Exposure to mixed metals/metalloids in early childhood: a cross-sectional cohort study in children from Sevilla, Southern Spain

Exposure to mixed metals/metalloids in early childhood: a cross-sectional cohort study in children from Sevilla, Southern Spain María Quintana-Mejía a , María G. Hinojosa a,* , Ana I. Garrido b , Marta Gonz´ alez b , Antonio Mill´ an c , Laura Acosta c , ´ Angela Peria˜ nez c , Isabel M. Moreno a a Department of Nutrition, Bromatology, Toxicology and Legal Medicine. Faculty of Pharmacy, Universidad de Sevilla, 41012, Seville, Spain b Pediatrics and Specialized Areas Unit. Virgen del Rocío University Hospital, 41013 Seville, Spain c Pediatric Clinical Management Unit. Virgen de Valme University Hospital, 41014, Seville, Spain ARTICLE INFO Keywords: Neurodevelopment Early exposure Metals Metalloids Biomonitoring Infant ABSTRACT The synergy of exposure to neurotoxic substances, including some metals and metalloids has emerged as a global concern due to its effects on neurodevelopment. Thus, this study determined the presence of Al, Cr, Mn, Ni, Cu, Zn, As, Se, Cd, and Pb in hair and their relationship with the developmental profile in a cohort of 254 children at 6, 12, 18, and 24 months of age in Seville, Southern Spain. A cross-sectional examination was conducted, including the measurement of metal-metalloid levels in hair using mass spectrometric analysis with inductively coupled plasma (ICP-MS). The children’s developmental profile was assessed using the Battelle Developmental Inventory (BDI). The results showed that each hair sample contained 2 to 10 metals or metalloids. A multiple regression analysis found that a model including all elements—along with factors such as age, number and levels of detected metals/metalloids, maternal education, and developmental measures—was significantly associated with overall development (BDI score) as well as the personal-social, cognitive, and language domains. In this sense, Pb, Al, Mn and As were demonstrated to be the elements with more negative correlations to the different parameters. The interaction effects of metal-metalloid mixtures reinforce global concerns about their differentiated impacts on early childhood, depending on sex and the affected developmental domain. These findings are alarming due to their potential implications as predictors of developmental deficits, psychomotor skills, and future school performance. Thus, environmental surveillance and infant biomonitoring in non-industrial urban settings, such as Seville, highlight the need to redefine territorial strategies for reducing everyday environmental pollutants. This study underscores the importance of addressing invisible chronic exposures and identifying social determinants that modulate neurotoxicity and neurodevelopmental disorders. 1. Introduction From a global perspective, millions of children are susceptible to developmental delaysand neurological disorders after exposure to metals, metalloids, and some other chemical agents (Grandjean and Landrigan, 2014; Guo et al., 2020). The neurotoxic action of these combinations can induce harmful effects in the central nervous system (CNS), affecting brain development and increasing the susceptibility to neurodegenerative disorders later in life (Pistollato et al., 2021). Some metals/metalloids are essential for human health, and their lack and/or excess can cause homeostatic imbalance (Awadh et al., 2023). Several studies have linked exposure to metals and metalloids with decreased IQ, lower academic achievement, lack of focus, hyperactivity, autism spectrum disorder, and mental health issues (Kou et al., 2025; Ouisselsat et al., 2023; Shen et al., 2021; Tabatadze et al., 2015). These findings highlight the urgent need to monitor the impact of such exposures—especially in children—to guide public health policies and protect their well-being and rights (Zdunek et al., 2019). In this regard, human biomonitoring (HBM) constitutes an efficient method for quantifying exposure to toxic substances and risk management in biological samples such as blood, urine, and hair (Zare et al., 2021). In particular, hair is considered an analytical and non-invasive matrix that reflects chronic exposure to xenobiotics and provides insights into environmental and epigenetic factors relevant to pediatric neurodevelopment (Kirichuk et al., 2023; Varrica et al., 2014; Zheng et al., 2021). Moreover, HBM plays a predominant role in diagnosis, therapeutic guidance, * Corresponding author. E-mail address: [email protected] (M.G. Hinojosa). Contents lists available at ScienceDirect Environmental Pollution journal homepage: www.elsevier.com/locate/envpol https://doi.org/10.1016/j.envpol.2025.127261 Received 10 June 2025; Received in revised form 4 September 2025; Accepted 11 October 2025 Environmental Pollution 386 (2025) 127261 Available online 13 October 2025 0269-7491/© 2025 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/bync-nd/4.0/ ). and risk prevention strategies (Islam et al., 2021). Despite increasing evidence of the adverse effects of metals and metalloids, most HBM studies have focused on single-element exposures. Very few investigations have explored the simultaneous exposure to multiple metals/metalloids from gestation to early life stages (Lin et al., 2024). This knowledge gap limits our understanding of how mixed exposures affect early neurodevelopment and health outcomes, particularly during the critical period of the first two years of life. Therefore, this study is motivated by the need to generate evidence on the health impact of combined metal/metalloid exposure in early childhood. Thus, this study aims to (1) assess the levels of exposure to multiple metals and metalloids in children aged 6–24 months using hair as a biomonitoring matrix, (2) evaluate the associations between mixed exposures and neurodevelopmental parameters, (3) explore the relationship between exposure patterns, health information, and lifestyle factors (4) provide evidence that may support preventive strategies and public health policies in Southern Spain. 