Bioaccessibility of some minerals in infant formulas

Abstract

To guarantee the adequate intake of nutrients a variety of food supplementation (including infant formulas) has been used to ensure the nutrition of infants. Considering that the total concentration of nutrients is not enough to determine whether the food provides all the nutritional needs, the objective of this study was to evaluate the total concentration and bioaccessibility of some elements in thirty commercial infant formulas consumed in Brazil. A standardized in vitro gastrointestinal digestion method was used to obtain the soluble fraction of each mineral, which was analyzed by ICP OES after microwave oxidative digestion to obtain the bioaccessibility values. The total concentration and the bioaccessibility of the elements varied considerably according to the sample type (traditional infant formulas, formulas for infants with gastrointestinal problems, formulas for premature and soy-based). The bioaccessibility values are 3–43% (Ca), 53–97% (Cu), 35–100% (Fe), 70–114% (K), 47–90% (Mg), 52–95% (P), 31–92% (Zn). In general, the total concentration values for the elements were higher than that declared by the manufacturers, also than the current legislation as well, regarding the DRI. Although these results, it is important to emphasize that the consumption of infant formulas can provide an adequate intake of minerals for the infants.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13197-021-05215-0.

Introduction

The World Health Organization recommends breastfeeding as the sole source of nutrients for infants up to 6 months of age, thus ensuring multiple benefits for children’s health. However, less than 40% of the world's children are exclusively breastfed (WHO 2017). It is important to highlight that inadequate or incorrect ingestion of nutrients may directly affect infant growth and development (Hoffman et al. 2003), therefore alternatives to replace breastfeeding are necessary, with a special emphasis on infant formulas.

The infant formulas should provide a sufficient amount of energy and nutrients to support the high growth rate during the first months of life. Even though efforts are being continuously done to study infant formulas, concerns about the nutritional adequacy of commercial products have still led to study the mineral content of infant foods. Regarding studies in Brazil, Vieira da Silva et al. (2013) evaluated the level of Ca, Cu, Fe, K, Mg, Mn, Na, and Zn in 10 commercially available milk-based infant formulas widely consumed in the country. The authors observed that the level in the starting formulas was, in general, in accordance with levels established by the Codex Alimentarius. Fernández-Menéndez et al. (2016) evaluated the Zn content and its speciation in human milk and in milk commercial formulas. The speciation analysis in formulas showed that the Zn was bound almost exclusively to low molecular weight ligands while in human milk the element is bound to specific high molecular weight proteins. In works with a more analytical focus, Gamela et al. (2019) developed analytical methods for the determination of Ca, Cu, Fe, K, Mg, Na, and Zn in infant formulas using high-resolution continuum source flame atomic absorption spectrometry.

There is already a consensus that the determination of the total concentration is not enough to affirm whether foods, when ingested, will provide an adequate amount of nutrients (Minekus et al. 2014). For this purpose, bioaccessibility and/or bioavailability studies must be carried out. Bioavailability can be evaluated by determining the amount of a nutrient or toxic species that is effectively absorbed by the body to be used in physiological functions or stored for future use. On the other hand, bioaccessibility studies may be carried out by determining the fraction of the element that is released from the food matrix and is soluble in the gut, and they are usually much easier to be evaluated compared to bioavailability studies. Nevertheless, bioaccessibility studies give very important information about the possible amount of a chemical species that will be absorbed during gastrointestinal digestion (do Nascimento da Silva et al. 2018). Such studies may be performed in vivo, with humans or animals, or using in vitro methods. In vitro assays are successfully performed to find conditions similar to those in vivo, once the methods are simple, accessible, and have good reproducibility (Thakur et al. 2020). Finally, the bioaccessibility is affected by some factors, such as the chemical form of the target compounds or elements, the interaction between the target specie and the food matrix, the synergy or antagonism with other compounds ingested concomitantly, etc. (Drago and Valencia 2004; Thakur et al. 2020).

