Technology-Related Trace Elements in Paediatric Scalp-Hair Biomonitoring: Biomarker Validity, Environmental Relevance, and Implications for Risk Assessment Authors: Antonio Peña-Fernández, Rafael Moreno-Gómez-Toledano, Borja Martínez-Alonso, M. Ángeles Peña Fernández Published: 2026 ABSTRACT Technology-related trace elements (TTEs), including rare earth elements (REEs), platinum-group elements (PGEs), silver, antimony, bismuth, barium, strontium, vanadium and radionuclide-associated elements, are increasingly relevant to environmental exposure assessment. Children and adolescents are important sentinel populations, and scalp hair offers a non-invasive matrix for multielement analysis. However, a reported hair concentration may combine follicular incorporation, sweat and sebum deposition and external particles, and may also be influenced by collection, washing, cosmetic treatment and analytical contamination. Separately, left-censoring and the statistical treatment of non-detects can materially affect population summaries and comparisons. This critical narrative review integrates paediatric, occupational, environmental and multimatrix evidence through a structured evidence matrix and an element-by-element audit of biomonitoring assessment values. Hair can identify spatial contrasts and source-related patterns, but current evidence generally does not support direct inference of absorbed dose or clinical risk. In long-term occupational REE exposure, urine shows stronger exposure-response performance than blood; paired hair–urine data for thorium discriminate exposed groups without establishing quantitative equivalence; antimony findings vary by exposure setting; and barium biomonitoring equivalents are available for urine and plasma, but not hair. No health-based guidance values validated specifically for paediatric scalp hair were identified for any of the target TTEs. A seven-stage weight-of-evidence framework is proposed, positioning hair as a complementary screening and hypothesis-generating matrix within multimatrix assessment. 1. Introduction The expanding use of specialised elements in digital technologies, renewable-energy systems, low-carbon transport, advanced materials and medical applications is changing the profile of environmental metal and metalloid exposure. Rare earth elements (REEs), such as lanthanum and cerium, platinum-group elements (PGEs), silver (Ag), antimony (Sb), bismuth (Bi), vanadium (V), gallium (Ga), germanium (Ge), niobium (Nb), tantalum (Ta), tungsten (W) and related elements are used in magnets, semiconductors, batteries, catalysts, photovoltaics, electronics and high-performance alloys. Their strategic importance is reflected in current critical raw materials policy, including Regulation (EU) 2024/1252 [1]. In this review, these substances are considered collectively as technology-related trace elements (TTEs), an operational exposure category that does not imply shared toxicokinetics, common toxicity or exclusively anthropogenic origin. The term TTE is therefore used here as an operational umbrella category rather than as a formal regulatory or toxicological classification. It overlaps with concepts such as critical raw materials, technology-critical elements, emerging contaminants, potentially toxic elements and essential trace elements, but technological relevance does not imply toxicity, and a toxic element need not be technologically critical. Environmental-health interpretation must remain element- and species-specific and consider concentration, chemical form, particle size, bioavailability, exposure route and duration and developmental susceptibility. Human exposure may arise across the full technological life cycle, from extraction, beneficiation and refining to manufacturing, product use, mechanical wear, recycling and disposal. These releases coexist with geogenic inputs from rocks, soils, water and mineral dust. Consequently, the detection of a TTE in an environmental or biological sample is not, by itself, evidence of technological contamination. Source interpretation requires information on local geology, environmental media, particle characteristics, product use, exposure pathways and the spatial and temporal correspondence between the suspected source and the biological sample. Children and adolescents are relevant to this emerging exposure landscape for two distinct reasons. First, their exposure patterns may differ from those of adults because of greater contact with floors, settled dust, soil and outdoor surfaces, age-specific hand-to-mouth activity, diet and, for some exposure media, higher intake relative to body weight. Second, developmental stage may modify susceptibility to adverse effects. These considerations are related but distinct and do not imply that paediatric exposure is universally higher for every TTE. Mobility, school and transport environments, cosmetic practices and consumer-product use also change across childhood and adolescence [2]. Paediatric biomonitoring must therefore distinguish developmental groups and account for the environments in which young people actually spend time. Human biomonitoring can integrate exposure from several sources and routes, but its interpretative value depends on the suitability of the measured biomarker for the element, chemical species, exposure route, biological half-life and decision question [3,4]. Blood and urine remain the most established matrices for many inorganic contaminants. Scalp hair is nevertheless attractive in paediatric research because collection is non-invasive, generally well accepted, compatible with repeated or school-based sampling and suitable for storage and retrospective multielement analysis [5,6]. These practical advantages have sometimes encouraged a stronger interpretation than the evidence supports. A concentration measured in scalp hair may include elements incorporated during follicular growth, material deposited through sweat or sebum and particles or products retained on or within the hair shaft. Washing does not necessarily remove all external material, while cosmetic treatment, pigmentation, collection tools, digestion vessels, procedural blanks and analytical detection capability can materially affect the reported result [6,7,8]. Hair is therefore not a single, universally defined biomarker: depending on the element and study design, it may function as an internal-exposure indicator, an environmental-contact matrix, a passive particle collector or a composite of these processes. The distinction between analytical detectability and biomarker validity is particularly important for TTEs, many of which occur at ultra-trace concentrations and lack matrix-matched reference materials, established toxicokinetic models or paediatric health-based values. Direct comparisons show that hair does not consistently correspond with internal matrices. In children, hair concentrations have shown important limitations as biomarkers for several toxic and essential elements [9]. In adults, hair Pb concentrations were only weakly related to blood Pb concentrations, while hair Cu, Mn and Sr concentrations were not significantly related to the corresponding blood concentrations. Moreover, geographical differences observed in hair were not consistently mirrored in blood, indicating that an area-related pattern observed in one matrix should not automatically be interpreted as the same systemic exposure pattern [10,11]. Thus, an analytically valid hair concentration may still be unsuitable for estimating absorbed dose or health risk. The literature is currently fragmented across environmental-exposure studies, analytical-method papers, studies reporting population reference distributions or intervals and biomonitoring-guidance frameworks. Population reference distributions describe concentration variability within a defined population