Science of the Total Environment 777 (2021) 146013 Contents lists available at ScienceDirect Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv Reductions in blood concentrations of persistent organic pollutants in the general population of Barcelona from 2006 to 2016 Miquel Porta a,b,c,⁎, José Pumarega a,b,c,1, Luis A. Henríquez-Hernández d,e,1, Magda Gasull a,b,c,1, Xavier Bartoll f, Juan P. Arrebola c,g,h, Eva Morales c,i, Jesús Ibarluzea c,j,k,l, Juan Alguacil c,m, Usama Bilal n,o, Octavio P. Luzardo d,e a Hospital del Mar Medical Research Institute (IMIM), Barcelona, Spain b School of Medicine, Universitat Autònoma de Barcelona, Barcelona, Spain c CIBER de Epidemiología y Salud Pública (CIBERESP), Madrid, Spain d Toxicology Unit, Research Institute of Biomedical and Health Sciences (IUIBS), Department of Clinical Sciences, Universidad de Las Palmas de Gran Canaria, Canary Islands, Spain e CIBER de Obesidad y Nutrición (CIBEROBN), Madrid, Spain f Agència de Salut Pública de Barcelona, Spain g Instituto de Investigación Biosanitaria Ibs.-Granada, Granada, Spain h Department of Preventive Medicine and Public Health, University of Granada, Spain i IMIB-Arrixaca, Department of Public Health Sciences, University of Murcia, Spain j Basque Government, Subdirectorate for Public Health and Addictions of Gipuzkoa, San Sebastián, Spain k Faculty of Psychology, University of the Basque Country, San Sebastián, Spain l Biodonostia Health Research Institute, Environmental Epidemiology and Child Development Group, San Sebastián, Spain m Universidad de Huelva, Huelva, Spain n Urban Health Collaborative, Drexel Dornsife School of Public Health, Philadelphia, PA, USA o Department of Epidemiology and Biostatistics, Dornsife School of Public Health Drexel University, Philadelphia, PA, USA H I G H L I G H T S G R A P H I C A L A B S T R A C T • Concentrations of the nine most preva- lent POPs decreased markedly during the 10-years in almost all sex, age and BMI subgroups. • A relevant component of the success is a reduction of differences (convergence) by gender. • For some POPs the decrease was larger in the younger groups. • Reductions in PCBs and DDE were ob- served in individuals with normal weight but not in individuals with obe- sity. • Barcelona is one of few cities that bio- monitor POP levels through periodic health surveys representative of the population. Abbreviations: BHS, Barcelona Health Survey; BMI, body mass index; CI, confidence interval; DDD, dichlorodiphenyldichloroethane; DDE, dichlorodiphenyldichloroethene; DDT, dichlorodiphenyltrichloroethane; GLM, General Linear Model; HCB, hexachlorobenzene; HCH, hexachlorocyclohexane; LOD, limit of detection; LOQ, limit of quantification; OCs, organochlorine compounds; OCPs, organochlorine pesticides; PAHs, polycyclic aromatic hydrocarbons; PBDEs, polybrominated diphenyl ethers; PCBs, polychlorinated biphenyls; POPs, persistent organic pollutants; TL, total serum lipids. ⁎ Corresponding author at: Hospital del Mar Medical Research Institute (IMIM), Universitat Autònoma de Barcelona, Carrer del Dr. Aiguader 88, E-08003 Barcelona, Catalonia, Spain. E-mail address: mporta@imim.es, Twitter: @miquelporta (M. Porta). 1 These authors contributed equally. https://doi.org/10.1016/j.scitotenv.2021.146013 0048-9697/© 2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). M. Porta, J. Pumarega, L.A. Henríquez-Hernández et al. Science of the Total Environment 777 (2021) 146013 a r t i c l e i n f o a b s t r a c t Article history: Background: Few cities in the world biomonitor changes in human levels of persistent organic pollutants (POPs) Received 22 December 2020 through periodic health surveys representative of the general population. Received in revised form 17 February 2021 Objectives: To analyze changes in serum concentrations of POPs in Barcelona from 2006 to 2016, and to analyze Accepted 17 February 2021 socio-demographic correlates of concentrations of 62 POPs in 2016. Available online 2 March 2021 Methods: Participants in the Barcelona Health Surveys of 2006 and 2016 (N = 231 and 240, respectively) were interviewed face-to-face, gave blood, and underwent a physical exam. POPs were analyzed by gas chromatogra- Editor: Adrian Covaci phy - mass spectrometry. Results: Concentrations of all nine most prevalent compounds decreased markedly during the 10-years . Reduc- Keywords: tions occurred in almost all sex, age and BMI subgroups. For most organochlorine compounds the reduction was Human biomonitoring larger in women than men (for HCB, −77% and − 62%, respectively). For β-HCH, PCBs, naphthalene and phen- Health survey anthrene the decrease was larger in the younger groups. Large reductions in concentrations of PCBs and DDE Persistent organic pollutants (POPs) were observed in individuals with normal weight but not in individuals with obesity. Pesticide residues Conclusions: While concentrations of most POPs are decreasing in Barcelona, significant sociodemographic differ- Environmental pollutants ences in such reductions warrant strengthening public and private policies towards groups making slower prog- Environmental exposure/human samples ress. A relevant component of the success in the current decreasing is a reduction of differences (convergence) by Polybrominated diphenyl ethers (PBDEs) Polycyclic aromatic hydrocarbons (PAHs) gender. © 2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 1. Introduction In addition to the organochlorine pesticides (OCPs) and polychlorinated biphenyls (PCBs) analyzed in the 2006 BHS study on A solid body of evidence shows that persistent organic pollutants serum concentrations of POPs, the BHS of 2016 also included (POPs) are highly lipophilic and degradation-resistant synthetic polybrominated diphenyl ethers (PBDEs) and polycyclic aromatic hy- chemicals that bioaccumulate in the environment, food webs and living drocarbons (PAHs). PBDEs are POPs used as flame retardants, and organisms. Exposed to such compounds throughout life, mostly from some of them are banned (UNEP - United Nations Environment the ingestion of animal fats, virtually all humans store POP mixtures in Program, 2020). PAHs are not banned; they are produced by natural adipose tissues –at highly different concentrations. Production and use and anthropogenic pyrogenic processes. The continuous presence of of numerous POPs have been periodically banned or restricted during PAHs in the environment supports their consideration as POPs (EEA - past decades. However, large differences in the decline or stabilization European Environment Agency, 2019). of environmental and human levels of the pollutants have been ob- Therefore, the aim of the present study was two-fold: first, to ana- served across populations, largely due to their physico-chemical and lyze changes in serum concentrations of POPs in the non- economic persistence. Today, human contamination and health effects institutionalized adult population of Barcelona city from 2006 to 2016; from POPs remain significant (Gore et al., 2015; Henkler and Luch, and second, to analyze socio-demographic correlates of concentrations 2011; Kassotis et al., 2020; LaMerrill et al., 2020; Lee et al., 2014; of 62 POPs in 2016. Marraudino et al., 2019; Peinado et al., 2020). Biomonitoring POPs and other environmental chemical agents is thus 2. Methods relevant because throughout the lifecourse and history individuals and societies experience regular changes in their contamination (Demeneix, 2.1. Study population and health interview survey 2017; Trasande, 2019; Porta, 2018; Henríquez-Hernández et al., 2020; Porta et al., 2008). Documenting such changes by sociodemographic fac- Methods of the Barcelona Health Survey (BHS) have been described tors, for instance, can identify groups where regulation is less effective to in detail (Bartoll et al., 2018; Gasull, 2019; Porta et al., 2012a). Briefly, prevent exposure and, thus, in whom private and public policies need the BHS is a health survey that is representative of the adult non- strengthening (Porta, 2004). When biomonitoring shows that POP body institutionalized population of the city of Barcelona. At the end of the concentrations decrease, we can see the beneficial effects of such policies 2006 and 2016 BHS interviews, participants ≥15 years old were offered (Henríquez-Hernández et al., 2020; Porta et al., 2008; Porta, 2004; to take part in the study on POPs (Porta et al., 2009); the youngest per- Angerer et al., 2007; Schulz et al., 2011; Bjerregaard-Olesen et al., 2017; sons of the 2006 and 2016 BHS who actually participated in the POP Buekers et al., 2018; Calafat, 2016; Joas et al., 2017; Dennis et al., 2017; study were 18 and 19 years old, respectively. Subsequently, a nurse per- Haines et al., 2017; Li et al., 2020; Nøst et al., 2019; Schoeters et al., sonally interviewed again each person who accepted to participate in 2017; Schröter-Kermani et al., 2013; Stubleski et al., 2018). the POP study, measured the weight, height, and the hip and waist cir- Often guided by global initiatives as the Stockholm convention cumference, and collected blood and urine samples. Participants had and similar frameworks, an increasing number of countries and been asked to fast for at least eight hours before blood extraction. As some cities in the world produce valid analyses of trends in POP in 2006, in 2016 blood was collected in a vacuum system tube and cen- levels based on studies representative of the general population trifuged for 15 min × 3000 rpm at 4 °C to obtain serum. Right after cen- (UNEP - United Nations Environment Program, 2020; Porta and trifugation, serum was divided in 1–3 mL aliquots and stored at −80° Zumeta, 2002). In some instances, analyses are integrated within until 2018–2019, when POP concentrations were analyzed. The addi- health surveys, which is a strength from scientific and policy stand- tional interview for the POPs study included structured questions points (Porta et al., 2008). Barcelona is one of the few such cities. about recent and past changes in body weight (last 6 months), and on The Barcelona Health Survey (BHS) conducted by the Barcelona whether the person had been breastfed. Women were also asked Agency for Public Health is the oldest of such surveys in Spain, about parity, breastfeeding their children, and abortion histories. with more editions and scientific productivity (ASPB, 2020). It is The average age of the 240 participants of the 2016 BHS was also the only study in the country based on representative samples 50 years, and 53% were women (Supplemental Table 1). Body mass of the general population in which blood has been collected on two index (BMI) was computed by measured weight [kg] divided by mea- time points to analyze changes in biomarkers of environmental sured height squared [m2]. 