Food

Chemistry
Food Chemistry 92 (2005) 481–489
www.elsevier.com/locate/foodchem

Bioaccessibility of minerals in school meals: Comparison

between dialysis and solubility methods
F. Cámara a, M.A. Amaro a, R. Barberá b,*
, G. Clemente c

a
Food Science and Nutrition, Campus de Rabanales, Edificio C-1, University of Córdoba, Carretera N-IV, Km 396, 14014 Córdoba, Spain
b
Nutrition and Food Chemistry, Faculty of Pharmacy, University of Valencia, Avda. Vicente Andrés Estellés s/n, 46100 Burjassot, Valencia, Spain
c
Statistics, Campus de Tarongers, University Polytechnical of Valencia, Valencia, Spain

Received 14 June 2004; received in revised form 2 August 2004; accepted 2 August 2004

Abstract

Determinations have been made of content and bioaccessibility of Ca, Fe, Zn and Cu in 13 dishes collected from a catering serv-
ice delivering to a school. Bioaccessibility was estimated by measuring the soluble or dialyzable mineral fraction resulting from in
vitro gastrointestinal digestion of the meal. The analyzed dishes had mineral contents (lg/g) in the following ranges: Ca (74.1–913),
Fe (2.8–17.9), Zn (2.8–13.1), Cu (0.28–1.90). Mineral solubility and dialysis percentages were as follows: Ca (1.7–96.2; 0.75–61.3), Fe
(16.0–97.8; 0.23–19.0), Zn (22.6–93; 5.78–31.45), Cu (35.7–92.3; 0.66–25.0). The highest bioaccessible Ca content corresponded to
fish-based dishes, while vegetables were poor sources of Ca. The lowest Fe and Zn bioaccessible percentages corresponded to veg-
etable-based dishes, particularly spinach omelet, whereas dishes having meat as the main ingredient exhibited the highest bioacces-
sible percentages. In the case of Cu, vegetable-based dishes could be considered the best sources.

2004 Elsevier Ltd. All rights reserved.

Keywords: Bioaccessibility; Calcium; Iron; Zinc; Copper; School meals

1. Introduction Zn, are often marginal in certain population groups,

e.g., toddlers or female adolescents (Räsänen & Ylönen,
It is widely accepted that, in childhood and adoles- 1992; Southon, Bailey, Wright, Belsten, & Finglas,
cence, diet influences, not only the immediate health of 1993).
children, but may also have an important impact on Calcium is an essential element. A good Ca intake
adult health. The childhood diet must be adequate to during childhood is associated with increased peak bone
support normal growth and development, and appropri- mass during adolescence, reduced development of pre-
ate amounts of minerals are required since a deficient in- cancerous cells in colonic mucosa, breast cancer, bone
take of certain minerals can produce diseases and lead to fragility and hip facture, with a possible delay in the
abnormal development. Mineral deficiency is usually development of high blood pressure and a reduction in
caused by a low mineral content in the diet when rapid the prevalence of osteoporosis (Barrer-Lux & Heany,
body growth is occurring and/or when there is poor 1994; Hallfrisch et al., 2000).
absorption of minerals from the diet (Favier, 1993). Iron serves metabolic and enzymatic functions (e.g.,
However, even in the ‘‘affluent’’ diet found in western Fe for the synthesis of hemoglobin or myoglobin, and
societies, intakes of some minerals, such as Ca, Fe and Fe-containing enzymes which participate in electron
transfer and redox reactions) (Yip, 2001). According
*
Corresponding author. Fax: +34 96 3864954. to WHO/UNICEF (1998), iron deficiency is the most
E-mail address: [email protected] (R. Barberá). prevalent single nutritional deficiency in the world and

0308-8146/$ - see front matter
2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.foodchem.2004.08.009
482 F. Cámara et al. / Food Chemistry 92 (2005) 481–489