under a specified analytical workflow and may indicate whether an observation is common or unusual in that context; they are not health-based toxicity thresholds. A focused synthesis is needed to determine which conclusions are supported for paediatric TTE hair data, which require corroboration in blood, urine or environmental media, and which remain scientifically unjustified. The primary aim of this critical narrative review is to evaluate the validity and appropriate interpretative role of paediatric scalp hair for technology-related trace elements. Specifically, the review addresses four linked questions: (i) what environmental and biological processes determine the concentration reported in hair; (ii) which TTEs and exposure settings currently support occurrence, spatial or source-pattern interpretation, and which provide evidence of internal exposure; (iii) how analytical quality, left-censoring, population reference distributions and matrix-specific biomonitoring values constrain risk interpretation; and (iv) what evidence is required to progress from analytical detection to source attribution, toxicological relevance and proportionate public-health action. On this basis, the review develops a structured evidence map, audits available biomonitoring assessment values and proposes a seven-stage weight-of-evidence framework for research and practice. 2. Materials and Methods This structured critical narrative review used an author-curated full-text evidence corpus that was deduplicated and supplemented through targeted searches of PubMed, Google Scholar and publisher databases, together with backward and forward citation tracking; searches were updated to 6 August 2026. Search concepts combined scalp or human hair; children, adolescents and paediatric populations; human biomonitoring and exposure biomarkers; technology-related or technology-critical elements and the individual target elements (REEs, PGEs, Ag, Sb, Bi, Ba, Sr, V, U, Th, Ga, Ge, Nb, Ta and W); and terms addressing external contamination, washing, ICP-MS or ICP-MS/MS, analytical quality assurance, reference distributions, biomonitoring guidance values, toxicokinetics, health-risk assessment and multimatrix comparisons. Priority was given to peer-reviewed paediatric studies, environmentally or occupationally exposed populations, multimatrix investigations, analytical-validity studies, paediatric reference-distribution studies and authoritative or peer-reviewed biomonitoring frameworks; adult and occupational evidence was retained where paediatric evidence was unavailable or where it provided important toxicokinetic, analytical or exposure–response information. The final structured evidence matrix comprised 61 unique core sources: 32 human or exposure-setting studies (Table S1A), 20 multimatrix, analytical or statistical-validity sources (Table S1B) and 9 biomonitoring or risk-interpretation frameworks (Table S1C). Contextual, policy and assessment-value sources were used separately where relevant and were not assigned the same role as core biomarker-validity evidence. Because literature identification was iterative and targeted rather than a prospectively logged systematic-screening workflow, candidate records and exclusion decisions were not recorded in a form that permits a defensible retrospective count of all records retrieved and excluded; accordingly, we report the exact composition of the retained core corpus rather than reconstructing a PRISMA-style flow retrospectively. Studies were critically appraised according to design, population, sample size, age and demographic structure, the definition and magnitude of differences between exposure groups or levels, comparison populations, spatial and temporal correspondence, individual or pooled sampling, collection and washing protocols, cosmetic treatment, analytical validation, procedural-blank control, matrix-matched reference materials, LOD/LOQ definitions, treatment of censored observations, complementary matrices, dietary and behavioural factors that could influence exposure or confound biomarker associations, source attribution and the distinction between statistical association and toxicological interpretation. Greater evidential weight was assigned to studies combining independent biological or environmental matrices, clearly characterised differences in exposure, appropriate QA/QC and biologically plausible exposure–response relationships. A separate element-by-element audit assessed health-based biomonitoring values, biomonitoring equivalents, occupational biological assessment values and values validated specifically for paediatric scalp hair; assessment values were retained only for the biological matrix and interpretative purpose for which they had been derived. Statistical or background reference values, air occupational exposure limits and commercial laboratory interpretive ranges not derived as authoritative health-based biomonitoring thresholds were not classified as health-based hair values (Tables S2 and S3). Quantitative meta-analysis and numerical risk-of-bias scoring were not applied because of substantial methodological heterogeneity. During preparation of the review, ChatGPT (OpenAI; GPT-5.6 and the integrated OpenAI image-generation capability) was used to support evidence organisation, language drafting, structural refinement and graphical development. Scientific concepts, evidence selection, interpretation and intended figure content were defined and verified by the authors; no generative-AI tool was used to generate or fabricate experimental data, numerical results or statistical analyses or to determine study eligibility autonomously, and all AI-assisted outputs were critically reviewed by the authors. 3. Environmental Relevance of Technology-Related Trace Elements Table 1 summarises the principal uses, environmental sources, defensible interpretative roles of scalp-hair measurements and major limitations for the technology-related element groups considered in this review. Across the evidence summarised in Table 1 and Figure 1, several recurring exposure contexts emerge. Reviews of human REE exposure and toxicity provide the broader context for interpreting hair findings [12,13]. Under pronounced exposure differences associated with REE mining or smelting, scalp hair can preserve spatial and geochemical signals, but available multimatrix data do not establish proportionality to systemic uptake; in long-term occupational REE exposure, urine currently shows stronger exposure–response performance than blood [14,15,16,17,18]. Additional hair-based REE studies in children, adolescents and maternal populations broaden the evidence base but do not establish quantitative hair–dose validity [19,20,21]. For PGEs, environmental-fate and human-hair evidence supports occurrence and source-pattern interpretation but not dose inference [22,23,24]. In urban, industrial and consumer-product settings, Ag, Sb, Bi, Ba, Sr and V can show geographical or source-related patterns, but external deposition, cosmetic or product contact, geogenic and dietary inputs, analytical censoring and matrix-specific kinetics constrain dose interpretation [25,26,27,28,29,30,31,32,33,34]. At the ultra-trace concentrations relevant to PGEs, instrumental carry-over or memory can additionally affect results when analyte retained from preceding measurements is not removed adequately during washout. For U and Th, total hair concentrations do not determine isotope composition, chemical species or radiological dose; paired hair–urine data for Th can distinguish exposure groups without demonstrating quantitative equivalence [35,36]. Evidence for Ga, Ge, Nb, Ta and W remains sparse, whereas electronic-waste exposure represents a complex mixture that requires environmental characterisation and validated complementary matrices rather than hair alone [37]. Across these settings, scalp hair is therefore most defensible as a screening and source-pattern matrix rather than as a stand-alone surrogate of absorbed dose or clinical risk. 4. Scalp Hair as a Paediatric Biomonitoring Matrix 4.1. Practical Advantages Hair sampling is painless, non-invasive and generally acceptable in school, community and longitudinal research. It does not require venepuncture, immediate centrifugation or urine-collection facilities. Samples require little storage space and may be transported under less demanding conditions than many fluid matrices. Advances in minimally invasive biospecimen collection are particularly relevant to children’s exposome research [38]. Hair can complement dried capillary blood, saliva, deciduous teeth, urine and other matrices by providing a different exposure window and practical collection profile. A small mass of hair can support multielement analysis using ICP-MS or ICP-MS/MS. Archived samples can also be reanalysed as analytical technology improves or new environmental questions emerge. 