Abstract
The neonatal period is highly susceptible to metabolic impairments that may persist throughout lifespan and predispose to increased obesity risk and related complications. During these first stages of life, trace elements and heavy metals play a central role in regulating health status and participating in obesity pathophysiology. Thus, we hypothesize that neonatal metrics might serve as reliable predictors of obesity-related metal alterations occurring later in childhood. This study relies on a population comprising children with obesity (N = 79, age range: 6–14 years), from whom birth metrics (i.e., gestational age, length, and weight at birth) were registered from medical records and blood samples were collected to evaluate classical metabolic markers (insulin and glucose metabolism, inflammatory status) and metal biodistribution by inductively-coupled plasma mass spectrometry. Interestingly, higher gestational age and length at birth were associated with lower inflammation, insulinemia, and glycemia in childhood, as well as with lower levels of toxic heavy metals (i.e., arsenic, cadmium, lead) and greater levels of essential trace elements (i.e., zinc, selenium). Conversely, higher body mass index at birth predicted exacerbated failures in glucose homeostasis and an unfavorable multi-elemental profile, as reflected in negative associations with minerals involved in endocrine control (i.e., zinc, chromium, molybdenum, selenium). In summary, this study supports that even slight deviations in neonatal parameters might influence metabolic health later in life, considering both classical clinical markers (e.g., insulin resistance, inflammation) and complementary metallomics assessments.
Supplementary Information
The online version contains supplementary material available at 10.1038/s41598-026-42460-9.
Keywords: Childhood obesity, Heavy metals, Neonatal metrics, Trace elements
Subject terms: Biomarkers, Diseases, Endocrinology, Medical research, Risk factors
Introduction
Obesity is a chronic and multi-faceted disorder affecting more than 330 million children and adolescents worldwide, according to recent estimates1. Among the numerous risk factors associated with its early onset, adverse neonatal outcomes have consistently been reported to bear a substantial impact on cardiometabolic health later in life. In particular, accumulating evidence suggests that both prematurity (i.e., birth before 37 weeks of gestation) and extreme weight at birth (i.e., small for gestational age, birth weight < 10th percentile; large for gestational age, birth weight > 90th percentile) increase the odds of developing childhood obesity and related complications, such as insulin resistance, glucose intolerance, dyslipidemia, and cardiovascular events2,3. Although genetic determinants (e.g., chromosomal disorders, ethnicity) have been proven to occasionally participate in the etiology of these neonatal conditions, it is well recognized that parental (e.g., gestational diabetes mellitus, family history of obesity) and environmental factors (e.g., maternal malnutrition, smoking, alcohol consumption, exposure to endocrine disruptors) are the most important players in shaping fetal programming mechanisms4. Thus, stressful perinatal milieus could be responsible for dysfunctions in adipose tissue, pancreas, liver, and microbiota leading to inflammation, insulin impairments, and oxidative stress in the fetus and the newborn, complications that may persist throughout the lifespan and predispose the offspring to obesity and its comorbidities during childhood2,3. In this respect, it should however be noted that existing literature has normally focused on studying the impact of unfavorable neonatal conditions (e.g., prematurity, being small for gestational age), but the relationship between birth metrics and metabolic risk in general populations (i.e., newborns showing birth metrics within-normal ranges) has scarcely been explored.
Trace elements and heavy metals have been described to be tightly involved in obesity-related molecular disturbances5. A recent study has demonstrated that children with obesity suffer from profound failures in metal homeostasis, alterations that were found to be associated with typical pathogenic events accompanying this metabolic disorder (i.e., oxidative stress, inflammation, impaired glucose metabolism, dyslipidemia)6. Furthermore, this characteristic multi-elemental profile has interestingly been reported to be modulated by other factors that may influence the susceptibility to develop obesity, such as dietary habits7, sex dimorphism8, puberty9, concomitance of insulin resistance10, and family history of obesity11. Thus, on the basis of this background, we hypothesize that birth parameters might be another inter-individual variability factor impacting metal homeostasis, thereby serving as reliable predictors of obesity-related alterations occurring later in childhood. In this respect, it is known that adequate intake of essential micronutrients during pregnancy is crucial for the offspring health12, while exposure to toxic species has deleterious consequences13. However, no prospective data are available on the association between neonatal metrics and metal levels in children with obesity. Accordingly, in this study, we have investigated the potential involvement of neonatal determinants (i.e., gestational age, length, and weight at birth) in shaping future obesity-related molecular disturbances.
Methods
Study population
The study participants were children with obesity presenting a body mass index (BMI) over two standard deviations above the age/sex-adjusted mean of the Spanish reference population (N = 79, age range: 6–14 years), who underwent an oral glucose tolerance test (OGTT) in the morning after overnight fasting to assess their carbohydrate metabolism. At inclusion, anthropometric characteristics (i.e., weight, height) were evaluated by pediatric endocrinologists, and neonatal variables (i.e., gestational age at birth, length at birth, weight at birth) were registered from medical records. Complementarily, a secondary healthy lean group (BMI Z-score below 1) was enrolled among children who needed a blood test for routine health monitoring (N = 25). The study was performed in accordance with the principles contained in the Declaration of Helsinki. The Ethical Committee of Hospital Universitario Puerta del Mar (Cádiz, Spain) approved the study protocol (Ref. PI22/01899), and all participants and/or legal guardians provided written informed consent.
