Determination of Major Phenolic Compounds in Echinacea spp. Raw Materials and Finished Products by High-Performance Liquid Chromatography with Ultraviolet Detection: Single-Laboratory Validation Matrix Extension (2024)

The publisher's final edited version of this article is available at J AOAC Int

Principle

This HPLC method is used to detect and quantify five phenolic compounds commonly found in Echinacea spp. raw materials, powdered extract finished products, and tinctures. The phenolic compounds are caftaric acid, chlorogenic acid, cichoric acid, cynarin, and echinacoside.

Test Materials

All test materials were stored at room temperature. Whole root and aerial parts of E. angustifolia, E. purpurea, and E. pallida were harvested in 2008 under the supervision of Wendy Applequist (Missouri Botanical Gardens, St. Louis, MO) and provided by Naturex (South Hackensack, NJ). The herbarium specimens for these three species were deposited with the Missouri Botanical Garden Herbarium, Voucher Nos. 217, 218, and 216, respectively.

1. E. purpurea whole root.—Naturex.

2. E. purpurea aerial parts.—Naturex.

3. E. angustifolia whole root.—Naturex.

4. E. angustifolia aerial parts.—Naturex.

5. E. pallida whole root.—Naturex.

6. E. purpurea aerials powdered extract.—Naturex.

7. E. angustifolia root powdered extract.—Naturex.

8. E. angustifolia root powdered extract in softgel capsule.—Supplied by commercial manufacturer.

9. E. purpurea aerials powdered extract.—Supplied by commercial manufacturer.

10. E. purpurea root glycerite tincture.—Supplied by commercial manufacturer.

11. E. angustifolia aerials tincture.—Supplied by commercial manufacturer.

12. E. angustifolia/E. purpurea roots and aerials tincture.—Supplied by commercial manufacturer.

13. E. purpurea leaf, stem, and flower in combination with elderberry and zinc in hardshell capsule.—Supplied by commercial manufacturer.

14. E. purpurea leaf, stem, flower in combination with Astragalus and Reishi in hardshell capsule.—Supplied by commercial manufacturer.

15. E. purpurea leaf, stem, flower in combination with vitamin C in hardshell capsule.—Supplied by commercial manufacturer.

Reagents and Supplies

1. Acetonitrile (C₂H₃N).—Formula weight (FW): 41.05, CAS No. 75-05-08, purity: ≥99.8% (GC), HPLC grade or equivalent. Colorless liquid, flammable, and poisonous (Fisher Scientific, Ottawa, ON, Canada).

2. Methanol (CH₃OH).—FW: 32.04, CAS No. 67-56-1, purity: 99.97%, HPLC grade or equivalent. Colorless liquid, volatile, flammable, and poisonous (Fisher Scientific).

3. Water (H₂O).—FW: 18.01, CAS No. 7732-18-5, purity: submicron filtered, HPLC grade or equivalent. Colorless liquid (Fisher Scientific).

4. o-Phosphoric acid (H₃PO₄).—FW: 98.00, CAS No. 7664-38-2, purity: ≥85.0%, HPLC grade or equivalent. Colorless liquid, irritant, and poisonous (Fisher Scientific).

5. Magnesium stearate (C₃₆H₇₀MgO₄).—FW: 594.27, CAS No. 557-04-0, vegetable grade or equivalent. White powder (Fisher Scientific).

6. Maltodextrin (C_6nH_(10n+2)O_(5n+1)).—FW: variable, CAS No. 9050-36-6, Maltrin M100 (grain derived) or equivalent. Hygroscopic white powder provided by Natural Factors Nutritional Products Ltd (Coquitlam, BC, Canada).

7. Panax quinquefolius L. root.—Obtained from the Ontario Ginseng Grower’s Association (Simcoe, ON, Canada) for use as a matrix blank (negative control) for spiking studies.

Solutions

1. Extraction solvent.—Methanol–water (60 + 40).

2. Mobile phase A.—0.1% o-Phosphoric acid in water (filtered through 0.2 µm nylon filter).

3. Mobile phase B.—Acetonitrile.

Reference Standards

For caftaric acid, cichoric acid, and echinacoside, 1000 ppm stock solutions were prepared by dissolving individual reference materials in extraction solvent. Chlorogenic acid and cynarin stock solutions were also prepared at 1000 ppm, but then diluted to 200 and 100 ppm, respectively. These stock concentrations were then diluted to appropriate concentrations to establish retention time and combined at different concentration levels for external calibration.