2. Materials and methods 2.1. Study population The present study is based on prospective birth cohorts in the south of Spain. Participants in this study were children born during July 2020–2022 in two hospitals located in the city of Seville (Spain): Virgen del Rocío University Hospital (VRUH) and Virgen de Valme University Hospital (VVUH), as can be observed in Supplementary Fig. 1. One hundred newborns were selected to participate in this study, and samples were collected every six months (6, 12, 18 and 24 months of age) to study the influence of dietary changes over time on metal content and its possible relationship with neurodevelopmental alterations. The cohorts were established considering the hospital where the participants were born. Participation was offered to all pregnant women that went into delivery at the time and fit in the inclusion criteria, enrolling those who accepted and signed informed consent. Parents were informed about the objectives, scope, importance, limitations, and benefits of the project. To be included, parents had to be residents in the hospital reference area for at least the last 5 years and be able to communicate in Spanish. Also, women must not have followed any assisted reproduction program and have a single pregnancy, regardless of whether they are nulliparous or multiparous. Women who suffered from chronic hypertension, any type of diabetes, thyroid disorders or chronic renal or cardiac disease during pregnancy were excluded. Regarding newborns, the exclusion criteria were established in those who presented risks in their psychomotor development: fetal suffering due to problems in childbirth, birth weight <1500 g and/or gestational age <32 weeks, Apgar <4 at 5 min and/or umbilical arterial pH <7, those with chromosomopathies, CNS malformations, symptomatic neonatal hypoglycemia, needing of prolonged mechanical ventilation, and children with psychoneurosensory risk (large premature babies, congenital intrauterine infections, bilirubin levels>25 mg/dL), as can be seen in Supplementary Fig. 2. A questionnaire on demographic characteristics (age, place of residence and birth, etc.) was provided to the parents via a QR code at the time of delivery. Subsequently, a more detailed questionnaire regarding the dietary habits of their children was administered at all the timepoints (6, 12, 18 and 24 months of age). This questionnaire included questions such as type of breastfeeding (maternal, formula or both), type of cereals (gluten or gluten free), age at which they started eating fruits, vegetables, meat, and fish; amount consumed per week and whether it was organic or not; consumption of baby food jars (organic or not), and consumption of canned fish. The study was approved by the Andalusian Biomedical Research Ethics Coordinating Committee in June 2020 (1479-N-20). 2.2. Study region The study was conducted in Seville, a major urban center in Southern Spain with nearly 1.9 million inhabitants in its metropolitan area. The region shows marked socioeconomic inequalities that may contribute to differential vulnerability to environmental exposures. From an environmental perspective, Seville is affected by traffic-related air pollution with significant concentrations of trace metals such as Pb, Cd, Ni, Zn, and Cu in fine particles (Fern´ andez-Espinosa et al., 2001). Moreover, its proximity to the Iberian Pyrite Belt, one of the largest polymetallic sulfide provinces worldwide, has left a legacy of soil and water contamination (Nieto et al., 2013; Olías et al., 2004), and associations between exposure to these metals and neurodevelopmental alterations have been reported (Capelo et al., 2022). In addition, large rural areas of the province are dedicated to agriculture and livestock farming, where metal exposure may also occur through the use of pesticides containing metallic compounds, particularly copper-based fungicides applied in olive groves and other crops (Casanova et al., 2024; L´ opez et al., 2019). Participants were recruited at two referral hospitals: Virgen del Rocío University Hospital (VRUH) and Virgen de Valme University Hospital (VVUH). VRUH mainly serves areas corresponding to the urban population, directly exposed to traffic-related air pollution and those geographically close to Huelva, a province historically impacted by mining activities, including the Aznalc´ ollar mine accident and contamination from the Iberian Pyrite Belt. VVUH covers a more rural zone characterized by intensive agriculture (olive groves, cereals, and horticultural crops). 2.3. Collection and storage of hair samples Hair samples were taken from the back of the scalp, stored in paper envelopes, labeled with the identification and codes assigned for analysis. The hair samples were taken aseptically to avoid contamination during the analysis phases: sampling of biological material, transport, storage, preparation, and corresponding analysis. 2.4. Standards and chemicals All solutions were prepared using deionized water (18 MΩ-cm at 25 ◦C) obtained from a Milli-Q system (Millipore, Bedford, MA, USA). All laboratory-use material was cleaned and treated using the considerations described by Dahiri et al. (2023). Briefly, the laboratory ware was cleaned with 20 % v/v HNO 3 for 4 h and rinsed three times with deionized water. Suprapur® quality HNO 3 (65 %), HCl (30 %), and H 2 O 2 (30 %) were purchased form Merck (MerckMillipore, Merck KGaA, Darmstadt, Germany) and used as received. The reagent blank solution consisted of 1 % HNO 3 , 0.2 % HCl, and gold. All internal standards were added online in the form of a 1 μ g/mL multielement solution in 1 % HNO 3 . 