4.2. Biological Incorporation Hair is produced within follicles through cellular proliferation and keratinisation. Elements circulating in blood may enter the growing shaft and interact with keratin, sulfhydryl groups, melanin, structural proteins and intracellular constituents. This pathway provides biological plausibility for hair as an exposure matrix. It does not establish a constant or linear relationship between hair concentration and circulating concentration. Elemental incorporation depends on chemical form, protein and melanin binding, circulating concentration, hair-growth phase, follicular physiology, nutritional status, age, sex and individual hair characteristics. Validation for one element cannot be transferred automatically to another. 4.3. Sweat and Sebum After emerging from the scalp, hair is exposed to sweat and sebum. Elements excreted through these pathways may deposit on or diffuse into the cuticle. This contribution may indirectly reflect internal exposure, but its magnitude is influenced by perspiration, scalp physiology, hygiene and hair products. It is not equivalent to follicular incorporation. 4.4. External Deposition External deposition is a major determinant of hair-associated elements. Particles from soil, road dust, mining emissions, industrial aerosols, smoke, water and cosmetics may adhere to the cuticle or become embedded within damaged structures. External material may still be environmentally informative. Hair-associated dust may indicate that a child occupied a contaminated environment and could have inhaled or ingested particles from the same source. However, the bulk hair concentration cannot then be assumed to represent systemic absorption. The endogenous–exogenous distinction is therefore not binary. Interpretation should consider the operationally measured fraction, whether the research question concerns internal dose or environmental contact, whether surface particles or wash fractions were characterised, whether the elemental pattern resembles local soil or dust, whether blood or urine supports systemic uptake and whether the finding persists in a newly collected sample. 4.5. Exposure Window and Segmental Interpretation Hair is often described as a long-term or cumulative matrix. More precisely, a hair segment integrates processes affecting that segment during and after formation. Approximate scalp-hair growth rates can support broad exposure windows, but growth varies among individuals, anatomical regions and growth phases. Segmental analysis should not be assigned exact calendar dates without individual growth information. Distal segments also have a longer opportunity for external deposition than proximal segments. 4.6. Hair Colour, Melanin and Fibre Morphology Some elements have an affinity for melanin, so pigmentation may influence incorporation independently of environmental exposure. Hair-fibre morphology also deserves consideration. Fibre geometry, curvature, cross-sectional characteristics, mechanical behaviour and water-swelling properties vary among individuals and have been shown to differ in their distributions among hair phenotypes from populations of different ancestry [39,40]. Such physical characteristics could influence surface interactions, retention of externally deposited material and responses to washing or cosmetic treatment, although direct evidence quantifying their effect on trace-element biomonitoring remains limited. Broad ethnic categories should therefore not be used as proxies for elemental uptake. Where relevant, studies should instead record directly observable hair characteristics, including natural colour, curvature or texture, treatment history and visible damage, together with appropriate demographic information. Across the evidence reviewed here, detailed hair morphology and ancestry or ethnicity were not reported or incorporated consistently into washing validation or multivariable interpretation. This represents a limitation of the current evidence base and should be addressed in future paediatric biomonitoring studies. 4.7. Cosmetic Treatment Dyeing, bleaching, straightening and permanent treatments alter cuticle integrity, porosity and chemical composition. They can introduce elements directly or modify susceptibility to environmental deposition and washing. Treated hair should be excluded where appropriate, analysed separately, included as a covariate or examined in sensitivity analyses. Cosmetic treatment is especially important for Ag and Bi and may also alter results for elements not directly present in the product by changing the structure of the hair shaft. 4.8. Paediatric Ethics and Communication Although hair sampling presents minimal physical risk, paediatric studies require ethical oversight, parental or guardian consent and age-appropriate assent. Cultural and personal sensitivities concerning hair should be respected. Sampling should be discreet and avoid visible cosmetic alteration. The return of individual results requires particular care. Families may interpret an elevated numerical result as a diagnosis even where no health-based value exists. Consent documentation should clarify whether individual results will be returned, which comparison distribution will be used, whether the result has recognised clinical meaning, when repeat analysis or recollection will occur and what environmental or clinical follow-up is available. 4.9. Archived Hair Archived hair can provide unique historical evidence. Its use requires documentation of the original collection date, anatomical location, segment length, storage container, storage environment, previous handling, sample mass, chain of custody, ethical authority for secondary analysis and comparability between historical and contemporary methods. Harmonisation of storage, shipment and associated data management is a wider priority for human biomonitoring and biobanking [41]. Metals are not expected to undergo biological degradation in stored keratin, but contamination, repeated handling and undocumented preparation remain possible. Archived hair should therefore be described as a retrospective occurrence baseline rather than an exact reconstruction of absorbed historical dose. 5. Pre-Analytical and Analytical Validity 5.1. Biomarker Fitness Before hair is collected, investigators should define the decision question. Hair may be suitable for occurrence characterisation, demographic comparison, geographical screening, source investigation, intervention follow-up, identification of unusual observations and selection for complementary biomonitoring. Suitability for estimation of internal dose or clinical interpretation requires substantially stronger element-specific evidence. 