Polychlorinated biphenyls (PCBs) are a class of chemical compounds in which 2–10 chlorine atoms are attached to the biphenyl molecule. There are 209 possible chlorinated compounds called congeners based on chlorine positions on the benzene ring (1, 2). During the 20th century, PCBs were extensively produced worldwide, primarily in the form of commercial mixtures like Aroclor^®, Clophen^®, Phenclor^®, Kanechlor^®, or Delor^®. Consequently, exposure to PCBs typically occurs through mixtures of PCBs rather than individual congeners (3, 4). In addition, there is evidence that humans and animals are commonly exposed to mixtures of multiple persistent organic pollutants (POPs), such as PCBs, furans, and dioxins (5).

PCBs have been widely recognised as POPs due to their environmental persistence and adverse health effects (1, 2). PCBs are part of the Stockholm Convention on Persistent Organic Pollutants due to their toxicity. The treaty, adopted in 2001 and enforced in 2004, aims to eliminate or restrict the production and use of POPs, including PCBs, protecting human health and the environment (6).

PCBs are lipophilic, therefore, they can bioaccumulate in the food chain and the human body. Exposure to PCBs can result in adverse health effects (7). PCBs are categorized into two primary groups, dioxin-like (DL) and non-dioxin-like (NDL) PCBs, based on their chemical structure and different toxicological effects. DL PCBs possess a planar configuration, enabling them to bind to the aryl hydrocarbon receptor (AhR), similarly to compounds in the dioxin family. This initiates specific biological pathways related to the modulation of AhR signaling (8, 9). DL PCBs are generally regarded as more toxic, as evidenced by animal studies. They are involved in carcinogenesis, immunotoxicity, neurodevelopmental disorders, thyroid hormone disruption, reproductive toxicity, and hepatotoxicity. NDL PCBs exert their toxic effects via other mechanisms, such as oxidative stress, disruption of calcium signaling, and interaction with estrogen and androgen hormone receptors, as well as thyroid hormone transport proteins. Compared to DL PCBs, NDL PCBs have demonstrated more pronounced effects on neurodevelopment and thyroid hormone regulation, in addition to exhibiting reproductive toxicity and hepatotoxicity (9, 10).

Epidemiological studies from Taiwan showed that higher levels of PCBs were associated with more frequent medical problems, such as abnormal menstrual bleeding, skin diseases, goitre, anaemia, joint and spine diseases (11, 12). In addition, a study by Codru et al. (13) reported that there is a higher odd of being diagnosed with diabetes in adult Native Americans with higher levels of PCBs. The results of cohort studies from eastern Slovakia indicate the associations between exposure to PCBs and impaired cochlear function (14), neurodevelopment (15), immune (16), and thyroid function, and metabolic activity (17) in children and adults. Due to these concerns, their production has been banned or significantly restricted in many countries (1, 2).

Once released, PCBs strongly bind to soil and sediment, causing them to remain in the environment for extended periods, with most types breaking down over months to years. They slowly seep out of the soil, especially those more heavily chlorinated congeners, and their transfer to plants through soil is minimal. The process of PCBs moving through the environment includes their evaporation from land and water surfaces into the air, followed by their return to the earth through wet or dry falling, and then evaporation again. For the general population, inhalation of PCBs present in the air is one way people are exposed, besides ingesting PCBs through food (1). Animal-origin foods, such as milk, dairy products, eggs, and mussels, are particularly susceptible to contamination with persistent organic pollutants (POPs) due to their bioaccumulation potential. This has been demonstrated in the Province of Taranto, Southern Italy, where industrial activities have led to elevated levels of POPs, especially PCBs, in food products, with egg and mussel samples exceeding the EU’s recommended limits (18–21). Similarly, a study by Agudo et al. (22) examined serum PCB concentrations in adult participants from both northern and southern Spain as part of the EPIC-Spain cohort. They found that dietary habits, particularly fish consumption, were strongly associated with increased serum PCB levels, with notable regional differences in exposure. Further study using polytopic vector analysis (PVA) revealed distinct congener patterns across participants. The PVA identified five unique patterns of PCB mixtures, one of which resembled air samples collected near contaminated sediment deposits at Akwesasne, indicating possible recent inhalation exposure to volatilized Aroclor 1,248. Another pattern was linked to unaltered Aroclor 1,254, while a third resembled Aroclor 1,262 but lacked labile congeners. The remaining two patterns reflected subsets of major persistent congeners, suggesting differences in bioaccumulation profiles or individual toxicokinetics (23).