Inorganic elements are present in biological systems and play important roles in maintaining the proper functioning of living organisms, participating as co-factors of enzymes, oxygen transport, combating the formation of free radicals, hormonal activity and structural organization of macromolecules (IOM 2006; Thakur et al. 2020). Considering that, it can be seen that the scientific community is still interested in bioaccessibility studies of inorganic elements in infant formulas. Roig et al. (1999) studied the total and the soluble concentration of Ca in human milk in soy-based infant formulas, as well as the dialyzable fraction. Drago and Valencia (2004) evaluated the effect of some components on the dialysability of Fe, Zn, and Ca. Perales et al. (2007) evaluated the Fe solubility, dialysability, and transport by Caco-2 cells in milk-based formulas, while Gomez et al. (2016) studied the Zn solubility and dialysability in soy-based, lactose-free and milk-based formulas, as well as the fractionation of Zn containing proteins. It is important to emphasize that in the last two works, the authors denominate the solubility and dialysability assays as bioaccessibility studies. Finally, do Nascimento da Silva et al. (2018) evaluated the bioaccessibility of Cu, Fe, Mg, Mn, and Zn in skimmed milk-based infant formulas.

These are practically all bioaccessibility studies with infant formulas found in the literature, and although there are these important contributions, they are still not enough to end bioaccessibility studies of metals and non-metals in infant formulas. Furthermore, there is no information on the bioaccessibility of K, Na, and P, elements that are also evaluated in the present work. Considering the importance of infant formulas in supplying nutrients, and that the total concentration does not give enough information about the nutritional value, this study aimed to evaluate the total concentration and bioaccessibility of Ca, Cu, Fe, K, Mg, Na, P and Zn in infant formulas (0–6 months), as well as checking its compliance with the reference values and with the current legislation.

Materials and methods

Samples

Infant formulas for infants aged from 0 to 6 months from different manufacturers were purchased at the local market in the city of Campinas, SP, Brazil. The samples were: traditional infant formulas (coded as A1, A2, A3, A4, A5); formulas for infants with gastrointestinal problems (coded as B1, B2, B3); formulas for preterm infants (named C1); and soy-based infant formulas (coded as D1, D2). For each sample, three batches were collected, except for samples C1, D1, and D2, which were analyzed in two batches, thus totaling 30 samples.

All samples were packaged in aluminum cans, labeled according to the manufacturer, in powder form, and within the shelf life. In addition to the commercial samples, a certified reference material, infant formula (NIST 1846), was analyzed to assess the accuracy and precision of the method. Table 1 shows the composition and the main ingredients of the infant formulas, according to product labels.

Table 1.

Labeled composition for the purchased infant formulas

Samples	Element (mg/kg)								Ingredients
	Ca	K	Mg	Na	Fe	Zn	P	Cu
A1	3600	4700	480	1450	51	54	2000	3.90	Whey, skimmed milk
A2	4100	4900	340	1350	60	39	2030	2.92	Whey protein, lactose, skimmed milk powder, GOS/FOS, ARA, fish oil, lecithin, and antioxidants
A3	5340	7370	380	2240	44	41	3810	3.84	Whey protein, lactose, lecithin, DHA, ARA, carotenoids, nucleotides, and tocopherols
A4	3500	5750	420	1360	62	52	2250	2.30	Lactose, skimmed milk powder, whey protein, DHA, ARA, corn syrup, and antioxidants
A5	3600	5650	510	2000	55	52	1900	4.60	Partially hydrolyzed whey protein, lactose, nucleotides, fish oil, ARA, and DHA
B1	3200	4900	630	1450	52	53	1750	4.00	Demineralized whey, skimmed milk, starch, lactose, DHA, ARA
B2	5560	5360	420	1840	57	57	3140	3.20	Skimmed milk powder, lactose, jatan gum
B3	4260	6340	410	1770	58	51	2740	3.19	Skimmed milk powder, starch, lactose, ARA, DHA, and antioxidants
C1	7600	6050	500	2690	100	63	4450	7.70	Whey protein, skimmed milk, ARA, DHA, nucleotides, lactose
D1	5300	5900	600	1850	53	67	3500	5.60	Soyproteinisolate, DHA, AR
D2	5320	5780	410	2430	52	60	3800	3.57	Soy protein isolate, corn syrup, DHA, ARA, antioxidants, and tocopherols