1. Caftaric acid (C₁₃H₁₂O₉).—FW: 312.23, CAS No. 67879-58-7, purity: 88.6%, acquired from ChromaDex (Irvine, CA), Part No. 3028, stored at −20°C and desiccated. Purity was adjusted for HPLC and presence of water and other solvents.

2. Chlorogenic acid (C₁₆H₁₈O₉).—FW: 354.31, CAS No. 327-325-6, purity: 93.9%, acquired from ChromaDex, Part No. 3450, stored at room temperature and desiccated.

3. Cichoric acid (C₂₂H₁₈O₁₂).—FW: 474.37, CAS No. 70831-56-0) purity: 97.0%, acquired from ChromaDex, Part No. 3640, stored at −20°C and desiccated.

4. Cynarin (C₂₅H₂₄O₁₂).—FW: 516.47, CAS No. 1182-34-9, purity: 95.44%, acquired from ChromaDex, Part No. 3990, stored at −20°C and desiccated.

5. Echinacoside (C₃₅H₄₆O₂₀).—FW: 786.74, CAS No. 82854-37-3, purity: 97.0%, acquired from ChromaDex, Part No. 5020, stored at −20°C and desiccated.

Apparatus

1. Centrifugal mill or grinder.—Retsch Ultra centrifugal mill ZM 100 (Retsch GmbH, Haan, Germany) or equivalent mill capable of grinding root samples to 60 mesh.

2. Analytical balance.—Mettler Toledo AE 260 analytical range (±0.1 mg; VWR International, Edmonton, AB, Canada) or equivalent.

   See Also

   00-81. Non-Substantive Revisions to the Table of Frequency Allocations, Federal RegisterTONE, EDUART Et Al v. NAUGATUCK COCA-COLA AKA COCA-COLA BOTTLING COMPANY Et Al, UWY-CV23-6070155-S, 24846827 (Conn. Super. Ct. Apr. 10, 2023)

3. Centrifuge.—Eppendorf 5804 table top centrifuge (VWR International) or equivalent

4. Wrist action shaker.—Burrell Model BT wrist action shaker (VWR International) or equivalent.

5. Syringes.—3 mL Luer-lok^® fitted with PTFE filter, 0.45 and 0.2 µm pore size, 25 mm diameter (Fisher Scientific).

6. Vortex mixer.—Table top (VWR International) or equivalent.

7. Micropipets.—Eppendorf Reference Series, 100, 200, and 1000 µL (VWR International) or equivalent.

8. HPLC system.—Agilent 1100 Series liquid chromatograph equipped with quaternary pump and degasser (G1354A), temperature-controlled column compartment (G1316A), temperature-controlled autosampler (G1327A), standard flow-cell 10 mm, 13 µL, 120 bar (G1315-60012), diode-array detector (G1315B), HPLC 2D ChemStation software (G2175AA), and online degasser (1322A; Agilent Technologies, Mississauga, ON, Canada) or equivalent.

HPLC Conditions

1. Autosampler temperature.—5°C.

2. Analytical column.—Cosmosil 5C18-AR-II, 150 × 4.6 mm id.

3. Column temperature.—25°C.

4. Detector conditions.—Monitor at 330 nm (8 nm bandwidth), no reference.

5. Flow rate.—1.5 mL/min.

6. Run time.—14.5 min with 3.5 min post time for column equilibration.

7. Injection volume.—5 µL.

8. Gradient conditions.—SeeTable 1.

   Table 1

   Gradient conditions for the separation of five phenolic compounds in echinacea test samples

   Time, min	Mobile phase B, %
   0	10
   13	22
   14	40
   14.5	40

   Open in a separate window

Calculations

The equations used to determine the average weight of finished products, based on 20 capsules, are as follows:

• Average capsule fill weight (g): AFW (g) = (C − S)/20

where C = weight of capsule with shell and fill content (g), and S = weight of empty shell (g).

• Average tablet weight (g): ATW,

• g = weight of 20 tablets/20

The calculation used to determine phenolic concentration is as follows:

• Concentration (µg/mL): C, µg/mL = (A − B)/D

where A = peak area (mAu × s), B = intercept of the calibration curve, and D = slope of the calibration curve.

