Food and Chemical Toxicology 71 (2014) 254–263 Contents lists available at ScienceDirect Food and Chemical Toxicology journal homepage: www.elsevier.com/locate/foodchemtox Heavy metal and metalloids intake risk assessment in the diet of a rural population living near a gold mine in the Peruvian Andes (Cajamarca) Marta Barenys a,b, Nuria Boix a,f, Andreu Farran-Codina c, Imma Palma-Linares c, Roser Montserrat c, Ariadna Curto a, Jesus Gomez-Catalan a,f, Pedro Ortiz d, Nilton Deza e, Juan M. Llobet a,f,⇑ a GRET-CERETOX and Toxicology Unit, Public Health Department, School of Pharmacy, University of Barcelona, Av. Joan XXIII s/n, 08028 Barcelona, Spain IUF – Leibniz Research Institute for Environmental Medicine, Auf’m Hennekamp 50, 40225 Düsseldorf, Germany Nutrition and Food Science Department, School of Pharmacy, Food and Nutrition Torribera Campus, University of Barcelona, Av. Prat de la Riba 171, 08921 Santa Coloma de Gramenet, Spain d Facultad de Ciencias Veterinarias, Universidad Nacional de Cajamarca, Carretera Baños del Inca km 3.5, Cajamarca, Peru e Escuela de Postgrado, Universidad Nacional de Cajamarca, Carretera Baños del Inca km 3.5, Cajamarca, Peru f INSA-UB, Public Health Department, School of Pharmacy and Food and Nutrition Torribera Campus, University of Barcelona, Prat de la Riba, 171, 08921 Santa Coloma de Gramenet, Barcelona, Spain b c a r t i c l e i n f o Article history: Received 14 April 2014 Accepted 21 June 2014 Available online 30 June 2014 Keywords: Heavy metal Cajamarca Gold mine Rural diet Nutrient intake Risk assessment a b s t r a c t This study evaluates the diet composition of a rural population near a gold mine in the Cajamarca district of Peru. The main consumed items by this population were tubers and cereals, and the mean energy intake (1990 kcal) was shown not to cover the recommended intake values for the male population. The concentrations of As, Cd, Hg, Pb, Zn, Al, Cr and, Cu in drinking water and food samples of items contributing to 91% of this diet (145 samples, 24 different items) were determined and used to calculate their daily intakes for risk assessment. The As, Cd and Pb daily intakes exceeded the limit values established by the European Food Safety Authority (EFSA), entailing serious concerns for the population’s health. Moreover, the intake values of As and Pb were shown to be higher, the closer to the gold mine the studied population was. Ó 2014 Elsevier Ltd. All rights reserved. 1. Introduction From a public health point of view, it is always important to know whether a diet covers the recommended amount of nutrients, as well as to ensure that a population does not ingest a quantity of toxic compounds above the maximum values established by official health organizations. Mining activities causing water pollution are a widespread international problem (Jarvis and Younger, 2000; Donato et al., 2007; Blacksmith Institute, 2012) which can produce health impacts in miners and communities living in the surrounding areas (Coelho et al., 2011). In Cajamarca, a department situated in the North of Peru, the biggest gold mine in South America is operating in an area close to where rural population is still living. The impact of the gold mine operations on the environment, health and soil and water rights has become a concern for the surrounding rural ⇑ Corresponding author. Address: Toxicology Unit, School of Pharmacy, University of Barcelona, Av. Joan XXIII s/n, E-08028 Barcelona, Spain. Tel.: +34 93 403 58 17; fax: +34 93 403 58 16. E-mail address: jmllobet@ub.edu (J.M. Llobet). http://dx.doi.org/10.1016/j.fct.2014.06.018 0278-6915/Ó 2014 Elsevier Ltd. All rights reserved. populations (Bury, 2005). This is reflected in the numerous social conflicts related to the gold mine during the last years, the appearance of social tensions between communities, and the confrontations and social mobilizations led by peasants against the mining exploitation (Bebbington et al., 2008; Bury, 2002). In this sense, previous studies have reported heavy metal pollution in sediments of the Rejo River at the closest points to this gold mine site entailing a risk for the aquatic life living in the headwaters (Yacoub et al., 2012). However, the risk of these water sources for the peasant population living in this area through their diet has never been studied. The diet of this population is partly covered by local food production in their own crops. Additionally, other non-local items are obtained from the city market. In this population there is a social concern about the food and water safety and the possibility of health related problems. Previous case-study interviews reflected serious concerns about the quality of the water, a decline in animal health and productivity, a reduction in fish catches and a decay in agricultural productivity (Bury, 2002). At the beginning of this study, there were no official documents neither published articles with data about the composition of the M. Barenys et al. / Food and Chemical Toxicology 71 (2014) 254–263 diet of this rural area. A dietary survey among this population was interesting by itself because it contributed to bring information on the composition of an unidentified diet. Moreover, the dietary survey was also necessary as a working tool for the heavy metals risk assessment, to know which foods should be sampled, the contribution of each food to the diet and later be used to calculate the total amount of metal intakes. During the development of this study, an official dietary survey was published by the Peruvian National Institute of Statistics and Informatics (INEI, 2012), whose results were used to compare and validate our own survey. The aims of this research work were (1) to study the diet composition of the rural population surrounding the gold mining area, (2) to calculate the energy and nutrients intake and compare it to the recommended values, (3) to determine some selected heavy metals and metalloids in drinking water and in the main foods contributing to the diet (4) to calculate the daily intake of heavy metals and metalloids and to compare it with the officially established limits. 255 duced to overcome the fact that same food could have been named different), the recipe name, the intake time during the day, the date, the week day, the food quantity in grams and the food quantity in domestic measure. A qualitative food frequency questionnaire including 50 single commonly consumed items was also performed to the same individuals by the same interviewer to contrast and compare the obtained information with the 24 h dietary recall. This would make possible to detect frequently consumed items which did not appear in the 24 h dietary recall due to, for example, seasonality. At the same time, the origin of the consumed items was recorded distinguishing between locally produced food and items obtained from the market. 