Antioxidant activity and phenolic composition of sumac (Rhus coriaria L.) extracts

Food Chemistry 113 (2009) 568–574

Contents lists available at ScienceDirect

Food Chemistry

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

Antioxidant activity and phenolic composition of herbhoneys

Robert Socha, Lesław Juszczak *, Sławomir Pietrzyk, Teresa FortunaDepartment of Analysis and Evaluation of Food Quality, University of Agriculture, Balicka 122 Street, 30-149 Krakow, Poland

a r t i c l e i n f o

Article history:Received 21 March 2008Received in revised form 4 June 2008Accepted 11 August 2008

Keywords:HerbhoneysAntioxidant activityPhenolic contentFlavonoid composition

0308-8146/$ - see front matter � 2008 Elsevier Ltd. Adoi:10.1016/j.foodchem.2008.08.029

* Corresponding author. Fax: +48 12 6624746.E-mail address: [email protected] (L. Juszczak

a b s t r a c t

A study of 10 herbhoneys of various origin revealed differences in their antioxidant activity and profilesof phenolic acids and flavonoids. The total phenolic content of herbhoneys determined spectrophotomet-rically varied between 21.7 and 75.3 mg gallic acid equivalents per 100 g product, and the total flavonoidcontent ranged from 6.9 to 28.5 mg quercetin equivalents per 100 g. Dark herbhoneys, such as raspberry,thyme, hawthorn and black chokeberry, exhibited a high antioxidant activity and contained high totallevels of polyphenols and flavonoids. There was a significant linear correlation between total phenolicand flavonoid contents and antioxidant activity in the reactions with DPPH� and ABTS�+ free radicals.The profiles of phenolic acids and flavonoids determined by HPLC depended on the variety of herbhoney.Among the products studied, raspberry and thyme herbhoneys were the richest in phenolic acids andflavonoids. The dominant phenolic acid in most samples was p-coumaric acid; the herbhoneys containedalso a considerable amount of gentisic acid. The dominant flavonoids were hesperetin and naringenin.Thyme herbhoney had an especially high quercetin content.

� 2008 Elsevier Ltd. All rights reserved.

1. Introduction The characteristic polyphenols present in honeys that are able

Herbhoney is a honey-like product obtained from bees fed on asaccharose-based food containing herbal extracts or fruit juices.The great variety of potential sources of such plant-derived addi-tives makes it possible to give specific sensory characteristics (col-our, aroma, taste) to herbhoneys, and thus considerably expand thecommercial range of bee products. Herbhoneys have the medicinalproperties of herbs or other raw materials contained in the food fedto bees, which makes them usable in pharmacy and medicine, bothas drug components, prophylactic agents and diet supplements.

The basic chemical composition of herbhoneys is similar to thatof natural honeys and most of them meet the quality criteria con-cerning acidity, water content, hydroxymethylfurfural content anddiastase number, applied to the latter (Juszczak, Socha, Ro _znowski,Fortuna, & Nalepka, 2008). Due to a different method of produc-tion, and especially a high availability of such biologically activesubstances as antioxidants in the food fed to bees, herbhoneysmay somewhat differ from natural honeys in the profile of phenolicacids and flavonoids. Being natural antioxidants, these compoundsplay a major role in human nutrition. Phenolic acids and flavonoidsmay also be used as biomarkers for the origin of a honey (Tomás-Barberán, Martos, Ferreras, Radovic, & Anklam, 2001). It should betaken into account, however, that the levels of individual phenoliccompounds are also closely connected with the climatic conditionsof an area (Kenjeric, Manic, Primorac, Bubalo, & Perl, 2007; Cek-steryte, Kazlauskas, & Racys, 2006).

ll rights reserved.

).

to perform the role of biomarkers are flavonoids such as hespere-tin, kaempferol, quercetin and chrysin, and phenolic acids: abscisic,ellagic, p-coumaric, gallic and ferulic (Ferreres, Andrade, & Tomás-Barberán, 1994; Kenjeric et al., 2007; Soler, Gil, Garcia-Viguera, &Tomás-Barberán, 1995; Tomás-Barberán et al., 2001; Yao et al.,2003; Yaoa, Jiang, Singanusong, Datta, & Raymont, 2005). The pres-ent study focuses on herbhoneys of various origins with the aim ofinvestigating their antioxidant properties and phenolic acid andflavonoid profiles.

2. Materials and methods

2.1. Materials

The study used 10 herbhoneys: aloe, black chokeberry, camo-mile, hawthorn, marigold, mint, nettle, pine, raspberry and thyme,produced in 2006 and supplied by P.P. Apipol (Krakow, Poland).The raspberry herbhoney sample came from an experimental api-ary, the rest were commercial samples. All samples were stored ina refrigerator at 4� C. Immediately before analyses they wereheated at a temperature below 40� C to dissolve crystals.