is the main cause of anemia in infants, children, adoles- In vitro estimation of the bioavailability of minerals
cents, and women of childbearing age. Iron deficiency and trace elements from foods involves the simulation
anemia is related to delayed cognitive development of gastrointestinal digestion and measurement of the
and intellectual impairment in children (McGregor & mineral soluble fraction or the mineral fraction that
Ani, 2001), reduced work capacity (Haas & Brownlie, dialyses across a semipermeable membrane of a certain
2001), an increased risk of maternal and neonatal mor- pore size (Wienk, Marx, & Beynen, 1999). In fact, these
tality (Allen, 2002), and altered immune function (Cook methods measure only the amount of mineral available
& Lynch, 1986). in the gastrointestinal tract for absorption, i.e., its so-
Zinc, in turn, is essential for normal growth and called bioaccessibility (Salovaara, Sandberg, & Andlid,
development, because it plays an important role in gene 2002). However, these methods are widely used because
expression, regulation of cellular growth and differentia- of their good correlation with in vivo studies (Roig,
tion (Hambidge, 2000), and development of the immune Alegrı´a, Barberá, Farré, & Lagarda, 1999; Wienk
response. It has a recognized action on more than 300 et al., 1999). Bioaccessibility values must be taken as rel-
enzymes implied in the metabolism of nucleic acids, car- ative indices of bioavailability, which means that the
bohydrates and proteins, participating as a cofactor method used provides a good basis for establishing ten-
(Salgueiro et al., 2002). Marginal Zn deficiency is a com- dencies, comparisons and the determination of effects
mon nutritional problem in developing and developed caused by different factors (Azenha & Vascondecelos,
countries (Michelson, Samuelson, Graham, & Lonner- 2000).
dal, 1994). Among children, it produces serious conse- In recent years, the mentioned in vitro methods have
quences for health, such as retarded growth, an been used to estimate the bioaccessability of minerals
increase in infectious diseases, and impaired cognitive from different foods (Bosscher, Van Cauwenbergh,
function (Rosado, 1998). Van der Auwera, Robberecht, & Deelstra, 2002; Jovanı´
Copper play a pivotal role in cell physiology as a cat- et al., 2000; Roig et al., 1999; Sahuquillo, Barberá, &
alytic cofactor in the redox chemistry of enzymes for Farré, 2003; Sebastiá, Barberá, Farré, & Lagarda,
proteins that carry out fundamental biological functions 2001) and also from dishes and composite diets (Kenne-
required for growth and development (Linder, 1991). It fick & Cashman, 2000; Luccarini, Canali, Cappelloni,
is needed for mitochondrial respiration, Fe absorption, Di Lullo, & Lombardi-Boccia, 1999; Pushpanjali &
free radical scavenging and elastin cross-linking (Tapi- Khokhar, 1996; Van Dyck et al., 1996).
ero, Townsend, & Tew, 2003). Copper deficiency is less The present study uses two in vitro methods (solubil-
frequent than other element deficiencies and has been ity and dialysis) to assess the bioaccessible amounts of
described mainly in premature infants and children Ca, Fe, Zn and Cu provided by different dishes usually
recovering from malnutrition (Cordano, 1998), or who distributed to a Spanish school lunchroom with the pur-
change to a poor intake of this mineral in the diet. pose to compare mineral bioaccessibility of the dishes as
Most school age children and adolescents have their a function of the ingredients used for their elaboration.
main meal of the day in the school lunchroom, and this
meal often constitutes the principal mineral contribution
to the daily recommended intake. 2. Materials and methods
To estimate the quality of a dietetic source of a given
mineral, it is necessary to precisely define the amount of 2.1. Materials
mineral available for absorption and utilization, i.e., its
bioavailability (Barberá & Farré, 1992). The bioavaila- Digestive enzymes and bile salts were supplied by Sig-
bility of minerals and trace elements has generated ma Chemical Co. (St. Louis, MO, USA). The working
increasing interest in the field of nutrition. solutions of these enzymes were prepared immediately
Bioavailability should be determined by in vivo meas- before use.
urements (Van Dyck, Tas, Robberecht, & Deelstra, Pepsin solution was obtained by dissolving 1.6 g of
1996). Ideally, this kind of research should have been per- pepsin (P-7000 from porcine stomach) in 10 ml of HCl
formed in humans; however, such studies are difficult, (0.1 M). The solution of pancreatin and bile salts was
expensive, and provide limited data with each experiment prepared by dissolving 0.4 g of pancreatin (P-170 from
(Hansen, Sandstrom, & Lonnerdal, 1996). While animal porcine pancreas) and 2.5 g of bile salt (B-8631 of por-
assays are less expensive, they are somewhat limited by cine origin) in 100 ml of 0.1 M NaHCO3.
uncertainties with regard to differences in metabolism be- The dialysis membranes, with a pore size (MMCO) of
tween animals and humans. As an alternative to human >12,000 Å (Dia. Inf. 36/3200 –28.6 mm, 30 m, Bestl no.
and animal in vivo studies, the availability of minerals 1063F09, Medicell Int. LTD, England), were rinsed sev-
or trace elements has also been estimated, based on sim- eral times with distilled deionized water before use.
ple, rapid and inexpensive in vitro methods (Miller, Schr- Standard Ca, Fe, Zn and Cu solutions were prepared
icker, Rasmussen, & Van Campen, 1981). immediately before use by dilution (with distilled deion-
 F. Cámara et al. / Food Chemistry 92 (2005) 481–489 483

ized water) of a 1000 mg/l standard solution (Titrisol, of NaHCO3 equivalent to the titrable acidity (previously
Merck, Darmstadt, Germany). The lanthanum solution measured) was placed in the flasks, together with 30 g
(5 g/100 ml) was prepared with La2O3 (Merck). aliquots of the pepsin digest. Incubation was continued
All reagents used were of analytical grade, and Milli- for 45 min, the pancreatic-bile salt mixture (7.5 ml) was
pore-MilliQ distilled deionized water was used through- added, and incubation was continued up to 2 h.
out. Glass and polyethylene material was soaked in After incubation, the segments of dialysis tubes were
HNO3 (sp. gr. 1.38) for 15 min and then rinsed three removed from the flasks, washed and weighed.
times with distilled deionized water. The titratable acidity was defined as the number of
equivalents of NaOH required to titrate the combined
2.2. Samples pepsin digest pancreatin–bile salts mixture to pH 7.5.
The mineral contents of the dialysis tubes were ana-
Thirteen dishes habitually included in the monthly lyzed by flame atomic absorption spectrophotometry
programme of the school menu with different formula- (FAAS) (Perkin–Elmer. Model 2380) for Ca, Fe and
tions and forms of preparation were supplied by a cater- Zn elements and by graphite furnace-atomic absorption
ing establishment (Table 1). spectrophotometry (GFAAS) with Zeeman effect (Per-
Aliquots of dishes were frozen in a still-air freezer at kin–Elmer Analyst 600) in the case of Cu.

18 C until required for analyses and processing.
2.3.1.2. Solubility method. The method described by
2.3. Analytical procedure Sahuquillo et al. (2003) with slight modifications was ap-
plied. Thirty grammes of each dish were homogenized
2.3.1. In vitro digestion with 70 ml of deionized distilled water, and the pH
2.3.1.1. Dialysis method. The method described by was adjusted to 2.0 with 6 N HCl. Pepsin–HCl digestion
Jovanı́ et al. (2000), with slight modifications, was ap- was carried out as mentioned in the dialysis method.
plied to estimate dialyzable minerals. Forty grammes Prior to the intestinal digestion step, the pH of the gas-
of each dish were homogenized with 60 g of deionized- tric digests was raised to 5 by drop-wise addition of 1 M
distilled water, and the pH was adjusted to 2.0 with 6 NaHCO3. Then 18.8 ml of the pancreatin–bile salt mix-
N HCl. To carry out pepsin–HCl digestion, 0.5 g of pep- ture were added and incubation was continued up to 2
sin solution per 100 g of sample were added. The mix- h. To stop intestinal digestion, the sample was maintained
ture was then incubated for 2 h at 37 C in a shaking for 10 min in an ice bath. The pH was adjusted to 7.2 by
water bath. A dialysis bag (molecular mass cut-off value drop-wise addition of 0.5 M NaOH. Aliquots of 20 g of
10–12,000 Da) containing 25 ml of water and an amount the digested sample were transferred to polypropylene

