Antioxidant activity in different parts of roselle ( Hibiscus sabdariffa L.) extracts and potential exploitation of the seeds

Food Chemistry 122 (2010) 1055–1060

Contents lists available at ScienceDirect

Food Chemistry

journal homepage: www.elsevier .com/locate / foodchem

Antioxidant activity in different parts of roselle (Hibiscus sabdariffa L.) extractsand potential exploitation of the seeds

Norhaizan Mohd-Esa *, Fong Shin Hern, Amin Ismail, Chew Lye YeeDepartment of Nutrition and Dietetics, Faculty of Medicine and Health Sciences, University Putra Malaysia, 43400 Serdang, Selangor, Malaysia

a r t i c l e i n f o a b s t r a c t

Article history:Received 2 June 2009Received in revised form 11 February 2010Accepted 18 March 2010

Keywords:RoselleHibiscus sabdariffa L.AntioxidantDPPHb-carotenePhenolicLipid peroxidation

0308-8146/$ - see front matter � 2010 Elsevier Ltd. Adoi:10.1016/j.foodchem.2010.03.074

* Corresponding author. Tel.: +60 3 89472427; fax:E-mail address: [email protected] (N. M

The calyx of the roselle plant has long been recognised as a source of antioxidants. The objective of thisstudy was to evaluate antioxidant activity, free radical-scavenging and total phenolic content in otherparts of the roselle plant. Roselle seed extracts were found to have the highest antioxidant activity andstrongest radical-scavenging activity of all plants tested. Methanol extracts showed a positive correlationbetween phenolic content and antioxidant activity, as measured by b-carotene bleaching assay and DPPHradical-scavenging activity. The antioxidant efficacy of roselle seeds in a whole food system was investi-gated by testing the effect on lipid oxidation in cooked beef patties stored at 4 �C for 14 days. Resultsshowed that patties treated with roselle seeds had reduced lipid oxidation compared to patties treatedwith BHT. This study suggests that roselle seeds have the potential to be used as food antioxidants.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction having the above mentioned activities, roselle extract can also per-

Hibiscus sabdariffa L. or roselle is also known locally as asamsusur, asam paya or Ribena Malaysia, and closely resembles cran-berries in flavour. It is also called karkade in Switzerland and inArab-speaking countries. Roselle is used in jams, jellies, saucesand wines. The young leaves and tender stem are eaten raw in sal-ads and chutney. They are also added to curries and some Malay-sian dishes as seasoning. The seeds are somewhat bitter but, inAfrica, they are ground into meal for human food due to their highprotein content. They have also been roasted to use as a substitutefor coffee (Morton, 1987). In northern Nigeria, roselle seeds are fer-mented in the presence of some spices to prepare a food known asMungza Ntusa (Balami, 1998). Its seeds also contain a substantialamount of oil that resembles cotton seed oil (Mohammed, Fernan-dez, Pineda, & Aguilar, 2007).

Many medicinal applications of this plant have been developedaround the world. For example, in China it is used to treat hyperten-sion, pyrexia, liver damage and leukaemia due to its high content ofprotocatechuic acid (Tseng et al., 2000). Studies by Muhammad andShakib (1995) have shown that roselle can prevent cancer, lowerblood pressure and improve the digestive system in humans. Its ca-lyx extract has also been used as an effective treatment for patientswith kidney stones due to its uricosuric effect (Prasongwatana,Woottisin, Sriboonlue, & Kukongviriyapan, 2008). In addition to

ll rights reserved.

+60 3 89426769.ohd-Esa).

form as an antioxidant. For example, it protects against low densitylipoprotein (LDL)-oxidation and has hypolipidemic effects in vivo(Hirunpanich et al., 2006). In some instances, it is also used to pre-serve food. Up until now, most of the scientific study has only fo-cused on the calyces of roselle. Very few data on antioxidantactivity of extracts obtained from other parts of the roselle plantare available. Thus, the aim of this work was to evaluate the antiox-idant activity of extracts from calyxes, leaves, seeds and stems of ro-selle plants, and compare these to a commercial syntheticantioxidant, butylated hydroxytoulene (BHT). In addition, the anti-oxidant activity of roselle seeds was also tested in a whole food sys-tem by determining effects of roselle on lipid oxidation in cookedbeef stored at 4 �C for up to 14 days.

