Do cooking techniques influence copper bioaccesibility in foods after in vitro digestion/fermentation in adults and children?

Do cooking techniques influence copper bioaccesibility in foods after in vitro digestion/fermentation in adults and children? Úrsula García-Conde a , Miguel Navarro-Moreno a,b , Beatriz Navajas-Porras a,b , Daniel Hinojosa-Nogueira a,b , Adriana Delgado-Osorio a,b , Silvia Pastoriza a , Dafni Moriki c , Konstantinos Douros c , Miguel Navarro-Alarc´ on a,b,* , Jos´ e ´ Angel Rufi´ an-Henares a,b,d a Departamento de Nutrici´ on y Bromatología, Facultad de Farmacia, Universidad de Granada, Spain b Instituto de Nutrici´ on y Tecnología de los Alimentos, INyTA (IBS), Universidad de Granada, Spain c Pediatric Allergy and Respiratory Unit, 3rd Department of Pediatrics, “Attikon” University Hospital, National and Kapodistrian University of Athens, School of Medicine, 11527 Athens, Greece d Instituto de Investigaci´ on Biosanitaria (IBS), Universidad de Granada, Granada, Spain ARTICLE INFO Keywords: Copper bioaccessibility in the small and large intestine Raw and cooked foods Celiac children Obese children Allergic children ABSTRACT Cu is essential for the growth and organism health. Classically, its available fraction has been studied by in vitro digestion studies as a measure of bioaccessibility of Cu (Cu-BA). In this work we applied a novel in vitro digestion/fermentation method to multiple foods subjected to different home cooking techniques (raw form vs. frying, roasting, toasting, boiling and grilling) by metabolization with faecal inoculate from healthy adults (HEAD), and healthy children (HE-CH) and sick children (children with gluten related disorders, GRD-CH; children with obesity, OB-CH; and children with allergy/intolerance to cow’s milk proteins, AICM-CH). In raw and cooked foods the bioaccessibility of Cu in the small intestine (Cu-BASI) was higher vs. that in the Cu bioaccessibility in the large intestine (Cu-BALI) (30.8 ±15.4 and 28.2 ±14.7 vs. 18.4 ±21.2 and 22.8 ±22.1 %, respectively; p < 0.001). Total Cu-BA in cooked foods (51.0 ±24.4 %) was higher than that in raw foods (49.0 ±25.1 %). In cereals, total Cu-BA was higher in the raw form, as well as in whole grains and those with gluten (p <0.05). In vegetables, the most drastic cooking techniques (roasting and frying) show higher values of total Cu-BA compared to raw form and boiling. The Cu-BALI in HE-CH (37.7 ±23.7 %) is higher than that determined in HE-AD (14.1 ±18.5 %) and sick children (GRD-CH: 14.6 ±19.8; OB-CH: 15.5 ±17.8; and AICM-CH: 26.9 ± 19.3 %; p <0.001). In conclusion, cooking techniques influence Cu-BA depending on the food group. Total Cu- BA, as well as that determined in the large and small intestine varied according to the category, group and specific foods, which is related to their different composition and species of the element. In healthy children, total Cu-BALI is higher than in adults and celiac, obese and allergic children probably due to growth requirements and specific microbiota. 1. Introduction Cu is an essential element in growth, immune system. bone mineralization, or iron metabolism (Arreondo Olguin et al., 2024). Therefore, its deficiency is related to the appearance of anaemias, skin depigmentation, bone malformations and alterations in cholesterol metabolism (Arreondo Olguin et al., 2024; Bost et al., 2016Bost et al., 2016). This element is absorbed in the duodenum through an active transport mechanism (12–60 %; Henriksen and Arnesen, 2023). Multiple studies have evaluated the bioaccessibility of Cu (Cu-BA) in food, after an in vitro process of simulating oral-gastrointestinal digestion, mostly in the small intestine, evaluating the levels of Cu that remained solubilized (BellailFn et al., 2013; Erdemir and Gucer, 2015; Omar et al., 2015; Ortiz and C´ amara-Martos, 2018; Regula et al., 2018; Kumari and Platel, 2022; Iaquinta et al., 2023; Zhang et al., 2023; Nascimento et al., 2024) or dialyzed (Feitosa et al., 2018).However, the fraction of non-bioavailable Cu in the small intestine that is obtained as residue and that would reach the colon, would be subjected to the action of fermenting microorganisms of the colonic microbiota, which would originate the genesis and modification of different components, which can influence its * Corresponding author at: Departamento de Nutrici´ on y Bromatología, Facultad de Farmacia, Universidad de Granada, Spain. E-mail address: [email protected] (M. Navarro-Alarc´ on). Contents lists available at ScienceDirect Food Research International journal homepage: www.elsevier.com/locate/foodres https://doi.org/10.1016/j.foodres.2024.115238 Received 16 July 2024; Received in revised form 30 September 2024; Accepted 18 October 2024 Food Research International 197 (2024) 115238 Available online 20 October 2024 0963-9969/© 2024 Universidad de Granada. Published by Elsevier Ltd. This is an open access article under the CC BY-NC license ( http://creativecommons.org/licenses/by-nc/4.0/ ). bioaccessibility in the colon (Cai et al., 2017; Nissen et al., 2022). In vivo studies require long times for their development and exhaustive compliance with strict conditions by the subjects, being expensive and having ethical limitations (Ting et al., 2015). However, in vitro studies are more widely used to evaluate the bioaccessibility of Cu and the possible factors that may affect it, constituting an initial screening, for subsequent extrapolation to more representative in vivo studies of human physiology. To date no studies have been carried out on a large number of food samples of both vegetable and animal origin to address the influence of different methods of home cooking such as frying, roasting, toasting, boiling and grilling, compared to their raw state, on the Cu-BA in both the small and large intestine. Zhang et al. (2023) studied the effects on the bioaccessibility of Cu in the small intestine (Cu-BASI) of boiling, steaming, baking and frying, but only in 12 species of different food ingredients. Feitosa et al. (2018) evaluated the Cu-BA also in the small intestine, in black beans subjected to soaking treatments, exerted by the anti-nutritive compounds present in the raw form (phytates, polyphenols and tannins). Erdemir et al. (2019) studied the Cu-BA in rice and vegetable-based baby foods using the in vitro method of gastrointestinal digestion. Kumari and Platel, (2022) evaluated the effect of different thermal processing techniques (microwave cooking, pressure cooking and simmering) in different