Introduction

Lead (Pb) is a non-essential heavy metal and a widespread environmental pollutant originating from both natural and anthropogenic sources. It is a highly toxic element and human exposure has been associated to numerous adverse health effects, such as cardiovascular, renal and hepatic system disorders in adults1,2,3,4,5,6, as well as immunotoxicity and intellectual and behavioural deficits in children7,8,9,10,11. In addition, Pb is considered an endocrine-disrupting chemical, due to its oestrogen-like properties evidenced in both in vitro and in vivo studies12,13,14,15,16. Pb compounds have also been classified as probably carcinogenic to humans (IARC, Group 2 A)17. Therefore, Pb exposure constitutes a significant environmental health challenge worldwide.

Currently, humans are exposed to Pb through numerous sources, including paints, tobacco, lead-acid batteries, contaminated foods, lead drinking water pipes and conduits, herbal remedies and spices, certain toys, cosmetics, electronics, and emissions from some industrial facilities, incinerators and old heating systems, among other industrial activities18,19. Furthermore, Pb is highly persistent in the environment, and humans are typically exposed via ingestion, inhalation and dermal contact, with the metal accumulating in the brain, liver, kidney, and, over time, in the teeth and bones20. Early developmental exposure to Pb may occur through both transplacental and lactation transmission9,20,21,22.

The exposure to and absorption of Pb is not homogeneous throughout the population and depends on different factors like the chemical characteristics of the Pb (chemical speciation, particles), exposure routes, and lifestyle and sociodemographic characteristics of the population, such as age, sex, residence, socioeconomic status, and educational level, among others23,24,25.

While three decades ago the World Health Organization (WHO) suggested a safe level of exposure to Pb of 100 µg/L26, in 2010, the risk assessment carried out by the European Food Safety Authority (EFSA)9 established a much lower blood reference level for the general European population, taking into account characteristics such as age and sex, as well as different adverse effects. The lowest benchmark dose derived for neurodevelopmental effects was set at 12 (0.50) µg/L Pb blood levels (corresponding dietary intake values in µg/kg body weight per day), at 15 (0.63) µg/L for chronic kidney disease and at 36 (1.50) µg/L for effects on systolic blood pressure27,28.

However, more recent epidemiological studies suggest that blood Pb concentrations lower than those indicated above (1–2 µg/dL) could also cause adverse health effects, so there is currently no evidence to establish a threshold concentration above which critical Pb exposure-induced effects develop3,29,30,31,32,33. It is therefore necessary to continue monitoring and implementing measures to reduce Pb exposure.

Blood Pb concentrations constitute a valuable tool, and the preferred biomarker for biomonitoring recent human exposures, representing recent exposure to Pb from various sources, such as food, water, soil, house dust, and polluted air. The half-life of Pb in erythrocytes of adult population ranges from 28 to 36 days20. Thus, some studies have explored the presence of Pb among the general population in some European countries over the past decades20,23,24,25,34,35,36,37,38. Information on Pb levels in children and teenagers after 2017 is more limited. Overall, the available studies indicate that the downward trend that started after the phase-out of leaded gasoline reached a plateau in 2010, stabilizing at approximately 10 µg/L20,35,38,39.

Furthermore, despite different regulatory policies implemented in many countries in recent decades, the population remains exposed, with levels currently estimated to be around 100 times higher than in the pre-industrial era6. The introduction of new sources of exposure (batteries, electronics, globalization of world trade, etc.) could be contributing to persistent variations in Pb exposure among European residents, suggesting the need for comprehensive and geographically targeted assessments.

Given the substantial evidence reporting associations between human exposure to Pb and adverse health effects40,41, significant efforts have been made globally to mitigate Pb exposure, with notable regulatory measures implemented worldwide. The European Union (EU) has played a pioneering role in addressing Pb pollution by imposing strict regulations and bans. The establishment of the Registration, Evaluation, Authorization, and Restriction of Chemicals (REACH) regulation and the ban on leaded gasoline (EU Directive 98/70/EC)42 have been significant milestones in the EU commitment in this regard9,28.

