Ethanol extraction preparation of American ginseng (Panax quinquefolius L) and Korean red ginseng (Panax ginseng C.A. Meyer): Differential effects on postprandial insulinemia in healthy individuals

Journal of Ethnopharmacology 159 (2015) 55–61

Contents lists available at ScienceDirect

Journal of Ethnopharmacology journal homepage: www.elsevier.com/locate/jep

Research paper

Ethanol extraction preparation of American ginseng (Panax quinquefolius L) and Korean red ginseng (Panax ginseng C.A. Meyer): Differential effects on postprandial insulinemia in healthy individuals Leanne R. De Souza a,c, Alexandra L. Jenkins a, Elena Jovanovski a, Dario Rahelić b, Vladimir Vuksan a,c,d,n a Clinical Nutrition and Risk Factor Modification Center and Li Ka Shing Knowledge Institute, St. Michael's Hospital, 70 Richmond St. E., Toronto, Ontario, Canada M5C1N8 b Division of Endocrinology, Diabetes and Metabolic Disease, Dubrava University Hospital, Zagreb, Croatia c Department of Nutritional Sciences and Department of Medicine, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada M5S3E2 d Division of Endocrinology and Metabolism, St. Michael's Hospital, Toronto, Ontario, Canada

art ic l e i nf o

a b s t r a c t

Article history: Received 22 April 2014 Received in revised form 8 October 2014 Accepted 27 October 2014 Available online 4 November 2014

Ethnopharmacological relevance: Ginsenosides are the proposed bioactive constituent of ginseng, especially for the attenuation of postprandial glycemia (PPG). The efficacious proportion of total and specific ginsenosides, remains unknown. Alcohol extraction of whole ginseng root can be used to selectively manipulate the ginsenoside profile with increasing alcohol concentrations producing high yields of total ginsenosides and varying their individual proportions. Aim of the study: We aimed to compare the acute efficacy of different ethanol-extraction preparations of American ginseng (AG) and Korean red ginseng (KRG), with their whole-root origins, on PPG and insulin parameters in healthy adults. Materials and methods: Following an overnight fast, 13 healthy individuals (Gender: 5M:8F, with mean7SD, age: 28.979.2 years, BMI: 26.372.7 kg/m2 and fasting plasma glucose: 4.2170.04 mmol/L) randomly received 3 g of each of the following 10 different ginseng treatments on separate visits: whole root KRG and AG; 30%, 50% or 70% ethanol extracts of KRG and AG and 2 cornstarch placebos. Treatments were consumed 40 min prior to a 50 g oral glucose challenge test with capillary blood samples collected at baseline, 15, 30, 45, 60, 90 and 120 min. Insulin samples were collected at 0, 30, 60 and 120 min. Results: There was no difference in attenuation of PPG among the tested ginseng preparations. Measures of Insulin Sensitivity Index (ISI) showed increased insulin sensitivity (IS) with KRG-30% and AG-50% extracts compared to placebo (po0.05). Conclusions: The insulin sensitizing effects of KRG-30% and AG-50% extracts suggest that other root parts, including other ginsenosides not typically measured, may influence PPG and insulin parameters. There is potential for AG and KRG extracts to modulate IS, an independent predictor of type 2 diabetes. & 2014 Elsevier Ireland Ltd. All rights reserved.

Keywords: American ginseng Korean red ginseng Alcohol extracts Postprandial glycemia Insulin sensitivity Complimentary alternative medicine

1. Introduction American ginseng (AG) (Panax quinquefolious L.) and Korean red ginseng (KRG) (Panax ginseng C.A. Meyer) are the two most commonly consumed ginsengs worldwide, and collectively they are among the most extensively studied alternative medicines for management of type 2 diabetes mellitus (T2DM) (Franz et al., 2002; Yin et al. (2008)). Both animal (Franz et al., 2002) and clinical studies (Vuksan et al., 2000a, 2000b, 2001, 2006) have n Correspondence to: Clinical Nutrition and Risk Factor Modification Center, St. Michael's Hospital, 30 Bond St., Toronto, Ontario, Canada M5C 1W8. Tel.: þ 1 416 864 5525; fax: þ1 416864 5538. E-mail address: [email protected] (V. Vuksan).

http://dx.doi.org/10.1016/j.jep.2014.10.057 0378-8741/& 2014 Elsevier Ireland Ltd. All rights reserved.

demonstrated the glycemic-lowering potential of these ginsengs in healthy individuals and in those with T2DM. While the clinical utility of ginseng as an anti-hyperglycemic agent is discernible, the apparent natural compositional variation across ginseng species (Vuksan and Sievenpiper, 2005; Wang et al., 2005), introduces practical limitations to formulating a consistent viable ginseng therapy. This characteristic variability extends to different ginseng batches of a given species (Sievenpiper et al., 2004a; Vuksan and Sievenpiper, 2005) due to individual plant diversity, where bioactive constituents increase from berries to the branching of the radix. (Li et al., 1996; Sievenpiper et al., 2004a, 2004b; Wang et al., 2005; De Souza et al., 2011). The total concentration and relative proportions of triterpene glycosidic saponins (ginsenosides) alter the pharmacological activity of ginseng (Attele et al., 1999; Lee et al., 2010)