2.5. Quantification of metals/metalloids in hair The concentrations of 10 metals/metalloid (Al, As, Cd, Cr, Cu, Mn, Ni, Pb, Se, and Zn) were evaluated in hair samples from children. An Anton Paar, Multiwave 3000 system was used to digest hair samples. All digestions were carried out in polytetrafluoroethylene (PTFE) digestionvessels. For predigesting, 0.003–0.3g of hair were used and 4 mL of HNO 3 65 % were added and incubated for 24 h. Afterwards, they were kept at 4 ◦C. The digestion program for 8 vessels consisted of an initial power ramp from 0 to 800W of 10 min, followed by a continuous power stage of 800W for 20 min, and a final cooling stage of 15 min, with no power applied. For digestion, the sample was transferred to a digestion vessel and another 1 mL of deionized water was used to gather the remains. All the measured samples were prepared in triplicate. M. Quintana-Mejía et al. Environmental Pollution 386 (2025) 127261 2 2.6. Instrumentation Mass spectrometric measurements using inductively coupled plasma (ICP-MS) were carried out with an Agilent 7500c ICP-MS system (Agilent Technologies, Tokyo, Japan), which includes an Integrated Autosampler and an Octupole Reaction System (ORS). Helium gas was used in collision mode for the measurement of Mn and Cd, while argon was employed in standard mode for Pb analysis. Standard Ni cones for sampling and skimming (with internal diameters of 1.0 mm and 0.7 mm, respectively) were used. Samples were introduced through a Babington PEEK (poly-ether-ether-ketone) nebulizer paired with a quartz Scotttype double-pass spray chamber (Agilent Technologies, Tokyo, Japan). To maintain temperature stability and reduce vapor interference, the spray chamber was cooled to 2 ◦C with water. The ICP torch was composed of a three-cylinder structure, featuring an injector diameter of 2.5 mm. A Shield Torch was employed during the entire analysis process, and all instrumental parameters were optimized daily by aspirating the tuning solution. 2.7. Assessment of child development The assessment of child development was performed using the Battelle Development Inventory (BDI) applicable from birth to 8 years of age with an average duration between 10 and 30 min. The test allows the measurement of development in global domains and motor, language, cognitive, personal/social, and adaptive skills. The test has been widely used for its reliability, content and criterion validity (Putnick et al., 2024). The inventory is a reliable tool for early detection of developmental delays, as well as developmental strengths or weaknesses based on comprehensive observation of skills (Troxel et al., 2024). 2.8. Statistical analysis Information regarding sociodemographic traits and hair metal/ metalloid levels in children was examined by age and gender. The variables were shown as percentages or mean ±standard deviation of the mean (mean ±SD). The normality of quantitative variables was tested using the Kolmogorov Smirnov method. The Mann-Whitney U test or Kruskal Wallis test was employed for comparing means among the groups, respectively. Spearman correlation was employed to assess the relationship between levels of metals and metalloids, sociodemographic information, and scores from the BDI. A simple linear regression model was used to determine whether there was a significant linear relationship between BDI scores, age, maternal education, and the number of metals and metalloids. Multiple linear regression was employed to assess if the BDI score outcomes aligned with levels of metals-metalloids and mother’s education. The levels of significance were p <0.0001, p < 0.001 and p <0.05. Statistical analysis was performed using GraphPad Prism 8.0 (GraphPad Software, Inc., San Diego, USA). All the equations can be found in Supplementary Table 1. 3. Results 3.1. Sociodemographic characteristics of the participants General sociodemographic characteristics of the 254 participants are indicated in Table 1, divided according to their distribution at different timepoints. 3.2. Metal and metalloid detection 3.2.1. Detection of metals and metalloids in children’s hair samples The overall distribution of metals and metalloids in children’s hair is shown in Table 2. The frequency of detection was the highest in Cu, and the lowest in Cd. Furthermore, at least 2 metals/metalloids were detected for each participant, being 10 the highest number of compounds detected in a sample (with a mean value of 8.37 metals per sample). 3.2.2. Detection of metals and metalloids by age group Concerning the detection frequency divided according to the different timepoints, the data is presented in Supplementary Table 2. The table contains the detection of various metals and metalloids in hair samples collected at 6, 12, 18, and 24 months old. Observations revealed varying detection patterns over time. Thus, Zn, Cr, Pb, Mn, Al, Ni, As, and Cd generally showed a decrease in levels as age increased. The most significant differences across timepoints were observed for Cr, Mn, Al, Ni, As, and Se. In addition, the frequency of detection of these metals and metalloids by sexes are presented in Fig. 1. Only significant differences in Cu, Pb, and Al levels were detected between sexes, with higher values in females. Furthermore, Fig. 2 shows the distribution of total metal and metalloid concentrations in hair samples by age groups (6, 12, 18, and 24 months). Significant age-related decreased levels were observed for Zn, Cr, Pb, Mn, Al, Se, Ni, As, and Cd with the highest concentrations at 6 months old. 3.3. BDI scores The distributions, means, and SD for BDI scores are presented in Table 3, showing the distribution of overall performance and by sexes. Table 1 Characteristics and lifestyle data from participants. Data given in percentages or mean ±SD, depending on the variable. Variable Category Total (n) Months (%) 6 12 18 24 Participants Number 254 37.01 28.35 18.90 15.75 VVUH 93 41.94 24.73 17.20 16.13 VRUH 161 34.16 