ARTICLE IN PRESS

JOURNAL OF
FOOD COMPOSITION
AND ANALYSIS
Journal of Food Composition and Analysis 21 (2008) 134–143
www.elsevier.com/locate/jfca

Original Article

Effects of various traditional processing methods on the

all-trans-b-carotene content of orange-fleshed sweet potato
A. Bengtssona,, A. Namutebib, M. Larsson Almingera, U. Svanberga
a
Department of Chemical and Biological Engineering, Food Science, Chalmers University of Technology, SE-412 96 Göteborg, Sweden
b
Department of Food Science and Technology, Makerere University, P.O. Box 7062, Kampala, Uganda
Received 30 March 2007; received in revised form 20 September 2007; accepted 21 September 2007

Abstract

The effects of traditional preparation methods and drying procedures on the provitamin A carotenoid content of orange-fleshed sweet
potato (OFSP) roots was determined by a high-performance liquid chromatography (HPLC) method. All-trans-b-carotene was the major
provitamin A carotenoid and the mean content of seven improved OFSP cultivars ranged from 108 to 315 mg/g dry matter. The retention
of all-trans-b-carotene was 78% when OFSP were boiled in water for 20 min. When OFSP were steamed for 30 min the retention was
77%, whereas deep-frying OFSP roots for 10 min resulted in retention levels of 78%. Drying slices of OFSP roots at 57 1C in a forced-air
oven for 10 h reduced the all-trans-b-carotene content by 12%. Solar drying and open-air sun drying OFSP slices to a moisture content of
p10% resulted in all-trans-b-carotene losses of 9% and 16%, respectively. The cis-isomer 13-cis-b-carotene was found in noticeable
amounts in all processed samples, but not in any raw samples. The formation of 13-cis-b-carotene correlated with the original amount of
all-trans-b-carotene found in the raw OFSP root. The high content of all-trans-b-carotene in the investigated improved OFSP varieties
and the moderately low losses due to degradation and isomerization renders OFSP a suitable food source of provitamin A.
r 2007 Elsevier Inc. All rights reserved.

Keywords: Orange-fleshed sweet potato; Ipomoea batatas; b-Carotene; Food processing; Drying; Retention; Vitamin A deficiency (VAD); Bio-fortified
staple foods

1. Introduction use of bio-fortified staple foods (Welch, 2002). However,

The prevalence of vitamin A deficiency (VAD) is high in
sub-Saharan Africa and particularly so in Uganda (Aguayo
and Baker, 2005). In 2001, a Ugandan Demographic and
Health Survey (UDHS, 2001) recorded the prevalence of
VAD to 28% for 0–59-month-old children and to 52% for
15–50-year-old women. VAD is thus one of the major
micronutrient deficiencies in the country. A number of
different intervention strategies to address VAD are being
promoted in developing countries. These include fortifica-
tion (Dary and Mora, 2002), supplementation (Beaton et
al., 1993), diet diversification (Gibson and Hotz, 2001) and
Corresponding author. Tel.: +46 31 772 38 22; fax: +46 31 772 38 30.
E-mail addresses: anton.bengtsson@chalmers.se (A. Bengtsson),
asnamutebi@agric.mak.ac.ug (A. Namutebi),
marie.alminger@chalmers.se (M.L. Alminger),
ulf.svanberg@chalmers.se (U. Svanberg).
an inadequate health infrastructure and limited financial
resources of many poor rural households calls for an
alternative strategy to supplementation and fortification
(Hagenimana and Low, 2000). The use of bio-fortified
staple foods (i.e. varieties bred for increased mineral and
vitamin content) has been justified as a sustainable food-
based approach to reach a large section of the rural
population who may not be reached by other intervention
strategies (FAO/ILSI, 1997).
Orange-fleshed sweet potato (OFSP) is among the bio-
fortified staples bred for high provitamin A carotenoid
content (CIP, 2006). Sweet potato (Ipomoea batatas) is a
major food crop in developing countries (Woolfe, 1992),
and it is mainly consumed as boiled roots. Sweet potato is
also commonly processed into dried slices and flour to
preserve the roots for household use during off-season.
OFSP may have the potential to prevent and combat VAD,
as was indicated by a South African efficacy study of