56

L.R. De Souza et al. / Journal of Ethnopharmacology 159 (2015) 55–61

and its associated physiological effects (Attele et al., 1999, 2002; Sievenpiper et al., 2006; De Souza et al., 2011). The specific role of ginsenosides in reducing glycemia remains contentious as differential effects occur in the presence of individual saponins, ratios of particular saponins, and total saponin concentration. Moreover, these proposed bioactive constituents may exert their effects independently or in combination with other components and ginsenosides that are not typically measured (Awang, 2000). Despite the bioactive variability of ginseng, well-studied species like AG have demonstrated some consistency in modulating postprandial glucose (PPG) within and across batches (Vuksan et al., 2000a, 2000d, 2001). Similarly, the potential for KRG to lower PPG levels is apparent, (Xie et al., 2005; Vuksan et al., 2006). In a previous issue of this journal, we reported reductions in PPG with KRG whole root (De Souza et al., 2011). Our prior research aligns with the speculation that AG and KRG may operate via complementary mechanisms where AG increases insulin release and KRG improves insulin sensitivity (IS) (Sievenpiper et al., 2006; Vuksan et al., 2006); though these proposed mechanisms remain to be confirmed. Standardization efforts target the identification of efficacious ginseng profiles that can be reproduced to achieve consistent therapeutic effects. A standardized ginseng product could provide a safe and accessible complementary therapy (Vuksan et al., 2000d; Xie et al., 2005). Alcohol extraction techniques may offer an approach by which ginsenosides are extracted from their native matrix to derive particular proportions by increasing alcohol concentrations to distinguish ginsenosides according to relative solubilites. This method permits selective identification of bioactive ginsenoside profiles and can be reproduced. We followed our systematic acute clinical screening model to investigate for the first time, the effects of various ethanol extraction preparations of AG and KRG on PPG and IS. Based on previous studies attributing bioactivity to ginsenosides for attenuation of PPG, we anticipated that ginseng extracts derived from higher ethanol proportions will yield a greater concentration of ginsenosides thereby improving glycemic-lowering and insulinsensitizing effects (Awang, 2000; Sievenpiper et al., 2006). The identified efficacious preparations of AG and KRG will be selected for long-term clinical evaluation in individuals with T2DM.

2. Materials and methods 2.1. Subjects This study was approved by the Research Ethics Board of St. Michael's Hospital, Toronto, ON. Healthy non-diabetic adults were recruited from a research database at the Clinical Nutrition and Risk Factor Modification Centre (RFMC) at St. Michael's Hospital, Toronto, ON, Canada. Subjects who met the inclusion criteria provided written informed consent to participate in the study. The following inclusion criteria were applied: 18–65 years of age, fasting plasma glucose (FPG): 4–6 mmol/l, clinically euthyroid, normal hepatic and renal function, and normotensive (seated systolic BPo140 mm Hg or diastolic BPo90 mm Hg), non-pregnant, no current use of natural health products or supplements including ginseng, no major illnesses/ diseases, no gastrointestinal disorders, no excessive alcohol intake (no more than 2 drinks per day for men and no more than 1 drink a day for women), no excessive smoking (no more than 10 cigarettes per day). 2.2. Design Using computer generated random numbers subjects were randomized to a double-blind, placebo-controlled, multiple

crossover trial. Each subject randomly received each of the following 10 single-dose treatments on separate occasions: KRG whole root, KRG-30%, 50%, 70% alcohol extracts, AG whole root, AG-30%, 50%, 70% ethanol extracts, and a cornstarch placebo which was repeated twice. An individual who was independent of the study performed randomization and the treatments were blinded to the study team and study subjects. 2.3. Treatments KRG and AG samples were each derived from their respective whole root ginseng source from Seoul, South Korea and British Columbia, Canada respectively to develop the ginseng ethanol extracts. The KRG and AG ginseng extracts were proprietary preparations produced by Natural Factors, Burnaby, British Columbia, Canada. All samples were analyzed for ginsenoside content (see Section 2.8). The placebo consisted of a cornstarch control (Canada Cornstarch™). Placebo and ginseng treatments were encapsulated in identical #00 (500 mg) opaque capsules (Capsugels). Each treatment consisted of 6 capsules (500 mg in each capsule for a total of 3 g), stored at 15 1C in light-protective sealed plastic containers. The 3 g dose was selected according to our previous dose-finding study in which we demonstrated equivalent effects among doses ranging from 2–6 g (Sievenpiper et al., 2006). Extracts were equivalent to the 3 g whole root dose. 2.4. Protocol We followed our systematic acute clinical screening model, described previously (Sievenpiper et al., 2003, 2006; De Souza et al., 2011). Subjects were counseled to maintain consistent dietary (150 g minimum carbohydrate intake) and exercise patterns for the duration of the study. Compliance was assessed at each visit and included selfreported adverse symptoms. A minimum 3-day washout period separated each visit to minimize carry-over effects. Subjects arrived at the study center following a 10–12 h overnight fast. Anthropometric measures including height and weight were collected. Subjects remained seated for the duration of the testing period. A fasting capillary blood sample was obtained and then one of the 10 treatments was randomly assigned to each subject and was consumed with 300 mL of water (over a selfstandardized time within 5 min). Forty minutes after the capsules were taken, another capillary sample was obtained and subjects consumed 50 g of anhydrous glucose (Fisher Scientific™) mixed into 250 ml of water. Additional capillary blood samples were obtained at 15, 30, 45, 60, 90 and 120 min after the consumption of the glucose. 2.5. Glucose analysis Finger-prick blood samples were collected in 7 mL tubes (Sarstedt, St Leonard, Quebec) that were prepared with anticoagulant. Collected capillary blood samples were frozen immediately at  20 1C and analyzed within 3 days of collection. Glucose analysis was conducted using a YSI 2300 STAT PLUS (Yellow Springs, OH, USA), calibrated with a standard 10 mmol/L glucose solution prior to analysis of each set of 8 samples. 2.6. Insulin analysis and the Insulin Sensitivity Index (ISI) An additional 6–8 drops of blood were collected into microvette tubes. Blood was allowed to clot and then centrifuged and the serum transferred to labeled polypropylene tubes and stored at  20 1C prior to analysis of serum insulin. Insulin was measured using Mercodia Insulin ELISA© Alpco Diagnostics Immunoassay Kits, for quantitative determination of human insulin. The Insulin

L.R. De Souza et al. / Journal of Ethnopharmacology 159 (2015) 55–61

Sensitivity Index (ISI) was used to quantify total body insulin sensitivity, calculated from the oral glucose challenge test, and provides a reasonable estimate of peripheral insulin sensitivity. We calculated ISI as described by Matsuda: 10,000/√ (FPG  FPI)  (mean PG  mean PI) (Matsuda and DeFronzo, 1999). Where FPG is the fasting plasma glucose and FPI is fasting plasma insulin for an individual at a particular time point. Mean PG and mean PI is the average plasma glucose and insulin respectively, for all individual time points within a sample.