Traditional infant formulas (IF): A1, A2, A3, A4, A5;

IF for infants with gastrointestinal problems: B1, B2, B3;

IF for preterm infants: C1; Soy-based IF: D1, D2;

In addition to the ingredients above, all samples contain minerals, vitamins, emulsifier and vegetable oils;

DHA docosahexaenoic acid, ARA arachidonic acid

In vitro simulated gastrointestinal digestion

The gastrointestinal digestion was performed according to (Minekus et al. 2014), by simulating the digestive stages that occur in the mouth, stomach, and intestine. A mass of about 0.75 g of the samples was mixed with 5 mL water, followed by stirring. Thus, the samples were incubated for 2 min at 37 °C with 4 mL of a solution simulating the saliva (plus 1 mL of CaCl₂ 7.5 mM) in a mixing water bath (model Q226M1, Quimis, Diadema, São Paulo, Brazil); then, 9.1 mL of the gastric juice was added (plus 0.7 mL of CaCl₂ 2 mM), and the samples were kept in the mixing water bath for 2 h. After gastric digestion, 18.5 mL of the intestinal juice was added (plus 1.35 mL of CaCl₂ 9 mM), and the samples were incubated for 2 h more. CaCl₂ was added only during the digestion procedure to avoid the precipitation of the fluids. The resulting hydrolysates were centrifuged at 10,600 g for 25 min, the supernatant (chyme) was then collected and stored at fridge until the microwave digestion. Procedural blanks were run in parallel to check the presence of analytes in the reagents. The in vitro gastrointestinal digestion was carried out in triplicate for all samples. The reagents used in the fluids and the concentrations are in supplementary material (Table S1).

Ca, Cu, Fe, K, Mg, Na, P and Zn determination

The oxidative digestion was carried out in a microwave oven (model Ethos 1600, Milestone, Sorisole, Bergamo, Italy) as described by do Nascimento da Silva et al. (2013) with adaptations. For the total analyte content in the samples, approximately 0.5 g of the dried sample was weighed into Teflon® flasks and 3 mL of HNO₃ (concentrated, ultrapure), 1 mL of 31% (v/v) H₂O₂ and 4 mL of deionized water were added. To measure the bioaccessible fractions (or bioaccessibility), 5 mL of the supernatants obtained in the gastrointestinal digestion were transferred to mineralization flasks, to which 3 mL of concentrated HNO₃ and 1 mL of 30% v/v H₂O₂ were added. The following temperature program was performed: 4 min until reaching 80 °C; 6 min at 80 °C; 80–120 °C for 4 min; 6 min at 120 °C; 120–180 °C for 10 min; 5 min at 180 °C; and finally cooling down for 15 min. After cooling, the samples were transferred to 15 mL tubes and the volume was completed with deionized water and analyzed using an ICP OES instrument (Perkin Elmer, model 8300DV, Waltham, MA, EUA).

The instrumental parameters used were as follows: radio frequency power of 1400 W; plasma argon flow rate of 15 L/min; auxiliary argon flow rate of 0.7 L/min; and nebulization gas flow rate of 0.7 L/min; concentric nebulizer with cyclonic chamber; sample aspiration flow of 1 mL/min; delay time of 30 s; and integration time of 3 s. The wavelengths of analytes studied were: Ca (317.933 nm); Cu (324.752 nm); Fe (238.204 nm); K (766.474 nm); Mg (279.077 nm); Na (589.592 nm); P (214.914 nm); and Zn (213.857 nm). The instrument was operated in the axial configuration for most of the analytes, except for Ca, K, and Na, for which the radial configuration was used.