2. Materials and methods

2.1. H. sabdariffa L.

Roselle (H. sabdariffa L.) calyces, stems, leaves and seeds werecollected from the Department of Agriculture, Rhu Tapai, KualaTerengganu, Malaysia.

2.2. Chemicals

b-Carotene, linoleic acid, Tween-20, butylated hydroxytoluene(BHT), 2,2-diphenyl-2-picrylhydrazyl (DPPH�), a-tocopherol,

1056 N. Mohd-Esa et al. / Food Chemistry 122 (2010) 1055–1060

thiobarbituric acid (TBA) and gallic acid were purchased from Sig-ma Chemical Co. (St. Louis, MO, USA). Folin–Ciocalteu reagent andmethanol were obtained from Merck (Darmstadt, Germany), so-dium bicarbonate and chloroform were purchased from BDHChemicals (Poole, England).

2.3. Preparation of extracts

The calyces, stems and leaves were separated and cut into smallpieces. They were then lyophilized using a freeze-dryer (Virtis Co.,Inc., New York, USA) at �50 �C, at 35 mm Hg, and stored in an air-tight container at �20 �C prior to further use. The extracts wereprepared according to the method of Guo, Lee, Chiang, Lin, andChang (2001) with modifications. H. sabdariffa L. was extractedwith distilled water or 80% (v/v) methanol for 2 h at room temper-ature, using an orbital shaker (Unimax 1010, Heidolph, Germany).The ratio of sample to extraction medium was 1:1000. The mixturewas filtered through a filter paper (Whatman no. 4). The filtratewas considered to be roselle extract and was used for the antioxi-dant assays.

2.4. Measurement of antioxidant activity

The antioxidant activities of different parts of the roselle plantwere evaluated, using the b-carotene bleaching method describedby Velioglu, Mazza, Gao, and Oomah (1998) with modifications.One millilitre of a 0.2 mg/ml b-carotene solution in chloroformwas added to flasks containing 0.02 ml of linoleic acid and0.2 ml of Tween-20. The chloroform was removed at 40 �C usinga rotary evaporator (Laborota 4000, Heidolph, Japan) for 5–10 min. The resultant mixture was immediately diluted with100 ml of distilled water and mixed for 1–2 min to form an emul-sion. A mixture, prepared similarly but without b-carotene, wasused as a blank. A control containing 0.2 ml of 80% (v/v) methanolinstead of extract was also prepared. A 5 ml aliquot of the emul-sion was added to a tube containing 0.2 ml of the sample extractat 1 mg/ml. BHT (1 mg/ml), in methanolic solution, was used as astandard. The tubes were placed in a water bath (Sb-16, Techne,England) at 40 �C for 2 h. Absorbance was read at 470 nm at15 min intervals, using a UV–visible spectrophotometer (UV-1601; Shimadzu Corp., Kyoto, Japan). The antioxidant activity ofeach sample was calculated as percent inhibition relative to thecontrol, using the following equation (Jayaprakasha, Singh, &Sakariah, 2001):

Antioxidant activity ð%Þ ¼ 1� ðA0 � AtÞA�

0 � A�

t

" #� 100;

where,A 0 and A

�

0 are the absorbance values measured at zero time ofincubation for sample extract and control, respectively.

At, and A�

t are the absorbance values for sample extract and con-trol, respectively, at t = 120 min.