2.3. Intake data processing 2. Materials and methods The 24 h dietary recall data was used to calculate the contribution of each raw food to the total diet. Each recipe was broken down into its ingredients, and ingredient raw weights estimated using food yields previously determined in the laboratory (see Section 2.2). A list of the single foods in decreasing consumption order was obtained. The intake of each food was calculated for each individual as well as the contribution of the food to the total intake expressed in percentage. The consumed foods were classified in 20 groups: water and broths, soft drinks, vegetables, meat and eggs, rice, fruits, potato, other tubers, milk, other dairy products, bread, other cereals and cereal products, sugars and sweets, oils and other fats, legumes, herbs and spices, fish, coffee and tea, alcoholic beverages, and nuts. 2.1. Study area 2.4. Energy and nutrients intake calculation The study area was defined as the rural area between Cajamarca city and the gold mine operating at the north of the Cajamarca district, covering the river basins of: Porcón River, Rejo River, and Azufre River (Fig. 1). The daily individual energy and nutrients intake was calculated using the table generated from the 24 h dietary recall. The food weights were converted into energy and nutrient intake values using the following sources (listed in preference order): Peruvian food composition tables (Reyes et al., 2009), Industrialized foods composition tables (Bejarano et al., 2002), Latinfoods (http://www.inta.cl/ latinfoods), and Food composition tables of CESNID-UB (Farran et al., 2004). Nutrients included in the intake calculation were selected from a Peruvian public health point of view according to the ‘‘Sala Situacional Alimentaria Nutricional 3’’ study from the Peruvian Health Ministry (Ministerio de Salud del Perú, 2012a), where Ca, Fe and vit. A intakes are reported to be under recommended intakes for more than half of the Peruvian population. Protein, lipid and carbohydrates were also included as they are needed to calculate energy provision, which is the minimum nutritional information required to evaluate intake adequacy. 2.2. Dietary survey Thirty-six people (twenty-eight males and eight females) between 13 and 55 years old, from 11 populations located in the study area were interviewed by the same interviewer for a 24 h dietary recall. All consumed foods and recipes were registered using household measures and food servings in order to obtain the information as precisely as possible. The household measures and food servings were converted to gram weights measuring volumes and weights of local utensils with a scale (Soehnle, Germany; 1 g precision) or following conversions detailed in Palma et al. (2008). Ingredients and quantities from several local recipes were prepared and weighted to determine weight changes due to cooking process. Products sold in not common bulk units, like peanut bags, sweets, and chocolate balls were locally bought to convert these units to gram weights. A table was created including for each interviewed person, the individual identification, registered name of the consumed food or ingredient, the unified name of the food or ingredient (intro- 2.5. Food sample collection Food samples were collected covering the 91% of the diet according to the results of the 24 h dietary recall. Four sampling areas were defined depending on the distance to the gold mine and the presence of gold mine influenced streams Fig. 1. Map of the study area showing the localization of Cajamarca city, part of the gold mine operations area, the populations included in the dietary survey (d) and the food sampling areas (s). Sources: modified from Google earth using information from the Education Ministry of Peru and the Peruvian National Institute of Statistics and Informatics. 256 M. Barenys et al. / Food and Chemical Toxicology 71 (2014) 254–263 (from closer to further): La Pajuela, Tual, La Ramada and Porcón Bajo (Fig. 1). A fifth area was established corresponding to the central market of Cajamarca city, where the rural population obtains the non-local food. These five areas, were considered zones of study and none of them was considered as a control area. Depending on the representativeness of the food on the total diet, different numbers of samples per each rural sampling area were collected. The more a food item was consumed, the more samples were collected. For local foods representing more than 5% of the diet, 10–15 samples were collected; for local items representing between 1% and 5% of the diet, 4–10 samples were collected and, for items representing less than 1% in the diet, at least 3 samples were collected. Each sample was composed of three different units of each food. Two or three samples of non-local foods were collected at the central Cajamarca market. Samples were transported in a 4 °C box to the laboratory and were cleaned using distilled water. Equivalent portions of edible parts of the foods composing one analytical sample were pooled and triturated in a domestic apparatus to homogeneity avoiding cross contamination. The samples were divided in triplicated aliquots, labelled and stored at À20 °C using plastic bags. Finally, one of the triplicates was sent in dry ice to the Toxicology Unit of the Pharmacy School at the University of Barcelona, and stored again at À20 °C until the analysis of metal content. 2.6. Sample preparation and analysis About 0.25 g of each sample were pre-digested with a solution of 4 mL of 65% Nitric acid (Suprapur, E. Merck. Darmstadt, Germany) and 4 mL of 30% Hydrogen peroxide (Suprapur, E. Merck) using TeflonÒ vessels in a Milestone Start D Microwave digestion system. The characteristics of the selected program consisted of ten intervals of 5–10 min, heating up to a maximum temperature of 210 °C. The digested samples were diluted to a final volume of 50 mL with ultrapure water (Milli-Q grade). The accuracy of the instrumental methods and analytical procedures was checked by duplication of the samples. Analytical grade reagents were used for blanks and calibration curves. Quality control of the methodology was assured by analysing a certified reference material (Lobster hepatopancreas, NRC Canada, TORT-2, Ottawa, ON, Canada). Every seven samples, a blank sample involving all reagents was run to check for interference and cross-contamination. The concentrations of As, Cd, Hg, Pb, Zn, Al, Cr, and Cu were determined with inductively coupled plasma-mass spectrometry (ICP-MS Perkin–Elmer Elan 6000, Woodbrige, ON, Canada). Rhodium was used as internal standard. Replicate measurements were performed. The limits of detection (LOD, fresh weight) were the following: As 0.0024 lg/g, Cd 0.0018 lg/g, Hg 0.0052 lg/g, Pb 0.01 lg/g, Zn 0.1 lg/g, Al 0.2 lg/g, Cr 0.072 lg/g, Cu 0.04 lg/g. Recovery rates for the elements analysed under the different experimental conditions ranged between 98.4% and 113.5%. 2.7. Heavy metal and metalloids exposure The results of the heavy metal and metalloids analysis in each food and water sample and the contribution of this food to the total diet, according to the 24 h dietary recall results, were used to calculate the daily intake of each metal per individual expressed in lg of metal/kg of body weight (b.w.) per day. For samples in which the metal was not detected, two estimations were calculated, the lowerbound estimation, where the values under the LOD were substituted by zero, and the upperbound estimation, where they were substituted by the LOD value. The obtained result was expressed in relation to the body weight of each of the individuals in the interviewed population. The mean, the minimum and the maximum metal intake were calculated for the total population of study and for the populations living in every sampling area. In this case, the exposure was calculated using the dietary data of the whole population but considering only the metal concentration of the specific products consumed in each sampling area. For those foods that were not of local origin, the mean value obtained from the market food was used. The obtained values were compared to the toxicological guidance values established by official health agencies, and in those cases where the mean intake exceeded official values, a statistical comparison between the results of each sampling area was performed. To support the results of the risk assessment performed using our own 24 h dietary recall data, a second evaluation was performed using the intake data obtained from the official Peruvian dietary survey (INEI, 2012). In this case heavy metals and metalloids intake in lg/kg body weight per day were represented as the mean value and the 80th percentile value for the lowerbound (LB) and the upperbound (UB) estimation for the total population and the estimations for the population in each sampling area. 2.8. Statistics The number of samples exceeding official recommended values for metals respect the total number of samples collected per area was compared and statistically analysed with Fisher’s exact test. Metal intake between the different sampling areas was statistically compared using the Student’s t-test. A probability of p < 0.05 was considered statistically significant. 