2.7. Study outcomes and statistical analyses Incremental glucose values were calculated for each test at the capillary blood sampling time points (baseline, 15, 30, 45, 60, 90, 120 min) as the difference in glycemia from the 0 min value to derive postprandial glucose. In case of no significant difference between the 40 min baseline and 0 min glucose samples, iAUC was calculated using the 0–120 min samples only. Incremental area under the curve (iAUC) was calculated geometrically using positive area under the glucose and insulin curves ignoring areas below the 0 value (Wolever et al., 1991) for each subject and then averaged for each treatment. Peak incremental blood glucose was calculated by averaging the maximum incremental glucose values for each test treatment. All results were expressed as mean7standard error of the mean (SEM) and differences were considered significant at a two-sided po0.05. Statistical analyses were done using the Number Cruncher Statistical System (NCSS 2000 statistical software, Kaysville, Utah) and SPSS release 16.0 (SPSS Inc, Chicago, IL). The primary outcome, incremental area under the glucose curve (iAUC glucose), and secondary outcomes, incremental area under the insulin curve (iAUC insulin), peak postprandial glycemia and ISI were used to compare the 8 ginseng treatments (whole root KRG and AG, 30%, 50% and 70% alcohol extracts of KRG and AG) to each other and to the placebos (2 cornstarch controls), using ANOVA and the Tukey Kramer post-hoc test. The normality of the data was determined prior to the analysis. The interactive time-treatment effects on plasma glucose and insulin were assessed using 2-factor ANOVA and were considered significant at po0.05. One-factor repeated measures ANOVA and Tukey Kramer post-hoc test were used to explore significant interactions at individual time points. Absolute baseline ( 40 min) differences between individual treatment groups and absolute pre-test treatment effects (0 min) were also assessed. Based on an expected coefficient of variation (CV) of 20%, 12 subjects would be required to detect an 18% difference in glucose iAUC (paired T, 80% power, α ¼0.05). Given an attrition rate of 20%, a total of 16 subjects were recruited.

57

2.8. Ginseng analyses Major active ginsenosides were extracted from American and Korean red ginseng samples with 50:50 MeOH/water. Both ginseng species were analyzed using PPD-type (Rb1, Rb2, Rc, Rd) and PPT-type (Rg1, Re, Rf) ginsenoside standards. Residues were quantified by external calibration versus standards containing Rg1, Re, Rb1, Rc, Rb2 and Rd. Analysis was conducted using HPLC-UV techniques at 203 nm on a C18 column a reverse-phase Beckman ultrasphere C-18, 5 μm octadecylsilane, 250  4.6 mm column; mobile phases – de-ionized water and acetonitrile. Confirmation of Ginsenosides by API-ES  ve ion MS. Analysis was conducted by Natural Factors, Burnaby, B.C.

3. Results 3.1. Participant characteristics, compliance and symptoms Sixteen individuals provided written informed consent to participate in the study. Three participants dropped out due to time conflicts. Thirteen individuals completed the blood glucose testing (5 males, 8 females), mean 7SD age: 28.9 7 9 years, BMI: 26.3 73 kg/m2, FBG: 4.21 70.04 mmol/L. According to the selfreport questionnaire, subjects remained within the prescribed inclusion criteria for the duration of the study and there were no reported adverse post-test symptoms across treatments and controls. There was no difference in participant fasting blood glucose levels prior to glucose loading. All subjects reported consumption of at least 150 g of carbohydrate on each of the three days preceding study visits. Baseline anthropometry, FBG, exercise, diet and lifestyle patters, were maintained throughout the study.

3.2. Effects of ginseng treatments on postprandial glycemia Plasma glucose for Korean and American ginseng treatments are shown in Tables 1 and 2 respectively. Absolute peak glucose measures showed no significant difference between treatments. Two-factor repeated measures ANOVA indicated no significant time-treatment interactive effects on glucose metabolism. Incremental peak PPG was not significantly changed by any of the treatments compared to control. AG 70% extract was 16% lower than control, though this difference did not reach significance. Overall, the extract treatments did not show differential effects on PPG parameters.

Table 1 Plasma glucose for Korean red ginseng (absolute values prior to and incremental values after the start of the 50 g oral glucose challenge) and related glycemic parameters in 13 healthy subjects who consumed ten treatments in random sequence. KRG whole root and control were compared to 3 aqueous ethanol extracts, with 30%, 50% and 70% ethanol in a crossover design. PG parameters Absolute PG (values reported in mmol/L) Baseline Pre-OGTT Incremental PG (mmol/L)

Peak PG IAUC (mmol/L  min/L) Reported values are mean 7 SEM, 95% CI.