2.5. Measurement of DPPH radical-scavenging activity

Free radical-scavenging was evaluated using the 1,1-diphenyl-2-picrylhydrazyl (DPPH) free radical according to the method ofShimada, Fujikawa, Yahara, and Nakamura (1992). One millilitreof freshly prepared 1 mM DPPH in methanol was added to testtubes containing 5 ml of the sample extracts. A control was pre-pared by adding 1 ml of DPPH� solution to 5 ml of 80% methanol.Following storage in the dark for 30 min, the absorbancewas read at 517 nm using a UV–visible spectrophotometer(UV-1601; Shimadzu Corp., Kyoto, Japan). The percentage of free

radical-scavenging activity was calculated, based on the follow-ing equation.

Free radical-scavenging activity ð%Þ

¼ 1� Absorbance of sample at 517 nmAbsorbance of control at 517 nm

� �� 100:

2.6. Measurement of total phenolic content

Total phenolic content was determined with Folin–Ciocalteu re-agent by the method of Kahkonen et al. (1999). Briefly, 0.2 ml ofeach sample extract (1 mg/ml) was mixed with 1 ml of a 10-folddilution of Folin–Ciocalteu reagent and 0.8 ml of 7.5% (w/v) bicar-bonate solution, and allowed to stand at room temperature for30 min. The absorbance was measured at 765 nm, using a UV–vis-ible spectrophotometer (UV-1601; Shimadzu Corp., Kyoto, Japan).The total phenolic content was expressed as gallic acid equivalents(GAE) in milligrammes per gramme of dry material.

2.7. Preparation of ground beef

Beef meat was prepared according to the method of Tang, Kerry,Sheehan, Buckley, and Morrissey (2001) with modifications. Com-mercial Australian beef (eyeround type) was purchased from a lo-cal market and was cut into small pieces and homogenised in ablender (Braun ZK 100, Germany). Ground beef (40 g) was thor-oughly mixed by hand with 7.5 g/kg of roselle seed. BHT and a-tocopherol (7.5 g/kg) were used for comparative purposes. Mincedmeat, containing no additive, was run as a control. Beef sampleswere pressed into a mould and cooked for 30 min in a water bath(Sb-16, Techne, England) at 75 ± 1 �C until an internal temperatureof �75 �C was reached. Following this step, cooked beef wascooled, covered with oxygen-permeable polyvinyl chloride filmand stored at 4 �C prior to further analysis.

2.8. Measurement of thiobarbituric acid (TBA) number

The extent of lipid oxidation in cooked beef was monitored,using the TBA assay, on days 1, 3, 7 and 14 according to the methodof Kirk and Sawyer (1991) with modifications. Cooked beef (10 g)was chopped into small pieces and homogenized with 100 ml ofdistilled water for 2 min. A few drops of 4 M HCl were slowlyadded to bring the mixture to pH 1.5. The mixture then was heatedand the distillate (5 ml) was collected in a conical flask. A blankwas prepared similarly, using 100 ml of distilled water withoutadded sample. Five millilitre of distillate were transferred into aglass-stoppered tube and 5 ml of TBA reagent (0.02 M) were added.The mixture was incubated in boiling water for 35 min. After beingcooled in water for 10 min, the absorbance was measured againstthe blank at 538 nm, using a UV–visible spectrophotometer (UV-1601; Shimadzu Corp., Kyoto, Japan). Lipid oxidation was ex-pressed as 2-thiobarbituric acid-reactive substances (TBAR) in mgmalondialdehyde (MDA)/kg beef.

2.9. Statistical analysis

Data were analysed using Statistical Package for Social Science(SPSS) for Windows version 11.0. Significant differences in antiox-idant activity and total phenolic content of different parts of the ro-selle plant and different extracts were determined using one-wayANOVA. The same test was used to compare the means of TBAnumbers in cooked beef patties. The statistical probability wasconsidered to be significantly different at the level of P < 0.05.The relationship between total antioxidant activity and totalphenolic content was analysed by Pearson’s correlation.