5.2. Sampling Location and Segment Samples should usually be collected close to the scalp from a consistent region, commonly the posterior vertex. The proximal end should be identified. The anatomical location, segment length, sample mass, proximal–distal orientation, collection date, collection device, storage container and storage conditions should be reported. 5.3. Collection-Tool Contamination Collection tools can contaminate ultra-trace analyses. Runkel et al. demonstrated that stainless-steel scissors may contribute materially to some elemental signals [42]. Tool validation should include unused-tool blanks, repeated cutting of clean materials, assessment of tool-alloy composition, cleaning-efficiency experiments, field-transport controls and evaluation of ceramic or alternative materials. A tool suitable for conventional metals may not be suitable for Nb, W, Co, Cr or other technology-related elements. 5.4. Participant Metadata Minimum participant-level information should include age, sex, natural hair colour, cosmetic treatment, washing frequency, swimming, water contact, hair and scalp products, medication, supplements, diet, seafood consumption, drinking-water source, residence, residential history, school location, activity spaces, proximity to traffic or industry, parental occupations and relevant hobbies or consumer-product use. Failure to collect these variables limits source attribution and may leave residual confounding that cannot be corrected statistically. 5.5. Washing Procedures No washing method has been shown to remove all external contamination while preserving all internally associated elements. Morton et al. demonstrated that different procedures remove externally bound elements with varying efficiency [43]. Verrey et al. proposed a sequential washing approach for a broad inorganic panel [44]. David et al., using LA-ICP-MS, showed that pre-cleaning did not improve every application and could alter spatial information [45]. These findings are not necessarily contradictory because bulk digestion and spatially resolved analysis measure different operational quantities. A washing protocol should report reagent identity and purity, sequence, number of cycles, reagent volume, duration, agitation, rinsing, drying, vessel material and whether wash solutions were analysed. Washed hair should not be described as a purely endogenous matrix unless this has been demonstrated experimentally. 5.6. Digestion and Instrumentation Closed-vessel microwave digestion with high-purity acids is commonly used for complete matrix destruction. Methods should report sample mass, acid composition and grade, vessel material, digestion temperature and pressure, final dilution, isotopes, collision or reaction gases, correction equations, internal standards, calibration range, drift correction, carry-over, washout time and result-acceptance criteria. Ultra-trace REEs, PGEs and technology-related elements are especially susceptible to contamination, instrumental carry-over/memory, incomplete washout and polyatomic interferences. Method performance at conventional-metal concentrations does not establish performance at the much lower concentrations of many emerging elements. Conventional complete acid digestion followed by ICP-MS or ICP-MS/MS provides an operational total-element concentration and does not preserve the original chemical species. This distinction is important for elements such as Sb, V and U because oxidation state and chemical speciation can influence environmental behaviour, bioavailability, toxicokinetics and toxicological interpretation [6,27,46]. When the decision question concerns internal dose or toxicological relevance rather than occurrence or source screening, species-resolved analysis or complementary speciation in an appropriately validated internal matrix should therefore be considered. Speciation requires dedicated sample preservation and separation procedures and cannot be reconstructed retrospectively from a complete acid digest. 5.7. Procedural Blanks Full procedural blanks should accompany each relevant preparation batch. Reagent blanks alone do not capture contamination introduced by cutting, washing, vessels, handling, digestion, dilution and the laboratory environment. Blank assessment should examine central tendency, dispersion, temporal drift, contaminated vessels, reagent changes, preparation batches and anomalous blank populations. A single global blank correction may be inappropriate where batches differ materially. 5.8. Negative Blank-Corrected Values Negative blank-corrected concentrations can occur when the sample signal is lower than the estimated blank contribution. These values should generally be retained in the internal analytical dataset. Replacing negative corrected values with zero alters the distribution and may bias summary statistics, correlations and censored-data models. For presentation, values may be classified as non-detected or below quantification, provided that the original corrected data and decision rules remain auditable. 5.9. Certified Reference Materials and Proficiency Testing Certified or reference materials should be selected according to the matrix and analyte. A hair certified reference material with assigned values for Pb, Cd or Hg does not validate the accuracy of REEs, PGEs or other elements lacking assigned values. Where appropriate assigned values are unavailable, validation should combine blank stability, spike recovery, replicate precision, calibration verification, isotope agreement, alternative reference materials, interference experiments, independent-method confirmation and interlaboratory comparison. Tadić et al. identified continuing gaps in matrix reference materials and proficiency-testing schemes across human biomonitoring [47]. 5.10. Multimatrix Analytical Validation Goullé et al. developed and analytically validated ICP-MS methods for whole blood, plasma, urine and hair and reported population distributions for numerous elements, including V, Ga, Ge, Rb, Sr, Pd, Ag, Sb, Ba, W, Pt, Bi and U [48]. Their workflow included method-specific linearity, precision, detection and quantification limits and participation in an interlaboratory quality-control programme. The study demonstrates that multielement measurements can be analytically validated across several biological matrices. It does not demonstrate that the resulting concentrations are biologically interchangeable. The reference datasets comprised 100 whole-blood, 100 plasma and 100 urine measurements and 45 hair measurements, and the study did not establish individual-level hair–blood or hair–urine relationships for the technology-related elements. The reported hair distributions are valuable analytical and population comparators, but they are not health-based thresholds. They should also not be treated as direct multimatrix validation because measurement feasibility across matrices is distinct from toxicokinetic equivalence. Olmedo et al. likewise validated methods for Cr, Cd, Mn, Ni and Pb in blood, urine, saliva and hair [49]. Analytical comparability is a prerequisite for multimatrix research, but biological equivalence must be evaluated separately. 5.11. Limits of Detection and Quantification Because this article is a review, no new LOD or LOQ was calculated by the authors. For each source study, the reported definition and calculation of LOD/LOQ were considered when available; values were not retrospectively recalculated because the necessary raw blank, calibration and recovery data were generally unavailable and analytical procedures differed among studies. Failure to report how LOD/LOQ was derived was treated as an analytical-reporting limitation. For future studies, LOD and LOQ should be derived from the complete analytical procedure, expressed in final hair units and incorporate procedural-blank variation, sample mass, dilution, recovery, instrument response and predefined decision criteria. Results should distinguish non-detected observations, detections below LOQ, reliably quantified results and analytically invalid observations. 