Sample collection and biochemical characterization
Venous blood samples were collected at standardized time points along the OGTT (i.e., 0, 30, 60, 90, 120 min) using BD Vacutainer tubes. After centrifuging at 1500 g for 10 min at 4 °C to separate the plasma, the cell pellet was subjected to three washing cycles with cold saline solution (9 g/L NaCl, 4 °C) and subsequent centrifugation (1500 g, 5 min, 4 °C) to recover RBCs. Then, samples were aliquoted and stored at − 80 °C until multi-elemental analysis.
Glucose and insulin plasma levels along the OGTT curve, as well as fasting contents of glycated hemoglobin (HbA1c), lipids (i.e., total cholesterol, TC; low-density lipoprotein cholesterol, LDL-C; high-density lipoprotein cholesterol, HDL-C; triglycerides, TG), and C-reactive protein (CRP) were measured using an Alinity automatic analyzer (Abbot, Madrid, Spain). The homeostasis model assessment of insulin resistance (HOMA-IR), the whole-body insulin sensitivity index (WBISI), the area under the curve for glucose (AUCGlc), and the area under the curve for insulin (AUCIns) were calculated by applying the formulas (1–4):
 |
1 |
 |
2 |
 |
3 |
 |
4 |
where Glc0 and Ins0 refer to fasting plasma levels of glucose and insulin; Glc30, Glc60, Glc90, and Glc120 to plasma glucose levels measured at 30, 60, 90, and 120 min along the OGTT; Ins30, Ins60, Ins90, and Ins120 to plasma insulin levels measured at 30, 60, 90, and 120 min along the OGTT; Glcmean and Insmean to mean glucose and insulin levels measured in plasma along the OGTT. Glucose and insulin concentrations are expressed as mg/dL and µU/mL, respectively.
An automated hematology analyzer was employed to determine white blood cell counts with the aim of computing different inflammatory indices, namely the systemic immune inflammation index (SII), the systemic inflammation response index (SIRI), the aggregate index of systemic inflammation (AISI), the platelet-to-lymphocyte ratio (PLR), and the neutrophil-to-lymphocyte ratio (NLR), using the formulas (5–9):
 |
5 |
 |
6 |
 |
7 |
 |
8 |
 |
9 |
where, N, P, L, and M refer to neutrophil counts, platelet counts, lymphocyte counts, and monocyte counts, respectively.
Multi-elemental analysis
To determine total metal contents, plasma (150 µL) and RBCs (50 µL) were diluted to a final volume of 3 mL using an alkaline solution composed of 2% 1-butanol (w/v), 0.05% EDTA (w/v), 0.05% Triton X-100 (w/v), and 1% NH4OH (w/v). Moreover, a second aliquot was subjected to protein precipitation under non-denaturing conditions with 300 µL of cold acetone, which allows separating high-molecular mass (HMM) and low-molecular mass (LMM) metal species, as detailed in our previous publication14. After protein precipitation and centrifugation at 10,000 g for 10 min at 4 °C, the supernatants were taken to dryness using a SpeedVac system. Then, dried supernatants (i.e., LMM – labile species) and protein pellets (i.e., HMM – metalloproteins) were reconstituted in 3 mL of the aforementioned alkaline solution. Afterward, multi-elemental analysis was performed in an inductively-coupled plasma mass spectrometer, using the operational conditions described elsewhere14. For quantification purposes, multi-elemental stock solutions containing 100 mg L−1 chromium, manganese, iron, cobalt, copper, zinc, arsenic, selenium, molybdenum, cadmium, and lead in 5% nitric acid (v/v) were employed for preparing calibration curves within the concentration range 0.5–2500 µg/L, containing 1 µg/L rhodium as the internal standard. The analytical performance of the method, in terms of limits of detection, intra-, and inter-assay precisions, is summarized in Table S1. The assessment of data quality was accomplished by implementing the QComics guidelines15.
Statistical analysis
Descriptive statistics of demographic, anthropometric, biochemical, neonatal, and multi-elemental data were expressed as the mean ± standard deviation. Prior to any statistical modeling, variables containing more than 20% missing values were removed, and the remaining data were imputed using the kNN algorithm. Then, logarithmic transformation and Pareto scaling was applied to increase the symmetry of skewed distributions and to adjust for differences in concentration scales, respectively. After this data pre-processing, FDR-adjusted linear regression was applied to analyze the association between neonatal metrics, anthropometric/biochemical variables, and metal levels. Additional models with covariate adjustment were computed to control for the influence of age, sex, and BMI as potential confounding factors. All the statistical analyses were conducted using the MetaboAnalyst 5.0 web tool (https://www.metaboanalyst.ca/). Using a sample size of 79 children, and considering an alpha risk of 0.05, the statistical power of our comparisons was above 80%, as calculated using the GRANMO 7.12 webtool.
Results
Demographic, anthropometric, biochemical, and neonatal characteristics of the study population are listed in Table 1. The participants (N = 79) were on average 11.1 years-old and 57% were male. As defined by inclusion criteria, children had a BMI above age- and sex-adjusted reference values (i.e., BMI Z-score > 2). The biochemical parameters were within the ranges reported in other Spanish cohorts of childhood obesity, indicative of hyperinsulinemia (e.g., HOMA-IR > 3.5), dyslipidemia (e.g., HDL-C < 45 mg/dL), and pro-inflammatory status (e.g., CRP > 3.0 mg/L)16,17. Regarding neonatal parameters, the study subjects presented a normal distribution in terms of gestational age (average 39.4 weeks) and anthropometry (length/weight Z-scores − 2 to 2) at birth.