Table 1
Description of the dishes of the school menu
Type Dish Formulation/preparation
With cereal base Cuban style rice Rice*, fried egg*, chicken broth, fried tomato*, sausage*, sunflower
oil, garlic and salt
Rice with lean meat Rice*, lean meat*, mushrooms, peppers, green peas, tomato,
onion*, sunflower oil, garlic and salt
Spaghetti with sausage Spaghetti*, sausage, tomato, sunflower oil and salt
Macaroni with tuna Macaroni*, tuna, tomato, sunflower oil and salt
With leguminous base Lentils with sausage Lentils*, sunflower oil, mashed tomato, salt, coloring, laurel,
(chorizo) chicken broth, carrot, sausage*, onion, pepper, garlic, potatoes*
Stew Chickpeas*, green beans, carrot, pork bone, chicken, potato*, veal
ragout*, salt pork, chicken broth and salt
With tuber base Potato stew Potatoes*, veal ragout*, white wine, mashed tomato, carrot,
chicken broth, onion, pepper, sunflower oil, laurel, coloring and
salt
With meat base Chicken in breadcrumbs Chicken in breadcrumbs*, vegetables and sunflower oil
with vegetable stew
Chicken in sauce Chicken breast*, chicken broth, flour, onion, almonds, sunflower
oil, potatoes, coloring matter and salt
Fish Fried hake Hake filet and sunflower oil
Precooked hake filet Hake in breadcrumbs and sunflower oil
With egg Spanish/potato omelet Potatoes, egg, sunflower oil and salt
Spinach omelet Spinach, eggs, sunflower oil and salt
*
Main ingredients.
484 F. Cámara et al. / Food Chemistry 92 (2005) 481–489

centrifuge tubes (50 ml, Costar Corning Europe, Badho- calculated as follows: Solubility (%) = 100 · S/C, where
evedorp, The Netherlands) and centrifuged (3500g) for S = soluble mineral content (lg/g sample) and C = total
1 h at 4 C. Then, the supernatant (soluble fraction) was mineral content of the sample (lg/g sample).
collected, its organic matter destroyed by dry ashing,
and the mineral content measured by FAAS. 2.5. Statistical analysis

2.4. Mineral content determination A simple regression analysis (Statgraphics Plus 4.0
for Microsoft Windows) was applied between Ca, Fe,
Ca, Fe, Zn and Cu total contents and Ca, Fe and Zn and Cu contents and their solubility and dialysis
Zn in soluble and dialyzed fractions were measured percentages.
by FAAS. The Cu content of the dialysates was deter-
mined by (GFAAS) with Zeeman effect. Previously
reported instrumental conditions were applied (Garcı́a, 3. Results and discussion
Alegria, Barberá, Farré, & Lagarda, 1998; Roig et al.,
1999). Total Ca, Fe, Zn and Cu contents in the analyzed
Prior to the atomic absorption spectrophotometry dishes are listed in Table 2. Values found for the four
(AAS), determination of Ca, Fe, Zn and Cu from dishes measured minerals were in the ranges given by other
and mineral soluble fractions, the organic matter was authors for composite dishes (Table 3). In five of
destroyed by ashing in a temperature-programmed fur- the analyzed dishes (rice with lean meat, chicken with
nace (Heraeus M1100/3, Hanau, Germany) at 450 C vegetables, potato omelet, macaroni and potato stew),
for 48 h (the temperature being slowly increased at a rate two different mean contents are mentioned, because
of 50 C/h). To the black ashes, 3 ml of HNO3 were two samplings in different seasons (spring and au-
added; the sample was then heated to dryness and tumn) were carried out. The differences found could
placed in a muffle furnace at 450 C for 24 h. The proc- be due to the ingredients or food origin (Gibson,
ess was repeated as many times as necessary to obtain a 1994), and, even in the case of the same ingredients
white residue. After cooling, the residue was dissolved and recipes, to preparation on different days and by
with 1 ml of HCl (sp. gr. 1.19) and 10 ml of distilled different people (Torelm, Danielsson, Appelqvist, &
deionized water. Bruce, 1997).
Dialysis mineral percentages were calculated as fol- The best models obtained in the correlation of each
lows: Dialysis (%) = 100 · D/C, where D = dialyzed total mineral content (Ca, Fe, Zn and Cu) to the corre-
mineral content (lg/g sample) and C = total mineral sponding soluble and dialyzable fractions are shown in
content (lg/g sample). Soluble mineral percentages were Table 4.

Table 2
Total mineral content (lg/g) (mean ± standard deviation, n = 3)
Ca Fe Zn Cu
Lentils with sausage (chorizo) 184 ± 15.8 10.7 ± 1.15 7.81 ± 0.76 1.48 ± 0.25
Stew 140 ± 13.4 9.79 ± 2.18 7.11 ± 0.70 1.90 ± 0.18
Cuban style rice 85.60 ± 3.86 7.56 ± 0.30 4.35 ± 0.13 1.05 ± 0.03
Rice with lean meat* 89.9 ± 1.44 15.9 ± 0.17 7.60 ± 0.16 1.14 ± 0.02
102 ± 4.61 2.92 ± 0.07 7.03 ± 0.13 1.46 ± 0.03
Chicken with vegetable stew* 114 ± 9.26 8.15 ± 1.60 6.59 ± 0.83 0.69 ± 0.02
323 ± 4.82 7.14 ± 1.96 5.13 ± 0.09 0.75 ± 0.05
Chicken in sauce 106 ± 3.22 10.5 ± 0.86 13.1 ± 0.66 0.76 ± 0.04
Potato* omelet 315 ± 6.58 13.7 ± 0.20 7.01 ± 0.17 0.63 ± 0.01
294 ± 7.67 8.92 ± 0.93 6.53 ± 0.19 1.01 ± 0.02
Spinach omelet 913 ± 31.1 17.9 ± 0.51 9.00 ± 0.27 0.79 ± 0.01
*
Macaroni with tuna 109 ± 3.83 6.48 ± 0.56 3.27 ± 0.08 0.83 ± 0.04
137 ± 8.50 6.28 ± 0.65 4.68 ± 0.05 1.16 ± 0.01
Spaghetti with sausage 145 ± 1.53 7.03 ± 0.31 6.29 ± 0.46 0.99 ± 0.02
Pre-cooked hake filet 190 ± 18.0 4.40 ± 0.51 3.70 ± 0.19 0.61 ± 0.01
Fried hake 357 ± 52.0 3.95 ± 0.08 3.88 ± 0.07 0.28 ± 0.01
Potato stew* 74.1 ± 1.85 2.82 ± 0.30 3.12 ± 0.07 0.76 ± 0.01
99.0 ± 1.84 4.85 ± 0.09 2.82 ± 0.17 0.80 ± 0.02
*
Dishes samples in two different time periods (spring–autumn).
 F. Cámara et al. / Food Chemistry 92 (2005) 481–489 485