It has been shown that eastern Slovakia represented one of the most heavily polluted sites by PCBs worldwide (24–29). In high-volume ambient air samples collected in residential areas at 1 to 30 km from the former PCB plant Chemko (Michalovce District), extremely high PCB values of the sum of 6 PCBs concentrations of 1700 pg./m³ were found (26). Sediment samples from rivers and creeks collected in the Michalovce district showed contamination levels of 1,700–3,100 ng/g d.w. (total PCBs). Soils collected near the Chemko waste storage and dump site showed contamination levels of 400–5,800 ng/g (total PCBs) (24). Comparison of the mean serum concentration of the sum of PCBs of 2,871 ng/g lipids for adults (n = 149, age (mean ± SD) 45.3 ± 8.4) of Michalovce district in 2015 with the value of 3,105 ng/g lipids for a corresponding population sample (n = 1,008, age (mean ± SD) 44.6 ± 12.5) of Michalovce district in year 2001 (27) shows that the exposure of adults in Michalovce district decreased insignificantly over 14 years. An increasing emission trend of PCBs from 2010 to 2017 was observed (30). Despite the persistent source of environmental contamination, the storage of hundreds of tonnes of PCBs in the buildings of the former Chemko Strážske company had not been officially confirmed for years. In 2021, the removal of barrels with PCBs was initiated.

The exposure routes of the population living in PCB-polluted areas have not been comprehensively elucidated to date. Previous studies, such as Strémy et al. (28), did not find a significant association between food consumption and PCB serum concentrations in a 2016 analysis. Our current study expands on this by investigating additional and previously unaddressed factors contributing to PCB exposure, beyond dietary sources, for populations in eastern Slovakia near the Chemko plant. Using principal component analysis (PCA), we identified groups of PCB congeners based on their similarities and developed a novel scoring system to quantify exposure from multiple sources, including the frequency and origin of food consumption, as well as residential proximity to contamination sites. Additionally, we incorporated spatial analysis through mapping to assess the impact of geographical location on exposure risk. This approach not only provides deeper insights into localized PCB exposure pathways but also presents a scalable methodology that can be applied to other PCB-polluted regions globally. By identifying critical exposure factors, our findings offer valuable implications for understanding PCB contamination in various environmental contexts, contributing to broader efforts in mitigating PCB exposure worldwide. The findings have broader implications for understanding how environmental exposure to PCBs affects human health, especially in regions impacted by long-term industrial pollution.

Baseline characteristics of subjects are shown in Table 1. Figure 2 presents the component plot of 20 PCB congeners analysed in this study. PCB 123 did not show a significant contribution to any of the three principal components in this PCA analysis. The analysis resulted in three principal components (PCs), which accounted for 72.03% (PC1), 8.72% (PC2), and 5.14% (PC3) of the total variance. PC1 was found to be positively controlled by PCB congeners with six or more chlorine atoms, including PCB 138, PCB 153, PCB 156, PCB 157, PCB 167, PCB 170, PCB 180, PCB 187⁺¹⁸², PCB 189, PCB 194, PCB 196⁺²⁰³, and PCB 199⁺²⁰¹. These congeners are represented by blue bullets in the component plot. PC2 was positively controlled by PCB congeners with three to five chlorine atoms, specifically PCB 28, PCB 52, and PCB 101, which are indicated by red bullets in the plot. PC3 was positively controlled by PCB congeners with four to five chlorine atoms, including PCB 74, PCB 99, PCB 105, PCB 114, and PCB 118. These congeners are represented by yellow bullets in the plot.