Table 1.
Demographic, anthropometric, biochemical, and neonatal characteristics of the study population.
| Demographic variables (at inclusion) | |
|---|---|
| N | 79 |
| Sex (% male) | 57.0 |
| Age (years) | 11.1 ± 2.2 |
| Anthropometric variables (at inclusion) | |
| Weight (kg) | 69.5 ± 19.0 |
| Weight Z-score | 5.0 ± 2.2 |
| Body mass index (BMI, kg/m2) | 30.3 ± 5.3 |
| Body mass index Z-score (zBMI) | 4.4 ± 2.1 |
| Biochemical variables: Carbohydrate metabolism (at inclusion) | |
| Fasting glucose (Glc0, mg/dL) | 86.3 ± 9.0 |
| Glucose, t = 30 min (Glc30, mg/dL) | 136.8 ± 26.1 |
| Glucose, t = 60 min (Glc60, mg/dL) | 132.1 ± 29.6 |
| Glucose, t = 90 min (Glc90, mg/dL) | 122.9 ± 27.8 |
| Glucose, t = 120 min (Glc120, mg/dL) | 122.0 ± 26.5 |
| Mean glucose (Glcmean, mg/dL) | 116.6 ± 19.4 |
| Area under the curve for glucose (AUCGlc, mg·h/dL) | 207.6 ± 59.6 |
| Fasting insulin (Ins0, µU/mL) | 22.3 ± 18.5 |
| Insulin, t = 30 min (Ins30, µU/mL) | 145.1 ± 102.5 |
| Insulin, t = 60 min (Ins60, µU/mL) | 155.9 ± 85.3 |
| Insulin, t = 90 min (Ins90, µU/mL) | 158.5 ± 120.8 |
| Insulin, t = 120 min (Ins120, µU/mL) | 142.7 ± 92.6 |
| Mean insulin (Insmean, µU/mL) | 122.6 ± 65.6 |
| Area under the curve for insulin (AUCIns, µU·h/mL) | 256.0 ± 142.4 |
| Homeostasis model assessment of insulin resistance (HOMA-IR) | 4.4 ± 2.6 |
| Whole-body insulin sensitivity index (WBISI) | 0.3 ± 0.2 |
| Glycated hemoglobin (HbA1c, %) | 5.3 ± 0.3 |
| Hemoglobin (g/dL) | 13.5 ± 1.0 |
| Biochemical variables: Lipid profile (at inclusion) | |
| Total cholesterol (TC, mg/dL) | 153.5 ± 24.9 |
| Low-density lipoprotein cholesterol (LDL-C, mg/dL) | 91.8 ± 22.4 |
| High-density lipoprotein cholesterol (HDL-C, mg/dL) | 43.8 ± 14.8 |
| Triglycerides (TG, mg/dL) | 95.5 ± 41.9 |
| Biochemical variables: Inflammatory markers (at inclusion) | |
| C-reactive protein (CRP, mg/L) | 5.2 ± 7.2 |
| Systemic immune inflammation index (SII) | 529.6 ± 388.5 |
| Systemic inflammation response index (SIRI) | 1.1 ± 1.3 |
| Aggregate index of systemic inflammation (AISI) | 342.5 ± 331.7 |
| Platelet-to-lymphocyte ratio (PLR) | 127.8 ± 43.9 |
| Neutrophil-to-lymphocyte ratio (NLR) | 1.8 ± 1.6 |
| Neonatal variables | |
| Gestational age at birth (GA, weeks) | 39.4 ± 2.0 |
| Birth length (BL, cm) | 50.3 ± 2.7 |
| Birth length Z-score (zBL) | 0.7 ± 1.4 |
| Birth weight (BW, kg) | 3.2 ± 0.5 |
| Birth weight Z-score (zBW) | 0.1 ± 1.2 |
| Birth body mass index (BBMI, kg/m2) | 12.6 ± 1.4 |
Results are expressed as mean ± standard deviation (except for sex, expressed as percentage).
Regression analysis was applied to analyze the association between birth metrics and anthropometric/biochemical variables measured later during childhood (Fig. 1). Gestational age at birth was found to negatively associate with inflammatory (AISI, SII, SIRI, NLR) and insulin-related (Ins30, Ins120, AUCIns, HbA1c) markers. In contrast, a positive relationship was observed with cholesterol levels (TC, LDL-C). Birth length showed negative associations with the BMI and variables related to inflammation (AISI, SII, SIRI, NLR, CRP), insulin (Ins120, Insmean, AUCIns), and glucose (Glc30, Glc60, Glc90, Glcmean, AUCGlc), as well as a positive association with TC. Similarly, birth length Z-scores were robustly associated with inflammation (CRP) and glucose metabolism (Glc30, Glc60, Glc90, Glcmean, AUCGlc) in a negative fashion. A positive association was found between the BMI at birth and postprandial glucose (Glc60, Glc90), but no significant relationships were detected for other weight-related metrics (i.e., birth weight and Z-scores).
Fig. 1.