Table 3
Calcium, Fe, Zn and Cu contents (lg/g) of different types of diet
Sample Ca Fe Zn Cu References
Kuwaiti composite dishes 1230–99.7 51.2–1.7 41.6–1.0 12.5–0.3 Dashti et al. (2004)
French dishes 1594–85 20.6–1.28 26.6–0.160 1.22–0.046 Noël et al. (2003)
Oman dishes 345–41 16.0–0.5 22.5–0.5 2.9–0.2 Musaiger et al. (1998)
Western dishes 2684–40.1 – 79.9–0.17 7.39–0.12 Ekmekcioglu et al. (1999)
Polish fast food 2481–95 27–5.9 – – Grajeta et al. (2002)

Table 4
Solubility mineral (SM) and dialyzed mineral (DM) percentage (%) (mean ± standard deviation of n = 6)
Ca Fe Zn Cu
SM DM SM DM SM DM SM DM
Lentils with sausage 35.3 ± 3.15 10.3 ± 1.32 16.0 ± 2.89 3.43 ± 0.35 69.9 ± 3.61 25.3 ± 1.38 70.9 ± 4.12 25.0 ± 3.42
Stew 91.7 ± 9.83 30.5 ± 4.18 49.7 ± 0.91 2.14 ± 0.31 37.1 ± 0.52 26.0 ± 0.99 35.7 ± 4.76 13.9 ± 4.03
Cuban style rice 67.6 ± 7.47 23.8 ± 3.94 81.7 ± 4.88 8.17 ± 0.79 58.9 ± 5.92 20.9 ± 2.79 71.0 ± 4.70 19.1 ± 2.63
Rice with lean meat 32.8 ± 5.70 29.3 ± 1.70 97.1 ± 0.98 1.62 ± 0.08 45.5 ± 2.63 7.45 ± 0.83 42.5 ± 3.47 3.53 ± 0.97
Chicken vegetable stew 38.0 ± 3.01 24.8 ± 2.76 53.8 ± 6.03 8.29 ± 0.99 67.0 ± 4.21 19.1 ± 1.68 60.1 ± 6.07 1.58 ± 0.17
Chicken in sauce 26.5 ± 1.89 61.3 ± 2.06 55.1 ± 2.98 5.12 ± 0.41 39.2 ± 6.98 12.0 ± 1.72 57.2 ± 4.57 0.66 ± 0.10
Potato omelet 35.9 ± 4.77 11.3 ± 2.68 73.7 ± 1.48 1.95 ± 0.19 66.9 ± 12.5 8.53 ± 1.16 92.3 ± 4.38 22.0 ± 0.51
Spinach omelet 1.73 ± 0.29 0.75 ± 0.14 45.4 ± 2.44 0.23 ± 0.01 46.7 ± 1.64 5.78 ± 1.11 88.3 ± 5.70 7.64 ± 1.11
Macaroni with tuna 40.8 ± 3.84 12.0 ± 0.47 80.0 ± 11.0 2.31 ± 1.01 61.6 ± 3.18 7.64 ± 1.00 64.4 ± 2.14 14.8 ± 0.92
Spaghetti with sausage 45.0 ± 2.59 17.5 ± 1.46 59.9 ± 4.95 3.39 ± 0.25 22.6 ± 2.55 9.57 ± 0.55 92.12 ± 7.47 3.70 ± 0.41
Pre-cooked hake filet 47.7 ± 3.82 21.1 ± 4.23 79.1 ± 4.83 15.5 ± 2.97 58.7 ± 3.97 8.98 ± 0.93 87.5 ± 3.71 N.D.
Fried hake 46.8 ± 7.41 17.4 ± 0.83 97.9 ± 3.65 3.39 ± 0.84 29.8 ± 9.20 7.67 ± 1.45 87.0 ± 6.09 N.D.
Potato stew 96.2 ± 7.95 28.1 ± 2.98 96.5 ± 9.20 19.0 ± 0.93 93.0 ± 5.52 31.5 ± 1.61 66.5 ± 8.04 15.7 ± 0.99
N.D., non detectable.

In the multiple regression analysis applied to evaluate corresponding to potato stew and spinach omelet,
the relationship between Ca, Fe, Zn and Cu total con- respectively (Table 2). Solubility and dialysis percent-
tents, and mineral solubility and dialysis, no statistically ages ranged from 1.7% to 96.2%, and from 0.75% to
significant correlations were found. 61.3%, respectively (Table 4)
Solubility and dialysability, expressed as lg/g,
3.1. Calcium showed a significant (p < 0.05) correlation with Ca
content – soluble Ca (r = 0.76) and dialyzable Ca
The total Ca content in the analyzed dishes ranged (r = 0.72) – the values being greater with increasing to-
from 74.1 to 913 lg/g, the lowest and highest contents tal Ca content (Table 4). Values corresponding to