The development of biomonitoring programs is essential to determine pollutant concentrations both in the general population and in different subpopulations. Similar to other human biomonitoring (HBM) programs established in various countries around the world43,44,45, the European Human Biomonitoring Initiative (HBM4EU, www.hbm4eu.eu)46 listed Pb as a priority substance for HBM to contribute to the strengthening of policy ambitions on Pb in the European Union, which represents a crucial endeavour to investigate and understand the current landscape of Pb exposure across European populations47. With the above-mentioned, the aims of this study were to explore internal Pb exposure in a pool of nine studies from four EU countries participating in the HBM4EU project, as well as to investigate sociodemographic determinants and temporal variations over the last decades. This work is framed as a continuation of the research lines initiated in HBM4EU, now included within the objectives of the European project “Partnership for the Assessment of Risks from Chemicals-PARC” (www.eu-parc.eu).

Materials and methods

Study design and sampling frame

Within HBM4EU, metadata were collected from existing HBM studies through IPCHEM (https://ipchem.jrc.ec.europa.eu/hbm4eu_overview.html). Inclusion criteria for this study were datasets with available information on Pb concentrations in whole blood samples as well as information on age, sex, country of residence, period of data collection and degree of uncertainty in the biomarker measurement. Datasets focusing on clinical population (those recruited in-hospital) were excluded. The nine participating studies were asked to provide individual harmonized data according to a codebook developed within the HBM4EU project. The data template developed and provided by HBM4EU ensured a standardized format and included, for example, defined variables and variable names, concentration units and limits of quantification and detection. Harmonized data on basic demographic characteristics, such as age, sex, country or region of residence, and period of biological sample (and data) collection was obtained from the nine selected HBM studies. The use of the data was in accordance with the information provided in the informed consent of each study. For the purposes of this study, children were defined as those individuals aged ≤ 12 years old, adolescents as those aged 13–17 years old, and adults as those aged ≥ 18 years old.

The Pb concentrations studied were from the following 9 studies from 4 European countries:

  • Belgium: FLEHS I (2002–2006), II (2007–2011), III (2012–2015) and IV (2016–2020) studies, including both teenager and adult populations.

  • Czech Republic: SZU study (2003–2019), including children and adult populations.

  • Germany: ESB (2003–2019) (www.umweltprobenbank.de), GerES IV (2003–2008) and GerES V (2009–2019) (www.uba.de/GerES) studies, including all three age groups, and.

  • Spain: BIOAMBIENT.ES study (2009–2014), including adult population.

Germany, with three HBM studies (GerES IV, V, and ESB) provided an individual dataset consisting of 1,753 children, 524 teenagers and 6,233 adults. Belgium, with three studies (FLEHS I, II, III, and IV) provided individual data of 3,090 teenagers and 1,822 adults. The SZU study (Czech Republic) added 984 children, 1 teenager and 1,522 adults; and the Spanish study (BIOAMBIENT.ES) provided a dataset with 1,868 adults. All study participants (or legal tutors in the case of individuals < 18 years old) gave written informed consent to participate at time of sampling. Research protocols of the selected studies were approved by their respective ethics committees: Scientific Ethical Committee and the legal department of IBERMUTUAMUR from the Spanish study BIOAMBIENT.ES, the Ethical Committee of the University of Antwerp and the University Hospital of Antwerp and additionally by the Ethical Committee of the local hospitals if relevant for the FLEHS I, II, III and IV studies, the ethical committee of the National Institute of Public Health in Prague for the SZU study, the Berlin Chamber of Physicians for the GerES IV and GerES V studies, and the ethics committees of the Medical Associations Saarland and Westfalen-Lippe and the Medical Faculty of the Westphalian Wilhelms-University Münster for the German Environmental Specimen Bank (ESB). All research was conducted in accordance with the Declaration of Helsinki, and collection, storage, transfer, and use of data were carried out according to the European General Data Protection Regulation (GDPR, Regulation-EU 2016/679).