To quantify the individual phenolic compounds on a % (w/w) basis, the following calculation was used:

• Phenolic, % (w/w) = [(C)(FV)(D)(100%)]/(W)

where C = concentration (µg/mL) from linear regression analysis, FV = the final volume (mL) of the sample preparation, D = the dilution factor of the sample preparation, and W = the sample weight (mg).

To quantify the individual phenolic compounds on a part per million (µg/mL) basis for tinctures, the following calculation was used:

• Phenolic in tincture, µg/mL = [(C)(FV)(D)]

where C = concentration (µg/mL) from linear regression analysis, FV = the final volume (mL) of the sample preparation, D = the dilution factor of the sample preparation (25), and W = the sample weight (mg).

For the validation study, the following equations were used for evaluating precision:

• RSD_r (found, %): RSD_r = SD(r)/mean × 100

where SD(r) = population SD (σ/n, where σ = sum of squares and n = number of replicates).

• PRSD_r (RSD_r calculated, %): PRSD_r = 2 × C^−0.15

• HorRat value: HorRat = RSD_r (found, %)/PRSDr (calculated, %)

Within-day: average and SDs of four data points within-day.

Within-laboratory: average and SDs of 12 data points over 3 days (separate batches on 3 days).

Chromatographic resolution, Rs, was calculated using the following equation:

• Rs = 2[(t_R)_A − (t_R)_B]/(W_A + W_B)

where t_R = retention time, min, W = width of peak at baseline, min, A = earlier-eluting peak, and B = later-eluting peak. Baseline resolution requires an Rs >1.5.

Preparation of Test Materials

1. Raw materials and powdered extracts.—

   1. Raw materials (root, aerials) were ground to 60 mesh powder in a grinding mill, hom*ogenized, and stored in separate plastic bags. Extract powders were mixed to ensure hom*ogeneity.

   2. 125 mg (±0.1 mg) of each powdered test material was weighed into separate 50 mL conical tubes.

   3. Using volumetric pipets, 25 mL extraction solvent (60% aqueous methanol solution) was added to each conical tube containing the test materials.

   4. Each sample tube was then mixed with a wrist action shaker for 30 min at room temperature.

   5. Tubes were centrifuged at 5000 rpm for 5 min.

   6. A portion of the supernatant was filtered through a 3 mm syringe fitted with a 0.45 µm Teflon membrane filter into an amber glass HPLC vial and capped.

2. Soft shell dry filled capsule test materials.—

   1. The total capsule content weight was determined by weighing 20 capsules. The 20 capsules were cut open with a suitable instrument.

   2. The capsule contents were emptied and combined in a conical tube. Weights of the capsule contents and the empty capsule shells were obtained and recorded. The average fill weight/capsule was calculated.

   3. The capsule contents were transferred into a 50 mL conical tube and mixed using a spatula to hom*ogenize the samples.

   4. The same procedure described in the Raw Materials and Powdered Extracts section, Steps (2) to (6), was followed.

3. Tinctures (ethanol and glycerite types).—

   1. Tincture vessels were inverted several times to ensure hom*ogeneity before aliquotting.

   2. Using a volumetric pipet, 1 mL tincture was accurately transferred into a 50 mL conical tube.

   3. Using volumetric pipets, 24 mL extraction solvent (60% aqueous methanol solution) was added the conical tube.

   4. Tubes were vortexed for 30 s to mix.

   5. Tubes were centrifuged at 5000 rpm for 5 min.

   6. A portion of the solution was filtered through a 3 mm syringe fitted with a 0.45 µm Teflon membrane filter into an amber glass HPLC vial and capped.

Preparation of Calibration Solutions

Refer to Table 2 for the approximate concentration of the individual phenolic compounds for each linearity determination (calibration level). All solutions not immediately used were stored at −20°C.

1. Linearity 1.—Refer to Table 3 for preparation instructions.

   Table 3

   Mixed calibration standard (linearity) preparation details

   Phenolic compound	Calibration std ID	Lin 1	Lin 2	Lin 3	Lin 4
   Caftaric acid	Stock solution IDa	Caf 1000	Caf 1000	Caf 1000	Caf 1000
   	Volume, µL	150	100	50	25
   Chlorogenic acid	Stock solution ID	Chloro 1000	Chloro 200	Chloro 200	Chloro 200
   	Volume, µL	50	100	50	25
   Cichoric acid	Stock solution ID	Cich 1000	Cich 1000	Cich 1000	Cich 1000
   	Volume, µL	200	100	50	25
   Cynarin	Stock solution ID	Cyn 1000	Cyn 100	Cyn 100	Cyn 100
   	Volume, µL	50	200	100	50
   Echinacoside	Stock solution ID	Ech 1000	Ech 1000	Ech 1000	Ech 1000
   	Volume, µL	300	200	100	50
   Extraction solvent volume, µL		250	300	650	825

   Open in a separate window

   ^aID = Identification code.