2.2. Determination of antioxidant activity in the reaction with DPPH�

(1,1-diphenyl-2-picrylhydrazyl) radical

The antioxidant activity of herbhoneys was determined using aprocedure described by Turkmen, Sari, and Velioglu (2006). One-gram samples of herbhoneys were dissolved in 5 ml of distilled

R. Socha et al. / Food Chemistry 113 (2009) 568–574 569

water, centrifuged at 4350g, and filtered through paper filters.Then, 0.5 ml of the solutions were mixed with 1.5 ml of 0.1 mMmethanol solution of DPPH� (Sigma–Aldrich, Steinheim, Germany).The control test was made with distilled water in place of herbhon-ey solution. The reaction mixtures were vortex-mixed and left inthe dark at room temperature for incubation during 60 min.Absorbance was measured at k = 517 nm against methanol, usinga UV/Vis V-530 spectrophotometer (Jasco, Tokyo, Japan). Theantioxidant activity of methanolic extracts of herbhoneys againstDPPH� was determined analogically. In this case 0.5 ml of thesolution containing 100 ll of methanolic extract and 400 ll ofmethanol was applied instead of the water solution of herbhoney.Antioxidant activity was expressed as a percent inhibition ofDPPH radical, and calculated from the equation:

AA½%� ¼ ðAbscontrol � AbssampleÞ=Abscontrol � 100

2.3. Determination of antioxidant activity in the reaction with ABTS�+

cation radical

The antioxidant activity of herbhoneys was determined inaccordance with Baltrusaityte, Venskutonis, and Ceksteryte(2007). ABTS�+ was obtained in the reaction of a 2 mM stock solu-tion of 2,20-azino-di(3-ethylbenzothiazoline-6-sulfonic acid) diam-monium salt (ABTS) (Sigma–Aldrich, Steinheim, Germany) inphosphate-buffered with potassium persulfate (POCh, Gliwice,Poland). The mixture was left to stand for 24 h. Prior to analysis,the ABTS�+ solution was diluted with phosphate buffer to producea solution with an absorbance of 0.80 ± 0.03 at k = 734 nm. One-gram samples of herbhoneys were diluted with distilled water to5 ml, centrifuged at 4350 g, and filtered through paper filters. Then,100 ll of a herbhoney solution was mixed with 6 ml of the ABTS�+

cation radical solution and after 15 min absorbance was measuredat a wavelength of 734 nm by using a UV/Vis V-530 spectropho-tometer (Jasco, Tokyo, Japan). The antioxidant activity of methano-lic extracts of herbhoneys in the reaction with ABTS�+ wasdetermined analogically. In this case 50 ll of methanolic extractswas applied instead of the water solution of herbhoney. Antioxi-dant activity was expressed as a percent inhibition of ABTS�+, calcu-lated from the same equation as for DPPH�.

2.4. Determination of total phenolic content

To determine the total phenolic content of herbhoneys, themethod described by Meda, Lamien, Romito, Millogo, andNacoulma (2005) was employed. Herbhoney solutions with a con-centration of 5 g/50 ml were centrifuged at 4350 g and filteredthrough paper filters. After that, 0.5 ml of the obtained solutionwas mixed with 2.5 ml of 0.2 N solution of Folin–Ciocalteaureagent (Sigma–Aldrich, Steinheim, Germany), and then 2 ml ofsodium carbonate solution (POCh, Gliwice, Poland) was added.Following incubation for 2 h, absorbance of the reaction mixturewas measured at k = 760 nm against a methanol blank using aUV/Vis V-530 spectrophotometer (Jasco, Tokyo, Japan). Thestandard curve was produced for gallic acid (Sigma–Aldrich,Steinheim, Germany) within the concentration range from 0 to200 mg/100 ml. The total phenolic content was expressed as gallicacid equivalents in mg/100 g of herbhoney sample.

2.5. Determination of total flavonoid content

The total flavonoid content of herbhoneys was established inthe reaction with aluminum chloride using the methods describedby Zhishen, Mengcheng, and Jianming (1999), Ardestani andYazdanparast (2007) Herbhoney solutions with a concentrationof 0.2 g/ml were centrifuged at 4350g and filtered through paper

filters. To 1 ml of the solution, 4 ml of distilled water and0.3 ml of sodium nitrate(III) (POCh, Gliwice, Poland) solution with15 g/100 ml concentration were added. This was followed by add-ing 0.3 ml of aluminum chloride (POCh, Gliwice, Poland) solution(10 g/100 ml) and then 4 ml of sodium hydroxide (POCh, Poland)solution (4 g/100 ml). The whole was made up to 10 ml with dis-tilled water, stirred and left to stand for 15 min. The total flavonoidcontent was calculated on the basis of the standard curve forquercetin solutions (Sigma–Aldrich, Germany) and expressed asquercetin equivalents in mg/100 g of herbhoney sample.

2.6. Extraction and analysis of phenolic compounds

Phenolic compounds were extracted by using ethyl acetate asextracting agent (Wahdan, 1998). Herbhoney solutions with a con-centration of 10 g/50 ml were acidified with HCl solution to reachpH 2, and then saturated with sodium chloride (POCh, Gliwice,Poland) in the amount of 3 g/10 ml. The resulting solutions wereextracted with three portions of ethyl acetate (POCh, Gliwice,Poland): one of 50 ml and two of 25 ml each. The extracts werecombined, and ethyl acetate was vacuum-evaporated at 40� Cunder argon atmosphere. The dry residue was dissolved in 5 mlof methanol (POCh, Gliwice, Poland) and stored at �18� C. Priorto the chromatographic analysis, the extracts were purified byusing Millex-LCR syringe filters (PTFE). The recovery of phenoliccompounds for such extraction was from 82% for chlorogenic acidto 106% for gallic acid.