N. Mohd-Esa et al. / Food Chemistry 122 (2010) 1055–1060 1057

3. Results and discussion

3.1. Phenolic content

Results obtained in the present study revealed that the level oftotal phenolic compounds in the extracts of roselle was consider-able (Table 1). The seeds had higher total phenolic content thanhad other parts of the roselle plant in both methanol and water ex-tracts. The leaves had a higher concentration of phenolic com-pounds than had the stems. Previous studies have shown thatthe developmental stage of the plant may affect biosynthetic path-ways of phenolic compounds; this could then affect the total phe-nolic and flavonoid contents (Krizman, Baricevic, & Prosek, 2007).It was also shown that the biosynthesis of polyphenol is acceler-ated by light exposure and serves as a filtration mechanism againstUV-B radiation (Harborne & Williams, 2000). Similar results havebeen demonstrated in potatoes. Total phenolic concentration washighest in flowers, followed by leaves and stem. A likely explana-tion is that large amounts are needed to protect the flowering plantagainst attack by phytopathogens (Im et al., 2008).

This study showed that selected parts of the roselle plant, ex-tracted by different solvents, had different total phenolic contents.The mean total phenolic content of seeds, calyces, leaves and stemsof the roselle plant was shown to decrease in the order of metha-nolic >water extracts. This indicated that phenolic compounds ofroselle plants are better extracted with methanol than with water.

3.2. Antioxidant activity

The comparative b-carotene bleaching rates of the control, BHT(standard) and methanolic and water extracts of different parts ofthe roselle plant are shown in Figs. 1 and 2. The decrease in absor-

Table 1Means of total phenolic content of selected samples.

Extracts Roselle plant Total phenolic contentA (mg of GAE/g)

Distilled water Seed 2.97 ± 0.17aCalyx 1.85 ± 0.11bLeaf 1.71 ± 0.04bStem 0.90 ± 0.03c

80% (v/v) Methanol Seed 4.87 ± 0.14aCalyx 2.91 ± 0.07bLeaf 2.20 ± 0.02cStem 1.31 ± 0.27d

A Values are presented as means standard deviation (n = 3)); different letters aresignificantly different at P < 0.05.

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0 20 40 60 8Time (min)

Abs

orba

nce

(470

nm)

Fig. 1. Comparison of absorbance values of water extracts of samples at 1 mg/ml u

bance was due to the oxidation of linoleic acid and b-carotene. Itshows a slower decrease in absorbance of b-carotene in thepresence of methanol extract than in the presence of water extract.

Table 2 shows the mean antioxidant activity, based on the b-carotene bleaching rate of methanol and water extracts of differentparts of roselle (seeds, leaves, calyxes and stems). The highest anti-oxidant activity, among different parts of roselle extracted withwater, was observed in calyces whereas the lowest was observedin stems. These results are consistent with the fact that calycesare rich in vitamin C (141 mg/100 g), anthocyanins (2.52 mg/100 g), b-carotene (1.88 mg/100 g), lycopene (164 lg/100 g)(Wong, Yusof, Ghazali, & Che Man, 2002), polyphenols and otherwater-soluble antioxidants (Duke & Atchley, 1984).

The highest antioxidant activity in the methanol extract wasobserved in roselle seeds, whereas roselle stems had the lowestantioxidant capacity of all the sample extracts. The high total anti-oxidant activity of seeds is possibly due to the presence of phytos-terols and tocopherols, particularly b-sistosterol and c-tocopherol,as shown by Mohammed et al. (2007).

Results also showed that the antioxidant activities of selectedparts of roselle plants in water and methanol were not consistent.Water extracts of roselle plant parts and activity of the standardexhibited the following order of strength of antioxidant activity:BHT > calyces > seeds > leaves > stems. Some of these samplesmay contain water-soluble polyphenolic compounds, whereas oth-ers do not. Methanol extracts and standard showed a different or-der of antioxidant activity: seeds > calyces > BHT > leaves > stems.From this sequence, it can be postulated that seeds and calycescontain more antioxidant than do leaves and stems of the roselleplant.