cereal and legume samples on the in vitro Cu-BA. He et al. (2010) evaluated the influence on the Cu-BA in the small intestine in two species of marine fish after boiling, steaming, frying and grilling, concluding that it decreased with processing. Others (Cai et al., 2017) evaluated the influence of the human gut microbiota on the bioavailability and Cu-BA, but only in seven classes of leafy vegetables using different in vitro digestion methods. On the other hand, others researchers have reported that the way of cooking vegetables (dark purple eggplant and red-skinned onion) could differently affect the phenolic profiles of foods and their impact on human colon microbiota (Nissen et al., 2022; Cattivelli et al., 2023; respectively). Additionally, it is also intended to study the possible influence that different pathologies in children (celiac disease, obesity and allergy to cow’s milk proteins,) may have on the bioaccessibility of Cu in the large intestine (Cu-BALI). Cu deficiency is a commonly unreported complication of celiac disease (Halfdanarson et al., 2009; Idris et al., 2019) and foods made with gluten-free cereals are of lower nutritional value (Pedron et al., 2016; García-Conde et al., 2024) and Cu content (rice) (Punshon and Jackson, 2018). Meta-analysis studies have referred to the inverse association between dietary Cu and metabolic syndrome (Ding et al., 2022). In patients with morbid obesity, Cu deficiency has been reported (Boullata et al., 2017). Moreover, children with gastrointestinal allergy induced by dietary proteins are at a higher risk of presenting Cu deficiency overall in children who do not consume hypoallergenic milk formula or multivitamin and multi-mineral supplements (Meyer et al., 2014). Additionally, Cu-BA in children could be conditioned by the stage of life. In childhood, despite a lower dietary and energy intake than in adulthood, the Cu-BA may be increased to cope with the growth characteristic of this stage of life. Considering the above and using the in vitro digestion/fermentation method, the aim of the study is to determine the total amount of Cu and its distribution between bioaccessible fractions in the small and large intestine (Cu-BASI and Cu-BALI, respectively) and non-bioaccessible fractions in different foods cooked with different home cooking techniques (frying, raw form, roasting, toasting, boiling and grilling). Additionally we evaluate if the stage of life (adulthood vs. infancy) or the presence of pathologies in childhood (celiac disease, obesity and allergy) influence Cu-BALI. 2. Materials and methods 2.1. Food samples and home culinary techniques A total of 54 different foods (Table S1) corresponding to 159 food samples were analysed (García-Conde et al., 2023, 2024), namely: a) plant foods: nuts (mixed nuts and peanuts), cereals (bread, whole wheat bread, penne, whole grain penne, rice, whole grain rice, cookies, whole grain cookies, breakfast cereal and whole grain breakfast cereal), fruits (apple, banana, orange, grapes, plum, peach and olive), vegetables (zucchini, capsicum, carrot, potato, sweet potato, eggplant, onion, cauliflower, spinach, garlic, tomato, cabbage and lettuce), legumes (lentils and kidney beans), oils (sunflower and olive oil), beverages and infusions (beer, red wine, cola (regular cola and light cola; and coffee (regular coffee and instant coffee) and other plant foods (dark chocolate and hazelnut spread); b) foods of animal origin: meats (chicken, beef, lamb and pork), fish (salmon and cod), dairy products (butter, milk (cow milk and fermented cow milk) and gouda cheese) and whole egg. The foods were submitted to different culinary treatments: frying, raw form, roasting, toasting (only for bread and whole-grain bread), boiling and grilling (Navajas-Porras et al., 2020). Some foods were analysed only in their raw form, since they are usually consumed in this state. In general, the amounts, cooking times as well as the food/medium ratios were adapted from those established by other researchers (Ramírez-Anaya et al., 2015). Extra virgin olive oil was used for frying and grilling. For frying, an oil/food ratio of 5/1 was used with heating at 180˚C for 8 min. For grilling, an oil:food ratio of 0.5/1 was used with heating between 220 and 250˚C for 3 min. In broiling the food was heated at 180˚C for 10 min. In boiling the water:food ratio was 5/1 with heating at 100˚C for 20 min. For roasting, the Grunkel TS140H toaster was used for 3 min at its fourth level for 3 min at 900 W. In general, the cooking times as well as the food/medium ratios were adapted from those established by other researchers (Ramírez-Anaya et al., 2015). The ratio of samples and the treatments to which they were subjected and the utensils used in the preparation are listed elsewhere (Navajas- Porras et al., 2020). After preparation, the samples were homogenized and stored in a nitrogen atmosphere at −80˚C, until their use in the analysis. 2.2. Stool samples from adults and children Healthy adults (HE-AD) were compared with healthy children (HECH). Additionally, unhealthy children (with gluten related disorders, GRD-CH; obesity, OB-CH; and allergy/intolerance to cow’s milk proteins, AICM-CH) were also compared with HE-CH. For each group, 10 individuals were considered, whose characteristics were previously described (Dello Russo et al., 2022; García-Conde et al., 2024). Enough faecal material was collected to perform the in vitro fermentation procedure. The informed consent document was signed by adults or legal representatives of children. That form included all of the information of the study as well as the exclusion and inclusion criteria. The study was conducted according to the guidelines of the Declaration of Helsinki. It was approved by the Ethics Committee of the University of Granada (protocol code 1080/CEIH/2020). Also, in Greece by the Scientific Committee of the University Hospital of Ioannina (Protocol number 382, Date 4 June 2020, Decision number 10/3–6-2020), the Scientific Committee of the University Hospital “Attikon” (Decision Number: 546/1–10-2020), and the Scientific Committee of the University Hospital of Patras (Decision Number: 360/22–7-20). Stools were deposited in a 0.6 L sterile recipient. Immediately after collection, two anaerobic gas generator sachets (AnaeroGen Compact, Thermo Scientific) were included in the sterile recipient, avoiding direct contact with the stool. The recipients were stored in a home refrigerator and transported to the Greek hospitals in a cooler bag within 24 h. Upon arriving at the laboratories, the containers were opened inside an anaerobic chamber (80 % N 2 , 10 % CO 2 and 10 % H 2 ) and the feces were Ú. García-Conde et al. Food Research International 197 (2024) 115238 2 mixed with a water:glycerol solution (20 % vol/vol) and stored at −80 ◦C. The stools were then shipped to the Spanish laboratory in dry ice and again stored at −80 ◦C for no longer than 2 months. 