ABSTRACT
Seventeen Chinese adzuki (Vignaumbellata.) bean varieties were collected and milled. The nutritional compositions including starch, fat, protein and phytochemicals were analyzed. Results obtained showed that these adzuki beans have a high content of resistant starch which accounts for about 23.57% of total starch, and have suitable amino acid constitutions. Palmitic acid (27.68%), linoleic acid (33.11%) and linolenic acid (26.61%) were the dominant fatty acids. Besides, four bound phenolic acid and two free phenolic acids were identified by HPLC. The antioxidant activity of 70% ethanol extracts and α-glucosidase inhibition activity of water extracts were evaluated. All the adzuki beans possessed strong ABTS.+ free-radical-scavenging capacity and α-glucosidase inhibition activity. Significant positive correlations (P < .01) of the antioxidant activity with total phenolic acids, total flavonoids and free caffeic acid contents were observed. These results are anticipated to providing useful information on the development of adzuki bean-based functional food.
1. Introduction
Adzuki bean (Vigna angularis) is a major pulse crop in the subgenus Ceratotropis and are mainly produced and consumed in Asia (Siriwardhane, Egawa, & Tomooka, Citation1992). Recently, adzuki beans have attracted attention for their considerable health benefits and functional components. It is reported that adzuki bean proteins have high α-glucosidase inhibitory activity, which would delay the digestion and absorption of carbohydrates and consequently suppress the postprandial hyperglycemia (Puls, Keup, Krause, Thomas, & Hoffmeister, Citation1977; Yao, Cheng, & Ren, Citation2014). Adzuki beans polysaccharides were reported to have significant antioxidant and immunoregulatory activities (Yao, Xue, Zhu, Gao, & Ren, Citation2015). Besides, Han et al. (Citation2005) examined the effects of adzuki-bean-resistant starch on serum cholesterol in rats fed a cholesterol diet and observed that adzuki-resistant starch has a serum cholesterol-lowering function via enhancement of the hepatic low-density lipoprotein-receptor mRNA and cholesterol 7α-hydroxylase mRNA levels. In addition, phytochemicals including phenolic acids and flavonoids extracted from adzuki bean showed significant antioxidant and radical-scavenging properties (Amarowicz, Estrella, Hernández, & Troszyska, Citation2008).
In China, adzuki beans is widely cultivated and bred. As mentioned above, most studies on adzuki beans have focused on extraction of functional components from adzuki beans and on evaluation of their biological activities. There is little information on comparison of the nutritional compositions of Chinese adzuki beans varieties which are majorly planted in China. Also, research on differences in the antioxidant and antidiabetic potential of different adzuki bean varieties is limited.
The present study was designed (1) to compare the nutritional compositions, (2) to evaluate the antioxidant activity and α-glucosidase inhibition activity of 17 Chinese adzuki bean (Vigna umbellata) varieties and (3) to investigate correlations between the phytochemicals and biological activities.
2. Materials and methods
2.1. Materials
Seventeen Chinese adzuki bean varieties (Longxiaodou 3, Xiaofeng 2, Zhonghong 7, Jihong 10, Nihewan adzuki bean, Tangshan adzuki bean, Jingnong 6, Jingnong 8, Suhong 2, Baihong 2, Baihong 5, Jihong 9218, Jihong 352, Baohong 947, Bao 876–16, Pinhong 2000–47 and Pinhong 2000–107) were collected. All samples were dried at 40°C, ground in a laboratory mill and passed through an 80-mesh screen sieve to obtain adzuki bean flour.
2.2. Chemicals
Standards of 3-(3,4-dihydroxyphenyl 2-propenoic acid (caffeic acid), 4-Hydroxy 3,5-dimethoxybenzoic acid (syringic acid), 4-hydroxycinnamic acid (p-coumaric acid), 4-hydroxy-3-methoxycinnamic acid (ferulic acid), 6-hydroxy-2,5,7,8-tetramethylchromane-2-carboxylic acid (Trolox), Folin–Ciocalteu phenolic reagent, 2,2′-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) ammonium salt (ABTS), potassium, rutin and aluminm chloride hexahydrate were purchased from Sigma-Aldrich (St. Louis, MO, USA). Mixed amino acid standard H was produced by Wako Pure Chemical Industries. All the other chemicals are of analytical grade and were obtained from Beijing Chemical Reagent (Beijing, China). All of the analytic grade solvents for high-performance liquid chromatography (HPLC) were purchased from Fisher Chemicals (Shanghai, China).
