ( 6.32 2.2 mmHg, p . 0.001) and not in the control arm

( 2.22 2.4 mmHg, p . 0.36). No difference between interventions was observed in the office diastolic BP ( 2.41 2.2 mmHg,p . 0.26).

Ginseng supplementation was associated with greater reductions in mean 24-h BP and office BP in individuals with higher baseline BP (R2 . 0.24, p < 0.01). However, the changes in the 24-hsystolic BP were similar in subgroup analysis by gender (male vs. female), age ( 60 vs. > 60 years), BMI ( 25 kg/m2 vs. >25 kg/m2),baseline BP and HbA1c values, and antihypertensive or antihyperglycemic medication type

Glycemic control

A 12-wk administration of Rg3-KRG þ AG led to a signiﬁcant treatment difference in HbA1c levels of -0.35 ± 0.2% (p ¼ 0.02), where change from baseline of -0.25 ± 0.2% (p ¼ 0.02) was observed in the combined ginseng group and no change in the control group (0.09 ± 0.2%, p ¼ 0.61) (Table 2). Fasting plasma

glucose was not affected by the ginseng intervention relative to control (treatment difference: -0.02 ± 0.3 mmol/l, p ¼ 0.95). Similarly, no difference in fasting insulin at 12 weeks was observed between treatment groups (2.72 ± 18.7 pmol/l, p ¼ 0.88).

Lipids

At study end, the reduction of the combined ginseng versus control treatment was signiﬁcantly greater for total cholesterol (treatment difference -0.50 ± 0.2 mmol/l (p ¼ 0.01) and for non-

Table 2

HDL (-0.54 ± 0.2 mmol/l, p ¼ 0.01), but approaching signiﬁcance for LDL-C (-0.35 ± 0.2 mmol/l, p ¼ 0.06). Similarly, mean TG levels were signiﬁcantly reduced in the ginseng group by -0.43 ± 0.2 mmol/l (p ¼ 0.02), relative to control. No difference was observed in HDL-C (0.06 ± 0.04 mmol/l, p ¼ 0.19).

Safety variables

Oral intervention was well tolerated. There were no serious adverse events and no differences in adverse events

between the two groups. Commonly reported adverse events

included headache (three participants on ginseng and four control),

nausea (two in ginseng and three in the control) and abdominal

discomfort (six in ginseng and four in control) of transient nature,

with no differences in number of events between treatment arms

(p . 0.62).

Table 3 displays measured safety variables. There was no significant

difference between the ginseng treatment and control with

respect to changes in creatinine ( 1.17 1.6 mmol/l, p . 0.47), ALT

(0.46 2.2 U/l, p . 0.92) and the QT interval (0.62 5.5 seconds,

p . 0.89). Furthermore, hematological measures of PT

( 0.06 0.05 seconds, p . 0.22), APTT (0.05 0.5 seconds,

p . 0.91) and INR (0.002 0.02, p . 0.94) did not differ between

the combined ginseng and control. Overall compliance, as assessed

by manual pill count, was 94.3 9.2% in the treatment arm and

96.3 9.3% in the control arm.

Discussion

Combination therapy has been advocated as a more effective means to meet clinical targets in the contemporary management of type-2 diabetes. The present study supports our hypothesis that coadministration of Rg3-KRG þ AG for 12 weeks can result in improvement in BP in already treated type-2 diabetes patients with hypertension. Additional beneﬁts included tighter glycemic control and reduction in plasma lipids with no adverse safety concerns.