Time point (min)

Control

KRG whole root

KRG 30% extract

KRG 50% extract

KRG 70% extract

 40 0 15 30 45 60 90 120 — —

4.36 70.10 4.36 70.07 1.53 70.21 2.92 70.21 2.86 70.38 1.84 70.35 1.22 70.23 0.47 70.26 7.7 70.26 211.0 720.2

4.2 70.12 4.4 70.11 1.6770.031 3.05 70.24 2.65 70.29 2.25 70.30 1.45 70.26 0.20 70.25 7.7 70.26 210.5 721.4

4.32 70.09 4.12 70.09 1.79 70.25 2.87 70.31 2.49 70.30 2.23 70.33 1.60 70.23 0.23 70.23 7.5 70.20 209.9 719.3

4.34 7 0.06 4.26 7 0.08 2.09 7 0.27 2.777 0.34 2.75 7 0.28 2.177 0.36 1.717 0.23 0.117 0.32 7.5 7 0.20 215.7 7 22.0

4.2 70.11 4.1 70.10 2.00 70.20 2.98 70.22 2.42 70.28 2.23 70.38 1.47 70.33 0.6770.29 7.5 70.27 217.1 723.1

58

L.R. De Souza et al. / Journal of Ethnopharmacology 159 (2015) 55–61

Table 2 Plasma glucose for American ginseng (absolute values prior to and incremental values after start of the 50 g oral glucose tolerance test) and related glycemic parameters in 13 healthy subjects who consumed in random sequence. AG whole root and control were compared to 3 aqueous ethanol extracts, with 30%, 50% and 70% ethanol in a crossover design. PG parameters Absolute PG (values reported in mmol/L) Baseline Pre-OGTT Incremental PG (mmol/L)

Peak PG IAUC (mmol/L  min/L)

Time point (min)

Control

AG whole root

AG 30% extract

AG 50% extract

AG 70% extract

 40 0 15 30 45 60 90 120 — —

4.36 7 0.10 4.36 7 0.07 1.53 7 0.21 2.92 7 0.21 2.86 7 0.38 1.84 7 0.35 1.22 7 0.23 0.47 7 0.26 7.7 7 0.26 211.0 7 20.2

4.007 0.16 4.23 7 0.10 1.98 7 0.20 2.78 7 0.28 2.29 7 0.26 1.777 0.35 1.577 0.34 0.277 0.30 7.4 7 0.25 201.3 7 22.7

4.127 0.21 4.187 0.14 1.647 0.21 2.90 7 0.15 2.477 0.28 1.92 7 0.26 1.22 7 0.29 0.197 0.0.19 7.3 7 0.21 189.5 7 18.2

4.38 7 0.09 4.137 0.13 2.337 0.26 2.82 7 0.32 2.46 7 0.37 1.747 0.45 1.32 7 0.36 0.357 0.25 7.5 7 0.23 208.0 7 25.6

3.93 70.13 4.05 70.13 1.54 70.24 2.71 70.26 2.29 70.33 1.56 70.32 1.2470.30 0.48 70.25 7.1 70.22 180.7 717.3

Reported values are mean 7 SEM, 95% CI. Table 3 Plasma insulin for Korean red ginseng (absolute values prior to and incremental values after start of the 50 g oral glucose tolerance test) and related glycemic parameters in 12 healthy subjects who consumed ten treatments in random sequence. KRG whole root and control. were compared to 3 aqueous ethanol extracts, with 30%, 50% and 70% ethanol in a crossover design. PI parameters Absolute PI (values reported in mu/l) Pre-OGTT Incremental PI (mu/l)

Peak IAUC (mu/l  min/l)

Time point (min)

Control

KRG Whole

KRG 30%

KRG 50%

KRG 70%

0 30 60 120 — —

6.667 1.02 56.0 7 9.0 41.2 7 10.0 18.2 7 6.3 67.6 7 11.8 4095.5 7 886.0

6.5 70.8 42.7 78.2 37.4 77.8 9.6 72.5 54.4 79.5 3251.9 7607.3

6.187 1.06 44.6 7 9.4 40.3 7 9.1 9.4 7 3.1 55.4 7 10.2 3432.4 7 733.4

6.047 0.91 51.2 7 14.7 43.6 7 10.3 12.6 7 4.3 66.07 15.3 3875.17 872.1

6.4 7 0.9 50.6 7 10.1 48.27 9.7 11.7 7 4.1 69.9 7 10.4 4039.17 721.7

Reported values are mean þSEM, 95% CI. Table 4 Plasma insulin for American ginseng (absolute values prior to and incremental values after start of the 50 g oral glucose tolerance test) and related glycemic parameters in 12 healthy subjects who consumed ten treatments in random sequence. AG whole root and control were compared to 3 aqueous ethanol extracts, with 30%, 50% and 70% ethanol in a crossover design. PI parameters Absolute PI (values reported in mu/l) Pre-OGTT Incremental PI (mu/l)

Time point (min)

Control

AG Whole

AG 30%

AG 50%

AG 70%

0 30 60 120

6.667 1.02 56.0 7 9.0 41.2 7 10.0 18.2 7 6.3 67.6 7 11.8 4095.5 7 886.0

6.98 7 1.31 56.6 7 12.2 37.5 7 9.1 13.4 7 3.9 68.3 7 13.9 3790.4 7 725.2

6.53 70.93 54.1 712.7 38.6 711.8 9.0 73.2 66.5 711.4 3548.9 7900.0

6.277 1.22 49.07 11.5 45.6 7 16.1 11.17 4.4 58.4 7 16.2 3854.5 7 1137.3

5.94 7 0.84 43.8 7 10.5 28.4 7 8.6 11.17 3.6 56.5 7 11.4 2956.6 7 690.6

Peak IAUC (mu/l  min/l) Reported values are mean þSEM, 95% CI.