2.3. In vitro digestion and fermentation method In order to mimic physiological processes in the human gut, all food samples were subjected to an in vitro digestion process (P´ erez-Burillo et al., 2018a, 2021) with oral, gastric and intestinal phase. Food was added to falcon tubes together with simulated salivary fluid (1:1, w/v) composed of salts and α -amylase (75 U/mL). The mix was kept at 37 ◦C for 2 min in oscillation. Right after, 10 mL of simulated gastric fluid was added, mimicking the gastric juices content in salts and pepsin (2000 U/ mL). The mix was kept at 37 ◦C for 2 h, at pH 3 in oscillation. Finally, 20 mL of simulated intestinal fluid was added, with equal content in salts, bile salts, and enzymes (here, we used 67.2 mg/mL pancreatine) than the intestinal juices. The mix was kept at 37 ◦C for 2 h, at pH 7, in oscillation. Once the intestinal phase was finished, tubes were kept in ice to stop enzymatic reactions and centrifuged at 3500 rpm for 10 min. The supernatant, which represents the fraction available for absorption in the small intestine, was stored in 1 mL tubes at −80 ◦C until analysis. Next, an in vitro fermentation was carried out using faecal samples from donors (HE-AD, HE-CH, GRD-CH, OB-CH and AICM-CH). For this procedure, 0.5 g of the solid pellet previously obtained after in vitro oralgastrointestinal digestion plus 10 % of the digestion supernatant of foods was fermented under oscillation at 37 ◦C for 20 h in an anaerobic workstation (80 % N 2 , 10 % CO 2 and 10 % H 2 ) as described in P´ erez- Burillo et al. (2021). The fermentation medium included cysteine 312 mg/L, resazurin 0.1 % v/v, peptone 14 g/L and hydrogen sulphide 312 mg/L. 7.5 mL of this fermentation medium were added to the fermentation tube. Inoculum was made from faecal material above described. Each faecal material was mixed with phosphate buffer saline pH =7.0 (at 33 % concentration). Two mL of inoculum were added to the fermentation tube (each food sample was fermented 3 times, once for each digestion). Then, in order to reach anaerobic conditions nitrogen was bubbled, leaving a transparent solution (contrary to the pink color obtained under the presence of oxygen). Once the in vitro fermentation was finished, tubes were kept in ice to stop microbial reactions and thereafter centrifuged at 3500 rpm for 10 min. The supernatant, which represents the fraction available for absorption in the large intestine, was stored in 1 mL tubes at −80 ◦C until analysis. The solid pellet, which represents the fraction not fermented and excreted with faeces, was also stored in 1 mL tubes at −80 ◦C until analysis. After in vitro digestionfermentation, three different fractions were obtained: a) digestion supernatant as fraction available for Cu absorption at the small intestine, expressed as Cu-BASI; b) fermentation supernatant as fraction available for Cu absorption at the large intestine, expressed as Cu-BALI; c) fermentation solid residue as Cu fraction not available for absorption and excreted with faeces, expressed as non-bioaccesible Cu fraction. The sum of the Cu levels corresponding to the 3 fractions analysed allowed the determination of the average amount of Cu present in the analysed foods. The reported digestion-fermentation method was performed in triplicate for any food sample analyzed. Results were expressed as mg Cu/kg of food (fresh weight). 2.4. Food mineralisation and analysis of Cu For the determination of Cu content in the digestion and fermentation supernatants of foodstuffs studied, a previously described procedure was applied (García-Conde et al. 2023, 2024). Specifically, 250 mg were weighed into 25 mL falcon tubes. Then, 3.5 mL of 33 % HNO 3 was added and the samples were kept for 96 h for complete mineralisation. The mineralised samples were diluted to 25 mL with reagent grade water (Milli-Q water prepared with the R015 Milli-Q system, Waters, Medford, MA, USA) to obtain the analytical dissolution. However, for Cu determination in fermentation solids and in some fermentation supernatants from foods (which were not completely mineralised by the cold mineralisation technique described above) between 50–100 mg were weighed into borosilicate tubes, followed by the addition of 3 mL of 65 % HNO 3 . On the other hand, 0.5 mL of 65 % HNO 3 together with 2.5 ml of Milli-Q water was added to the used microwave teflon digestion vessels, into which the borosilicate tubes with the samples to be mineralised were placed. Teflon digestion vessels were used and placed in the rotor of the microwave digester (Multiwave 5000 with Rotor 24HVT50, Anton Parr GmbH, Graz, Austria). Then, a previously optimized time–temperature programme was applied for sample mineralization (García-Conde et al., 2023). Next, in order to prepare the final analytical dissolution, the mineralized samples were diluted to 40 mL with reagent grade water. In each of the batches, four blanks were prepared with the reagents used in the mineralization process described. Cu concentrations were measured in the analytical solution by inductively coupled plasma mass spectrometry technique (ICP-MS/MS; Agilent 8900, Agilent Technologies Inc., Santa Clara, CA, USA). A calibration curve (0.10, 0.50, 1.0, 5.0, 10, 25, 50, 100, 250, 500 and 1,000 ppb) was prepared from a standard Cu solution of 1,000 mg/L in HNO 3 at 1 % (Merck; Darmstadt, Germany). For the Cu measurement, the Internal Standard Kit (Ge, Ir, Rh Sc; ISC Science, batch 20210712) was used. The measurements were carried out using the linear calibration method and in triplicate for each of the samples analyzed. Analytical parameters of the procedure for the Cu analysis in the food samples considered were carried out prior to the determination of element concentrations by ICP-MS. The limit of detection (LOD) was 0.175 µg/L. For reference standards certified in Cu such as “Bovine muscle powder number 8414 ″ , and ”Citrus leaves powder number 1515 ″ , both certified by the National Institute for Standards and Technology (NIST; Gaithersburg, MD, USA), we determined concentrations of 2.82 ±0.07 and 16.55 ±1.33 for certified levels of 2.84 ±0.45 and 16.50 ± 1.02 µg/g, respectively. Specifically to evaluate the accuracy of the method to which the above values correspond, 10 fractions of 40 and 8 g of each of the reference standards, respectively, were used and subjected to the in vitro process of oral-gastro-intestinal digestion, followed by fermentation with the faecal inoculum from healthy children. Additionally, recovery experiments were performed, obtaining calculated recoveries for Cu that ranged between 97.1 and 101.8 %. 