Reference values of per- and poly- fluoroalkyl substances concentration in serum and related factors in Korean adults: Korean National Environmental Health Survey
Article information
Abstract
This study determined reference values of per- and polyfluoroalkyl substance (PFAS) exposure in the general Korean population. Serum samples from 2,993 adults in the fourth Korean National Environmental Health Survey (KoNEHS) (2018 –2020) were analyzed for five PFAS: perfluorooctanoic acids (PFOA), perfluorooctane sulfonic acid (PFOS), perfluorohexane sulfonic acid (PFHxS), perfluorodecanoic acid (PFDA), and perfluorononanoic acid (PFNA). The geometric means (GMs) and 95th percentile concentrations of serum PFOA were 6.43 and 16.55 μg/L, respectively; those of PFOS were 15.07 and 43.96 μg/L; 4.17 and 14.91 μg/L for PFHxS; 2.06 and 5.98 μg/L for PFNA; and 0.91 and 2.40 μg/L for PFDA. Higher serum PFAS concentrations were observed in older adults, men, former smokers, and frequent seafood consumers. Exposure levels also varied based on socioeconomic factors such as income and education. Additionally, participants residing in coastal areas exhibited higher serum PFAS concentrations, whereas higher PFHxS levels were observed in those living near industrial complexes. Higher concentrations of PFDA and PFNA were detected in participants consuming local drinking water (GMs, 3.29, 2.86 and 2.82 μg/L for local-based water, tap water and purifier or mineral water for PFNA; 1.43, 1.22 and 1.20 μg/L for PFDA; p-values were <0.05). These findings suggest that the Korean PFAS exposure level is relatively high, and may be related with residential and lifestyle characteristics.
Introduction
Today, individuals are exposed to a wide range of chemicals from various sources, including food, water, consumer products, clothing, cosmetics, medicines, and environmental media. Such exposures can have a significant impact on health, hence making the monitoring of chemical exposure vital for the development of effective environmental and health policies. Perfluoroalkyl and polyfluoroalkyl substances (PFAS) are environmental pollutants of particular concern due to their high persistence, bioaccumulation, and toxicity, and managing them is crucial for public health [1-3]. In the general population, exposure to PFAS can occur through food and packaging materials [4-9], consumer or industrial products such as cosmetics, cookware, furniture, cleaning goods, firefighting forms, and carpets [8-12], and environmental media, including indoor and drinking water [4-6, 11, 13]. Recent epidemiological and experimental studies have strengthened the evidence linking various health effects, including cancer to PFAS exposure [14-15]. Therefore, monitoring PFAS levels in the general population is essential, and several countries have incorporated PFAS into their biomonitoring programs [16-18].
Reference value (RV) is defined a value derived from a series of corresponding measured values of a random sample from a defined population group according to a specified statistical procedure [19]. RV provides a representative value of exposure to environmental pollutants in a defined population and upper RV(RV95) indicates an assessment level that is rarely exceed in the reference population [19]. RVs derived from a biomonitoring program offer information of internal exposures of environmental pollutants and scientific evidences for environmental policies [20]. Several national biomonitoring programs, such as US National Health and Nutrition Examination Survey (NHANES), Canadian Health Measures Survey (CHMS), German Environmental Survey (GerES), and Human Biomonitoring for Europe (HBM4EU), have been in operation for a long time, and RVs representing the people of each country are derived from these programs.
The Korean National Environmental Health Survey (KoNEHS) is a national biomonitoring program conducted by the National Institute of Environmental Research (NIER) under the Environmental Health Act [21]. Initiated in 2009, KoNEHS operates on a 3-year cycle. In the fourth cycle (KoNEHS cycle 4), conducted from 2018 to 2020, serum PFAS concentrations were included as analytes [22]. KoNEHS cycle 4 provided RVs for serum PFAS concentrations in the general Korean population. A previous study reported Korean RVs for serum PFAS using data from 2009–2010 [23]. However, given the strengthened regulations following Korea’s adoption of the UN’s Stockholm Convention on persistent organic pollutants, it is essential to update the information on PFAS exposure in the general population. In this study, we present the RVs of serum PFAS concentrations and identify factors related to PFAS exposure in the general Korean population.