The mean reduction in 24-h systolic BP of -4.0 mmHg over 12 weeks is clinically meaningful, particularly against the backdrop of antihypertensive therapy in this subject group. Systolic BP offers consistent and proportional CVD risk decrease in type-2 diabetes, where a 5 mmHg change results in a pooled estimate of 13% stroke reduction in BP lowering trials [14,15]. Improvement of BP is pro- posed to have a greater potential to reduce CVD than lowering plasma glucose concentrations in this population, thus represent- ing a relevant primary therapeutic target [16]. Importantly, the 24- h BP reduction was not accompanied by an increase in heart rate and appears to be attributed to change in daytime rather than night-time BP, possibly an effect of 3x/day administration of Rg3- ginsensoide component and a reported, short half-life of its me- tabolites (0.46 ± 0.1 hours) [17]. The achieved ambulatory BP reduction is similar to that of ofﬁce BP (i.e., -4.0 mmHg vs. -4.4 mmHg); nevertheless, the difference in ofﬁce BP only approached statistical signiﬁcance (p < 0.07) on account of large variability in the measure. The use of ambulatory BP monitoring in this trial, as an established better predictor of clinical outcomes than conven- tional ofﬁce BP, offers an advantage of providing a more accurate estimate of true vascular beneﬁts, underscoring the importance of ambulatory BP monitoring for deﬁning BP control [18].

Selection of Rg3-enriched KRG as a complementary agent stems from a line of investigations into the vascular effects of this ginseng species. We have previously shown that the effect of KRG supple- mentation on ﬂow-mediated vasodilation and pulse wave reﬂec- tion was attributable to the ginsenoside fraction of the root [12]. In particular, ginsenoside 20(S) Rg3 appears most potent to induce nitric oxideedriven vasodilatation in vivo [19] and may attenuate the rise in intracellular calcium levels [20], while a preliminary evaluation of Rg3-KRG in hypertensive rats demonstrated signiﬁ- cant improvement in BP [21]. A pilot study of single oral bolus dose administration of 400 mg Rg3-KRG in humans corroborated pre- clinical evidence with attained reductions in hemodynamic indices [12,21]. Furthermore, the same formulation of Rg3-KRG, assessed over eight weeks, may effectively improve erectile dysfunction of- fering additional support for the underlying vasodilating activity of Rg3-KRG [22]. The antihypertensive effect of Rg3 might be also exerted through the inhibition of angiotensin-converting-enzyme thereby modulating the activity of the renin-angiontensin system. Concurrent to the observed BP beneﬁt, the combined ginseng supplementation resulted in a reduction of HbA1c level, where baseline concentration was proportional to HbA1c decline. The observed mean difference of 0.35% is within the range of >0.3%, proposed by the FDA guidelines as clinically meaningful lowering of HbA1c [23]. The modest decrease, which was adjunctive to oral antihyperglycemic agents, may be particularly relevant in patients who are near to achieving their treatment goal.

The choice of coadministration of two ginseng species rests with

the understanding that they differentially affect glycemic control. Our group has previously shown that 3 g KRG administration resulted in 33% increase in both, HOMA-IS and oral glucose chal- lengeederived insulin sensitivity indices, over 12 weeks [4]. Several pathways are implicated, including modulation of glucose trans- port via glucose transporter protein and regulation of glucose disposal via glycolytic enzymes, further supported by the primary effect of KRG in intracellular signaling pathways [24,25]. Conversely, 6 g AG supplementation, improved HbA1c by

0.29 ± 0.04% (3.2 ± 0.4 mmol/mol), which was concomitant with a 30% rise in fasting insulin without associated weight gain or hy- poglycemia [26], a ﬁnding reproduced in this study. The post- prandial effect of AG on glucose-stimulated insulin secretion was observed previously [27]. Preclinical evidence also suggests enhanced insulin release in diabetic mice fed with AG, underwrit- ten by morphological evidence of islet area increase [28,29]. The reduction in HbA1c using combined ginseng appears stronger than those noted previously using single ginseng supplementation. However, lack of direct comparison and factors such as concomitant therapy preclude an interpretation beyond a speculative one.

The observed decrease in HbA1c was not reﬂected in changes in fasting glucose, despite a noted reduction in our previous trial using

the AG root [5] and observations from a metaanalysis reporting a pooled fasting glucose lowering effect across ginseng sources [30]. However, resultant HbA1c values, although partially determined by fasting glucose, are thought to more strongly reﬂect postprandial glycemia [31]. Both AG and KRG have repeatedly been demon- strated to lower plasma glucose area under the curve ranging from 26% to 38% that were consistent across different doses (3e12 g) and times of administration (i.e., immediately or 40 min preprandially) [6], and may therefore have been the main mechanism of the HbA1c reduction seen in this study.