Network representation of the association between birth metrics and anthropometric/biochemical variables measured later in childhood. Positive and negative associations are represented as red and blue lines, respectively (the thicker the line, the stronger the association). Abbreviations: AISI, aggregate index of systemic inflammation; AUCGlc, area under the curve for glucose; AUCIns, area under the curve for insulin; BBMI, birth body mass index; BL, birth length; BMI, body mass index; CRP, C-reactive protein; GA, gestational age at birth; Glc30, glucose levels at 30 min along the OGTT; Glc60, glucose levels at 60 min along the OGTT; Glc90, glucose levels at 90 min along the OGTT; Glcmean, mean glucose levels along the OGTT; HbA1c, glycated hemoglobin; Ins30, insulin levels at 30 min along the OGTT; Ins120, insulin levels at 120 min along the OGTT; Insmean, mean insulin levels along the OGTT; LDL-C, low-density lipoprotein cholesterol; NLR, neutrophil-to-lymphocyte ratio; SII, systemic immune inflammation index; SIRI, systemic inflammation response index; TC, total cholesterol; zBL, birth length Z-score.
The application of a state-of-the-art analytical approach enabled us quantifying the major fractions comprising the plasma and erythroid metallome, including total metal contents, HMM species (i.e., metalloproteins), and LMM species (i.e., labile ions), as summarized in Table 2. Total metal levels were detected within the normal concentration ranges for human plasma and RBC samples collected from occupationally non-exposed individuals18. Complementarily, speciation analysis evidenced that most trace elements and heavy metals were mainly present in the form of metalloproteins, whereas labile ions normally accounted for a minor fraction of the total content (< 10%), as reported elsewhere6. Interestingly, strong associations were observed between neonatal metrics and metal levels (Table 3). The statistical significance of most of these associations was maintained after adjusting for age, sex, and BMI as confounding factors. On the one hand, gestational age at birth was negatively associated with plasma levels of various toxic heavy metals (As-Total, Cd-Total, Pb-Total, Pb-HMM). Although no significant associations were detected with birth length, Z-scores were found to positively associate with plasma metalloproteins of zinc and selenium. Birth weight showed positive associations with erythroid arsenic (Total) and iron (LMM). In addition, birth weight Z-scores positively associated with total and HMM contents of heavy metals in plasma (Cd-Total, Cd-HMM) and RBCs (As-Total, Pb-Total), as well as with labile iron in both biological matrices. Finally, negative associations were observed between the BMI at birth and levels of various essential trace elements in plasma (Zn-HMM, Cr-LMM, Mo-HMM) and RBCs (Se-LMM).
Table 2.
Concentrations of metals in plasma and red blood cells (expressed as mean ± standard deviation, in µg/L).
| Metal | Fraction | Plasma | Red blood cells |
|---|---|---|---|
| Chromium | Total | 7.1 ± 4.8 | ND |
| HMM | 5.9 ± 4.0 | ND | |
| LMM | 1.5 ± 2.8 | ND | |
| Manganese | Total | 4.5 ± 1.5 | 20.9 ± 14.4 |
| HMM | 4.2 ± 0.9 | 20.6 ± 11.5 | |
| LMM | 0.4 ± 0.5 | 0.7 ± 0.2 | |
| Iron | Total | 804.8 ± 364.8 | 766456.1 ± 86301.6 |
| HMM | 773.6 ± 416.7 | 807986.6 ± 188591.6 | |
| LMM | 29.5 ± 15.9 | 24.5 ± 27.8 | |
| Cobalt | Total | 1.6 ± 0.7 | ND |
| HMM | 0.8 ± 0.3 | ND | |
| LMM | 0.4 ± 0.1 | ND | |
| Copper | Total | 1376.2 ± 240.4 | 594.8 ± 89.3 |
| HMM | 1258.9 ± 182.4 | 573.0 ± 57.2 | |
| LMM | 24.4 ± 28.8 | 1.7 ± 0.7 | |
| Zinc | Total | 768.2 ± 328.7 | 9534.5 ± 1746.6 |
| HMM | 724.9 ± 423.5 | 10589.4 ± 5274.0 | |
| LMM | 36.7 ± 22.2 | 61.3 ± 7.8 | |
| Arsenic | Total | 0.9 ± 1.2 | 50.1 ± 20.1 |
| HMM | 0.1 ± 0.1 | 23.4 ± 26.9 | |
| LMM | 0.5 ± 0.7 | 22.6 ± 10.6 | |
| Selenium | Total | 123.5 ± 20.1 | 161.1 ± 43.3 |
| HMM | 137.0 ± 18.1 | 143.2 ± 141.6 | |
| LMM | 2.0 ± 2.7 | 0.2 ± 0.1 | |
| Molybdenum | Total | 2.6 ± 0.8 | 22.6 ± 7.1 |
| HMM | 2.2 ± 2.2 | 21.7 ± 2.4 | |
| LMM | ND | 0.2 ± 0.1 | |
| Cadmium | Total | 0.0020 ± 0.0064 | 1.4 ± 1.4 |
| HMM | 0.0021 ± 0.0065 | 1.4 ± 1.4 | |
| LMM | ND | ND | |
| Lead | Total | 0.023 ± 0.0049 | 62.8 ± 9.2 |
| HMM | 0.023 ± 0.0049 | 61.6 ± 11.5 | |
| LMM | ND | 1.3 ± 0.2 |
LMM, low molecular mass; HMM, high molecular mass; ND, non-detected.