3.3. Effects of ginseng treatments on postprandial insulinemia Plasma insulin for Korean and American ginseng treatments are shown in Tables 3 and 4 respectively. Two-factor repeated measures ANOVA indicated no significant time-treatment interactive effects on insulin (p¼ 0.61). Therefore, no statistical analysis was carried out on individual time points.

ginsenoside concentration among the AG 30%, 50% and 70% extracts. KRG 30% and 70% extracts showed large differences in ginsenoside content. Ginsenoside analyses of whole root AG and KRG were not provided by the manufacturer, precluding our ability to compare the ginsenoside concentrations between extracts and their whole-root origins.

3.4. Effects of ginseng treatments on insulin sensitivity

4. Discussion

Repeated measures analysis of variance showed a significant difference in total body ISI between treatments. Higher levels corresponded to improved insulin sensitivity. ISI measures indicate that KRG 30% and AG 50% extracts were higher than control (po0.05) (Fig. 1).

We detected insulin-sensitizing effects with KRG 30% and AG 50% extracts. Insulin sensitivity is a composite indicator of glucose and insulin response and their interaction (NCEP, 2001). Moreover as a concomitant feature of glycemic control, insulin sensitivity is considered an independent predictor of T2DM and cardiovascular disease (Wallace and Matthews, 2002), and is therefore an important target therapeutic outcome. Despite our use of a wide range of ethanol extraction methodologies, we were unable to detect a significant reduction in PPG or

3.5. Ginsenoside results Detailed information on the ginsenoside data was not available; however, limited information revealed comparable levels of total

L.R. De Souza et al. / Journal of Ethnopharmacology 159 (2015) 55–61

Fig. 1. Insulin Sensitivity Index (ISI) (Matsuda and DeFronzo, 1999) calculated from fasting glucose and insulin levels and mean postprandial glucose and insulin levels for 12 healthy subjects. Treatments were administered at 3 g doses, 40 min prior to a 50 g oral glucose challenge following a double-blind, randomized crossover sequence. Treatments consisted of whole root and aqueous ethanol extracts: (A) KRG whole root; 70% KRG extract; 50% KRG extract; 30% KRG extract; control (cornstarch); KRG 30% was significantly different from control (p o 0.05) and (B) AG whole root; 70% AG extract; 50% AG extract; 30% AG extract; control (cornstarch); AG 50% was significantly different from control (p o 0.05).

plasma insulin with any of the AG and KRG treatments. The absence of PPG and insulin lowering effects might be explained by the influence of bioactive components, or degradable plant parts that are typically removed in the extraction process. These potentially bioactive constituents may operate independently or in complement with saponin concentrations. Indeed, previous research has demonstrated the adjuvant activity of saponins, where active compounds interact to enhance ginsenoside bioactivity (Sparg et al., 2004; Sun et al., 2007; Leung and Wong, 2010). The insulin sensitizing effects detected with KRG 30% supports this notion given that the lower alcohol content would retain more of the native plant matrix including unmeasured bioactive components. The findings reported herein are consistent with our previous study showing higher efficacy with the body fraction of the KRG root despite its characteristically lower ginsenoside content (De Souza et al., 2011). In another study, we found that water-extracted KRG had the lowest PPG reduction despite having the highest ginsenoside concentration across tested treatments (Sievenpiper et al., 2006). In our previous dose–response sub-study, within the therapeutic dose range for KRG, the lowest dose yielded the highest reductions in PPG while higher doses exhibited null or equivalent effects (Sievenpiper et al., 2006) and a similar trend was identifiedobserved in a previous study with AG in patients with T2DM (Vuksan et al., 2000c). Research from other groups has demonstrated deleterious effects of high ginsenoside concentrations, describing anti-complementary (Dong et al., 2001; Kim et al., 1998, 2009) actions of some ginsenosides, or paradox effects (Murray et al., 2003). Therefore, the insulin sensitizing results observed with KRG 30% compared to KRG 70% extract, supports the possibility that lower ginsenoside concentrations may be more favorable in this species and that other bioactive constituents may also be relevant. Contrary to variations in ginsenoside profiles across KRG extracts, the AG extracts had similar ginsenoside profiles regardless of the ethanol methodology applied. This may have been due

59

to a depressed ginsenoside profile in the AG whole root sample from which ethanol extracts were derived, such that lower ethanol concentrations may have sufficiently extracted the maximum ginsenoside content available in the intact root (Gafner et al., 2004; Zhong and Yue, 2005). Previous studies comparing saponin extraction methods show maximum extraction of ginsenosides Rb1 with a 50% aqueous ethanol preparation (Sievenpiper et al., 2003, 2004b; Dascalu et al., 2007) and similar effects were observed with ginsenoside Rd when compared to both lower and higher ethanol preparations (Gafner et al., 2004). The insulin sensitizing effect of AG 50% extract, despite a similar total saponin profile further supports the possibility that other, thus far unidentified bioactive moieties may exert a pharmacological role in modifying metabolic parameters. Our findings with AG extracts are not consistent with an earlier study proposing an insulin-secretagogue action in a different batch of AG root with a reported non-significant increase of 23% in fasting insulin, after long-term intake of an AG root extract (Suzuki and Hikino, 1989; Attele et al., 2002). The acute administration of AG in the present study, may not adequately describe these insulin-secretory effects detected over a long-term period. The mechanistic action of ginseng remains to be elucidated. Animal studies in glucose-loaded healthy and diabetes-induced states (Kimura et al., 1981; Yokozawa et al., 1985; Ohnishi et al., 1996; Xu et al., 2003; Kang et al., 2008) demonstrate hypoglycemic effects with ginseng. Proposed mechanisms include enhanced insulin secretion (Waki et al., 1982), increased insulin sensitivity (Ohnishi et al., 1996), and reduced glucose absorption (Onomura et al., 1999) or a combination of these (Chang et al., 2002). Other preclinical investigations suggest a glucose-lowering action of particular ginsenosides such as Rh2, specifying an insulin secretory action (Lee et al., 2010). Our findings should be interpreted in light of several limitations. The similar ginsenoside profiles across AG extracts may be explained by a depressed ginsenoside profile in the AG whole root; however, limited information on the proprietary extract samples precludes any detailed analyses to confirm this possibility. In addition, the majority of high-quality, commercial AG is supplied by crops from Ontario, Canada, whereas in the present study AG was obtained from a non-specified source in British Columbia, Canada, an area with higher moisture levels and elevated temperatures that are not congruent with the typical, ideal growth conditions for AG (Yeh et al., 2003; Dascalu et al., 2007). These growth conditions may have affected the ginsenoside profile of the AG whole root sample. Ginsenoside analyses were conducted in duplicate instead of the usual triplicate approach which would have permitted a detailed statistical comparison of individual and total ginsenosides. This information could help elucidate both the particular role of individual ginsenosides in attenuation of PPG and the distinct effects of extraction methodologies on the ginsenoside profile. Finally, we used a 50 g glucose load instead of a standardized 75 g glucose load for the ISI calculation which may have influenced the magnitude and kinetics of insulin secretion, such that individual measures of PPG and PPI were not sufficiently captured. Nonetheless, it is noteworthy that we were able to detect a significant association between IS and KRG 30% and AG 50%, despite our use of a 50 g glucose load, therefore our findings may underestimate the true effect of these extracts on IS. Despite the aforementioned limitations, the increased IS observed with both KRG 30% and AG 50% is indicative of the potential for ginseng extracts to modify metabolic parameters. Moreover, particular ginsenosides, including those not typically measured and other compositional variables may exert glycemialowering and insulin-sensitizing actions in compliment to ginsenosides. Commercially available ginseng products are predominantly prepared through alcohol methodologies, though none