0889-1575/$ - see front matter r 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jfca.2007.09.006
 ARTICLE IN PRESS
A. Bengtsson et al. / Journal of Food Composition and Analysis 21 (2008) 134–143
school-aged children that consumption of boiled and
mashed OFSP improved their vitamin A status (van
Jaarsveld et al., 2005). In order for communities to benefit
from the high provitamin A carotenoid content of OFSP,
preparation and processing methods should be optimized
since these steps influence the retention levels of provitamin
A carotenoids (Rodriguez-Amaya, 1997). Promotion of
OFSP to farmers in Uganda has started and therefore
knowledge about the retention levels of b-carotene in the
processed sweet potato products is of great concern.
Provitamin A carotenoids are degraded by heat treat-
ment and exposure to light during processing and during
prolonged storage, but there is a lack of information
regarding the retention of provitamin A carotenoids from
traditional processing methods. Also, there is concern over
conflicting values for carotenoid retention due to the
intricate nature of the carotenoid analysis (Rodriguez-
Amaya, 1997). The following study was undertaken to
acquire knowledge about the retention of provitamin A
carotenoids in boiled, steamed, deep-fried, and dried OFSP
roots. In addition, the aim was to assess the natural
variability of b-carotene content in improved OFSP
cultivars and to measure the formation of cis-isomers
during processing and drying. An additional objective was
to evaluate color measurements of OFSP flour as a rapid
screening method for the b-carotene content of fresh OFSP
roots.
2. Materials and methods
2.1. Plant material
OFSP (I. batatas) root samples, which are improved
sweet potato cultivars under field tests, were harvested 4.5
months after planting from Namulonge Agricultural and
Animal Production Research Institute (NAARI), Uganda.
The variety Ejumula was harvested 4 months after planting
from a farmer site in the vicinity of NAARI. Table 1 shows
the OFSP cultivars (one commercially released variety
Ejumula and six field test varieties) used for the present
study.
Table 1
Orange-fleshed sweet potato cultivars procured from Namulonge Agri-
cultural and Animal Production Research Institute, Uganda
Sweet potato variety
Ejumula
SPK004
SPK004/6/6a
SPK004/6a
SPK004/1/1a
SPK004/1a
Sowola 6/94/9b
a
Advanced yield trial.
b
Intermediate yield trial.
135
2.2. Chemicals and standards
All chemicals were obtained from Sigma–Aldrich (Stock-
holm, Sweden) or Fischer Scientific GTF (Göteborg,
Sweden). The water used for extraction and high-perfor-
mance liquid chromatography (HPLC) analysis was
generated by Millipore Milli-Q plus ultra-pure water
system (Millipore, Solna, Sweden). All-trans-b-carotene
standard (synthetic, crystalline, Type II, product C-4582)
was purchased from Sigma–Aldrich (Stockholm, Sweden).
2.3. Preparation of samples
Four roots from each OFSP cultivar were randomly
selected for each preparation method. Each root was
washed, peeled and quartered longitudinally (from the stem
end to the root end). Two opposite quarters were combined
and kept frozen as a reference sample, while the remaining
two quarters were prepared either by boiling, steaming or
deep-frying. The weight was recorded before and after
processing. Sweet potato samples were prepared as ready-
to-eat products. Therefore, the processing times varied for
the different methods. Sweet potatoes used for boiling were
immersed in tap water contained in an open aluminum pot
and boiled for 20 min. Sweet potatoes used for steaming
were wrapped in banana leaves and placed on top of cut
banana leaf stem ends to serve as a separator from the
added tap water. The sweet potatoes were then steamed at
93 1C for 30 min in an aluminum pot with the lid on. Sweet
potatoes used for deep-frying were immersed in locally
produced sunflower oil with a temperature between 160
and 170 1C and deep-fried for 10 min.
2.3.1. Drying
Sweet potato roots (variety Ejumula) used for drying
experiments were sliced to 1–2 mm thickness using a food
processor (Braun Combi Max K600). A portion of
approximately 50 g, taken as a reference sample, was
placed in an amber polystyrene bottle, screw capped and
stored in a freezer (–20 1C) prior to subsequent carotenoid
analysis. The remaining slices (450 g for each method)
were used for drying. The procedure was repeated for four
consecutive days. Sliced sweet potato for oven drying was
spread out on aluminum foil lined trays, which were placed
in an HT 4 forced-air cabinet oven dryer (Innotech,
Altdorf, Germany) and dried at 57 1C for 10 h to a brittle
texture. Sliced sweet potato for solar drying was spread out
on 0.97 m  0.77 m-sized plastic-meshed trays and placed in
a tunnel solar dryer. The base of the solar dryer was lined
with a black high-gauge polyethylene sheet. Drying
temperatures varied between 45 and 63 1C in the solar
dryer. Sliced sweet potato for open-air sun drying was
spread out on a transparent low-gauge polyethylene sheet
placed on a papyrus mat. Sliced sweet potato was dried
under direct sunlight and occasionally turned to improve
the drying process. Drying temperatures varied between 30
and 52 1C, measured directly above the sweet potato slices.
 ARTICLE IN PRESS
136 A. Bengtsson et al. / Journal of Food Composition and Analysis 21 (2008) 134–143
Solar and open-air sun-dried slices were dried to a moisture
content of p10%. The time of drying varied between 6 and
10 h depending on weather conditions.
2.3.2. Treatment after processing
Boiled, steamed and deep-fried samples were drained on
absorbent paper, cooled for 15 min at room temperature,
packaged in thick polyethylene bags, and frozen at –20 1C.
All dried slices were packaged and sealed in similar
polyethylene bags and immediately frozen at –20 1C. All
samples were transported to Sweden under frozen condi-
tions in cool boxes of expanded polystyrene. Both
processed and reference samples were freeze-dried in a
Heto DW6-55 Dry winner (Ninolab, Upplands Väsby,
Sweden) before determination of carotenoid content. To
ensure that the freeze-drying of the OFSP samples did not
affect the b-carotene content this pre-treatment step was
evaluated prior to the study. No significant difference was
found in all-trans-b-carotene content of seven OFSP
samples analyzed before and after freeze-drying. The mean
difference was less than 2.7% between non-freeze-dried
and freeze-dried samples. The freeze-dried samples were
overlaid with nitrogen prior to storage at –20 1C until
carotenoid analysis in order to avoid degradation of b-
carotene due to presence of oxygen. All analyses were
performed within 7 months on receipt of fresh sweet potato
roots. Prior to the carotenoid measurements all freeze-
dried samples were homogenized into a finely ground flour
(particle size 0.2 mm) with a coffee grinder (Braun
CaféSelect KMM 30).
2.3.3. Dry matter determination
Dry matter (DM) content was determined in the fresh
samples by measuring sample weights before and after
freeze-drying. Solids content of the boiling water was also
measured by drying in an oven at 105 1C until constant
weight.
2.3.4. Extraction of fat
Determination of oil content in the deep-fried samples
was done according to the method of Lee et al. (1996) with
chloroform and methanol (1:1, v/v) as the extraction
solvent. The mean fat content of 12 samples was 5.5% with
a standard deviation of 0.9%.
2.4. Sweet potato flour L*a*b* color value determination
Color values of the fresh samples were determined using
a Konica Minolta CR-400 chroma meter (Konica Minolta
Sensing Inc., Osaka, Japan). The equipment was calibrated