Table 3.
Association between neonatal variables and metal levels in plasma and red blood cells.
| Gestational age at birth | Birth length Z-score | Birth weight | Birth weight Z-score | Birth body mass index | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| β (p-value) | β (p-value), adjusted model | β (p-value) | β (p-value), adjusted model | β (p-value) | β (p-value), adjusted model | β (p-value) | β (p-value), adjusted model | β (p-value) | β (p-value), adjusted model | |
| As (Plasma – HMM) | − 2.0 (4.7 × 10−2) | − 1.7 (9.9 × 10−2) | NS | NS | NS | NS | NS | NS | NS | NS |
| As (RBC – Total) | NS | NS | NS | NS | 3.0 (4.1 × 10−3) | 2.8 (6.4 × 10−3) | 3.3 (1.6 × 10−3) | 3.1 (2.8 × 10−3) | NS | NS |
| Cd (Plasma – Total) | − 1.9 (5.5 × 10−2) | − 1.9 (5.7 × 10−2) | NS | NS | NS | NS | 2.0 (4.5 × 10−2) | 1.9 (5.8 × 10−2) | NS | NS |
| Cd (Plasma – HMM) | NS | NS | NS | NS | NS | NS | 2.2 (3.3 × 10−2) | 2.1 (4.2 × 10−2) | NS | NS |
| Pb (Plasma – Total) | − 2.8 (6.6 × 10−3) | − 2.3 (2.2 × 10−2) | NS | NS | NS | NS | NS | NS | NS | NS |
| Pb (Plasma – HMM) | − 2.6 (1.1 × 10−3) | − 2.1 (4.2 × 10−2) | NS | NS | NS | NS | NS | NS | NS | NS |
| Pb (RBC – Total) | NS | NS | NS | NS | NS | NS | 2.2 (3.2 × 10−2) | 2.2 (3.4 × 10−2) | NS | NS |
| Zn (Plasma – HMM) | NS | NS | 2.2 (3.0 × 10−2) | 2.0 (4.9 × 10−2) | NS | NS | NS | NS | − 2.2 (3.1 × 10−2) | − 2.2 (3.1 × 10−2) |
| Se (Plasma – HMM) | NS | NS | 2.0 (4.9 × 10−2) | 2.0 (5.2 × 10−2) | NS | NS | NS | NS | NS | NS |
| Se (RBC – LMM) | NS | NS | NS | NS | NS | NS | NS | NS | − 1.9 (5.9 × 10−2) | − 1.9 (5.7 × 10−2) |
| Fe (Plasma – LMM) | NS | NS | NS | NS | NS | NS | 2.2 (2.8 × 10−2) | 2.0 (4.9 × 10−2) | NS | NS |
| Fe (RBC – LMM) | NS | NS | NS | NS | 2.1 (3.9 × 10−2) | 2.2 (3.3 × 10−2) | 2.1 (3.9 × 10−2) | 2.1 (4.1 × 10−2) | NS | NS |
| Cr (Plasma – LMM) | NS | NS | NS | NS | NS | NS | NS | NS | − 2.2 (2.8 × 10−2) | − 2.3 (2.6 × 10−2) |
| Mo (Plasma – HMM) | NS | NS | NS | NS | NS | NS | NS | NS | − 2.0 (4.4 × 10−2) | − 2.6 (1.0 × 10−2) |
LMM, low molecular mass; HMM, high molecular mass; NS, non-significant; RBC, red blood cell.
Although the aim of this study was to elucidate the influence of birth determinants in shaping obesity-related molecular disturbances during childhood, we also performed secondary analyses to check whether these associations between neonatal, clinical, and metallomics variables can be replicated in a healthy lean population (clinical and metallomics data are shown in Tables S2 and S3, respectively). In regression analysis between neonatal and clinical data, we only found significant associations between CRP levels and birth length (p-value = 6.9 × 10−3) and its Z-score (p-value = 7.7 × 10−3). Moreover, we also failed to replicate most of the aforementioned metallomics findings (Table S4), being solely detected a few significant associations between weight-related metrics and some essential trace elements (i.e., selenium, zinc), as similarly reported in the childhood obesity cohort.
Discussion
A number of studies have proven that adverse birth outcomes (e.g., prematurity, being small/large for gestational age) associate with deleterious repercussions on metabolic health later during childhood, encompassing impairments in insulin function, glucose disposal, and inflammation3,19. However, we report for the first time that even slight deviations in neonatal metrics could predispose to pathogenic exacerbations in children showing birth metrics within-normal ranges. As expected, most of these associations between neonatal determinants and obesity-related complications occurring during childhood, in terms of classical metabolic risk factors and complementary metallomics assessments, could not be replicated in a healthy lean population.
Herein, gestational age at birth was negatively associated with various inflammatory and insulin-related markers, which reinforces the importance of proper gestation duration in shaping homeostatic mechanisms in the offspring20. Regarding anthropometric variables, lower birth length was found to be associated with increased BMI and related complications (i.e., inflammation, insulinemia, glycemia). However, no significant associations were detected between birth weight and the biochemical parameters under study, whereas the BMI at birth was solely associated with postprandial glucose in a positive fashion. Altogether, these findings suggest that birth length could be a better predictor of obesity-related metabolic risk than birth weight itself. Finally, gestational age and length at birth were also found to be associated positively with TC and/or LDL-C levels. This concurs with previous studies reporting lower cholesterol in preterm21 and small for gestational age22 children, which has been attributed to prenatal adaptations leading to disturbed synthesis and breakdown of this essential lipid.