Chemical analyses

Exposure assessment was conducted by analysing Pb concentrations in whole blood samples. Recruitment and sampling were carried out according to the respective guidelines and standard operating procedures (SOPs) established in each study. Pb concentrations were analysed by using validated methodologies further detailed elsewhere, i.e., Cañas et al.34, for the Spanish BIOAMBIENT.ES study, Schoeters et al.35,39, for the four FLEHS studies from Belgium, Hrubá et al.48, for the SZU study from the Czech Republic, Vogel et al.38, and UBA (German Environment Agency)49for the GerES IV, V studies, and Lermen et al.35, for the ESB study from Germany. All the studies analysed Pb concentrations using inductively coupled plasma mass spectrometry (ICP-MS), with the exception of SZU, that used flame atomic absorption spectrometry24. Limits of quantification (LOQ) ranged between 0.048 and 2.1 µg/L33,34,35,37,39,50. The data sets did not contain any Pb measurements below the respective LOQ.

Covariables

The selection of determinants and sources of Pb exposure was limited to those for which information was available in the selected HBM studies, i.e., age (years), sex (male or female), country of residence (Belgium, Czech Republic, Germany and Spain), sampling year, degree of urbanization [Eurostat definition based on number of inhabitants https://ec.europa.eu/eurostat/web/degree-of-urbanisation/methodology): thinly populated area (rural area), intermediate density area (towns or suburbs), and densely populated area (cities)], education level (ISCED classification: https://ec.europa.eu/eurostat/statistics-explained/index.php?title=International_Standard_Classification_of_Education_(ISCED), explained/index.php? title = International_Standard_Classification_of_Education_(ISCED), low, medium, and high education), and smoking habit (smoker and non-smoker). Active smoking was included as an independent variable in all the models, including those for individuals under 18 years of age. Passive smoking exposure was not systematically assessed across all the included studies and was therefore not included in the statistical analyses.

Statistical analysis

A joint dataset of individual participants´ information was created. Categorical variables were described as percentages, and continuous variables as arithmetic means, standard deviations, maximum and minimum values, and percentiles (25th, 50th and 75th). To explore the variables associated with blood Pb concentrations, we conducted linear mixed-effects analyses. This approach allows for the examination of both fixed and random effects, providing a comprehensive analysis of the factors influencing blood Pb levels while accounting for potential within-group variations. The model included the following fixed effects: age, sex, sampling year, smoking habit, degree of urbanization, and educational level, which was included only in analyses restricted to participants aged 18 years or older. Since most participants under the age of 18 had not completed their education, this variable could not be considered an adequate indicator for socioeconomic status. The individual study from which each individual data was obtained was included as a random effect. This approach is preferable to considering the ‘study’ variable as a fixed effect, as it allows for modelling the variability between studies, separating the variance attributable to specific factors of each study (design, age, sex, region, time of collection, etc.), and accounting for the intrinsic heterogeneity inherent in study’s design, conduct, and data collection procedures.

We assessed the distribution of Pd levels in the population observing that it was moderately skewed, consistent with the known right-skewed distribution of blood Pb concentrations. The natural logarithm transformation of the values did not change the model estimates or the statistical significance of the coefficients. The analyses were therefore maintained using the non-transformed independent variable, facilitating comparison with established public health thresholds. Each beta coefficient indicates the change in the dependent variable for a one-unit increase in continuous variables or the difference between the reference category and other categories for categorical variables.

Given the potential differences in exposure determinants according to age, and taking into account the lack of information on the educational level of participants under 18 years of age, separate analyses were conducted for those under 18 years old —overall, as well as for children (≤ 12 years) and adolescents (13–17 years)— and for adults (≥ 18 years). The models for individuals < 18 years were not adjusted for educational level. In addition, as the degree of urbanization was not available in two of the nine studies (BIOAMBIENT.ES and FLEHS I), a sensitivity analysis was also performed by fitting the model without adjustment for this variable (shown in supplementary material).

Multicollinearity among the fixed effects included in the mixed-effects models was assessed using the Generalized Variance Inflation Factor (GVIF). To allow comparability across variables with different degrees of freedom, adjusted GVIF values were computed as GVIF^(1/(2*Df)), where Df represents the number of coefficients associated with each variable. All adjusted GVIF values were below 1.03, indicating no relevant multicollinearity among the fixed predictors of the models.