3. Results 3.1. Characteristics of the population answering the dietary survey Thirty-six people were included in the dietary survey, 28 males (78%) and 8 females (22%). The age of the participants was comprised between 13 and 55 years old, being the 63% of them teenagers between 14 and 17 years old. The mean body weight of the participants was 50.9 kg ± 10.8 (n = 25) [mean males: 50.4 kg ± 10.9 (n = 21); mean females: 53.8 kg ± 11.5 (n = 4)]. The participants were living in 11 different populations comprised in the study area. The number of participants from each population were: Tual: 10, Porcón Bajo: 7, La Ramada: 4, Cince-Las Vizcachas: 3, Quilish 38: 3, La Apalina: 3, Quishuar Corral: 2, Combayo-La Florida: 1, Manzanas Alto: 1, Hualtipampa Baja: 1 and Huambo Cancha: 1 (see map in Fig. 1). 3.2. Dietary survey From the 24 h dietary recall, the total consumption of each food in grams per day was obtained. Water consumed in different ways was the most consumed item followed by potatoes, milk, soft drinks, rice and, bread. A list of the most consumed items per capita in g/day is presented in Table 1, for all items contributing to at least 0.8% of the total intake. The results of the qualitative food frequency questionnaire were very similar to the 24 h dietary recall and showed that the 4 most consumed items after water were the same in both surveys: potatoes, soft drinks, milk and rice, but in a slightly different position. When comparing the first 14 items of the qualitative food frequency questionnaire, 9 of them were also present in the 14 first positions after water of the 24 h dietary recall: potatoes, soft drinks, milk, rice, olluco (Ullucus tuberosus), onion, bread, carrot, and cabbage, while the following five items were different: oca (Oxalis tuberosa), chili, sweet potatoes (Ipomea batatas), tomato and, banana. The individual foods obtained from the 24 h dietary recall, were classified in 20 food groups described in Section 2.3. The result of this classification was compared to the result of an official Peruvian dietary survey reporting food consumption in kilograms per capita per year (INEI, 2012). This official dietary survey collected information from questionnaires of 1536 households at the rural area and 34,698 households at the urban area covering a period of seven Table 1 Most consumed items per capita according to the 24 h recall dietary survey, including items contributing to at least 0.8% of the total intake. Results are expressed in g/day and in% of total intake in decreasing consumption order. Intake results of the same items in the official Peruvian dietary survey (INEI, 2012) for population living in the mountains is detailed in the next column in g/day for comparison. 24 h dietary recall INEI 2012 Item % of total intake g/day g/day Water Potatoes Milk Soft drinks Rice Bread Chicken Onion Egg Carrot Orange Olluco Noodles Cabbage Apple 44 15.6 6.8 4.4 3.6 2.2 1.6 1.4 1.3 1.3 1.1 1.1 0.9 0.9 0.8 1153.7 408.9 179.6 130.7 94.2 58.2 42.0 37.3 33.0 32.8 28.1 27.9 23.5 23.2 21.9 – 319.2 25.2 99.2 66.8 70.1 23.3 29.6 13.2 27.1 19.5 15.9 35.3 7.9 19.7 M. Barenys et al. / Food and Chemical Toxicology 71 (2014) 254–263 days and three geographical areas: coast, mountains and rainforest. Our results showed similar intakes for our studied population and the INEI population group living in the mountains area when comparing individual items (expressed in grams per capita per day; Table 1) or comparing all items classified in the 20 food groups (Fig. 2), except for four groups: milk, soft drinks, sugar and sweets and legumes. The following groups: water and broths, coffee and tea, alcoholic beverages, and nuts were not represented in the official Peruvian dietary survey and they have therefore not been included in the comparative Fig. 2. The values of these 4 groups in g/day for the present study were: water and broths 1173.5; coffee and tea 1.3; alcoholic beverages 1.1; and nuts: 0.3. The low number of participants included in our 24 h dietary recall reflects practical work difficulties in the field associated with cultural mistrust of foreigners. Besides, the low number of women included among participants is related with gender roles established within these rural communities. There, the image of masculinity is associated with men acting as mediators between their family or community and the outside world, while the femininity image keeps women firmly inside the community (Harvey, 2008). According to these roles, great effort was needed to obtain women report, because in most cases only men agreed to give information to foreigner interviewers. These practical limitations were overcome by using the qualitative frequency questionnaire and the official Peruvian dietary survey as internal and external validations to support our own results, which represented a specific population living in the mountains and in a rural area. 3.3. Energy and nutrient intake The mean energy intake of the interviewed population was 1990 kcal per day. Table 2 shows the results of the energy and nutrient intake values calculated from the 24 h dietary recall for the interviewed population depending on sex and age. In the same table, the values of energy and nutrient intake requirements for Peruvian population living in rural areas recommended by the Health Ministry of Peru (Ministerio de Salud del Perú, 2012b) and by the FAO/WHO/UNU recommendations (FAO, WHO & UNU, 2001) are given. 257 3.4. Samples obtained in each sampling area In total, 145 samples of 24 different foods and beverage items were collected (Table 3). During the collection of the samples, the peasants from Tual and La Pajuela, complained about a decrease in the agricultural productivity and crops diversity. This fact is reflected in Table 3 showing a lower variety of foods available in these populations compared to Porcón Bajo and La Ramada. In Tual and La Pajuela, the population was, therefore, more dependent on the local market products. 