In a second step, we analyzed the association between neonatal parameters and blood metal levels in children with obesity. One of our most robust findings was a negative association between gestational age at birth and plasma total and HMM contents of various toxic metal species, including arsenic, cadmium, and lead. In this line, numerous studies have reported that exposure to heavy metals raises the risk of prematurity13, but our results interestingly suggest that these low gestational age-related metal alterations could persist throughout childhood. Mechanistically, the toxicity of these metals has been linked to their ability to exacerbate oxidative stress, modulate the release of inflammatory cytokines, induce endoplasmic reticulum stress, and disrupt hormonal control, which altogether may provoke placental damage and lead to preterm delivery23–25, but also predispose to obesity and comorbidities5. Furthermore, it should be noted that heavy metals, especially cadmium, have high affinity for binding to metallothioneins, proteins that are involved in maintaining the homeostasis of essential trace elements such as zinc. Thus, even at low concentrations, heavy metals may induce deficiencies in nutrients that are crucial for optimal health26. Concurring with this, we found birth length Z-scores to positively associate with plasma levels of metalloproteins containing zinc and selenium. As stated above, these results could be explained by the pivotal role that trace elements play in developmental biology, mainly because of their participation in regulating growth hormone (GH) signaling. In particular, zinc can favor the formation of the more stable dimeric form of GH and its aggregation in secretory granules, prolong its binding to receptors, and strength signal transduction; whereas selenium, in the form of type II deiodinase selenoenzyme, is required for GH biosynthesis27. Supporting this hypothesis, other authors have previously reported that higher availability to essential micronutrients results in faster and more stable growth patterns during the first stages of life28,29. Moreover, maternal and neonatal supplementation with zinc and selenium has been proven to elicit significant benefits on the newborn growth30,31.
Unlike the aforementioned results supporting that higher gestational age and birth length (i.e., healthier neonatal metrics) associate with lower levels of potentially deleterious metal species and greater levels of vital micronutrients, BMI at birth was found to predispose to an unfavorable multi-elemental profile. In particular, we found a negative association with plasma contents of zinc (HMM), chromium (LMM), and molybdenum (HMM), as well as erythroid selenium (LMM). Trace elements have been reported to play a protective role against the characteristic pathophysiological mechanisms underlying obesity. For instance, zinc is known to be essential for the correct processing, storage, secretion, and action of insulin in pancreatic β-cells32. Chromium can regulate blood glycemia and dyslipidemia by modulating a myriad of interrelated biological processes involved in insulin signaling (e.g., tyrosine kinase receptor activation, glucose transporter stimulation) and lipid metabolism (e.g., β-oxidation, cholesterol synthesis)33. In this respect, it should be stressed that the low-molecular-weight chromium-binding substance is the most biologically active form of this metal, in line with our results showing a significant negative association between the BMI at birth and LMM chromium species, but not with total nor HMM contents. As a cofactor of multiple enzymes, molybdenum also participates in glucose-induced insulin secretion, signal transduction, glycogen synthase inactivation, and lipid peroxidation32. Finally, selenium is a pivotal component of the antioxidant defense, so its proper homeostasis is crucial for mitigating oxidative stress and inflammatory processes34. Altogether, these beneficial effects over obesity-related molecular disturbances could explain the negative association that we have found between birth BMI and trace elements. On the other hand, birth weight Z-scores showed a positive association with potentially harmful metal species, such as plasma cadmium (total and HMM fractions), erythroid lead and arsenic (total content), and labile iron in both biological matrices, which could be allocated to their endocrine disrupting properties and induction of weight gain35.
The main strength of this study is its retrospective nature, which has enabled us to decipher the association between neonatal metrics and obesity-related molecular disturbances occurring in childhood. Unlike most existing literature, normally focusing on populations that suffer from adverse birth conditions (e.g., prematurity, small for gestational age), we investigated how slight deviations in neonatal metrics within-normal ranges may influence later metabolic risk. Another strength was the implementation of a state-of-the-art analytical approach aimed at studying metal biodistribution instead of simply determining total contents to better understand the complex involvement of metal homeostasis in health status. However, some limitations deserve to be mention as well, such as the relatively small sample size. The analysis of paired blood samples collected at birth (unavailable in our study population) would have been of great interest to establish more robust links between early molecular disruptions experienced during the perinatal period and those persisting along childhood. Finally, we have to stress that, considering the retrospective nature of our study, information on maternal exposures that may affect birth outcomes (e.g., smoking) was not recorded, which makes necessary future investigations aimed at comprehensively elucidating the interplay between prenatal and neonatal determinants in shaping metabolic health of the offspring.