16% of participants were obese contaminants as POPs. (BMI ≥ 30 kg/m2), 85% were breastfed in childhood, 23% were born 2 M. Porta, J. Pumarega, L.A. Henríquez-Hernández et al. Science of the Total Environment 777 (2021) 146013 outside Spain, 90% had completed at least primary schooling, and 31% 2.2.1. Sample preparation and instrumentation were from occupational social classes IV or V (Supplemental Table 1). Half-milliliter aliquots of serum samples were mixed with 0.4 mL of While occupational and educational indicators were slightly more fa- water/n-propanol (85:15) and subsequently centrifuged at 3000 rpm vorable for women, there were no significant differences in the distribu- for 5 min. Then, 0.1 mL of acetic acid was added to each sample and tion of age, having been breastfed in childhood, birth place, educational loaded to 200 mg (3 mL) Chromabond® C18ec columns (Macherey- level, social class and employment status by sex. However, a slightly Nagel, Germany) mounted in a vacuum manifold (Waters Corporation, greater proportion of women than men had a normal weight (47% vs. USA). The columns were previously conditioned with 2 × 1 mL metha- 41%, respectively, p < 0.09). The most relevant differences between par- nol followed by 2 × 1 mL isopropanol:water (15:85). After passing the ticipants in the 2016 BHS and the 231 participants in the 2006 BHS samples the columns were washed with 1 mL of isopropanol:water (Porta et al., 2012a; Porta et al., 2009) were birth place, social class, ed- (15:85), and to a drying thereof under vacuum for 30 min. Finally, the ucational level, and employment status (Supplemental Table 1). There- analytes were eluted with 1 mL of dichloromethane. Briefly, we fore, such differences will subsequently be controlled through employed a Gas Chromatography (GC) System 7890B equipped with a multivariate adjustment (Section 2.3. Statistical analyses). The Ethics 7693 Autosampler (Agilent Technologies, Palo Alto, CA, USA) for gas Committee of the Parc de Salut Mar approved the study protocols, and chromatographic separations. Two fused silica ultra-inert capillary col- all participants signed an informed consent before completing question- umns Agilent J&WHP-5MS (Crosslinked 5% phenylmethylpolysiloxane, naires at baseline (Porta et al., 2009). Agilent Technologies) each with a length of 15 m, 0.25 mm i.d., and a film thickness of 0.25 μm were connected in series and used as the sta- tionary phase. Both columns were connected by a Purged Ultimate 2.2. Analytical chemical methods Union (PUU; Agilent Technologies). Helium (99.999%) at a constant flow rate of 1.0 mL/min for column 1 was used as the carrier gas. A Analyses of serum samples of the 2016 BHS were carried out in the Re- back-flushing technique was incorporated to the GC. The oven temper- search Institute of Biomedical and Health Sciences (IUIBS) in the Univer- ature program was programmed as follows: a) an initial temperature of sity of Las Palmas de Gran Canaria, Spain. Serum concentrations of 62 60 °C held for 1 min; b) increase to 170 °C at a rate of 40 °C/min; c) in- compounds were measured (Supplemental Table 2 and Supplemental crease to 310 °C at a rate of 10 °C/min with 3 min hold time; and d) cool Fig. 1). Of these, 38 were organochlorine compounds (OCs): 20 were or- down to 60 °C. Injector and transfer line were set at 280 °C. Standards ganochlorine pesticides (OCPs): o,p'-dichlorodiphenyltrichloroethane and samples were injected (1 μL) in the splitless mode using a 4-mm (DDT), p,p'-DDT, o,p'-dichlorodiphenyldichloroethene (DDE), p,p'-DDE, o, ultrainert liner with glass wool (Agilent Technologies). The detection p'-dichlorodiphenyldichloroethane (DDD), p,p'-DDD, hexachlorobenzene of the analytes was performed using a Triple Quad 7010 mass spectrom- (HCB), hexachlorocyclohexane (isomers α-, β-, γ-, and δ-HCH), aldrin, en- eter (Agilent Technologies, Palo Alto, CA, USA). The quantification was drin, dieldrin, endosulfan (α-, and β-isomers), endosulfan sulphate, done using point calibration curves, which were constructed using a mirex, methoxychlor, and heptachlor; the OCs analyzed included as well least-squares linear regression from the injection of standard solutions 18 polychlorinated biphenyl (PCB) congeners: marker-PCBs (IUPAC num- ranging from 0.025 to 25 ng/mL. The limits of quantification (LOQ) bers 28, 52, 101, 138, 153, and 180), and dioxin-like PCBs (congeners 77, were 0.03 to 0.1 ng/mL (Supplemental Table 2). The details of validated 81, 105, 114, 118, 123, 126, 156, 157, 167, 169 and 189). Analyses included chromatographic method and quality control have been previously re- also 8 polybrominated diphenyl ethers (PBDEs): the tribrominated conge- ported (Luzardo et al., 2019; Henríquez-Hernández et al., 2017). ner BDE-28, the tetrabrominated BDE-47, the pentabrominated BDE-85, To be conservative, the main statistical analyses were limited to -99 and -100, the hexabrominated BDE-153 and -154, and the compounds that were detected above the detection limit in >75% of heptabrominated BDE-183; and the 16 most environmentally relevant participants of BHS 2016. The limits of detection (LODs) for the 11 com- polycyclic aromatic hydrocarbons (PAHs) listed by the United States Envi- pounds included in the main analyses ranged from 0.009 to ronmental Protection Agency (Keith and Telliard, 1979; Loibner et al., 0.021 ng/mL, and the LOQs from 0.030 to 0.070 ng/mL (Supplemental 2004) (Supplemental Table 2 and Supplemental Fig. 1). Table 2). When the assigned value of an analyte was below the LOD of In serum samples of the 231 participants of the 2006 BHS we ana- IUIBS, it was assigned the mid-value of this limit; and when a POP con- lyzed the 8 PBDEs –congeners #28, 47, 85, 99, 100, 153, 154, and 183– centration was between the LOD and the LOQ, the mid-value between and the 16 PAHs not analyzed in the previous study (Porta et al., LOD and LOQ was used (Porta et al., 2012a; Porta et al., 2009; Porta 2012a) (Supplemental Table 3 and Supplemental Fig. 1): naphthalene, et al., 2010). acenaphthene, acenaphthylene, fluorene, phenanthrene, anthracene, Concentrations of total cholesterol and triglycerides were deter- fluoranthene, pyrene, benzo[a]anthracene, chrysene, benzo[b]fluoran- mined enzymatically, using serum obtained at the same time as the thene, benzo[k]fluoranthene, benzo[a]pyrene, dibenzo[a,h]anthracene, serum used for POP analyses. Total serum lipids (TL) were calculated benzo[ghi]perylene and indeno[1,2,3,-cd]pyrene (Luzardo et al., 2019). by the standard formula 2 of Phillips et al. (Gasull, 2019; Porta et al., Serum samples from the 2006 BHS had been stored at IMIM frozen at 2012a). POP concentrations were individually corrected for TL and are −80 °C, with no incidents. expressed in nanograms of analyte per gram lipid (ng/g of lipid). In addition, in 2018–2019 the laboratory at IUIBS analyzed the con- Among the 240 participants in the 2016 study, mean (standard devia- centrations of the 18 POPs previously analyzed by CSIC in the 231 subjects tion) serum concentrations of total cholesterol, triglycerides and TL from the 2006 BHS (Porta et al., 2012a) (Supplemental Fig. 1). IUIBS did so were 193.4 (40.3), 95.9 (65.5) and 597.3 (132.8) mg/dL, respectively. as a methodological control in samples of 30 (13%) such individuals, since it was not justified to repeat the analyses of these 18 POPs for all 231 sub- 2.2.2. Quality of analyses and quality control (QA/QC) jects (biological samples from a population sample that is representative All measurements were performed in triplicate, and the geomet- of the general population are particularly precious). To compare the con- ric mean was used for the calculations. In each batch of samples, centrations of the compounds analyzed in 2016 and 2006 we computed three controls were included for every 18 vials: a reagent blank by linear regression a conversion factor from values obtained in the consisting of a vial containing only cyclohexane; a vial containing above-mentioned 30 individuals of the 2006 BHS. The conversion factor 2 ng/mL of each of the pollutants in cyclohexane; and an internal lab- was applied to concentrations of the 18 POPs previously analyzed in the oratory quality control sample (QC) consisting of melted butter 231 participants of 2006 BHS (Supplemental Fig. 1 and Supplemental spiked at 10 ng/mL of each of the analytes, which was processed Methods). After applying the conversion factor to the concentrations of using the same method of extraction as the serum samples. The re- 2006, we applied to these concentrations the same LODs and LOQs used sults were considered to be acceptable when the concentration of by IUIBIS to the samples of 2016. the analytes determined in the QC sample was within 15% of the 3 M. Porta, J. Pumarega, L.A. Henríquez-Hernández et al. Science of the Total Environment 777 (2021) 146013 deviation of the theoretical value. Further details on quality of analy- 2016). In the analyses we included the 9 POPs that had been detected ses and quality control (QA/QC) were provided previously (Cabrera- (each) in >70% of the study subjects of 2006 BHS (Table 1). Rodríguez et al., 2019). Generalized Linear Models (GLM) were used to analyze variations in lipid-corrected and log-transformed concentrations of POPs (Kleinbaum et al., 2007). The main effects of all predictors were independently ex- 2.3. Statistical analyses plored in base models (Lash et al., 2021). Sociodemographic differences between participants in the 2016 and 2006 BHS were also controlled Univariate statistics were computed as customary (Kleinbaum et al., with GLM to compute geometric means adjusted for the relevant con- 2007). Kruskal-Wallis' test and Mann-Whitney's U test were used to as- founders. The level of statistical significance was set at 0.05 and all sess differences in concentrations of POPs by sociodemographic charac- tests were two-tailed. The statistical significance, the precision of the es- teristics of the participants. The Kolmogorov-Smirnov test for normality timates, and the magnitude of the effects were all taken into consider- was used to check the distributions of POPs; as none was normal, log- ation to assess significance (Lash et al., 2021; Amrhein et al., 2019). transformed values were used in regression analyses. The selection of Statistical analyses were conducted using SPSS version 22.0.0.0 (IBM a representative sample of the