Table 5
Regression analysis: Models and correlation coefficients
Element Factor Model p r
Total Ca, lg/ g Soluble Ca, lg/g y = 24.24 + 36x 0.0042 0.76
Soluble Ca, % y = exp (4.52
3.93 · 10
3x) 0.0001
0.88
Dialyzed Ca, lg/ g y = 1/(2.16 · 10
2 + 1.11 · 10
4x) 0.0053 0.72
Dialyzed Ca, % y = exp (3.70
4.18 · 10
3x) 0.0000
0.91
Total Fe, lg/g Soluble Fe, lg/ g y = 2.49 + 0.32x 0.0002 0.87
Soluble Fe, % y = 101.16
3.77x 0.0024
0.79
Dialyzed Fe, lg/g y = 1/(
1.40 + 0.71x) 0.0618* 0.53
Dialyzed Fe, % y = exp (3.09
0.20x) 0.0009
0.80
Total Zn, lg/g Soluble Zn, lg/g y = 1.06 + 0.34x 0.0077 0.70
Soluble Zn, % y = 30.73 + 122.28/x 0.0907* 0.48
Dialyzed Zn, lg/ g y = 0.16X0.88 0.0496 0.55
Dialyzed Zn, % y = 17.44
0.44x 0.6466*
0.14
Total Cu, lg/g Soluble Cu, lg/g y = 0.66X0.64 0.0016 0.78
Soluble Cu, % y = 1/(6.61 · 10
3 + 8.78 · 10
3x) 0.0060 0.72
Dialyzed Cu, lg/ g y =
8.46 · 10
2 + 0.21x 0.0032 0.75
Dialyzed Cu, % y = 0.96 + 9.66x 0.1263* 0.45
r, Correlation coefficient.
*
p > 0.05.
486 F. Cámara et al. / Food Chemistry 92 (2005) 481–489

spinach omelet were not included. However, when Ca (r =
0.80) (Table 4) were found. A similar correlation
solubility and dialysability were expressed as percent- coefficient (r =
0.72) between total Fe content and
ages, negative correlations were found (r =
0.88 and dialysis Fe percentage was reported in a dialysis assay
r =
0.91) for solubility and dialysability, respectively of Fe from weaning dishes containing different vegetable
(Table 4). ingredients (Bosscher et al., 2002).
Spinach omelet, the dish with the highest Ca content, The highest Fe contents corresponded to spinach
showed the lowest Ca solubility and dialysability per- omelet, potato omelet, rice with lean meat, chicken in
centages (Table 5). The high oxalic acid content in spin- sauce, lentils and stew (Table 2) and - with the exception
ach (6.62 g/kg; Chawla, Saxena, & Seshadri, 1988) of chicken in sauce – these dishes yielded the lowest dial-
favors Ca precipitation and thus decreases its bioacces- yzable Fe percentages (Table 4).
sibility. Kennefick & Cashman (2000) observed that Spinach has a high Fe content (Chiplonkar et al.,
oxalate addition to semisynthetic diets decreases Ca dia- 1999), but also contains oxalic acid (Yadav & Sehgal,
lysability in a dose-dependent manner. 2003) that affects Fe bioaccessibility. This could explain
Vegetable components, such as fibre and phytic acid, why spinach omelet presented one of the lowest percent-
negatively affect Ca dialysability (Kennefick & Cash- ages of soluble Fe (45.4%), and the lowest percentage of
man, 2000; Periago et al., 1997). The presence of these dialyzable Fe (0.23%) – despite having the highest Fe
inhibitors could explain the relatively low dialysis per- content (17.9 lg/g) among the dishes analyzed. Similar
centages obtained for pasta dishes (12.0% and 17.5% results were reported by Chawla et al. (1988), who indi-
for macaroni and spaghetti, respectively) and lentils cated that, among the analyzed foods, spinach presented
(10.3%) – these values being similar to those reported the lowest solubility (2.8%) percentage and the highest
by Luccarini et al. (1999) for macaroni with broccoli oxalic acid content (6.62 g/kg).
(18.07%), and for macaroni with cauliflower (19.0%), The low solubility and dialysability of Fe from spin-
where the Ca dialysis percentages were lower than those ach omelet could also be related to the egg used as addi-
corresponding to broccoli (22.9%) and cauliflower tional ingredient. The low dialysability percentage of Fe
(23.4%). (1.95%) from potato omelet, whose main ingredient is
Calcium dialysabilities (lg/g) of pre-cooked (40.0) egg, would support this assertion. The negative effect
and fried hake (62.2) were in the lowest end of the range of egg proteins upon Fe bioavailability has been re-
of dialysability percentages (63.6–83.1%) reported by ported elsewhere (Beard, Dawson, & Pinero, 1996; Fair-
Martı́nez, Santaella, Ros, & Periago (1998) in fish-based weather-Tait, 1989; Kapsokefalou & Miller, 1995).
infant food. It has been reported that different fish spe- Some egg proteins (ovotransferrin from egg white and
cies are a good dietetic source of Ca (Larsen, Thilsted, phosphovitin from yolk) form complexes with Fe. Phos-
Kongsbak, & Hansen, 2000; Losso, Munene, & Moody, phovitin binds more than 50% of Fe (III) in yolk, thus
2003), and the data obtained in the present study indi- reducing Fe bioavailability (Thapon, Audiot, Proteis,
cate both fish Ca content and bioaccessibility to be & Sauveur, 1994).
good. Iron dialysis percentages of pasta dishes are low (2–
3%) (Table 4), in agreement with the low Fe bioavaila-
bility (3%) obtained by Galán, Cherouvier, Fernández-
3.2. Iron Ballart, Marti-Henneberg, & Hercberg (1990) in a study
among volunteers fed Spanish dishes labelled with Fe
Iron contents in the analyzed dishes ranged from 2.8 radioisotopes.
to 17.9 lg/g, corresponding to potato stew and spinach The highest Fe dialysis percentages corresponded to
omelet, respectively. Solubility and dialysability percent- dishes having meat as the main ingredient, such as
ages ranged from 16.0% to 97.8%, and from 0.23% to breadcrumb chicken with vegetables (15.5%) or potato
19.0%, respectively (Table 5). stew (19.0%). Similar dialysis percentages have been re-
A linear positive correlation (p < 0.05) was found be- ported for meat-based beikost (4.9–8.6%) (Santaella,
tween Fe content and Fe soluble fraction content (lg/g) Martı´nez, Ros, & Periago, 1997) and fish-based beikost
(r = 0.87; Table 4) when values corresponding to lentils (9.3–11.6%) (Martı´nez et al., 1998).
were excluded from the model, since the high Fe content Meat protein favors Fe bioavailability (Engelmann
of lentils was associated with a low soluble Fe content. et al., 1998) – this effect having been observed in in
Lentils yielded lower soluble Fe contents than chick peas vitro (Kapsokefalou & Miller, 1991; Mulvihill & Mor-
or white beans (Sahuquillo et al., 2003), despite the high- risey, 1998) and in vivo (Baech et al., 2003;
er Fe content of lentils. This could be explained by the Kapsokefalou & Miller, 1995) – and this could be re-
higher phytate content of lentils (Málñez, Alegrı́a, lated to the Fe (III) to Fe (II) reducing ability of sul-
Farré, & Frigola, 2002). fhydryl groups in sulfur-containing amino acids from
Negative correlations between Fe content and Fe sol- meat proteins (Mulvihill, Kirwan, Morrisey, & Flynn,
ubility percentage (r =
0.79) or Fe dialysis percentage 1998).
 F. Cámara et al. / Food Chemistry 92 (2005) 481–489 487