60

L.R. De Souza et al. / Journal of Ethnopharmacology 159 (2015) 55–61

have been clinically and systematically evaluated. Alcohol extraction methods can facilitate the identification of bioactive saponins in a consistent, reproducible manner. The present study is the first to explore the effects of different ethanol extraction methods on AG and KRG to improve metabolic parameters in healthy subjects. The insulin-sensitizing effects observed with ginseng alcohol extracts, illustrates a promising potential for the application of extraction methodologies to produce a viable complementary therapy in the management of glycemic excursions in chronic conditions like T2DM.

Contributions of authors Leanne R. De Souza, M.Sc. conducted the study, completed experimental statistical analyses and composed the manuscript. Alexandra A. Jenkins, Ph.D., RD received the grant from the Canadian Diabetes Association, provided significant advice and consultation, and edited the manuscript. Elena Jovanovski, M.Sc. edited the manuscript. Dario Rahelić, M.D. provided significant advice and consultation. Vladimir Vuksan, Ph.D. received the grant from the Canadian Diabetes Association, provided the clinic space, resources for the study, protocol development, significant advice and consultation, and edited the manuscript.

Acknowledgments This study was funded by an Operating Grant from the Canadian Diabetes Association (Toronto, ON, Canada) (Grant no. OG-2-103038). The KRG was generously donated by the Korean Ministry of Agriculture, Seoul, Korea. Ethanol extracts and supplemental funding for insulin analysis were provided by Natural Factors, Burnaby, B.C., Canada. References Attele, A.S., Wu, J.A., Yuan, C.S., 1999. Ginseng pharmacology: multiple constituents and multiple actions. Biochemical Pharmacology 58 (11), 1685–1693. http://dx. doi.org/10.1016/S0006-2952(99)00212-9 (PMID:10571242). Attele, A.S., Zhou, Y.P., Xie, J.T., Wu, J.A., Zhang, L., Dey, L., et al., 2002. Antidiabetic effects of Panax ginseng berry extract and the identification of an effective component. Diabetes 51 (6), 1851–1858. http://dx.doi.org/10.2337/diabetes.51.6.1851 (PMID:12031973). Awang, D.V.C., 2000. The neglected ginsenosides of North American ginseng (panax quinquefolius L.). Journal of Herbs, Spices & Medicinal Plants 7, 103–109. Chang, T.K.H., Chen, J., Benetton, S.A., 2002. In vitro effect of standardized ginseng extracts and individual ginsenosides on the catalytic activity of human CYP1A1, CYP1A2, and CYP1B1. Drug Metabolism and Disposition 30, 378–384. Dascalu, A., Sievenpiper, J.L., Jenkins, A.L., Stavro, M.P., Leiter, L.A., Arnason, J.T., Vuksan, V., 2007. Five batches representative of Ontario-grown American ginseng root produce comparable reductions of postprandial glycemia in healthy individuals. Canadian Journal of Pharmacology 85 (9), 856–864 (PMID: 18066131). Dong, M., Feng, X.Z., Wu, L.J., Wang, B.X., Ikejima, T., 2001. Two new steroidal saponins from the rhizomes of Dioscorea panthaica and their cytotoxic activity. Planta Medica 67 (9), 853–857. http://dx.doi.org/10.1055/s-2001-18856 (PMID:11745024). De Souza, L.R., Jenkins, A.L., Sievenpiper, J.L., Jovanovski, E., Rahelic, D., Vuksan, V., 2011. Korean red ginseng (Panax ginseng C.A. Meyer) root fractions: differential effects on postprandial glycemia in healthy individuals. Journal of Ethnopharmacology 137 (1), 245–250. http://dx.doi.org/10.1016/j.jep.2011.05.015. Franz, M.J., Bantle, J.P., Beebe, C.A., Brunzell, J.D., Chiasson, J.L., Garg, A., et al., 2002. Evidence-based nutrition principles and recommendations for the treatment and prevention of diabetes and related complications. Diabetes Care 25 (1), 148–198. http://dx.doi.org/10.2337/diacare.25.1.148 (PMID:12502619). Gafner, S., Bergeron, C., McCollom, M.M., Cooper, L.M., McPhail, K.L., Gerwick, W.H., Angerhofer, C.K., 2004. Evaluation of the efficiency of three different solvent systems to extract triterpene saponins from roots of Panax quinquefolius using high-performance liquid chromatography. Journal of Agricultural and Food Chemistry 52, 1546–1550. http://dx.doi.org/10.1021/jf0307503. Kang, K.S., Yamabe, N., Kim, H.Y., Park, J.H., Yokozawa, T., 2008. Therapeutic potential of 20(S)-ginsenoside Rg(3) against streptozotocin-induced diabetic