To identify the key factors contributing to PCB exposure, a stepwise multivariate regression analysis was conducted, examining the associations between the sums of PCB congeners represented by PCs and diet, residency location, and potential confounding variables – sex, age, active and second-hand smoking, alcohol consumption, work in the industry, and contact with toxic material. The following models were chosen according to the lowest AIC values as the models with the best fit:

As the results of stepwise multivariate regression showed, the score of residency location and age of our subjects are the two strongest predictors of PCB concentrations represented by PC1-3. Figure 6 illustrates the scores plots showing the mild incline of the score of residency location along the PC1 and PC3 axes. Based on these scores plots, it can be inferred that subjects with higher scores for residency location were predominantly exposed to congeners from the PC1 and PC3 groups. Conversely, subjects with lower scores for residency location were mainly exposed to PCB congeners from the PC2 group. This distinction in exposure patterns underscores the importance of residency location in the distribution of PCB congener profiles. Next, the PCA score plots for PC1 and PC2, as well as PC2 and PC3 (Figures 7A,C), did not reveal any clear patterns regarding the association between age and these PCs. However, the PCA score plot for PC1 and PC3 (Figure 7B) showed a noticeable pattern. It indicated that older individuals tended to have higher exposure to PCB congeners represented by PC1, while younger individuals had relatively higher exposure to PCB congeners represented by PC3.

To explore the potential relationship among specific PCB congeners in our study, we conducted PCA. The PCA yielded three PCs, with PC1 influenced by highly chlorinated PCBs, while low chlorinated PCBs influenced PC2 and PC3. Previous studies utilising PCA to identify PCB components have also observed that the categorisation of PCBs is mainly based on their physicochemical properties, precisely the number and location of chlorine atoms. In a study by Ravenscroft et al. (38) at the Akwesasne Mohawk Nation, United States, PCA analysis resulted in four components. PC1 consisted of PCB congeners with lower levels of chlorination (52, 70, 90_101, and 123_149), explaining 50.1% of the variance. PC2 encompassed highly chlorinated congeners (164_138, 153, 180, 99, and 187), explaining 15.9% of the variance. Additionally, PC3 and PC4 included highly chlorinated congeners (105, 118, and 110, as well as 153, 164_138, and 99), each contributing to 6% of the variance. Another study investigating PCB exposure in a heavily contaminated area in Anniston, United States, observed a similar pattern. The results of the PCA analysis revealed the division of PCB congeners into four groups based on their degree of chlorination. Group A consisted of congeners 28, 66, and 74; group B consisted of congeners 99, 118, 138, 146, 153, 167, 183, and 187; and groups C1 and C2 comprised highly chlorinated congeners (157, 156, 170, 177, 178, 180, 189, 194, 196, 199, 206, and 209). Furthermore, an observation indicated that older individuals manifested a greater incidence of highly chlorinated PCBs (39). Similarly, Megson et al. (40) employed PCA to examine the distribution of PCB congeners based on their degree of chlorination. Group 1 consisted of low-chlorinated congeners (28, 44, 49, 52, 66, 87, 101, 110, 149, and 151), predominantly found in younger subjects. Group 2 comprised highly chlorinated congeners (156, 170, 172, 180, 194, 195, 199, 203, 206, and 209). In a study by Plaku-Alakbarova et al. (41), the PCA results demonstrated the division of PCB congeners into various groups based on the number and position of chlorine atoms. Groups 4, 7, and 1 contained tri/tetra-, tetra/penta-, and hexa/hepta-chlorinated PCBs with chlorine atoms at the 4,4′ positions. Groups 2, 8, and 5 contained tetra- to hexa-, hexa/hepta-, and hepta/octa-chlorinated PCBs with chlorine atoms at the 2,2′ positions.