Materials and Methods
Study participants
KoNEHS cycle 4 encompassed all age groups, including children aged three and older, adolescents, and adults. However, blood sampling was applied to adolescents and adults, but excluded for children. This study includes serum PFAS concentrations of only adults.
Data collection
This study performed a secondary analysis using raw data from KoNEHS cycle 4, provided by the NIER. The data analysis was conducted according to NIER guidelines. The KoNEHS cycle 4 dataset included a comprehensive questionnaire covering personal information, environmental exposures, clinical tests, and biomonitoring data [22]. For this study, the questionnaire data and biomonitoring data for serum PFAS were used. The KoNEHS cycle 4 questionnaire includes residential characteristics, indoor environment, use of chemical products, dietary habits, socioeconomic statuses, living habits, medicine use, personal components such as smoking, job history, etc., and more. We used the information about residential characteristics, dietary habits, and social economic statuses to find related factors for PFAS exposure, and personal components as confounders to analyze in the statistical models. In this study, variables of income and education represent the social economic status. The income variable refers to monthly income, when <3 million Korean Won (₩) (approximately 2,070 US dollar) it was defined “Low”, “Middle” for 3 – 5 million ₩, and “High” for ≥5 million ₩. The variable of education means the highest schooling. In KoNEHS cycle 4 stratified the whole country to extract representative samples from across the country, defined city, rural, and coastal areas, and allocated and distributed the number of samples. Due to concerns about metal pollution in the ambient air, some areas have operated airborne metal-monitoring systems such as industrial complexes, and we defined this as “metal-polluted area”. In addition, KoNEHS cycle 4 asked the source of household drinking water. If it was from local underground water and/or mineral spring, not metropolitan tap water, commercial mineral bottle water and/or purifier, we classified it “Local-based water”.
KoNEHS cycle 4 was conducted with the approval of the Ethical Review Board of NIER (NIER-2018-BR-003-02), and this study was approved by the Seokyeong University Institutional Review Board (SKUIRB-2024-01-028).
Serum PFAS analysis
KoNEHS cycle 4 included the analysis of five PFAS compounds in serum samples: perfluorooctanoic acid (PFOA, CAS no. 335-67-1), perfluorooctane sulfonic acid (PFOS, CAS no. 1763-23-1), perfluorohexane sulfonic acid (PFHxS, CAS no. 355-46-4), perfluorodecanoic acid (PFDA, CAS no. 335-76-2), and perfluorononanoic acid (PFNA, CAS no. 375-95-1). The serum PFAS analysis was performed by a commercial laboratory (Smartive Co. Ltd., Gyeonggi-do, Korea) assigned by the NIER as a partner of KoNEHS. Serum samples (0.5 mL) were used for PFAS analysis. Following sampling, serum samples were stored under −70℃ until analysis. The protein precipitation method was adopted as a preprocessing step to remove proteins from the serum samples. After adding 0.01 mL of the internal standard material and 0.8 mL of acetonitrile to the sample, the mixture was stirred for 10 seconds to denature the serum proteins. After centrifugation at 13,000 rpm for 10 min, all proteins were sedimented, and 0.8 mL of the supernatant was dispensed. The dispensed sample was concentrated for 15 –20 minutes using N2 gas, and then redissolved in 0.025 mL of 50% methanol, and used for analysis. Quantitative analysis was performed using high-performance liquid chromatography-tandem mass spectrometry (Q-Sight Triple Quad, Perkin-Elmer, Waltham, MA, US) with a C18 column (2.1×100 mm).
Calibration curves were constructed using six standards, with an R2 value greater than 0.995 in all tests. Accuracy and precision were assessed through internal quality assurance and control (QAQC) schemes. NIST (National Institute of Standards and Technology) 1957 and 1958 were used for matrix reference materials in accuracy tests in the internal QAQC. Three concentrations of QC samples were applied in every batch. For external QAQC, Smartive participated in the NIER QAQC program for KoNEHS and the German External Quality Assessment Scheme (G-EQUAS), successfully passing all evaluations.