2.3. Nutritional composition analyses
Total starch, amylose and resistant starch were determined using Megazyme kits (Megazyme International Ireland, Bray Business Park, Bray, Co. Wicklow, Ireland). Results of amylose and resistant starch were expressed as the percentage of total starch. Total fat content was determined by the method 985.29 according to AOAC (Citation1990). Fatty acid analyses were carried out according to the method of Miao et al. (Citation2010) on a gas chromatography (GC) (Agilent 6890. USA) equipped with a flame ionization detector (FID). Protein was determined by the Kjeldahl method (979.09; AOAC, Citation1990), using a nitrogen to protein conversion factor of 5.71. The amino acid profile analyses were performed by reverse phase-high performance liquid chromatography (RP-HPLC) after 22 h hydrolysis at 110°C with 6 M HCl and further democratisation with o-phthaldialdehyde (OPA) and fluorenylmethyl chloroformate (FOMC chloride) (Qin et al., Citation2014).
2.4. Determination of total flavonoid content (TFC)
One gram of adzuki bean flour was extracted with 40 mL of 70% methanol by shaking in a water bath maintained at 70°C for 2 h. The solution was centrifuged at 1500 g for 10 min. A 1 mL supernatant was dried by a freeze drier. The appropriate dilution of extraction (0.5 mL) was mixed with 1.5 mL of 95% ethanol, 0.1 mL of 10% aluminium chloride hexahedron (AlCl3), 0.1 mL of 1 M potassium acetate (CH3COOK) and 2.8 mL of demonized water. After incubation at room temperature for 40 min, the absorbance of the reaction mixture was measured at 415 nm against a demonized water blank at a Smart SpecTM Plus spectrophotometer (Bio-RAD USA). The TFC was determined on the basis of a calibration curve of authentic rutin (Woisky & Salatino, Citation1998).
2.5. Determination of total phenolic content (TPC)
One gram of adzuki bean flour was extracted with 40 mL of 70% methanol by shaking in a water bath maintained at 70°C for 2 h. The solution was centrifuged at 1500 g for 10 min. One milliliters of supernatant were dried by a freeze drier. TPC was evaluated using the Folin–Ciocalteu method described previously (Slinkard & Singleton, Citation1977) with some modification. Briefly, 50 μL of the extract was mixed with 5 mL of distilled water followed by the addition of 500 μL of 1 M Folin–Ciocalteu reagent and 500 μL of a 20% (w/v) Na2CO3 solution. The mixture was thoroughly mixed and allowed to stand for 60 min at room temperature before the absorbance was measured at 765 nm (Bio-Rad Smart Spec Plus Spectrophotometer, Hercules, USA). Quantification was done with respect to the standard curve of gallic acid. The results were expressed as milligrams of gallic acid equivalent (GAE) per gram of flour.
2.6. HPLC analysis of individual phenolic acids
Bound phenolic acids were extracted according to the method of Zhang, Wang, Yao, Yan, and He (Citation2012) and free phenolic acids in the sample flour were extracted according to a reported back method (López et al., Citation2013). Two kinds of extracts were stored at –20°C until HPLC analyses. The HPLC system was equipped with an Agilent-1100 UV detector and an Agilent TC-C18 (2504.60 mm, 5 um). The wavelength of the UV detector was set at 280 and 320 nm. The mobile phase was a mixture of solvent A (HPLC water containing 0.05% TFA) and solvent B (acetonitrile: MeOH: TFA = 30:10:0.05). The gradient elution was programmed as follows: from 10% to 12% B in 16 min; from 12% to 38% B in 9 min; from 38% to 70% B in 7 min; from 70% to 85% B in 8 min and from 85% to 100% B in 10 min. The flow rate was fixed at 1.0 mL/min, and the injection volume was 20 μL. Each phenolic acid was quantified according to its calibration curve.