30.43 19.88 15.53 Sex Female 136 37.50 27.94 19.12 15.44 Male 116 37.07 28.45 18.10 16.38 Breastfeeding Yes 121 43.80 27.27 15.70 13.22 Time (months) 6.01 ± 4.40 4.70 ± 2.22 6.26 ± 2.98 6.83 ± 4.37 9.13 ± 9.07 No (only formula) 15 25.00 31.25 31.25 6.25 Combined feeding 73 46.58 24.66 15.07 13.70 Dietary habits Gluten-free Cereals 108 37.04 28.70 20.37 13.89 Glutencontaining Cereals 107 30.84 32.71 20.56 15.89 Ecological fruit 130 39.23 29.23 16.92 14.62 Ecological vegetables 126 37.30 30.16 17.46 15.08 Chicken 113 30.97 32.74 20.35 15.93 Beef 105 24.76 35.24 21.90 18.10 Vegetable oil 114 29.82 33.33 20.18 16.67 Canned seafood servings (times a week) 1.75 ± 1.04 1.95 ± 1.00 1.57 ± 0.60 1.42 ± 0.51 2.00 ± 2.05 Canned nonfish servings (times a week) 1.59 ± 0.68 1.63 ± 0.54 1.68 ± 0.79 1.56 ± 0.86 1.18 ± 0.40 Mother’s education Primary education 11 2.20 4.17 10.00 2.20 Secondary education 51 24.18 19.44 20.00 24.18 Higher education 192 73.63 76.39 70.00 73.63 Abbreviations: VVUH: Virgen del Varme University Hospital; VRUH: Virgen del Rocio University Hospital. M. Quintana-Mejía et al. Environmental Pollution 386 (2025) 127261 3 The values represent the proportion of children scoring within the expected range for their age (±1), in the process of acquiring the domain (−1.5), and those with suspected delays in developmental domain acquisition (−2). Overall, about 55 % of children scored within ±1 SD of expected development on the BDI, with domain-specific rates of 52.6 % (personal-social), 79.6 % (adaptive), 86.9 % (motor), 63.5 % (language) and 68.6 % (cognitive). Girls tended to score slightly higher than boys in most domains, but none of the sex differences reached statistical significance (all p >0.13). 3.4. Correlation analysis 3.4.1. Correlations between the sociodemographic characteristics of participants and the detection of metals and metalloids Notably, age negatively correlated with Al, Mn, Ni, As, and Pb. The sex variable exhibits a significant negative correlation with Pb. Exclusive breastfeeding for up to six months showed strong correlations with Al, Cr, Ni, and Cd (Table 4). 3.4.2. Correlation between metals and metalloids, month old, mother’s education and BDI The correlation analysis revealed distinct patterns in the association between metal-metalloid levels and neurodevelopmental scales (Table 5). Thus, Al, Mn, Ni, and As demonstrated consistently negative correlations across multiple domains. A strong inverse relationship was observed with As across all scales. Similarly, Al and Mn exhibited significant associations. Conversely, Cu and Zn presented a mixed correlation profile, with Cu showing positive associations with motor development, while Zn displayed weaker trends across domains. The presence of Pb correlated negatively with cognitive and motor functions. In addition, maternal education and sex showed varied associations, with negative correlations emerging in specific contexts, potentially indicating differential exposure patterns or biological susceptibility. 3.4.3. Correlations between metal-metalloid levels with BDI scores by sexes As it is showed in Table 6 the strongest and most consistent negative correlations across all BDI domains, particularly among females. Al and Mn also exhibited significant inverse associations with multiple domains in females, while in males, correlations were generally weaker. Only As remained significantly associated with poorer scores across several domains in males. In the overall sample, As, Al, and Mn were negatively correlated with BDI total and most subscales. Ni and Pb also showed weaker negative correlations with the adaptive and motor domains, respectively. Month of life correlated positively with all BDI scores, reflecting expected developmental progress, as can be observed in Supplementary Fig. 3. Mother’s education was weakly but significantly correlated with the language domain. 3.4.4. Relationship between BDI scores, metal exposure, and sociodemographic factors Using a simple linear regression model, it was observed that the age of the participants was related to the global and specific domain scores. The multiple regression model that characterized the interaction factor between the adjusted metals and metalloids index and age, along with maternal education, showed a significant relationship with the global development scale and the personal-social, cognitive, and language domains (Table 7). As expected, month of life showed a consistent and positive association with all BDI domains. Mother’s education also exerted a positive effect, with stronger associations in motor and personal-social domains. In contrast, the adjusted metals and metalloids index demonstrated small but consistent negative associations across all domains, with a particularly marked inverse effect on the total BDI score, as shown in Supplementary Fig. 4. 4. Discussion In the present study, we measured the hair concentration of 10 metals and metalloids (Al, Cr, Mn, Ni, Cu, Zn, As, Se, Cd, and Pb). Also, associations were established between metal-metalloid mixture parameters, lifestyle information, and the infant developmental profile. Concerning our results, 7 of the 10 elements studied (Cu, Zn, Cr, Pb, Mn, Al, Se) were detected in over 90 % of participants, the remaining elements (As, Cd, Ni) being present in around 50–70 % of the participants. Cu was found in 98.4 % of participants (mean 17.38 μ g/g), showing a slight temporal increase. These levels aligned with those in Madrid children aged 0–5 years (16.3 μ g/g) and Korean preschoolers (15.51 μ g/g) (Llorente Ballesteros et al., 2017; Park et al., 2007). They were comparable to Indian ADHD cases (14.01 μ g/g) but exceeded their controls (7.43 μ g/g) (Nayak et al., 2021), as well as Chinese toddlers with recurrent respiratory infections (8.8 μ g/g) and Japanese children aged 3–6 (9.95 μ g/g) (Kusanagi et al., 2023; Mao et al., 