Statistical analysis
In this study, statistical analysis was divided into two. The first is descriptive statistics for deriving RVs, applied geometric means (GMs) and survey-weighting. We defined RVs as means representing the concentrations in serum samples of all investigated population and derived as representative values of whole Korean population. In addition, RV95 refers to the guidance value that distinguished high exposure groups. Given that serum PFAS concentrations showed a right-skewed distribution, GMs were used as representative values. Thus, the GMs of serum PFAS concentrations from KoNEHS cycle 4 are presented as RVs for serum PFAS in Korean adults. These RVs were calculated with survey-weighting, because KoNEHS samples were designed to calculate an official national reference value, KoNEHS surveyed based on a two-stage proportionally stratified sampling design [24]. The RVs obtained in this study were compared to those reported in the United States and Germany [25-26]. To determine high exposure groups, this study referred two guidance values – 95th percentile from the NHANES and Human Biomonitoring (HBM) values of the German Human Biomonitoring Commission [25-26]. HBM values are derived from health-related guidance values and used as advices for planning and execution of related research and policies of environmental health [26].
The second is analytical statistics to find out factors related to PFAS exposure of Korean general population. To compare serum PFAS concentrations across different general characteristics of KoNEHS participants, Student’s t-test and analysis of variance (ANOVA) were used. In addition, covariate-adjusted means were calculated using analysis of covariance (ANCOVA).
Descriptive statistics analyses for deriving used log-transferred values using “survey package” of R version 4.3.2 (R Foundation for Statistical Computing, Vienna, Austria). Analytical statistics including analyses used the same program with statistical significance set at p < 0.05.
Results
Study participants
Serum PFAS analysis in KoNEHS cycle 4 involved 2,993 participants (1,298 male; 1,695 female), representing a total of 43,863,499 Korean adults aged 20 years and older (21,848,320 male; 22,015,179 female). Among the participants, 41.3% were in their 40s and 50s, and 64.5% were never smokers. The majority of participants (71.6%) resided in urban areas (Table 1).
General characteristics of KoNEHS participants.
RVs and RV95 of serum PFAS from KoNEHS cycle 4
Table 2 presents the RVs for the concentrations of five serum PFAS compounds from the KoNEHS cycle 4. For PFOA, the GM was 6.43 μg/L, with a 95th percentile value of 16.55 μg/L, which is higher than the 4.27 μg/L reported in the United States. Notably, 81.9% and 25.9% of participants exceeded the 95th percentile of the United States NHANES (4.27 μg/L), and the HBM value (10 μg/L), respectively. The GM and 95th percentile of serum PFOA concentrations were 15.07 and 43.96 μg/L, respectively, with 42.7% of participants exceeding the 95th percentile of the United States NHANES (19.1 μg/L). Additionally, 36.9% and 30.4% of male and female participants exceeded HBM, respectively. For PFHxS, PFNA, and PFDA, which lack suggested HBM values, the GMs were 4.17, 2.06, and 0.91 μg/L, respectively, with corresponding 95th percentiles of 14.91, 5.98, and 2.40 μg/L, respectively. Between 42.6% and 74.7% of participants had levels exceeding the 95th percentiles of the United States NHANES.
Serum PFAS concentrations of KoNEHS participants and comparisons with other reference values (μg/L, n=2,993).
Serum PFAS concentrations by general characteristics
Serum concentrations of all five PFAS compounds were significantly higher in male participants than those of female participants (GMs of female and male: 6.74 and 7.43 μg/L for PFOA; 15.71 and 18.15 μg/L for PFOS; 4.00 and 5.18 μg/L for PFHxS; 2.21 and 2.60 μg/L for PFNA; 0.98 and 1.11 μg/L for PFDA). Additionally, PFAS concentrations increased with age (GMs of 2–30s, 4–50s, and ≥60s: 4.56, 7.19, and 8.75 μg/L for PFOA; 9.83, 16.00, and 23.71 μg/L for PFOS; 2.97, 4.56, and 5.53 μg/L for PFHxS; 1.30, 2.38, and 3.32 μg/L for PFNA; 0.63, 1.02, and 1.38 μg/L for PFDA).