The results were expressed as the mean and standard deviation. Tukey’s test was used to compare the concentrations of the elements in different samples. Differences were considered statistically significant at P < 0.05.

Results and discussion

The total concentration of Ca, Cu, Fe, K, Mg, Na, P and Zn

The experimental conditions for the analysis by ICP OES were optimized by evaluating the recovery of analytes from a certified reference material. Table S2 shows the recovery values for Ca, Cu, Fe, K, Mg, Na, P, and Zn, which were ranging from 94 to 108%. The recoveries were considered satisfactory, i.e. the experimental conditions were adequate and provided quantitative recoveries for all elements. For analytes at concentration levels of mg/kg and μg/kg, recovery ranges of 80–115% and 70–125%, respectively, are acceptable (AOAC 2002).

The results considering the total concentration of the elements for the infant formulas studied are presented in Table 2. In general, the findings for total concentration in the present work are in agreement with data presented in the literature for infant formulas from different regions (Lesniewicz et al. 2010; Papachristodoulou et al. 2018; Vieira da Silva et al. 2013). Interestingly, most samples showed higher contents than those declared on the labels, for example, the values for Ca were about 17% higher than those reported by the manufacturers. In Brazil, there is a tolerance of ± 20% for the variation in the concentration for micronutrients concerning the labeled (BRASIL 2003); however, it is allowed values greater than 20% if the manufacturer can justify it. Besides that, the formulas for preterm infants (C1) showed the highest values for Ca (9433 mg/kg), while those for infants with gastrointestinal problems (B1) the lowest (3913 mg/kg), as expected. However, the values are 24 and 22% higher than those declared on the labels for samples C1 and B1, respectively.

Table 2.

The total measured concentration of the analytes in the infant formulas (in mg/kg)

Infant formula
Analyte	A1	A2	A3	A4	A5	B1	B2	B3	C1	D1	D2
Ca	4140 ± 57de	4631 ± 350de	6849 ± 24b	4345 ± 397de	4045 ± 270de	3738 ± 219e	7490 ± 1193b	5230 ± 30bc	9434 ± 370a	6436 ± 11b	7606 ± 220b
Cu	4.6 ± 0.2b	3.2 ± 0.1de	3.8 ± 0.1c	3.6 ± 0.3 cd	4.1 ± 0.1c	4.0 ± 0.1c	3.0 ± 0.1e	3.0 ± 0.1e	7.8 ± 0.4a	5.0 ± 0.1b	3.2 ± 0.1de
Fe	79 ± 8b	78 ± 3b	63 ± 4b	68 ± 5b	73 ± 13b	57 ± 3b	69 ± 7b	78 ± 10b	122 ± 29a	67 ± 3b	111 ± 3a
K	4672 ± 64d	4680 ± 21d	8953 ± 80a	7364 ± 379ab	5562 ± 177bcd	4741 ± 221bcd	6126 ± 238bc	6117 ± 225bc	6706 ± 29abc	6652 ± 152abc	7039 ± 54abc
Mg	526 ± 34bc	354 ± 15e	483 ± 3 cd	484 ± 21 cd	523 ± 26bc	709 ± 36a	505 ± 18bcd	474 ± 5 cd	575 ± 13b	660 ± 13a	445 ± 10d
Na	1783 ± 118d	1792 ± 39d	3077 ± 4a	1738 ± 8d	2254 ± 65b	1964 ± 22 cd	2146 ± 98bc	2307 ± 49b	3013 ± 271a	2171 ± 70bc	2811 ± 120a
P	2257 ± 102de	2751 ± 135d	5236 ± 12a	3441 ± 267c	2441 ± 50de	2114 ± 30e	4335 ± 194b	4422 ± 91b	5642 ± 604a	4295 ± 81b	4967 ± 144ab
Zn	65 ± 1b	46 ± 3d	50 ± 1 cd	63 ± 1b	55 ± 3c	55 ± 4c	64 ± 4b	47 ± 1d	67 ± 1ab	75 ± 1a	64 ± 1b

