JFS: Sensory and Nutritive Qualities of Food

Blueberry and Grape Anthocyanins as

Breakfast Cereal Colorants
M.E. CAMIRE, A. CHAOVANALIKIT, M.P. DOUGHERTY, AND J. BRIGGS

ABSTRACT: Fruit anthocyanins provide color and health benefits, thus cereals containing these pigments could be
used as functional foods. White corn meal with 10% sucrose was twin-screw extruded with either corn syrup, lowbush
blueberry concentrate, or Concord grape-juice concentrate. Extrusion reduced anthocyanins, but there was no
change due to storage at room temperature for 3 mo. Polymeric color was higher in the blueberry cereal. The grape
cereal was lighter and less red than the blueberry product. Bulk density was highest in the corn-syrup cereal. Overall
acceptability was higher for the syrup and grape cereals. Sweetness and flavor acceptability were correlated with
overall liking.
Key Words: anthocyanins, extrusion, hedonic, blueberry, grape

Introduction ing Co., Chicago, Ill., U.S.A.), sucrose (Domino Sugar Corpo-

A NTHOCYANINS ARE WATER-SOLUBLE PIGMENTS RESPONSIBLE

for the red, blue, and purple colors in many food crops.
They are typically found in fruits such as blueberries, straw-
ration, New York, N.Y., U.S.A.), wild blueberry concentrate
and spent (a by-product of osmotic dehydration) ( Jasper Wy-
man & Son, Milbridge, Maine, U.S.A.), grape-juice concen-
berries, cranberries, and grapes. Recently researchers have trate (Welch’s, Concord, Mass., U.S.A.), and Karo light corn
focused on the health benefits of these pigments, especially syrup (CPC International Inc., Englewood Cliffs, N.J., U.S.A.)
their antioxidant activity. Frankel and others (1998) found were used to produce the extruded cereals.
that anthocyanin concentration of Concord grape juice cor-
related with the inhibition of copper-catalyzed oxidation of Sample Preparation
human low-density lipoprotein in vitro. Heinonen and others White corn meal was mixed with 10% sugar (w/w) then
(1998a) studied the antioxidant activity of several berries on stored at room temperature prior to extrusion. Corn syrup
low-density lipoprotein and liposome oxidation. Blueberry (96 8Brix), blueberry concentrate (75 8Brix), and blueberry
extract inhibited hexanal formation in low-density lipopro- spent (85 8Brix) were adjusted to have the same total solids
tein oxidation better than strawberries. In liposome oxida- as grape juice (69 8Brix) by adding tap water. The ratio of wa-
tion, blueberries inhibited hexanal formation better than red ter to corn syrup was 1.36 kg:4.08 kg. The ratio of water to
raspberries, blackberries, and strawberries. Heinonen and blueberry concentrate was 0.77 kg:9.07 kg. However, blue-
others (1998b) also found that grape extract could inhibit hy- berry spent could be not pumped at an even rate, so no ex-
droperoxide and hexanal formation. Aglycones of anthocya- truded products were obtained. Grape-juice concentrate was
nins could inhibit low-density lipoprotein oxidation (Satué- used as is.
Gracia and others 1997). The oxygen radical absorbance
capacity (ORAC) test has been used by Prior and others Extrusion Cooking
(1998) to evaluate the antioxidant value of acetonitrile/4% A Werner & Pfleiderer ZSK-30 (Werner and Pfleiderer,
acetic acid extracts of foods. ORAC value for blueberries and Ramsey, N.J., U.S.A.) co-rotating twin-screw extruder was
bilberries were correlated with anthocyanin content (r = used to extrude corn meal with corn syrup, blueberry con-
0.77) and total phenols (r = 0.92) (Prior and others 1998). centrate, and grape-juice concentrate. The length and dia of
Other health benefits of blueberries have been reviewed each screw were 936 and 30 mm, respectively. A single die
(Camire 2000). Estimates of anthocyanin consumption are not with the dimensions of 4 mm in dia by 9.7 mm long was
yet available, but it seems desirable to include these pigments used. The barrel temperature profile used was: 38 - 49 - 116 -
in the diet. Breakfast cereals could become a convenient 138 - 113 8C, from feed to die end. Screw speed was 300 rpm.
source of these pigments. Many commercial products are ar- Barrel temperatures and screw speed during extrusion were
tificially colored, and a few cereals are now colored with fruit monitored through the control cabinets of the ZSK-30 ex-
juices and other natural sources. Most cereals are processed truder.
by extrusion cooking, yet little has been published about the The mixture of corn meal with sugar was fed with an auto-
Sensory and Nutritive Qualities of Food