non-institutionalized population in SPSS Statistics, Armonk, NY, USA, 2013), R version 3.5.2 (2018), and each BHS followed a complex design (Bartoll et al., 2018; Gasull, 2019; Stata 15.0. Porta et al., 2012a). In order to guarantee that the demographic struc- ture of the population is preserved with respect to age and gender, and thus the representativeness of the study population with respect 3. Results the general population of Barcelona, sampling weights were used in the statistical analyses of the present study. Concentrations of all nine most prevalent compounds decreased We calculated the number of POPs detected in each person at high markedly during the 10-year study period. This was so for all four concentrations (nPhc) in the two BHS studies as follows: for each subject main measures: lipid-adjusted (ng/g lipid) and crude (ng/ml) concen- we added the number of POPs whose respective serum concentrations trations assessed by medians and by adjusted geometric means, with were equal to or greater than the percentile 75 of each POP distribution most changes from about −30% to −80% or more (all p < 0.05) in the 2006 BHS (Porta et al., 2012a; Porta et al., 2012b; Pumarega et al., (Table 1 and Fig. 1). Table 1 Change in serum concentrations of persistent organic pollutants in the general population of Barcelona, 2006 to 2016. Barcelona Health Survey of 2006 Barcelona Health Survey of 2016 Change 2006–2016 (N = 231) (N = 240) % > LOD (p-value)a Medians (%)b aGM (%)c % > LOD Median (P25-P75) aGM (CI 95%) % > LOD Median (P25-P75) aGM (CI 95%) p,p’-DDE ng/g lipid 323.1 (142.0–977.5) 379.7 (330.3–436.5) 177.9 (98.9–549.3) 226.8 (197.7–260.2) −0.4 −45 −40 ng/ml 100.0 2.13 (0.86–6.53) 2.37 (2.06–2.73) 99.6 1.02 (0.55–3.48) 1.33 (1.16–1.52) (>0.999) −52 −44 HCB ng/g lipid 111.0 (41.5–313.8) 119.6 (108.0–132.4) 25.5 (15.6–70.7) 34.9 (31.6–38.6) +1.2 −77 −71 ng/ml 98.4 0.75 (0.24–2.08) 0.75 (0.68–0.83) 99.6 0.16 (0.09–0.43) 0.20 (0.18–0.23) (0.208) −79 −73 β-HCH ng/g lipid 69.6 (35.9–220.4) 89.8 (80.0–100.8) 17.3 (7.08–52.7) 20.1 (18.0–22.5) −5.2 −75 −78 ng/ml 96.9 0.48 (0.20–1.51) 0.56 (0.50–0.63) 91.6 0.11 (0.03–0.32) 0.12 (0.10–0.13) (0.016) −78 −79 PCB 118 ng/g lipid 12.5 (9.30–27.4) 15.5 (13.7–17.6) 1.71 (1.28–15.5) 3.96 (3.50–4.47) −41.6 −86 −74 ng/ml 82.5 0.06 (0.06–0.17) 0.09 (0.09–0.11) 40.9 0.01 (0.01–0.09) 0.02 (0.02–0.03) (<0.001) −87 −76 PCB 138 ng/g lipid 59.6 (35.0–107.6) 56.9 (51.8–62.6) 40.1 (21.4–74.8) 36.9 (33.7–40.6) −0.3 −33 −35 ng/ml 97.8 0.38 (0.19–0.68) 0.36 (0.32–0.39) 97.5 0.24 (0.12–0.44) 0.22 (0.20–0.24) (>0.999) −38 −39 PCB 153 ng/g lipid 81.0 (48.4–144.2) 77.8 (70.7–85.6) 64.6 (31.6–119.6) 59.0 (53.7–64.9) +1.4 −20 −24 ng/ml 97.8 0.52 (0.27–0.97) 0.49 (0.44–0.54) 99.2 0.38 (0.18–0.70) 0.35 (0.31–0.38) (0.277) −27 −29 PCB 180 ng/g lipid 80.7 (45.3–124.9) 72.1 (64.9–80.1) 64.6 (24.2–121.1) 46.4 (41.8–51.4) −3.0 −20 −36 ng/ml 98.1 0.53 (0.25–0.82) 0.45 (0.41–0.50) 95.1 0.37 (0.13–0.73) 0.27 (0.24–0.30) (0.073) −31 −40 Naphthalene ng/g lipid 294.9 (27.5–444.9) 80.5 (60.5–107.2) 1.71 (1.26–81.3) 8.16 (6.16–10.8) −37.1 −99 −90 ng/ml 82.5 2.03 (0.20–2.61) 0.50 (0.38–0.67) 45.4 0.01 (0.01–0.47) 0.05 (0.04–0.06) (<0.001) −100 −91 Phenanthrene ng/g lipid 115.4 (2.34–206.5) 39.3 (30.7–50.5) 51.4 (6.61–101.4) 26.2 (20.5–33.4) +2.9 −55 −33 ng/ml 73.2 0.70 (0.01–1.26) 0.25 (0.19–0.31) 76.2 0.30 (0.05–0.57) 0.15 (0.12–0.20) (0.459) −57 −38 The Table shows results for the 9 POPs that had been detected (each) in >70% of the study subjects of the 2006 BHS. LOD: limit of detection. % > LOD: percent of subjects with concentrations above the LOD. P25, P75: 25th, and 75th percentile, respectively. aGM: Geometric mean adjusted for age, sex and body mass index. CI: confidence interval. Concentrations of the seven first compounds in the 2006 Barcelona Health Survey (BHS) reflect the conversion factor explained in the Methods section. a Difference in percentage points between the 2006 and 2016 BHS in the percent of subjects with concentrations above the limit of detection (Fisher's exact test, two-tail test). b Percent change in the median of the 2006 and 2016 BHS; the differences between such pairs of medians were statistically significant (p-value <0.05, Mann-Whitney's U test, two-tail) for all 9 compounds: for HCB, β-HCH, PCBs 118, and 138, naphthalene and phenanthrene the p-values were ≤ 0.001. c Percent change in the aGM of the 2006 and 2016 BHS; the differences between such pairs of aGM were statistically significant (p-value ≤0.001, Wald test) for all 9 compounds. 4 M. Porta, J. Pumarega, L.A. Henríquez-Hernández et al. Science of the Total Environment 777 (2021) 146013 Fig. 1. Change from 2006 to 2016 in the geometric mean (adjusted for age, sex and BMI) of the serum concentration of p,p’-DDE, HCB, β-HCH, PCB 138, and naphthlene in the corresponding representative samples of inhabitants of the city of Barcelona. Footonote: See figures in the last right column of Table 1. Many POPs were still detected in 2016, and the change in the Seven of the 8 PBDEs and 8 of the 16 PAHs were detected in ≤7.5% of % > LOD was only significant (e.g., larger than −35% and p < 0.001) individuals in both 2006 and 2016. BDE-153 and acenaphthylene were for two of the nine compounds (PCB 118 and naphthalene) (Table 1). detected >20 percentage points more in 2016 than in 2006 (both Nevertheless, in 2016 p,p’-DDE, HCB, β-HCH, PCBs 138, 153, 156 and p < 0.001) (Supplemental Tables 2 and 3). Mixtures found in human 180, fluorene, phenanthrene, pyrene and fluoranthene were detected bodies change incessantly. in >75% of the population of Barcelona, whereas in 2006 PCB 118 and Happily, in the 10-year study period, substantial reductions in con- naphthalene were also detected in >75% of participants; by contrast, centrations (most, greater than 30%, and up to 80% and 90%) occurred in 2006 fluorene, pyrene and fluoranthene were not detected in almost all sex, age and BMI subgroups. For instance, from 2006 to (p < 0.001 in the % > LOD for all five compounds) (Table 1 and Supple- 2016 there was a 77% decrease in the concentrations of HCB among mental Tables 2 and 3). women, or a reduction of 91% for naphthalene among men (Table 2). Endrin, methoxychlor, o,p’-DDE and γ-HCH were not detected in any For most OCs the reduction was larger in women than in men; this participant in 2016; o,p’-DDT, p,p’-DDT, o,p’-DDD, p,p’-DDD, α- and β- was so, specifically, for β-HCH (−80% vs. –74% in women and men, re- endosulfan, heptachlor, α- and δ-HCH, and seven of the 18 PCBs were spectively), p,p’-DDE, HCB, and PCBs 138, 153 and 180 (Table 2). detected in <6% of participants (Supplemental Tables 2 and 3). However, in 2016 women still had significantly higher median Table 2 Percent change from 2006 to 2016 in the adjusted geometric mean of serum concentrations of persistent organic pollutants (ng/g lipid) in the population of the city of Barcelona, by sex, age, and body mass index. Sexa Ageb Body Mass Indexc Men Women 18–44 years 45–64 years ≥65 years Normal weight Over-weight Obese p,p’-DDE −38⁎ −42⁎ −38⁎ −55⁎ −18 −54⁎ −25 −34 HCB −62⁎ −77⁎ −63⁎ −77⁎ −72⁎ −69⁎ −75⁎ −73⁎ β-HCH −74⁎ −80⁎ −80⁎ −77⁎ −72⁎ −81⁎ −79⁎ −67⁎ PCB 118 −78⁎ −71⁎ −80⁎ −73⁎ −61⁎ −82⁎ −76⁎ −44⁎ PCB 138 −30⁎ −39⁎ −44⁎ −30⁎ −20 −49⁎ −33⁎ 5 PCB 153 −16 −31⁎ −31⁎ −20⁎ −14 −38⁎ −25⁎ 26 PCB 180 −32⁎ −39⁎ −56⁎ −12 −10 −47⁎ −40⁎ 21 Naphthalene −91⁎ −88⁎ −94⁎ −89⁎ −74⁎ −92⁎ −84⁎ −91⁎ Phenanthrene −33 −34 −57⁎ −18 27 −45⁎ 15 −58⁎ The Table shows results for the 9 POPs that had been detected (each) in >70% of the study subjects of the 2006 Barcelona Health Survey study on serum concentrations of POPs. a Models adjusted for age and body mass index. b Models adjusted for sex and body mass index. c Models adjusted for age and sex. ⁎ p-value <0.05 (Wald test) for the percent change from 2006 to 2016 in each stratum; e.g., the 42% decrease from 2006 to 2016 in the concentrations of p,p’-DDE among women was statistically significant (p-value <0.05). 5 M. Porta, J. Pumarega, L.A. Henríquez-Hernández et al. Science of the Total Environment 777 (2021) 146013 Fig. 2. Convergence by gender: Change from 2006 to 2016 in the geometric mean (adjusted for age and BMI) of the serum concentrations (in ng/g lipid) of six persistent organic pollutants in the population of the city of Barcelona by gender. Footnote: The figure shows a selection of the results included in the first two columns of Table 2. The geometric means of each Barcelona Health Survey were adjusted for age and body mass index of the two studies simultaneously. concentrations than men of DDE, HCB and β-HCH. Table 2 and Fig. 2 Among the latter, concentrations of PCBs 138, 153 and 180 did not show the reduction of differences (convergence) by gender. change or even increased non-significantly from 2006 to 2016 (Table 2). No compound showed a larger decrease in the 10 years in the oldest In 2006, 33.4% of participants had concentrations of 3 or more POPs than in the younger groups, whereas for β-HCH, PCBs, naphthalene and at high concentrations; this figure was 15.9% in 2016, a remarkable ab- phenanthrene the decrease was larger in the younger groups, often with solute reduction of 17.4 percentage points, and a 52% relative reduction. monotonic trends (Table 2). The number of participants with 6 or more POPs at high concentrations Larger reductions in concentrations of PCBs and DDE were observed was 13.7% in 2006 and 4.0% in 2016, a difference of 9.6 percentage in individuals with normal weight and overweight than in the obese. points, and a 71% relative reduction (Table 3). 