The results are expressed as mean ± standard deviation (n = 3);

Averages followed by the same lowercase letter in the row don’t differ significantly by the Tukey test (p < 0.05);

Traditional infant formulas (IF): A1, A2, A3, A4, A5;

IF for infants with gastrointestinal problems: B1, B2, B3;

IF for preterm infants: C1; Soy-based IF: D1, D2

The levels of Fe, Na, P, and Zn were, on average, 25, 18, 24, and 11% higher than those reported by the manufacturers. The greater discrepancy was observed for the Fe content in the soy-based infant formula (D2), which showed a value that is 53% higher than labeled. It is worth mentioning that Fe-fortified infant foods are important to meet high dietary requirements for children, especially considering the importance of Fe during the developmental stage of the brain (WHO 2017). On the other hand, excess iron supplementation in preterm infants might negatively impact brain development or even induce brain injury, besides higher risk of impaired growth and infection (Wu et al. 2019). For K and Mg, the samples A1 and A2 presented values similar to those declared on the labels. Finally, good accordance between the finds and the values labeled was observed for Cu for almost all samples.

Considering the literature, it was already been reported discrepancies in values for inorganic elements in infant formulas regarding the labeled. Vieira da Silva et al. (2013) observed differences of 41% for Cu, 60% for Zn, and 73% for Mn. It is important to highlight that the small concentrations of these elements in the food matrix could be susceptible to large variations in the content of these elements in the raw materials. Papachristodoulou et al. (2018) found values up to 100% higher than those labeled for Ca in some products, up to 90% for Zn, and 70% for Fe. Lesniewicz et al. (2010) obtained values that differed by 10–20% than the declared by the manufacturer, and also observed values much higher for Mn (at least 3 times higher than labeled).

It is interesting that the values for the total concentration are usually above the concentration in human milk and below those for cow’s milk, as can be seen in Table 3. Most infant formulas are based on cow’s milk, i.e., the milk powder (usually skimmed) is mixed with other ingredients diluting the milk components, then the concentrations are lower in the formulas. Still, the concentration of these elements must be higher than in human milk since their bioavailability in cow’s milk is much lower than in human milk (IOM 2006; Koletzko et al. 2005). An exception here is the concentration for Fe in the infant formulas, which proved to be much higher than in cow’s milk. This fact is due to the addition of Fe salts in large amounts, mainly as ferrous sulfate, aiming a better absorption of this element, which is very little bioavailable in cow’s milk or vegetable sources (Koletzko et al. 2005).

Table 3.

The total concentration of the analytes in the infant formulas studied compared to human and cow’s milk (in mg/L)

Analyte	Infant formula											FAO1		Institute of medicine2	Pietrzak-Fiecko et al. (2020)
	A1	A2	A3	A4	A5	B1	B2	B3	C1	D1	D2	Human milk	Cow milk	Human milk	Human milk	Cow milk
Ca	542	634	910	600	542	486	989	706	1519	850	1016	330	1156	280	276	1198
Cu	0.60	0.44	0.51	0.50	0.55	0.52	0.40	0.41	1.26	0.66	0.43	1.0	–	0.25	–	–
Fe	10.3	10.7	8.4	9.4	9.8	7.4	9.1	10.5	19.6	8.8	14.8	–	1.0	0.3	0.2	0.8
K	612	641	1190	1016	745	616	809	826	1080	878	940	526	1496	525	713	1479
Mg	69	48	64	67	70	92	67	64	93	87	59	31	114	35	38	126
Na	234	246	409	240	302	255	283	311	485	287	375	175	433	180	159	439
P	296	377	696	475	327	275	572	597	908	567	663	144	939	140	–	–
Zn	8.5	6.3	6.6	8.7	7.4	7.2	8.4	6.3	10.8	9.9	8.5	2.1	4.1	1.2	0.46	6.2