Serum PFAS concentrations were lower in “never-smokers” compared to “ex-“ and “current-smoker” (GMs of never-, ex-, and current-smokers: 6.79, 7.86, and 7.03 μg/L for PFOA; 16.13, 19.76, and 15.72 μg/L for PFOS; 4.10, 5.37, and 5.10 μg/L for PFHxS; 2.24, 2.89, and 2.38 μg/L for PFNA; 0.98, 1.24, and 1.00 μg/L for PFDA). Participants who frequently consumed seafood also had significantly higher PFAS concentrations, except for PFHxS (GMs of never-, monthly-, weekly-, and daily-intake: 5.47, 6.31, 7.73, and 9.44 μg/L for PFOA; 12.65, 14.88, 18.36, and 25.57 μg/L for PFOS; 3.94, 4.20, 4.69, and 5.60 μg/L for PFHxS; 1.67, 2.07, 2.68, and 3.61 μg/L for PFNA; 0.74, 0.91, 1.15, and 1.53 μg/L for PFDA). All p-values were <0.001, except for PFHxS in relation to seafood intake (Fig. 1 and Table S1).
Comparisons of serum PFAS concentrations by general characteristics of KoNEHS participants (A) For sex; (B) for age; (C) for smoking status; (D) for frequencies of seafood intake
PFOA, perfluorooctanoic acid; PFOS, perfluorooctane sulfonic acid; PFHxS, perfluorohexane sulfonic acid; PFNA, perfluorononanoic acid; PFDA, perfluorodecanoic acid
Never, never or almost never; Monthly, 1-2 in a month; Weekly, 1-2 in a week; Daily, almost every day for seafood intake. y-axis of all graphs mean serum concentration of each PFAS (μg/L). All cases were statistically significant by Student’s t-test or ANOVA (p<0.05 or p<0.01) except PFHxS in relation to seafood intake (p=0.549).
Factors associated with serum PFAS concentrations
Comparing serum PFAS concentrations in relation to socioeconomic status (SES) revealed that higher household monthly income was correlated with increased serum PFOA concentrations, although this relationship was not statistically significant (p =0.070). In contrast, after adjusting for confounders such as sex, age, pack year, and seafood intake frequency, serum concentrations of PFOS, PFHxS, PFNA, and PFDA were significantly higher among participants with lower educational attainment (Table 3).
Comparisons of serum PFAS concentrations by social economic statues and living environment of KoNEHS participants.
Participants residing in coastal areas exhibited the highest serum concentrations of PFOA, PFOS, PFNA, and PFDA. In contrast, those living in urban industrial areas showed the highest concentrations of PFHxS. In addition, those who consumed local water sources had the highest serum concentrations of PFNA and PFDA.
Discussion
This study, providing the first representative values (RVs) and guidance values (RV95) of serum PFAS concentrations in the Korean general population via the KoNEHS, revealed that PFAS concentrations in Korean population are notably high, with many participants exceeding the HBM guidelines. The RVs for PFAS concentrations in the serum of Korean adults were higher than those reported in other national and regional surveys, including the NHANES, HBM4EU, CHMS, GerES, Flemish Environment and Health Study (FLEHS), and other studies [18, 27-32]. While differences in analytical methods could partly explain these discrepancies, the fact that our study passed both the international and domestic external QA programs (G-EQUAS and NIER’s QAQC program) suggests that the differences in serum concentrations were not merely due to differences in the analytical process. This study showed that serum PFAS concentrations in the Korean population were higher than those in other countries. This elevated PFAS exposure in the Korean population may be attributed to manufacturing-based industrial structure, particularly in sectors such as semiconductors and automobiles [33]. This industrial structure contributes to both occupational and environmental exposure to PFAS. Furthermore, high PFAS exposure may be due to the increased use of contact lenses in Koreans [34-35].