2.5. Statistical analysis In the statistical analysis of the data obtained, the SPSS statistical program (SPSS 28.0, Chicago, IL) was used. Data were expressed as mean Cu values ±standard deviation (SD). The existence of statistically significant differences was set to a p value lower than 0.05 (p <0.05). For the ANOVA analysis, the Student’s t-test or the Mann-Whitney test were used. 3. Results and discussion 3.1. Comparison of the partial (small and large intestine) and total Cu-BA Fig. 1A shows that the Cu-BA in plant foods is significantly higher in the small intestine than in the large intestine (p <0.01). Furthermore, the Cu-BASI is significantly higher than that found in foods of animal origin (p <0.001; Fig. 1A). These results are related to the different composition of the 2 food groups considered, where those of vegetable origin have lower protein content and a higher content of non-digestible carbohydrates and fiber. Therefore, the total bioaccessibility is higher. In this sense, it has been suggested that some undigestible polysaccharides (inulin and oligofructose) could act as potent enhancers of mineral bioavailability in foods of plant origin (Vitali et al., 2008). Others reported that the most important factors inhibiting Cu absorption in rats are animal proteins, S-amino acids (Brown et al., 1990) and Ú. García-Conde et al. Food Research International 197 (2024) 115238 3 histidine (Harvey et al., 1981). Additionally, microbial fermentation is known to have a superior impact on plant foods (Navajas-Porras et al., 2020), given their higher content of undigested food matrix in the small intestine (Bellail et al., 2012; García-Conde et al., 2023;). That higher solid residue in plant foods (fiber, resistant carbohydrates, undigested plant proteins, lipids and phenolic compounds; Power et al., 2014), which upon reaching the large intestine are modified with greater intensity, by the colonic microbiota present in the fresh faecal inoculum, with respect to foods of animal origin. In spite of this, no differences in the Cu-BALI are observed in foods of vegetable origin with respect to those of animal origin. However, the higher protein content and being of animal origin in animal foods (Brown et al., 1990), and probably of sulphur amino acids and histidine (Harvey et al., 1981), acting as inhibitors of Cu dissolution (Vitali et al., 2008), and forming insoluble chelates with Cu could negatively influence the Cu-BASI and therefore the total bioaccessibility of this element, which, as we have found, is lower in foods of animal origin. In foods of vegetable origin, the total Cu-BA in nuts is significantly higher, with the exception of other plant foods (p <0.01; Table 1). It should also be noted that the total Cu-BA in nuts is significantly higher than in the other food groups of plant origin, with the exception of other plant foods, which in turn have significantly higher total Cu-BA values than the other groups (p <0.01; Table 1). For cereals, fruits, oils, beverages and infusions and other foods of plant origin, Cu-BASI is significantly higher (p <0.001; Fig. 1B). Additionally, there are statistically significant differences between Cu- BA values in both the small and large intestine between the different plant food groups (p <0.01) (Table 1). Grouping plant foods into those with high (nuts, legumes and cereals) and low protein content (vegetables and fruits) shows that both groups have statistically higher Cu-BA values in the small intestine (p < 0.05), and that Cu-BA is significantly higher in those with high protein content (p <0.001 Fig. 1C). However, protein content was not observed to influence the total bioaccessibility of Cu (Fig. 1C). Wapnir (1998) indicated that dietary proteins facilitate Cu absorption and bioavailability. Therefore, plant proteins could be related to the formation of soluble chelates with Cu in the small intestine and with the higher Cu- BASI found in plant foods with higher protein content. However, Vitali et al. (2008) did not appreciate influence of protein content on Cu- BA in whole grain tea-biscuits and rice varieties (Ortiz & C´ amara-Mar- tos, 2018). New standardized studies using the simulator of the human intestinal microbial ecosystem (SHIME) in vitro method are therefore required in order to test and overcome the existing controversy about the influence of protein content in specific foods or selected menus on Cu-BA. In animal foods, in egg and cod the total Cu-BA is significantly higher and lower than in most other animal food groups, respectively (p <0.01; Table 1). Likewise, in foods of animal origin, no significant differences in Cu-BA values between the small and large intestine were observed (p > 0.05; Fig. 1D). Iaquinta et al. (2022) in beef and chicken determined values of 43 and 59 % for Cu-BA, which differ from ours (49.5 and 37.2 %, respectively). In specific cereals and by-products (Fig. 2A) the Cu-BASI was significantly higher in whole wheat bread, penne, whole grain penne, whole grain rice, cookies, whole grain cookies and breakfast cereals (p <0.05). Similarly, higher Cu-BASI percentages were found for grouped whole-cereals and non-whole cereals (p <0.05; Fig. 2B). The values of total Cu-BA were higher in the different groups of whole grains considered individually (Table S2) or grouped (Fig. 2B), which is related to that fiber, phytic acid and polyphenols, contrary to what is established for other minerals, do not limit the Cu-BA (Vitali et al., 2008), but rather, as found in our study, increase it to a greater extent in the small intestine. Others (Li et al., 2018) determined in rice grains harvested in regions Fig. 1. Mean Cu bioaccessibility values (%) in the small and large intestine of: (1A) all plant and animal foods; (1B) plant foods; (1C) plant foods depending on their protein content (high protein content: nuts, legumes and cereals; low protein content: vegetables and fruits); (1D) animal foods. The presence of equal capital or lowercase letters on the