5.12. Left-Censored Data Technology-related elements frequently produce left-censored datasets. Arbitrary substitution with zero, LOD/2, LOD/√2 or LOD can bias medians, variances, correlations and group comparisons, particularly where censoring is substantial or differs among groups [50]. Depending on sample size, censoring proportion, distribution and study objective, defensible approaches may include detection-frequency reporting, Kaplan–Meier estimation, regression on order statistics, maximum-likelihood estimation, censored regression, Peto–Peto or related tests and sensitivity analyses. No method is universally optimal. When censoring is extreme, occurrence-based reporting is more defensible than estimation of a continuous distribution. An element that is 100% censored cannot support meaningful percentile, correlation, regression or principal-component analysis. 5.13. Multivariate and Mixture Analysis Technology-related exposures often occur as mixtures. Road dust may contain PGEs, Sb, Cu, Zn, Ba and other wear-related elements. Mining emissions may contain REEs, U, Th, Pb, Cd or As. Electronic waste can generate product-specific mixtures. Multivariate approaches may include correlation analysis, hierarchical clustering, principal-component analysis, elemental ratios, enrichment profiles, receptor models and supervised classification. Elements should not be included after arbitrary substitution merely to create a complete matrix, because this can generate artificial covariance and false source components. A statistical component should not be labelled as traffic, industry or geology without independent source evidence. Table 2 presents the minimum pre-analytical, analytical and reporting requirements for paediatric scalp-hair studies. 6. Human Evidence Across Paediatric Populations 6.1. Population Patterns Across Contrasting Environmental Settings Population-based studies consistently demonstrate that scalp hair can reveal spatial or demographic heterogeneity, but they also show why geographical proximity should not be treated as a direct surrogate for internal exposure. Drobyshev et al. identified multielement differences among 166 schoolchildren living near a toxic-waste site and in comparison areas, illustrating the feasibility of school-based screening while also highlighting limitations arising from small geographical groups, regional background variability, absence of internal biomarkers and possible external contamination [51]. In a much larger study of 1595 schoolchildren from an oil-producing region of Kazakhstan, Batyrova et al. observed age-, sex- and area-related differences, but elemental patterns did not consistently follow proximity to oil production [52]. This result suggests that mobility, diet, geology, household conditions and local activity spaces may be more informative than distance to a nominal source. Beltrán-Ardila et al. identified school-level heterogeneity in pooled hair samples from children attending 14 schools in Bogotá [53]. Pooling provided an efficient means of geographical surveillance but prevented assessment of individual distributions, determinants or extreme observations. Taken together, these studies support hair as a population-screening matrix while demonstrating that spatial contrasts cannot be interpreted automatically as individual absorbed dose. The evidential value increases when the sampling design captures relevant activity spaces, analytical preparation is harmonised and environmental or internal matrices are collected contemporaneously. 6.2. European Paediatric Evidence and the Alcalá de Henares Programme The Alcalá de Henares research programme provides one of the most extensive internally coherent series of paediatric hair studies in Europe. Early investigations characterised distributions of conventional metals and metalloids in children and adolescents and examined differences related to age, sex, residential area and urban topsoil, the latter referring to topsoil sampled within the urban environment of Alcalá de Henares rather than to rural comparison soils [54,55,56,57]. These studies established several principles that remain directly relevant to emerging elements: hair distributions are usually skewed; age and sex may alter central tendency and upper tails; residential contrasts generate source hypotheses but do not prove source attribution; and environmental and hair concentrations do not necessarily correspond directly. Subsequent reanalysis of the archived samples extended the programme to Ag, REEs, PGEs, Ba, Sr, V and other technology-related elements [25,30,36]. The later studies addressed an analytically more difficult concentration range and placed greater emphasis on detection frequency, procedural contamination, censoring and the distinction between environmental occurrence and toxicological interpretation. The archive is scientifically valuable because it permits comparison across developmental groups and element classes within a common historical population. Its limitations are equally important: the samples were collected in 2001, the environmental datasets are not always temporally matched, blood and urine were not collected, analytical inventories vary among publications and the exposure profile predates the rapid expansion of several contemporary technologies. The earlier studies generally included 117 children and 96 adolescents, whereas the later analytical programme included up to 120 children and 97 adolescents; the exact number contributing to each element must therefore be reported rather than assuming an identical population across all publications. Independent Spanish studies broaden the comparison. Llorente Ballesteros et al. reported reference distributions in 648 children and adolescents from Madrid, while Ruiz et al. analysed 28 elements in 419 children from a Mediterranean population without a recognised major industrial source [58,59]. Torrente et al. combined childhood hair measurements with exploratory neuropsychological assessments, and Esplugas et al. provided a rare longitudinal community perspective through repeated biomonitoring near a hazardous-waste incinerator [60,61]. The Sicilian PGE study and the Kola Peninsula REE study further extend European evidence to traffic-industrial and mining settings [20,23]. Table 3 provides a structured evidence map of selected European paediatric studies rather than a pooled reference interval. Original population characteristics, age groups, sampling periods, analytical features, units, summary statistics and detection or censoring information were retained to minimise inappropriate direct comparison among methodologically heterogeneous datasets. Arithmetic means, geometric means, medians, percentiles and censoring-aware estimates were not treated as interchangeable, and converted values, where used, were reported alongside rather than in place of the original units. Accordingly, between-study differences in reported concentrations should not be attributed to exposure alone without considering hair washing, digestion, cosmetic exclusions, sample mass, blank correction, analytical sensitivity and the treatment of non-detects. The value of this comparison lies in identifying broadly compatible concentration ranges, methodological sources of variation and technology-related elements for which independent European paediatric evidence remains unavailable. The value of the European comparison is therefore not the derivation of a single “normal range”, but the identification of reproducible concentration ranges, methodological sources of variation and elements for which independent paediatric evidence is absent. 