stability of anthocyanins during extrusion. The objective of matic feeder (Accurate, Whitewater, Wis., U.S.A.) through a
this project was to evaluate the stability and acceptability of hopper to the feeding section of the extruder at a constant
blueberry and grape anthocyanins in extruded cereals. rate of 13.6 kg/h. Liquid material (corn syrup, blueberry con-
centrate, or grape-juice concentrate) was pumped to the mix-
Materials and Methods ture of corn meal at the feed section at the rate of 4.4 kg/h.
The mixtures were continuously extruded without inter-
Materials ruption in a random sequence. Before and after each run,
White corn meal (9.3% moisture content, J.R. Short Mill- the extruder was allowed to self-clean and to stabilize oper-

438 JOURNAL OF FOOD SCIENCE—Vol. 67, Nr. 1, 2002 © 2002 Institute of Food Technologists
Anthocyanins in Cereals . . .

Table 1—Anthocyanin changes in corn-based cereals due to extrusion and storage at room temperature for 3 mo a
Storage time Anthocyanins Degradation Anthocyanin
Sample (mo.) (mg/100g, d.b.) Color density Polymeric color % tannin index degradation index
Nonextruded – 40.71 ± 0.94 a 4.22 ± 0.03 a 0.99 ± 0.05 a 23.52 ± 1.11 c 1.26 ± 0.01 b 2.07 ± 0.11 a
blueberry
Extruded 0 4.03 ± 0.99 cd 0.89 ± 0.08 c 0.29 ± 0.02 c 32.46 ± 3.22 b 1.46 ± 0.08 ab 1.03 ± 0.04 b
blueberry
Extruded 3 5.13 ± 1.34 c 0.96 ± 0.24 c 0.50 ± 0.05 b 54.86 ± 17.19 a 1.76 ± 0.24 a 0.96 ± 0.08 b
blueberry
Nonextruded – 9.94 ± 0.25 b 1.43 ± 0.07 b 0.12 ± 0.00 d 8.32 ± 0.62 d 1.30 ± 0.03 b 1.20 ± 0.01 b
grape
Extruded grape 0 2.64 ± 0.17 d 0.45 ± 0.03 d 0.11 ± 0.01 d 23.79 ± 1.54 bc 1.61 ± 0.18 ab 1.09 ± 0.05 b
Extruded grape 3 3.50 ± 0.25 d 0.41 ± 0.08 d 0.13 ± 0.01 d 32.13 ± 4.67 bc 1.61 ± 0.31 ab 1.16 ±0.23 b
a Means 6 standard deviations of 2 runs, N = 6
b Different letters in the same column indicate significant difference at p # 0.05.