3.5. Heavy metals and metalloids concentration in food and water All food samples were checked to be under the maximum levels of Cd, Hg and Pb established by the EU1881-2006 (Commission of the European Communities, 2006) for each foodstuff, as no official local regulation exists. Water samples were checked to be under the recommended levels in the WHO Guidelines for drinking-water quality concerning As, Cd, Hg, Pb, Al, Cr, and Cu (WHO, 2008). There were six samples with an analytical value over official established maximum values. Metals exceeded per sample are detailed in Table 3. The samples over the maximum values were distributed over the sampling areas as follows: zero samples in Porcón Bajo, zero in La Ramada, two in Tual (milk and oca), two in La Pajuela (two water samples) and two in the Cajamarca market (rice and carrot) (see Table 3). There was a statistically significant difference in the frequency of samples exceeding limits between La Pajuela, the closest sampling area to the gold mine, and Porcón Bajo or La Ramada (p = 0.038 and p = 0.036 respectively), the two furthest sampling areas to the gold mine. There were no significant differences between any other groups. 3.6. Heavy metals and metalloids exposure The total intake of each metal through the diet per each food type was calculated per individual following two estimations (LB and UB) and using the intake results of our 24 h dietary recall. The mean (and max–min values) of all individual calculations for the LB and the UB estimations, for the total population and the Consumed foods per capita in g/day 500 400 300 Study area (rural population of the mountains) Rural area (INEI, 2012) Mountains area (INEI, 2012) Lima Metropolitan area (INEI, 2012) 200 100 0 Fig. 2. Consumed foods per capita classified in groups and expressed in grams per day. The graph shows the result of the 24 h dietary recall of the study area (black, rural population of the mountains area) and the results of an official Peruvian dietary survey (INEI, 2012) in rural population (grey), in population of the mountains (white) and, in population from the Lima metropolitan area (striped). 258 M. Barenys et al. / Food and Chemical Toxicology 71 (2014) 254–263 Table 2 Recommended and calculated mean values of energy and nutrient intake for male and female population between 12 and 59 years old.% Protein,% Lipid, and% Carbohydrate are obtained from Expert Consultation on diet, nutrition and the prevention of chronic diseases, WHO/FAO 2003; the range shown in total Energy comprises the recommended values for Peruvian rural population in Requerimientos de energía para la población peruana, Ministerio de Salud de Perú 2012 and the recommended values for light physical activity in Human energy requirements, FAO/WHO/UNU, 2001; Calcium, Iron and Vit. A are obtained from Human vitamin and mineral requirements, FAO/WHO, 2001. RECOMMENDED DAILY VALUES Males 12 to 17 Females 18 to 29 30 to 59 Protein% of total energy Lipid% of total energy Carbohydrate% of total energy Total energy (kcal) Protein (g) Lipid (g) Carbohydrate (g) Ca (mg) Fe (mg) Vit. A (lg) 18 to 29 30 to 59 1905–2125 47.6–79.7 31.8–70.8 262–398 1300 9–65 600 1900–2154 47.5–80.8 31.7–71.8 261–404 1000 20–59 500 1975–2216 49.4–83.1 32.9–73.9 272–416 1000–1300 8–59 500 10–15% 15–30% 55–75% Total energy (kcal) Protein (g) Lipid (g) Carbohydrate (g) Ca (mg) Fe (mg) Vit. A (lg) CALCULATED VALUESa 12 to 17 2118–2900 53.0–109 35.3–96.7 291–544 1300 13–38 600 2375–2871 59.4–108 39.6–95.7 327–538 1000 9–27 600 2350–2836 58.8–106 39.2–94.5 323–532 1000 9–27 600 All males 12–17 N = 20 18–29 N=2 30–59 N=6 All females 12–17 N=4 18–29 N=2 30–59 N=2 1969* 62.5 45.9 333 621* 13.5 686 1930* 63.7 46.2 322 652* 13.5 834 2127* 63.0 45.7 374 596* 15.6 360* 2046* 58.2* 44.9 356 525* 12.8 303* 2014 63.3 49.7 368 661* 19.7 721 2516$ 81.6$ 71.2$ 404$ 851* 24.4 916 1633* 48.2 34.2 295 386* 11.7* 122* 1751* 41.5* 22.2* 367 555* 18.2 931 Males Females a Values calculated with the information from the 24 h dietary recall performed in this study. Indicates values under the recommendations. Indicates values above the recommendations. * $ Table 3 Number and type of samples collected in each sampling area and in total. Number of samples exceeding the maximum official values. Total number of samples collected Water samples collected Food samples collected Types of food items collected Water samples exceeding official maximum valuesa Food samples exceeding official maximum valuesb Sample and heavy metal exceeding maximum values a b * Porcón Bajo La Ramada Tual La Pajuela Cajamarca market Total 39 4 35 14 0 0 – 40 4 36 13 0 0 – 24 4 20 7 0 2 Milk Pb/Oca Pb 10 3 7 6 2* 0 Water As, Pb/Water As 32 0 32 13 0 2 Rice Cd/Carrot Cd 145 15 130 23 2 4 WHO Guidelines for drinking-water quality (WHO, 2008). EU1881-2006 (Commission of the European Communities, 2006) for at least one heavy metal. Statistically significant (p < 0.05) compared to Porcón Bajo and La Ramada. population in each sampling area, were compared to official reference values for each metal as detailed in Table 4. Mean intake results for the total population exceeded the established limit values for Cd, As and Pb; while for Hg, Al, Zn, Cu and Cr mean intakes, limits were not exceeded. Main food items contributing to the mean intake of As, Cd and Pb in the total population where: rice and water (As), rice and potatoes (Cd) and water and potatoes (Pb). When analysing As, Cd and Pb results obtained in each of the four separated sampling areas, a different pattern of regional metal intake distribution and a different pattern of main contributing items was observed for each metal. To illustrate these differences, the contribution of potatoes, water and rice to the total metal intake depending on the sampling area is presented in Fig. 3. In the As case, and following EFSA directives (EFSA, 2009), total As was regarded as inorganic As because fish and shellfish consumption was lower than 0.2% of the total intake in the studied population. Arsenic mean intakes were higher than the BMDL01 for cancers and skin lesions in humans (0.3 lg/kg b.w. per day; EFSA, 2009) in all sampling areas, ranging from 0.32 lg/kg b.w. per day in Porcón Bajo (mean LB) to 0.71 lg/kg b.w. per day in La Pajuela (mean UB). Moreover, in La Pajuela, which is the closest sampling area to the gold mine, mean intakes were significantly higher than in the other three areas (p < 0.05; Table 4). Mean LB intakes of Tual and La Ramada, were also significantly higher than the mean LB intake of Porcón Bajo, the furthest sampling area from the gold mine. Cd mean intakes were above reference values (0.35 lg/kg b.w. per day: TWI/7; EFSA, 2012a) in all sampling areas except in Porcón Bajo and La Pajuela when considering the LB estimation. Intake values covered a small range from 0.32 lg/kg b.w. in Porcón Bajo (LB) to 0.45 lg/kg b.w. in La Ramada (UB), and in this case there was no statistical correlation with closeness to the gold mine. Mean LB and UB Pb intakes were in general under reference values, except in La Pajuela (BMDL10 for nephrotoxicity and BMDL01 for cardiovascular effects in humans; EFSA, 2010), where the whole studied population (LB min–max and UB min–max) presented intakes higher than the BMDL01 and where in addition, the results were significantly higher than in all other studied areas. Statistical analysis of the LB mean intakes revealed that Pb intakes were significantly higher, the closest the population to the gold mine was. It is important to remark that in this study total Hg was regarded as inorganic Hg due to the low consumption of fish and M. Barenys et al. / Food and Chemical Toxicology 71 (2014) 254–263 259 Table 4 Heavy metals and metalloids intake in lg/kg body weight per day represented as mean, maximum and minimum value for the lowerbound (LB) and the upperbound (UB) estimation for the total population and the population in each sampling area. Intake data were