Chromatographic analyses of phenolic compounds were madeby high-performance liquid chromatography (HPLC–LaChrom, Hit-achi, Tokyo, Japan). Phenolic acids such as caffeic, chlorogenic, p-coumaric, ferulic, gallic, sinapic and syringic acid were detectedat k = 280 nm, while gentisic acid and flavonoids (chrysin, galangin,hesperetin, kaempferol, naringenin and quercetin) were deter-mined at k = 330 nm. Gradient elution was conducted at a flow rateof 1 ml/min using a solution of acetic acid (POCh, Gliwice, Poland),2.5 g/100 ml, and acetonitrile (Merck, Darmstadt, Germany) as amobile phase. The solvent system was a linear gradient from 3%of acetonitrile then increased to 8% for 10 min, this was increasedto 15% at 20 min, 20% at 30 min, 30% at 40 min and 40% at 50 min.Finally, the column was eluted isocratically with acetonitrile be-fore the next injection.

The compounds studied were separated on a RP18 Lichrosorbcolumn (250 � 4.5 lm) (Merck, Darmstadt, Germany) at a temper-ature of 25� C. Quantification of individual phenolic acids andflavonoids was based on comparison with standards from Sigma–Aldrich (Steinheim, Germany) and Fluka Chemie AG (Buchs,Switzerland).

2.7. Statistical analysis

For the estimation of the significant of differences between themeans, a one-way analysis of variance and the test of least signif-icant differences (LSD) at p = 0.05 was applied. Calculations wereperformed with statistical software package Statistica 8.0 (StatSoftInc. Tulsa, USA).

3. Results and discussion

3.1. Antioxidant activity

Figs. 1 and 2 show the antioxidant activity of herbhoney solu-tions in the reactions with DPPH� and ABTS�+ radicals, respectively.For the former, the values of this parameter varied between 27.2%(camomile herbhoney) and 86.8% (hawthorn herbhoney), whereasfor the latter they ranged from 21.9% (camomile herbhoney) to

0

20

40

60

80

100

aloe

black

chok

eber

ry

camom

ile

hawth

orn

marigo

ldmin

tnett

lepin

e

rasp

berry

thym

e

antio

xida

nt a

ctiv

ity [

%]

Fig. 1. Antioxidant activity of herbhoney solutions in reaction with DPPH� radical(LSD0.05 = 2.02).

0

20

40

60

80

100

aloe

black

chok

eber

ry

camom

ile

hawth

orn

marigo

ldmin

tnett

lepin

e

rasp

berry

thym

e

antio

xida

nt a

ctiv

ity [

%]

Fig. 2. Antioxidant activity of herbhoney solutions in reaction with ABTS�+ cationradical (LSD0.05 = 2.31).

0

20

40

60

80

100

aloe

black

chok

eber

ry

camom

ile

hawth

orn

marigo

ldmin

tnett

lepin

e

rasp

berry

thym

e

antio

xida

nta

ctiv

ity[%

]

Fig. 3. Antioxidant activity of herbhoney extracts in reaction with DPPH� radical(LSD0.05 = 3.22).

570 R. Socha et al. / Food Chemistry 113 (2009) 568–574

79.2% (raspberry herbhoney). Dark red herbhoneys (raspberry,black chokeberry) and intensely green herbhoney (thyme) tendedto be highly active in the reaction involving DPPH�. The highestantioxidant activity in the reaction with ABTS�+, more than twicethat of most samples, was exhibited by raspberry herbhoney. Theother samples with dark red (black chokeberry) or vividly greencolours (pine and thyme) were much less active. In both reactions,pale yellow–green herbhoneys (camomile, mint, and aloe) showedthe lowest antioxidant activity.

A relationship between colour parameters and antioxidantactivity in natural Slovenian honeys was observed by Bertoncelj,Doberšek, Jamnik, and Golob (2007), who found light honeys(e.g., acacia) to be least active and dark honeys to be most activeagainst DPPH. The linear correlation between the results of thetwo assays (R = 0.61) was significant at p = 0.05. A somewhat stron-ger linear correlation between the antioxidant activities deter-mined with DPPH� and ABTS�+ was found by Baltrusaityte et al.(2007). The honeys examined by the latter authors displayed agreater antioxidant activity for ABTS�+. The differences in the anti-oxidant activities between the DPPH� and ABTS�+ assays, observedin the present study, may be attributed to the different concentra-tions of the substrates (herbhoney samples) on the one hand, andto the different kinetics of the two reactions, on the other hand(Baltrusaityte et al., 2007).

The antioxidant activity of natural herbhoneys depends largelyon their chemical composition, i.e., phenolic and flavonoid content,hence on their origin. Antioxidant activity is also affected by theprocessing (e.g., heating) and storage method of the material. Asshown by Turkmen et al. (2006), the antioxidant activity of naturalhoneys rises when the temperature and time of heating are in-creased. In natural honeys, antioxidant activity is due to the pres-ence of many various substances such as enzymes, organic acids,amino acids, Maillard reaction products, phenolic compounds,flavonoids, tocopherols, catechins, ascorbic acid, and carotenoids(Meda et al., 2005). To produce herbhoneys, bees receive food con-taining plant extracts or fruit juices. Since antioxidant substancesor other bioactive components are present in higher concentra-tions in such a food, they may be more easily available to bees.In addition, the total antioxidant activity of herbhoneys may partlybe due to substances that do not occur, or occur in tiny amounts innatural honeys. Baltrusaityte et al. (2007) showed that honeys ob-tained using plant (birch, pine and nettle) extracts have a higherantioxidant activity and contain much more apigenin than naturalhoneys.