3.3. Zinc follows: calcium > iron = zinc > copper. An increase in
solubility (lg/g) with increasing element content was ob-
In the analyzed dishes, zinc contents ranged from 2.8 served for all four minerals in the analyzed dishes. When
to 13.1 lg/g (Table 2), these values corresponding to po- Ca and Fe contents increased, the corresponding dialysis
tato stew and chicken in sauce, respectively. Solubility amount and percentage decreased. In the case of Zn and
and dialysability percentages (Table 4) ranged from Cu, increased mineral content implied a larger dialyzed
22.6% to 93.0%, and from 5.78% to 31.4%, respectively. amount.
Significant correlations (p < 0.05) between Zn content In the analyzed dishes, no significant correlations
and soluble Zn (r = 0.70) and dialyzable Zn fractions were found between the Ca, Fe, Zn and Cu solubilities
(r = 0.55) were found, showing soluble and dialyzable and dialysis percentages. High mineral solubility was
Zn to increase with the Zn content of the meal. not always related to a high dialysis percentage, because
The highest Zn dialysis percentages corresponded to the mineral may be bound to compounds of molecular
dishes having tubers or legumes as the main ingredient, sizes in excess of the pore size of the dialysis membrane.
but also containing meat – such as potato stew, lentils Only for Ca in chicken in sauce was the dialysis percent-
and stew, followed by dishes having meat as the main age seen to be greater than the solubility percentage.
ingredient (chicken with vegetables, chicken in sauce). The main dietetic Ca sources, on considering Ca con-
The latter yielded Zn dialysis percentages similar to tent and dialysability, were fish dishes. The negative ef-
those reported for meat-based weaning foods (13.7– fect of oxalates on Ca bioaccessibility was evident in
16.1%) (Santaella et al., 1997). Several studies have spinach omelet, where a high Ca content was neverthe-
shown Zn absorption to increase with the animal pro- less associated with the lowest soluble and dialyzable
tein intake (Sandström, 1992). Ca percentages.
Zinc dialysis percentages in cereal-based dishes (rice The highest Fe contents corresponded to cereal, leg-
with lean meat, macaroni and spaghetti) were low, with umes and spinach omelet dishes. However, the dialys-
the exception of Cuban style rice (20.9%), in which the ability percentages were low and unrelated to the Fe
presence of egg as a main ingredient could explain the contents. The low solubility percentage of Fe from len-
higher Zn bioaccessibility. tils, and the low dialysis percentage of Fe from spinach
omelet were salient findings – particularly considering
3.4. Copper that the corresponding Fe contents were high.
The highest Zn dialysis percentage corresponded to
Copper contents were in a narrow range in the ana- legume-based dishes, despite their high phytic acid con-
lyzed dishes, from 0.28 to 1.90 lg/g, corresponding to tents; however, heat treatment was applied to legumes,
fried hake and stew, respectively (Table 2). Copper sol- thus reducing their phytic acid contents (Mañez et al.,
ubility and dialysis percentages ranged from 35.9% to 2002).
92.3%, and from non-detectable to 25.0%, respectively Finally, Cu showed opposite behaviour to Ca, Fe, and
(Table 4). The lowest dialysis percentages correspond Zn in the sense that vegetable-based foods had higher
to dishes having meat as the main ingredient, dialyzable mineral bioaccessibility (solubility and dialysability) per-
Cu being non-detectable in hake. On the other hand, the centages than meat-based dishes – even dialyzable Fe
latter yielded Cu solubility percentages of over 80%, being non-detectable in fish. This implies that phytic acid
thus showing Cu to be bound to compounds larger than and other well known inhibitors of Ca, Fe and Zn
the pore of the dialysis membrane used. absorption would not affect Cu bioaccessibility.
Significant correlations between Cu content and solu-
ble Cu (r = 0.78) and dialyzable Cu fractions (r = 0.75)
were found (Table 4).
Acknowledgement
Available information on the solubility and dialys-
ability of Cu from foods is scarce, particularly in refer-
F. Cámara benefits from a Grant given by the Minis-
ence to Cu from dishes and menus. According to the
try of Education and Culture (Spain).
results obtained, meat and fish were poorer Cu dietetic
sources than vegetable-based foods, since low dialysis
percentages must be added to the low Cu contents (Ta-
bles 2 and 5). References

Allen, L. H. (2002). Iron supplements: Scientific issues concerning

4. Conclusions efficacy and implications for research and programs. Journal of
Nutrition, 132, 813S–819S.
Azenha, M., & Vascondecelos, M. (2000). Assessment of the Pb and
The mineral contents of the analyzed dishes, usually Cu in vitro availability in wines by jeans of speciation procedures.
found in Spanish school lunchrooms, can be ranked as Food Chemical Toxicology, 38, 899–912.
488 F. Cámara et al. / Food Chemistry 92 (2005) 481–489