Traditional infant formulas (IF): A1, A2, A3, A4, A5;

IF for infants with gastrointestinal problems: B1, B2, B3;

IF for preterm infants: C1; Soy-based IF: D1, D2;

⁽¹⁾FAO (FAO 2013). ⁽²⁾Institute of Medicine (IOM 1996)

The total concentration of analytes of the different batches of infant formulas is shown in Table S3. Overall, there was a statistical difference between batches, probably due to the ingredients used in infant formula formulations.

Daily intake of Ca, Cu, Fe, K, Mg, Na, P and Zn from infant formulas

From the total concentration obtained experimentally, it was calculated the contribution to the daily intake of each element from the infant formulas evaluated in the present study. The daily consumption of infant formulas was defined according to the label recommendations, once exists distinct preparation modes among the analyzed brands and types of infant formulas (0–6 months).

The values are presented in Table 4, and ranged (in mg/day) between 406 to 1018 for Ca, 484 to 945 for K, 35 to 70 for Mg, 173 to 336 for Na, 6 to 15 for Fe, 5 to 8 for Zn, 0.3 to 0.8 for Cu, and 195 to 554 for P. Similar values were found by Papachristodoulou et al. (2018) in formulas marketed in Greece (in mg/day), for Ca (348–717), K (230–610), Fe (3.2–10.5) and Zn (4.6–9.0). Considering the Dietary Reference Intakes (DRIs) (Table 4) by the Food and Nutrition Board (IOM 2006), the daily intake of the elements from the studied formulas is greater than the adequate for infants up to 6 months. However, it is not that alarming, as many of these elements usually are not fully soluble and show low bioavailability from infant formulas. In addition, given the lower bioavailability of Ca from infant formulas than from cows’ milk, a Ca intake up to approximately 700 mg per day from infant formula is recommended by the European Food Safety Authority—EFSA (Koletzko et al. 2005), considering infants up to 6 months of age. Still, EFSA emphasizes that in view of possible untoward effects of unbalanced ratios between Ca and P contents, the calcium-phosphorus-ratio (weight/weight) should not be less than 1:1 and not be greater than 2:1.

Table 4.

Daily intake of the elements from infant formulas and the adequate intake for infants up to 6 months of age (mg/day)

Analyte	“Traditional” IF		IF for infants with problems gastrointestinal		IF for preterm infants		Soy based IF		Adequate intake (1, 2)
	Min	Max	Min	Max	Min	Max	Min	Max
Ca	406	722	675	815	963	1018	675	815	210
K	484	945	687	743	702	706	687	743	400
Mg	35	53	46	70	59	61	46	70	30
Na	173	323	223	304	296	336	223	304	120
Fe	6	9	7	12	11	15	7	12	0.27
Zn	5	7	7	8	7	7	7	8	2
Cu	0.3	0.5	0.3	0.5	0.8	0.8	0.3	0.5	0.2
P	195	479	387	463	476	554	387	436	100

Min: minimum; Max: maximum. ⁽¹⁾ IOM 2006. ⁽²⁾ Koletzko et al. 2005

The results obtained in the present study were also compared with data from the Brazilian Health Regulatory Agency (BRASIL 2011) and from the Codex Alimentarius Commission (CAC 2007), which are shown in Table S4. The data for both references include the minimum quantity for all the elements in 100 kcal of the formulas, and the maximum only for K, Na, Fe, and P. All samples showed the amount of the elements in 100 kcal above the minimum and below the maximum quantity required by legislation, except for Fe. Samples A1, A2, A4, A5, B2, B3, C1, and D2 showed values above the maximum allowed for Fe (1.3 mg per 100 kcal). However, according to EFSA, in the case of soy-based formulas (samples D1 and D2), the maximum for Fe should be 2.0 mg per 100 kcal due to its very low bioavailability in this type of food matrix (Koletzko et al. 2005).