bars shows the existence of statistical significant differences (p <0.05). Ú. García-Conde et al. Food Research International 197 (2024) 115238 4 highly contaminated by heavy metals Cu-BA values by using the in vitro system of evaluation with oral, gastric and small intestine phases (21.2 ±11.2 %), lower than those measured in the present study for Cu-BASI (Table S2) in whole-grain rice and rice (39.6 ±26.5 and 25.3 ±5.5 %, respectively). Also, gluten cereals have significantly higher mean values of total and partial Cu-BALI compared to gluten-free cereals (Table 2; p < 0.001). Besides, noteworthy is the lower Cu content in gluten-free cereals (Table 2), as indicated Rogaska et al. (2020). These results express the importance of recommending non-adherence to gluten-free diets, unless the subjects have celiac disease (García-Conde et al., 2024). Despite this, a study in children with celiac disease found that at diagnosis ≅10 % had blood Cu levels below the reference ranges, and that these did not change after follow-up <6, 6–12 and >12 months (McGrogan et al., 2021). For this reason these investigators recommended intervention by dietary advice and supplementation (McGrogan et al., 2021). In gluten-free rice and buckwheat breads it was observed that after supplementation with milk and milk with seeds, the Cu concentration present doubled and tripled, respectively, but conversely there was a decrease in small intestinal bioaccessibility by ≅30 % (Regula et al., 2018). This finding would be related to the low Cu-BASI found in this study for milk 17.8 ±18.1 %. In nuts we have determined total Cu-BA values of 73.8 ±19.2 %, considerably higher than those determined in Turkish hazelnuts of 23.5 ±4.1 % by others (Erdemir, & Gucer, 2015). On the other hand, Arpadjan et al. (2013) found Cu bioaccessibilities of 62 and 14 % in hazelnuts and walnuts, respectively. In Fig. 2C, 2D and 2E, the Cu-BASI and Cu-BALI for vegetables, fruits and dairy products, respectively, are shown for the specific foods analyzed and those for which there are statistically significant differences (p <0.05; the corresponding Cu concentrations and BA values are shown in Tables S3, S4 and S5, respectively). As seen in most cereals (Fig. 2A and 2B) and in specific vegetables (potatoe, sweet potatoe and garlic; Fig. 2C), fruits (plum and peach; Fig. 2D) and dairy products (milk and fermented milk; Fig. 2E) the Cu- BASI is higher. This finding is in agreement with that reflected by Cai et al. (2017) in vegetables, who report a decrease of Cu-BA in the colon (between 24.8–100 %) with respect to that established in the small intestine. These researcher report Cu-BA values between 11.4–77.6 and “not detected”-53.0 % in the small and large intestine phases, compared to values of 16.5–43.4 and 11.5–34.1 % determined in the vegetables analyzed in our study, respectively. Therefore, we have observed a great variability in the total Cu-BA, in the small and large intestine, among the different food groups of plant origin. Such variability was also found in the foods included in each subgroup as found in cereals and vegetables, as indicated by other researchers also in vegetables (Cai et al., 2017) or in different foods (Zhang et al., 2023). The great disparity in the bioaccessibility of Cu would be associated with the different composition and different species of this element present in foods, as indicated by other researchers (Cai et al., 2017). These can be, on one hand, inorganic forms by combination with different anions and in different oxidation states of the mineral (in the case of Cu, as Cu o , Cu 1+ , or Cu 2+ ); on the other hand, they can also be different organic forms after formation of complexes (soluble/insoluble chelates) with different natural components of the food (i.e. such as amino acids, short chain fatty acids, phytates, oxalates, tannins), or those generated-released during culinary treatment (Maillard reaction compounds). 3.2. Effect of home culinary technologies Cu-BA in foods For practically all the home cooking techniques applied in foods of plant origin, the Cu-BASI is higher (Fig. 3A), with the exception of grilling, which is the only one in which this result is appreciated in foods of animal origin (Fig. 3B). Furthermore in this work we have found that the total Cu-BA, in all cooked foods and all grouped cooking techniques studied (51.0 ±24.4 %), is higher than that found in raw foods (49.0 ± 25.1 %; Fig. 3C). Contrarily, Zhang et al. (2023) found a higher bioaccessible fraction for Cu (for small intestinal phase) for different raw foods (76–80 %) than when previously cooked by boiling and frying (41–50 %). In our study, when considering the total Cu-BA only in foods of vegetable origin, the bioaccessibility of raw foods is lower than that of frying, roasting and grilling (Table S6). When culinary techniques are grouped according to whether the heating was carried out in hot liquids (frying in oil and boiling in water) or hot air (roasting and grilling), compared to the raw form of the food, it is observed that Cu-BALI in foods of plant origin (Fig. 3D) is significantly lower in the latter (p <0.05). In foods of animal origin (Fig. 3E) the Table 1 Mean total Cu concentrations (±standard deviation, SD; mg/kg) and bioaccessibility values (±SD, %) in the small and large intestine in (digestionfermentation method was performed in triplicate for any food sample analyzed). Food group Total Cu (ppm) Cu bioaccesibility in the small intestine † Cu bioaccesibility in the large intestine † Total Cu bioaccesibility Vegetalbased foods     Nuts*6.86 ±2.56 42.4 ± 9.27 ABCDEF 31.4 ± 19.1 ABCDEFG 73.8 ±19.2 ABCDEF Cereals*4.56 ±1.81 34.7 ±19.5 aAGHI 12.9 ± 16.4 aAHIJKLM 47.2 ± 24.9 AGHIJ Fruits*3.60 ±0.68 30.6 ±12.6 bBJKL 22.2 ± 21.1 bBHNO 52.6 ±22.4 BGKL Vegetables * 4.15 ±1.38 29.0 ±12.4 CMNO 25.2 ±22.2 CIPQ 54.1 ± 23.5 CHMN Legumes*3.39 ±0.71 26.6 ±5.05 DPQR 25.5 ±22.8 DJRS 52.1 ± 22.7 DOPQ Oils*1.03 ±0.37 47.4 ± 14.6 cGJMPS 1.80 ± 4.85 cEKNPRT 49.1 ± 15.3 EORS Beverages and infusions * 2.80 ±0.96 20.4 ± 6.96 dEHKNQST 0.46 ± 0.64 dFLOQSU 20.9 ± 7.14 FIKMPRT Others*6.35 ±0.17 54.0 ± 8.10 eFILORT 16.1 ± 12.3 eGMTU 70.2 ± 17.1 JLNQST All vegetal foods* 4.05 ±1.65 31.1 ±14.1 f 21.4 ±21.5 f 52.2 ±24.1 Animal foods     Dairy products * 0.27 ±0.06 18.2 ±13.9 26.8 ±28.7 45.0 ±30.0 AB Chicken*1.87 ±0.53 15.8 ±15.1 21.4 ±24.0 37.2 ±24.2 CDEF Beef*3.31 ±1.38 29.2 ±28.1 20.3 ±21.8 49.5 ±25.2 CGH Salmon*2.51 ±0.25 23.8 ±5.45 A 21.9 ±21.8 45.7 ±20.4 DIJK Cod*2.15 ±0.32 13.0 ±1.19 