before each measurement using a calibration plate. Color
measurements were performed on sweet potato flour and
the values were averaged from four consecutive readings.
The color was described based on the values of L*, a* and
b*, where L* is a measure of lightness, a* defines
components on the red–green axis, and b* defines
components on the yellow-blue axis. The a* parameter
was used in the comparison with the all-trans-b-carotene
HPLC measurements of the fresh OFSP samples.
2.5. Preparation of all-trans-b-carotene standard
A stock solution of all-trans-b-carotene in n-hexane
(100 mg/mL) was prepared and stored at 20 1C. The purity
of the standard was verified by HPLC and photodiode
array detection. Concentration of the standard solution
was calculated using the absorbance at 450 nm against n-
hexane as blank by spectrophotometric measurement in a
Cary 50 Series UV–visible spectrophotometer (Varian
Australia Pty Ltd., Mulgrave, Australia) and with the
molar absorption coefficient of 2592 for all-trans-b-
carotene. A standard curve was constructed with eight
different concentrations for all-trans-b-carotene in dupli-
cate. The detector’s response showed good linearity and
reproducibility. One-point calibration on each day of
analysis was performed, verifying any change in the
detector’s response.
2.6. Carotenoid extraction
Finely ground flour samples (0.2 g) in duplicate were
reconstituted with 1 mL of ultra pure water in a test tube
for 20 min followed by addition of 2 mL acetone containing
0.1% (w/v) butylated hydroxytoluene. The test tube was
vortexed and then centrifuged at 4750 g for 3 min. The
resulting supernatant was saved in a new test tube. The
residue was re-extracted with 2 mL acetone and centrifuged
again. This was repeated up to four times until the residue
was colorless. To the resulting acetone extract 3 mL
petroleum ether was added together with 5 mL ultra pure
water to aid in the separation of two phases. The organic
and water phases were separated by centrifugation at 4750g
for 4 min and the organic phase was transferred to a new
tube. This step was repeated once. The pooled organic
phases were evaporated in a heating block at 35 1C under a
stream of nitrogen. The residue was dissolved in 5 mL
(processed samples) or 10 mL (raw samples) mobile phase
(methanol:methyl tert-butyl ether (1:1, v/v)). Precautionary
measures such as exclusion of oxygen, protection from
light, avoiding temperatures above 35 1C, avoiding contact
with acids and use of high-purity solvents were taken to
prevent carotenoid losses during analysis (Schiedt and
Liaaen-Jensen, 1995).
2.7. HPLC analysis of carotenoids
Carotenoids were analyzed by reversed phase HPLC
using a Waters 600 system equipped with auto sampler
injector, degasser, pump and a Waters 996 UV–visible
photodiode array detector operating at 450 nm. The data
were stored and processed by means of Millennium 4.00
Software (Waters, Stockholm, Sweden). Absorption spec-
tra were recorded between 250 and 500 nm. Separations
were carried out on a C30 carotenoid column (5 mm,
 ARTICLE IN PRESS
A. Bengtsson et al. / Journal of Food Composition and Analysis 21 (2008) 134–143
250 mm  4.6 mm i.d.) (YMC Europe GMBH, Scherm-
beck, Germany). The mobile phase used for isocratic
elution consisted of methanol:methyl tert-butyl ether:water
(55:41:4, v/v/v). The flow rate was 1 mL/min and the
injection volume 20 mL.
2.8. Carotenoid identification and quantification
Carotenoids were identified using retention times and
UV/vis absorption spectra and the absorption spectra were
compared with literature data. Quantification of all-trans-
b-carotene was achieved using a calibration curve with
eight concentrations and with the correlation coefficient of
0.995. Quantification of cis-isomers was achieved using
relative response factors. Response factors used were 0.806
for 13-cis-b-carotene and 0.702 for 15-cis-b-carotene
(Mulokozi and Svanberg, 2003). The analysis of carote-
noids was carried out on duplicate samples from each root,
and the coefficient of variation (CV) for the analysis was
below 3%. The concentration of each carotenoid was
expressed as micrograms per milliliter, given as the mean of
duplicate extractions.
2.9. Method validation
Our analytical method was validated in an interlabora-
tory proficiency study that included analysis of a lyophi-
lized sweet potato test material that was distributed to 15
laboratories. Both total b-carotene and all-trans-b-carotene
were analyzed in three replicate samples and our results
(with a CV of 3%) did not differ significantly from the
values reported by the reference laboratory that used the
HPLC method described for sweet potato in HarvestPlus
handbook for carotenoid analysis (Rodriguez-Amaya and
Kimura, 2004).
2.10. Calculation of carotenoid retention
Retention of all-trans-b-carotene was calculated based
on the following equation:
Table 2
The contents of dry matter, all-trans-b-carotene, a* color value, and vitamin A activity in seven different improved OFSP cultivars
Cultivara Dry matter contentb
(%)
SPK004/1 34.672.6
SPK004/6 30.772.3
SPK004 35.072.4
Sowola 6/94/9 35.070.8
SPK004/1/1 33.571.8
SPK004/6/6 30.372.8
Ejumula 34.671.6
a
Duplicate analysis of each sweet potato sample (n ¼ 12) for each cultivar except cultivar Sowola 6/94/9 (n ¼ 4).
b
Mean7standard deviation.
c
Mean (lowest and highest value shown in parentheses).
d
Values in the same column not sharing the same letters (a–d) are significantly different (Po0.05) using ANOVA and Tukey’s HSD multiple ranks test.
e
Conversion factor 1 mg RAE ¼ 12 mg all-trans-b-carotene.
137
% Retention
carotenoid content per g of cooked food ðdry basisÞ
¼  100.
carotenoid content per g of fresh food ðdry basisÞ
As each OFSP root was quartered, the retention value
calculated for each preparation method was obtained by
comparing the processed value with its own reference
sample. Normally, the retention is overestimated when it is
calculated on a dry weight basis (Rodriguez-Amaya, 1997).
However, in the present study changes in DM due to loss of
soluble solids, loss or uptake of water, and uptake of fat in
the deep-fried samples were accounted for. As a result the
calculated retention values can be regarded as close to true
retention.
2.11. Statistical analysis
The statistical comparison of data was performed by
ANOVA using SPSS software to reveal significant differ-
ences among the cultivars and processing methods studied.
The correlation coefficients and their probability levels
were obtained from linear regression analysis. Determina-
tion of significance of differences among processing
methods was obtained by Tukey’s HSD multiple rank test.
P values ofo0.05 were considered significant.
3. Results
The DM content of the storage roots was high with
mean values ranging from 30.3% to 35.0% for the different
varieties. Cultivars SPK004, SPK004/1, Sowola 6/94/9 and
Ejumula had the highest DM content while cultivars
SPK004/6 and SPK004/6/6 had the lowest (Table 2).
The mean all-trans-b-carotene content ranged from
108.1 to 314.5 mg/g DM. Cultivar SPK004/6 had the
highest content whereas the lowest content was found in
SPK004 and SPK004/1 (Table 2). The natural variability in
the all-trans-b-carotene content was considerable among
the 12 root samples from each cultivar as can be seen in the
large range in Table 2. All-trans-b-carotene comprised the
All-trans-b-carotenec,d a* color valueb Vitamin A activitye
(mg/g DM) (mg RAE/100 g FW)
108.1 a (44.3–192.7) 6.4172.4 311
314.5 d (206.3–460.3) 13.7472.0 804
116.3 a (48.2–191.7) 6.4172.1 339
212.0 b, c (150.6–251.1) 12.3571.3 619
143.6 a, b (85.6–219.3) 8.3971.8 400
246.2 c (97.9–363.8) 12.1472.6 622
261.9 c, d (185.6–324.8) 12.3171.0 755