ating conditions as indicated by steady torque. Upon release major pigment in blueberries, the reported molar absorbance
from the extruder, extrudates were cut with a motorized sin- is questionably low (Wrolstad 1976). The molecular weight is
gle-blade cutter. The cutter speed for corn-syrup samples was 562.5 (Koeppen and Basson 1966). The maximum absorption
700 rpm, while the cutter speed for blueberry and grape sam- is at 538 nm. The extinction coefficient is 29,500 L cm-1 mg-1.
ples was 900 rpm because of the stickiness of samples. After The absorbance was measured with a Spectronic 20D1 (Spec-
extrusion had reached equilibrium as indicated by stable tronic Instrument, Rochester, New York, N.Y., U.S.A.) at 538
torque and product appearance, samples were collected on a nm and 700 nm.
metal screen in a single layer. Samples were cooled to room Color density, polymeric color, and the percent of tannin
temperature then dried at 150 8C for 5 min in a rotating contribution were measured using the bisulfite solution
Hobart convection oven (Model DRO Series 6, Troy, N.Y., method (Somers 1971; Wrolstad 1976). Color density was
U.S.A.). After drying, they were allowed to cool to room tem- measured by adding 0.6 mL of distilled water to 9 mL of ex-
perature, packed into tri-laminate bags (Cadhill Products, tracting solvent. The absorbance was measured with a Spec-
Paris, Ill., U.S.A.), and sealed (Model 24-AS, Sencorp Systems tronic 20D1 (Spectronic Instrument) at 420 nm, 538 nm, and
Inc., Hyannis, Mass., U.S.A.). All samples were stored at room 700 nm. Polymeric color was measured by adding 0.6 mL of
temperature for further analysis. 20% potassium metabisulfite (FisherChemical, Fair Lawn,
N.J., U.S.A.) to 9 mL of extracting solvent. The percent con-
Methods of Analysis tribution of tannin was calculated as the ratio of polymeric
One bag of extrudate from each treatment was milled in a color to color density.
Thomas Wiley Laboratory mill (Model 4, Arthur H. Thomas Degradation index was measured by the proportion of to-
Co., Philadelphia, Pa., U.S.A.) with a U.S. No. 20 sieve. The tal anthocyanin by the single pH method to total anthocya-
ground samples were placed in plastic zip bags and stored at nin by the pH differential method (Fuleki and Francis 1968).
room temperature. Anthocyanin content in blueberry and The anthocyanin-degradation index was measured as the ra-
grape samples were analyzed in triplicate twice: after extru- tio of the absorbance at the anthocyanin lambda max to the
sion and after storage at room temperature for 3 mo. Sam- absorbance at 420 nm (Wrolstad 1976).
ples of 20-g ground product were blended at low speed for All extruded samples were ground prior to analysis for
10 min with 200 mL of 0.1% hydrochloric acid in methanol Hunter L*a*b* color. The average dia of 10 extrudates per
(w/v) solution (EM Science, Gibbstown, N.J., U.S.A.) (Koep- extrusion run were measured for expansion ratio. Bulk den-
pen and Basson 1966). The contents were then placed into sity was measured by lightly packing cereals in a 1-L tared
250-mL aluminum foil jacketed plastic centrifuge bottles beaker. The moisture content of all nonextruded and extrud-
(Nalgene Labware, Nalge Company, Rochester, N.Y., U.S.A.). ed samples were analyzed by drying 16 h in a 105 8C oven
A magnetic stir bar was placed into each centrifuge bottle. (AOAC method 950.46 1990).
Nitrogen gas was flushed into the centrifuge bottles to purge
oxygen. The contents were stirred overnight on individual Sensory Evaluation
stir plates at room temperature. The bottles were then cen- Ninety-two Univ. of Maine staff and students who eat
trifuged for 10 min in a Sorvall RC-5B Refrigerated Super- sweetened cereals were recruited by flyers and on the First
speed Centrifuge (DUPONT Co., Wilmington, Del., U.S.A.) Class Computer Conference System for a central location
with a GSA head at 8,000 rpm. The relative centrifugal force test for acceptability. The test protocol was approved by the
was 10410 3 g. The supernatant was filtered under vacuum Human Subjects Protection Committee of the College of
through Whatmann No. 1 filter paper then poured into a Natural Sciences, Forestry, and Agriculture, Univ. of Maine.
Sensory and Nutritive Qualities of Food

weighed 500-smL flask. Another 100 mL of extraction solvent The test was conducted at the Memorial Union, Univ. of
was poured into the centrifuge bottle for re-extraction of the Maine, Orono. Since each treatment had extruded in dupli-
pellet. The extracts were treated similarly and combined. The cate, only 1 duplicate per treatment was tested in order to
total weight of the supernatant was recorded and converted minimize sensory overload, thus each panelist evaluated 3
to volume (100 mL of supernatant weighed 79.44 g). samples (1 each of corn syrup, blueberry, and grape). Cereals
Monomeric anthocyanin pigment content was determined were served without milk in 56-mL plastic cups. No replica-
with a pH differential method (Wrolstad 1976). Anthocyanin tion of the sensory test was performed due to the large num-
content was reported as mg of malvidin-3-glucoside/100 g on ber of subjects that participated. The panelists answered de-
a dry-weight basis. Although delphinidin-3-glucoside is the mographic questions on the 1st page of the evaluation packet