2.7. Antioxidant activity assay
All dried samples were ground in a laboratory mill and passed through a sieve (80 meshes). Bean samples (1 g) were extracted twice in 10 mL of 70% ethanol for 2 h at room temperature. The ABTS.+ free-radical-scavenging activity was determined using the method used by Re et al. (Citation1999) and Wojdyło, Oszmiański, and Czemerys (Citation2007). The determination of ABTS.+ free-radical-scavenging activity was determined as described in Yao, Sang, Zhou, and Ren (Citation2010). Briefly, ABTS was removed from redistilled water at a concentration of 7 μM/L. ABTS.+ radical cation was obtained by reacting ABTS stock solution with 2.45 mM/L potassium persulfate and kept at room temperature in the dark for 16 h. For the study of infusion, the resulting solution containing the ABTS.+ solution was diluted with redistilled water to an absorbance of 0.70 (0.02) at 734 nm and equilibration at 30°C. Then a reagent blank reading was made. After the addition of 3.0 mL of diluted ABTS.+ solution (A 734 nm = 0.70 ± 0.02) to 30 μL of the extracts or Trolox (prepared in DMSO for use as standard), the absorbance was taken exactly 6 min after initial mixing. The results were reflected in μM of Trolox equivalents (TE) per gram. All determinations were made in triplicate.
2.8. α-Glucosidase inhibition assay
One gram of adzuki bean flour was extracted with 10 mL of distilled water by shaking in a water bath maintained at 45°C for 2.5 h. The solution was centrifuged at 3000 g for 10 min. One milliliters of supernatant were used to analyze the α-glucosidase inhibition activity. The α-glucosidase inhibition activity was calculated as previously described (Nishioka, Kawabata, & Aoyama, Citation1998; Yao, Sang, Zhou, & Ren, Citation2009). The α-glucosidase solution was prepared by mixing 0.1 g rat intestinal acetone powder with 3 mL 0.1 M phosphate buffer (pH 7.0) in a plastic test tube extracted with ultrasound at 4°C for 0.5 min, and repeated 12 times, then centrifuged at 3500 × g for 5 min. The supernatant was the α-glucosidase solution. The reaction mixtures contained 50 μL of total phenolic compound extracts and 100 μL of 0.1 M phosphate buffer (pH 7.0) containing α-glucosidase solution and incubated in 96 well plates at 37°C for 10 min. After preincubation, 50 μL of 5 mM p-nitrophenyl-α-D-glucopyranoside in a 0.1 M phosphate buffer (pH 7.0) was added to each well. The reaction mixtures were incubated at 37°C for 5 min. Absorbance readings were recorded at 490 nm on a microplate reader before and after incubation (BioRad, IMAX, Hercules, USA). The results were expressed as percent of α-glucosidase inhibition, and the inhibition activity was calculated according to the following equation: (Acontrol––asample)/Acontrol*100%.
2.9. Statistical analysis
All values were expressed as mean ± SD. Statistical analysis was performed using SAS (version 9.1.3) and the correlation analysis was using SPSS (version 17.0). Dunnett’s multiple range tests were used to determine the significant differences between a group means at P < .05.
3. Results and discussion
3.1. Nutritional composition
As shown in , the total starch content of these s17 adzuki bean varieties ranged from 44.55% to 53.92% of seed. Consistent with these results, Tjahjadi and Breene (Citation1984) reported the total starch content of adzuki bean is 49.98%. The amylose and resistant starch account for 11.08–26.19% and 19.92–26.90% of total starch, respectively. The variety of Jingnong 6 contained the highest resistant starch content. Resistant starch has recently attracted much interest in its non-digestibility in the small intestine. It is fermented in the gut and is generally recognized as the main components in cereals that could improve gut microbiota composition (Nielsen, Theil, Purup, Nørskov, & Bach Knudsen, Citation2015). Therefore, Jingnong 6 might be a major source of prebiotic food.
Table 1. The content of total starch, resistant starch and amylose in adzuki beans.
The content of fat and fatty acid is shown in . The variety of Jihong 352 showed the highest fat content (6.31 mg/g) and the Tangshan adzuki bean showed the lowest fat content (5.45 mg/g). There was a distinct difference in the fatty acid content of the 17 adzuki bean varieties. The palmitic acid, linoleic acid and linolenic acid were the dominant fatty acid in all beans and the contents are 24.03 (Xiaofeng 2) ∼29.41% (Tangshan), 30.11(Nihewan) ∼36.12% (Xiaofeng 2), and 23.52 (Tangshan) ∼27.76% (Bao 876–16), respectively. These three kinds of fatty acid account for 87.4% of the total fatty acid that determined. Kim, Nam, Kim, Hayes, and Lee (Citation2014) have reported that linolenic acid has the effects of cardiovascular-protective, anti-cancer, neuro-protective, anti-osteoporotic, anti-inflammatory and antioxidative activities, and it may be beneficial as a nutraceutical/pharmaceutical candidate and is safe for use as a food ingredient.