6.3. Nutritional, Behavioural and Health-Outcome Context Elemental distributions in hair do not depend exclusively on environmental source intensity. The Malagasy REE study found generally weak relationships with age and nutritional status, suggesting that hair growth, mineral metabolism, dietary insufficiency, socioeconomic circumstances and soil or dust contact may influence the observed concentration [19]. Urinary biomonitoring in Spanish children likewise demonstrated associations between dietary groups and several metals [34]. Diet, nutritional status, drinking water and product use should consequently be treated as core exposure determinants rather than optional covariates, regardless of the biological matrix selected. Evidence linking technology-related elements in hair with health outcomes remains sparse and methodologically heterogeneous. Maternal hair REEs were not associated with neural-tube-defect risk after adjustment for confounding at the relatively low concentrations examined [21]. Exploratory associations have been reported between hair REEs and nervous-system disease categories in children from the Kola Peninsula and between conventional hair metals and neuropsychological measures in Spanish children [20,60]. These studies differ in timing, outcome definition, exposure range and analytical panel and cannot establish a coherent dose–response relationship. Health-outcome research will require prospective or clearly temporally ordered designs, prespecified hypotheses, sufficient power, control of multiple comparisons, assessment of nutritional and socioeconomic determinants, characterisation of co-exposures and independent replication. Most importantly, the biological matrix used as the exposure variable must first be shown to represent the relevant exposure process. 7. Multimatrix Evidence and Biomarker Validity 7.1. Analytical Feasibility Does Not Establish Biological Equivalence The ability to measure an element in several biological specimens does not establish that those specimens represent the same exposure interval or biological process. Goullé et al. validated multielement ICP-MS procedures for whole blood, plasma, urine and hair, as well as reported population distributions for V, Ga, Ge, Rb, Sr, Pd, Ag, Sb, Ba, W, Pt, Bi, U and other elements [48]. This work is important because it demonstrates analytical feasibility across matrices, including elements rarely considered in routine biomonitoring. However, the blood, plasma and urine datasets included 100 measurements each, whereas the hair dataset included 45, and the study did not establish paired individual-level relationships between hair and internal fluids for the technology-related elements. The reported values are analytical and population comparators, not evidence that hair can substitute for blood or urine. Olmedo et al. similarly demonstrated that several conventional metals can be measured reproducibly in blood, urine, saliva and hair [49]. Direct biological comparisons nevertheless reveal limited correspondence. Rodrigues et al. found only a weak hair–blood association for Pb and no meaningful relationships for Cu, Mn or Sr, while Semenova et al. showed that geographical patterns could differ between blood and hair [10,11]. In children, paired blood and urine measurements displayed only weak correspondence for As, Pb and Sr and essentially no relationship for Ba [33]. These findings do not invalidate any one matrix; rather, they demonstrate that matrix selection must be guided by toxicokinetics, exposure timing and the decision question. Table 4 summarises the principal multimatrix studies informing the distinction between analytical feasibility, cross-matrix correspondence and validation of hair as a surrogate for internal exposure. 7.2. Element-Group Lessons from the Available Multimatrix Evidence REEs and Th currently provide the clearest illustration of why group discrimination and quantitative dose estimation must be separated. Hair REEs repeatedly distinguish strongly contrasting mining and reference populations, but the most persuasive exposure–response evidence presently favours urine in occupationally exposed adults [14,15,16,17,18]. For Th, hair and urine both differentiated mine workers from local residents, yet only urinary measurements were linked to radiological dose through an established biokinetic model and no simple quantitative relationship between the two matrices was demonstrated [35]. The principal missing evidence is a paediatric study collecting hair, urine and blood from the same participants across a well-characterised REE or Th exposure gradient. Sb and Ba show a second form of matrix specificity. Hair Sb was informative near an active mine but less informative in a geogenic exposure setting, while urinary speciation provided more direct evidence of internal processing [27,28,29]. For Ba, biomonitoring equivalents exist for urine and plasma, but paired paediatric blood and urine measurements showed no meaningful correspondence [31,33]. The implication is not merely that hair lacks a guidance value; it is that guidance values and biological interpretations cannot be transferred even between established internal matrices without evidence. For PGEs and Ag, the evidence remains primarily descriptive. Hair can reveal geographical or demographic differences and may identify direct product contact, but no paediatric study currently links environmental concentration, hair concentration, an internal biomarker, absorbed dose and adverse effect [23,25,26]. Mn provides a useful cautionary analogue. Despite a much larger literature, a review of 86 studies found no universally reliable biological matrix for non-occupational environmental Mn exposure, and occupational studies indicate that group discrimination does not necessarily support individual classification [62,63]. Emerging technology-related elements should therefore not be expected to acquire individual-level interpretability solely through increased analytical sensitivity. The overall evidence supports an element- and matrix-specific hierarchy. Hair is useful for exposure-pattern screening; blood or urine may provide stronger internal-exposure information for selected elements; and quantitative health interpretation requires a separate toxicokinetic and epidemiological evidence chain. Table 5 integrates the most defensible interpretation of scalp-hair measurements for each element group with the complementary matrices currently supported by the evidence and the status of available matrix-specific biomonitoring assessment values [64,65,66,67,68]. Mercury (Hg) is included in the table solely as a comparison element and not as a target TTE. Methylmercury provides an instructive example in which hair interpretation is supported by a comparatively coherent evidence chain linking a defined chemical species and exposure pathway with blood–hair kinetics and health-risk interpretation [69,70]. Its inclusion therefore illustrates the level of element- and species-specific validation that is currently lacking for most TTEs; mercury-derived hair values are not transferable to these elements. 