Vol. 67, Nr. 1, 2002—JOURNAL OF FOOD SCIENCE 439

 Anthocyanins in Cereals . . .

before evaluating samples by using a paper ballot. Questions Table 2—Moisture and expansion propertiesa
included gender, age, frequency of cereal purchase and con- Expansion Bulk density
sumption, preference for sweetened or unsweetened cereal, Sample % Moisture ratio (g/cc)
and important cereal quality characteristics. A 9-point he-
Nonextruded 17.66 6 0.01b – –
donic scale (1 = dislike extremely, 5 = neither like nor dislike, corn-syrup mixture
9 = like extremely) (Peryam and Pilgrim 1957) was used to Extruded corn- 4.76 6 0.58c 1.90 6 0.17b 0.33 6 0.05a
rate the 3 samples for overall acceptability and acceptability syrup mixture
of color, sweetness, hardness, and flavor. Nonextruded 18.65 6 0.03b – –
Fifty-two percent of panelists were female, and 48% were blueberry mixture
Extruded blueberry 3.66 6 0.84c 2.05 6 0.16a 0.18 6 0.03b
male. The ages of panelists ranged from 16 to 70 y old, with a
mixture
mean of 26 y and a median of 22 y. Most panelists ate cereals Nonextruded 20.53 6 0.04a – –
often; 37.4% of panelists ate cereals every d. More panelists grape mixture
preferred sweetened breakfast cereals than unsweetened ce- Extruded grape 3.80 6 0.02c 1.96 6 0.13ab 0.23 6 0.03b
reals. Quality was the important factor considered when mixture
buying breakfast cereals, followed by price, taste, and nutri- a Means 6 standard deviations of 2 extrusion runs; different letters within
columns indicate significant differences (p # 0.05, Tukey’s HSD).
tion, respectively.

Experiment Design and Statistical Methods

A full factorial design using the General Linear Model was
used for chemical variables, and 1-way ANOVA was used in well as from added metals due to screw wear. The browning
physical variables by using SYSTAT 7.0 (SYSTAT, Evanston, Ill., reaction could be caused by the Maillard reaction and sugar
U.S.A.) to find the difference among treatments. If a difference hydrolysis.
was detected, Tukey’s Honest Significant Difference test was Degradation index and anthocyanin degradation index
used to compare means. A probability level of p # 0.05 was were used to measure the degradation of anthocyanins (Ta-
considered significant for all statistical procedures. One-way ble 1). Extrusion increased the degradation of blueberry an-
analysis of variance was used to determine the effect of fruit thocyanins, which might be caused by browning; storage
concentrates on the sensory properties of cornmeal extru- time had no degradative effect on blueberry extrudates.
dates. Pearson correlation coefficients were also calculated.
Physical Properties
Results and Discussion Expansion ratio and bulk density are important character-
istics for extruded puffed products. Blueberry products had
Extrusion Cooking higher expansion ratio and lower bulk density than those
Blueberry concentrate and grape-juice concentrate did not with just corn syrup added (Table 2). Moisture content may
affect melt temperature, torque, and die pressure during ex- affect expansion ratios (Ilo and others 1996). Since the pump
trusion. Torque ranged from 35% to 44%. Blueberry spent rate and total solids were held constant, the type of fruit
could not be pumped steadily, thus no product was obtained. concentrate added had no effect on moisture content in ex-
truded products. Prior to extrusion, however, the grape con-
Anthocyanins centrate mix did have a higher moisture content. Extrusion
Blueberry concentrate (69 8Brix) had a higher anthocya- decreased moisture content. Expansion was inversely corre-
nin content (125.4 g/100 g, fresh weight) than did grape-juice lated with moisture and bulk density (r # -0.85).
concentrate (28.3 g/100g, fresh weight). Extrusion decreased
blueberry anthocyanins by 90% and grape anthocyanins by Color
74% (Table 1). Storage for 3 mo did not change anthocyanin The color of white corn meal with corn-syrup extrudates
content in blueberry and grape extrudates. was high in lightness, low in red color, and high in yellow (Ta-
Color density, polymeric color, and % contribution of tan-
nin were used to indicate polymerization and browning reac-
tions. Nonextruded blueberry samples had higher color den-
sity and polymeric color than nonextruded grape samples
(Table 1). Extrusion decreased color density in blueberry
samples (78%) and grape samples (70%). The storage time
did not have any effect on color density. Extrusion decreased
polymeric color in blueberry samples but not in grape sam-
ples. Longer storage times increased polymeric color only in
blueberry samples. The interaction among types of samples,
the extrusion, and storage time had a significant effect on
polymeric color. Different mechanisms for pigment break-
Sensory and Nutritive Qualities of Food