Table 2. The content of fat and fatty acid in adzuki beans.
The adzuki bean varieties contained 22.83 (Jingnong 8) ∼25.41% (Jinhong 352) protein (), which was consistent with the data of Tjahjadi et al. The value of adzuki bean proteins was noted in the introduction. Adzuki bean proteins have been demonstrated to have strong α-glucosidase inhibition activity, and this may explain the antidiabetic effect of water extract from adzuki beans (Yao et al., Citation2009). The relatively high content of protein in Jingnong 8 indicates its potential for development of antidiabetic food. Therefore, the following research on the α-glucosidase inhibition activity of water extracts from these 17 adzuki bean varieties was conducted. The amino acids composition is shown in . The amino acid composition is similar in these adzuki bean varieties and is characterized by a high content of glutamine acid, accounting for about 32.84 mg/g protein.
Table 3. The content of protein and amino acid profile in adzuki beans.
3.2. Total flavonoid, phenolic acid contents and individual phenolic acids
In this study, the Longxiaodou and Xiaofeng 2 varieties with the same average TFC content of 70.41 mg/g were found to possess the highest TFC among all the studied varieties. The Suhong 2 contained the lowest TFC (53.54 mg/g). TPC as measured by the Folin–Ciocalteu method varied widely in adzuki beans. Phenolic compounds are regarded as the major compounds that contribute to the total antioxidant activities of the grains (Yao et al., Citation2010). The highest TFC level was also found in the Longxiaodou 3 variety, with an average of 2.75 mg/g. The lowest was Suhong 2 (2.11 mg/g). This observation is in agreement with that of Yao, Cheng, Wang, Wang, and Ren (Citation2011) who observed that the content of the TPC was 2.7 mg/g.
In the present study, four bound phenolic acids (syringic acid, caffeic acid, p-coumaric acid and ferulic acid) and two free phenolic acids (caffeic acid and ferulic acid) were found in those beans. The contents of individual phenolic acid in the different bean varieties are shown in . It was found that caffeic acid was the dominant bound phenolic acid in all varieties and accounted for about 39.2% of the total of this class. Caffeic acid is an effective ABTS.+ and 1,1-diphenyl-2-picrylhydrazyl radical (DPPH) scavenging (Gülçin, Citation2006), so it may be a good antioxidant to human. Free phenolics can survive upper gastrointestinal digestion and be released from the colon through microflora digestion activity. The average concentration of free phenolic acids in adzuki bean varieties was 124.71 μg/g. The Baohong 947 has the highest free phenolic acid content in adzuki beans (202.5 μg/g). In summary, phenolic acid, TPC and TFC content of tested adzuki beans have significant differences, while those compositions are antioxidant ingredient. So we determinate the antioxidant activity of the adzuki beans.
Table 4. The content of total flavonoid (TFC), total phenolic (TPC), bound phenolic acid and free phenolic acid in adzuki beans.
3.3. Biological activities
ABTS.+ radical cation assays were used for the evaluation of free-radical-scavenging properties of the 17 varieties of adzuki bean. The results are presented in . The synthetic nitrogen-centered ABTS.+ extremist is not biologically relevant but is often used as an “indicator compound” in testing hydrogen-donation capacity and thus antioxidant activity (Re et al., Citation1999). Xiaofeng 2 showed the highest antioxidant activity (3.62 μM Trolox/g). Itoh et al. (Citation2009) investigated the antidiabetic effects of adzuki beans on streptozotocin (STZ)-induced diabetic rats, and they suggested that the active fraction of adzuki beans which suppressed the postprandial blood glucose was phenolic compounds. There were significant differences in these adzuki beans on the α-glucosidase inhibition activities. The Jihong 10 was the most active (27.43%), followed by the Baihong 2 (19.9%) and the last one is 11.2% (Longxiaodou 3).
Table 5. Biological activities of adzuki beans.