138 A. Bengtsson et al. / Journal of Food Composition and Analysis 21 (2008) 134–143
500
450
400
β-carotene (μg/g DM)
350
300
250
200
150
100
50
0
0 2
Fig. 1. Correlation between all-trans-b-carotene content and a* color value of individual samples from seven OFSP varieties. Each point represents one
sample (n ¼ 76).
major provitamin A carotenoid found in the OFSP roots.
Neither a-carotene or b-cryptoxanthin were found in
detectable amounts in any of the samples.
The flesh color of the varieties varied from yellow or
light orange to deep orange. The a* color value was
measured on sweet potato flour and ranged from 6.41 for
cultivars SPK004 and SPK004/1 to 13.74 for cultivar
SPK004/6 (Table 2). This is consistent with the all-trans-b-
carotene content for each individual sample of seven
cultivars measured by HPLC as shown in Fig. 1 by a
significant relationship (r ¼ 0.96, Po0.001).
The retention of the all-trans-b-carotene in processed
samples of OFSP is affected by losses due to degradation,
trans-cis-isomerization and potential leakage. All three
processing methods caused a significant reduction in the
all-trans-b-carotene content of the sweet potato roots
(Po0.001). However, there were no significant differences
between the effects of boiling, steaming and deep-frying.
The retention of all-trans-b-carotene varied between 70%
and 81% for the boiled samples, between 69% and 81% for
the steamed roots, and between 76% and 80% for the deep-
fried samples (Table 3). Of the boiled samples, the retention
in SPK004/6/6 was significantly higher than in SPK004
(Po0.05), and of the steamed samples the retention in
SPK004/6 was close to significantly higher (Po0.07) than
in SPK004/1 and SPK004. There was no correlation
between percentage retention and all-trans-b-carotene con-
tent of the fresh roots of OFSP (data not shown).
The isomer 13-cis-b-carotene was found in all processed
samples. The amount of 13-cis-b-carotene ranged from 4.7
to 18.1 mg/g DM in the boiled sweet potato samples, the
steamed roots contained between 4.5 and 15.8 mg/g DM 13-
cis-b-carotene, while the 13-cis-b-carotene content in the
deep-fried samples ranged from 3.6 to 23.5 mg/g DM (Table
3). Fig. 2 shows the correlation between all-trans-b-
carotene in the fresh and 13-cis-b-carotene in the processed
samples (r ¼ 0.88, Po0.001). The amount of 13-cis-b-
carotene formed during processing is clearly related to the
ARTICLE IN PRESS
y = 25.4 x - 55.6
R = 0.96
4 6 8 10 12 14 16 18
a∗ color value
amount of all-trans-b-carotene in the fresh sweet potato
root. The 15-cis-b-carotene was identified in a few
processed samples but the amounts were small. The
maximum amount of 15-cis-b-carotene was 4.0 mg/g DM
in steamed and deep-fried roots of cultivar SPK004/6. The
fraction of 13-cis-b-carotene to total b-carotene varied
from 5.7% to 8.8% in the boiled samples, 4.0–10.3% in the
steamed roots, and 3.0–8.5% in the deep-fried samples,
respectively.
Drying slices of Ejumula variety significantly (Po0.05)
reduced the concentration of all-trans-b-carotene (Table 4).
The reduction of the all-trans-b-carotene content was 12%
in oven-dried samples and 9% in solar-dried samples.
Open-air sun drying reduced the all-trans-b-carotene
content with 16%, however not significantly different from
the two other methods. Only small amounts of 13-cis-b-
carotene were found in the dried samples (Table 4). The
moisture content was 10.8%, 6.8% and 10.0% for oven-
dried, solar-dried and open-air sun-dried sweet potato
slices, respectively.
4. Discussion
The high DM content observed for all cultivars was in
accordance with the expected DM content above 30%. The
values found in the present study were higher than
previously reported values for six orange-fleshed varieties
(Hagenimana et al., 1999). Additional breeding targets to
high provitamin A carotenoid content are good resistance
against sweet potato virus and high DM content as desired
in sub-Saharan Africa. Although cultivars SPK004/6 and
SPK004/6/6 had the lowest DM content (Table 2), on
average 30%, this still complies with the breeding target of
more than 30%.
Freeze-drying has been shown to cause no degradation
or isomerization of b-carotene in carrots (Regier et al.,
2005; Sweeney and Marsh, 1971). This is as expected since
oxygen is excluded during the process (Desobry et al.,
 ARTICLE IN PRESS
A. Bengtsson et al. / Journal of Food Composition and Analysis 21 (2008) 134–143 139