6 M. Porta, J. Pumarega, L.A. Henríquez-Hernández et al. Science of the Total Environment 777 (2021) 146013 Table 3 Absolute and relative change from 2006 to 2016 in the prevalence of subjects with POPs at high concentrations. Number of POPs at 2006 (N = 231) 2016 (N = 240) Cumulative 2006 2016 2016 vs. 2006 high concentrations number of POPs (%) (%) N (%) Cumulative N (%) Cumulative Absolute Relative percent percent differencea differenceb 0 74 (31.8) 31.8 136 (56.6) 56.6 1 42 (18.1) 50.0 41 (17.1) 73.7 ≥1 68.2 43.4 −24.7 −36.4% 2 39 (16.8) 66.8 25 (10.5) 84.1 ≥2 50.0 26.3 −23.7 −47.4% 3 17 (7.5) 74.2 8 (3.5) 87.7 ≥3 33.2 15.9 −17.4 −52.1% 4 10 (4.3) 78.6 10 (4.3) 92.0 ≥4 25.8 12.3 −13.4 −52.3% 5 18 (7.8) 86.3 10 (4.0) 96.0 ≥5 21.4 8.0 −13.4 −62.6% 6 10 (4.3) 90.6 2 (0.9) 96.9 ≥6 13.7 4.0 −9.6 −70.8% 7 18 (7.9) 98.5 5 (1.9) 98.8 ≥7 9.4 3.1 −6.3 −67.0% 8 3 (1.5) 100.0 0 (0.0) 98.8 ≥8 1.5 1.2 −0.3 −20.0% 9 0 (0.0) 100.0 3 (1.2) 100.0 9 0 1.2 +1.2 –– Based on concentrations above percentile 75 of the nine most frequently detected POPs in the 2006 Barcelona Health Survey study of POPs (see Table 1, Table 2, and Section 2.3. Statistical analyses). N, number of subjects. a Absolute difference from 2006 to 2016 (in percentage points) of the percent of subjects with POPs at high concentrations (>P75). b Relative difference or change from 2006 to 2016 (in %) of the percent of subjects with POPs at high concentrations (>P75). We next focus exclusively on findings in 2016. Concentrations of were seen for OCPs (>70% for concentrations of HCB and β-HCH), OCPs and PCBs increased with increasing age (Tables 4 and 5). This followed by PCBs (>70% for PCB 118 and >20% for the other three was not so for PAHs, neither in women nor in men. In men, most POP most detected PCBs). Reductions were more marked among women, concentrations tended to increase with increasing BMI, with the excep- younger individuals, and people who were not obese. We found re- tion of PAHs, which were lower in obese men (Table 4). In women, markable equality of POP levels in the different social classes. adjusting for age, concentrations of all OCPs increased with increasing Slightly larger reductions of concentrations in women than men BMI, whereas this pattern was not apparent for PCBs, which showed have been previously reported (Fernyhough et al., 1999; Partearroyo lower levels in women who were overweight (Table 5). et al., 2019). Perhaps women reduced more the intake of dietary animal Concentrations of OCPs and PCBs were slightly higher in men who fats. Also, concentrations of all compounds were lower in women hav- lost 4–5 kg in the previous 6 months. In men, concentrations of all com- ing breastfed their children longer (Song et al., 2018). Nevertheless, in pounds were non-significantly lower in participants who reported 2016 women still had significantly higher concentrations than men of being breastfed as a child (Table 4); in women this was also so for DDE, HCB and β-HCH, as also observed before (Zubero et al., 2017; most POPs (Table 5). Concentrations of all compounds were non- Carreño et al., 2007). Hence, a relevant component of the success seen significantly lower in women who breastfed their children longer. in the current decreasing human concentrations of POPs in Barcelona In men, we mutually adjusted BMI and weight-change (during pre- appears to be a convergence of such concentrations between women vious 6 months) to assess whether concentrations were highest in sub- and men (i.e., a reduction in gender differences). Available studies jects with high BMI and higher weight-loss. This was so for DDE, HCB, β- could be reanalyzed to formally assess this possibility while accounting HCH, and the four PCBs, with figures similar to those already shown in for confounders; hence, to identify common and separate causes of the Table 4. These associations were not apparent for PAHs. In women, we decrease –by gender. further adjusted for duration of lactation. Only concentrations of DDE Both in women and men concentrations of OCs (OCPs and PCBs) in- were highest in women with high BMI and higher weight-loss. creased with age, as often reported. The observation probably results Figures for the other POPs were similar to those already shown in from age-cohort-period effects (Porta, 2018; Porta et al., 2008; Porta, Table 5. 2004; Quinn and Wania, 2012; Gasull et al., 2013). In fact, no compound Men and women born outside Spain showed higher concentrations showed a larger decrease from 2006 to 2016 in the oldest than in the of p,p’-DDE and lower concentrations of the rest of OCs. Men born younger groups, whereas for β-HCH, PCBs, naphthalene and phenan- abroad also had lower concentrations of fluorene. threne the decrease was larger in the younger groups, mostly with No simple patterns were apparent for occupational social class. In monotonic trends. This is also excellent news. Younger cohorts have men (Table 4), no compounds showed statistically significant differ- probably been less exposed to OCs because of policies operating during ences in the geometric means adjusted for age (aGM). Concentrations recent periods (Porta et al., 2008; Porta, 2004; Porta, 2012). By contrast, of PCBs were often higher in men and women of the more affluent social the decrease in the concentrations of these highly persistent pollutants classes I-II but, again, none was statistically significant. Higher concen- was less marked among older individuals, whom obviously had more trations in classes IV-V (less affluent) were only apparent for p,p’-DDE, years of exposure and (influenced as well by the higher OC contamina- fluorene, phenanthrene and pyrene in men, but none was statistically tion in their specific historical period), higher body concentrations. But significant. In men the aGM was never higher in class III than in the the observation is also likely due to age-related physiological factors and other class categories. In women (Table 5), concentrations of most behaviors, such as decreased detoxification activity of liver enzymes POPs were similar across classes; they were almost never higher in clas- (Sotaniemi et al., 1997), and higher consumption of POP-rich foods by ses IV-V than in classes I-II; and concentrations of p,p’-DDE, HCB, β-HCH, older individuals (Porta, 2018; Henríquez-Hernández et al., 2020; and PAHs were highest in class III (statistically significant for HCB and Porta et al., 2008; Partearroyo et al., 2019; Gasull et al., 2013). pyrene). Larger reductions in concentrations of PCBs and DDE were observed in individuals with normal weight and overweight than in the obese; 4. Discussion among the latter, concentrations of some PCBs did not change or even increased from 2006 to 2016. These findings reinforce reasons to fight During the 10-year period from 2006 to 2016, there were marked re- obesity through individual behavior and policies, and suggest that re- ductions in blood concentrations of most POPs across most population ductions in internal POP concentrations are less likely among the subgroups defined by gender, age, and BMI. The largest decreases obese even when levels decrease in society. In agreement, it has been 7 M. Porta, J. Pumarega, L.A. Henríquez-Hernández et al. Science of the Total Environment 777 (2021) 146013 Table 4 Serum concentrations of POPs (ng/g lipid) (compounds detected in over 75% of the population of Barcelona city) in 2016 by sociodemographic characteristics, men. p,p’-DDE HCB β-HCH Median (P25-P75) aGM (CI 95%) Median (P25-P75) aGM (CI 95%) Median (P25-P75) aGM (CI 95%) All men 145.4 (87.6–355.0) 189.5 (154.6–232.2) 23.2 (13.9–44.6) 25.4 (21.5–29.9) 10.8 (6.0–35.9) 15.9 (13.4–18.9) Age (years) 18–44 103.0 (73.2–151.6)⁎ 112.4 (83.8–150.7) 15.4 (10.5–24.2)⁎ 15.8 (12.6–20.0) 6.6 (5.3–8.3)⁎ 6.6 (5.1–8.4) 45–64 176.8 (112.0–377.6) 176.8 (124.1–251.9) 23.9 (13.1–43.0) 22.7 (17.2–30.1) 17.6 (8.6–36.5) 16.2 (12.0–21.8)⁎⁎ ≥65 543.3 (183.6–910.2) 456.0 (292.2–711.8)⁎⁎ 55.0 (30.3–84.3) 61.0 (42.9–86.8)⁎⁎ 52.8 (36.0–123.0) 59.5 (40.9–86.6)⁎⁎ Body Mass Index Normal 94.4 (49.4–145.9)⁎ 122.6 (87.4–172.1) 15.2 (10.2–30.8)⁎ 26.5 (20.0–35.2) 7.8 (5.6–11.3)⁎ 17.0 (12.7–22.7) Overweight 268.8 (130.1–815.2) 249.7 (185.6–336.1)⁎⁎ 23.2 (15.3–49.7) 21.4 (16.7–27.3) 17.7 (5.9–39.0) 13.1 (10.1–16.9) Obese 275.5 (128.2–491.5) 231.1 (142.4–375.0)⁎⁎ 39.3 (26.1–68.8) 36.5 (24.4–54.5) 26.5 (18.3–55.1) 23.2 (15.3–35.2) Weight change in last 6 months Lost 4–5 Kg 200.3 (91.4–1417) 386.9 (192.6–777.2)⁎⁎ 18.0 (13.6–54.4) 30.5 (17.1–54.3) 8.9 (5.1–38.9) 18.5 (10.2–33.8) Little or no change (± 3 Kg)a 151.6 (89.2–455.3) 183.0 (147.3–227.4) 23.5 (13.9–45.8) 25.3 (21.1–30.3) 10.8 (6.1–35.9) 15.8 (13.1–19.1) Gained 4–5 Kg 129.0 (72.0–138.6) 134.9 (67.1–271.2) 18.7 (9.3–26.4) 22.0 (12.3–39.3) 12.4 (4.7–18.3) 14.6 (8.0–26.7) Was breastfed No 139.9 (117.4–555.8) 228.8 (138.5–377.8) 24.4 (18.5–53.5) 35.5 (23.4–53.7) 8.1 (6.8–35.4) 19.2 (12.4–29.7) Yes 151.6 (87.4–355.0) 185.0 (146.4–233.7) 22.2 (13.3–46.3) 24.8 (20.4–30.1) 12.1 (6.1–36.2) 15.7 (12.9–19.3) Birth place Catalonia 146.6 (86.5–347.8) 169.6 (133.4–215.5) 27.6 (15.3–49.0)⁎ 35.5 (23.4–53.7) 17.1 (7.4–36.2)⁎ 17.3 (14.1–21.3) Rest of Spain 138.3 (74.2–306.3) 152.7 (87.1–267.7) 24.7 (10.7–67.8) 27.9 (17.9–43.5) 16.6 (5.6–54.5) 14.8 (9.1–24.0) Abroad 148.6 (92.7–531.2) 324.5 (206.9–508.8)⁎⁎ 11.9 (7.0–18.1) 14.2 (9.9–20.3)⁎⁎ 6.5 (4.3–9.3) 12.2 (8.3–18.0) Occupational social class I-II (more affluent) 132.0 (74.3–473.6) 169.9 (127.9–225.8) 132.0 (74.3–473.6) 28.1 (22.3–35.4) 9.1 (5.6–46.5) 16.0 (12.6–20.4) III 136.7 (82.5–206.8) 179.5 (115.7–278.3) 136.7 (82.5–206.8) 24.7 (17.3–35.3) 10.6 (6.6–22.9) 16.2 (11.2–23.5) IV-V (less affluent) 183.8 (121.5–531.2) 238.5 (163.5–347.8) 183.8 (121.5–531.2) 21.7 (15.9–29.4) 15.1 (6.7–36.2) 15.5 (11.2–21.3) PCB 138 PCB 153 PCB 156 PCB 180 Median aGM Median aGM Median aGM Median aGM (P25-P75) (CI 95%) (P25-P75) (CI 95%) (P25-P75) (CI 95%) (P25-P75) (CI 95%) Men 39.2 36.4 64.9 60.9 7.4 7.2 54.7 53.2 (21.3–75.3) (30.9–42.8) (30.3–122.9) (52.3–70.9) (4.9–14.3) (6.1–8.4) (24.7–134.7) (44.1–64.2) Age (years) 18–44 21.4 16.0 32.4 29.1 5.2 3.9 25.3 18.3 (8.1–28.6)⁎ (12.6–20.4) (18.4–47.9)⁎ (23.2–36.5) (1.5–7.0)⁎ (3.1–4.8) (8.4–32.6)⁎ (13.9–24.0) 45–64 56.6 47.8 91.7 77.2 10.6 8.4 112.1 85.5 (34.4–76.3) (35.7–64.0)⁎⁎ (60.0–137.4) (58.8–101.4)⁎⁎ (5.2–16.4) (6.5–10.9)⁎⁎ (71.4–145.1) (61.4–118.9)⁎⁎ ≥65 82.7 86.4 131.1 134.4 16.0 15.9 152.4 141.5 (74.9–129.2) (59.9–124.7)⁎⁎ (113.8–202.8) (95.4–189.3)⁎⁎ (11.0–26.2) (11.5–21.9)⁎⁎ (122.9–184.2) (93.5–214.4)⁎⁎ Body Mass Index Normal 22.3 34.9 37.7 58.2 5.7 7.5 30.2 57.1 (13.5–47.6)⁎ (26.3–46.3) (21.3–75.3)⁎ (44.7–75.8) (3.3–8.1)⁎ (5.7–9.9) (13.3–102.6)⁎ (41.3–79.0) Overweight 50.1 34.3 81.2 58.3 8.1 6.6 82.4 47.2 (28.5–78.3) (26.8–43.9) (33.3–132.7) (46.2–73.5) (4.1–16.8) (5.2–8.4) (27.8–152.0) (35.5–62.8) Obese 56.9 46.7 87.2 75.4 11.9 8.2 86.1 62.8 (29.7–95.0) (31.2–69.9) (46.0–162.8) (51.7–109.9) (5.3–17.2) (5.6–12.1) (43.1–145.5) (39.5–99.9) Weight change in last 6 months Lost 4–5 Kg 45.4 45.3 78.4 85.9 6.0 8.4 67.8 68.4 (20.1–69.8) (25.5–80.2) (35.1–118.9) (50.5–146.3) (5.2–13.6) (4.8–14.6) (25.5–150.5) (35.5–132.0) Little or no 40.1 35.2 67.0 58.5 7.8 7.3 55.1 51.3 change (20.8–77.4) (29.5–42.1) (30.3–127.9) (49.6–69.1) (4.8–15.5) (6.1–8.6) (24.7–139.5) (41.9–62.9) (± 3 Kg)a Gained 4–5 Kg 27.7 41.3 50.2 66.5 5.1 5.6 42.7 61.7 (23.1–40.2) (23.3–73.3) (30.5–68.5) (39.0–113.2) (3.2–6.6) (3.2–9.7) (20.8–80.0) (31.9–119.0) Was breastfed No 50.0 40.2 79.8 71.7 7.9 8.6 56.3 54.3 (8.1–77.0) (26.2–61.8) (18.4–127.1) (48.2–106.6) (5.3–12.3) (5.7–12.9) (8.1–123.8) (33.5–88.2) Yes 40.1 35.7 65.2 59.4 7.5 7.1 58.3 54.0 (21.5–75.7) (29.2–43.6) (30.8–113.9) (49.3–71.4) (4.3–15.2) (5.9–8.6) (25.5–138.8) (43.1–67.6) Birth place Catalonia 47.8 41.5 79.5 68.9 7.8 8.4 83.3 66.6 (22.1–78.3)⁎ (34.6–49.9) (41.2–127.9)⁎ (57.9–81.9) (5.2–15.6)⁎ (7.0–10.0) (30.3–147.2)⁎ (54.7–81.1) Rest of Spain 54.2 49.6 84.2 78.8 6.9 7.4 72.7 74.2 (23.2–76.5) (32.3–76.3) (33.8–155.0) (52.6–117.9) (5.2–13.4) (4.8–11.3) (31.5–153.5) (46.8–117.9) Abroad 23.5 18.5 29.4 33.1 3.2 4.2 21.7 19.2 (3.9–32.3) (13.1–26.1)⁎⁎ (10.1–57.9) (23.9–45.8)⁎⁎ (1.4–7.8) (3.0–5.9)⁎⁎ (3.2–33.5) (13.2–27.8)⁎⁎ 8 M. Porta, J. Pumarega, L.A. Henríquez-Hernández et al. Science of the Total Environment 777 (2021) 146013 Table 4 (continued) PCB 138 PCB 153 PCB 156 PCB 180 Median aGM Median aGM Median aGM Median aGM (P25-P75) (CI 95%) (P25-P75) (CI 95%) (P25-P75) (CI 95%) (P25-P75) (CI 95%) Occupational social class I-II (more 42.9 39.2 72.2 64.4 7.8 7.9 54.5 59.7 affluent) (21.4–77.4) (31.1–49.3) (33.8–127.0) (51.9–79.9) (5.2–14.5) (6.3–9.8) (25.3–144.1) (45.8–77.6) III 27.3 30.9 41.5 55.9 5.7 6.4 53.3 46.3 (8.9–70.1) (21.7–44.0) (16.9–118.6) (40.1–77.9) (1.5–13.1) (4.5–9.0) (8.8–107.3) (30.9–69.6) IV-V (less 39.2 35.9 63.0 58.7 6.2 6.7 62.5 48.1 affluent) (24.0–65.1) (26.5–48.8) (30.2–123.3) (44.1–78.1) (4.0–14.6) (5.0–9.0) (23.9–129.4) (33.9–68.2) Fluorene Phenanthrene Pyrene Fluoranthene Median aGM Median aGM Median aGM Median aGM (P25-P75) (CI 95%) (P25-P75) (CI 95%) (P25-P75) (CI 95%) (P25-P75) (CI 95%) Men 18.3 15.4 46.8 28.1 23.6 22.3 21.3 17.2 (11.0–35.7) (12.5–18.9) (7.1–104.0) (20.2–39.0) (16.9–32.1) (19.6–25.4) (13.4–37.2) (14.0–21.2) Age (years) 18–44 22.3 19.5 71.8 31.3 24.2 24.7 21.3 18.5 (13.6–39.0)⁎ (14.5–26.2) (7.0–123.0)⁎ (19.8–49.6) (15.6–33.0) (205–29.6) (13.6–36.0) (13.8–24.8) 45–64 14.0 11.7 29.3 15.9 20.5 18.1 18.7 12.5 (8.9–24.1) (8.2–16.7)⁎⁎ (1.8–65.9) (9.1–27.6) (15.0–29.8) (14.5–22.6) (6.8–30.2) (8.8–17.9) ≥65 23.4 16.6 76.5 55.3 27.5 26.5 33.2 25.1 (10.0–39.5) (10.6–26.0) (36.6–137.9) (27.6–110.9) (20.6–44.6) (20.0–35.0) (18.7–44.5) (16.0–39.2) Body Mass Index Normal 22.4 19.7 69.3 33.5 25.6 25.4 21.8 18.8 (14.1–39.5)⁎ (13.9–28.1) (7.1–137.6) (18.9–59.4) (17.0–33.3) (20.3–31.7) (15.2–41.1) (13.0–27.0) Overweight 17.4 14.3 36.7 26.0 22.2 21.3 20.3 16.9 (8.8–29.0) (10.5–19.5) (7.5–103.5) (15.7–42.9) (17.0–34.0) (17.5–25.9) (12.5–38.4) (12.3–23.2) Obese 15.1 11.0 43.6 23.6 19.8 19.2 18.9 15.1 (5.8–25.1) (6.6–18.1)⁎⁎ (7.0–72.0) (10.4–53.4) (13.4–30.2) (14.0–26.4) (10.0–30.3) (9.0–25.4) Weight change in last 6 months Lost 4–5 Kg 16.7 11.4 35.5 16.7 25.0 21.3 19.4 20.3 (11.9–25.5) (5.6–23.2) (1.9–86.6) (5.3–52.4) (16.5–29.2) (13.6–33.5) (16.3–29.0) (9.8–42.3) Little or no change (± 3 21.2 16.5 55.6 31.6 23.8 22.8 21.8 17.4 Kg)a (11.6–37.3) (13.2–20.6) (7.8–107.1) (22.1–45.1) (17.1–34.5) (19.8–26.3) (12.4–41.4) (13.9–21.9) Gained 4–5 Kg 14.7 9.8 20.0 13.1 16.9 18.3 16.9 12.7 (6.1–26.5) (7.2–18.7) (4.8–20.0) (1.4–77.4) (4.2–41.0) (14.2–23.4) (11.7–28.9) (14.2–23.8) Was breastfed No 23.4 21.6 89.6 46.4 31.9 28.8 41.1 25.9 (14.2–39.6) (12.6–37.3) (23.7–120.2) (19.8–108.6) (14.7–57.4) (20.6–40.4) (14.4–42.0) (15.4–43.7) Yes 17.5 14.6 41.0 25.6 23.0 21.2 20.1 16.1 (10.3–36.8) (11.3–18.7) (6.1–114.3) (17.2–38.1) (16.8–31.4) (18.1–24.8) (12.9–32.9) (12.7–20.6) Birth place Catalonia 21.7 17.5 69.1 33.9 25.3 24.1 22.1 19.7 (12.8–38.8) (13.7–22.4) (7.2–119.0)⁎ (23.0–50.2) (17.1–35.9) (20.7–28.2) (16.3–41.4) (15.4–25.4) Rest of Spain 15.3 14.7 11.8 10.1 22.2 20.5 15.2 12.4 (10.0–35.7) (8.3–26.0) (1.8–33.3) (4.0–25.2)⁎⁎ (17.0–26.6) (14.3–29.5) (9.8–32.8) (6.9–22.3) Abroad 15.4 10.1 36.7 27.7 22.0 17.8 18.1 13.1 (5.6–27.4) (6.4–16.0)⁎⁎ (15.9–99.8) (13.3–57.7) (14.0–30.2) (13.3–23.8) (10.6–28.4) (8.2–20.9) Occupational social class I-II (more affluent) 17.0 12.6 51.2 27.5 23.6 22.1 21.3 20.4 (5.4–31.3) (9.5–16.7) (4.2–117.5) (17.3–43.8) (16.8–31.4) (18.4–26.5) (15.2–41.4) (15.2–27.3) III 20.7 18.1 45.4 24.1 24.0 21.6 21.7 13.8 (12.0–37.4) (11.7–28.1) (6.0–82.7) (11.8–49.4) (17.9–32.0) (16.3–28.6) (8.6–35.0) (8.8–21.7) IV-V (less affluent) 17.6 19.5 47.3 32.4 22.9 23.2 20.2 15.1 (12.5–40.2) (13.4–28.5) (26.0–104.5) (17.5–60.1) (15.6–39.3) (18.2–29.5) (13.1–33.4) (10.2–22.2) aGM: Geometric mean adjusted for age. N = 112. a Reference category (except where otherwise noted, the reference category is the first category mentioned above). ⁎ p-value <0.05 (median, Kruskal-Wallis test). ⁎⁎ p-value <0.05 (aGM, Wald test; compared against the reference category). suggested that bioaccumulated OCPs are frequently at higher concen- In innovative studies, concentrations of DDE and PCBs were both pos- trations and tend to persist longer in the adipose tissue of obese than itively associated with BMI and recent weight-loss; they were highest in non-obese individuals (Arrebola et al., 2009; Wolff et al., 2005; with higher BMI, higher weight-loss, and (in women) shorter lactation Wolff et al., 2007). (Wolff et al., 2005; Wolff et al., 2007; Wolff and Teitelbaum, 2020). In Men with obesity and overweight had the highest levels of p,p’-DDE. our study, this was so in men for DDE, HCB, β-HCH, and the four PCBs; Women with obesity and overweight had the highest levels of HCB and in women, for DDE only. However, numbers were small, and more mea- β-HCH. Women with overweight had the lowest levels of PCBs, as also sures per person would be needed to capture influences during the life observed in other populations (Ibarluzea et al., 2011). course, especially in women. Further progress to elucidate these issues 9 M. Porta, J. Pumarega, L.A. Henríquez-Hernández et al. Science of the Total Environment 777 (2021) 146013 Table 5 Serum concentrations of POPs (ng/g lipid) (compounds detected in over 75% of the population of Barcelona city) in 2016 by sociodemographic characteristics, women. p,p’-DDE HCB β-HCH Median (P25-P75) aGM (CI 95%) Median (P25-P75) aGM (CI 95%) Median (P25-P75) aGM (CI 95%) Women 221.1 (114.6–968.3) 273.4 (230.4–324.4) 36.4 (17.5–104.7) 45.5 (39.2–52.7) 26.7 (10.7–81.1) 26.5 (21.7–32.4) Age (years) 18–44 135 (71.1–195.4)⁎ 131.4 (99.8–172.9) 17.5 (13.6–25.4)⁎ 19.2 (15.2–24.3) 10.5 (3.6–17.6)⁎ 9.3 (6.8–12.8) 45–64 317.8 (125.2–640.3) 280.9 (207.4–380.5)⁎⁎ 57.8 (24.3–95.2) 51.2 (39.5–66.3)⁎⁎ 36.4 (16.2–60.0) 33.8 (23.7–48.2)⁎⁎ ≥65 1204 (871.6–1796) 936.4 (661.1–1326)⁎⁎ 139.4 (96.7–327.1) 170.7 (126.9–229.7)⁎⁎ 109.9 (68.4–208.5) 115.8 (77.1–173.8)⁎⁎ Body Mass Index Normal 146.0 (70.3–293.9)⁎ 228.7 (173.8–301.0) 21.4 (15.5–35.0)⁎ 34.2 (27.4–42.7) 14.0 (8.0–30.6)⁎ 21.1 (15.4–28.8) Overweight 404.7 (149.0–1202) 300.7 (220.2–410.7) 84.4 (21.5–137.3) 50.2 (39.0–64.6)⁎⁎ 44.0 (16.2–106.3) 27.5 (19.3–39.2) Obese 898.0 (260.3–1989) 360.7 (229.2–567.5) 132.1 (70.9–220.0) 72.2 (50.1–104.2)⁎⁎ 85.5 (38.5–151.8) 44.1 (26.3–73.6)⁎⁎ Weight change in last 6 months Lost 4–5 Kg 329.5 (69.6–733.1) 322.4 (155.6–668.2) 24.8 (16.4–64.0) 39.1 (20.9–73.1) 32.8 (1.4–76.5) 22.7 (9.8–52.8) Little or no change (± 3 Kg)a 223.8 (118.9–969.7) 269.7 (224.1–324.7) 38.8 (17.5–110.5) 44.1 (37.7–51.7) 24.8 (10.8–75.8) 25.2 (20.3–31.2) Gained 4–5 Kg 142.1 (91.0–1074) 282.0 (150.6–528.2) 47.8 (15.6–251.9) 71.1 (41.5–121.8) 71.9 (16.5–176.2) 54.0 (26.2–111.6) Was breastfed No 180 (118.2–1208) 310.1 (188.7–509.6) 48.4 (21.4–91.9) 48.9 (32.6–73.6) 26.0 (11.7–66.2) 28.2 (15.9–49.8) Yes 222.7 (120.1–757.1) 278.5 (230.7–336.1) 37.9 (16.7–126.0) 48.5 (41.6–56.6) 32.2 (10.8–106.3) 29.0 (23.4–36.0) Birth place Catalonia 206.3 (98.8–683.7) 250.5 (198.7–315.7) 52.8 (22.4–118.8)⁎ 55.4 (45.9–66.9) 29.8 (10.7–92.7)⁎ 28.4 (21.6–37.5) Rest of Spain 349.9 (179.9–1192) 211.7 (143.9–311.5) 108.1 (56.7–177.7) 57.9 (42.2–79.4) 67.9 (26.6–117.5) 29.1 (18.3–46.1) Abroad 193.4 (101.9–471.7) 403.7 (283.9–573.9)⁎⁎ 15.6 (12.1–21.4) 24.8 (18.6–33.1)⁎⁎ 14.6 (4.3–19.0) 21.3 (13.9–32.5) Occupational social class I-II (more affluent) 186.1 (101.9–376.5)⁎ 256.4 (195.2–336.7) 25.7 (17.3–59.5) 45.1 (35.9–56.6) 15.0 (8.9–39.8)⁎ 23.5 (17.6–31.4) III 256.7 (121.8–779.6) 320.5 (225.8–454.8) 72.0 (17.3–139.4) 65.7 (49.0–88.0)⁎⁎ 42.9 (12.2–109.0) 36.8 (25.4–53.4) IV-V (less affluent) 614.1 (143.4–1208) 279.2 (203.7–382.7) 50.7 (18.6–137.3) 35.4 (27.2–46.1) 35.5 (15.4–109.9) 23.4 (16.8–32.8) Breastfed her children Never 739.5 (154.2–1421) 468.0 (232.7–941.1) 78.8 (24.2–181.1) 63.8 (35.6–114.5) 60.6 (14.9–126.2) 44.9 (21.9–92.0) ≤6 months 374.8 (172.6–1021) 342.2 (232.4–503.9) 54.8 (22.6–115.2) 50.3 (36.4–69.5) 32.8 (16.2–80.8) 32.1 (21.6–47.8) >6 months 367.3 (105.3–1192) 277.9 (208.9–369.7) 53.3 (18.8–139.1) 46.7 (36.8–59.5) 35.7 (10.4–109.9) 27.2 (20.3–36.5) PCB 138 PCB 153 PCB 156 PCB 180 Median aGM Median aGM Median aGM Median aGM (P25-P75) (CI 95%) (P25-P75) (CI 95%) (P25-P75) (CI 95%) (P25-P75) (CI 95%) Women 42.2 42.1 63.9 64.6 8.7 7.1 67.8 48.7 (22.7–73.6) (37.1–47.7) (34.2–117.2) (57.3–72.8) (4.0–16.6) (6.1–8.2) (24.1–110.4) (41.5–57.0) Age (years) 18–44 20.3 20.6 31.4 31.4 5.6 3.9 20.7 19.3 (14.4–31.0)⁎ (16.9–25.2) (21.9–48.8)⁎ (25.8–38.1) (1.5–7.4)⁎ (3.1–5.0) (11.4–43.9)⁎ (14.9–25.0) 45–64 57.7 49.4 86.7 79.3 9.1 7.5 80.3 71.3 (39.1–74.7) (39.6–61.5)⁎⁎ (59.4–118.0) (63.9–98.3)⁎⁎ (5.2–15.5) (5.7–9.8)⁎⁎ (59.5–112.6) (53.5–95.0)⁎⁎ ≥65 112.6 115.6 168.8 170.5 18.4 18.7 127.9 143.7 (63.7–252.4) (89.8–148.8)⁎⁎ (103.5–365.0) (133.1–218.4)⁎⁎ (12.7–34.6) (13.7–25.4)⁎⁎ (98.7–344.1) (103.3–199.8)⁎⁎ Body Mass Index Normal 29.0 43.1 48.0 71.6 6.3 7.8 43.2 62.4 (18.5–50.4)⁎ (35.3–52.5) (27.0–83.9)⁎ (59.4–86.4) (1.7–10.7)⁎ (6.1–9.8) (18.7–80.9)⁎ (48.8–79.6) Overweight 59.7 37.6 94.0 53.0 12.6 6.1 84.0 33.9 (31.1–80.2) (30.0–47.1) (39.2–122.7) (42.9–65.6)⁎⁎ (6.0–15.3) (4.6–8.0) (24.1–112.4) (25.8–44.7)⁎⁎ Obese 77.8 42.9 107.2 63.4 13.3 6.3 89.7 43.3 (40.3–148.7) (30.9–59.6) (73.1–198.4) (46.5–86.3) (6.4–22.2) (4.2–9.3) (50.6–144.0) (28.9–64.6) Weight change in last 6 months Lost 4–5 Kg 41.4 37.0 45.7 56.6 6.6 6.2 31.4 31.6 (16.3–61.2) (12.6–59.6) (21.8–63.1) (26.5–88.1) (34.1–94.0) (2.8–9.0) (3.2–11.7) (14.3–80.8) Little or no change (± 3 43.0 42.1 70.9 65.3 9.1 7.2 68.8 52.1 Kg)a (25.3–75.6) (36.8–48.2) (38.4–118.9) (57.4–74.3) (4.4–17.3) (6.1–8.5) (25.1–112.2) (44.0–61.6) Gained 4–5 Kg 25.2 45.8 40.9 63.2 7.8 6.8 43.3 30.9 (17.3–107.4) (28.9–72.4) (23.9–150.5) (40.8–97.8) (1.6–17.9) (3.9–11.8) (4.8–102.3) (17.5–54.6)⁎⁎ Was breastfed No 45.0 55.8 63.9 95.0 15.8 11.1 72.6 80.4 (31.0–121.2) (39.4–78.9) (49.3–201.3) (68.8–131.2) (6.3–28.0) (7.1–17.2) (34.7–146.7) (52.1–123.9) Yes 43.1 42.7 72.5 64.1 8.7 6.8 69.2 47.2 (20.0–75.4) (37.4–48.7) (31.1–117.6) (56.7–72.4)⁎⁎ (1.7–14.5) (5.7–8.0)⁎⁎ (20.5–112.2) (40.1–55.7)⁎⁎ Birth place Catalonia 50.4 51.5 80.8 83.8 9.0 8.3 80.9 71.3 (29.0–77.7)⁎ (43.9–60.4) (48.4–134.8)⁎ (72.7–96.5) (5.8–17.9)⁎ (6.8–10.2) (44.3–113.0)⁎ (59.6–85.3) Rest of Spain 71.8 46.5 104.0 70.9 12.8 7.4 98.3 57.3 (42.6–132.8) (35.6–60.6) (75.1–199.8) (55.9–89.8) (6.5–26.9) (5.3–10.4) (67.9–178.2) (42.5–77.3) 10 M. Porta, J. Pumarega, L.A. Henríquez-Hernández et al. Science of the Total Environment 777 (2021) 146013 Table 5 (continued) PCB 138 PCB 153 PCB 156 PCB 180 Median aGM Median aGM Median aGM Median aGM (P25-P75) (CI 95%) (P25-P75) (CI 95%) (P25-P75) (CI 95%) (P25-P75) (CI 95%) Abroad 18.6 25.6 23.9 35.2 1.7 4.9 15.7 19.5 (12.8–31.1) (20.1–32.6)⁎⁎ (14.8–45.9) (28.4–43.7)⁎⁎ (1.4–8.7) (3.6–6.7)⁎⁎ (5.6–33.3) (14.9–25.6)⁎⁎ Occupational social class I-II (more affluent) 35.9 44.9 60.0 73.2 6.8 8.0 65.0 60.6 (23.9–71.9) (36.8–54.7) (37.2–114.8) (60.6–88.3) (5.1–13.8) (6.3–10.1) (23.9–102.8) (47.3–77.6) III 60.8 48.0 89.9 67.8 12.1 7.4 83.2 44.8 (21.3–101.0) (37.3–61.9) (31.8–143.4) (53.3–86.4) (1.7–19.0) (5.5–10.0) (18.0–112.3) (32.6–61.6) IV-V (less affluent) 49.8 36.5 75.5 54.7 12.7 6.0 69.2 42.4 (24.5–101.9) (29.0–45.9) (36.0–144.5) (44.0–68.0) (1.7–19.3) (4.6–7.8) (24.5–114.0) (31.8–56.5) Breastfed her children Never 71.9 58.0 114.4 91.3 11.7 9.1 101.8 84.2 (41.3–115.5) (34.9–97.9) (69.4–180.0) (55.2–150.9) (6.4–19.0) (5.3–15.8) (82.1–147.1) (43.9–161.4) ≤6 months 59.1 50.8 83.2 75.4 11.8 8.4 77.2 54.7 (33.1–85.0) (38.1–67.9) (51.2–139.8) (57.1–99.6) (6.8–16.6) (6.2–11.4) (44.3–121.2) (38.1–78.4) >6 months 54.0 41.5 88.8 68.1 9.6 7.4 83.2 52.6 (23.2–107.0) (33.5–51.4) (42.2–150.0) (55.4–83.6) (5.1–22.0) (5.9–9.3) (36.8–147.8) (40.3–68.6) Fluorene Phenanthrene Pyrene Fluoranthene Median aGM Median aGM Median aGM Median aGM (P25-P75) (CI 95%) (P25-P75) (CI 95%) (P25-P75) (CI 95%) (P25-P75) (CI 95%) Women 21.0 13.4 53.6 25.2 23.3 22.0 20.0 15.5 (5.7–40.2) (10.5–17.0) (2.9–95.0) (18.4–34.6) (13.6–33.6) (19.3–25.0) (11.4–31.3) (12.6–19.2) Age (years) 18–44 25.4 15.4 50.8 22.2 23.8 24.4 22.9 17.9 (6.7–44.4) (10.6–22.3) (2.4–107.7) (13.5–36.4) (14.5–39.0) (19.9–29.8) (11.9–34.8) (12.9–25.0) 45–64 18.1 14.2 51.7 29.9 23.6 19.9 19.7 13.7 (6.2–39.9) (9.4–21.5) (9.4–91.8) (17.3–51.7) (14.0–28.2) (15.9–24.8) (10.1–29.2) (9.5–19.7) ≥65 13.3 9.7 64.7 25.0 18.6 21.1 16.2 14.3 (4.8–24.6) (6.0–15.5) (1.7–88.2) (13.3–46.8) (13.1–35.5) (16.4–27.3) (6.9–27.8) (9.4–21.8) Body Mass Index Normal 20.3 12.1 49.9 25.0 23.0 20.4 19.5 14.0 (6.3–36.8) (8.3–17.8) (2.9–99.0)⁎ (15.3–40.8) (12.9–29.6) (16.7–24.9) (11.9–28.4) (10.0–19.6) Overweight 24.5 18.1 78.4 41.2 25.5 26.8 26.6 19.6 (11.9–46.9) (11.7–28.0) (18.9–116.1) (23.6–72.0) (15.2–38.8) (21.4–33.6) (13.4–45.6) (13.4–28.7) Obese 16.6 10.5 8.8 11.1 21.9 20.9 17.5 15.0 (1.5–32.9) (5.6–19.7) (1.6–78.0) (4.9–24.9) (13.2–38.5) (15.1–29.1) (6.9–29.0) (8.6–25.9) Weight change in last 6 months Lost 4–5 Kg 9.6 5.1 55.0 23.2 26.8 28.7 19.2 17.8 (1.2–33.4) (1.9–13.8) (7.1–111.1) (6.0–89.7) (13.3–62.9) (16.7–49.2) (15.0–34.7) (7.2–43.7) Little or no change (± 3 21.0 13.7 51.4 24.9 21.8 21.0 19.7 15.4 Kg)a (6.2–40.2) (10.7–17.7) (2.6–92.2) (17.7–35.1) (13.1–29.1) (18.3–24.1) (11.2–29.2) (12.3–19.4) Gained 4–5 Kg 28.4 20.0 71.2 31.1 29.3 31.1 30.4 15.2 (18.6–51.1) (8.5–47.3) (6.5–141.5) (9.7–99.4) (27.3–64.7) (19.6–49.6) (5.4–46.8) (7.0–33.2) Was breastfed No 19.6 20.2 9.6 15.3 23.5 26.3 16.2 9.9 (9.2–54.5) (9.9–40.9) (1.6–137.6) (6.0–39.4) (13.1–42.3) (17.9–38.6) (1.4–29.2) (5.4–18.2) Yes 21.0 12.5 51.4 25.5 21.8 20.9 19.7 16.0 (5.2–39.4) (9.6–16.4) (5.9–91.4) (17.8–36.4) (13.3–29.2) (18.1–24.2) (12.3–29.2) (12.7–20.2) Birth place Catalonia 16.7 11.9 52.8 27.2 21.2 22.1 20.2 16.9 (3.5–44.6) (8.5–16.5) (2.9–112.7)⁎ (17.5–42.1) (12.4–33.6) (18.6–26.4) (8.9–36.2) (12.6–22.6) Rest of Spain 15.3 13.8 45.0 16.7 21.8 18.2 18.6 14.5 (5.5–24.6) (8.0–23.9) (1.9–75.3) (8.1–34.7) (12.9–26.1) (13.6–24.4) (14.4–25.6) (8.9–23.5) Abroad 25.4 16.6 76.9 30.5 27.5 25.4 23.9 13.9 (16.9–43.0) (10.0–27.3) (5.9–103.3) (15.7–59.2) (17.4–39.5) (19.4–33.2) (8.6–43.0) (8.9–21.6) Occupational social class I-II (more affluent) 21.2 13.9 46.4 22.7 21.8 19.4 17.8 15.1 (6.4–40.3) (9.5–20.4) (2.5–93.5)⁎ (13.9–37.2) (10.4–30.4) (15.9–23.7) (9.9–29.1) (10.8–21.1) III 25.2 14.8 88.4 48.2 27.3 28.1 27.1 20.4 (2.9–51.1) (9.1–24.2) (25.0–137.7) (25.6–90.9) (18.0–38.6) (21.7–36.4)⁎⁎ (13.7–38.5) (13.3–31.4) IV-V (less affluent) 17.0 11.5 41.0 16.4 20.9 21.6 18.9 13.1 (5.5–28.4) (7.4–17.9) (1.7–78.4) (9.3–29.1) (13.4–35.1) (17.1–27.3) (6.9–29.0) (8.9–19.3) (continued on next page) 11 M. Porta, J. Pumarega, L.A. Henríquez-Hernández et al. Science of the Total Environment 777 (2021) 146013 Table 5 (continued) Fluorene Phenanthrene Pyrene Fluoranthene Median aGM Median aGM Median aGM Median aGM (P25-P75) (CI 95%) (P25-P75) (CI 95%) (P25-P75) (CI 95%) (P25-P75) (CI 95%) Breastfed her children Never 44.7 29.6 72.7 44.9 26.4 25.9 24.0 13.9 (11.0–52.7) (11.8–74.0) (14.6–211.7) (13.6–148.6) (15.2–48.2) (16.3–41.3) (5.7–36.5) (6.0–32.1) ≤6 months 20.6 13.1 58.3 (9.4–86.0) 27.5 26.2 23. 24.1 16.5 (8.4–25.9) (7.9–21.8) (14.2–53.3) (17.9–39.5) (18.3–30.7) (14.2–38.6) (10.4–26.2) >6 months 19.2 11.4 53.4 21.5 19.4 19.9 16.1 13.5 (5.4–31.8) (7.8–16.6) (2.5–93.4) (13.2–35.0) (13.1–27.4) (16.4–24.0) (7.3–26.7) (9.6–19.1) aGM: Geometric mean adjusted for age. N = 128. a Reference category (except where otherwise noted, the reference category is the first category mentioned above). ⁎ p-value <0.05 (median, Kruskal-Wallis test). ⁎⁎ p-value <0.05 (aGM, Wald test; compared against the reference category). could be made integrating empirical information on actual levels and population-based study in Sweden (Bjermo et al., 2017), while in the trends of POP exposure during the specific historical periods preceding BHS of 2016, BDE-153 was detected in 24% of participants, being the each analysis (Porta, 2018; Henríquez-Hernández et al., 2020; Porta most detected PBDE compound. et al., 2008; Li et al., 2020; Wolff et al., 2005; Wolff et al., 2007; Wolff p,p’-DDT was detected (by IUIBS) in only 3% of the 240 samples from and Teitelbaum, 2020). 2016, whereas in 2006 it was detected (by IQAB-CSIC) in 97% of the 231 The reduction in mean concentrations of most POPs was not inevita- samples (median, 0.14 ng/mL) (Porta et al., 2012a). The nature of this bly to be accompanied by a substantial reduction in the number of par- difference is uncertain because p,p’-DDT was not detected in the 30 ticipants with a high number of POPs at high concentrations (Porta, samples from 2006 re-analyzed by IUIBS in 2016 (see 2.2. Analytical 2004; Porta et al., 2012b; Pumarega et al., 2016), but in this study it chemical methods). The long period of storage could explain the differ- was: in the 10 years, the percent of participants with concentrations ence. DDT can undergo dechlorination in frozen blood, with the subse- of 3 or more and 6 or more POPs each at high concentrations decreased quent increase in the concentrations of its metabolites (DDE and by 52% and 70%, respectively. Again, good news. DDD) (Ecobichon and Saschenbrecker, 1967). DDT was detected in 4% There are local studies, like those carried out in areas where people are of 447 newborns from the island of La Palma (Spain) (Cabrera- exposed to polluted environments or the emissions of industrial facilities, Rodríguez et al., 2019), in 23% of 121 adults from Romania, a country that have monitored different kinds of pollutants, including POPs, and that only recently banned DDT (Luzardo et al., 2019), or in 13% of which also support the main conclusions of the present study. When in 1135 adolescents from Germany (Bandow et al., 2020). Such currently 2005 a municipal solid waste plant started to operate in Bilbao (Basque observed percentages contrast with those observed more than Country, Spain), a biomonitoring project identified and quantified POPs 10 years ago, with DDT detection frequencies close to 100% in some from 2006 to 2013 in three cross-sectional studies, including 127 adults cases (Porta et al., 2008; Schoeters et al., 2017; Saoudi et al., 2014). prospectively monitored, living in areas near to and further away from While there have been significant improvements in the analytical tech- the plant. OCs decreased significantly during the study period: 80% niques over the last decades, there remains a need for re-assessment of (PCDD/Fs), 80% (dl-PCBs), and 37% (marker-PCBs) (Zubero et al., 2017). historical samples. The two PAHs most frequently detected in 2006 samples, naphtha- Compared to citizens born in Spain, citizens born abroad had higher lene and phenanthrene, showed decreases in concentrations greater concentrations of p,p’-DDE and lower of HCB and PCBs. These facts have than 90% and 30%, respectively. They were detected in 83% and 73% of also been observed in different cohorts in Spain (Ibarluzea et al., 2011). participants, respectively, in 2006, and in 45% and 76%, respectively, in Citizens from Latin America represent a high percentage of citizens born 2016. While in 2006 fluorene, pyrene and fluoranthene were not de- abroad residing in Barcelona. The specific country of origin of citizens tected in anybody, in 2016 they were detected in 85%–99% of citizens. born abroad and the duration of their stay in the city are key to under- These changes were significant for fluorene, pyrene, fluoranthene, and standing differences in concentrations with the rest of the population naphthalene, but not for phenanthrene. Uses and physicochemical char- of Barcelona. In comparison to Europe, where OCPs were banned in acteristics of PAHs may explain these variations. Although PAHs are the mid- to late 1970s, in Latin American countries they were banned rightly considered as POPs (EEA - European Environment Agency, later, and in many their use is still allowed, albeit with restrictions 2019), PAHs have a lower liposolubility coefficient than other OCs, (UNEP, 2002). Conversely, although HCB was formerly used as fungi- they are partially metabolized in the organism (which limits their accu- cide, non-dietary sources of HCB and PCBs are more common in indus- mulation), and have less resistance to degradation in the environment trialized areas (ATSDR, 2020). Also, concentrations of OCPs were higher (Mumtaz et al., 1996). Despite this, their constant presence in the envi- in a group of 131 subjects from Morocco than in 100 subjects from Ca- ronment and the continuous exposure of humans to them (mainly nary Islands, a region of Spain 100 km off the coast of Morocco through diet) (Yebra-Pimentel et al., 2015), make PAHs relevant for bio- (Henríquez-Hernández et al., 2016), demonstrating the influence of so- monitoring studies, where they are frequently detected. Thus, phenan- cial and life conditions in the concentrations of OCPs