Baech, S. B., Hansen, M., Bukhave, K., Jensen, M., Sorensen, S. S., Hansen, M., Sandstrom, B., & Lonnerdal, B. (1996). The effect of
Kristensen, L., Purslow, P. P., Skibsted, L. H., & Sandström, B. caseinphosphopeptides on zinc and calcium absorption from high
(2003). Nonheme-iron absorption of small amounts of pork meat. phytate infant diets assessed in rat pups and Caco-2 cells. Pediatric
American Journal of Clinical Nutrition, 77, 173–179. Research, 40(4), 547–552.
Barberá, R., & Farré, R. (1992). Biodisponibilidad de los elementos Jovanı´, M., Alegrı́a, A., Barberá, R., Farré, R., Lagarda, M. J., &
traza. Revista Española de Ciencia y Tecnologı´a de los Alimentos, Clemente, G. (2000). Effect of proteins, phytates, ascorbic acid and
32(4), 381–399. citric acid on dialysability of calcium, iron zinc and copper in soy-
Barrer-Lux, M. J., & Heany, R. P. (1994). The role of calcium intake in based infant formulas. Nahrung, 44(2), 114–117.
preventing bone fragility, hypertension and certain cancers. Journal Kapsokefalou, M., & Miller, D. D. (1991). Effects of meat and selected
of Nutrition, 124, 1406S–1411. food components on the valence of nonheme iron during in vitro
Beard, J. L., Dawson, H., & Pinero, D. J. (1996). Iron metabolism: A digestion. Journal of Food Science, 56, 352–355.
comprehensive review. Nutrition Review, 54, 295–317. Kapsokefalou, M., & Miller, D. D. (1995). Iron speciation in intestinal
Bosscher, D., Van Cauwenbergh, R., Van der Auwera, J. C., contents of rats fed meals composed of meat and nonmeat sources
Robberecht, H., & Deelstra, H. (2002). Calcium, iron and zinc of protein and fat. Food Chemistry, 52, 47–56.
availability from weaning meals. Acta Paediatrica, 91, 761–768. Kennefick, S., & Cashman, K. D. (2000). Investigation of an in vitro
Chawla, S., Saxena, A., & Seshadri, S. (1988). In vitro availability of model for predicting the effect of food components on calcium
iron in various green leafy vegetables. Journal of Science Food and availability from meals. International Journal of Food Sciences and
Agriculture, 46, 125–127. Nutrition, 51, 45–54.
Chiplonkar, S. A., Tarwadi, K. V., Kavedia, R. B., Mengale, S. S., Larsen, T., Thilsted, S. H., Kongsbak, K., & Hansen, M. (2000).
Paknikar, K. M., & Agte, V. V. (1999). Fortification of vegetarian Whole small fish as a rich calcium source. British Journal of
diets for increasing bioavailable iron density using green leafy Nutrition, 83, 191–196.
vegetables. Food Research International, 32, 169–174. Linder, M. C. (1991). Biochemistry of copper. New York: Plenum
Cook, J. D., & Lynch, S. R. (1986). The liabilities of iron deficiency. Press.
Blood, 68, 803–809. Losso, J. N., Munene, C. N., & Moody, M. W. (2003). Inductively
Cordano, A. (1998). Clinical manifestations of nutritional copper coupled plasma optical emission spectrometric determination of
deficiency in infants and children. American Journal of Clinical minerals in catfish frame. Nahrung, 47, 309–311.
Nutrition, 67, 1012S–1016. Luccarini, M., Canali, R., Cappelloni, M., Di Lullo, G., & Lombardi-
Dashti, B., Al-Awadi, F., Alkandari, R., Ali, A., & Al-Otaibi, J. Boccia, G. (1999). In vitro calcium availability from brassica
(2004). Macro- and microelements contents of 32 Kuwaiti com- vegetables (Brassica oleracea L.) and as consumed in composite
posite dishes. Food Chemistry, 85, 331–337. dishes. Food Chemistry, 64, 519–523.
Ekmekcioglu, C., Anderle, H., Strauss-Blasche, G., Steffan, I., Málñez, G., Alegrı́a, A., Farré, R., & Frigola, A. (2002). Effect of
Feyertag, J., & Marktl, W. (1999). Calcium, magnesium, copper traditional, microwave and industrial cooking on inositol phos-
and zinc content of menu components: Comparison of analysed phate content in beans, chickpeas and lentils. International Journal
with calculated values. Nahrung, 43, 311–316. of Food Science and Nutrition, 53, 503–508.
Engelmann, M. D. M., Davidsson, L., Sandström, B., Walczyk, T., Martı́nez, I., Santaella, M., Ros, G., & Periago, M. J. (1998). Content
Hurell, R. F., & Michaelsen, K. F. (1998). The influence of meat on and in vitro availability of Fe, Zn, Mg, Ca and P in homogenized
nonheme iron absorption in infants. Pediatric Research, 43(6), fish-based weaning foods after bone addition. Food Chemistry, 63,
768–773. 299–305.
Favier, A. E. (1993). Nutritional and clinicals factors affecting the McGregor, G., & Ani, C. (2001). A review of studies on the effect of
bioavailability of trace elements in humans. In U. Schlemmer (Ed.), iron deficiency on cognitive development in children. Journal of
Proceedings of the international conference bioavailability Õ93. Nutrition, 131, 649S–666S.
Nutritional, chemical and food processing implications of nutrient Michelson, K. F., Samuelson, G., Graham, T. W., & Lonnerdal, B.
availability (pp 202–212). Karlsruhe. (1994). Zinc intake, zinc status and growth in a longitudinal study
Fairweather-Tait, S. J. (1989). Iron in food and its availability. Acta of healthy Danish infants. Acta Paediatrics, 83, 1115.
Paediatrica Scandinavia Supplement, 361, 12–20. Miller, D. D., Schricker, B. R., Rasmussen, B. S., & Van Campen, D.
Galán, P., Cherouvier, F., Fernández-Ballart, J., Marti-Henneberg, (1981). An in vitro method for estimation of iron availability from
C., & Hercberg, S. (1990). Bioavailable iron density in French meals. American Journal of Clinical Nutrition, 34, 2248–2256.
and Spanish meals. European Journal of Clinical Nutrition, 44, Mulvihill, B., & Morrisey, P. A. (1998). Influence of the sulphydryl
157–163. content of animal proteins on in vitro bioavailability of non-haem
Garcı´a, R., Alegria, A., Barberá, R., Farré, R., & Lagarda, M. J. iron. Food Chemistry, 61, 1–7.
(1998). Dialyzability of iron zinc and copper of different types of Mulvihill, B., Kirwan, F. M., Morrisey, P. A., & Flynn, A. (1998).
infant formulas marketed in Spain. Biological Trace Element Effect of myofibrillar muscle proteins on the in vitro bioavailability
Research, 65, 7–17. of non-haem iron. International Journal of Food Sciences and
Gibson, R. S. (1994). Content and bioavailability of trace elements in Nutrition, 49, 187–192.
vegetarian diets. American Journal of Clinical Nutrition, 59, Musaiger, A. O., Ahmed, M. O., & Rao, M. V. (1998). Chemical
1223S–1232S. composition of some dishes of Oman. Food Chemistry, 61, 17–22.
Grajeta, H., Prescha, A., & Biernat, J. (2002). Fe, Ca and Mg contents Noël, L., Leblanc, J. C., & Guérin, T. (2003). Determination of several
in selected fast food products in Poland. Nahrung, 1, 7–10. elements in duplicate meals from catering establishments using
Hambidge, M. (2000). Human zinc deficiency. Journal of Nutrition, closed vessel microwave digestion with inductively coupled plasma
130, 1344S–1349S. mass spectrometry detection: Estimation of daily dietary intake.
Hallfrisch, J., Veillon, C., Patterson, K., Hill, A. D., Benn, I., Holiday, Food Additives and Contaminants, 20(1), 44–56.
B., Burns, R., Zhonnie, S., Price, F., & Sorenson, A. (2000). Bone- Periago, M. J., Ros, G., Rincón, F., & Martı́nez, C. (1997). Nutritional
related mineral content of water samples collected on the Navajo meaning of dietary fibre and phytic acid in meat-based homoge-
reservation. Toxicology, 149, 143–148. nised weaning foods. Food Research International, 30(3), 223–230.
Haas, J. D., & Brownlie, T. (2001). Iron deficiency and reduced work Pushpanjali y Khokhar, S. (1996). In vitro availability of iron and zinc
capacity: A critical review of the research to determine a causal from some Indian vegetarian diets: correlations with dietary fibre
relationship. Journal of Nutrition, 131, 676S–688S. and phytate. Food Chemistry, 56, 111–114.
 F. Cámara et al. / Food Chemistry 92 (2005) 481–489 489