Serum PFAS concentrations have been shown to increase with age, sex (male), smoking (ex-smoker), and seafood consumption, as reported in previous studies worldwide [17, 27, 29, 36-39]. Interestingly, this study found that ex-smokers had higher PFAS exposure levels than current smokers. Smoking may be related with many environmental pollutants not only for PFAS, some studies have also reported higher PFAS concentrations among the non-smoking group, indicating that the relationship between smoking status and PFAS levels remains unclear [40-41].
SES is another important factor influencing exposure to chemicals like PFAS as already reported in several studies, however [39, 42], this study did not find any relationship. The relationship between education level and PFAS levels remains unclear [42], but, a higher education level may generally reduce chemical exposure and interpret health-related information [42].
In this study, residents living coastal areas have higher serum PFAS concentrations. And this study also indicated that high PFAS exposure levels may be related to seafood consumption. In European biomonitoring studies, serum PFAS levels increased with higher consumption of fish and seafood [18, 39, 43]. Known as “forever chemicals,” PFAS can accumulate in marine ecosystems, making seafood consumption a significant contributor to PFAS exposure in the general population [44-45]. This result is thought to be related to the persistence of PFAS, and higher levels of PFAS with increasing age can be observed in this context. In contrast, the PFHxS concentration was the highest in the urban industrial area, whereas the other four compounds were the highest in the coastal area. Recently, the use of PFHxS has increased as a substitute for PFOA and/or PFOS [46]. PFHxS, a perfluorosulfonic acid (PFSA) with seven or fewer fluorinated carbons [47], and primarily detected in wastewater discharged from industrial complexes [46, 48]. In Korea, PFHxS, along with PFOA and PFOS, has been included in tap water quality inspections since 2018, following a rise in PFHxs concentrations at filtration plants near industrial complexes [49]. This study may suggest that PFHxS exposure is more likely driven by industrial pollution in drinking water rather than by seafood consumption. Although not statistically significant, the highest PFHxS concentration was found in participants using tap water as their primary drinking source. Of course, there is a limitation for this assumption in this study itself. The population consuming local water, sourced from mineral springs and groundwater, exhibited the highest concentrations of PFDA and PFNA. A previous study reported that high concentrations of PFDA were detected in subsoil samples [50]. The concentration of PFDA in the soil is generally high, and both PFNA and PFOA concentrations increase in the soil after rainfall [51]. Thus, populations relying on local water sources are likely affected by PFDA and PFNA contamination in the soil.
Establishing strong scientific evidence for the relationships between PFAS exposure in this study is challenging due to the limitations inherent in national biomonitoring, which is designed as a cross-sectional study. Furthermore, this study did not adequately consider various factors related to PFAS exposure, such as diet, consumer products, clothes, and occupational exposures.
Despite these limitations, this study suggests that PFAS exposure levels in the Korean general population are high and are related with residential environment. Therefore, PFAS exposure can be a community-based environmental health issue as we already experienced [52].
Conclusions
This study highlights the PFAS exposure levels in Korean adults, using data from a national biomonitoring program that accurately represents the general population. In addition to other known factors, seafood intake, residence, and drinking water intake may be related to PFAS exposure in the general population.
Serum PFAS concentrations by general characteristics
Notes
Acknowledgement
This work was supported by the Korea Environment Industry & Technology Institute (KEITI) through the Core Technology Development Project for Environmental Diseases Prevention and Management, funded by the Korea Ministry of Environment (MOE) [Grant No. 2480000101/ RS-2021-KE001378].
Conflict of interest
The authors declare that they have no known competing financial interests or personal relationships.
CRediT author statement
YMC: Conceptualization, Writing-Original draft preparation, Writing-Review & Editing, Methodology, Software, Visualization, Supervision, Project administration, Funding acquisition. DH: Software, Formal analysis. JJ: Formal analysis, Writing-Original draft preparation. KHC: Writing-Review & Editing. YSH and WJK: Investigation.
Supplementary Material
This material is available online at www.eaht.org.