The antioxidant activity of methanolic extracts of herbhoneys inthe reaction with DPPH� and ABTS�+ is shown in Figs. 3 and 4. Thehighest antioxidant activity in the reaction with DPPH was exhib-ited by the extracts obtained from thyme, raspberry and hawthornherbhoneys, analogically to the water solutions of herbhoneys. Thesignificantly lower activity was exhibited by the extract of blackchokeberry herbhoney, that may be attributed to the significantparticipation in antioxidant activity of compounds which werenot extracted under the employed conditions. The lowest antioxi-dant activity in the reaction with DPPH� was exhibited by the ex-tract from aloe herbhoney, where water solution also exhibitedlow activity in this reaction. A significant linear correlation was ob-served (R = 0.91) between the antioxidant activity against DPPH�,which was carried out using water solutions of herhoneys andtheir methanolic extracts. Analogically to the water solutions ofherbhoneys, the highest antioxidant activity against ABTS�+ wasexhibited by extracts obtained from raspberry, hawthorn, andthyme herbhoneys, whereas the extract from aloe herbhoney char-acterized the lowest activity. In the reaction with ABTS�+ the corre-lation between the results for the water solutions and methanolicextracts of herbhoneys was lower (R = 0.79), but statistically signif-icant (p = 0.05). Whereas, considerably higher linear correlation(R = 0.96) was observed between the reactions of methanolic ex-tracts of herbhoneys against DPPH� and ABTS�+, in comparison tothe water solutions of herbhoneys.

0

20

40

60

80

100

aloe

black

chok

eber

ry

camom

ile

hawth

orn

marigo

ldmin

tnett

lepin

e

rasp

berry

thym

e

antio

xida

ntac

tivity

[%]

Fig. 4. Antioxidant activity of herbhoney extracts in reaction with ABTS�+ cationradical (LSD0.05 = 5.91).

0

10

20

30

aloe

black

chok

eber

ry

camom

ile

hawth

orn

marigo

ldmin

tnett

lepin

e

rasp

berry

thym

e

Que

rcet

ineq

uiva

lent

s[m

g/10

0g]

Fig. 6. Total flavonoid content of herbhoneys expressed as quercetin equivalent(LSD0.05 = 1.58).

R. Socha et al. / Food Chemistry 113 (2009) 568–574 571

3.2. Total phenolic and flavonoid content

The total phenolic content of herbhoneys, expressed as gallicacid equivalent, ranged from 21.7 mg/100 g for the camomile sam-ple to 75.3 mg/100 g for the raspberry product (Fig. 5). Being ingeneral within the broad range reported in the literature, these val-ues often were even higher than those of natural honeys (Al-Mam-ary, Al-Meeri, & Al-Habori, 2002; Aljadi & Kamaruddin, 2004;Baltrusaityte et al., 2007; Bertoncelj et al., 2007; Blasa et al.,2006; Küçük et al., 2007; Meda et al., 2005). As with antioxidantactivity, the total levels of phenolic compounds were highest indark red herbhoneys (raspberry, black chokeberry) and intenselyyellow–green ones (hawthorn, thyme), and lowest in light-col-oured products (camomile, mint). This supports the observationsmade by others on the correlation between the phenolic contentand antioxidant activity of natural honeys and their colour(Bertoncelj et al., 2007). In particular, Meda et al. (2005) found thatdark honeys, such as honeydew ones, have a higher phenolic con-tent. It should be taken into account, however, that herbhoneysmay somewhat differ from natural honeys in the profile of com-pounds reducing the Folin–Ciocalteu reagent, among them pheno-lics. Besides, they may contain other compounds, e.g., amino acids,that contribute to the total antioxidant activity of the product. Asignificant correlation between the levels of individual amino acidsand antioxidant activity was found by Pérez, Iglesias, Pueyo,González, and de Lorenzo (2007) for natural Spanish honeys. Also

0

20

40

60

80

aloe

black

chok

eber

ry

camom

ile

hawth

orn

marigo

ldmin

tnett

lepin

e

rasp

berry

thym

e

Gal

licac

ideq

uiva

lent

s[m

g/10

0g]

Fig. 5. Total phenolic content of herbhoneys expressed as gallic acid equivalent(LSD0.05 = 2.50).