renal damage in rats. European Journal of Pharmacology 591, 266–272 (PMID: 18611400). Kim, D.S., Oh, S.R., Lee, I.S., Jung, Y., Jong, D.P., Shin, I.K., et al., 1998. Anticomplementary activity of ginseng saponins and their degradation products. Phytochemistry 47 (3), 397–399. http://dx.doi.org/10.1016/S0031-9422(97) 00580-3 (PMID:9433814). Kim, M., Ahn, B.Y., Lee, J.S., Chung, S.S., Lim, S., Park, S.G., Jung, H.S., Lee, H.K., Park, K.S., 2009. The ginsenoside Rg3 has a stimulatory effect on insulin signaling in L6 myotubes. Biochemical and Biophysical Research Communications 389, 70–73 (PMID: 19699714). Kimura, M., Waki, I., Chujo, T., Kikuchi, T., Hiyama, C., Yamazaki, K., Tanaka, O., 1981. Effects of hypoglycemic components in ginseng radix on blood insulin level in alloxan diabetic mice and on insulin release from perfused rat pancreas. Journal of Pharmacobio-Dynamics 4, 410–417 (PMID: 7026762). Lee, S.Y., Kim, Y.K., Park, N.I., Kim, C.S., Lee, C.Y., Park, S.U., 2010. Chemical constituents and biological activities of the berry of Panax ginseng. Journal of Medicinal Plants Research 4, 349–353 (ISSN 1996-0875). Leung, K.W., Wong, A.S., 2010. Pharmacology of ginsenosides: a literature review. Chinese Medicine 5 (20), 1–7 (PMID:20537195). Li, T.S.C., Mazza, G., Cottrell, A.C., Gao, L., 1996. Ginsenosides in roots and leaves of American ginseng. Journal of Agricultural and Food Chemistry 44 (3), 717–720. http://dx.doi.org/10.1021/jf950309f. Matsuda, M., DeFronzo, R.A., 1999. Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care 22 (9), 1462–1470. Murray, R.K., Granner, D.K., Mayes, P.A., Rodwell, V., 2003. Harper's Illustrated Biochemistry. L.M. Publications, McGraw-Hill Companies, New York. NCEP, 2001. Executive summary of the third report of the National Cholesterol Education Program (NCEP) Expert Panel on detection, evaluation and treatment of high blood cholesterol in adults (Adult treatment panel III). Journal of the American Medical Association 285 (19), 2486–2497 (PMID: 11368702). Ohnishi, Y., Takagi, S., Miura, T., Usami, M., Kako, M., Ishihara, E., Yano, H., Tanigawa, K., Seino, Y., 1996. Effect of ginseng radix on GLUT2 protein content in mouse liver in normal and epinephrine-induced hyperglycemic mice. Biological & Pharmaceutical Bulletin 19, 1238–1240 (PMID: 8889050). Onomura, M., Tsukada, H., Fukuda, K., Hosokawa, M., Nakamura, H., Kodama, M., Ohya, M., Seino, Y., 1999. Effects of Ginseng Radix on sugar absorption in thesmall intestine. American Journal of Chinese Medicine 27, 347–354 (PMID: 10592843). Sievenpiper, J.L., Arnason, J.T., Leiter, L.A., Vuksan, V., 2003. Variable effects of American ginseng: a batch of American ginseng (Panax quinquefolius L.) with a depressed ginsenoside profile does not affect postprandial glycemia. European Journal of Clinical Nutrition 57 (2), 243–248. http://dx.doi.org/10.1038/sj. ejcn.1601550 (PMID:12571655). Sievenpiper, J.L., Arnason, J.T., Leiter, L.A., Vuksan, V., 2004a. Decreasing, null and increasing effects of eight popular types of ginseng on acute postprandial glycemic indices in healthy humans: the role of ginsenosides. Journal of the American College of Nutrition 23 (3), 248–258 (PMID:15190050). Sievenpiper, J.L., Arnason, J.T., Vidgen, E., Leiter, L.A., Vuksan, V., 2004b. A systematic quantitative analysis of the literature of the high variability in ginseng (Panax spp): should ginseng be trusted in diabetes? Diabetes Care 27 (3), 839–840 (PMID:14988315). Sievenpiper, J.L., Sung, M.K., Di Buono, M., Seung-Lee, K., Nam, K.Y., Arnason, J.T., et al., 2006. Korean red ginseng rootlets decrease acute postprandial glycemia: results from sequential preparation- and dose-finding studies. Journal of the American College of Nutrition 25 (2), 100–107 (PMID:16582025). Sparg, S.G., Light, M.E., van Staden, J., 2004. Biological activities and distribution of plant saponins. Journal of Ethnopharmacology 94 (2-3), 219–243. http://dx.doi. org/10.1016/j.jep.2004.05.016 (PMID:15325725). Sun, J., Hu, S., Songa, X., 2007. Adjuvant effects of protopanaxadiol and protopanaxatriol saponins from ginseng roots on the immune responses to ovalbumin in mice. Vaccine 25 (6), 1114–1120. http://dx.doi.org/10.1016/j.vaccine.2006.09.054 (PMID:17069940). Suzuki, Y., Hikino, H., 1989. Mechanisms of hypoglycemic activity of panaxans A and B, glycans of Panax ginseng roots: effects on plasma-level, secretion, sensitivity and binding of insulin in mice. Phytotherapy Research 3 (1), 20–24. http://dx.doi.org/10.1002/ptr.2650030106. Vuksan, V., Sievenpiper, J.L., 2005. Herbal remedies in the management of diabetes: lessons learned from the study of ginseng. Nutrition, Metabolism and Cardiovascular Diseases 15 (3), 149–160. http://dx.doi.org/10.1016/j.numecd.2005.05.001 (PMID:15955462). Vuksan, V., Sievenpiper, J.L., Koo, V.Y., Francis, T., Beljan-Zdravkovic, U., Xu, Z., Vidgen, E., 2000a. American ginseng (Panax quinquefolius L.) reduces postprandial glycemia in nondiabetic subjects and subjects with type 2 diabetes mellitus. Archives of Internal Medicine 160 (7), 1009–1013. http://dx.doi.org/ 10.1001/archinte.160.7.1009 (PMID:10761967). Vuksan, V., Sievenpiper, J.L., Xu, Z., Beljan-Zdravkovic, U., Jenkins, A.L., Arnason, J.T., et al., 2000b. Ginseng and Diabetes: a new way to use an old medicine. Canadian Journal of Diabetes 25 (Suppl. 2), S111–S1120. Vuksan, V., Stavro, M.P., Sievenpiper, J.L., Beljan-Zdravkovic, U., Leiter, L.A., Josse, R.G., et al., 2000c. Similar postprandial glycemic reductions with escalation of dose and administration time of American ginseng in type 2 diabetes. Diabetes Care 23 (9), 1221–1226. http://dx.doi.org/10.2337/diacare.23.9.1221 (PMID:10977009). Vuksan, V., Stavro, M.P., Sievenpiper, J.L., Koo, V.Y., Wong, E., Beljan-Zdravkovic, U., et al., 2000d. American ginseng improves glycemia in individuals with normal