3.4. Correlation of antioxidant activity with TPC, TFC and individual phenolic acids
Correlation coefficients for TPC and TFC with ABTS+ assay are shown in . High correlation between the content of total phenolic compounds and their antioxidant capacity has been previously demonstrated by Zhou et al. (Citation2009). The results obtained in our study showed that TFC and TPC significantly correlate with the ABTS.+ assay (P < .01). Besides, a positive correlation (P < .01) of ABTS.+ assay with the free caffeic acid content was also observed.
Table 6. Correlation of antioxidant activity with TPC, TFC and individual phenolic acids.
4. Conclusion
In conclusion, there are significant differences in nutritional composition analyses and biological activities among the adzuki bean varieties investigated. All these adzuki beans have high content of resistant starch and have suitable amino acid constitution. Palmitic acid, linoleic acid and linolenic acid were the dominant fatty acids. Besides, four bound phenolic acid and two free phenolic acids were identified by HPLC. The antioxidant activity of 70% ethanol extracts and α-glucosidase inhibition activity of water extracts were evaluated. All the adzuki beans possessed strong ABTS.+ free-radical-scavenging capacity and α-glucosidase inhibition activity. Significant positive correlations (P < .01) of the antioxidant activity with total phenolic acids, total flavonoids and free caffeic acid contents were observed. These results are used as information when selecting bean varieties for daily diet or better design of potential functional food that can treat diseases such as cancer and associated cardiovascular diseases.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes on contributors
Zhengxing Shi, majors in food science, is a master of Engineering.
Yang Yao, majors in functional food, is a doctor of agriculture.
Yingying Zhu, majors in the quality and safety of agricultural products, is a doctor of agriculture.
Guixing Ren, majors in cereal chemistry, is a doctor of agriculture.
References
- Amarowicz, R., Estrella, I., Hernández, T., & Troszyska, A. (2008). Antioxidant activity of extract of adzuki bean and its fractions. Journal of Food Lipids, 15(1), 119–136. doi: 10.1111/j.1745-4522.2007.00106.x
- AOAC. (1990). Official methods of analysis (15th ed.). Arlington, VA: Association of Official Analytical Chemists.
- Gülçin, İ. (2006). Antioxidant activity of caffeic acid (3, 4-dihydroxycinnamic acid). Toxicology, 217(2), 213–220. doi: 10.1016/j.tox.2005.09.011
- Han, K., Iijuka, M., Shimada, K., Sekikawa, M., Kuramochi, K., Ohba, K., … Fukushima, M. (2005). Adzuki resistant starch lowered serum cholesterol and hepatic 3-hydroxy-3-methylglutaryl-coa mRNA levels and increased hepatic LDL-receptor and cholesterol 7alpha-hydroxylase mRNA levels in rats fed a cholesterol diet. British Journal of Nutrition, 94(6), 902–908. doi: 10.1079/BJN20051598
- Itoh, T., Kobayashi, M., Horio, F., & Furuichi, Y. (2009). Hypoglycemic effect of hot-water extract of adzuki (Vigna angularis) in spontaneously diabetic kk-a(y) mice. Nutrition, 25(2), 134–141. doi: 10.1016/j.nut.2008.08.001
- Kim, K. B., Nam, Y. A., Kim, H. S., Hayes, A. W., & Lee, B. M. (2014). α-Linolenic acid: Nutraceutical, pharmacological and toxicological evaluation. Food and Chemical Toxicology, 70, 163–178. doi: 10.1016/j.fct.2014.05.009
- López, A., El-Naggar, T., Dueñas, M., Ortega, T., Estrella, I., Hernández, T., & Carretero, M. E. (2013). Effect of cooking and germination on phenolic composition and biological properties of dark beans (Phaseolus vulgaris). Food Chemistry, 138, 547–555. doi: 10.1016/j.foodchem.2012.10.107
- Miao, X. F., Zhu, M. X., Wen-Ping, X. U., Shen, H. B., Sheng-Wei, D. U., & Pei, Y. F. (2010). Rapid determination on fatty acids content by gas chromatography in soybean. Soybean Science, 23(11), 1037–1044.