Table 3
The contents of all-trans-b-carotene and 13-cis-b-carotene, and the retention of all-trans-b-carotene in seven different OFSP cultivars after boiling for
20 min, steaming for 30 min, and deep-frying for 10 min

Cultivara/preparation method All-trans-b-caroteneb (mg/g DM) 13-cis-b-carotene Retention of

all-trans-b-carotene (%)d
c
(mg/g DM) (% of total b-carotene)

SPK004/1
Fresh 108.1 (44.3–192.7)
Boiled 73.0 (41.5–83.9) 4.770.5 6.1 78.1
Steamed 73.7 (27.7–119.7) 4.572.7 5.8 72.0
Deep-fried 103.1 (59.4–161.0) 9.072.9 8.0 80.4
SPK004/6
Fresh 314.5 (206.3–460.3)
Boiled 252.6 (223.0–275.4) 18.171.8 6.8 79.5
Steamed 249.9 (161.0–404.4) 15.876.5 5.9 84.4
Deep-fried 253.1 (194.9–320.3) 23.572.9 8.5 76.7
SPK004
Fresh 116.3 (48.2–191.7)
Boiled 90.2 (71.8–109.6) 8.771.8 8.8 70.1e
Steamed 59.3 (33.3–89.4) 6.872.1 10.3 68.6
Deep-fried 106.2 (60.1–158.7) 6.572.1 5.8 79.4
Sowola 6/94/9
Fresh 212.0 (150.6–251.1)
Boiled 160.6 (114.3–199.9) 12.172.5 7.0 75.8

SPK004/1/1
Fresh 143.6 (85.6–219.3)
Boiled 107.1 (72.7–129.1) 6.472.0 5.7 79.5
Steamed 116.3 (99.7–132.9) 6.671.3 5.4 79.8
Deep-fried 114.3 (70.8–191.8) 3.670.9 3.0 76.1

SPK004/6/6
Fresh 246.2 (97.9–363.8)
Boiled 201.3 (152.2–224.3) 15.972.6 7.3 81.2e
Steamed 215.7 (126.4–266.1) 15.776.9 6.8 77.4
Deep-fried 168.9 (80.8–258.2) 10.674.1 5.9 79.7
Ejumula
Fresh 261.9 (185.6–324.8)
Boiled 199.4 (159.2–229.4) 12.472.4 5.8 77.9
Steamed 213.5 (201.7–222.9) 8.971.5 4.0 80.5
Deep-fried 210.3 (185.6–244.7) 11.571.8 5.2 79.5
Mean valuef
Boiled 11.2 6.7 77.6
Steamed 9.7 5.9 77.0
Deep-fried 10.8 6.3 78.3
a
Duplicate analysis of each sweet potato sample (n ¼ 4) for each cultivar and preparation method.
b
Mean (lowest and highest value shown in parentheses).
c
Mean7standard deviation.
d
Retention values calculated for each preparation method were obtained by comparing the processed value with its own reference value.
e
Retention values are significantly different (Po0.05) using ANOVA and Tukey’s HSD multiple ranks test.
f
Mean values calculated based on 28 samples for boiling and 24 samples for steaming and deep-frying.

1998). In a subset of samples we found no degradation of
all-trans-b-carotene due to freeze-drying. During storage at
–20 1C in the presence of oxygen degradation of b-carotene
may occur. Therefore, the samples in the present study
were stored under nitrogen. No losses of all-trans-b-
carotene were found in the freeze-dried OFSP samples in
the present study during storage at –20 1C. After repeated
analysis of a set of samples after 3 months the difference
was less than 1.3%.
4.1. Variability in all-trans-b-carotene content of OFSP
The mean total all-trans-b-carotene content was found to
vary between 108.1 and 314.5 mg/g DM in the cultivars
analyzed in the present study. These values correspond well
with previously reported values using HPLC ranging from
less than 1 mg/g fresh weight (FW) in white-fleshed varieties
to 265 mg/g FW in orange-colored varieties (Huang et al.,
1999; K’osambo et al., 1998; Kidmose et al., 2006; Kimura
 ARTICLE IN PRESS
140 A. Bengtsson et al. / Journal of Food Composition and Analysis 21 (2008) 134–143

500
450
y = 14.0 x + 50.5
400 R = 0.88

β-carotene (μg/g DM)

350
300
250
200
150
100
50
0
0 5 10 15 20 25 30
13-cis-β-carotene (μg/g DM)

Fig. 2. Correlation between all-trans-b-carotene in fresh and 13-cis-b-carotene in processed OFSP samples (n ¼ 76).

Table 4
The contents of all-trans-b-carotene and 13-cis-b-carotene, and the retention of all-trans-b-carotene in Ejumula OFSP variety after three different drying
proceduresa

OFSP sample Drying temperature All-trans-b-caroteneb 13-cis-b-carotene Retention of all-trans-

(1C) (mg/g DM) b-carotene (%)
(mg/g DM)b (% of total b-
carotene)

Fresh 310.8716.3
Oven-dried 57 274.3722.8 3.570.3 1.3 88.2
Fresh 290.574.3
Solar-dried 45/63c 264.677.6 2.071.3 0.7 91.1
Fresh 301.7718.3
Open-air sun-dried 30/52c 252.7715.0 0.970.2 0.4 83.8
a
Duplicate analysis of each sweet potato sample; each drying method was repeated four times.
b
Mean7standard deviation.
c
Lowest/highest temperature during drying.