in two populations threne and pyrene were the most frequently detected PAHs (both in that are geographically close. more than 55% of 447 newborns) in the island of La Palma (Spain); Studies biomonitoring time trends in human contamination from naphthalene was the PAH detected at highest concentration (Cabrera- environmental chemical agents may or not be representative of the Rodríguez et al., 2019). A similar profile was reported in 121 target population; be more or less limited to a range of ages, socio- Romanian subjects (Luzardo et al., 2019). In a series of 121 women economic positions and geographies; include larger or smaller num- from China, fluorene, phenanthrene, pyrene and fluoranthene showed bers of individuals, chemicals, and periods of sample collection; frequencies of detection greater than 95% (Wang et al., 2015). integrate or not personal and social predictors of the concentrations; In the present study the percentages of detection of PBDEs were and be linked or not to institutions and policies to decrease exposure. lower than in other studies in Europe; e.g. the HELIX study detected They must, in any case, use meaningful biomarkers and, critically, BDE-153 in 73% of participating mothers (Montazeri et al., 2019), and apply accurate procedures for sample collection, processing, analy- the same compound was detected in all 170 participants in a sis, and quality assurance (Henríquez-Hernández et al., 2020; Porta 12 M. Porta, J. Pumarega, L.A. Henríquez-Hernández et al. Science of the Total Environment 777 (2021) 146013 et al., 2008; Porta, 2004; Angerer et al., 2007; Schulz et al., 2011; Acknowledgments Bjerregaard-Olesen et al., 2017; Buekers et al., 2018; Calafat, 2016; Joas et al., 2017; Haines et al., 2017; Li et al., 2020). We think our The work was supported in part by research grants from Instituto de study fulfils the latter set of requirements (Luzardo et al., 2019; Salud Carlos III, Ministry of Health, Government of Spain – co-funded by Henríquez-Hernández et al., 2017). We also think we were reason- FEDER (FIS PI13/00020, FIS PI17/00088, and CIBER de Epidemiología y ably rigorous and pragmatic to maximize the comparability of chem- Salud Pública - CIBERESP); the Hospital del Mar Medical Research Insti- ical analyses performed in 2006 and 2016 in the two laboratories tute (IMIM), Barcelona; the Government of Catalonia (2014 SGR 1012, (e.g., re-analyzing at IUIBS in 2016, 30 samples from the 2006 2017 SGR 439); Fundació La Marató de TV3 (20132910); CRUE- BHS). It is often not possible to use the same laboratory in different Santander Fondo Supera Covid-19 (15072020). UB was supported by studies; even when it is, protocols change (Porta et al., 2008; Porta, the Office of the Director of the National Institutes of Health under 2004; Angerer et al., 2007; Schulz et al., 2011; Bjerregaard-Olesen award number DP5OD26429. The authors gratefully acknowledge tech- et al., 2017; Buekers et al., 2018; Calafat, 2016; Joas et al., 2017; nical and scientific assistance provided by Laura Sánchez, Natàlia Dennis et al., 2017; Haines et al., 2017; Li et al., 2020; Nøst et al., Pallarès, Marc Domínguez, and Yolanda Rovira. Special thanks to Paco 2019; Schoeters et al., 2017; Schröter-Kermani et al., 2013; Bolúmar. Stubleski et al., 2018). Therefore, we think the reported decreases in POP concentrations have a high degree of validity. References Our study used a cross-sectional design to study a representative Amrhein, V., Greenland, S., McShane, B., 2019. Scientists rise up against statistical signifi- sample of the target population, it included a small number of individ- cance. Nature 567, 305–307. uals (471 in total) from 18 to 80 years of age and of all socioeconomic Angerer, J., Ewers, U., Wilhelm, M., 2007. Human biomonitoring: state of the art. Int. positions, it measured a fair number of chemicals (62) in two periods, J. Hyg. Environ. Health 210, 201–228. it integrated personal and social predictors of the concentrations, and Arrebola, J.P., Martin-Olmedo, P., Fernandez, M.F., Sanchez-Cantalejo, E., Jimenez-Rios, J.A., Torne, P., et al., 2009. Predictors of concentrations of hexachlorobenzene in human it is linked to policies developed by the Barcelona Agency for Public adipose tissue: a multivariate analysis by gender in southern Spain. Environ. Int. 35, Health. The common presence of results on POP studies in social net- 27–32. works and media has likely influenced citizens' behaviors, and policies ASPB, 2020. Agència de Salut Pública de Barcelona. Enquestes de salut. Barcelona, ASPB Available at:. https://www.aspb.cat/arees/la-salut-en-xifres/enquestes-de-salut/. of public and private organizations. (Accessed 12 December 2020). Finally, we can look further backwards to concentrations we mea- ATSDR - Agency for Toxic Substances and Disease Registry, 2020. Toxic Substances Portal. sured in 2002 in the first representative sample of citizens of Barce- https://www.atsdr.cdc.gov/substances/index.asp. Accessed 12 Dec 2020. lona:(Porta et al., 2012a) concentrations decreased markedly between Bandow, N., Conrad, A., Kolossa-Gehring, M., Murawski, A., Sawal, G., 2020. Polychlorinated biphenyls (PCB) and organochlorine pesticides (OCP) in blood 2002 and 2016, with changes of −72% for p,p’-DDE, −86% for HCB, plasma - results of the German environmental survey for children and adolescents −89% for β-HCH, −92% for PCB 118, and of more than −60% for the 2014-2017 (GerES V). Int. J. Hyg. Environ. Health 224, 113426. other three main PCBs (138, 153 and 180). Bartoll, X., Pérez, K., Pasarín, M., Rodríguez-Sanz, M., Borrell, C., 2018. Resultats de l’Enquesta de Salut de Barcelona 2016/17. Barcelona, Agència de Salut Pública de Bar- celona Available:. https://www.aspb.cat/wp-content/uploads/2018/12/ASPB_ 5. Conclusions Enquesta-Salut-Barcelona-2016.pdf. (Accessed 12 December 2020). Bjermo, H., Aune, M., Cantillana, T., Glynn, A., Lind, P.M., Ridefelt, P., et al., 2017. Serum While concentrations of most POPs are decreasing in Barcelona, sig- levels of brominated flame retardants (BFRs: PBDE, HBCD) and influence of dietary factors in a population-based study on Swedish adults. Chemosphere 167, 485–491. nificant sociodemographic differences in such reductions warrant Bjerregaard-Olesen C, Long M, Ghisari M, Bech BH, Nohr EA, Uldbjerg N, et al. Temporal strengthening public and private policies towards groups making trends of lipophilic persistent organic pollutants in serum from Danish nulliparous slower progress. A relevant component of the success in the current de- pregnant women 2011-2013. Environ. Sci. Pollut. Res. Int. 2017, 24, 16592–16603. Buekers, J., David, M., Koppen, G., Bessems, J., Scheringer, M., Lebret, E., et al., 2018. Devel- creasing is a reduction of differences (convergence) by gender. For sev- opment of policy relevant human biomonitoring indicators for chemical exposure in eral pollutants the decrease was larger in the younger groups, mostly the European population. Int. J. Environ. Res. Public Health 15, 2085. with monotonic trends, a finding coherent with the possibility that Cabrera-Rodríguez, R., Luzardo, O.P., Almeida-González, M., Boada, L.D., Zumbado, M., younger cohorts are being less exposed to POPs. Even in a society with Acosta-Dacal, A., et al., 2019. Association between prenatal exposure to multiple per- sistent organic pollutants (POPs) and growth indicators in newborns. Environ. Res. decreasing levels, reductions were less frequent in individuals with obe- 171, 285–292. sity than with normal weight. Calafat, A.M., 2016. Contemporary issues in exposure assessment using biomonitoring. Supplementary data to this article can be found online at https://doi. Curr. Epidemiol. Rep. 3 (2), 145–153. org/10.1016/j.scitotenv.2021.146013. Carreño, J., Rivas, A., Granada, A., Lopez-Espinosa, M.J., Mariscal, M., Olea, N., et al., 2007. Exposure of young men to organochlorinepesticides in southern Spain. Environ. Res. 103, 55–61. Funding Demeneix, B., 2017. Toxic Cocktail. How Chemical Pollution is Poisoning our Brains. Ox- ford University Press, New York. Dennis, K.K., Marder, E., Balshaw, D.M., Cui, Y., Lynes, M.A., Patti, G.J., et al., 2017. Biomon- The work was supported in part by research grants from Instituto de itoring in the era of the exposome. Environ. Health Perspect. 125 (4), 502–510. Salud Carlos III, Ministry of Health, Government of Spain – co-funded by Ecobichon, D.J., Saschenbrecker, P.W., 1967. Dechlorination of DDT in frozen blood. Sci- FEDER (FIS PI13/00020, FIS PI17/00088, and CIBER de Epidemiología y ence 156, 663–665. Salud Pública - CIBERESP); the Hospital del Mar Medical Research Institute EEA - European Environment Agency. Persistent organic pollutant emissions. Indicator assesment. Copenhagen, EEA, 2019. Available at: https://bit.ly/35hkcoQ. Accessed (IMIM), Barcelona; the Government of Catalonia (2014 SGR 1012, 2017 12 Dec 2020. SGR 439); Fundació La Marató de TV3 (20132910); and the CRUE- Fernyhough, L., Horwath, C., Campbell, A.J., Robertson, M.C., Busby, W.J., 1999. Changes in Santander Fondo Supera Covid-19 (15072020). UB was supported by dietary intake during a 6-year follow-up of an older population. Eur. J. Clin. Nutr. 53, 216–225. the Office of the Director of the National Institutes of Health under Gasull, M., 2019. Analysis of the Contamination of the General Population by Persistent award number DP5OD26429. The funders had no role in study design, Toxic Compounds and of some of their Adverse Health Effects [Doctoral Dissertation]. data collection and analysis, decision to publish, or preparation of the Universitat Autònoma de Barcelona, Barcelona Available:. http://www.imim.es/ manuscript. programesrecerca/epidemiologia/es_documentsgrecm.html. (Accessed 12 December 2020). Gasull, M., Pumarega, J., Rovira, G., López, T., Alguacil, J., Porta, M., 2013. Relative effects of Declaration of competing interest educational level and occupational social class on body concentrations of persistent organic pollutants in a representative sample of the general population of Catalonia. The authors declare that they have no known competing financial Spain. Environ. Int. 60, 190–201. interests or personal relationships that could have appeared to influ- Gore, A.C., Chappell, V.A., Fenton, S.E., Flaws, J.A., Nadal, A., Prins, G.S., et al., 2015. EDC-2: the Endocrine Society’s second scientific statement on endocrine-disrupting ence the work reported in this paper. chemicals. Endocr. Rev. 36, E1–E150. 