2. Linearity 2.—Refer to Table 3 for preparation instructions.

3. Linearity 3.—Refer to Table 3 for preparation instructions.

4. Linearity 4.—Refer to Table 3 for preparation instructions.

5. Linearity 5.—100 µL of Linearity 2 solution was diluted with 900 µL extraction solvent and mixed well.

6. Linearity 6.—100 µL Linearity 3 solution was diluted with 900 µL extraction solvent and mixed well.

7. Linearity 7.—100 µL Linearity 5 solution was diluted with 900 µL extraction solvent and mixed well.

SLV Parameters—Matrix Extension

This method was validated according to AOAC INTERNATIONAL guidelines for conducting an SLV (49).

1. Selectivity.—The selectivity of the method for the targeted phenolic compounds was established by injecting each individual reference analyte and comparing to the chromatographic profile of the test articles to establish that resolution between each analyte was achieved. An Rs of >1.5 between closely eluting components was deemed acceptable.

2. Linearity.—The linearity for each analyte was evaluated using seven-point standard calibration curves. Calibration solutions for constructing calibration curves were prepared as described above. The calibration curves for each phenolic compound were plotted, and simple linear regression used to determine the slope and y-intercept of each curve for each analyte. Calibration curves were visually inspected to confirm linearity and r² values for the regression curves were calculated for each quantified phenolic compound. An r² of ≥99.5% was deemed acceptable for quantification.

3. Method detection limit (MDL) and LOQ.—The absence of a suitable sample blank precluded the use of the International Union for Pure and Applied Chemistry method for determination of detection limits for the analytes. As an alternative, the detection limit for each analyte was determined using the U.S. Environmental Protection Agency MDL protocol (50). The MDL is defined as the minimum concentration of substance that can be measured and reported with 99% confidence that the analyte concentration is greater than zero. Seven replicates of a standard solution containing very low concentrations of each analyte were used to determine the MDL. The concentration of each replicate was calculated. The LOQ was calculated as 10 times the sample SD of the replicates used to determine MDL.

4. Precision.—Precision was evaluated by analyzing multiple replicates of each test sample. Four replicates of each material were prepared and analyzed on each of 3 separate days, resulting in 12 replicates for each test material. The within-day, between-day, and total SDs were calculated for each individual phenolic compound for each of the 15 materials using single factor analysis of variance (ANOVA) with an α-value of 0.5. In all cases, null hypothesis (H₀) for the analysis was that determination of analyte content does not differ between days; the alternative hypothesis (H_A) was that determination of analyte content does differ between days. A “pass” under ANOVA indicates acceptance of H₀. Finally, the HorRat value for each analyte in each material was calculated (51, 52). Values between 0.3 and 1.3 were considered acceptable for SLV.

5. Accuracy.—Method extraction efficiency in raw materials was evaluated by performing a spike recovery study at three levels of the five analytes [1.8, 4.7, and 8.6% (w/w) total phenolics] onto 125 mg of P. quinquefolius L. root powder ground to 60 mesh. Seven replicates were prepared/level. As a check on extraction efficiency from a commercial extract, a matrix model of 99% maltodextran and 1% magnesium stearate was used. Three different levels each of caftaric acid, chlorogenic acid, cynarin, echinacoside, and cichoric acid were spiked onto 125 mg of a negative control material consisting of 99% maltodextran and 1% magnesium stearate. While some commercially available echinacea extracts (loose or in capsule form) claim to contain as much as 4% “echinacosides,” our laboratory has observed actual levels between 1.0 and 2.8% (w/w) total phenolics. Typical extracts contain E. purpurea root (no cynarin or echinacoside) or aerials (no cynarin), and occasionally E. angustifolia aerial parts (all analytes). The spike study was meant to capture the range of 0.9–3.8% (w/w) total phenolics.