8. From Population Distributions to Health-Risk Interpretation 8.1. Reference Distributions and Health-Based Biomonitoring Values A population reference distribution indicates how common or unusual a measurement is within a defined population, time period and analytical workflow. It does not indicate whether that measurement is safe, harmful or causally related to disease. Hair distributions are particularly sensitive to age, sex, geography, geology, hair pigmentation, nutritional status, cosmetic treatment, collection protocol, washing procedure, analytical platform, detection capability and statistical treatment. Health-based biomonitoring values require a substantially stronger evidential chain. Frameworks developed through HBM4EU emphasise the need for a suitable biomarker, a toxicological or epidemiological point of departure, toxicokinetic linkage, explicit exposure assumptions, assessment factors, population applicability and confidence evaluation [64]. Santonen et al. similarly concluded that biomonitoring contributes to risk assessment only when the biomarker is fit for purpose, sampling and quality assurance are adequate and the relationship with exposure or effect is sufficiently characterised [65]. Biomonitoring equivalents illustrate this matrix specificity. Health Canada has used such values in screening-level assessments where adequate toxicokinetic information exists [66], and published equivalents for Ba relate specifically to urine and plasma [31]. The German Human Biomonitoring Commission likewise distinguishes statistical reference values from health-related HBM values [67]. Moreover, health-based interpretations may be revised when new evidence emerges, as demonstrated by the suspension of former blood-Pb HBM values when a defensible effect threshold could no longer be supported [68]. Hair Hg can contribute to methylmercury risk assessment because a comparatively coherent evidence chain connects a predominant chemical species, dietary exposure, blood and hair kinetics and epidemiological effects [69,70]. Clinical reference-interval principles also require carefully selected populations, sufficient sample size, appropriate partitioning and external verification [71]. Matrix-specific HBM guidance for Cd further demonstrates that values derived for blood or urine cannot be transferred to hair, even for the same element [72]. Within the literature and official biomonitoring frameworks reviewed, no health-based guidance value, biomonitoring equivalent, clinical action level, occupational biological limit or toxicity threshold validated specifically for paediatric scalp hair was identified for REEs, PGEs, Ag, Sb, Bi, Ba, Sr, V, U, Th, Rb, Ga, Ge, Nb, Ta or W. Matrix-specific assessment values were identified for selected elements in blood, plasma or urine, including biomonitoring equivalents for Ag, Ba and Bi and occupational biological reference values for Sb, Ba and V; however, these values were derived for other matrices, populations and interpretative purposes, and none can be transferred to paediatric scalp-hair interpretation without an element- and matrix-specific toxicokinetic derivation. 8.2. Why Hair Concentrations Cannot Be Converted into Hazard Quotients A conventional non-cancer hazard quotient compares an estimated exposure dose with a health-based reference dose:HQ = estimated exposure dose/health-based reference dose Both quantities must represent compatible exposure metrics, commonly expressed as mass per kilogram of body weight per day. A hair concentration expressed as ng/g or µg/g does not directly represent daily intake, inhaled concentration, absorbed dose, circulating concentration, target-organ dose or cumulative body burden. Dividing a hair concentration by an oral reference dose therefore produces a numerical result without validated toxicological meaning. Environmental guidelines are equally matrix-specific. Soil screening values, drinking-water limits and air guidelines apply to their corresponding media under defined assumptions concerning intake, frequency, duration and body weight. They cannot be compared numerically with hair concentrations. Environmental exposure modelling can legitimately accompany a hair study, but the resulting estimate should be described as a pathway-specific assessment, such as a soil-ingestion screening assessment, rather than as validation of a safe or unsafe hair value. Direct hair-based quantitative risk assessment would require an element- and species-specific relationship between external exposure and hair, a validated relationship between hair and internal dose, a defined exposure window, quantitative control of external contamination, characterisation of population variability, relevant dose–response evidence and a validated hair guidance value. These conditions are not currently met for the technology-related elements considered here. 9. Weight-of-Evidence Interpretation and Public-Health Application 9.1. A Seven-Stage Interpretative Framework The proposed framework is intended to prevent progression from analytical detection to a toxicological or clinical conclusion without the necessary intermediate evidence. The initial stage defines the decision question, distinguishing occurrence surveillance, population comparison, source investigation, intervention evaluation, selection for additional testing, formal risk assessment and clinical management. Hair cannot be assumed to simultaneously represent a historical exposure record, an environmental particle sampler, an internal-dose biomarker and a diagnostic test. Analytical validity must then be established through verified sample identity, appropriate collection tools, standardised preparation, procedural blanks, interference control, calibration, precision, recovery, suitable LOD and LOQ, batch comparability and explicit censoring classification. Once analytical validity is secure, the population distribution can be described using detection and quantification frequencies, censoring-aware descriptive statistics, relevant age and sex strata, cosmetic-treatment information and review of extreme observations. Source plausibility should subsequently be evaluated through geology, residence, school and other activity spaces, traffic, industrial activity, diet, drinking water, consumer products, medication, household occupations and measurements in soil, dust, air or water. Source attribution should remain probabilistic because a compatible hair pattern rarely proves a source independently. Multi-matrix corroboration can then be sought using blood, urine, nails, deciduous teeth, repeated hair, wash fractions or relevant environmental media, with the objective of achieving convergent evidence rather than identical numerical results. Only after these stages should toxicological relevance be considered in relation to chemical species, particle characteristics, solubility, bioavailability, exposure route, critical effects, developmental susceptibility, human evidence and matrix-specific guidance values. Where the link with dose remains absent, interpretation should stop at exposure prioritisation or hazard identification. Proportionate action may range from no immediate intervention to analytical confirmation, recollection, environmental investigation, exposure reduction, validated clinical testing or specialist referral. Within Stage 6, “validated clinical testing” refers to confirmation using a clinically accepted, element- and species-appropriate biological test with an authoritative interpretative basis, typically in blood or urine where such testing exists, rather than by repeat hair analysis alone. “Specialist referral” refers to evaluation by an appropriate paediatric, clinical toxicology or environmental/occupational medicine service when confirmed exposure, compatible clinical findings or an authoritative matrix-specific action criterion warrants further clinical assessment. At every stage, insufficient evidence should trigger verification or additional data collection rather than automatic progression. No single methodological feature or cross-matrix observation is sufficient by itself to validate scalp hair as a surrogate for internal dose; progression through the proposed framework therefore requires convergent evidence across analytical validity, source plausibility, complementary matrices and toxicological relevance. These seven stages, together with their core questions, evidential requirements and permissible conclusions or actions, are summarised in Table 6 and visualised in Figure 2. 