As seen in this study, the selected parts of the roselle plant ex-tracted by different solvents exhibited varying degrees of antioxi-dant activity. The methanol extracts from the four parts of theroselle plant exhibited higher total antioxidant activity than didthe water extracts. The difference is probably due to the character-istic of the solvent: this could affect which compounds are ex-tracted from the plant matrix. This phenomenon can beexplained by a change in polarity of the antioxidant compounddue to the particular solvent used for extraction. As the polarityof the solvent increases, higher extraction yields of total solublesolids and total extractable polyphenolic compounds were ob-tained. So, methanol extracts offer a more realistic determinationof antioxidant activity than do distilled water extracts of rosellein this study. Other factors that might strongly influence the anti-oxidant activities are the antioxidant concentration, the tempera-ture and pH of the media processing treatment and storage(Gazzani, Papetti, Massolini, & Daglia, 1998). In our study, how-ever, variation in media pH and temperature were not taken into

0 100 120 140

control

BHT

seed

calyx

leaf

stem

sing b-carotene bleaching method. Results are means of three determinations.

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0 20 40 60 80 100 120 140Time (min)

Abs

orba

nce

(470

nm)

control

BHT

seed

calyx

leaf

stem

Fig. 2. Comparison of absorbance values of methanol extracts of samples at 1 mg/ml using b-carotene bleaching method. Results are means of three determinations.

Table 2Means of total antioxidant activity of selected samples and standard, assayed by b-carotene linoleate bleaching.

Extracts Roselle plant Total antioxidant activity* (%)

Distilled water Seed 45.9 ± 2.32aCalyx 54.1 ± 5.80aLeaf 27.9 ± 0.00bStem 10.7 ± 3.50c

80% (v/v) Methanol Seed 78.7 ± 2.32aCalyx 61.5 ± 5.80bLeaf 54.9 ± 8.11bStem 22.1 ± 1.16cBHT 57.4 ± 11.60b

* Antioxidant activity = % inhibition relative to control. Values are presented asmeans ± standard deviation (n = 3); different letters are significantly different atP < 0.05.

Table 3Means of DPPH radical-scavenging activity of selected samples and standard.

Extracts Roselle plant DPPH radical-scavenging activity (%)A

Distilled water Seed 65.1 ± 2.58aCalyx 30.8 ± 1.43bLeaf 46.5 ± 0.96cStem 43.3 ± 4.69c

80% (v/v) Methanol Seed 91.8 ± 1.05aCalyx 87.9 ± 0.76bLeaf 89.8 ± 0.33bStem 81.8 ± 9.02a,bBHT 24.6 ± 6.75c

A Values are presented as means ± standard deviation (n = 3); different letters aresignificantly different at P < 0.05.

1058 N. Mohd-Esa et al. / Food Chemistry 122 (2010) 1055–1060

account. So, it is not known whether these factors could affect theresult.

Pearson correlation analysis showed that there was a positivecorrelation between phenolic content and antioxidant activity, asmeasured by b-carotene bleaching assay for both water(r = 0.733) and methanol (r = 0.904) extracts. Another possible rea-son why roselle seeds exhibited strong antioxidant activity is be-cause of their high phenolic content.

ANOVA showed a significant difference (P < 0.05) in antioxidantactivity between methanol and water extracts of roselle seeds,leaves and stems. However, there was no significant difference be-tween methanol and water extracts of calyces. Methanol extractsof seeds also exhibited significantly (P < 0.05) higher antioxidantactivity than did BHT.

3.3. DPPH radical-scavenging activity

The antioxidant activity of a sample is strongly dependent onthe model system in which it is evaluated. A single analytical assay(b-carotene bleaching method) may be inadequate for measuringantioxidant activity. For this reason, the selected parts of roselleplants were also examined for their radical-scavenging capacitiesusing the stable radical, DPPH�, which has been widely used to testthe free radical-scavenging ability of various chemicals. DPPH rad-ical-scavenging activities of methanol and water extracts (1 mg/ml) were tested. As shown in Table 3, DPPH radical-scavengingactivities of the extracts did not show the same patterns. The waterextracts of selected parts of the roselle plant and standard de-creased in the order: seeds > leaves > stems > calyces > BHT.However, the methanol extracts of the selected parts of the roselleplant and standard decreased in the order: seeds > leaves >calyces > stems > BHT.