down or differences in anthocyanin composition in the 2

types of fruit products may explain these changes.
For the contribution of tannin, type of fruit concentrate,
extrusion, and storage time increased the polymerization–
browning reaction of products. Blueberry concentrate pre-
sented higher polymerization than did grape-juice concen- Figure 1—Mean hedonic scores (1 = dislike extremely, 5 =
trate. The extrusion and storage time increased the neither like nor dislike, 9 = like extremely) for extruded
cereals evaluated by a consumer panel (N = 99). Means
polymerization and browning reaction in blueberry samples. within an attribute with different letters are significantly
The polymerization could have occurred due to the heat as different (Tukey’s HSD Test, p # 0.05).

440 JOURNAL OF FOOD SCIENCE—Vol. 67, Nr. 1, 2002

Anthocyanins in Cereals . . .

Table 3—Color analyses of extruded productsab sufficient amounts of the colorants remain to produce a pur-
Cereal sample L a b ple color. Extrusion conditions should be optimized to mini-
mize pigment loss, and adequate sugar should be added to
Corn syrup 78.50 6 0.78 a 1.92 6 0.09 c 14.91 6 0.25 a
reduce tartness. Other sources of these pigments may prove
Blueberry 49.26 6 2.39 c 8.82 6 0.94 a 7.65 6 0.86 b
Grape 60.04 6 0.85 b 5.87 6 0.14 b 7.95 6 0.13 b to be more stable during processing.
a Means 6 standard deviations of 2 runs, N=6.
b Different letters in the same column indicate significant difference at References
p # 0.05. AOAC. 1990. Official methods of analysis. 15th ed. Washington, D.C.: Association
of Official Analytical Chemists. p 931.
Camire ME. 2000. Bilberries and blueberries as functional foods & pharmaceu-
ticals. In: Mazza G, Ooomah BD, editors. Functional foods: herbs, botanicals
and teas. Lancaster, Pa.: Technomic Press. p 289-319.
Frankel EN, Bosanek CA, Meyer AS, Siliman K, Kirk LL. 1998. Commercial grape
ble 3). The addition of blueberry concentrate and grape-juice juices inhibit in vitro oxidation of human low-density lipoprotein. J Agric Food
concentrate had significant effects on lightness, redness, and Chem 46(3):834-838.
Fuleki T, Francis FJ. 1968. Quantitative methods for anthocyanins 2. Determina-
yellowness of cornmeal extrudates. Blueberry samples con- tion of total anthocyanin and degradation index for cranberry juice. J Food Sci
tained higher anthocyanins (4 mg/100 g, dry basis) and pre- 33:78-83.
sented darker and redder color than grape samples (2.6 mg/ Heinonen MI, Meyer AS, Frankel EN. 1998a. Antioxidant activity of berry phe-
nolics on human low-density lipoprotein and liposome oxidation. J Agric Food
100 g anthocyanins, dry basis). Both fruit decreased yellow- Chem 46(10):4107-4112.
ness. The metals from screw wear during extrusion may have Heinonen M, Rein D, Satué-Gracia MT, Huang SW, German JB, Frankel EN. 1998b.
Effect of protein on the antioxidant activity of phenolic compounds in a leci-
contributed to increased darkness due to anthocyanin-metal thin-liposome oxidation system. J Agric Food Chem 46(3):917-922.
complexation or browning reactions. Ilo S, Tomschik U, Berghofer E, Mundigler N. 1996. The effect of extrusion oper-