obtained from the 24-h dietary recall. Reference values: Provisional Tolerable Weekly Intake (PTWI), Benchmark Dose Low Confidence Limits of the 1% and 10% effects (BMDL01, BMDL10), Tolerable Weekly Intake (TWI) and Tolerable Upper Intake Level (TU) of the respective heavy metal or metalloid established by official health agencies. Calculations were performed if necessary to express all the values in lg/kg b.w. per day. Heavy metals and metalloids intake in lg/kg b.w. per day mean LB (min–max) mean UB (min–max) (Calculations based on the 24 h dietary recall data) Total population As Porcón Bajo La Ramada Tual La Pajuela 0.32 (0.02–0.67) 0.42 (0.14–0.83) 0.39 (0.10–0.75) 0.67 (0.25–1.29)pb,r,t 0.50 (0.18–0.94) 0.42 (0.14–0.81) 0.46 (0.16–0.87) 0.45 (0.15–0.83) 0.71 (0.27–1.31)pb,r,t 0.36 (0.11–0.67) 0.32 (0.08–0.63) 0.40 (0.17–0.72)pb 0.37 (0.11–0.69)pb 0.34 (0.08–0.69) 0.42 (0.16–0.75) 0.37 (0.15–0.71) pb 0.45 (0.20–0.80) 0.43 (0.16–0.76) 0.40 (0.14–0.75) Hg 0.004 (0–0.018) 0.25 (0.12–0.40) 0.006 (0–0.038) 0.25 (0.13–0.39) 0.003 (0–0.022) 0.25 (0.13–0.39) 0.005 (0–0.025) 0.25 (0.12–0.39) 0.003 (0–0.013) 0.24 (0.12–0.38) 0.57 lg/kg b.w. per day (TWI/7), EFSA (2012b) Pb 0.34 (0.18–0.70) 0.03 (0–0.16) 0.05 (0–0.24)pb 0.13 (0.04–0.40)pb,r 1.52 (0.74–2.93)pb,r,t 0.91 (0.40–1.51) 0.49 (0.24–0.79) 0.51 (0.27–0.78) 0.55 (0.28–0.95) 1.89 (0.93–3.42)pb,r,t 3.57 lg/kg b.w. per day (PTWI/7), WHO (1986). Withdrawn by EFSA (2010) 1.5 lg/kg b.w. per day (BMDL01 cardiovascular effects in humans) 0.63 lg/kg b.w. per day (BMDL10 nephrotoxicity in humans), EFSA (2010) 118 (28–553) 129 (9–703) 127 (29–504) 65 (7–660) 118 (39–552) 286 lg/kg b.w. per day (PTWI/7), FAO/WHO (2011) 125 (31–562) 134 (13–707) 129 (29–505) 71 (11–667) 123 (43–505) Zn 125 (37–278) 126 (38–279) 138 (36–344) 140 (37–345) 130 (42–286) 131 (43–287) 121 (35–286) 123 (36–287) 109 (27–226) 112 (30–228) 416 lg/kg b.w. per day (TU/60 kg), EFSA (2006) Cu 15 (5–33) 16 (6–35) 18 (7–36) 19 (8–38) 13 (4–34) 14 (5–35) 14 (5–29) 15 (6–31) 11 (3–29) 12 (3–30) 83 lg/kg b.w. per day (TU/60 kg), EFSA (2006) Cr 3.31 (0.99–6.14) 5.58 (2.49–9.63) 3.37 (0.73–10.18) 5.93 (2.47–13.77) 2.62 (0.50–4.81) 5.07 (2.43–8.15) 3.44 (1.39–5.89) 5.55 (2.73–8.41) 2.06 (0.16–5.54) 4.92 (2.30–9.37) No reference value can be derived. No evidence of adverse effects associated with supplementary intake of chromium up to a dose of 16.7 lg/kg b.w. per day (1 mg/d/60), EFSA (2006) Al pb 2.14 lg/kg b.w. per day (PTWI/7), FAO/WHO (1989). Withdrawn by EFSA (2009) 0.3–8 lg/kg b.w. per day (BMDL01 cancers and skin lesions in humans), EFSA (2009) 0.44 (0.14–0.86) Cd pb Reference value 0.35 lg/kg b.w. per day (TWI/7), EFSA (2012a) Underlined values: values exceeding at least one of the respective listed references. Statistical analysis between the 4 different sampling areas was performed with Student’s t test and only for those metals whose mean intakes exceeded reference values. Significance threshold established at p < 0.05. pb Significantly different from Porcón Bajo. r Significantly different from La Ramada. t Significantly different from Tual. shellfish (<0.2%; EFSA, 2012b) and therefore the guidance PTWI was not exceeded (0.57 lg/kg b.w. per day: PTWI/7; EFSA, 2012b). Potential daily Al intakes exceeded PTWI/7 reference value only when the maximum values were considered. To verify the results of the heavy metal and metalloids risk assessment performed using our own 24 h dietary recall data, a second evaluation was performed using the intake data obtained from the official Peruvian dietary survey (INEI, 2012) as detailed in Table 5. The obtained results were based on dietary information of a much broader population, and indeed, reflected the same pattern of heavy metals and metalloids intake, thus supporting our own calculations. 4. Discussion 4.1. Dietary survey The results of the 24 h dietary recall show that the main consumed food groups in the studied population are potatoes, milk, soft drinks, vegetables, meat and eggs, and rice, while legumes, and fish are among the less consumed foods. The data obtained from this survey has several limitations: the number of interviewed people is low, the ages and sexes of the subjects are not homogeneously distributed and there is no seasonal monitoring. In general, using a 24 h dietary recall to collect intake data is also associated to two important limitations: the limitation of the memory of the interviewed person, and the lack of intraindividual monitorization (day-to-day variation). However, this kind of survey presents the advantage of capturing spontaneous intake with a high detail and allows monitoring of illiterate population. Besides, a qualitative frequency survey was also performed to double-check the information obtained, and indeed, the results of both surveys were very similar. Moreover, while the actual study was performed, an official dietary survey from the Peruvian National Institute of Statistics and Informatics (INEI) was published (INEI, 2012), supporting our own results. A specific group fulfilling the characteristics of our studied population was not present in the official Peruvian dietary survey, but two groups with similar features were included: the rural group and the mountains group. The diet of the rural group includes information on rural populations from the coast, from the mountains and from the rainforest geographical areas, and the diet of the mountain group includes information on rural and urban populations from the mountains. The results of our survey compare very well to the results of these two groups, being more similar to the mountain group. Between our group and the mountain group, the main food groups with similar consumption per capita in g/day are: potato, vegetables, rice, bread, oils and other fats, other dairy products, herbs and spices, and fish. The main differences between these two populations lay in the milk and soft drinks groups, being the consumption of the studied population higher than the consumption of the mountain population for both. The high consumption of soft drinks compares very well with the results obtained in Lima Metropolitan Area in the official Peruvian dietary survey (INEI, 2012), showing that our results are high, but not different than in other parts of 260 M. Barenys et al. / Food and Chemical Toxicology 71 (2014) 254–263 Contribution to metal intake (µg/day) Other foods Potatoes Water Rice Pb Cd As Fig. 3. Contribution of rice, water, potatoes and all other analysed food items to the mean intake of As, Cd and Pb in lg/day, considering the mean intake of each food item and the lowerbound estimation of the metal analysis. the country. Specific evaluations of fluid intake patterns among adolescents in Brazil also reported high intakes of milk and soft drinks (age group 11–17 years old: milk and dairy products = 0.32 L/day; carbonated soft drinks = 0.48 L/day; Feferbaum et al., 2012). Besides, Cajamarca is one of the main milk and dairy producer regions in Peru (Ministerio de Agricultura del Perú, 2009). It is also important to note that soft drinks are a prototypical seasonal product, and their sales tend to increase during the dry season (May–October; Banco Wiese Sudameris, 2002), period when our survey was conducted. Therefore, it is reasonable to interpret that the high consumption of milk and soft drinks in our 24 h dietary recall arise from the overrepresentation of teenagers, and possibly from regional/seasonal variations in the consumption of soft drinks. Summarizing, despite the