Meda et al. (2005) observed a significant relationship betweenproline content and antioxidant activity. Many authors report thattotal phenolic content is strongly linearly correlated with antioxi-dant activity (Buratti, Benedetti, & Cosio, 2007; Meda et al.,2005). In the present study, a significant (p = 0.05) linear correla-tion was observed between total phenolic content and antioxidantactivity of herbhoney solutions and extracts against DPPH�

(R = 0.73, R = 0.70) and ABTS�+ (R = 0.96, R = 0.81), respectively.The total flavonoid content of herbhoneys, expressed as

quercetin equivalent, was the highest for raspberry herbhoney(28.5 mg/100 g; Fig. 6). Hawthorn, thyme and black chokeberryherbhoneys also contained a considerable amount of flavonoids.This confirms the results indicating a high antioxidant activity ofthose products in the reactions with DPPH� and ABTS�+ (Figs. 1–4).The flavonoid content was lowest for camomile herbhoney(6.9 mg/100 g). Again, flavonoid content appeared to be closelyconnected with colour; herbhoneys with strong red or greencolours (raspberry, black chokeberry, thyme) had a high flavonoidcontent, while those with pale yellow or pale green colours(camomile, mint) contained a small amount of these compounds.

The total flavonoid content determined in this study is signifi-cantly higher than the values reported for natural honeys (Blasaet al., 2006; Meda et al., 2005). This may be due to the fact thatthe food fed to bees in the production of herbhoneys constitutesa richer source of flavonoids; therefore, the products obtained insuch a way have a much higher flavonoid content than naturalhoneys. In addition, intensely red herbhoneys, such as raspberryand black chokeberry, may also contain anthocyanin pigments.

Although the determination of phenolics by using the Folin–Ciocalteu reagent and the determination of flavonoids by usingaluminum chloride are based on different mechanisms of thereaction, and the reactants exhibit different affinities to individualsubstrates, a significant (p = 0.05) linear correlation (R = 0.83)between total phenolic content and total flavonoid content wasobserved in the present study. There was also a significant(p = 0.05) linear correlation between total flavonoid content andantioxidant activity of herbhoney solutions and extracts: R = 0.70,R = 0.90 for ABTS�+, and R = 0.95, R = 0.91 for DPPH, respectively.Blasa et al. (2006) report that total flavonoid content is stronglylinearly correlated with antioxidant activity.

3.3. Profile of phenolic acids and flavonoids

The levels of individual phenolic acids in herbhoneys are spec-ified in Table 1. Among the phenolic acids, p-coumaric acid waspresent in the largest amounts, followed by gentisic acid; the

Table 1Phenolic acid content of herbhoneys

Herbhoney variety Caffeic acid(lg/100 g)

Chlorogenic acid(lg/100 g)

p-Coumaric acid(lg/100 g)

Ferulic acid(lg/100 g)

Gallic acid(lg/100 g)

Gentisic acid(lg/100 g)

Sinapic acid(lg/100 g)

Syringic acid(lg/100 g)

Aloe 10.1a 794.6a 112.8a 39.3a,b 189.2a

Black chokeberry 118.9b 13.7a 446.4b 41.4b 111.3b,c 229.2a,b 44.6a 7.8a

Hawthorn 517.4 336.4 361.0b 113.2a 16.3a 1012.8 274.9 58.8Raspberry 35.8c 24.9a 3676.7 55.6b 1126.0 136.8a 24.9b

Mint 25.7c 240.8c 61.3b 92.7c 1645.8 24.3b

Marigold 106.7b 316.7b 13.2c 74.3b,c 576.9 104.6 14.4a

Nettle 20.4a 350.8b 18.3c 61.0a 123.3a 23.6b

Camomile 35.5c 199.6c 46.3b 12.1a 337.8b 43.4a 1.4Pine 27.7c 629.7a 22.4c 153.3c 204.6a 48.6a 9.7a

Thyme 52.0 1141.9 178.6 23.1a 142.3a 119.0

Means in the same columns with the same superscript letters are not significantly different at p = 0.05.

572 R. Socha et al. / Food Chemistry 113 (2009) 568–574

syringic and chlorogenic acid contents were lowest. Raspberryherbhoney proved the richest in phenolic acids (5082 lg/100 g intotal), followed by hawthorn herbhoney, while nettle and camo-mile products were poorest in this respect. This corresponds withthe high antioxidant activity of raspberry and hawthorn herbho-neys in the DPPH� (Fig. 1) and ABTS�+ (Fig. 2) reactions and theirhigh phenolic content (Fig. 3), on the one hand, and the low valuesof these parameters for nettle and camomile products, on theother. Chlorogenic acid, which was found only in three herbhoneys,reached its highest level in hawthorn herbhoney. It was also pres-ent in black chokeberry herbhoney. The presence of small amountsof chlorogenic acid in the latter has been mentioned in the litera-ture (Oszmianski & Wojdyło, 2005). The chlorogenic acid was alsodetermined in Australian and New Zealand honeys (Yao et al.,2003; Yaoa et al., 2005). Mint and raspberry herbhoneys provedrich in gentisic acid, and thyme herbhoney in ferulic acid. The gallicacid level was highest for raspberry herbhoney which contained 10times more acid than most of the other products.

Gallic acid has already been identified in natural honeys. Aljadiand Yusoff (2003) found its levels to range from 82 to 330 lg/100 gin Malaysian honeys, whereas Yaoa et al. (2005) determined muchhigher values in Australian honeys. The latter authors establishedthat gallic acid is the main component of the phenolic acid profileand a potential marker of the origin of a honey. Yao et al. (2003)found gallic acid to be the dominant acid in some Australian andNew Zealand honeys and observed a high level of abscisic acid inthe natural honeys they examined. In the present study, significant(p = 0.05) linear correlations were observed between gallic acidcontent and total phenolic content (R = 0.88), and between the for-mer and the antioxidant activity of herbhoney solutions againstABTS�+ (R = 0.93).