The bioaccessibility of Ca, Cu, Fe, K, Mg, Na, P, and Zn from infant formulas

Bioaccessibility studies are good to give particularly important information about the possible amount of a chemical species that will be absorbed during gastrointestinal digestion. Besides, it is helpful to compare different food sources and determine the effect of inhibitors and enhancers on the bioaccessibility of elements. In the present study, it was determined the bioaccessibility of Ca, Cu, Fe, K, Mg, Na, P, and Zn, and the results are shown in Table 5. The bioaccessibility of Na was not determined because this element was present in large quantities in the reagents used to simulate the in vitro gastrointestinal digestion; thus, it was impossible to determine the small amount of the element released from the sample to the chyme.

Table 5.

Bioaccessibility of the analytes from the infant formulas (in percentage)

Infant formula
Analyte	A1	A2	A3	A4	A5	B1	B2	B3	C1	D1	D2
Ca	10 ± 1	10 ± 2	22 ± 1	3 ± 2	24 ± 3	3 ± 1	18 ± 2	3 ± 1	43 ± 1	5 ± 1	23 ± 1
K	97 ± 1	77 ± 5	89 ± 5	74 ± 3	102 ± 24	101 ± 15	70 ± 3	114 ± 15	83 ± 11	96 ± 26	72 ± 5
Mg	71 ± 5	50 ± 3	53 ± 2	47 ± 6	90 ± 8	51 ± 3	50 ± 5	62 ± 6	63 ± 6	72 ± 3	49 ± 3
Na	nd	nd	nd	nd	nd	nd	nd	nd	nd	nd	nd
Fe	69 ± 6	100 ± 33	65 ± 4	78 ± 5	35 ± 9	52 ± 4	38 ± 4	81 ± 9	69 ± 5	83 ± 4	35 ± 6
Zn	60 ± 2	51 ± 1	48 ± 2	50 ± 5	92 ± 12	37 ± 8	49 ± 3	58 ± 3	73 ± 7	78 ± 4	31 ± 2
Cu	96 ± 5	96 ± 10	97 ± 4	78 ± 4	98 ± 5	53 ± 5	67 ± 3	81 ± 3	80 ± 4	97 ± 8	82 ± 6
P	52 ± 6	60 ± 1	52 ± 5	73 ± 4	95 ± 9	93 ± 1	79 ± 8	90 ± 7	92 ± 4	93 ± 6	78 ± 6

The results are expressed as mean ± standard deviation (n = 3);

Traditional infant formulas (IF): A1, A2, A3, A4, A5;

IF for infants with gastrointestinal problems: B1, B2, B3;

IF for preterm infants: C1; Soy-based IF: D1, D2

Calcium bioaccessibility values were between 3 and 24%, except for IF for preterm infants, which showed a value of 43%. It is worth mentioning that this formula exhibited the highest total Ca concentration (9434 mg/kg) when compared to all formulas studied. The Ca concentration in infant formulas can establish the free Ca content, once the higher the amount of total Ca, the greater its bioavailability in this type of food matrix. The correlation between the total concentration and bioaccessibility values showed r = 0.73. Also, similar results were observed in the literature (Drago and Valencia, 2004; Roig et al. 1999). Concerning the elements Mg and K, they are known to be bioavailable and well utilized for animals (about 80%), i.e., these elements are also more soluble during the gastrointestinal conditions (Ammerman et al.1995). Thus, these Mg and K bioaccessibility values higher than for Ca are understandable.