et al., 2007; Takahata et al., 1993; van Jaarsveld et al.,
2006). There were large differences in the all-trans-b-
carotene content of sweet potato roots from the same
cultivar. Similar findings have been reported by van
Jaarsveld et al. (2006). Analyzing the OFSP variety
Resisto, they observed a variation in the b-carotene content
between 132 and 194 mg/g FW for medium-sized sweet
potatoes from the same harvest batch. The fact that sweet
potatoes of not exactly the same size were analyzed in the
present study may explain some of the large variation seen
within the same cultivar.
4.2. Provitamin A activity of OFSP
In 2001, the US Institute of Medicine (IOM, 2001)
derived new conversion factors for estimating the amount
of retinol activity equivalents (RAE) obtained from
provitamin A carotenoids in foods, stating that 12 mg of
all-trans-b-carotene is equivalent to 1 mg RAE. This value is
twice as high as the previously recommended factor (FAO/
WHO, 1988). Thus, the vitamin A value calculated in RAE
of the OFSP cultivars in the present study ranged from 311
to 804 mg RAE/100 g FW (Table 2). With an average
retention of about 78% after cooking, these OFSP cultivars
will provide about 243–627 mg RAE/100 g ready-to-eat
portion. Hence, a medium-sized sweet potato root in the
present study (mean weight 177 g) could more than well
cover the safe daily recommended vitamin A intake level of
400 mg RE for pre-school children (FAO/WHO, 1988),
even after processing. Moreover, in a recent intervention
study in South Africa (van Jaarsveld et al., 2005), the
vitamin A status in school children was significantly
improved when the children were served a daily meal of
cooked OFSP that provided 830 RAE/100 g cooked root.
4.3. Correlation between flour color and all-trans-b-carotene
content of OFSP
The flour color of the sweet potato cultivars was
measured with a chroma meter and the mean a* color
 ARTICLE IN PRESS
A. Bengtsson et al. / Journal of Food Composition and Analysis 21 (2008) 134–143
values ranged from 6.41 to 13.74 based on 12 root samples
for each cultivar. These values were found to highly
correlate (r ¼ 0.96, Po0.001) with the amount all-trans-b-
carotene found in the fresh sweet potato roots. This
supports earlier findings that color measurements of sweet
potato flours could be used as a screening tool to estimate
the total b-carotene content in fresh OFSP roots. Takahata
et al. (1993) reported a correlation coefficient of 0.90
between color value a* and the b-carotene content in
orange-fleshed varieties, which is in accordance with the
results in this study. In white-fleshed varieties the b* color
value has been reported as the best measure of correlation
between color variable and b-carotene content as shown by
a correlation coefficient of 0.74 (Ameny and Wilson, 1997).
Thus, it is less reliable to estimate the b-carotene content
from color measurements of pale cultivars.
4.4. Effect of degradation and isomerization on the all-trans-
b-carotene content of OFSP
Factors such as variety, growing conditions, stage of
maturity, harvesting and post-harvest handling, processing
and storage of OFSP may have a large influence on the b-
carotene content (Rodriguez-Amaya, 2000). In addition,
the analysis of fresh and processed samples may introduce
further variability of the results due to difficulties in
achieving a complete extraction of carotenoids. While
processing softens the cell wall and facilitates carotenoid
extraction, incorporation of oil and formation of degrada-
tion products may have the opposite effect (de Sá and
Rodriguez-Amaya, 2004). However, the extraction of
carotenoids from both fresh and processed freeze-dried
samples in the present study was repeated until no color
was noticed in the acetone extract. Retention values could
be overestimated due to the fact that the b-carotene content
in fresh samples is decreased as a result of enzymatic
oxidation. However, van Jaarsveld et al. (2006) observed
no loss of b-carotene in chopped raw OFSP that were
allowed to stand for 4 h, indicating that enzymatic
oxidation is not a major issue in sweet potato. Although
processing reduces the carotenoid content, heat processing
has the potential to enhance the bioavailability of
carotenoids in cooked vegetables (van het Hof et al.,
1998), which may compensate for the loss.
Structural alterations of carotenoids during processing
are attributed to geometrical isomerization and enzymatic
or non-enzymatic oxidation (Rodriguez-Amaya, 2002).
These changes result in reduced provitamin A activity of
the sweet potato. Chandler and Schwartz (1988) investi-
gated isomerization and degradation of all-trans-b-caro-
tene in sweet potato during different processing treatments.
Heat processing induced the formation of 13-cis-b-caro-
tene, and the quantity formed was related to the severity
and length of the heat treatment. The amount of 13-cis-b-
carotene found in the processed samples ranged from 25.2
to 101 mg/g DM, corresponding to 5.2–28.9% of the total
b-carotene content. However, the raw samples were also
141
reported to contain the 13-cis isomer. This was explained
by a 9 months long storage period employed for the sweet
potatoes prior to the initiation of the study. Kidmose et al.
(2006) reported similar findings with significantly higher
content of cis-isomers when frying 3 min compared with
1 min, and besides formation of 13-cis-b-carotene, 15-cis-b-
carotene and 9-cis b-carotene were also detected. In the
present study 13-cis-b-carotene was formed during proces-
sing in amounts ranging from 3.6 to 23.5 mg/g DM. The
fraction of 13-cis-b-carotene to the total b-carotene content
varied between 3.0% and 10.3%. Boiling generally resulted
in the largest fraction of 13-cis-b-carotene to the total
b-carotene content. The amount of 13-cis-b-carotene was
positively correlated to the amount of all-trans-b-carotene
found in the fresh roots (r ¼ 0.88, Po0.001). Furthermore,
15-cis-b-carotene was found in small quantities in a few of
the processed samples.
4.5. Retention of all-trans-b-carotene in processed OFSP
It is very important to analyze paired samples (i.e.
comparable raw and cooked samples) in retention studies
to be able to preclude differences due to non-uniform
distribution of b-carotene in the sweet potato root
(Rodriguez-Amaya, 1997). Consequently, by using paired
samples sampling errors could be avoided. Retention is
defined as the proportion of carotenoids remaining in the
processed sweet potato root in relation to the amount of