The amount of ferulic acid in most herbhoneys was at the levelreported by Tomás-Barberán et al. (2001) for natural Europeanhoneys (20–100 lg/100 g), but in thyme herbhoney it was higher.Hawthorn herbhoney exhibited the highest caffeic acid contentthat was more than five times higher than the level found in mostof the other products which were similar to European honeys inthis respect. To compare, Tomás-Barberán et al. (2001) report thelevels of caffeic acid in European honeys of different origin to varybetween 20 and 160 lg/100 g, with the highest value being foundfor sunflower honey.

The p-coumaric acid content was largest in raspberry herbhon-ey (more than 10 times larger than in most of the samples). Alsothyme herbhoney had a high p-coumaric acid content which, how-ever, was more than three times lower than in raspberry herbhon-ey. The fact that among the acids under study p-coumaric acidshowed highest levels in 9 of the 10 products suggests that thisacid is characteristic of the herbhoneys studied. A high level of p-coumaric acid was also observed by Baltrusaityte et al. (2007) innatural honeys and in honeys with plant extracts; the level was

the highest in those with birch extracts. Compared to our results,the p-coumaric acid contents of various natural European honeys,established by Tomás-Barberán et al. (2001), are much lower(10–180 lg/100 g). The level of p-coumaric acid was found toinfluence the antioxidant activity and total phenolic content ofherbhoneys. There were significant (p = 0.05) linear correlationsbetween p-coumaric acid content and total phenolic content(R = 0.91), and between acid content and antioxidant activity ofherbhoneys solutions against ABTS�+ (R = 0.92), confirming thefindings on the high antioxidant activity of raspberry and thymeherbhoneys in the reactions with DPPH and ABTS�+. The highp-coumaric acid content of thyme herbhoney is supported by theliterature data on thyme. Namely, Wojdyło, Oszmianski, andCzemerys (2007), investigating the antioxidant activity of variousherb species observed a high level of this acid in thyme.

Sinapic acid reached its highest level in hawthorn herbhoney,and was not observed in two herbhoneys. Syringic acid contentwas largest in thyme herbhoney, while the acid did not occur infour products. Studies by other authors have also found that syrin-gic acid occurs only in some varieties of honey (Yao et al., 2003;Yaoa et al., 2005).

The levels of individual flavonoids determined in herbhoneysare shown in Table 2. The totals were largest for thyme herbhoney,followed by raspberry herbhoney. This corresponds with the highantioxidant activity of both products in the reactions with DPPH�

and ABTS�+ (Figs. 1–4). The two herbhoneys had also a high totalflavonoid content (Fig. 6). Aloe herbhoney contained the smallestamount of the flavonoids determined, which corresponds with itsrelatively low antioxidant activity (Figs. 1–4) and average total fla-vonoid content (Fig. 4). By contrast, black chokeberry herbhoney,containing a low amount of the determined flavonoids, had a rela-tively high antioxidant activity (Figs. 1 and 2) and total phenoliccontent (Fig. 6). In this case, the high antioxidant activity may beconnected with the presence of anthocyanin pigments responsiblefor the dark red colour of the product. The same is true for dark redraspberry herbhoney whose high antioxidant activity and totalphenolic content are due to anthocyanin pigments. Of all samples,the latter product contained the largest amount of chrysin(Table 2), while six herbhoneys showed its absence. The presenceof chrysin in Australian and New Zealand honeys is rather limited(Yao et al., 2003), but it appears to be the main flavonoid (alongwith kaempferol and apigenin) of Lithuanian honeys (Baltrusaityteet al., 2007). Kenjeric et al. (2007) reported a much larger range ofchrysin contents than that found in the present study and notedthat chrysin made more than 50% of the total flavonoid contentof Robinia honeys from Croatia. They also observed significant dif-ferences in flavonoid profiles due to climatic conditions; honeysproduced in a year with a sunny and warm (high UV–B radiation)summer season with low rainfalls contained much moreflavonoids. The range of chrysin contents determined by Tomás-

Table 2Flavonoid content of herbhoneys

Herbhoney variety Chrysin(lg/100 g)

Galangin(lg/100 g)

Hesperetin(lg/100 g)

Kaempferol(lg/100 g)

Naringenin(lg/100 g)

Quercetin(lg/100 g)

Aloe 25.7a 14.3a 2.9a 21.7a,b

Black chokeberry 3.5a 10.5 14.5a 31.3b 11.6a

Hawthorn 23.3b 5.8a,b 28.5a 102.4b 122.7c 32.1a,c

Raspberry 104.0 7.2b 212.1 98.3b 252.5 87.5d

Mint 130.9 3.94a 110.5c 43.9b,c

Marigold 37.2a,b 95.4b 24.4b 46.7b,c

Nettle 4.7a 2.6a 172.8 244.7 189.7d 72.5d

Camomile 257.8 61.6c 195.3d 64.3d

Pine 6.7a 7.7b 50.0b 68.4c 2.8a 20.7a

Thyme 27.9b 48.5 279.9 101.9b 14.6a,b 1199.9

Means in the same columns with the same superscript letters are not significantly different at p = 0.05.