6. Stability.—To determine if cichoric acid degrades during extraction and analysis, dried root samples of E. angustifolia DC and E. purpurea (L.) Moench were ground to 60 mesh. Each was extracted four times with four different extraction solvents: 100% water; methanol–water (60 + 40); methanol–water (60 + 40), with 1 mM ascorbic acid; and ethanol–water (20 + 80). Following the validated method, 25 mL extraction solvent was added to 125 mg ground echinacea and vortex-mixed for 30 s. Exactly 1 mL suspension was removed, filtered through a 0.2 µm syringe filter into an HPLC vial, and immediately analyzed by HPLC (time 0). Exactly 1 mL extraction solvent was added to the tube to replace the lost volume, and the mixture was shaken for 20 min on a wrist-action shaker at 385 osc./min. Again, 1 mL suspension was removed, filtered into an HPLC vial, and analyzed immediately (time 20). Another 1 mL of fresh extraction solvent was added back to the tube. This process was repeated at 20 min intervals until aliquots from 0, 20, 40, 60, and 80 min of shaking were prepared for HPLC analysis. Cichoric acid was quantified against an 11-point external standard calibration curve ranging from 1 to 200 µg/mL cichoric acid prepared in methanol.

In a separate experiment, stability of extracted test solutions was assessed by combining 125 mg E. purpurea with 25 mL methanol–water (60 + 40), vortexing 30 s, and shaking for 20 min on a wrist-action shaker. A portion of the final extracted solution was filtered into seven separate HPLC vials kept at room temperature. From six of these vials, 30 consecutive HPLC injections were made over the course of 9 h. The seventh vial was stored at room temperature for 6 days and then analyzed. The concentration for each of the five analytes was calculated for each sample. Degradation would be indicated if the actual concentrations were less than the expected concentrations.

Method Validation Results—Performance Characteristics

Identification of analytes in test materials was performed by comparing peak t_R values and UV profiles to the individual reference standards diluted to within the method calibration curve concentrations. A gradient elution was used for the analysis of the five major phenolic compounds in Echinacea spp. (Table 1). The order of elution was caftaric acid (4.18–4.23 min), chlorogenic acid (4.50–4.57 min), cynarin (7.57–7.69 min), echinacoside (7.81–7.92 min), and cichoric acid (12.96–13.13 min). A representative chromatogram of a mixed calibration standard illustrating this elution order can be seen in Figure 1.

Quantification of the analytes was carried out by linear regression analysis using quadruplicate samples prepared on three separate days at seven concentration levels. The analytical range used for each phenolic is listed in the section Preparation of Calibration Solutions. Test matrixes included E. purpurea, E. angustifolia, and E. pallida in root and/or aerial raw materials; E. purpurea and E. angustifolia in powdered extract (loose, in capsule as single botanical ingredient, or in capsule with other botanical ingredients); and E. purpurea and/or E. angustifolia in tincture (ethanol or glycerite).

Selectivity

Baseline resolution (R_s >1.5) was achieved for each analyte within the calibration range. R_s >4.0 and R_s >3.0 was achieved between caftaric acid/chlorogenic acid and cynarin/echinacoside peaks, respectively. There was no evidence of chromatographic interference with analytes of interest by goldenseal, zinc, vitamin C, reishi, astragalus, or elderberry in formulations.

Linearity

All of the calibration curves generated over the course of the study appeared linear upon visual inspection. All of the RSDs were above 99.5%. These results confirm that the curves were linear over the expected concentration range for echinacea materials.

MDL and LOQ

Variance checks showed that the method used was applicable for the analytes. The MDL and LOQ for each of the analytes are reported in Table 4.

Precision

For some of the test materials, chlorogenic acid, cynarin, and echinacoside were not detected; for that reason, no response values were reported for these analytes, and no precision analysis was performed. The responses observed for all other analytes, in all test articles, were above the detection limit of the method and were thus reported as detected. For all reported materials, ANOVA indicated no significant differences for between-day precision. Average HorRat values for raw materials and dry finished products (Tables 5a and b) were 0.49 and 0.55, respectively, while average HorRat values for the tinctures (Table 5c) were lower than expected at 0.19. The low tincture HorRat values could be attributed to the use of volumetric glassware and high within-laboratory precision since they were consistently low among all analytes.