SPSS Statistics 23.0 (IBM, Chicago, IL, USA) and R statistical computing environment v4.4.1 (http://www.r-project.org/) were used for data analyses. Mixed-effects analyses were performed by using lme4 R library v1.1-35.151.

Results

The studies selected for the present work were conducted in Belgium, Czech Republic, Germany, and Spain between 2003 and 2019 and included a total of 17,790 participants (6,352 subjects < 18 years old and 11,445 adults) with available Pb concentrations (Table 1). Of all participants, 47.8% were from Germany (n = 8,505), followed by Belgium (n = 4,906), and, to a lesser extent, the Czech Republic (n = 2,499) and Spain (n = 1,880). Men and women were almost equally represented, with women accounting for 50.7% of the total population. Most participants lived in cities (overall 69.9%), and 15.9% reported being smokers. A large percentage of participants had a high level of education (67.8%), which is often the case in biomonitoring studies based on volunteers from the general population.

Table 1 Characteristics of the study populations.
Full size table

All participants showed detectable blood Pb concentrations (the data sets did not contain any Pb measurements below the respective LOQ, as all Pb concentrations exceeded detection and quantification limits in all samples), with an overall mean concentration ± standard deviation (SD) of 20.6 ± 16.7 µg/L. Mean ± SD for each age group were: 18.3 ± 10.1 µg/L for individuals < 13 years old, 18.5 ± 14.1 µg/L for individuals aged 13–17 years, and 21.8 ± 18.4 µg/L for individuals > 17 years old (Fig. 1).

Fig. 1
figure 1

Overview of countries included in the pooled analysis, variables associated with blood Pb levels (multivariable mixed-effects models), and blood Pb levels by age group. The variable “smoker” refers only to active smoking.

Full size image

The results of multivariable mixed-effects models analysing the determinants of Pb concentrations for each age group are summarized in Tables 2, 3, 4 and 5. Among individuals younger than 18 years of age (Table 2), each additional year of age was associated with an average 0.4 µg/L decrease in Pb concentrations. Pb concentrations showed a significant reduction over time compared to the reference period (2003–2005). Overall, on average, females < 18 years of age had Pb concentrations 1.23 µg/L lower than males, although this difference was not statistically significant in the children group (Table 3). The positive associations between Pb levels with age and smoking found in the above population groups were not observed in the teenager’s group (Table 4). In adults, age was positively associated with Pb concentrations, with each additional year corresponding to an average 0.3 µg/L increase in Pb levels (Table 5). Year of sampling continued to show a significant decreasing trend, with reductions in Pb concentrations over time. Female participants had average Pb levels 3.1 µg/L lower than those of males. Smoking was also significantly associated with higher Pb concentrations, with smokers showing average Pb levels 2.5 µg/L higher than nonsmokers. Educational level was inversely associated with Pb concentrations, with individuals with higher education showing average Pb levels 4.1 µg/L lower than those with fewer years of education. Finally, residents in cities or suburbs had average Pb levels between 2.9 and 3.5 µg/L higher than those in rural areas, with this difference being statistically significant only among suburban residents. The sensitivity analyses without adjustment for the degree of urbanization yielded similar results, except that the positive effect of smoking on Pb concentrations lost statistical significance in individuals < 18 years old (Supplementary Tables S1, S2, S3, S4).

Table 2 Variables associated with Pb concentrations in individuals < 18 years old. Linear mixed-effects analyses (n=3692).
Full size table
Table 3 Variables associated with Pb concentrations in individuals ≤ 12 years old. Linear mixed-effects analyses (n=1749).
Full size table
Table 4 Variables associated with Pb concentrations in adolescents (13-17 years old). Linear mixed-effects analyses (n=1943).
Full size table
Table 5 Variables associated with Pb concentrations in adults (≥18 years old). Linear mixed-effects analyses (n=7369).
Full size table

Discussion

This work represents one of the largest analyses of blood Pb concentrations data among European citizens and provides valuable information on exposure levels and patterns in a pooled population from Belgium, the Czech Republic, Germany, and Spain, as well as on the temporal trends and associations with sociodemographic data.