2014). A Russian study of Down syndrome versus control children showed similar levels (9.54 vs. 9.62 μ g/g) with no significant difference (Grabeklis et al., 2019). Zn was detected in 96.5 % of participants (mean 121.8 μ g/g), with lower levels at 12, 18, and 24 months (102.8, 103.2, 107.8 μ g/g) than at 6 months (151.1 μ g/g). Similar age-related variation appeared in Russian children with Down syndrome (137.5 μ g/g at one year vs. 99.9 μ g/g at two years) and controls (57.1 vs. 93.5 μ g/g) (Grabeklis et al., 2019). Our values aligned with Turkish 1–3 year-olds (103.3 μ g/g) and Canadian 2–6 year-olds (115 μ g/g) (Razi et al., 2012; Vaghri et al., 2008), but exceed those from Madrid (68.2 μ g/g) and Korean preschoolers (70.0 μ g/g) (Llorente Ballesteros et al., 2017; Park et al., 2007), and slightly surpassed Chinese children prone to respiratory infections (85.0–98.5 μ g/g) (Mao et al., 2014). Cr was detected in 93.7 % of participants (mean 0.90 μ g/g), with similar levels at 6 months (1.14 Table 2 Descriptive statistics of metals and metalloid’s concentration in children’s hair samples (n =254). Values are expressed as μ g/g. Mann Whitney P value p <0.0001***, p <0.001**, p <0.05*. Elements Detection Range Mean ±SD Percentiles Mean ±SD P value n % 25th 75th 95th VVUH n =93 VRUH n =161 Cu 250 98.43 0.44–80.00 17.38 ±11.56 10.68 20.08 40.78 15.55 ±9.51 18.44 ±12.51 0.0303* Zn 245 96.46 1.95–1150.00 121.80 ±134.40 50.00 154.00 297.20 121.10 ±149.40 122.20 ±125.20 0.5428 Cr 238 93.70 0.04–14.20 0.90 ±1.38 0.27 1.00 2.77 0.72 ±0.60 1.02 ±1.69 0.8580 Pb 238 93.70 0.02–11.30 2.10 ±1.98 0.76 2.74 5.96 1.88 ±1.80 2.24 ±2.08 0.1288 Mn 235 92.52 0.03–7.21 0.90 ±1.00 0.37 1.01 2.78 0.81 ±0.84 0.95 ±1.09 0.1932 Al 234 92.13 2.80–670.00 62.71 ±75.63 22.50 74.18 192.80 58.22 ±68.28 65.52 ±79.99 0.1586 Se 231 90.94 0.05–3.60 0.66 ±0.45 0.46 0.70 1.41 0.62 ±0.39 0.68 ±0.48 0.7828 Ni 185 72.83 0.01–12.90 1.19 ±1.70 0.35 1.26 4.03 0.84 ±0.75 1.43 ±2.09 0.0495* As 140 55.12 0.00–0.72 0.10 ±0.09 0.06 0.12 0.24 0.09 ±0.07 0.11 ±0.11 0.7093 Cd 124 48.82 0.00–2.26 0.23 ±0.45 0.03 0.16 1.54 0.24 ±0.44 0.23 ±0.46 0.8720 MM-N 254 –2.00–10.00 8.37 ±1.72 7.00 10.00 10.00 8.85 ±1.29 8.10 ±1.87 0.0021** Abbreviations: MM-N: Number of detected metals/metalloids above the limit of detection (LOD). M. Quintana-Mejía et al. Environmental Pollution 386 (2025) 127261 4 μ g/g), 12 months (0.90 μ g/g) and 24 months (0.91 μ g/g), but significantly lower at 18 months (0.44 μ g/g), comparable to U.S. children (0.41 μ g/g) (Geier et al., 2012). These means were 4.7-fold higher than in Russian Down syndrome toddlers (0.245 μ g/g) and 6-fold higher than their controls (0.156 μ g/g) (Grabeklis et al., 2019). Concerning Pb, this element was detected in 93.7 % of samples (mean 2.10 μ g/g), peaking at 6 months (2.51 μ g/g) and dipping by 24 months (1.63 μ g/g), with 12 and 18 months near 2.00 μ g/g. The 24-month value parallels Korean preschoolers (Park et al., 2007) and exceeded Madrid children (1.10 μ g/g) by double (Llorente Ballesteros et al., 2017). At 12 and 24 months, our levels also surpassed those in Russian Down syndrome and control groups (0.892 vs. 0.688 μ g/g at 12 months; 1.079 vs. 0.460 μ g/g at 24 months) (Grabeklis et al., 2019). A higher mean Pb (2.98 μ g/g) was reported in Egyptian autistic children near gas stations (Mohamed et al., 2015). Concerning Mn, it was detected in 92.5 % of children (mean 0.90 μ g/g), declining from 1.21 μ g/g at 6 months to 0.48 μ g/g at 24 months—a 2.5-fold drop suggesting age-related regulation. Values at 24 months (0.48 μ g/g) approximated those in Korean preschoolers (0.29 μ g/g) and Romanian 3–15 year-olds (0.35 μ g/g) (Park et al., 2007; Senofonte and Caroli, 2000) but exceed Madrid (0.30 μ g/g) and Gipuzkoa (0.31 μ g/g) cohorts by 3 times-fold (Irizar et al., 2019; Fig. 1. Detection of metals and metalloids by sex. p <0.0001***, p <0.001**, p <0.05*, Mann Whitney test. Fig. 2. Detection of metals and metalloids after 6, 12, 18, and 24 months of life. p <0.0001***, p <0.001**, p <0.05*, Kruskal-Wallis test. M. Quintana-Mejía et al. Environmental Pollution 386 (2025) 127261 5 Llorente Ballesteros et al., 2017). Near a Brazilian ferromanganese plant, children 1–4 years presented much higher Mn levels (15.2 μ g/g) (Menezes-Filho et al., 2009). Regarding Al, it appeared in 92.1 % (mean 62.7 μ g/g), dropping from 86.8 μ g/g at 6 months to 26.8 μ g/g at 24 months (significant negative correlation with age). Our mean (~62 μ g/g) matched Egyptian autistic children (63.9 μ g/g) but was higher than Madrid (24.3 μ g/g), Japanese (8.2 μ g/g), Korean (8.8 μ g/g), and Romanian (10.2 μ g/g) data (Kusanagi et al., 2023; Llorente Ballesteros et al., 2017; Mohamed et al., 2015; Park et al., 2007; Senofonte and Caroli, 2000). Furthermore, Se was present in 90.9 % (mean 0.66 μ g/g) with stable levels across ages (0.60–0.70 μ g/g), similar to Romanian children (0.77 μ g/g) and slightly above Madrid (0.40 μ g/g) (Llorente Ballesteros et al., 2017; Senofonte and Caroli, 2000). Ni occurred in 72.8 % (mean 1.19 μ g/g), decreasing from 1.64 μ g/g at 6 months to 0.44 μ g/g at 24 months (significant age trend). Romanian 3–15-year-olds had higher Ni (1.49 μ g/g) (Senofonte and Caroli, 2000), while Madrid children showed lower levels (0.77 μ g/g), and U.S. autistic children much lower (0.17 μ g/g) (Geier et al., 2012; Llorente Ballesteros et al., 2017). Concerning As, it was detected in 55.1 % (mean 0.10 μ g/g), slightly declining from 0.15 μ g/g at 6 months to 0.07 μ g/g at 24 months. These matched Korean (0.11 μ g/g) and Romanian (0.09 μ g/g) figures (Park et al., 2007; Senofonte and Caroli, 2000) and exceed U.S. autistic levels (0.074 μ g/g) (Geier et al., 2012), but were far below Pakistani ranges (1.25–31.9 μ g/g) linked to groundwater contamination and gas extraction (Brahman et al., 2016; Kazi et al., 2011; Shaikh et al., 2019). Lastly, Cd