Accuracy

Two spike recovery studies were conducted to determine method accuracy. The first study was designed to emulate raw Echinacea spp. materials containing 1.8, 4.7, and 8.6% (w/w) total phenolics (sum of all five analytes) spiked onto P. quinquefolius L. root powder. Recovery (Table 6) over the three levels, averaged over all samples, was 99.8%. A second study used materials designed to resemble commercial Echinacea spp. extracts containing 0.9, 1.8, and 3.8% (w/w) total phenolics. The average recovery (Tables 7a–c) over these three levels was 100.0% (1.39% RSD).

Stability

Significant cichoric acid degradation was observed when E. purpurea and E. angustifolia root samples were extracted with highly aqueous extraction solvents [ethanol–water (20 + 80)]. No advantage was observed when 1 mM ascorbic acid was added to the methanolic extraction solution. The cichoric acid concentrations in liquid extracts of echinacea materials prepared by the validated method [methanol–water (60 + 40)] did not degrade over the entire 100 min test period.

Concentrations of each of the five analytes in an E. purpurea root sample extracted with methanol–water (60 + 40) were found to be stable through 30 injections over a 9 h period, based on a ≤4% difference in peak areas. A sample of this solution stored at room temperature for 6 days did not show any significant degradation of any analyte.

1. Binns SE, Baum BR, Arnason JT. Syst. Bot. 2002;27:610–632. [Google Scholar]

2. Upton R, Graff A, editors. American Herbal Pharmacopoeia and Therapeutic Compendium. Scott’s Valley, CA: AHP; 2005. [Google Scholar]

3. Bauer R. In: Immunomodulatory Agents from Plants. Wagner H, editor. Switzerland: Birkhauser, Basel; 1999. pp. 41–88. [Google Scholar]

4. Lopes-Lutz D, Mudge E, Ippolito R, Brown P, Schieber A. J. Agric. Food Chem. 2011;59:491–494. [PubMed] [Google Scholar]

5. Bauer R. J. Herbs Spices Med. Plants. 1999;6:51–62. [Google Scholar]

6. Timbo BB, Ross MP, McCarthy PV, Lin CJ. J. Am. Diet. Assoc. 2006;106:1966–1974. [PubMed] [Google Scholar]

7. Barnes PM, Bloom B, Nahin RL. National Center for Health Statistics. [accessed August 25, 2011];Vital Health Stat. 2008 Series No. 12, http://www.cdc.gov/nchs/data/nhsr/nhsr012.pdf.

8. Cavaliere C, Rea P, Lynch ME, Blumenthal M. HerbalGram. 2010;86:62–65. [Google Scholar]

10. Blumenthal M, Ferrier GKL, Cavaliere C. HerbalGram. 2006;71:64–66. [Google Scholar]

11. Li TSC. Hort. Technol. 1998;8:122–129. [Google Scholar]

12. Pellati F, Benvenuti S, Magro L, Melegari M, Soragni F. J. Pharm. Biomed. 2004;35:289–301. [PubMed] [Google Scholar]

13. Thygesen L, Thulin J, Mortensen A, Skibsted LH, Molgaard P. Food Chem. 2007;101:74–81. [Google Scholar]

14. Gilroy CM, Steiner JF, Byers T, Shapiro H, Georgian W. Arch. Intern. Med. 2003;163:699–704. [PubMed] [Google Scholar]

16. Luo X, Chen B, Yao S, Zeng J. J. Chromatogr. A. 2003;986:73–81. [PubMed] [Google Scholar]

17. Perry NB, Burgess EJ, Glennie V. J. Agric. Food Chem. 2001;49:1702–1706. [PubMed] [Google Scholar]

18. Bergeron C, Livesey JF, Awang DVC, Arnason JT, Rana J, Baum BR, Letchamo W. Phytochem. Anal. 2000;11:207–215. [Google Scholar]

19. Nusslein B, Kurzmann M, Bauer R, Kreis W. J. Nat. Prod. 2000;63:1615–1618. [PubMed] [Google Scholar]

20. Bergeron C, Gafner S, Batcha LL, Angerhofer CK. J. Agric. Food Chem. 2002;50:3967–3970. [PubMed] [Google Scholar]

21. Chen A, Chan M, Brown PN. A Study to Determine if Cichoric Acid Degradation Occurs During Extraction and Analysis of Echinacea Root Materials. Pharm. Biol; Gateway to Wellness: NHPRS 6th Annual Research Conference and Tradeshow; Vancouver, BC, Canada. 2009. p. 30. [abstract] [Google Scholar]