The mean DPPH radical-scavenging activity values of roselleseeds, calyces, leaves and stems were significantly higher(P < 0.05) in methanol extracts than in water extracts of the samesample. These results are in agreement with the results from b-car-otene bleaching assay in that the methanol extract of roselle seedshad a higher DPPH�-scavenging activity than had other part of theroselle plant or BHT.

Antioxidant activity in stems and leaves, as measured by DPPH�,is also significantly higher than that of BHT. According to Yen, Duh,and Tsai (2002), the presence of polyphenolic compounds, such asantroquinones, xanthones, proanthocyanidins and flavonols, couldaccount for the reasonably strong antioxidant activity in the ex-tracts of stems and leaves. Yen and Chen (1995) reported that poly-phenols were the major compounds in tea leaves and seem to beresponsible for its antioxidant activity. So it is suggested, in furtherstudy, to observe the UV–vis spectra of various extracts of roselleplant in order to identify possible antioxidant compounds.

For antioxidant activity, as measured by DPPH radical-scaveng-ing activity, a positive correlation with phenolic content was sig-nificant only for the methanol extracts (r = 0.636) but not for thewater extracts (r = 0.552).

3.4. TBA (thiobarbituric acid) numbers of cooked beef patties

Synthetic antioxidants, such as BHT and butylated hydroxyani-sole (BHA), have been shown to be very effective in the reductionof lipid oxidation in foods (Shahidi, Rubin, & Wood, 1987). Namiki(1990), however, revealed that synthetic antioxidants in food couldpromote of carcinogenesis, and that natural antioxidants, such astocopherol and ascorbic acid derivatives, may be used as BHT andBHA substitutes. The production of tocopherol and ascorbic acidderivatives, however, involve higher manufacturing costs (Moure

0

2

4

6

8

10

12

14

16

18

Day 1 Day 3 Day 7 Day 14

Storage time (days)

TBA

num

ber (

mg

mal

onal

dehy

de/ k

g sa

mpl

e) Control

Roselle seedextractBHT

Alpha-tocopherol

Fig. 3. Effect of roselle seed extracts, as compared to BHT and a-tocopherol on the TBA numbers of cooked beef patties stored at 4 �C for 14 days (CV < 1.65%).

N. Mohd-Esa et al. / Food Chemistry 122 (2010) 1055–1060 1059

et al., 2001) and these antioxidants are not as effective as the syn-thetic antioxidant. As a consequence, several studies have been doneon antioxidants from natural plant sources, including ground tea leafand tea extract (He & Shahidi, 1997), that have substantial antioxi-dant activity. Therefore, since the roselle seed has a higher phenoliccontent and higher antioxidant activity than have BHT and otherparts of the roselle plant, we would like to explore its potential asa food preservative.

Fig. 3 shows the effects of roselle seeds extract, as comparedto control, BHT and a-tocopherol, on TBA numbers of cookedbeef patties stored at 4 �C for 14 days. It shows that the additionof extracts and commercial antioxidants, at a concentrationequivalent to 750 mg in 100 g of beef, resulted in a significantreduction in the TBA numbers of cooked beef patties as com-pared to the control sample. The control sample, which was onlycooked beef without any treatment, consistently showed thegreatest TBA numbers. The results clearly illustrated that theaddition of roselle seed extracts, at a concentration equivalentto 750 mg/100 g of beef, resulted in a significant reduction inthe TBA numbers of cooked beef patties compared to the controlsample. It appears that the seed extracts used in this study pos-sess antioxidant properties that can inhibit lipid oxidation incooked beef patties stored at 4 �C for 14 days. Results also indi-cated that the roselle seeds possess antioxidative bioactive com-pounds that are more powerful than are commercial antioxidants(BHT and a-tocopherol).