9.2. Population Surveillance, Individual Results and Environmental Justice The most defensible current public-health role of scalp hair for technology-related elements is structured screening. At population level, hair may identify geographical contrasts, demographic patterns, unusual upper-tail distributions, historical baselines, possible source profiles and groups requiring more detailed investigation. A coherent shift across a school or community may justify environmental follow-up even when no individual health threshold exists. Individual interpretation requires greater caution. An isolated high result may reflect external deposition, cosmetic or product contact, unusual behaviour, sample misidentification, analytical interference, batch contamination or genuinely elevated exposure. Before communicating toxicological significance, the result should be confirmed analytically, participant metadata should be reviewed, a new sample should be considered and a validated internal matrix should be obtained where one exists. Terms such as “toxic level”, “poisoning”, “safe concentration” or “deficiency” should not be used without an authoritative matrix-specific basis. Exposure assessment should also extend beyond residential address. Children divide their time among homes, schools, playgrounds, transport routes, sports facilities and other indoor and outdoor environments. Environmental measurements should correspond to those activity spaces and to the approximate period represented by the analysed hair segment. These interpretative safeguards are particularly important in communities near mines, traffic corridors, industrial facilities or waste-management sites. Hair screening may reveal environmental inequalities, but overstating uncertain results can stigmatise communities, generate unnecessary anxiety and transfer responsibility for investigation to families. Community engagement, transparent explanation of uncertainty and access to appropriate environmental or clinical follow-up should therefore be integrated into study design. 10. Conclusions, Applications, and Prospects Technology-related trace elements are increasingly relevant to environmental health as their use expands across renewable energy, digital infrastructure, transport, electronics, advanced materials and consumer products. Scalp hair offers important practical advantages for paediatric biomonitoring because sampling is non-invasive, readily repeated, compatible with multielement analysis and suitable for archived material. The evidence reviewed here nevertheless supports a more limited biological interpretation than these practical advantages might suggest. Hair can identify population distributions, geographical differences and source-related patterns, particularly where exposure differences are pronounced, but a measured concentration cannot generally be interpreted as a quantitative surrogate of absorbed dose or individual health risk. The principal limitation is that the measured hair signal may combine follicular incorporation, deposition through sweat and sebum and external material retained on or within the shaft. The relative contribution of these processes is element-, species-, exposure- and hair-dependent and is further influenced by collection, washing, cosmetic treatment, analytical blanks, detection capability and statistical treatment. Multimatrix evidence reinforces this distinction: urine currently provides the clearest exposure–response evidence for long-term occupational REE exposure; hair and urine can both discriminate Th-exposed groups without demonstrating quantitative equivalence; Sb performance differs among exposure settings; and Ba has biomonitoring equivalents for urine and plasma but not for hair. Consequently, the most defensible present application of paediatric scalp-hair analysis is complementary screening and hypothesis generation within a weight-of-evidence assessment rather than stand-alone clinical or dose assessment. The major research priority is direct paediatric validation linking well-characterised environmental sources with contemporaneous hair, urine and, where ethically and scientifically justified, blood measurements. Longitudinal and intervention studies should be prioritised over additional isolated cross-sectional surveys because within-person responses to remediation, relocation, traffic reduction, altered water supply or removal of a relevant product would provide stronger evidence of biomarker responsiveness. REE studies should evaluate individual elements and determine whether occupational urinary exposure–response relationships extend to lower paediatric exposures; PGEs and Ag require direct comparison with internal matrices; Sb requires greater emphasis on chemical speciation; Ba and Sr require paired-matrix studies; and repeated V measurements are needed to distinguish recent from longer-term exposure. Electronic-waste studies should integrate TTE mixtures with household or worksite dust, water, diet, parental occupation and child activity spaces. Analytical development remains equally important. Future work should harmonise full procedural-blank control, contamination and interference assessment, LOD/LOQ definitions, reference materials, proficiency testing, washing protocols and censoring-aware statistical workflows. Hair-fibre morphology, treatment history and characteristics that may affect external retention or washing efficiency should be documented more consistently. Conventional total-element analysis should also be complemented, where toxicologically relevant, by particle-, species- or isotope-resolved approaches because chemical species, oxidation state, isotope composition, solubility, particle size and bioaccessibility may be more informative than total concentration alone. Species-resolved methods are particularly relevant to Sb, V and U, while radiological interpretation of U or Th requires appropriate isotope and biokinetic information rather than total hair concentration. This review has several strengths, including integration of paediatric and occupational evidence, multimatrix comparisons, analytical-validity studies and biomonitoring frameworks, together with an explicit separation of analytical detection, exposure interpretation and toxicological interpretation. Its limitations should also be recognised. It is a structured critical narrative review rather than a formal systematic review or meta-analysis; the underlying literature is highly heterogeneous in population, source intensity, sampling period, hair preparation, analytical capability and statistical treatment, and no numerical risk-of-bias score was applied. Several target elements have been investigated in only one or a few paediatric populations, some of the strongest multimatrix evidence derives from occupationally exposed adults, and several publications from the author group derive from the same archived Spanish cohort and should therefore not be regarded as independent replications. Furthermore, much of the available literature measures total elemental content, and detailed hair morphology or ancestry-associated fibre characteristics have not been characterised consistently. Population reference distributions may indicate whether a hair concentration is common or unusual within a specified population and analytical workflow, but they are not toxicity thresholds and should not be transferred between analytical methods or converted directly into dose or hazard quotients. Until direct paediatric multimatrix validation, stronger species-resolved evidence and matrix-specific health-based interpretation become available, scalp hair should not be presented as a stand-alone measure of absorbed dose, clinical toxicity or individual health risk. Used with explicit analytical safeguards, environmental context and proportionate multimatrix follow-up, however, it can provide a valuable and ethically attractive component of paediatric environmental surveillance.