limitations of our survey, the official Peruvian dietary survey (INEI, 2012), and specifically, its data on the mountains population group, supports our results and allows the validation of our data. Consequently, the results were considered representative enough to be used to calculate the energy and nutrients intake and to decide which food items should be sampled for the heavy metal and metalloid risk assessment. The mean energy intake of the studied population was 1990 kcal per day. When the energy intake was calculated depending on the sex and age, it was clear that the intake was not covering the energy needs for any of the male groups. Among females, the teenager group had an energy intake exceeding the recommendations while the other two groups did not reach them for some nutrients. In the group between 18 and 29 years old, iron intakes were almost 50% of the recommended values, and in all groups, the calcium intake was not enough to cover the recommendations. The low number of female participants does not allow drawing conclusions from this data, as the results could easily be biased. However, regarding the male group, the high number of teenagers participating allows confirming that the energy intake of this group is not correctly covered. It is also of special concern the calcium consumption which is not meeting the recommendations in none of the studied groups. The Peruvian Health Ministry already reported extremely low calcium intakes within Peruvian population (90% of population consuming 150 mg calcium per day or less; Ministerio de Salud del Perú, 2012a). In our study, calcium intakes did not meet recommendations, but were higher than reported by the Peruvian Health Ministry, mainly due to higher consumptions of milk and cabbage within our studied population. These intake differences, could account for about 200 mg of calcium per day more. Also important are the low intakes of vitamin A in the male groups between 18 and 59 years old, although the number of participants in these groups was smaller. The low energy intake combined with a low consumption of milk, other dairy products and vegetables, even if higher than in other Peruvian areas, could explain the low intakes of calcium and vitamin A observed in these groups. The energy intake of other rural communities living in the highlands in Peru, was reported to be also lower than the general recommendations or even the Peruvian recommendations for rural populations. For example, in two surveys at Vicos Hacienda (including 37 and 40 households), the mean energy intake reported was 1481 and 1509 kcal per day and in the Chacan community was 1194 and 1404 kcal per day (43 and 39 households; Collazos et al., 1954). On the other hand, the Nuñoa community, also living in the highlands, was reported to have a mean energy intake of 2671 or 3170 kcal per day (39 households; Mazess and Baker, 1964). Therefore, general conclusions cannot be extracted for these communities regarding the energy intake, as local characteristics and local food availabilities may cause important differences. Nevertheless, a common characteristic was observed in the previous reports of these three communities: the levels of calcium and vitamin A were really below the recommended values (Mazess and Baker, 1964) and this was also observed in some (vitamin A) or all age groups (calcium) of our study. 4.2. Risk assessment of heavy metals and metalloids exposure The final phase of this study evaluates the potential adverse health effects in the study population due to heavy metals and metalloids exposure through the diet. It was beyond the aim of this study to evaluate different routes of exposure as for example inhalation exposure or other scenarios like occupational exposure. In this sense, higher exposures than those reported by this study due to other exposure routes cannot be excluded. From our results, aluminium daily intakes exceeded PTWI values only when considering maximum calculations. However, this PTWI value is based on LOELs of animal studies, applying several safety factors and no clear health effects have been proved in humans due to Al through the diet (FAO/WHO, 2011). Calculated mean daily intakes of Hg, Al, Zn, Cu or Cr were under the tolerable intake references established by EFSA and FAO/WHO (EFSA, 2012b, 2006; FAO/WHO, 2011) and therefore did not suppose a threat to human health. The assessment of the risk associated to As, Cd and Pb intakes was, however, more disturbing. M. Barenys et al. / Food and Chemical Toxicology 71 (2014) 254–263 261 Table 5 Heavy metals and metalloids intake in lg/kg body weight per day represented as the mean value and the 80th percentile value for the lowerbound (LB) and the upperbound (UB) estimation for the total population and the population in each sampling area. Intake data were obtained from the official Peruvian dietary survey (INEI, 2012). Reference values: Provisional Tolerable Weekly Intake (PTWI), Benchmark Dose Low Confidence Limits of the 1% and 10% effects (BMDL01, BMDL10), Tolerable Weekly Intake (TWI) and Tolerable Upper Intake Level (TU) of the respective heavy metal or metalloid established by official health agencies. Calculations were performed if necessary to express all the values in lg/ kg b.w. per day. Heavy metals and metalloids intake in lg/kg b.w. per day mean LB/LB 80th percentile mean UB/UB 80th percentile (Calculations based on the official Peruvian dietary survey data) Reference value Total population Porcón Bajo La Ramada Tual La Pajuela 0.36/0.51 0.27/0.40 0.35/0.49 0.32/0.46 0.55/0.73 0.40/0.56 0.34/0.47 0.38/0.52 0.36/0.51 0.57/0.75 2.14 lg/kg b.w. per day (PTWI/7), FAO/WHO (1989). Withdrawn by EFSA (2009) 0.3–8 lg/kg b.w. per day (BMDL01 cancers and skin lesions in humans), EFSA (2009) 0.30/0.40 0.27/0.37 0.32/0.42 0.31/0.46 0.29/0.40 0.35 lg/kg b.w. per day (TWI/7), EFSA (2012a) 0.34/0.45 0.31/0.42 0.36/0.47 0.35/0.46 0.33/0.45 Hg 0.002/0.003 0.18/0.21 0.003/0.004 0.18/0.21 0.001/0.003 0.17/0.21 0.003/0.005 0.17/0.21 0.002/0.003 0.17/0.20 0.57 lg/kg b.w. per day (TWI/7), EFSA (2012b) Pb 0.30/0.34 0.04/0.04 0.06/0.05 0.08/0.07 1.25/1.47 0.65/0.80 0.36/0.42 0.37/0.43 0.38/0.43 1.50/1.77 3.57 lg/kg b.w. per day (PTWI/7), WHO (1986) Withdrawn by EFSA (2010) 1.5 lg/kg b.w. per day (BMDL01 cardiovascular effects in humans) 0.63 lg/kg b.w. per day (BMDL10 nephrotoxicity in humans), EFSA (2010) Al 126/90 130/97 160/97 163/101 124/109 125/110 45/35 50/40 121/99 124/103 286 lg/kg b.w. per day (PTWI/7), FAO/WHO (2011) Zn 101/120 102/122 113/135 115/137 104/125 105/126 92/116 93/118 90/109 92/111 416 lg/kg b.w. per day (TU/60 kg), EFSA (2006) Cu 15/15 15/16 17/17 18/18 14/15 14/15 13/14 14/15 12/13 12/14 83 lg/kg b.w. per day (TU/60 kg), EFSA (2006) Cr 2.62/3.26 4.29/5.19 3.27/3.66 5.07/5.78 2.05/2.76 3.83/4.83 2.39/3.02 3.94/4.95 1.99/2.57 3.92/4.79 No reference value can be derived. No evidence of adverse effects associated with supplementary intake of chromium up to a dose of 16.7 lg/kg b.w. per day (1 mg/d/60), EFSA (2006) As Cd Underlined values: values exceeding at least one of the respective listed references. Arsenic exposure can cause cancers of the urinary bladder, lung and skin (EFSA, 2009). On 2009 EFSA established a BMDL01 for cancers and skin lesions in humans at a range of 0.3–8 lg/kg b.w. per day. This value is slightly exceeded by the population living in the four studied areas, therefore, As intake is estimated to be high, and an increased risk of cancers and skin lesions in this population cannot be