The bioaccessibility values for Fe and Zn are, on average, lower than for the other elements. It is well known that the bioaccessibility of essential minerals is highly dependent on the food matrix. The sample D2, which is a soy-based formula, presented the lowest bioaccessible fraction of Fe (35%) and Zn (31%), probably due to the formation of insoluble complexes under gastrointestinal conditions. The presence of some food components, known as antinutritional factors, such as phytates and phenols present in soybeans, may decrease, or prevent the absorption of some essential elements, like Fe and Zn (Shilpa and Jyothi Lakshmi 2012). These antinutritional factors may bind with cations leading to the formation of insoluble complexes that precipitate during intestinal digestion (Gomez et al. 2016; Liang et al. 2009). Although formula D1 is also soy-based, the bioaccessibility of Fe and Zn was higher than in formula D2. The Zn bioaccessibility may be affected by other factors, such as different compounds used in the supplementation of infant formulas, amount and type of proteins, and the use of milk protein hydrolysates (Drago and Valencia 2004). The latter may explain the greater bioaccessibility in sample A5 when compared to the other formulas.

For Cu, the bioaccessibility values were higher than 80%, except for the formulas B1 and B2, destined to infants with gastrointestinal problems. The samples D1 and D2 did not show an inhibitory effect on Cu solubility, although the presence of phytates. In contrast to Fe and Zn, Cu-phytate complexes are more soluble at the pH of the gastrointestinal tract (Carbonaro et al. 2002; do Nascimento da Silva et al. 2018). Furthermore, the P bioavailability from food is usually very high with the exception of phytate phosphorus in plant sources, such as grains, legumes, and seeds, which is poorly digested (Ammerman et al. 1995). However, the high bioavailability for this element in all samples, including the soy-based formulas is probably due to the addition of some phosphate salts, like phosphates of Ca, Mg, K.

Finally, it is known that the bioaccessibility of minerals in infant formulas is lower than that observed in breast milk. Breast milk contains enzymes (amylase and lipase) that aid digestion; on the other hand, in infant formulas, the addition of some inorganic salts can lead the elements precipitation during the digestion, and the presence of various antinutritional factors derived from the ingredients can also decrease the solubility of some elements (Gomez et al. 2016; Hallberg et al. 1992; Thakur et al. 2020). However, the high bioaccessibility values for the infant formulas studied in the present work can be considered satisfactory, since the elements show to be quite soluble; wherefore, it could result in suitable absorption by the infants.

Conclusion

It was possible to evaluate the total concentration, the daily intake (for infants 0–6 mo), and the bioaccessibility of Ca, Cu, Fe, K, Mg, Na, P, and Zn from infant formulas marketed in Brazil. In this study it was observed that the total concentration of the elements measured was always higher than the values labeled; however, it is allowed by the legislation if the manufacturer can justify it. Also, the values for total concentration are usually above the concentration in human milk and below those for cow’s milk. Considering the daily intake, the study showed values higher than that recommended for infants 0–6 mo. However, it is not that alarming, as many of these elements usually are not fully soluble and show low bioavailability from infant formulas. Moreover, when the legislation for infant formulas is considered, almost all samples showed the amount of the elements (in 100 kcal) above the minimum and below the maximum quantity required by legislation. Finally, the bioaccessibility results showed that Ca is the element less soluble in gastrointestinal digestion, and K is the most soluble. Additionally, the values for almost all elements and samples are satisfactory, considering that for the element to be absorbed, first it must be soluble during the digestion.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Methodology, MRM, VL, and ENS; formal analysis, MRM and VLS; investigation, MRM, VLS, and ENS; data curation, MRM, VLS, and ENS; writing, MRM and ENS; manuscript revision, MRM, ENS, SC, and HTG; supervision, HTG and SC; and funding acquisition, SC and HTG

Funding

The authors gratefully acknowledge the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), and the National Institute of Advanced Analytical Science and Technology (INCTAA) (CNPq—Process n^o 573894/2008–6 and FAPESP Process n^o 2008/57808–1) for the financial support.

Availability of data and material

All data generated or analysed during this study are included in this published article [and its supplementary information files].

Declarations

Conflict of interest

The authors declare no conflict of interest.