We observed a reduction in plasma lipid concentrations, including total cholesterol, non-HDL-cholesterol and triglycerides and a tendency toward lower LDL-C. Non-HDL-C is a particularly desirable target in individuals with diabetes characterized with elevated triglycerides [32]. A recent systematic review and meta- analysis pooling RCT data on lipid outcomes illustrated reductions in total, LDL-C and triglycerides in favor of ginseng, but no change in HDL-C [33]. Plausibly, the steroidal structure of triterpene saponins may act to modify gene transcription, protein synthesis and hepatic cholesterol production via inhibition of HMG-CoA reductase [34,35]. It appears that the effect on lipids is not speciﬁc to our selected intervention types as the signal emerges across different species and preparations [26]. While intriguing, these ﬁndings remain exploratory and necessitate more direct investigation.

The adverse events and selected biochemical safety markers in the ginseng group were similar to those noted with the control. These data are relevant given that the quantity of total ginsenosides present in the intervention, 375 mg/day, is in the uppermost range of those administered to humans in RCT settings. Mucalo et al (2014) provided comprehensive evidence in type-2 diabetes on the safety proﬁle of three-month supplementation with AG extract standardized to 10% of ginsenosides, as used in our trial

[36]; conversely, no clinical data on ginsenoside Rg3-enriched KRG was available. Accordingly, our results contribute to the current understanding of the clinical toxicology related to ginseng use and remain in concordance with the previous reports on generic

ginseng monopreparations [37].

While preliminary, our study provides support for the potential clinical utility of promising herbal supplements as adjuncts to standard care. Use of botanical agents offers a distinct advantage because of its biochemical complexity and the possibility to act on multiple metabolic targets [38]. Conversely, lack of standardization and high compositional variability, as we demonstrated both be- tween and within ginseng species and quantiﬁcation assays, may prove to be prohibitive in their use as an alternative treatment option [39]. Hence, rigorous efﬁcacy-founded component stan- dardization of herbal supplements is a top priority to facilitate a reliable physiological response.

Our study has strength in its rigorous application of a controlled, adequately powered double-blind protocol and use of a well- characterized, reproducible intervention, derived from systematic preliminary screening studies. A high retention rate increases generalizability of our ﬁndings from ITT analysis. We opted to use a pragmatic, translational study design by selecting a population of participants with type-2 diabetes, who are already on medication and in whom assessed risk factors were clinically optimized. We further opted to use 24-h ambulatory BP recordings as the primary outcome, minimizing confounders of ofﬁce BP measures to allow for more coherent characterization of BP.

There are several limitations of the current study. First, use of the intervention capsules (i.e., six/day) may add to the poly- pharmacy burden, despite the reassuring compliance (95%) in the outpatient study setting. Second, the bioactivity of ginseng varies across different preparations and modes of intake and extrapo- lating the efﬁcacy and safety endpoints to other ginseng

preparations, therefore, poses a challenge. Similarly, the choice of ginseng dose was limited by the availability of previous data and regulatory controls; we are thus unable at this time to deﬁne a dose-response relationship. Third, lack of direct comparison of in- dividual ginseng species in the trial precludes comprehensive assessment of respective additive contributions. Fourth, using a time pointespeciﬁc end point at week-12 for ofﬁce BP instead of the change of BP over the entire duration of the trial did not use all of the data collected. Finally, the glycemic endpoints were analyzed within respective centers, introducing a source of measurement variability. Future research must further consider these points moving forward.

Conclusion

The present study provides quality evidence that combining botanicals, Rg3-KRG þ AG, with purported complementary activity, can promote clinically meaningful changes in BP when added to standard therapy in type-2 diabetes. Further changes in exploratory glycemic and lipid endpoints provide additional support for the overall cardiometabolic beneﬁts. Subsequent efforts should focus on further optimization of ginseng components and its reproduc- ibility for development of viable therapeutic entities. Given the staggering diabetes prevalence frequently clustered with hyper- tension and dyslipidemia, such promising multitargeted risk reduction aids should be increasingly sought to assist in alleviating CVD risk.