AB 19.9 ±18.5 32.9 ± 18.7 AGILM Egg*3.74 ±0.75 38.4 ±16.9 B 18.2 ±14.7 56.7 ± 17.7 BEJLN Lamb*2.24 ±0.18 23.1 ±13.5 27.1 ±24.7 50.1 ±27.4 FMO Pork*2.44 ±0.62 14.1 ±8.91 23.9 ±23.4 38.0 ± 24.6 HKNO All animal foods* 2.63 ±0.86 22.8 ±22.9 22.9 ±23.3 44.5 ±25.3 * Rows labelled with the same superscript lowercase letters for Cu-BA values in the small and large intestine in every food category/type denotes the existence of statistically significant differences (p <0.05). † Columns labelled with the same superscript capital letters for Cu-BA values for different food groups in plant- and animal foods denote the existence of statistically significant differences (p <0.05). Ú. García-Conde et al. Food Research International 197 (2024) 115238 5 behaviour is different, since the Cu-BASI in foods cooked with in hot liquids (frying and boiling) present significantly lower values with respect to the other 2 groups (p <0.005). However, the Cu-BALI in foods cooked in hot liquids is statistically higher only with respect to foods subjected to hot air heating (p <0.005). Therefore, in this study it is remarkable the different influence on Cu- Fig. 2. Mean Cu bioaccessibility values (%) in the small and large intestine of different samples of: (2A) cereals and by-products; (2B) whole vs. non-whole cereals; (2C) vegetables; (2D) fruits; (2E) dairy products. The presence of equal lowercase letters on the bars shows the existence of statistical significant differences (p <0.05). Table 2 Mean Cu contents (±standard deviation, SD; mg/kg) and Cu-BA values (±SD, %) in the large and small intestine in cereals with gluten vs. free-gluten cereals (digestion-fermentation method was performed in triplicate for any food sample analyzed). Cereal group Total Cu † Bioaccesibility in the small intestine † Bioaccesibility in the large intestine † Total bioaccessibility † Gluten cereals*4.89 ±1.75 A 35.1 ±21.1 a 14.0 ±16.5 Ba 49.1 ±25.9 C Gluten-free cereals*2.61 ±0.12 A 32.5 ±7.83 b 6.75 ±14.6 Bb 39.2 ±15.4 C * Rows labelled with the same superscript lowercase letters for Cu-BA values in the small and the large intestine for every cereal group denotes the existence of statistically significant differences (p <0.05). † Columns labelled with the same superscript capital letters for total Cu and Cu-BA values in different cereal groups denote the existence of statistically significant differences (p <0.001). Ú. García-Conde et al. Food Research International 197 (2024) 115238 6 BA in the small and large intestine exerted by the distinct home cooking techniques applied to foods, when comparing their raw state vs. frying and boiling or roasting and grilling (Fig. 3d and 3e, respectively). Consequently, in plant foods the Cu-BASI is not affected by culinary treatments. However, Cu-BALI is lower in raw foods, so their technological processing (frying and boiling or roasting and grilling) increases Cu-BALI (Fig. 3D). In this sense, prior culinary treatment could facilitate the release from the food matrix of components more easily transformed by the intestinal microbiota that would ultimately increase Cu-BALI, a fact that should be addressed in future research. On the other hand, it has been reported that more intense cooking methods such as roasting and grilling increase the abundance of beneficial bacterial genera in the microbiota (P´ erez-Burillo et al., 2018b), which could be related to the greater Cu-BALI found in foods of plant origin when cooked by frying and boiling or roasting and grilling vs. their raw form. However, in animal foods the effect is more notable, as it influences both Cu-BASI and Cu-BALI. Cu-BASI is lower in foods cooked by frying and boiling than in foods cooked by roasting and grilling or in the raw state (Fig. 3E). It could be that cooking in water or oil could lead to the loss or alteration of the original structure of certain components related to the formation of soluble chelates with Cu in the small intestine, a fact to be verified in future and more specific investigations. However, Cu- BALI is lower in foods of animal origin cooked by the more drastic techniques of roasting and grilling compared to frying and boiling. Therefore, in foods of animal origin in reference to Cu-BALI, the opposite happens to what has been established in reference to the influence of the Fig. 3. Mean Cu bioaccessibility values (%) in the small and large intestine of all foods depending on home culinary technique: (3A) in plant foods; (3B) in animal foods; (3C) in all raw and cooked foods; (3D) in raw plant foods compared to those heated by hot liquids or hot air; (3E) in raw animal foods compared to those heated by hot liquids or hot air; (3F) in vegetables. The presence of equal capital or lowercase letters on the bars shows the existence of statistical significant differences (p <0.05). Ú. García-Conde et al. Food Research International 197 (2024) 115238 7 culinary techniques considered in Cu-BASI. Vegetables constitute the largest food group studied in this research (n =60). Given their content of soluble fiber susceptible to fermentation by the microorganisms present in the feces used in this study, the influence of the culinary technologies used on Cu-BA was specifically evaluated (Fig. 3F). It is noteworthy that roasting and frying are the culinary techniques for which the total Cu-BA is significantly higher than that of raw form and boiling techniques (p <0.001). In this sense, it has been indicated that roasting and frying are the culinary techniques that involve the use of more drastic conditions in the cooking of vegetables, and that they are related to higher values of Maillard reaction indicators such as higher furosine and 5-hydroxymethylfurfural (P´ erez- Burillo et al., 2019). On the other hand, Nissen et al. (2022) also find that the intense heat treatments of frying (to a greater extent) and grilling of dark purple eggplant increase the levels of extractable phenolic compounds, relative to the raw state. This increase facilitated further metabolisation of phenolic compounds to 3-(3 ′ -hydroxyphenyl) propanoic acid, which was ultimately associated with increased growth of beneficial gut microbiota bacteria. Additionally, Cattivelli et al. (2023) using the INFOGEST protocol, and for lower gut fermentation, the short-term batch model, namely MICODE (multi-unit in vitro colon gut model), found that the degradation of red-skin onion flavonols by colonic microbiota facilitated the accumulation of 3-(3‘-hydroxyphenyl) propanoic acid, 3-(3 ′ -hydroxyphenyl)acetic