- Nielsen, T. S., Theil, P., Purup, S., Nørskov, N., & Bach Knudsen, K. E. (2015). Effects of resistant starch and arabinoxylan on parameters related to large intestinal and metabolic health in pigs fed fat rich diets. Journal of Agricultural & Food Chemistry, 63(48), 10418–10430. doi: 10.1021/acs.jafc.5b03372
- Nishioka, T., Kawabata, J., & Aoyama, Y. (1998). Baicalein, an α-glucosidase inhibitor from Scutellaria baicalensis. Journal of Natural Products, 61(11), 1413–1415. doi: 10.1021/np980163p
- Puls, W., Keup, U., Krause, H. P., Thomas, G., & Hoffmeister, F. (1977). Glucosidase inhibition. A new approach to the treatment of diabetes, obesity, and hyperlipoproteinaemia. Die Naturwissenschaften, 64(10), 536–537. doi: 10.1007/BF00483562
- Qin, P., Song, W., Yang, X., Sun, S., Zhou, X., & Yang, R. (2014). Regional distribution of protein and oil compositions of soybean cultivars in china. Crop Science, 54(3), 1139–1146. doi: 10.2135/cropsci2013.05.0314
- Re, R., Pellegrini, N., Proteggente, A., Pannala, A., Yang, M., & Rice-Evans, C. (1999). Antioxidant activity applying an improved ABTS radical cation decolorization assay. Free Radical Biology and Medicine, 26(9), 1231–1237. doi: 10.1016/S0891-5849(98)00315-3
- Siriwardhane, D., Egawa, Y., & Tomooka, N. (1992). Cross-compatibility of cultivated adzuki bean (Vigna angularis) and rice bean (V. umbellata) with their wild relatives. Plant Breeding, 107(4), 320–325. doi: 10.1111/j.1439-0523.1991.tb00555.x
- Slinkard, K., & Singleton, V. L. (1977). Total phenol analysis: Automation and comparison with manual methods. American Journal of Enology and Viticulture, 28(1), 49–55.
- Tjahjadi, C., & Breene, W. M. (1984). Isolation and characterization of adzuki bean (Vigna angularis cv Takara) starch. Journal of Food Science, 49(2), 558–562. doi: 10.1111/j.1365-2621.1984.tb12467.x
- Woisky, R. G., & Salatino, A. (1998). Analysis of propolis: Some parameters and procedures for chemical quality control. Journal of Apicultural Research, 37(2), 99–105. doi: 10.1080/00218839.1998.11100961
- Wojdyło, A., Oszmiański, J., & Czemerys, R. (2007). Antioxidant activity and phenolic compounds in 32 selected herbs. Food Chemistry, 105(3), 940–949. doi: 10.1016/j.foodchem.2007.04.038
- Yao, Y., Cheng, X., & Ren, G. (2014). α-glucosidase inhibitory activity of protein-rich extracts from extruded adzuki bean in diabetic KK-Ay mice. Food & Function, 5(5), 966–971. doi: 10.1039/c3fo60521c
- Yao, Y., Cheng, X., Wang, L., Wang, S., & Ren, G. (2011). Biological potential of sixteen legumes in China. International Journal of Molecular Sciences, 12(10), 7048–7058. doi: 10.3390/ijms12107048
- Yao, Y., Sang, W., Zhou, M., & Ren, G. (2009). Antioxidant and α-glucosidase inhibitory activity of colored grains in China. Journal of Agricultural and Food Chemistry, 58(2), 770–774. doi: 10.1021/jf903234c
- Yao, Y., Sang, W., Zhou, M., & Ren, G. (2010). Phenolic composition and antioxidant activities of 11 celery cultivars. Journal of Food Science, 75(1), C9–C13. doi: 10.1111/j.1750-3841.2009.01392.x
- Yao, Y., Xue, P., Zhu, Y., Gao, Y., & Ren, G. (2015). Antioxidant and immunoregulatory activity of polysaccharides from adzuki beans (Vigna angularis). Food Research International, 50(3), 1092–1095.
- Zhang, Y., Wang, L., Yao, Y., Yan, J., & He, Z. (2012). Phenolic acid profiles of Chinese wheat cultivars. Journal of Cereal Science, 56(3), 629–635. doi: 10.1016/j.jcs.2012.07.006
- Zhou, S. H., Fang, Z. X., Lü, Y., Chen, J. C., Liu, D. H., & Ye, X. Q. (2009). Phenolics and antioxidant properties of bayberry (Myrica rubra Sieb. et Zucc.) pomace. Food Chemistry, 112(2), 394–399. doi: 10.1016/j.foodchem.2008.05.104