R. Socha et al. / Food Chemistry 113 (2009) 568–574 573

Barberán et al. (2001) for a number of natural European honeyvarieties (9–415 lg/100 g) was also much broader than that estab-lished in herbhoneys in the present study, except for chestnut andcitrus honeys whose chrysin levels were similar to our values. Asignificant (p = 0.05) linear correlation occurred only betweenchrysin content and antioxidant activity in the ABTS�+ reaction ofherbhoneys solutions (R = 0.96).

Herbhoneys contained also a relatively small amount of galan-gin that was identified and determined in five samples amongwhich thyme herbhoney was richest. The other herbhoneys had alower galangin content than Robinia honeys from Croatia (Kenjericet al., 2007). Considerable levels of chrysin and galangin were alsoobserved in Portuguese heather honeys (Ferreres et al., 1994). Theamount of hesperetin in herbhoneys was quite large, the largest inthyme herbhoney, followed by camomile nad raspberry. In thethyme and raspberry products, the high level of this flavonoid cor-relates well with their high antioxidant activity and total flavonoidcontent. Soler et al. (1995) found hesperetin to be a characteristicflavonoid of French citrus honeys but not of other varieties ofFrench honeys.

Kaempferol constitutes one of the main flavonoids occurring innatural honeys and has already been determined by many authors(Baltrusaityte et al., 2007; Ceksteryte et al., 2006; Ferreres et al.,1998; Kenjeric et al., 2007; Martos et al., 2000; Soler et al., 1995;Yao et al., 2003). Among the herbhoneys studied, nettle herbhoneycontained the highest amount of kaempferol. The second richestproducts were herbhoneys with a high antioxidant activity, i.e.,hawthorn, raspberry and thyme. The levels of this flavonoid wereclose to those reported by Ferreres et al. (1998) for Spanish rose-mary honey. Besides apigenin and chrysin, kaempferol is one ofthe main flavonoids found in Lithuanian honeys (Baltrusaityteet al., 2007). Our values are similar to those obtained by Yaoet al. (2003) for natural Australian and New Zealand honeys, butslightly higher than the levels observed by Kenjeric et al. (2007)for Croatian honeys and Ceksteryte et al. (2006) for Lithuanianhoneys.

Investigating the flavonoid profile of natural European hon-eys, Tomás-Barberán et al. (2001) noted significant differencesin kaempferol contents. Among these honeys, heather andrapeseed ones were especially abundant in kaempferol, muchmore than our herbhoneys, while accacia honey was poorest inthis respect.

Another important flavonoid occurring in honeys is quercetin.Soler et al. (1995) found it to be a characteristic flavonoid of sun-flower honey. In our studies, the richest herbhoney was the thymeproduct with its over 20 times higher quercetin content comparedto the other herbhoneys studied (Table 2.). This is closely con-nected with the origin of the herbal extract used in its production.The quercetin content of the other herbhoneys fell within thebroad ranges reported by Tomás-Barberán et al. (2001) for natural

European honeys and by Yao et al. (2003) for Australian and NewZealand honeys, and was significantly higher that the quercetinlevels determined by Ceksteryte et al. (2006) in Lithuanian honeys.All the herbhoneys contained naringenin (Table 2). Its amount waslargest in raspberry herbhoney, and smallest in pine and aloe prod-ucts. In three products (raspberry, hawthorn, and black choke-berry), naringenin was the most abundant flavonoid determined.Whether naringenin occurs in natural honeys or not, has not yetbeen mentioned in the literature.

4. Conclusion

Herbhoneys differed in antioxidant activity and the profiles ofphenolic acids and flavonoids. Dark-coloured products, such asraspberry, thyme, hawthorn, and black chokeberry herbhoneys,showed high antioxidant activities (for both free radicals) and totalphenolic and flavonoid contents. There were significant linear cor-relations between these parameters. The phenolic acid and flavo-noid profiles depended on the kind of product, with raspberryand thyme herbhoneys being richest in phenolic acids and flavo-noids. Among the phenolic acids, p-coumaric acid prevailed inmost herbhoneys. The products contained also a considerableamount of gentisic acid. Among the flavonoids determined in thesestudies, hesperetin and naringenin were dominant. Thyme herb-honey contained an especially large amount of quercetin.

To determine the whole profile of polyphenols in herbhoneys, itwould be necessary to undertake further research consideringanthocyanin pigments and other biologically active compoundscharacteristic of the herbs or fruit juices used in herbhoney pro-duction. In herbhoneys, such compounds may have a significant ef-fect on antioxidant activity and total phenolic and flavonoidcontents

References

Aljadi, A. M., & Yusoff, K. M. (2003). Isolation and identification of phenolic acids inMalaysian honey with antibacterial properties. Turkish Journal of MedicalSciences, 33, 229–236.

Aljadi, A. M., & Kamaruddin, M. Y. (2004). Evaluation of the phenolic contents andantioxidant capacities of two Malaysian floral honeys. Food Chemistry, 85,513–518.

Al-Mamary, M., Al-Meeri, A., & Al-Habori, M. (2002). Antioxidant activities and totalphenolics of different types of honey. Nutrition Research, 22, 1041–1047.