L.R. De Souza et al. / Journal of Ethnopharmacology 159 (2015) 55–61

glucose tolerance: effect of dose and time escalation. Journal of the American College of Nutrition 19 (6), 738–744 (PMID:11194526). Vuksan, V., Sievenpiper, J.L., Wong, J., Xu, Z., Beljan-Zdravkovic, U., Arnason, J.T., et al., 2001. American ginseng (Panax quinquefolius L.) attenuates postprandial glycemia in a time-dependent but not dose-dependent manner in healthy individuals. American Journal of Clinical Nutrition 73 (4), 753–758 (PMID:11273850). Vuksan, V., Sung, M.K., Sievenpiper, J.L., Stavro, P.M., Jenkins, A.L., Di Buono, M., et al., 2006. Korean red ginseng (Panax ginseng) improves glucose and insulin regulation in well-controlled, type 2 diabetes: results of a randomized, doubleblind, placebo-controlled study of efficacy and safety. Nutrition, Metabolism and Cardiovascular Diseases 18 (1), 46–56 (PMID:16860976). Waki, I., Kyo, H., Yasuda, M., Kimura, M., 1982. Effects of a hypoglycemic component of ginseng radix on insulin-biosynthesis in normal and diabetic animals. Journal of Pharmacobio-Dynamics 5, 547–554 (PMID: 6759629). Wallace, T.M., Matthews, D.R., 2002. The assessment of insulin resistance in man. Diabetic Medicine 19 (7), 527–534 (PMID: 12099954). Wang, A.B., Wang, C.Z., Wu, J.A., Osinski, J., Yuan, C.S., 2005. Determination of major ginsenosides in Panax quinquefolius (American ginseng) using highperformance liquid chromatography. Phytochemical Analysis 1, 272–277. http://dx.doi.org/10.1002/pca.838 (PMID:16042154).

61

Wolever, T.M.S., Jenkins, D.J.A., Jenkins, A.L., Josse, R.G., 1991. The glycemic index: methodology and clinical implications. American Journal of Clinical Nutrition 54, 846–854 (PMID:1951155). Xie, J.T., McHendale, S., Yuan, C.S., 2005. Ginseng and diabetes. American Journal of Chinese Medicine 33, 397–404 (PMID:16047557). Xu, Q.F., Fang, X.L., Chen, D.F., 2003. Pharmacokinetics and bioavailability of ginsenoside Rb1 and Rg1 from Panax notoginseng in rats. Journal of Ethnopharmacology 84 (2-3), 187–192. http://dx.doi.org/10.1016/S0378-8741(02) 00317-3. Yin, J., Zhang, H., Ye, J., 2008. Traditional Chinese medicine in treatment of metabolic syndrome. Endocrine, Metabolic & Immune Disorders – Drug Targets 8 (2), 99–111 (PMCID: PMC2467395). Yeh, G.Y., Eisenberg, D.M., Kaptchuk, T.J., Phillips, R.S., 2003. Systematic review of herbs and dietary supplements for glycemic control in diabetes. Diabetes Care 26, 1277–1294. http://dx.doi.org/10.2337/diacare.26.4.1277. Yokozawa, T., Kobayashi, T., Oura, H., Kawashima, Y., 1985. Hyperlipemia-improving effects of ginsenoside-Rb2 in streptozotocin-diabetic rats. Chemical & Pharmaceutical Bulletin (Tokyo) 33, 3893–3898 (PMID:4092288). Zhong, J.J., Yue, C.J., 2005. Plant cells: secondary metabolite heterogeneity and its manipulation. Advances in Biochemical Engineering/Biotechnology 100, 53–88.