Conclusions
In this study, we have explored the impact that neonatal metrics may have on obesity-related molecular disturbances occurring later during childhood, namely inflammation and impairments in insulin and glucose homeostasis, and their association with altered blood levels of trace elements and heavy metals. Higher gestational age and length at birth were found to be associated with a healthier metabolic status, as mirrored in lower inflammation, insulinemia, and glycemia. This was in turn accompanied by a favorable multi-elemental profile, with gestational age at birth being negatively associated with deleterious heavy metals (i.e., arsenic, cadmium, lead) and birth length being positively associated with essential trace elements (i.e., zinc, selenium). In contrast, higher BMI at birth predisposed to exacerbated dysfunctions in glucose disposal, together with alterations in the levels of micronutrients participating in endocrine control (i.e., zinc, chromium, molybdenum, selenium). Altogether, these findings suggest that even slight deviations in neonatal parameters have a strong impact on later metabolic risk, considering both classical clinical markers (e.g., insulin resistance, inflammation) and complementary metallomics assessments, which could be allocated to the persistence of molecular disruptions experienced during fetal development throughout the lifespan.
Supplementary Information
Below is the link to the electronic supplementary material.
Author contributions
Conceptualization, R.G.-D.; methodology, Á.G.-D. and R.G.-D.; formal analysis, L.J.-S., Á.G.-D., and R.G.-D.; investigation, L.J.-S., Á.G.-D., and R.G.-D.; resources, J.D.-R. and R.G.-D.; data curation, R.G.-D.; writing—original draft preparation, L.J.-S. and R.G.-D.; writing—review and editing, L.J.-S., Á.G.-D., and R.G.-D.; supervision, R.G.-D.; project administration, R.G.-D.; funding acquisition, R.G.-D. All authors have read and agreed to the published version of the manuscript.
Funding
This research has been funded by Instituto de Salud Carlos III (PI22/01899), and supported by the Spanish Network in Maternal, Neonatal, Child and Developmental Health Research (RICORS-SAMID, RD24/0013/0020), co-financed by the European Union through FEDER founds 2021–2027. Á.G.-D. held an intramural grant from the Biomedical Research and Innovation Institute of Cádiz (LII19/16IN-CO24). J.D.-R. thanks the “Río Hortega” program from “Instituto de Salud Carlos III” (CM23/00026). R.G.-D. was recipient of a “Miguel Servet” fellowship (CP21/00120) funded by “Instituto de Salud Carlos III”.
Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Declarations
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Álvaro González-Domínguez, Email: alvaro.gonzalez@inibica.es.
Raúl González-Domínguez, Email: raul.gonzalez@inibica.es.
References
- 1.Di Cesare, M. et al. The epidemiological burden of obesity in childhood: A worldwide epidemic requiring urgent action. BMC Med.17, 212 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Larqué, E. et al. From conception to infancy—early risk factors for childhood obesity. Nat. Rev. Endocrinol.15, 456–478 (2019). [DOI] [PubMed] [Google Scholar]
- 3.Cauzzo, C., Chiavaroli, V., Di Valerio, S. & Chiarelli, F. Birth size, growth trajectory and later cardio-metabolic risk. Front. Endocrinol.14, 1187261 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Seneviratne, S. N. & Rajindrajith, S. Fetal programming of obesity and type 2 diabetes. World J. Diabetes13, 482–497 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.González-Domínguez, Á., Jurado-Sumariva, L. & González-Domínguez, R. Association between childhood obesity, trace elements, and heavy metals: Recent discoveries and future perspectives. Obes. Rev.25, e13764 (2024). [DOI] [PubMed] [Google Scholar]
- 6.González-Domínguez, Á. et al. Altered metal homeostasis associates with inflammation, oxidative stress, impaired glucose metabolism, and dyslipidemia in the crosstalk between childhood obesity and insulin resistance. Antioxidants11, 2439 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.González-Domínguez, Á., Domínguez-Riscart, J., Millán-Martínez, M., Lechuga-Sancho, A. M. & González-Domínguez, R. Exploring the association between circulating trace elements, metabolic risk factors, and the adherence to a Mediterranean diet among children and adolescents with obesity. Front. Public. Health. 10, 1016819 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.González-Domínguez, Á., Domínguez-Riscart, J., Millán-Martínez, M., Lechuga-Sancho, A. M. & González-Domínguez, R. Sexually dimorphic metal alterations in childhood obesity are modulated by a complex interplay between inflammation, insulin, and sex hormones. Biofactors49, 849–860 (2023). [DOI] [PubMed] [Google Scholar]
- 9.González-Domínguez, Á. et al. Trace elements as potential modulators of puberty-induced amelioration of oxidative stress and inflammation in childhood obesity. Biofactors49, 820–830 (2023). [DOI] [PubMed] [Google Scholar]
- 10.González-Domínguez, Á., Millán-Martínez, M., Domínguez-Riscart, J., Lechuga-Sancho, A. M. & González-Domínguez, R. Metal homeostasis and exposure in distinct phenotypic subtypes of insulin resistance among children with obesity. Nutrients15, 2347 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.González-Domínguez, Á., Jurado-Sumariva, L., Domínguez-Riscart, J., Saez-Benito, A. & González-Domínguez, R. Parental obesity predisposes to exacerbated metabolic and inflammatory disturbances in childhood obesity within the framework of an altered profile of trace elements. Nutr. Diabetes14, 2 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Thakur, G. K., Shankar, H., Arora, T. K. & Kulkarni, B. Role of mineral nutrients other than iron in pregnancy: Under recognized opportunities to improve maternal/fetal outcomes: A literature review. Arch. Gynecol. Obstet.309, 895–905 (2024). [DOI] [PubMed] [Google Scholar]