acid and 3-(3‘,4 ′ -dihydrox- yphenyl)acetic acid. The study compared three cooking methods (frying, grilling and raw state), showing that especially in the raw state the flavonols of red-skin onion after colonic fermentation increased the proportion of beneficial bacteria in the gut microbiota to a greater extent than the heat-treated ones, while inhibiting the development of harmful opportunistic bacteria (Clostridium perfringens and Escherichia coli). Besides regarding Cu-BALI that corresponding to frying only was significantly higher than that of raw form and boiling (p <0.05). Therefore, drastic techniques such as frying could cause a greater breakage of the cell wall of vegetables and consequently facilitate greater access of fermenting microorganisms (P´ erez-Burillo et al., 2018b) that would enable the appearance of compounds capable of forming soluble chelates with Cu that would finally increase its bioaccessibility, like the short chain fatty acids reported by others (Nissen et al., 2022; Cattivelli et al. (2023). In view of the results found in vegetables and cereals referred to above, it is contradictory the different effect that is established in the total Cu-BA, so that in addition to the influence exerted on it by the products of the Maillard reaction, other influential factors must also be involved, possibly related to the different composition of vegetables and cereals and by-products (12 and 10 different products, respectively) and the different effect that culinary treatments exert on the differential components. Therefore, future studies concerning specific vegetables and cereals with distinct culinary techniques, which also take into account the differential effect on the food matrix, should be planned in order to better understand the intimate mechanism driving Cu-BA. Kumari & Platel (2022) indicated that thermal processing of most cereals, and specifically for wheat and rice, produced a reduction in Cu- BA during microwave cooking, pressure cooking and cooking in an open pot. This finding is similar to our due to that the total Cu-BA values found specifically in the analyzed raw cereals (wheat and by-products and rice: 61.2 ±25.2 % were higher than those found when they were previously cooked (frying, toasting and boiling: 36.6 ±18.7 %). Mesías et al. (2012) indicated that the consumption of highly processed foods in adolescents decreases Cu bioavailability by decreasing serum and erythrocyte Cu levels. These researchers related this finding to the negative effect on Cu bioavailability exerted by the Maillard reaction products generated when comparing the “Brown diet” with the “White diet”. This result is in line with what was previously found and referred to in cereals where the Cu-BA in its raw state is higher. In fact, cereals are a group of foods in which especially the formation of Maillard reaction products during thermal processing is facilitated by the joint presence of the 2 substrates necessary for their development, such as reducing sugars and free amino groups from amino acids and dietary proteins. In black beans, it has been found that Cu-BA is high (≈70 %) after traditional cooking with a pressure cooker or regular pan, and that it decreased if the soaking water was discarded (Feitosa et al., 2018). However in the present study it was not observed in legumes that total Cu-BA is affected by cooking when comparing boiling (54.1 ±21.5 %) with roasting (52.1 ±24.6 %) and grilling (50.2 ±22.7 %). 3.3. Effect of life stage on the Cu-BALI Fig. 4A shows the distribution of Cu in bioavailable and nonbioavailable fractions for all foods, in studied individuals. Fig. 4B shows that Cu-BALI in HE-AD is significantly lower than that determined for HE-CH (p <0.001) as it can also be seen in practically all the subgroups included in plant-based and animal foods (Table 4; p <0.05). Similarly, in plant foods, regardless of their higher or lower protein content, Cu-BALI values are significantly higher in HE-CH than in HE-AD (Table S8; p <0.05). Additionally, for most of the foods considered individually (Tables S9, S10 and S11), especially in vegetables and fruits, the Cu-BALI values found are also significantly higher in healthy children than in healthy adults (p <0.05). In cereals grouped into whole-grain and non-whole grain, and in cereals with gluten (Tables S13 and S14, respectively) and free-gluten cereals also the Cu-BALI is significantly higher in children than in HE-AD (p <0.05). Consequently, the stage of life influences the total Cu-BA, being lower in adults (43.0 ±22.7) than in children (66.3 ±24.9 %) (Fig. 4A). This result could be related to the growth stage inherent to childhood aimed at satisfying the needs associated with it, despite a lower food and energy intake than that of healthy adults. An adequate intake of this element is therefore transcendental to guarantee an optimal growth as was reported for animals (Carpenter et al., 2019). Another aspect associated with the increased Cu-BALI in children (37.7 ±23.7) with respect to HE-AD (14.1 ±18.5 %; Fig. 4B), globally and for most of the food groups of plant and animal origin analyzed (Tables 4, S8, S9, S10, S11, S12, S13 and S14), would be the differential composition in their colonic microbiota. The microbiota can control the metabolism and transport of trace metals such as Cu, regulating their assimilation from food or by direct competition with the host (Pajarillo et al., 2021). These researchers have also reported that Cu can modulate the intestinal microbiota (Pajarillo et al., 2021). Therefore, future studies are missing to contemplate this bidirectional action regarding, on the one hand, how the composition of the specific microbiota is related to the Cu-BA in a different way according to the stage of life in specific foods, and on the other hand, the modulating capacity of dietary Cu, of the microbiota composition. Finally, the Cu-BALI is statistically higher for the different culinary techniques used, considered individually (Table 3) or grouped (Table S7) when the fermentation was carried out with inoculums of feces from HE-CH compared to HE-AD. 3.4. Effect of childrens pathology on the Cu-BALI Fig. 4B includes the mean values of Cu-BALI for all foods in healthy and diseased children. It is seen that Cu-BALI in HE-CH is significantly higher than that determined for celiac, obese and allergic children (p < 0.05). Likewise, the Cu-BA in AICM-CH is significantly higher than in the other 2 groups of sick children. Again, in most plant and animal