22. Fusco D, Liu X, Savage C, Taur Y, Xiao W, Kennelly E, Yuan J, Cassileth B, Salvatore M, Papanicolaou GA. Vaccine. 2010;28:3956–3962. [PMC free article] [PubMed] [Google Scholar]

23. Mrozikiewicz PM, Bogacz A, Karasiewicz M, Mikolajczak PL, Ozarowski M, Seremak-Mrozikiewicz A, Czerny B, Bobkiewicz-Kozlowska T, Grzeskowiak E. Phytomedicine. 2010;17:830–833. [PubMed] [Google Scholar]

24. Melchart D, Walther E, Linde K, Brandmaier R, Lersch C. Arch. Fam. Med. 1998;7:541–545. [PubMed] [Google Scholar]

26. Barrett BP, Brown RL, Locken K, Maberry R, Bobula JA, D’Alessio D. Ann. Intern. Med. 2002;137:939–946. [PubMed] [Google Scholar]

27. Taylor JA, Weber W, Standish L, Quinn H, Goesling J, McGann M, Calabrese C. J. Am. Med. Assoc. 2003;290:2824–2830. [PubMed] [Google Scholar]

28. Goel V, Lovin R, Barton R, Lyon MR, Bauer R, Lee TD, Basu TK. J. Clin. Pharm. Ther. 2004;29:75–83. [PubMed] [Google Scholar]

29. Cohen HA, Varsano I, Kahan E, Sarrell EM, Uziel Y. Arch. Pediatr. Adolesc. Med. 2004;158:217–221. [PubMed] [Google Scholar]

31. Yale SH, Liu K. Arch. Intern. Med. 2004;164:1237–1241. [PubMed] [Google Scholar]

32. Turner RB, Bauer R, Woelkart K, Hulsey TC, Gangemi JD. N. Engl. J. Med. 2005;353:341–348. [PubMed] [Google Scholar]

34. Barrett B, Brown R, Rakel D, Mundt M, Bone K, Barlow S, Ewers T. Ann. Intern. Med. 2010;153:769–777. [PMC free article] [PubMed] [Google Scholar]

35. Linde K, Barrett B, Woelkart K, Bauer R, Melchart D. Cochrane Database Syst. Rev. 2006;1 CD000530. [PubMed] [Google Scholar]

36. Schoop R, Klein P, Suter A, Johnston SL. Clin. Ther. 2006;28:174–183. [PubMed] [Google Scholar]

37. Gillespie EL, Coleman CI. Conn. Med. 2006;70:93–97. [PubMed] [Google Scholar]

38. Shah SA, Sander S, White CM, Rinaldi M, Coleman CI. Lancet Infect. Dis. 2007;7:473–480. Erratum(2007)in Lancet Infect. Dis.7, 580. [PMC free article] [PubMed] [Google Scholar]

40. Wolsko PM, Solondz DK, Phillips RS, Schachter SC, Eisenberg DM. Am. J. Med. 2005;118:1087–1093. [PubMed] [Google Scholar]

41. Gagnier JJ, DeMelo J, Boon H, Rochon P, Bombardier C. Am. J. Med. 2006;119:800.e1–800.e11. [PubMed] [Google Scholar]

43. Glowniak K, Zgórka G, Kozyra M. J. Chromatogr. A. 1996;730:25–29. [Google Scholar]

44. Wolf H, Zundorf J, Winckler T, Bauer R, Dingermann T. Planta Med. 1999;65:773–774. [PubMed] [Google Scholar]

45. Sloley BD, Urichuk LJ, Tywin C, Coutts RT, Pang PKT, Shan JJ. J. Pharm. Pharmacol. 2001;53:849–857. [PubMed] [Google Scholar]

46. Luo W, Ang CYW, Gehring TA, Heinze TM, Lin LJ, Mattia A. J. AOAC Int. 2003;86:202–208. [PubMed] [Google Scholar]

47. Castro J, Krishna M, Choiniere J, Marcus R. Anal. Bioanal. Chem. 2010;397:1259–1271. [PubMed] [Google Scholar]

48. Brown PN, Chan M, Betz JM. Anal. Bioanal. Chem. 2010;397:1883–1892. [PubMed] [Google Scholar]

51. Horwitz W. Anal. Chem. 1982;54:67A–76A. [Google Scholar]

52. Horwitz W, Albert R. J. AOAC Int. 2006;89:1095–1109. [PubMed] [Google Scholar]