appeared in 48.8 % (mean 0.23 μ g/g), fluctuating from 0.32 μ g/g at 6 months to 0.12 μ g/g at 24 months. Levels at 12–24 months aligned with Turkish wheezing (0.22 μ g/g) and controls (0.12 μ g/g), and Romanian (0.23 μ g/g) and U. S. autistic children (0.18 μ g/g) (Geier et al., 2012; Razi et al., 2012; Senofonte and Caroli, 2000), but exceed Madrid (0.02 μ g/g), Russian Down syndrome (0.026–0.022 μ g/g), and Korean (0.08 μ g/g) levels—and were 250 times higher than Japanese preschoolers (0.0009 μ g/g) (Grabeklis et al., 2019; Kusanagi et al., 2023; Llorente Ballesteros et al., 2017; Park et al., 2007). Elevated Cd is linked to traffic, industry, and contaminated environments (J¨ arup and Akesson, 2009). In general, as can be seen by this comparison, although the values obtained are similar to those in some other studies, especially those performed in Europe (such as Romania). However, it is noteworthy that they are higher than some other populations, especially the Asian ones. In this sense, the reports available from India, China, Japan, Russia and Korea led to lower values in most of the metals, except for Mn. In addition, the study performed in the capital of Spain, Madrid, which is in the center, and only 500 km away, led in general to lower values than the ones obtained in the present study. Concerning the implication of sex, in our study, girls showed significantly higher levels of Al, Cu, and Pb compared to boys, which was consistent with findings from Llorente Ballesteros et al. (2017) in children from Madrid, as well as from Długaszek and Skrzeczanowski (2017) in Polish children, where girls had higher Zn and Cu levels. These results suggest gender-related differences in metal exposure or metabolism. Similarly, Tinkov et al. (2020) reported ageand sex-related differences in metal concentrations in Russian children, with more pronounced differences in younger children. They found variations in essential trace elements such as Cu, Zn, and Mn, highlighting the influence of age and sex on metal levels. These findings further support the notion that gender and developmental stages may significantly affect trace metal concentrations. To further explore potential health implications, we assessed the neurological development of participants using the BDI and analyzed correlations between BDI scores and levels of detected metals and metalloids. Despite the scarcity of studies examining this relationship, our results revealed a significant negative correlation between Pb levels and BDI scores, particularly in the language domain. In boys, Pb was also negatively and significantly associated with cognitive and motor domains. Similar findings were reported in Spanish adolescents from Catalonia, where higher Pb levels in hair were significantly associated with poorer attentional performance (Torrente et al., 2005). Mn levels were negatively and significantly correlated with total BDI scores and all developmental domains: personal-social, cognitive, adaptive, language, and motor. This is consistent with previous literature. In Uruguayan children aged 14–45 months, lower Mn levels were positively associated with language development (Rink et al., 2014). In Canadian children aged 6–15 years, higher Mn levels in hair were linked to increased hyperactivity and oppositional behaviors (Bouchard et al., 2007), as well as poorer cognitive performance, particularly in the verbal domain (Menezes-Filho et al., 2011) and reduced memory and attention scores (Oulhote et al., 2014). Ni showed a negative, although mostly non-significant, correlation with total BDI scores and the cognitive, language, and motor domains; however, significant negative correlations were found for the personal-social and adaptive domains. In adults, Ni exposure has been associated with reduced verbal memory, processing speed, and executive function (Wurth et al., 2018). Cd levels were negative, though not significantly, correlated with BDI scores and developmental domains, with similar trends in both boys and girls. These results differed from findings in Bangladesh, where Cd exposure was associated with reduced intellectual performance, particularly in girls (Kippler et al., 2012). To our knowledge, no previous studies have examined correlations between BDI scores and most of the other elements included in this study (As, Se, Cr, Al, Cu and Zn), although existing literature suggests that these elements may also be associated with neurodevelopmental outcomes. In our population, As levels were negatively and significantly associated with total BDI scores and all developmental domains. This association was consistent across sexes. Prior studies have shown that chronic low-level As exposure is associated with reduced IQ and verbal functioning, especially in girls (Hamadani et al., 2011; Wang et al., 2022). Se exhibited a negative, but non-significant, correlation with BDI Table 3 BDI scores in the whole population and by sexes. Data given in mean ±SD and in % of SD (standard deviation of BDI). Significance between girls and boys studied with the Mann-Whitney test (p <0.05). Domain Total (n =137) Female (n =80) Male (n =57) Mean ±SD SD, % Mean ±SD SD, % Mean ±SD SD, % ±1−1.50 −2±1−1.50 −2±1−1.50 −2 BDI score 52.31 ±21.18 54.74 29.93 15.33 53.96 ±21.52 62.50 21.25 16.25 49.98 ±20 43.86 42.11 14.04 Personal–Social 11.58 ±5.13 52.55 24.82 22.63 11.94 ±5.18 81.25 15.00 3.75 11.07 ±5.06 50.88 19.30 29.82 Adaptative 10.49 ±4.52 79.56 10.95 9.49 10.93 ±4.49 53.75 28.75 17.50 9.88 ±4.53 77.19 5.26 17.54 Motor 10.71 ±5.11 86.86 8.76 4.38 10.90 ±5.06 88.75 8.75 2.50 10.44 ±5.22 84.21 8.77 7.02 Gross motor 5.02 ±2.58 95.62 2.92 1.46 5.13 ±2.68 98.75 0.00 1.25 4.88 ±2.45 91.23 7.02 1.75 Fine motor 5.54 ±2.50 85.40 13.14 1.46 5.75 ±2.43 91.25 7.50 1.25 5.25 ±2.59 77.19 21.05 1.75 Language 9.58 ±3.93 63.50 32.85 3.65 9.95 ±4.01 62.50 36.25 1.25 9.07 ±3.78 64.91 28.07 7.02 