excluded. EFSA report already warned that high consumers of rice would be included in the high exposure level groups and indeed, in our study main items contributing to As intake were rice and water. Rice is a highly consumed item which cannot be produced at Andean altitudes. Population from this area relies on the market supply and our samples were, therefore, collected in the Cajamarca market. Food products displayed in the market come from several regions of the Cajamarca department or even from other departments. For this reason, the origin of the rice sampled in the market could not be determined in our study, and the gold mine cannot be excluded as the origin of this contamination as rice productive areas are also established in downstream areas with water sources influenced by the gold mine. While rice contribution to As intake was the same in all studied areas because it was a non-local item, there was a variation on water contribution to As intake among the studied areas. This item was the responsible one for the statistical differences found among the As intake in la Pajuela and all other areas, or among La Ramada–Tual and Porcón Bajo (Table 4). These results indicate that the closer are the water sources of a population to the gold mine, the higher is the contribution of this water to the daily As intake (Fig. 3). Cd exposure can cause osteomalacia, renal disfunction and has been statistically associated with higher risk of cancer (EFSA, 2012a). Cd intake values of the studied population slightly exceeded the Tolerable Weekly Intake of 2.5 lg/kg b.w. established by EFSA. This reference value was set to ensure a high level of protection of all consumers, including exposed and vulnerable subgroups of the population. Based on EFSA Panel recommendations, our results would advise for a need to reduce cadmium exposure in the population because of the limited safety margin, and would confirm that in some areas, cadmium exposure could exceed health-based guidance values. It was already pointed by EFSA report that in general, foods that are consumed in larger quantities (and not the food with the highest cadmium levels) have the greatest impact on cadmium dietary exposure. In this sense, potatoes and rice were already identified as important contributors to Cd dietary exposure, supporting the results of our study where the main items contributing to Cd intake were rice and potatoes. Potatoes contribution to total Cd intake was similar in all studied areas and there were no statistical differences correlating with closeness to the gold mine (Fig. 3 and Table 4). Similarly to the arsenic situation mentioned above, rice contribution to Cd intake was the same in all areas because it was a non-local item. Although the origin of the contamination cannot be determined in the rice product particular case, it is important to remark that more control should be performed in market products in order to ensure the compliance of the Cd and As limits by commercialized food items. In 2010, the EFSA established two Benchmark Dose Low Confidence Limits (BMDLs) on respect to effects of lead exposure in humans: BMDL10 for nephrotoxicity at 0.63 lg/kg b.w. and BMDL01 for cardiovascular effects at 1.5 lg/kg b.w. According to our calculations, mean Pb intakes were under these low confidence limits for three out of the four areas of study, Porcón Bajo, La Ramada and Tual. But in La Pajuela, the intake values were above both 262 M. Barenys et al. / Food and Chemical Toxicology 71 (2014) 254–263 BMDLs for the whole studied population (LB min–max and UB min–max), being water the main contributing item to Pb intake (Fig. 3). In a Margin of Exposure (MOE) approach, MOE is the ratio of the low confidence limit (in this case the BMDL) divided by the calculated intake of the substance. In this manner, a MOE = 1 reflects that the amount of metal consumed is the same than the dose considered to have measurable adverse health effects. A MOE = 100 would reflect that the intake is hundred times lower than this dose. In general, it is not enough to check if intakes exceed BMDLs or not, instead, high MOEs would be desirable to ensure absence of effects in the population. In our particular case of study, the MOE in La Pajuela was between 0.41 and 0.33 regarding nephrotoxic effects and between 0.98 and 0.78 for cardiovascular effects. In this concrete population the calculated Pb intake was higher than the two established reference BMDLs, and therefore, the MOE was lower than 1, rising a clear concern on health effects derived from Pb exposure through the diet. Moreover, the EFSA establishes a third BMDL for developmental neurotoxicity, which would apply to foetus and infants and corresponds to 0.5 lg/kg b.w (EFSA, 2010). No children were included in our survey, and the low number of participant women at childbearing age, makes very difficult to rely on these results. Despite the limitations, a calculation of the Pb intake among women within 20–40 years included in our study, and eating La Pajuela diet, would result in a Pb intake of 1.40 for the LB (min:1.16; max:1.56) and 1.67 for the UB (min:1.37; max:1.90), values clearly surpassing the neurodevelopmental toxicity BMDL and leaving MOEs as low as 0.32 and 0.26. It is important to remark that concerning Pb intake risk assessment, MOEs where only lower than 1 in La Pajuela, the area where the streams receive the highest influence from the gold mine. From our results, and following official health organizations recommendations, it is reasonable to advice to La Pajuela population not to drink from the water sources in their area. There are several interactions between micronutrients and heavy metals, and a particular case is the role that dietary deficiencies play in metal toxicity (Peraza et al., 1998). For example, diets low in proteins or minerals increase Cd and Pb absorption and toxicity (Goyer, 1997; Peraza et al., 1998). In the present study, the Cd and Pb intakes are already higher than the recommended values but, the observed deficiencies in calcium, iron and proteins could be exacerbating the risk of specific subgroups as well. It is the first time that a risk for the populations’ health is reported in this area, but several previous reports highlighted the environmental risk in the same place (Yacoub et al., 2012, 2013). In the Rejo sub basin, lead concentrations were higher in sediments and waters of streams coming from the gold mine compared to sediments of non-influenced streams (Yacoub et al., 2012, 2013). From these environmental studies, it is clear that the gold mine presence has an impact on the Pb concentrations of the influenced streams and, now, we have shown evidences that the presence of Pb in this area is translated to an increased risk of nephrotoxicity and cardiovascular diseases in the populations living near the gold mine through the diet. 5. Conclusions After a dietary survey and heavy metals risk assessment in a rural population living near a gold mine in the Andes, the following conclusions can be extracted: The mean consumed items by this population are tubers and cereals. The energy, calcium and vitamin A intake values are really below the Peruvian and the WHO/FAO recommended intakes, suggesting a high probability of not covering the requirements of the male population subgroup. The dietary intakes of Pb and As are higher as closer the population to the gold mine is, and it is due to the consumption of local foods and water. There is a concern regarding the dietary