4. Conclusions

In Malaysia, the seeds of the roselle plant are discarded as by-product. About 50% of raw material is made up of seed capsules.This study shows that seeds normally discarded as by-productscan be used as a source of antioxidant; this indirectly increasesthe value of the seeds. Additionally, a study by Halimatul, Amin,Mohd.-Esa, Nawalyah, and Muskinah (2007) showed that boiledroselle seeds contain a good quality protein, similar to casein.The seeds can also reduce total cholesterol and low densitylipoprotein (LDL) when given to hypercholesterolemic rats (Emmy,Amin, Normah, Mohd.-Esa, & Ainul, 2008). Thus it can beconcluded that other parts of the roselle plant, such as seeds andleaves, are also useful for human consumption.

Acknowledgement

This research was made possible by grants and assistance fromthe Universiti Putra Malaysia and Jabatan Pertanian, Rhu Tapai,Terengganu, providing raw material.

References

Balami, A. (1998). The effect of processing conditions, packaging and storage on selectedquality attributes of Mungza Ntusa (M.Sc. thesis). Nigeria: University of Ibadan.

Duke, J. A., & Atchley, A. A. (1984). Proximate analysis. In B. R. Christie (Ed.), Thehandbook of plant science in agriculture. Boca Raton, Fla: CRC Press Inc.

Emmy, H., Amin, I., Normah, H., Mohd.-Esa, N., & Ainul, Z. A. B. (2008). Effects ofdefatted dried roselle (Hibiscus sabdariffa L.) seed powder on lipid profiles ofhypercholesterolemia rats. Journal of Science and Food Agriculture, 88,1043–1050.

Gazzani, G., Papetti, A., Massolini, G., & Daglia, M. (1998). Anti- and pro-oxidantactivity of water soluble components of some common diet vegetables and theeffect of thermal treatment. Journal of Agricultural and Food Chemistry, 46,4118–4122.

Guo, J. T., Lee, H. L., Chiang, H. L., Lin, F. I., & Chang, C. Y. (2001). Antioxidantproperties of the extracts from different parts of broccoli in Taiwan. Journal ofFood and Drug Analysis, 9, 96–101.

Halimatul, S. M. N., Amin, I., Mohd.-Esa, N., Nawalyah, A. G., & Muskinah, M. (2007).Protein quality of roselle (Hibiscus sabdariffa L.) seeds. ASEAN Food Journal, 14,131–140.

Harborne, J. B., & Williams, C. A. (2000). Advances in flavonoid research since 1992.Phytochemistry, 55, 481–504.

He, Y., & Shahidi, F. (1997). Antioxidant activity of green tea and its cathechins in afish meat model system. Journal of Agricultural and Food Chemistry, 45,4262–4266.

Hirunpanich, V., Utaipat, A., Molales, N. P., Bunyapraphtsala, N., Sato, H., Herunsale,A., et al. (2006). Hypocholesterolemic and antioxidant effects of aqueousextracts from the dried calyx of Hibiscus sabdariffa L. in hypercholesterolemicrats. Journal of Ethnopharmacology, 103, 252–260.

Im, H. W., Suh, B. S., Lee, S. U., Kozukue, N., Mayumi, O. K., Carol, E.L., & Friedman, M.(2008). Analysis of phenolic compounds by high-performance liquidchromatography and liquid chromatography/mass spectrometry in potatoplant flowers, leaves, stems and tubers and in home-processed potatoes.Journal of Agriculture and Food Chemistry, 56, 3341–3349.

Jayaprakasha, G. K., Singh, R. P., & Sakariah, K. K. (2001). Antioxidant activity ofgrape seed (Vitis vinifera) extracts on peroxidation models in vitro. FoodChemistry, 73, 285–290.