Several regions in the world have experienced significant PCB contamination due to industrial activities, such as Anniston, Alabama, United States (42); Grimsby, Ontario, Canada (43); Lanzhou, China (44); or Thane Creek, India (45). The next notable source of PCB exposure was the Chemko Strážske plant in eastern Slovakia. The plant produced PCBs utilized as insulating liquids and plasticisers in paints, adhesives, and seals. Between 1959 and 1984, the total production of PCBs at the Chemko Strážske plant reached 21,500 tons, and a significant amount of hazardous waste from this production is still being stored. Furthermore, a substantial quantity of PCBs was released into the environment during manufacturing, leading to extensive contamination of water, soil and living organisms in eastern Slovakia (26, 32). To investigate the relationship between PCB components and residency location, we calculated a residence score for the study subjects based on their residence. We devised an original approach (Figure 1), allowing us to assess the exposure to PCBs related to the locality of residence not only according to the proximity to Chemko Strážske but also by considering other factors such as the distance from Poša sludge bed, the direction of wind flow, the stream of Laborec river and the relief of Vihorlat mountains, that can significantly modulate the level of exposure. In such a way, we could more accurately evaluate the impact of these multifaceted elements on PCB exposure. The results demonstrated a positive association between the PCB components and the residence score, indicating that individuals residing in closer proximity to Chemko Strážske (considering also environmental factors mentioned above) showed higher levels of PCBs. In addition, the score plots (Figure 6) showed that subjects with higher scores for residence location were predominantly exposed to PCB congeners from the PC1 and PC3 groups. Conversely, subjects with lower scores for residence location were exposed to PCB congeners from the PC2 group. We assume that subjects living near the Chemko plant had higher levels of PCBs with highly chlorinated congeners from PC1, affected by natural sources, such as river flow, the prevailing direction of wind currents, and the mountainous terrain. Subjects living far from the Chemko plant had higher levels of PCBs with low chlorinated congeners from PC2. The volatility of low chlorinated congeners of PCBs can explain this. Therefore, they can reach a greater distance (28, 46). Previous research has indicated that areas near the Chemko Strážske plant in eastern Slovakia exhibit higher levels of pollution compared to more distant regions (28, 29). Years ago, a team of researchers from Slovakia analysed air, water, soil, and sediment samples in the area of Chemko Strážske. The results indicated substantially heightened PCB levels in the vicinity of the Chemko Strážske region, especially near the Laborec river, which receives effluent discharge from the plant, as well as within the Zemplínska Šírava reservoir. Research also pointed out that these elevated concentrations impacted aquatic life and wildlife in rivers and forests (26, 47).

Prior investigations in eastern Slovakia have unveiled that domestically produced food items, particularly meat and eggs, exhibit elevated levels of PCB contamination when juxtaposed with commercially available products (32). Studies from highly contaminated areas with PCBs, such as Anniston in Alabama (USA) and Lanzhou (China), confirmed the findings from Slovakia that contaminated food, mainly fish, meat, and eggs, can be a source of PCB exposure for residents living near industrial areas (42, 44). Building upon this knowledge and previous research findings, we evaluated our subjects’ consumption behavior to the concentrations of PCB congeners. We also assigned each study participant a food origin score and consumption frequency score based on the questionnaire. Those who consumed homemade food products and/or consumed with higher frequency received a higher score, while those consuming retail products and/or consumed with lower frequency received a lower score. The hypothesis stating that homemade food products and high food consumption frequency are associated with higher levels of PCBs in our subjects was not confirmed by our results. On the contrary, we did observe statistically significant negative associations between the ΣPC1, ΣPC2 and the consumption frequency score or food origin score among the subjects. Our results are not in line with the previous research on the population from eastern Slovakia (28); there was no significant association between levels of PCBs in human blood serum and the origin of food. On the contrary, studies by Donato et al. (48) and Helmfrid et al. (49) reported a positive association between PCB exposure and food consumption frequency of probands from highly polluted areas in Italy and Sweden, respectively.

Based on the findings obtained through stepwise multivariate regression analysis, the place of residence could impact PCB exposure more than food origin or food consumption frequency in subjects residing in eastern Slovakia. This suggests that inhalation of contaminated air might be a potential route of PCB exposure, as was shown in previous studies (28, 29, 43). While PCBs are commonly recognised for their tendency to bioaccumulate in the food chain, they can also enter the air through various industrial processes, combustion, and volatilisation from contaminated sources. Once released into the atmosphere, PCBs can be transported over long distances and eventually settle on surfaces such as soil and water. Exposure to airborne PCBs can occur in different scenarios. Individuals residing near industrial facilities involved in the production, usage, or disposal of PCBs may experience heightened levels of PCB-contaminated air. Furthermore, regions with a history of PCB contamination, such as hazardous waste sites or heavily industrialized areas, may exhibit elevated concentrations of airborne PCBs (28, 29, 50, 51).