Receptive 5.01 ±2.28 64.96 31.39 3.65 5.21 ±2.33 70.00 27.50 2.50 4.72 ±2.19 57.89 36.84 5.26 Expressive 4.58 ±1.76 89.05 6.57 4.38 4.73 ±1.80 91.25 7.50 1.25 4.39 ±1.71 85.96 5.26 8.77 Cognitive 9.79 ±3.41 68.61 17.52 13.87 10.10 ±3.48 72.50 15.00 12.50 9.35 ±3.29 63.16 21.05 15.79 M. Quintana-Mejía et al. Environmental Pollution 386 (2025) 127261 6 Table 4 Correlation analysis between the levels of metals and metalloids and the sociodemographic characteristics of the participants. Correlations Variable Al Cr Mn Ni Cu Zn As Se Cd Pb Month of life r −0.2703 −0.1134 −0.2526 −0.2693 0.09452 −0.1297 −0.3022 −0.06970 −0.1176 −0.1521 P value <0.0001 0.0807 <0.0001 0.0002 0.1361 0.0426 0.0003 0.2915 0.1932 0.0189 95 % CI −0.3852 to −0.1472 −0.2372 to 0.01392 −0.3686 to −0.1288 −0.3981 to −0.1301 −0.02990 to 0.2161 −0.2509 to −0.004420 −0.4457 to −0.1435 −0.1970 to 0.05992 −0.2880 to 0.05993 −0.2740 to −0.02544 Sex r−0.03781 −0.05433 −0.05716 −0.04772 −0.1288 −0.003083 −0.03967 −0.03558 −0.09418 −0.1900 P value 0.5666 0.4061 0.3851 0.5201 0.0427 0.9619 0.6429 0.5922 0.3021 0.0034 95 % CI −0.1658 to 0.09143 −0.1808 to 0.07388 −0.1843 to 0.07189 −0.1911 to 0.09761 −0.2494 to −0.004291 −0.1289 to 0.1228 −0.2048 to 0.1277 −0.1645 to 0.09450 −0.2675 to 0.08500 −0.3102 to −0.06388 Mother’s education r−0.08821 0.01466 −0.01035 −0.05704 0.007509 −0.001563 0.01704 0.04392 −0.1950 −0.05246 P value 0.1787 0.8219 0.8745 0.4406 0.9060 0.9806 0.8417 0.5065 0.0300 0.4204 95 % CI −0.2140 to 0.04050 −0.1127 to 0.1416 −0.1381 to 0.1178 −0.1997 to 0.08795 −0.1167 to 0.1315 −0.1269 to 0.1238 −0.1493 to 0.1824 −0.08564 to 0.1720 −0.3590 to −0.01938 −0.1784 to 0.07520 Breastfeeding r−0.1376 −0.1422 −0.06630 −0.1432 −0.08308 −0.04478 −0.01923 0.04535 −0.09057 −0.07881 P value 0.1260 0.1137 0.4608 0.1594 0.3399 0.6101 0.8708 0.6213 0.4493 0.3785 95 % CI −0.3058 to 0.03897 −0.3101 to 0.03428 −0.2384 to 0.1099 −0.3322 to 0.05679 −0.2492 to 0.08775 −0.2140 to 0.1271 −0.2466 to 0.2102 −0.1342 to 0.2221 −0.3156 to 0.1441 −0.2496 to 0.09673 Breastfeeding time r−0.1750 −0.1837 −0.1434 −0.1105 0.1776 −0.009651 −0.07960 −0.1272 −0.2150 −0.1340 P value 0.0687 0.0536 0.1350 0.3170 0.0554 0.9181 0.5318 0.1960 0.0933 0.1628 95 % CI −0.3515 to 0.01355 −0.3579 to 0.002763 −0.3220 to 0.04503 −0.3174 to 0.1064 −0.004075 to 0.3479 −0.1916 to 0.1730 −0.3192 to 0.1695 −0.3113 to 0.06609 −0.4411 to 0.03672 −0.3134 to 0.05462 Exclusive breastfeeding for up to 6 months r 0.6052 −0.5632 0.3635 −0.6243 −0.07434 0.3722 0.4857 0.3955 0.8356 0.2857 P value 0.0218 0.0149 0.1829 0.0226 0.7694 0.1283 0.1850 0.1042 0.0192 0.2505 95 % CI 0.1099 to 0.8597 −0.8156 to −0.1307 −0.1828 to 0.7383 −0.8745 to −0.1117 −0.5231 to 0.4066 −0.1146 to 0.7148 −0.2634 to 0.8694 −0.08749 to 0.7280 0.2227 to 0.9751 −0.2091 to 0.6640 Only formula r 0.01767 −0.1556 0.1254 −0.06513 −0.09488 0.08328 −0.1233 0.1125 −0.1821 −0.06491 P value 0.8601 0.1221 0.2047 0.5537 0.3287 0.3960 0.3356 0.2702 0.1637 0.5169 95 % CI −0.1774 to 0.2114 −0.3416 to 0.04210 −0.06886 to 0.3105 −0.2744 to 0.1501 −0.2789 to 0.09581 −0.1092 to 0.2698 −0.3601 to 0.1284 −0.08791 to 0.3041 −0.4168 to 0.07529 −0.2562 to 0.1312 Gluten-free cereals r 0.06863 −0.05164 0.1439 −0.05248 0.007891 0.01274 −0.007211 0.1120 −0.09780 −0.01580 P value 0.4803 0.5956 0.1355 0.6293 0.9336 0.8934 0.9545 0.2598 0.4383 0.8717 95 % CI −0.1219 to 0.2543 −0.2383 to 0.1387 −0.04544 to 0.3232 −0.2602 to 0.1599 −0.1763 to 0.1915 −0.1724 to 0.1970 −0.2507 to 0.2371 −0.08328 to 0.2991 −0.3337 to 0.1497 −0.2050 to 0.1746 Gluten-containing cereals r 0.2912 0.04003 0.2710 0.01136 0.06540 0.2148 0.05513 0.1115 0.05028 0.1842 P value 0.0019 0.6766 0.0038 0.9163 0.4855 0.0212 0.6552 0.2574 0.6885 0.0552 95 % CI 0.1108 to 0.4530 −0.1475 to 0.2247 0.08998 to 0.4347 −0.1986 to 0.2203 −0.1183 to 0.2448 0.03297 to 0.3828 −0.1857 to 0.2897 −0.08191 to 0.2968 −0.1941 to 0.2888 −0.004032 to 0.3598 Ecological fruit r 0.08618 −0.04504 0.07714 −0.03339 −0.04587 −0.05579 0.1719 0.07320 –−0.06199 P value 0.3452 0.6223 0.3964 0.7494 0.6057 0.5317 0.1487 0.4349 –0.4976 95 % CI −0.09300 to 0.2600 −0.2210 to 0.1338 −0.1013 to 0.2507 −0.2344 to 0.1704 −0.2170 to 0.1280 −0.2271 to 0.1189 −0.06223 to 0.3881 −0.1106 to 0.2522 –−0.2371 to 0.1171 Ecological vegetables r 0.01461 −0.06102 −0.009272 −0.06597 −0.04445 −0.05177 −0.001527 0.03219 −0.07649 −0.05770 (continued on next page) M. Quintana-Mejía et al. Environmental Pollution 386 (2025) 127261 7 Table 4 (continued) Correlations Variable Al Cr Mn Ni Cu Zn As Se Cd Pb P value 0.8731 0.5061 0.9189 0.5298 0.6184 0.5648 0.9898 0.7327 0.5291 0.5296 95 % CI −0.1636 to 0.1919 −0.2369 to 0.1188 −0.1860 to 0.1680 −0.2661 to 0.1396 −0.2163 to 0.1301 −0.2246 to 0.1243 −0.2315 to 0.2286 −0.1518 to 0.2140 −0.3060 to 0.1614 −0.2338 to 0.1221 Chicken r 0.3553 0.05246 0.2506 0.03822 0.1091 0.1752 −0.03064 0.2806 −0.09420 0.07825 P value 0.0001 0.5794 0.0069 0.7221 0.2357 0.0566 0.8012 0.0034 0.4519 0.4101 95 % CI 0.1833 to 0.5061 −0.1327 to 0.2341 0.07068 to 0.4146 −0.1714 to 0.2445 −0.07158 to 0.2828 −0.004902 to 0.3444 −0.2637 to 0.2058 0.09583 to 0.4466 −0.3287 to 0.1513 −0.1080 to 0.2592 Beef r 0.3218 0.04751 0.2617 0.04640 0.03612 0.2660 0.05002 0.07959 0.1431 0.03242 P value 0.0005 0.6189 0.0049 0.6696 0.6978 0.0038 0.6809 0.4196 0.2555 0.7355 95 % CI 0.1458 to 0.4781 −0.1393 to 0.2310 0.08174 to 0.4252 −0.1659 to 0.2546 −0.1456 to 0.2155 0.08871 to 0.4269 −0.1872 to 0.2817 −0.1138 to 0.2672 −0.1044 to 0.3739 −0.1549 to 0.2175 Vegetable oil r 0.3041 0.03896 0.1655 0.04268 0.006561 0.1341 −0.02624 0.01618 −0.08772 0.03147 P value 0.0011 0.6834 0.0784 0.6912 0.9435 0.1478 0.8293 0.8692 0.4872 0.7418 95 % CI 0.1265 to 0.4628 −0.1477 to 0.2229 −0.01896 to 0.3391 −0.1671 to 0.2487 −0.1736 to 0.1863 −0.04784 to 0.3074 −0.2596 to 0.2100 −0.1751 to 0.2063 −0.3247 to 0.1596 −0.1550 to 0.2158 Canned sea-food servings r 0.02515 −0.06335 0.01756 −0.03793 −0.04378 0.1171 −0.03298 0.01359 0.1877 −0.02165 P value 0.7860 0.4900 0.8490 0.7122 0.6237 0.1917 0.7803 0.8854 0.1197 0.8129 95 % CI −0.1555 to 0.2042 −0.2391 to 0.1165 −0.1622 to 0.1962 −0.2356 to 0.1627 −0.2157 to 0.1307 −0.05903 to 0.2861 −0.2595 to 0.1970 −0.1699 to 0.1962 −0.04943 to 0.4048 −0.1986 to 0.1567 Canned non-fish servings r 0.02686 0.03675 0.09885 0.01148 −0.06949 0.09101 0.1362 0.03900 0.04356 −0.07799 P value 0.7709 0.6890 0.2807 0.9112 0.4339 0.3088 0.2473 0.6763 0.7183 0.3932 95 % CI −0.1531 to 0.2051 −0.1427 to 0.2138 −0.08108 to 0.2725 −0.1884 to 0.2104 −0.2395 to 0.1046 −0.08454 to 0.2611 −0.09528 to 0.3537 −0.1435 to 0.2190 −0.1917 to 0.2741 −0.2523 to 0.1012 Abbreviations: r, represents the effect size; CI, represents the confidence interval. M. Quintana-Mejía et al. Environmental Pollution 386 (2025) 127261 8 scores and domains in our study. Previous research has suggested that prenatal Se exposure is linked to better cognitive and psychomotor outcomes, particularly in girls (Skr¨ oder et al., 2015, 2017). However, high Se levels have also been associated with lower verbal IQ in boys and higher verbal IQ in girls (Wang et al., 2022). Cr levels were negatively, though not significantly, correlated with BDI scores and all domains. Other studies have reported that intrauterine Cr exposure, particularly in genetically susceptible children, may increase the risk of cognitive developmental delays by 24 months (Jia et al., 2024). Al levels in our population were negatively and significantly correlated with total BDI scores and all developmental domains. This aligns with findings by Corkins (2019), who reported that prenatal Al exposure can impair brain development, potentially reducing developmental scores. Cu showed a positive, though non-significant, correlation with BDI scores and subscales, while Zn showed a negative, also non-significant, correlation. Finally, since more than one element was detected in all analyzed samples, we explored the potential influence of metal–metalloid mixtures on developmental outcomes by evaluating the relationship Table 5 Correlation analysis between metal and metalloid levels, months old, mother’s education and BDI. Variable Scales BDI score Personal–Social Cognitive Adaptative Language Motor Cu r 0.1004 0.1092 0.1039 0.1108 0.07046 0.07231 P value 0.2485 0.2091 0.2323 0.2023 0.4185 0.4063 95 % CI −0.07041 to 0.2654 −0.06153 to 0.2737 −0.06690 to 0.2687 −0.05987 to 0.2753 −0.1003 to 0.2372 −0.09848 to 0.2390 Zn r −0.09036 −0.09268 −0.03592 −0.07552 −0.1139 −0.09265 P value 0.3010 0.2887 0.6814 0.3876 0.1918 0.2888 95 % CI −0.2566 to 0.08112 −0.2588 to 0.07879 −0.2049 to 0.1351 −0.2426 to 0.09594 −0.2787 to 0.05744 −0.2588 to 0.07882 Pb r −0.1433 −0.1084 −0.1570 −0.08441 −0.1743 −0.1694 P value 0.1066 0.2234 0.0767 0.3435 0.0491 0.0560 95 % CI −0.3092 to 0.03099 −0.2767 to 0.06643 −0.3218 to 0.01696 −0.2542 to 0.09044 −0.3376 to −0.0007578 −0.3331 to 0.004275 Cr r −0.06229 −0.1088 −0.0002349 −0.07715 −0.04314 −0.1048 P value 0.4901 0.2270 0.9979 0.3924 0.6328 0.2447 95 % CI −0.2353 to 0.1146 −0.2791 to 0.06808 −0.1758 to 0.1754 −0.2494 to 0.09981 −0.2171 to 0.1335 −0.2753 to 0.07213 Al r −0.2437 −0.2438 −0.2219 −0.2333 −0.2156 −0.2421 P value 0.0060 0.0059 0.0125 0.0086 0.0153 0.0063 95 % CI −0.4015 to −0.07186 −0.4016 to −0.07195 −0.3820 to −0.04893 −0.3922 to −0.06084 −0.3763 to −0.04226 −0.4001 to −0.07020 Mn r −0.2182 −0.2339 −0.1745 −0.1938 −0.2206 −0.2040 P value 0.0137 0.0081 0.0497 0.0290 0.0127 0.0214 95 % CI −0.3781 to −0.04575 −0.3922 to −0.06226 −0.3385 to −0.0003402 −0.3560 to −0.02025 −0.3802 to −0.04825 −0.3652 to −0.03085 Se r −0.05051 −0.05237 −0.02424 −0.05328 −0.05315 −0.03603 P value 0.5806 0.5667 0.7910 0.5600 0.5610 0.6936 95 % CI −0.2262 to 0.1284 −0.2280 to 0.1266 −0.2011 to 0.1542 −0.2289 to 0.1257 −0.2287 to 0.1258 −0.2124 to 0.1426 Ni r −0.1937 −0.2057 −0.1446 −0.2225 −0.1732 −0.1963 P value 0.0559 0.0422 0.1553 0.0277 0.0880 0.0527 95 % CI −0.3777 to 0.004858 −0.3882 to −0.007544 −0.3335 to 0.05538 −0.4031 to −0.02514 −0.3593 to 0.02609 −0.3800 to 0.002168 As r −0.4160 −0.3838 −0.4484 −0.3791 −0.4158 −0.3954 P value 0.0002 0.0007 <0.0001 0.0008 0.0002 0.0004 95 % CI −0.5875 to −0.2088 −0.5618 to −0.1718 −0.6130 to −0.2465 −0.5581 to −0.1665 −0.5873 to −0.2085 −0.5711 to −0.1850 Cd r −0.08631 −0.06329 −0.09815 −0.03878 −0.1321 −0.09191 P value 0.4710 0.5974 0.4121 0.7464 0.2688 0.4426 95 % CI −0.3117 to 0.1483 −0.2907 to 0.1709 −0.3225 to 0.1366 −0.2680 to 0.1946 −0.3529 to 0.1027 −0.3168 to 0.1428 MM-N r 0.02637 0.03839 0.04944 0.03600 0.04804 0.005566 P value 0.7597 0.6560 0.5662 0.6762 0.5772 0.9485 95 % CI −0.1420 to 0.1932 −0.1302 to 0.2048 −0.1193 to 0.2154 −0.1325 to 0.2025 −0.1206 to 0.2140 −0.1623 to 0.1731 Month of life r 0.9538 0.9393 0.8652 0.9232 0.9090 0.9066 P value <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 95 % CI 0.9357 to 0.9668 0.9158 to 0.9563 0.8159 to 0.9020 0.8938 to 0.9446 0.8747 to 0.9343 0.8714 to 0.9325 Sex r −0.09298 −0.08359 −0.1086 −0.1146 −0.1109 −0.04468 P value 0.2798 0.3315 0.2064 0.1822 0.1970 0.6042 95 % CI −0.2567 to 0.07592 −0.2478 to 0.08532 −0.2714 to 0.06018 −0.2770 to 0.05411 −0.2735 to 0.05790 −0.2108 to 0.1240 Mother’s education r−0.1365 −0.1167 −0.1289 −0.1167 −0.1788 −0.1132 P value 0.1118 0.1746 0.1333 0.1744 0.0366 0.1877 95 % CI −0.2974 to 0.03197 −0.2789 to 0.05208 −0.2904 to 0.03966 −0.2790 to 0.05204 −0.3364 to −0.01139 −0.2757 to 0.05555 Breastfeeding time r−0.02361 −0.02473 −0.004059 −0.02464 0.005914 −0.03398 P value 0.8005 0.7913 0.9654 0.7920 0.9495 0.7161 95 % CI −0.2043 to 0.1586 −0.2053 to 0.1575 −0.1855 to 0.1776 −0.2053 to 0.1576 −0.1758 to 0.1872 −0.2142 to 0.1485 Abbreviations: r, represents the effect size; CI, represents the confidence interval; MM-N: number of metals and metalloids in hair. M. Quintana-Mejía et al. Environmental Pollution 386 (2025) 127261 9