Kahkonen, M. P., Hopia, A. I., Vuorela, H. J., Rauha, J. P., Pihlaja, K., Kujala, T. S., et al.(1999). Antioxidant activity of plant extracts containing phenolic compounds.Journal of Agricultural and Food Chemistry, 47, 3954–3962.

Kirk, R. S., & Sawyer, R. (1991). Pearson’s composition and analysis of foods (9th ed.).England: Addison Wesley Longman Ltd.. pp. 642–643.

Krizman, M., Baricevic, D., & Prosek, M. (2007). Determination of phenoliccompounds in fennel by HPLC and HPLC–MS using a monolithic reversed-phase column. Journal of Pharmacology and Biology Analysis, 43, 481–485.

1060 N. Mohd-Esa et al. / Food Chemistry 122 (2010) 1055–1060

Mohammed, R., Fernandez, J., Pineda, M., & Aguilar, M. (2007). Roselle (Hibiscussabdariffa) seed oil is a rich source of (-tocopherol. Journal of Food Science, 72,S207–S211.

Morton, J. F. (1987). Roselle. In Fruits of warm climates (pp. 281–286). Miami, USA:Florida Flair Books.

Moure, A., Cruz, J. M., Franco, D., Dominguez, J. M., Sineiro, J., Dominguez, H., et al.(2001). Review: Natural antioxidants from residual source. Food Chemistry, 72,145–171.

Muhammad, T. B., & Shakib, A. B. (1995). Jus hibiscus: Bukan sekadar minumanbiasa. Dewan Ekonomi, 12–14.

Namiki, M. (1990). Antioxidants/antimutagens in food. Critical Review of FoodScience and Nutrition, 29, 273–300.

Prasongwatana, V., Woottisin, S., Sriboonlue, P., & Kukongviriyapan, V. (2008).Uricosuric effect of roselle (Hibiscus sabdariffa) in normal and renal-stoneformer subject. Journal of Ethnopharmacology, 117(3), 491–495.

Shahidi, F., Rubin, L. J., & Wood, D. F. (1987). Control of lipid peroxidation in cookedground pork with antioxidants and dinitrosyl ferrohemochrome. Journal of FoodScience, 52, 564–567.

Shimada, K., Fujikawa, K., Yahara, K., & Nakamura, T. (1992). Antioxidativeproperties of xanthan on the autoxidation of soybean oil in cyclodextrinemulsion. Journal of Agricultural and Food Chemistry, 40, 945–948.

Tang, S., Kerry, J. P., Sheehan, D., Buckley, D. J., & Morrissey, P. A. (2001).Antioxidative effect of added tea catechins on susceptibility of cooked redmeat, poultry and fish patties to lipid oxidation. Food Research Institute, 34,651–657.

Tseng, T., Kao, T., Chu, C., Chou, F., Lin, W., & Wang, C. (2000). Induction of apoptosisby hibiscus protocatechuic acid in human leukaemia cells via reduction ofretinoblastoma (RB) phosphorylation and Bcl-2 expression. BiochemicalPharmacology, 60, 307–315.

Velioglu, Y. S., Mazza, G., Gao, L., & Oomah, B. D. (1998). Antioxidant activity andtotal phenolics in selected fruits, vegetables, and grain products. Journal ofAgricultural and Food Chemistry, 46, 4113–4117.

Wong, P. K., Yusof, S., Ghazali, H. M., & Che Man, Y. B. (2002). Physico-chemicalcharacteristics of roselle (Hibiscus sabdariffa L.). Nutrition and Food Science, 32,68–73.

Yen, G. C., & Chen, H. Y. (1995). Antioxidant activity of various tea extracts inrelation to their antimutagenicity. Journal of Agriculture and Food Chemistry, 43,27–32.

Yen, G. C., Duh, P. D., & Tsai, H. L. (2002). Antioxidant and pro-oxidant properties ofascorbic acid and gallic acid. Food Chemistry, 79, 307–311.