Funding

The trial was funded by the Canadian Diabetes Association Grant-ineAid funding, a contribution from BTGin Corp., Korea, and a donation from the Diabetes Alternative Research and Healthcare Foundation. Ontario Ginseng Growers provided an in-kind dona- tion of the raw, Ontario-grown P.quinquefolius root. JL Sievenpiper was funded by a PSI Graham Farquharson Knowledge Translation Fellowship, Canadian Diabetes Association (CDA) Clinician Scientist award, CIHR INMD/CNS New Investigator Partnership Prize, and Banting & Best Diabetes Centre Sun Life Financial New Investigator Award. None of the sponsors had a role in any aspect of the present study, including design and conduct of the study; collection, management, analysis and interpretation of the data; and prepa- ration, review, approval of the manuscript or decision to publish.

Authors’ contributions

All authors participated in the trial design. AK, LDS and EJ took part in carrying out the trial and the data collection. FAY, EJ and AK took part in the data analysis. EJ, AK, VV and LZ prepared the manuscript draft. All authors interpreted the data. All authors read the manuscript critically and approved the submitted version.

Declaration of competing interest

V Vuksan holds an American (No. 7,326,404 B2) and Canadian (No. 2,410,556) patent for use of viscous ﬁber blend in diabetes, metabolic syndrome and cholesterol lowering. VV was partial owner of Glycemic Index Laboratories, Inc. (Toronto, ON, Canada) during 2004e2015.

JL Sievenpiper has received research support from the Canadian Institutes of health Research (CIHR), Diabetes Canada, PSI Founda- tion, Banting and Best Diabetes Centre (BBDC), Canadian Nutrition Society (CNS), American Society for Nutrition (ASN), Calorie Control Council, INC International Nut and Dried Fruit Council Foundation, National Dried Fruit Trade Association, The Tate and Lyle

Nutritional Research Fund at the University of Toronto, and The Glycemic Control and Cardiovascular Disease in Type 2 Diabetes Fund at the University of Toronto (a fund established by the Alberta Pulse Growers). He has received speaker fees and/or honoraria from Diabetes Canada, Canadian Nutrition Society (CNS), Dr. Pepper Snapple Group, Dairy Farmers of Canada, Nutrition Foundation of Italy (NFI), C3 Collaborating for Health, Sprim Brasil, WhiteWave Foods, Rippe Lifestyle, mdBriefcase, Alberta Milk, FoodMinds LLC, Memac Ogilvy & Mather LLC, PepsiCo, The Ginger Network LLC, International Sweeteners Association, Pulse Canada, and GI Foun- dation. He has ad hoc consulting arrangements with Winston & Strawn LLP, Perkins Coie LLP, and Tate & Lyle. He is a member of the European Fruit Juice Association Scientiﬁc Expert Panel. He is on the Clinical Practice Guidelines Expert Committees of Diabetes Canada, European Association for the study of Diabetes (EASD), and Canadian Cardiovascular Society (CCS), as well as an expert writing panel of the American Society for Nutrition (ASN). He serves as an unpaid scientiﬁc advisor for the Food, Nutrition, and Safety Pro- gram (FNSP) and the Technical Committee on Carbohydrates of the International Life Science Institute (ILSI), North America. He is a member of the International Carbohydrate Quality Consortium (ICQC), Executive Board Member of the Diabetes and Nutrition Study Group (DNSG) of the EASD and Director of the Toronto 3D Knowledge Synthesis and Clinical Trials foundation. His wife is an employee of Unilever Canada.

AL Jenkins is vice president and partial owner of Glycemic Index

Laboratories, Inc. (Toronto, ON, Canada) a clinical research organi- zation. No other authors declared conﬂict of interest.

Acknowledgments

The authors would like to thank BTGin for donation of the Rg3- enriched KRG (OKBT), Canadian Phytopharmaceuticals for extrac- tion of AG and Roger Foods Ltd. for their donation of the wheat bran. We thank all the participants, investigators, and trial-site staff who were involved in this study. In addition, we thank Dr. Kristina Blaslov for data collection at the Zagreb site.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.jgr.2019.11.005.

References