carotenoids originally present in the sweet potato. Loss of
solids during processing may lead to an overestimation of
the retention values during processing if not accounted for
(Ogunlesi and Lee, 1979; Rodriguez-Amaya, 1997). In the
present study, the retention of all-trans-b-carotene in
processed and dried OFSP was calculated based on dry
weights of the sweet potato samples, and corrected for
changes in DM due to loss of soluble solids, loss or uptake
of water, and uptake of fat in the deep-fried samples.
Hence, these calculations are comparable to the expression
of true retention as suggested by Murphy et al. (1975).
Moreover, carotenoid measurements on a dry weight basis
allow for easier comparisons among different laboratories.
The mean retention of all-trans-b-carotene was found to be
77.6% in boiled, 77.0% in steamed, and 78.3% in deep-
fried OFSP. This is in contrast to the common idea that
deep-frying affects the b-carotene content to a larger extent
than steaming and boiling (Rodriguez-Amaya, 1997). One
explanation for the similar results obtained for the
retention values could be that the OFSP roots were
prepared as ready-to-eat products, hence different pre-
paration times were applied (20 min for boiling, 30 min for
steaming and 10 min for deep-frying). Using the same
preparation time might lead to a higher retention for the
steamed samples compared to the samples prepared by the
other processing methods. Van Jaarsveld et al. (2006)
reported a true retention of b-carotene ranging from 83%
to 92% in boiled OFSP under different conditions. The
highest retention was noticed when boiling for 20 min in a
 ARTICLE IN PRESS
142 A. Bengtsson et al. / Journal of Food Composition and Analysis 21 (2008) 134–143
pot with the lid on, whereas longer boiling times resulted in
a true retention ranging from 70% to 80%. Kidmose et al.
(2006) did not find any significant difference in true
retention of all-trans-b-carotene between varying frying
times. However, the frying times used varied only from 1 to
3 min for sweet potato, and they proposed that longer
frying times may result in lower retentions. K’osambo et al.
(1998) reported that boiling for 30 min reduced the total
carotenoid content by 14–59% in four different cultivars,
but no consideration was made regarding weight changes
due to loss or gain of water and soluble solids.
4.6. Retention of all-trans-b-carotene in dried OFSP
Solar drying of vegetables and fruits is an appropriate
technique to prolong shelf-life and improve quality of plant
foods for consumption during off-season periods (Ruel and
Levin, 2000). With solar drying, higher air temperatures
and lower humidity can be achieved compared with open-
air sun drying, thereby shortening the drying time. This
may result in larger b-carotene retention in solar-dried
compared with open-air sun-dried vegetables. The three
different drying methods evaluated did not bring about any
substantial decrease in the all-trans-b-carotene content.
Sweet potato slices, which were dried in a solar tunnel drier
or forced-air oven retained more all-trans-b-carotene than
open-air sun-dried sweet potato slices, but the difference
was not significant. The slight difference observed could be
attributed to the direct exposure to sunlight in the open-air
sun-dried samples as the drying time on average was the
same as for the solar-dried samples. Still, the retention of
all-trans-b-carotene was high with a mean value of 84% for
the open-air sun-dried samples. Isomerization was not
substantial in any of the dried samples and lower compared
with the processed samples. The time of drying seems to be
an important factor for the degradation of b-carotene, and
indeed, Mdziniso et al. (2006) showed that a higher
dehydration rate affected the b-carotene content in solar-
dried OFSP to a lesser extent, resulting in higher retention
values. They reported b-carotene losses of 4.0–5.8% in
solar-dried sweet potatoes when drying slices to moisture
content of 8–10%.
5. Conclusions
To conclude, this paper showed that the retention of all-
trans-b-carotene in OFSP after boiling, steaming and deep-
frying is similar in prepared, ready-to-eat products. Both
solar and oven drying can be considered as appropriate
drying methods, resulting in high retention values. The
retention of all-trans-b-carotene in OFSP after open-air
sun drying was somewhat lower due to the exposure to
direct sunlight. The retention was substantial, both after
processing and drying, which means that OFSP with high
b-carotene content can be considered as an excellent source
of provitamin A. Of interest for further study is to
determine the bioavailability of all-trans-b-carotene in
these improved cultivars, particularly SPK004/6, SPK004/
6/6 and Ejumula, which contained significantly higher
amounts of all-trans-b-carotene among the cultivars
investigated.
Acknowledgments
We thank Dr. Robert Mwanga at NAARI, Uganda for
providing the sweet potato roots, Emmauel Okalanyi for
assistance during collection and processing of the sweet
potatoes, and The Swedish Agency for Research Coopera-
tion in Developing Countries (SAREC) (Grant SWE-2004-
005) for funding.
References
Aguayo, V.M., Baker, S.K., 2005. Vitamin A deficiency and child survival
in sub-Saharan Africa: A reappraisal of challenges and opportunities.
Food and Nutrition Bulletin 26, 348–355.
Ameny, M.A., Wilson, P.W., 1997. Relationship between Hunter color
values and b-carotene contents in white-fleshed African sweetpotatoes
(Ipomoea batatas Lam.). Journal of the Science of Food and
Agriculture 73, 301–306.
Beaton, G.H., Martorell, R., Aronson, K.J., Edmonston, B., McCabe, G.,
Ross, A.C., Harvey, B., 1993. Effectiveness of vitamin A supplementa-
tion in the control of young child morbidity and mortality in
developing countries. ACC/SCN State-of-the-art Series Policy Discus-
sion Paper No. 13. World Health Organization, Geneva, Switzerland.
Chandler, L.A., Schwartz, S.J., 1988. Isomerization and losses of trans-b-
carotene in sweet potatoes as affected by processing treatments.
Journal of Agricultural and Food Chemistry 36, 129–133.
CIP, 2006. International Potato Center (CIP) Report. McKnight
Foundation CCRP East African Regional Sweetpotato Project Annual
Report (April 2005–March 2006). International Potato Center,
Nairobi, Kenya.
Dary, O., Mora, J.O., 2002. Food fortification to reduce vitamin A
deficiency: International vitamin A consultative group recommenda-
tions. Journal of Nutrition 132, 2927S–2933S.