foods statistically higher Cu-BALI for HE-CH than for different groups of sick children was found (Table 4). Similarly, in plant foods, regardless their protein content, Cu-BALI values are significantly higher in healthy children (Table S8; p <0.05). From the data included in Table S9 (for cereals and by-products), S10 (for vegetables), S11 (for fruits) and S12 (for dairy products), it can be seen that only eggplant, onion, cauliflower, spinach, garlic, tomato and Ú. García-Conde et al. Food Research International 197 (2024) 115238 8 cabbage Cu-BALI in healthy children was significantly higher than that corresponding to the 3 groups of sick children studied. Likewise, and as can be seen in Tables S13 and S14, Cu-BALI in HECH is also significantly higher than that of celiac, obese and allergic children, for cereals grouped according to fiber content (whole and nonwhole cereals) or gluten (gluten and gluten-free cereals), respectively (p <0.05). In GRD-CH the Cu-BALI (14.6 ±19.8) is lower than that determined in HE-CH (Fig. 4B; 37.7 ±23.7 %), a result that together with the lower Cu content of gluten-free cereals also referred by others (Pedron et al., 2016; Punshon and Jackson, 2018) and its lower bioaccessibility (Table 2), may notably compromise the nutritional status of this essential element in celiac children, and thus their growth and health status. This result is reinforced by the finding that Cu-BALI in celiac children is higher in gluten-containing cereals (Table S13). Possibly, the lower Cu-BALI in celiac children is directly related to a specific composition of the intestinal microbiota associated with the pathology itself (Zafeiropoulou et al., 2020; El Mouzan et al., 2022), as well as to the composition of the diet ingested, as it has been indicated that the higher consumption of meat and beans may be one of the causes of the decrease in beneficial bacteria in the microbiota in celiac patients (Soheilian-Khorzoghi et al., 2022). The differential composition of the colonic microbiota in celiac patients has been related to modified levels and proportions of SCFA as well as valeric, 2-ethylhexanoic, tetradecanoic acids, etc. (Federica et al., 2022) produced in the fermentative processes induced by the microbiota. These fermentative products could be related to a lower formation of soluble chelates in the celiac children studied and therefore to a lower Cu-BALI. Contrarily, others indicate that there is no scientific evidence of specific traits in the microbiota of celiac children (Abdukhakimova et al., 2021). In contrast, other studies did not associate celiac seropositivity in celiac children aged 6–9 years with serum Cu levels (Li et al., 2023). Nevertheless, Halfdanarson et al., (2009) reported that Cu deficiency is a complication of celiac disease, which is related to the results of our study. Along the same lines, (Ince et al., 2008) suggested that increased urinary losses of Cu together with malabsorption problems may be causes of Cu deficiency in celiac women. In addition, especially in several groups of foods of vegetable origin (Table 4) and in several vegetables (Table S10) and fruits (Table S11) the Cu-BALI is lower in OB-CH than in allergic children. Also, in OB-CH when considering all foods analyzed, lower Cu-BALI values were found than those determined in HE-CH (15.5 ±17.8 vs. 37.7 ±23.7 %), respectively; Fig. 4B) as was observed for most of the foods included as subgroups in the plant foods (Tables 4, S9, S10, S11, S13 and S14). This result, as previously mentioned, could be related to the specific composition of the intestinal microbiota associated with the pathology itself, and with the formation of compounds after the fermentative process that hinder the formation of soluble chelates with Cu, which warrants further research in this area. Similarly to our findings, Boullata et al. (2017) in morbidly obese people concluded the existence of Cu deficiency. Likewise, in obese patients with hepatic steatosis it has been reported that a decrease in the bioavailability of Cu increases the risk of vascular pathologies, when it was found that serum Fig. 4. (4A) Non-bioaccessible and bioaccessible Cu fractions in foods in the small and large intestine of healthy adults, and healthy and unhealthy children. (4B) Mean Cu bioaccessibility values (%) of all food in the large intestine of different groups of subjects. Table 3 Mean Cu bioaccessibility values (Cu-BA ±standard deviation; %) in the large intestine of healthy adults, and healthy and unhealthy children*of all plant and animal foods depending on the home culinary techniques used (digestionfermentation method was performed in triplicate for any food sample analyzed). Home culinary technique Cu-BA in HEAD Cu-BA in HE-CH Cu-BA in GRD-CH Cu-BA in OB-CH Cu-BA in AICM-CH Plant foods Raw form #† 13.2 ± 18.6 A 29.0 ± 24.0 Aabc 11.7 ± 16.8 ad 11.3 ± 14.3 be 21.9 ± 19.2 cde Frying #† 16.3 ± 20.6 B 37.9 ± 21.9 Babc 14.1 ± 20.3 ad 14.5 ± 16.3 be 29.9 ± 20.4 cde Roasting #† 15.9 ± 19.5 C 43.4 ± 20.8 Cabc 15.7 ± 20.6 ad 14.8 ± 18.5 be 31.3 ± 16.5 cde Toasting #† 5.97 ± 10.9 D 30.7 ± 13.9 Da 15.2 ± 20.0 5.95 ± 5.28 a 14.2 ± 17.1 Boiling #† 12.3 ± 19.4 E 39.3 ± 21.1 Eabc 15.6 ± 21.2 ad 10.8 ± 15.1 be 24.6 ± 16.9 cde Grilling #† 19.1 ± 20.7 F 42.2 ± 22.3 Fabc 14.5 ± 19.0 ad 15.9 ± 14.6 be 32.4 ± 16.7 cde Animalfoods Raw form #† 11.9 ± 15.4 37.1 ± 33.0 16.4 ± 23.5 30.6 ± 30.8 27.8 ± 27.1 Frying †# 14.5 ± 18.4 B 48.8 ± 27.0 Babc 24.9 ± 27.7 a 30.7 ± 23.8 b 36.0 ± 22.2 Roasting #† 9.32 ± 10.0 C 28.8 ± 22.0 Ca 12.8 ± 17.1 a 20.2 ± 16.7 18.8 ± 14.3 Boiling #† 11.5 ± 14.9 D 42.8 ± 25.5 Dab 19.1 ± 21.5 a 25.8 ± 20.1 26.4 ± 18.4 b Grilling # 6.90 ± 9.39 E 32.1 ± 24.0 Eabc 7.85 ± 11.4 a 11.1 ± 13.4 b 20.0 ± 22.4 * Healthy adults (HE-AD); healthy children (HE-CH); children with gluten related disorders (GRD-CH); children with obesity (OB-CH); children with allergy/intolerance to cow’s milk proteins (AICM-CH). # Rows labelled with the same superscript capital letters for Cu-BA values for the home culinary techniques used in every food category for healthy adults vs. healthy children denotes the existence of statistically significant differences (p < 0.05). † Rows labelled with the same superscript lowercase letters for Cu-BA values for the home culinary techniques used in every food category for different groups of children (healthy and unhealthy children) denotes the existence of statistically significant differences (p <0.05). Ú. García-Conde et al. Food Research International 197 (2024) 115238 9