The Pb concentrations reported in most of the included European studies (except for the German studies) were considerably higher than those reported in the United States during the same study periods (13.8 µg/L)52. The lower exposure levels found in the German participants (12.24 µg/L) could be a consequence of the relatively earlier implementation of risk reduction measures aimed at decreasing Pb exposure, such as the phase-out of leaded gasoline and air quality regulations53,54, compared to the other European countries. In contrast, Spanish participants showed the highest Pb concentrations. These results should, however, be interpreted with caution due to the relatively restricted sampling period (2009–2010) of the Spanish BIOAMBIENT.ES study, as well as the characteristics of the population, a Spanish occupational population sample33.

Multivariable regression models showed increasing Pb levels with increasing age in adults, which could be related to longer exposure periods and, consequently, to a higher bioaccumulation potential, as well as to higher environmental levels of Pb in the past, in agreement with previous studies23,55,56. In children, however, age was inversely associated with Pb levels. Several scientific studies support that, in children, blood Pb levels trend to decrease with age, in contrast to the trend observed in adults. This decrease is attributed to a cohort effect, reflecting a progressive reduction in environmental exposure to Pb in more recent generations57,58.

Previous studies have also shown higher Pb concentrations in the male population, in line with our results23,24,25,55,59,60,61,62,63. Differences in exposure levels between men and women could be related to sex-specific lifestyles, as for example, more frequent manual occupations among males64 could promote higher occupational exposure to this metal. In addition, higher exposure to tobacco smoke and alcohol among men65 could be also related to higher Pb levels, as this metal is present in tobacco smoke66. Alcohol consumption may also dysregulate hepatic metabolism, favouring the bioaccumulation of organic Pb67. However, the fact that these differences were also observed in teenagers and children suggests an additional possible sex-related effect. Although research on sex-related differences in Pb metabolism is limited, previous studies have reported sex-specific associations with health outcomes, which appear to be more pronounced in male children68. Moreover, inorganic Pb is mainly found in blood erythrocytes, and men have been shown to have higher blood haematocrit levels69. Blood loss due to menstruation in women may also contribute to increased Pb excretion and, therefore, contribute to the abovementioned sex differences70.

The results underscore the dynamic nature of Pb exposure, reflecting both historical and contemporary factors influencing environmental contamination. The observed temporal trends in Pb exposure reveal a complex interplay of historical industrial practices, regulatory interventions, and societal changes. Our findings indicate that 100% of the participants were exposed to Pb, with a trend toward lower Pb concentrations in the most recent studies. This trend is in line with previous findings71,72 and correspond with stringent regulatory measures such as phasing-out of Pb in petrol, removal of Pb from paints, ceramics, water pipes and food cans and a reduction of atmospheric emissions of anthropogenic Pb from 2000 to 201827,73. However, this decline in Pb levels is insufficient for public health purposes, given the growing concern on persistent sources of Pb exposure, particularly in low- and middle-income countries, and recent evidence linking Pb exposure to adverse health effects, even at low and very low concentrations20,29,74.

The Declaration of Brescia on the prevention of metals neurotoxicity lowered the maximum Pb concentration to which vulnerable groups, such as children and women of childbearing age75, could be exposed. Early Pb exposure can lead to adverse health effects in children, as some studies have reported decreased IQ in children with blood Pb levels ≥ 1 µg/dL76. Consequently, the Centers for Disease Control and Prevention (CDC) revised its reference value for Pb from 10 to 5 µg/dL in children77. Interestingly, despite obvious differences with the CDC populations, the percentage of children in our study (only from the Czech Republic and Germany) with blood Pb levels above 10 µg/dL and 5 µg/dL were 0.1% and 2.6%, respectively (data not shown). In adults, an increased risk of cardiovascular morbidity and mortality is also observed at apparently low concentrations (1 µg/dL)76. Globally, the contribution of human Pb exposure to health expenditures has been estimated at $6 trillion US Dollars, with 77% of this cost attributed to cardiovascular morbidity and mortality, and 23% to IQ loss78.