exposure to As, Cd and Pb in the studied population. Financial support This project was funded by a University of Barcelona Cooperation Project and a PCI from the Spanish Agency for International Cooperation and Development (AECID) A/025030/09. Conflict of Interest The authors declare that there are no conflicts of interest. Transparency Document The Transparency document associated with this article can be found in the online version. Acknowledgements The authors acknowledge the following groups for its support to this research work: the GRECDH, Development Cooperation and Human Development Research Group of the Polytechnic University of Catalonia, the Scientific-Technical Services of the University of Barcelona, the TECNATOX Center of the Rovira i Virgili University and GRUFIDES, Education and Intervention Group for the Sustainable Development in Cajamarca. References Banco Wiese Sudameris, 2002. Reporte sectorial: Gaseosas. Los bajos precios debilitan a la industria. Departamento de estudios económicos URL (accessed 12.03.2014). Bebbington, A., Humphreys, D., Bury, J., Lingan, J., Muñoz, J.P., Scurrah, M., 2008. Mining and social movements: struggles over livelihood and rural territorial development in the andes. World Dev. 36, 2888–2905. Bejarano, E., Bravo, M., Huamán, M., Huapaya, C., Roca, A., Rojas, E., 2002. Tabla de Composición de Alimentos Industrializados, first ed., first reimpression. Ministerio de Salud, Instituto Nacional de Salud, Lima. Blacksmith Institute. The World’s Worst Pollution Problems: Assessing Health Risks at Hazardous Waste Sites. 2012. (accessed 12.03.2014). Bury, J., 2002. Livelihoods, mining and peasant protests in the Peruvian Andes. JLAG 1, 3–19. Bury, J., 2005. Mining mountains: neoliberalism, land tenure, livelihoods, and the new Peruvian mining industry in Cajamarca. Environ. Plan. A 37, 221–239. Coelho, P.C.S., Teixeira, J.P.F., Gonçalves, O.N.B.S.M., 2011. Mining activities: health impacts. In: Nriagu, J.O. (Ed.), Encyclopedia of Environmental Health. Elsevier, Burlington, pp. 788–802. Collazos, C., White, H.S., Huenemann, R.L., Reh, E., White, P.L., Castellanos, A., Benites, R., Bravo, V., Loo, A., Moscoso, I., Caceres, C., Dieseldorff, A., 1954. Dietary surveys in Peru. III. Chacan and Vicos, rural communities in the Peruvian Andes. J. Am. Diet. Assoc. 30, 1222–1230. Commission of the European Communities, 2006. Commission regulation 1881/ 2006 setting maximum levels for certain contaminants in foodstuffs. Off. J. Eur. Union. L 364, 5–24. Donato, D.B., Nichols, O., Possingham, H., Moore, M., Ricci, P.F., Noller, B.N., 2007. A critical review of the effects of gold cyanide-bearing tailings solutions on wildlife. Environ. Int. 33, 974–984. European Food Safety Authority, 2006. Scientific Committee on Food and Scientific Panel on Dietetic Products, Nutrition and Allergies; Tolerable upper intake levels for vitamins and minerals. (accessed 10.04.2014). European Food Safety Authority, 2009. EFSA Panel on Contaminants in the Food Chain (CONTAM); Scientific Opinion on Arsenic in Food. EFSA J. 7, 1351–1354. European Food Safety Authority, 2010. EFSA Panel on Contaminants in the Food Chain (CONTAM); Scientific Opinion on Lead in Food. EFSA J. 8, 1570–1720. European Food Safety Authority, 2012a. Cadmium dietary exposure in the European population. EFSA J. 10, 2551–2588. European Food Safety Authority, 2012b. EFSA Panel on Contaminants in the Food Chain (CONTAM); Scientific Opinion on the risk for public health related to the presence of mercury and methylmercury in food. EFSA J. 10, 2985–3226. M. Barenys et al. / Food and Chemical Toxicology 71 (2014) 254–263 FAO, WHO, 1989. Evaluation of certain food additives and contaminants: Thirtythird report of the Joint FAO/WHO Expert Committee on Food Additives. WHO Technical Report Series, No. 776. (accessed: 10.04.2014). FAO, WHO, 2001. Human vitamin and mineral requirements. Report of a joint FAO/ WHO expert consultation. FAO Food and nutrition Division. Food and Agricultural Organization, Rome. FAO, WHO, 2011. Evaluation of certain food additives and contaminants: Seventyfourth report of the Joint FAO/WHO Expert Committee on Food Additives. WHO Technical Report Series, No. 966. (accessed 10.04.2014). FAO, WHO, UNU, 2001. Human energy requirements. Report of a joint FAO/WHO/ UNU expert consultation. FAO Food and nutrition technical report series 1. Food and Agricultural Organization, Rome. Farran, A., Zamora, R., Cervera, P., 2004. Food composition tables of CESNID (Centre for Superior Studies in Nutrition and Dietetics), second ed. McGraw-Hill/ Interamericana de España, S.A.U. Ediciones Universitat de Barcelona, Barcelona. Feferbaum, R., de Abreu, L.C., Leone, C., 2012. Fluid intake patterns: an epidemiological study among children and adolescents in Brazil. BMC Public Health. 12, 1005. Goyer, R.A., 1997. Toxic and essential metal interactions. Annu. Rev. Nutr. 17, 37– 50. Harvey, P., 2008. Speech in the communication of gender, in: Intercultural discourse and communication: the essential readings. Kiesling, S.F., Paulston, C.B. (Eds.), Online publication 2008 (Online pp. 176–177, ISBN: 9780470758434), John Wiley & Sons, Blacwell Publishing. Instituto Nacional de Estadística e Informática, INEI, 2012. Perú: Consumo per cápita de los principales alimentos. Encuesta nacional de presupuestos familiares (ENAPREF). INEI, Lima. (accessed 03.12. 2013). Jarvis, A., Younger, P., 2000. Broadening the scope of mine water term environmental impact assessment: a UK perspective. Environ. Impact Assess. Rev. 20, 85–96. 263 Latinfoods. Red Latinoamericana de Composición de Alimentos. (accessed 31.07.2013). Mazess, R.B., Baker, P.T., 1964. Diet of quechua indians living at high altitude: Nuñoa, Peru. Am. J. Clin. Nutr. 15, 341–351. Ministerio de Agricultura del Perú, 2009. Boletín mensual de leche. Ministerio de Agricultura, Lima. (accessed 27.05.2014). Ministerio de Salud del Perú, 2012a. Sala Situacional Alimentaria Nutricional 3: Micronutrientes. Ministerio de Salud, Instituto Nacional de Salud, Lima. (accessed 11.03.2014). Ministerio de Salud del Perú, 2012b. Requerimientos de energía para la población peruana. Ministerio de Salud, Instituto Nacional de Salud, Lima. Palma, I., Farran, A., Cantós, D., 2008. Tablas de composición de alimentos por medidas caseras de consume habitual, first ed. McGraw-Hill Interamericana de España, S.A.U. Ediciones Universitat de Barcelona, Barcelona. Peraza, M.A., Ayala-Fierro, F., Barber, D.S., Casarez, E., Rael, R.T., 1998. Effects of micronutrients on metal toxicity. Environ. Health Perspect. 106, 203–216. Reyes, M., Gómez-Sánchez, I., Espinoza, C., Bravo, F., Ganoza, L., 2009. Tablas peruanas de composición de alimentos, eigth ed. Ministerio de Salud, Instituto Nacional de Salud, Lima. WHO, 1986. Lead (evaluation of health risk to infants and children). World Health Organization. (accessed 10.04.2014). WHO, 2008. Guidelines for drinking-water quality, third ed. World Health Organization. WHO/FAO, 2003. Diet, nutrition and the prevention of chronic diseases. Report of a join WHO/FAO expert consultation. World Health Organization, Geneva. Yacoub, C., Pérez-Foguet, A., Miralles, N., 2012. Trace metal content of sediments close to mine sites in the Andean region. Sci. World J. 2012, 732519. Yacoub, C., Blazquez, N., Pérez-Foguet, A., Miralles, N., 2013. Spatial and temporal trace metal distribution of a Peruvian basin: recognizing trace metal sources and assessing the potential risk. Environ. Monit. Assess. 185, 7961–7978.