Räsänen, L., & Ylönen, K. (1992). Food consumption and nutrient Tapiero, H., Townsend, D. M., & Tew, K. D. (2003). Trace elements in
intake of one- to two-year-old Finnish children. Acta Paediatrics, human physiology and pathology. Copper. Biomedicine and Phar-
81, 7–11. macotherapy, 57, 386–398.
Roig, M. J., Alegrı́a, A., Barberá, R., Farré, R., & Lagarda, M. J. Thapon, J. L., Audiot, V., Proteis, Y. & Sauveur, J. (1994).
(1999). Calcium bioavailability in human milk, cow milk and infant Présentation générale de lÕoeve. In J. L. Thapon & C. M. Bourgeois
formulas- comparison between dyalisis and solubility methods. (Coordonnateurs). LÕoeuf et les ovoproduits. (pp. 6–15, 21–24).
Food Chemistry, 65, 353–357. Lavoisier Tec & Doc, Paris.
Rosado, J. L. (1998). Zinc deficiency and its functional implications. Torelm, I., Danielsson, R., Appelqvist, L. A., & Bruce, A. (1997).
Salud Publica Mexicana, 40, 181–191. Variations in major nutrients and minerals due to interindividual
Salgueiro, M. J., Zubilaga, M. B., Lysionek, A. E., Caro, R. A., Weill, preparation of dishes from recipes. Journal of Food Composition
R., & Boccio, J. R. (2002). The role of zinc in the growth and and Analysis, 10, 14–27.
development of children. Nutrition, 18, 510–519. Van Dyck, K., Tas, S., Robberecht, H., & Deelstra, H. (1996). The
Salovaara, S., Sandberg, A. S., & Andlid, T. (2002). Organic acids influence of different food components on the in vitro availa-
influence iron uptake in the human epithelial cell line Caco-2. bility of iron, zinc and calcium from a composed meal.
Journal of Agricultural and Food Chemistry, 50, 6233–6238. International Journal of Food Science and Nutrition, 47, 499–
Sandström, B. (1992). Dose dependence of zinc and manganese 506.
absorption in man. Proceedings of the Nutrition Society, 51, 211–218. WHO/UNICEF (1998). International Nutritional Anemia Consulta-
Santaella, M., Martı́nez, I., Ros, G., & Periago, M. J. (1997). tive Group. Guidelines for the use of iron supplements to prevent and
Assessment of the role of meat cut on the Fe, Zn, Cu, Ca and treat iron deficiency Anemia. Washington, DC: International Life
Mg content and their in vitro availability in homogenised weaning Sciences Institute.
foods. Meat Science, 45(4), 473–483. Wienk, K. J. H., Marx, J. J. M., & Beynen, A. C. (1999). The concept
Sahuquillo, A., Barberá, R., & Farré, R. (2003). Bioaccessibility of of iron bioavailability and its assessment. European Journal of
calcium, iron and zinc from three legume samples. Nahrung, 47(6), Nutrition, 38, 51–75.
438–441. Yadav, S. K., & Sehgal, S. (2003). Effect of domestic processing and
Sebastiá, V., Barberá, R., Farré, R., & Lagarda, M. J. (2001). Effects of cooking methods on total, HCl extractable iron and in vitro
legume processing on calcium, iron and zinc contents and dialysa- availability of iron in bathua and fenugreek leaves. Nutrition and
bilities. Journal of Science Food and Agriculture, 81, 1180–1185. Health, 17, 61–63.
Southon, S., Bailey, A. L., Wright, A. J. A., Belsten, J., & Finglas, P. Yip, R. (2001). In B. A. Bowman & R. M. Russell (Eds.), Present
M. (1993). Micronutrient undernutrition in British schoolchildren. knowledge in nutrition (pp. 311–328). Washington, DC: ILSI
Proceedings of Nutrition Society, 52, 155–163. Press.