ating conditions on the apparent viscosity and properties of extrudates in
twin-screw extrusion cooking of maize grits. Lebensm-Wiss u Technol
Sensory Evaluation 29(7):593-598.
Figure 1 shows that the corn-syrup samples received the Koeppen BH, Basson DS. 1966. The anthocyanin pigments of barlinka grapes.
Phytochem 5:183-187.
highest overall, color, sweetness, hardness, and flavor ac- Peryam DR, Pilgrim FJ. 1957. Hedonic scale method of measuring food prefer-
ceptability scores. The addition of fruit-juice concentrate sig- ence. Food Technol 11(9):9-14.
nificantly affected overall acceptability and acceptability of Prior RL, Cao G, Martin A, Sofic E, McEwen J, O’Brien C, Lischner N, Ehlenfeldt
M, Kalt W, Krewer G, Mainland CM. 1998. Antioxidant capacity as influenced
color, sweetness, and flavor but not hardness acceptability. by total phenolic and anthocyanin content, maturity, and variety of Vaccinium
The moisture of extruded products could be possibly related species. J Agric Food Chem 46(7):2686-2693.
Satué-Gracia MT, Heinonen M, Frankel EN. 1997. Anthocyanins as antioxidants
to hardness because all products received the same score of on human low-density lipoprotein and lecithin-liposome systems. J Agric Food
hardness. Blueberry samples received the lowest score on Chem 45(9):3362-3367.
overall acceptability. Panelists gave grape and corn-syrup Somers TC. 1971. The polymeric nature of wine pigments. Phytochem 10:2175-
2186.
samples a similar overall acceptability score. Overall accept- Wrolstad RE. 1976. Color and pigment analyses in fruit products. Corvallis, Ore.:
ability was highly correlated with sweetness, hardness, and Oregon Agric Expt Station Bull 624.
MS 20001572, Submitted 4/16/01, Accepted 4/18/01, Received 8/1/01
flavor ( r $ 0.99). Although grape-juice concentrate and corn
syrup were sweet, their sweetness scores were only about 5 Maine Agricultural and Forest Experiment Station external publication #2472. Partial fund-
ing was provided by the Maine Agricultural and Forest Experiment Station and by a USDA
(neither like nor dislike). Blueberry concentrate provided Special Grant (#95-34221-1415). We thank the sensory panelists and the staff of the Dept. of
tartness that may have interfered with the sweetness of the Food Science & Human Nutrition and the Natick Soldier Center for their assistance. This
work was completed in partial fulfillment of the requirements for a M.S. degree for Ms.
product. Thus, if bright color is the major objective, blueber- Chaovanalikit.
ry should be used, but improvements in sweetness and flavor
Authors Camire and Dougherty are with the Dept. of Food Science & Hu-
are needed. man Nutrition, Univ. of Maine, 5736 Holmes Hall, Orono, ME 04469-5736.
Author Chaovanalikit is with the Dept. of Food Science & Technology, Or-
Conclusions egon State Univ., 100 Wiegand Hall, Corvallis, OR 97331-6602. Author Briggs
is with the Natick Soldier Center, Natick, MA 01760. Direct correspondence

B REAKFAST CEREALS APPEAR TO BE A SUITABLE FOOD FOR AN -

thocyanin fortification. Despite losses during extrusion,
to Dr. Camire (E-mail: camire@maine.edu).

Sensory and Nutritive Qualities of Food

Vol. 67, Nr. 1, 2002—JOURNAL OF FOOD SCIENCE 441