Our findings showed that age is a significant factor in PCB levels in the serum of subjects. Our study aligns with previous research by Jursa et al. (25), Megson et al. (40), and Pavuk et al. (42) confirming a positive association between PCB levels and the age of the study subjects. This association can be attributed to two main factors: bioaccumulation of PCBs in the human body and slower metabolic activity in older individuals (52, 53). The historical use and accumulation of PCBs in the environment is a primary reason for higher PCB levels in older individuals. PCBs were extensively used in various industrial applications until their restrictions or bans. This led to the widespread release of PCBs into the environment through industrial processes and waste disposal. As a result, older generations had greater opportunities for exposure to PCBs over a more extended period. PCBs resist environmental degradation and persist in the environment, allowing them to bioaccumulate in organisms, including humans, through the food chain (1, 2). The second factor contributing to higher PCB levels in older individuals is slower metabolic activity. Research indicates that older humans metabolize PCBs more slowly than younger individuals (53–56). Metabolism of PCBs involves enzymatic processes, including oxidation and conjugation, which facilitate their breakdown and elimination from the body. Age-related changes in liver function, such as reduced enzyme activity, can contribute to a slower metabolism of PCBs in older individuals. Certain enzymes involved in PCB metabolism, such as cytochrome P450 enzymes, may exhibit decreased activity with age. These enzymes are crucial in transforming PCBs into more easily excreted metabolites (52, 53). The slower metabolism of PCBs in older individuals results in prolonged exposure to these compounds and an increased risk of accumulation in the body. In summary, the higher levels of PCBs in older individuals can be attributed to the historical accumulation of PCBs in the environment and the slower metabolic activity associated with aging. This interplay of factors fosters the enduring presence and accumulation of PCBs within the bodies of older people (53–56). Moreover, score plots for PCs according to age groups (Figure 7B) showed that older subjects had higher concentrations of highly chlorinated congeners of PCBs compared to the younger subjects. We confirmed the findings of Megson et al. (40) and Pavuk et al. (39) which can be explained by the longer persistence of highly chlorinated congeners of PCBs in the human body (46, 57).

The strengths of our study lie in the clear PCB exposure scoring methodology, which incorporates a comprehensive scoring system for food origin, frequency of consumption, and residential location of subjects. The development of a map that accounts for multiple environmental and geographic factors in determining the residential score represents an innovative approach as well.

It is important to note that our study has some limitations related to the recruitment strategy, potentially impacting the generalizability of the study results. Selection bias might have occurred by recruiting subjects solely from outpatient clinics. These individuals could differ systematically from the general population. Recall bias might have occurred when filling out the questionnaire, and subjects did not remember their consumption behavior in detail. Non-response bias may be present, as subjects who chose not to participate in our study could differ systematically from those who did (58). These factors could limit the generalizability of the study results.

Our study provides insights into the origins, pathways, and determinants of PCB exposure, with a focus on the population of eastern Slovakia, though its implications are also applicable for other highly polluted regions in the world. By analysing PCB congeners in detail, we identified the significant role of residence proximity to contamination sources, age, and lifestyle factors like smoking and alcohol consumption in influencing PCB levels. Our novel mapping approach, which considers environmental factors such as wind direction, river flow, and mountainous terrain, offers a more precise framework for assessing exposure, applicable to other regions worldwide facing PCB contamination.

The findings underscore the global relevance of residency location in determining PCB exposure levels, as individuals near industrial contamination sites, such as the Chemko Strážske plant, are particularly vulnerable to higher concentrations of highly chlorinated PCBs. This pattern is not unique to Slovakia and mirrors trends observed in other heavily contaminated regions around the world, including the USA, China, and other industrialized areas. Furthermore, our study suggests that inhalation of contaminated air could be a critical exposure pathway for PCB congeners in many parts of the world.

Given the persistence of PCBs in the environment and their ability to bioaccumulate, these results highlight the need for global, knowledge-based management strategies to address historical contamination, mitigate exposure, and protect vulnerable populations. The data generated in this study not only enhance our understanding of PCB distribution in a local context but also serve as a foundation for broader public health interventions aimed at reducing PCB exposure and its associated health risks worldwide.