de Sá, M.C., Rodriguez-Amaya, D.B., 2004. Optimization of HPLC
quantification of carotenoids in cooked green vegetables. Comparison
of analytical and calculated data. Journal of Food Composition and
Analysis 17, 37–51.
Desobry, S.A., Netto, F.M., Labuza, T.P., 1998. Preservation of b-
carotene from carrots. Critical Reviews in Food Science and Nutrition
38, 381–396.
FAO/ILSI, 1997. Preventing micronutrient malnutrition: A guide to food-
based approaches. Food and Agriculture Organization of the United
Nations/International Life Sciences Institute, Washington, DC.
FAO/WHO, 1988. Requirements of vitamin A, iron, folate and vitamin
B12. Report of a Joint FAO/WHO Expert Consultation. FAO Food
and Nutrition Series No. 23. Food and Agriculture Organization of the
United Nations, Rome, Italy.
Gibson, R.S., Hotz, C., 2001. Dietary diversification/modification
strategies to enhance micronutrient content and bioavailability of
diets in developing countries. British Journal of Nutrition 85,
S159–S166.
Hagenimana, V., Low, J., 2000. Potential of orange-fleshed sweet potatoes
for raising vitamin A intake in Africa. Food and Nutrition Bulletin 21,
414–418.
Hagenimana, V., Carey, E.E., Gichuki, S.T., Oyunga, M.A., Imungi, J.K.,
1999. Carotenoid contents in fresh, dried and processed sweetpotato
products. Ecology of Food and Nutrition 37, 455–473.
Huang, A.S., Tanudjaja, L., Lum, D., 1999. Content of alpha-, beta-
carotene, and dietary fiber in 18 sweetpotato varieties grown in
Hawaii. Journal of Food Composition and Analysis 12, 147–151.
 ARTICLE IN PRESS
A. Bengtsson et al. / Journal of Food Composition and Analysis 21 (2008) 134–143
IOM, 2001. Dietary reference intakes for vitamin A, vitamin K, arsenic,
boron, chromium, copper, iodine, iron, manganese, molybdenum,
nickel, silicon, vanadium, and zinc. US Institute of Medicine, Food and
Nutrition Board, Standing Committee on the Scientific Evaluation of
Dietary Reference Intakes. National Academy Press, Washington, DC.
K’osambo, L.M., Carey, E.E., Misra, A.K., Wilkes, J., Hagenimana, V.,
1998. Influence of age, farming site, and boiling on pro-vitamin A
content in sweet potato (Ipomoea batatas (L.) Lam.) storage roots.
Journal of Food Composition and Analysis 11, 305–321.
Kidmose, U., Yang, R.-Y., Thilsted, S.H., Christensen, L.P., Brandt, K.,
2006. Content of carotenoids in commonly consumed Asian vegetables
and stability and extractability during frying. Journal of Food
Composition and Analysis 19, 562–571.
Kimura, M., Kobori, C.N., Rodriguez-Amaya, D.B., Nestel, P., 2007.
Screening and HPLC methods for carotenoids in sweetpotato, cassava
and maize for plant breeding trials. Food Chemistry 100, 1734–1746.
Lee, C.M., Trevino, B., Chaiyawat, M., 1996. A simple and rapid solvent
extraction method for determining total lipids in fish tissue. Journal of
AOAC International 79, 487–492.
Mdziniso, P., Hinds, M.J., Bellmer, D.D., Brown, B., Payton, M.E., 2006.
Physical quality and carotene content of solar-dried green leafy and
yellow succulent vegetables. Plant Foods for Human Nutrition 61, 13–21.
Mulokozi, G., Svanberg, U., 2003. Effect of traditional open sun-drying
and solar cabinet drying on carotene content and vitamin A activity of
green leafy vegetables. Plant Foods for Human Nutrition 58, 1–15.
Murphy, E.W., Criner, P.E., Gray, B.C., 1975. Comparisons of methods
for calculating retention of nutrients in cooked foods. Journal of
Agricultural and Food Chemistry 23, 1153–1157.
Ogunlesi, A.T., Lee, C.Y., 1979. Effect of thermal processing on the
stereoisomerisation of major carotenoids and vitamin A value of
carrots. Food Chemistry 4, 311–318.
Regier, M., Mayer-Miebach, E., Behsnilian, D., Neff, E., Schuchmann,
H.P., 2005. Influences of drying and storage of lycopene-rich carrots
on the carotenoid content. Drying Technology 23, 989–998.
Rodriguez-Amaya, D.B., 1997. Carotenoids and Food Preparation: The
Retention of Provitamin A Carotenoids in Prepared, Processed, and
Stored Foods. John Snow, Inc./OMNI Project, Arlington, VA.
Rodriguez-Amaya, D.B., 2000. Some considerations in generating
carotenoid data for food composition tables. Journal of Food
Composition and Analysis 13, 641–647.
143
Rodriguez-Amaya, D.B., 2002. Effects of processing and storage on food
carotenoids. Sight and Life Newsletter 3, 25–35.
Rodriguez-Amaya, D.B., Kimura, M., 2004. HarvestPlus Handbook for
Carotenoid Analysis. International Food Policy Research Institute and
International Center for Tropical Agriculture, Washington, DC.
Ruel, M.T., Levin, C.E., 2000. Assessing the potential for food-based
strategies to reduce vitamin A and iron deficiencies: A review of recent
evidence. FCND Discussion Paper No. 92. Food Consumption and
Nutrition Division, International Food Policy Research Institute,
Washington, DC.
Schiedt, K., Liaaen-Jensen, S., 1995. Isolation and analysis. In: Britton,
G., Liaanen-Jensen, S.H., Pfander, E. (Eds.), Isolation and Analysis.
Birkhäuser, Basel, Switzerland, pp. 81–108.
Sweeney, J.P., Marsh, A.C., 1971. Effect of processing on provitamin A in
vegetables. Journal of the American Dietetic Association 59, 238–245.
Takahata, Y., Noda, T., Nagata, T., 1993. HPLC determination of b-
carotene content of sweet potato cultivars and its relationship with
color values. Japanese Journal of Breeding 43, 421–427.
UDHS, 2001. Uganda Demographic and Health Survey (UDHS) Report.
Uganda Bureau of Statistics and Macro International, Inc., Calverton,
MD.
van het Hof, K.H., Gärtner, C., West, C.E., Tijburg, L.B.M., 1998.
Potential of vegetable processing to increase the delivery of carotenoids
to man. International Journal for Vitamin and Nutrition Research 68,
366–370.
van Jaarsveld, P.J., Faber, M., Tanumihardjo, S.A., Nestel, P., Lombard,
C.J., Benade, A.J.S., 2005. b-Carotene-rich orange-fleshed sweet
potato improves the vitamin A status of primary school children
assessed with the modified-relative-dose–response test. American
Journal of Clinical Nutrition 81, 1080–1087.
van Jaarsveld, P.J., Marais, D.W., Harmse, E., Nestel, P., Rodriguez-
Amaya, D.B., 2006. Retention of b-carotene in boiled, mashed orange-
fleshed sweet potato. Journal of Food Composition and Analysis 19,
321–329.
Welch, R.M., 2002. Breeding strategies for biofortified staple plant foods
to reduce micronutrient malnutrition globally. Journal of Nutrition
132, 495S–499S.
Woolfe, J.A., 1992. Sweet potato—past and present. In: Sweet Potato: An
Untapped Food Resource. Cambridge University Press, Cambridge,
UK, pp. 15–40.