Although the present study does not allow us to reach a consistent conclusion on the distribution of Pb concentrations in different age groups, European regional areas, or temporal trends, our results highlight that Pb exposure in Europe remains a public health concern. It is therefore important to develop more targeted strategies to manage Pb exposure and its potential health implications across different age groups, particularly in the most vulnerable groups (i.e., children and pregnant women or women of childbearing age), and at vulnerable life stages, such as periods of development. Moreover, given that no threshold for Pb toxicity can be derived79, a thorough review of regulatory policies, the determination of a greater number of preventive measures, and the development of risk reduction strategies targeting vulnerable population groups are of vital importance. These actions will allow health agencies to prioritize interventions and optimize the implemented measures.

The results of this study also highlight the value of both consolidated and ongoing biomonitoring initiatives and emphasize the need for site-specific strategies to address sources of human exposure and contamination. The identification of population groups with sustained Pb exposure provides crucial information for policymakers to implement targeted interventions and allocate resources effectively. HBM programs, together with longitudinal studies, are crucial for tracking the effectiveness of regulatory actions, identifying emerging risks, and guiding evidence-based policy adjustments. In this context, the identification of determinants of Pb exposure also plays an important role for taking preventive measures. Our results are congruent with previous research80, and emphasize the importance of considering socio-economic factors, lifestyle behaviours, and regional industrial activities in understanding the drivers of Pb exposure, in more homogeneous study populations to assess the contribution of these specific variables.

The individual studies included in this work were heterogeneous and not harmonized both in the analytical methods (and LOQs) used and in the sociodemographic characteristics of the participants. Therefore, we cannot rule out that both methodological factors and those associated with differences in the recruitment strategy or the different regulatory policies applied, and/or the degree of restrictiveness of the measures implemented, may have influenced the results obtained. For example, the variable ‘study’ was included in the statistical models as a random effect. Although this approach would allow modelling and explain the variability between studies attributable to specific factors (design, age, sex, region, time of collection, etc.), may not fully disentangle the impact of contextual factors such as the national regulatory background, geographic distribution, or differences in urban development. These factors are likely to have contributed to the observed heterogeneity in Pb levels and deserve to be explored further in future research. Nevertheless, as mentioned above, all adjusted GVIF values were less than 1.03, indicating that the variable ‘study’ was not related to the other explanatory variables.

The results of the present study should therefore be interpreted with caution due to the inherent limitations, including the lack of harmonized recruitment protocols (e.g. different study designs, analysis methodologies, populations, and sampling years), and the lack of common available information on more exposure determinants, especially occupation and diet, limiting the consistency and robustness of the analyses performed. Although data on active smoking were available, even for some participants under the age of 18, not all studies included information on passive smoking, so this could not be taken into account despite its relevance, especially in younger populations. However, the pooled analyses of several EU cohorts and the large sample size of this study could provide an integrated perspective of Pb concentrations in the blood of a large sample of European residents, suggesting geographical and temporal differences. In addition, all the analysed samples showed concentrations > LOD (limit of detection), suggesting a low risk of bias derived from the uncertainty of values < LOQ.

Conclusions

This study sheds light on the evolving landscape of Pb exposure in Europe, emphasizing the successes achieved through regulatory interventions while highlighting the challenges that remain. Our data indicate the sustained presence of Pb levels in the population over recent years and highlight the need to establish harmonized criteria to update regulatory policies and implement preventive measures aimed at reducing or avoiding Pb exposure, thereby mitigating its adverse effects on human health. Collaborative efforts between public health agencies, environmental regulators, and communities are essential to comprehensively address the multifaceted nature of Pb exposure. Future research should continue to monitor trends, explore emerging sources of exposure, and refine strategies to ensure sustained progress in reducing Pb exposure and protecting the health of European populations. It would also be necessary to have more harmonized data to reduce the uncertainty in the joint analysis, one of the major obstacles for the integrated interpretation of human exposure data.