Ardestani, A., & Yazdanparast, R. (2007). Antioxidant and free radical scavengingpotential of Achillea santolina extracts. Food Chemistry, 104, 21–29.

Baltrusaityte, V., Venskutonis, P. R., & Ceksteryte, V. (2007). Radical scavengingactivity of different floral origin honey and beebread phenolic extracts. FoodChemistry, 101, 502–514.

Bertoncelj, J., Doberšek, U., Jamnik, M., & Golob, T. (2007). Evaluation of the phenoliccontent, antioxidant activity and colour of Slovenian honey. Food Chemistry, 105,822–828.

Blasa, M., Candiracci, M., Accorsi, A., Piacentini, M. P., Albertini, M. C., & Piatti, E.(2006). Raw Millefiori honey is packed full of antioxidants. Food Chemistry, 97,217–222.

574 R. Socha et al. / Food Chemistry 113 (2009) 568–574

Buratti, S., Benedetti, S., & Cosio, M. S. (2007). Evaluation of the antioxidant power ofhoney, propolis and royal jelly by amperometric flow injection analysis. Talanta,71, 1387–1392.

Ceksteryte, V., Kazlauskas, S., & Racys, J. (2006). Composition of flavonoids inLithuanian honey and beebread. Biologija, 2, 28–33.

Ferreres, F., Andrade, P., & Tomás-Barberán, F. A. (1994). Flavonoids fromPortuguese heather honey. European Food Research and Technology, 199, 32–37.

Ferreres, F., Juan, T., Pérez-Arquillué, C., Herrera-Marteache, A., Garcia-Viguera, C., &Tomás-Barberán, F. A. (1998). Evaluation of pollen as a source of kaempferol inrosemary honey. Journal of the Science of Food and Agriculture, 77, 506–510.

Juszczak, L., Socha, R., Ro _znowski, J., Fortuna, T., & Nalepka, K. (2008).Physicochemical properties and quality parameters of herbhoneys. FoodChemistry, doi:10.1016/j.foodchem.2008.07.098.

Kenjeric, D., Manic, M., Primorac, L., Bubalo, D., & Perl, A. (2007). Flavonoid profile ofRobinia honeys produced In Croatia. Food Chemistry, 102, 683–690.

Küçük, M., Kolaylı, S., Karaoglu, S., Ulusoy, E., Baltacı, C., & Candan, F. (2007).Biological activities and chemical composition of three honeys of different typesfrom Anatolia. Food Chemistry, 100, 526–534.

Meda, A., Lamien, C. E., Romito, M., Millogo, J., & Nacoulma, O. G. (2005).Determination of the total phenolic, flavonoid and proline contents in BurkinaFasan honey, as well as their radical scavenging activity. Food Chemistry, 91,571–577.

Martos, I., Ferreres, F., Yao, L., D’Arcy, B., Caffin, N., & Tomás-Barberán, F. A. (2000).Flavonoids in monospecific eucalyptus honeys from Australia. Journal ofAgricultural and Food Chemistry, 48, 4744–4748.

Oszmianski, J., & Wojdyło, A. (2005). Aronia melanocarpa phenolics and theirantioxidant activity. European Food Research and Technology, 221, 809–813.

Pérez, A. R., Iglesias, M. T., Pueyo, E., González, M., & de Lorenzo, C. (2007). Aminoacid composition and antioxidant capacity of Spanish honeys. Journal ofAgricultural and Food Chemistry, 55, 360–365.

Soler, C., Gil, M. I., Garcia-Viguera, C., & Tomás-Barberán, F. A. (1995). Flavonoidpatterns of French honeys with different floral origin. Apidologie, 26, 53–60.

Tomás-Barberán, F. A., Martos, I., Ferreras, F., Radovic, B. S., & Anklam, E. (2001).HPLC flavonoid profiles as markers for the botanical origin of European unifloralhoneys. Journal of the Science of Food and Agriculture, 81, 485–496.

Turkmen, N., Sari, F., & Velioglu, Y. S. (2006). Effects of prolonged heating onantioxidant activity and colour of honey. Food Chemistry, 95, 653–657.

Wahdan, H. A. L. (1998). Causes of the antimicrobial activity of honey. Infection, 26,30–35.

Wojdyło, A., Oszmianski, J., & Czemerys, R. (2007). Antioxidant activity and phenoliccompounds in 32 selected herbs. Food Chemistry, 105, 940–949.

Yao, L., Datta, N., Tomás-Barberán, F. A., Ferreres, F., Martos, I., & Singanusong, R.(2003). Flavonoids, phenolic acids and abscisic acid in Australian and NewZealand Leptospermum honeys. Food Chemistry, 81, 159–168.

Yaoa, L., Jiang, Y., Singanusong, R., Datta, N., & Raymont, K. (2005). Phenolic acids inAustralian Melaleuca, Guioa, Laphostemon, Banksia and Helianthus honeys andtheir potential for floral authentication. Food Research International, 38,651–658.

Zhishen, J., Mengcheng, T., & Jianming, W. (1999). The determination of flavonoidcontents in mulberry and their scavenging effects on superoxide radicals. FoodChemistry, 64, 555–559.