- 13.Issah, I. et al. Exposure to metal mixtures and adverse pregnancy and birth outcomes: A systematic review. Sci. Total Environ.908, 168380 (2024). [DOI] [PubMed] [Google Scholar]
- 14.González-Domínguez, Á., Millán-Martínez, M., Sánchez-Rodas, D., Lechuga-Sancho, A. M. & González-Domínguez, R. Characterization of the plasmatic and erythroid multielemental biodistribution in childhood obesity using a high-throughput method for size fractionation of metal species. In Mass Spectrometry for Metabolomics (Methods in Molecular Biology, vol. 2571). (ed. González-Domínguez R.) 123–132 (Humana Press, 2023). [DOI] [PubMed]
- 15.González-Domínguez, Á. et al. Recommendations and guidelines for robust, easily implementable and reportable quality control of metabolomics data. Anal. Chem.96, 1064–1072 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Martos-Moreno, G. Á., Martínez-Villanueva, J., González-Leal, R., Chowen, J. A. & Argente, J. Sex, puberty, and ethnicity have a strong influence on growth and metabolic comorbidities in children and adolescents with obesity: Report on 1300 patients (the Madrid Cohort). Pediatr. Obes.14, e12565 (2019). [DOI] [PubMed] [Google Scholar]
- 17.González-Domínguez, Á. et al. Blunted reducing power generation in erythrocytes contributes to oxidative stress in prepubertal obese children with insulin resistance. Antioxidants10, 244 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Heitland, P. & Köster, H. D. Human biomonitoring of 73 elements in blood, serum, erythrocytes and urine. J. Trace Elem. Med. Biol.64, 126706 (2021). [DOI] [PubMed] [Google Scholar]
- 19.Melville, J. M. & Moss, T. J. The immune consequences of preterm birth. Front. Neurosci.7, 79 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Berti, C. et al. Pregnancy and infants’ outcome: Nutritional and metabolic implications. Crit. Rev. Food Sci. Nutr.56, 82–91 (2016). [DOI] [PubMed] [Google Scholar]
- 21.Mortaz, M., Fewtrell, M. S., Cole, T. J. & Lucas, A. Cholesterol metabolism in 8 to 12-year-old children born preterm or at term. Acta Paediatr.92, 525–530 (2003). [PubMed] [Google Scholar]
- 22.Yu, X. et al. A comparison of the growth status, level of blood glucose, and lipid metabolism in small for gestational age and appropriate for gestational age girls with central precocious puberty: A retrospective study. Transl Pediatr.10, 783–789 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Wang, H. et al. Maternal serum arsenic level during pregnancy is positively associated with adverse pregnant outcomes in a Chinese population. Toxicol. Appl. Pharmacol.356, 114–119 (2018). [DOI] [PubMed] [Google Scholar]
- 24.Amegah, A. K., Sewor, C. & Jaakkola, J. J. K. Cadmium exposure and risk of adverse pregnancy and birth outcomes: A systematic review and dose-response meta-analysis of cohort and cohort-based case-control studies. J. Expo Sci. Environ. Epidemiol.31, 299–317 (2021). [DOI] [PubMed] [Google Scholar]
- 25.Ashrap, P. et al. Maternal blood metal and metalloid concentrations in association with birth outcomes in Northern Puerto Rico. Environ. Int.138, 105606 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Jacobo-Estrada, T., Santoyo-Sánchez, M., Thévenod, F. & Barbier, O. Cadmium handling, toxicity and molecular targets involved during pregnancy: Lessons from experimental models. Int. J. Mol. Sci.18, 1590 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Caputo, M. et al. Regulation of GH and GH signaling by nutrients. Cells10, 1376 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Liu, Y. et al. The association of serum toxic metals and essential elements during early pregnancy with body mass index trajectory of infants during the first years: A prospective study. Ecotoxicol. Environ. Saf.269, 115766 (2023). [DOI] [PubMed] [Google Scholar]
- 29.Zhang, X. et al. Maternal exposure to trace elements, toxic metals, and longitudinal changes in infancy anthropometry and growth trajectories: A prospective cohort study. Environ. Sci. Technol.57, 11779–11791 (2023). [DOI] [PubMed] [Google Scholar]
- 30.Sinha, B. et al. Enteral zinc supplementation in preterm or low birth weight infants: A systematic review and meta-analysis. Pediatrics150, e2022057092J (2022). [DOI] [PubMed] [Google Scholar]
- 31.Barman, M. et al. Maternal dietary selenium intake is associated with increased gestational length and decreased risk of preterm delivery. Br. J. Nutr.123, 209–219 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Barra, N. G. et al. Micronutrients impact the gut microbiota and blood glucose. J. Endocrinol.250, R1–R21 (2021). [DOI] [PubMed] [Google Scholar]
- 33.Genchi, G., Lauria, G., Catalano, A., Carocci, A. & Sinicropi, M. S. The double face of metals: The intriguing case of chromium. Appl. Sci.11, 638 (2021). [Google Scholar]
- 34.Barchielli, G., Capperucci, A. & Tanini, D. The role of selenium in pathologies: An updated review. Antioxidants11, 251 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Al Osman, M., Yang, F. & Massey, I. Y. Exposure routes and health effects of heavy metals on children. Biometals32, 563–573 (2019). [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.