13 M. Porta, J. Pumarega, L.A. Henríquez-Hernández et al. Science of the Total Environment 777 (2021) 146013 Haines, D.A., Saravanabhavan, G., Werry, K., Khoury, C., 2017. An overview of human bio- Porta, M., Puigdomènech, E., Ballester, F., Selva, J., Ribas-Fitó, N., Llop, S., et al., 2008. Mon- monitoring of environmental chemicals in the Canadian Health Measures Survey: itoring concentrations of persistent organic pollutants in the general population: the 2007–2019. Int. J. Hyg. Environ. Health 220 (2 Pt A), 13–28. international experience. Environ. Int. 34, 546–561. Henkler, F., Luch, A., 2011. Adverse health effects of environmental chemical agents Porta, M., Gasull, M., Puigdomènech, E., Rodríguez-Sanz, M., Pumarega, J., Rebato, C., et al., through non-genotoxic mechanisms. J. Epidemiol. Community Health 65, 1–3. 2009. Sociodemographic factors influencing participation in the Barcelona Health Henríquez-Hernández, L.A., Luzardo, O.P., Arellano, J.L.P., Carranza, C., Sánchez, N.J., Survey study on serum concentrations of persistent organic pollutants. Chemosphere Almeida-González, M., et al., 2016. Different pattern of contamination by legacy 76, 216–225. POPs in two populations from the same geographical area but with completely differ- Porta, M., Gasull, M., Puigdomènech, E., Garí, M., 2010. Bosch de Basea M, Guillén M, et al. ent lifestyles: Canary Islands (Spain) vs. Morocco. Sci. Total Environ. 541, 51–57. Distribution of blood concentrations of persistent organic pollutants in a representa- Henríquez-Hernández, L.A., Luzardo, O.P., Zumbado, M., Serra-Majem, L., Valerón, P.F., tive sample of the population of Catalonia. Environ. Int. 36, 655–664. Camacho, M., et al., 2017. Determinants of increasing serum POPs in a population Porta, M., López, T., Gasull, M., Rodríguez-Sanz, M., Garí, M., Pumarega, J., et al., 2012a. Dis- at high risk for cardiovascular disease. Environ. Res. 156, 477–484. tribution of blood concentrations of persistent organic pollutants in a representative Henríquez-Hernández, L.A., Ortiz-Andrelluchi, A., Álvarez-Pérez, J., Acosta-Dacal, A., sample of the population of Barcelona in 2006, and comparison with levels in 2002. Zumbado. M., Martínez-González. M.A., et al. Human biomonitoring of persistent or- Sci. Total Environ. 423, 151–161. ganic pollutants in elderly people from the Canary Islands (Spain): a temporal trend Porta, M., Pumarega, J., Gasull, M., 2012b. Number of persistent organic pollutants de- analysis from the PREDIMED and PREDIMED-plus cohorts. Sci. Total Environ. 2020 tected at high concentrations in a general population. Environ. Int. 44, 106–111. Nov 20:143637. doi: https://doi.org/10.1016/j.scitotenv.2020.143637. Pumarega, J., Gasull, M., Lee, D.H., López, T., Porta, M., 2016. Number of persistent organic Ibarluzea, J., Alvarez-Pedrerol, M., Guxens, M., Marina, L.S., Basterrechea, M., Lertxundi, A., pollutants detected at high concentrations in blood samples of the United States pop- et al., 2011. Sociodemographic, reproductive and dietary predictors of organochlorine ulation. PLoS One 11 (8), e0160432. https://journals.plos.org/plosone/article?id= compounds levels in pregnant women in Spain. Chemosphere 82, 114–120. 10.1371/journal.pone.0160432. Joas, A., Schwedler, G., Choi, J., Kolossa-Gehring, M., 2017. Human biomonitoring: Science Quinn, C.L., Wania, F., 2012. Understanding differences in the body burden-age relation- and policyfor a healthy future, April 17–19, 2016, Berlin, Germany. Int. J. Hyg. Envi- ships of bioaccumulating contaminants based on population cross sections versus in- ron. Health. 220 (2 Pt A), 299–304. dividuals. Environ. Health Perspect. 120, 554–559. Kassotis, C.D., Vandenberg, L.N., Demeneix, B., Porta, M., Slama, R., Trasande, L., 2020. En- Saoudi, A., Fréry, N., Zeghnoun, A., Bidondo, M.L., Deschamps, V., Göen, T., et al., 2014. docrine disrupting chemicals: economic, regulatory, and policy implications. Lancet Serum levels of organochlorine pesticides in the French adult population: the French Diabetes Endocrinol. 8, 719–730. National Nutrition and Health Study (ENNS), 2006-2007. Sci. Total Environ. 472, Keith, L.H., Telliard, W.A., 1979. Priority pollutants. I. A perspective view. Environ. Sci. 1089–1099. Technol. 13, 416–423. Schoeters G, Govarts E, Bruckers L, Den Hond E, Nelen V, De Henauw S, et al. Three cycles Kleinbaum, D.G., Kupper, L.L., Muller, K.E., Nizam, A., 2007. Applied Regression Analysis of human biomonitoring in Flanders - Time trends observed in the Flemish Environ- and Multivariable Methods. 4th. ed. Duxbury, Pacific Grove, CA. ment and Health Study. Int. J. Hyg. Environ. Health 2017, 220 (2 Pt A), 36–45. LaMerrill, M.A., Vandenberg, L.N., Smith, M.T., Goodson, W., Browne, P., Patisaul, H.B., et Schröter-Kermani, C., Müller, J., Jürling, H., Conrad, A., Schulte, C., 2013. Retrospective al., 2020. Consensus on the key characteristics of endocrine-disrupting chemicals as monitoring of perfluorocarboxylates and perfluorosulfonates in human plasma ar- a basis for hazard identification. Nature Rev. Endocrinol. 16, 45–57. chived by the German Environmental Specimen Bank. Int. J. Hyg. Environ. Health Lash TL, VanderWeele TJ, Haneuse S, Rothman KJ, Eds. Modern Epidemiology, 4th. ed., 216 (6), 633–640. Philadelphia, Walters-Kluwer, 2021. Schulz, C., Wilhelm, M., Heudorf, U., Kolossa-Gehring, M., 2011. Update of the reference Lee, D.H., Porta, M., Jacobs, D.R., Vandenberg, L.N., 2014. Chlorinated persistent organic and HBM values derived by the German Human Biomonitoring Commission. Int. pollutants, obesity, and type 2 diabetes. Endocr. Rev. 35, 557–601. J. Hyg. Environ. Health 215, 26–35. Li, L., Hoang, C., Arnot, J.A., Wania, F., 2020. Clarifying temporal trend variability in human Song, S., Ma, X., Pan, M., Tong, L., Tian, Q., 2018. Excretion kinetics of three dominant or- biomonitoring of polybrominated diphenyl ethers through mechanistic modeling. ganochlorine compounds in human milk within the first 6 months postpartum. Envi- Environ. Sci. Technol. 54, 166–175. ron. Monit. Assess. 190, 457. Loibner, A.P., Szolar, O.H.J., Braun, R., Hirmann, D., 2004. Toxicity testing of 16 priority Sotaniemi, E.A., Arranto, A.J., Pelkonen, O., Pasanen, M., 1997. Age and cytochrome P450- polycyclic aromatic hydrocarbons using Lumistox®. Environ. Toxicol. Chem. 23, linked drug metabolism in humans: an analysis of 226 subjects with equal histopath- 557–564. ologic conditions. Clin. Pharmacol. Ther. 61, 331–339. Luzardo, O.P., Badea, M., Zumbado, M., Rogozea, L., Floroian, L., Ilea, A., et al., 2019. Body Stubleski, J., Lind, L., Salihovic, S., Lind, P.M., Kärrman, A., 2018. Longitudinal changes in burden of organohalogenated pollutants and polycyclic aromatic hydrocarbons in persistent organic pollutants (POPs) from 2001 to 2009 in a sample of elderly Swed- Romanian population: influence of age, gender, body mass index, and habitat. Sci. ish men and women. Environ. Res. 165, 193–200. Total Environ. 656, 709–716. Trasande, L. Sicker, Fatter, Poorer: The Urgent Threat of Hormone-Disrupting Chemicals to Marraudino M, Bonaldo B, Farinetti A, Panzica G, Ponti G, Gotti S. Metabolism disrupting Our Health and Future... And What We Can Do About It. New York: Houghton, Mifflin, chemicals and alteration of neuroendocrine circuits controlling food intake and en- Harcourt; 2019. Reviewed in: Porta M. am. J. Public. Health 2020, 110(4), 423–424 ergy metabolism. Front. Endocrinol. (Lausanne) 2019, 9, 766. https://ajph.aphapublications.org/doi/10.2105/AJPH.2019.305560. Montazeri, P., Thomsen, C., Casas, M., de Bonta, J., Haug, L.S., Maitre, L., et al., 2019. Socio- UNEP - United Nations Environment Program, 2002. Chemicals. Regionally-Based Assess- economic position and exposure to multiple environmental chemical contaminants ment of Persistent Toxic Substances. Regional Report, Central America and the Carib- in six European mother-child cohorts. Int. J. Hyg. Environ. Health 222, 864–872. bean. Geneva, United Nations Environment Programme, 2002. https://bit.ly/ Mumtaz, M.M., George, J.D., Gold, K.W., Cibulas, W., DeRosa, C.T., 1996. ATSDR evaluation 3obgCVZ. Accessed 12 Dec 2020. of health effects of chemicals. IV. Polycyclic aromatic hydrocarbons (PAHs): under- UNEP - United Nations Environment Program, 2020. Stockholm, Rotterdam, and Basel standing a complex problem. Toxicol. Ind. Health 12, 742–971. Conventions. http://www.pops.int. Accessed 12 Dec 2020. Nøst T.H., Berg V., Hanssen L., Rylander C., Gaudreau E., Dumas P., et al. Time trends of per- Wang, B., Jin, L., Ren, A., Yuan, Y., Liu, J., Li, Z., et al., 2015. Levels of polycyclic aromatic hy- sistent organic pollutants in 30 year olds sampled in 1986, 1994, 2001 and 2007 in drocarbons in maternal serum and risk of neural tube defects in offspring. Environ. northern Norway: measurements, mechanistic modeling and a comparison of study Sci. Technol. 49, 588–596. designs. Environ. Res. 2019, 172, 684–692. Wolff, M.S., Teitelbaum, S.L., 2020. Using BMI as a chronometer for persistent chemical ex- Partearroyo, T., Samaniego-Vaesken, M.L., Ruiz, E., Aranceta-Bartrina, J., Gil, Á., González- posures and chronic disease. Environ. Res. 193, 110588. https://doi.org/10.1016/j. Gross, M., et al., 2019. Current food consumption amongst the Spanish ANIBES study envres.2020.110588. 33307085. population. Nutrients 11, 2663. Wolff, M., Britton, J.A., Teitelbaum, S.L., Eng, S., Deych, E., Ireland, K., 2005. Liu et al. im- Peinado FM, Artacho-Cordón F, Barrios-Rodríguez R, Arrebola JP. Influence of proving organochlorine biomaker models for cancer research. Cancer Epidemiol. Bio- polychlorinated biphenyls and organochlorine pesticides on the inflammatory milieu. markers Prev. 14, 2224–2236. A systematic review of in vitro, in vivo and epidemiological studies. Environ. Res. Wolff, M.S., Anderson, H.A., Britton, J.A., Rothman, N., 2007. Pharmacokinetic variability 2020, 186, 109561. and modern epidemiology–the example of dichlorodiphenyltrichloroethane, body Porta, M., 2004. Persistent toxic substances: exposed individuals and exposed popula- mass index, and birth cohort. Cancer Epidemiol. Biomark. Prev. 16, 1925–1930. tions. J. Epidemiol. Community Health 58, 534–535. https://doi.org/10.1136/ Yebra-Pimentel, I., Fernández-González, R., Martínez-Carballo, E., Simal-Gándara, J., 2015. jech.2004.021238. A critical review about the health risk assessment of PAHs and their metabolites in Porta, M., March 2012. Baja la ‘contaminación interior’. El País 13, 36. https://elpais.com/ foods. Crit. Rev. Food Sci. Nutr. 55, 1383–1405. sociedad/2012/03/12/actualidad/1331563333_025368.html. Zubero, M.B., Eguiraun, E., Aurrekoetxea, J.J., Lertxundi, A., Abad, E., Parera, J., et al., 2017. Porta, M., 2018. Vive Más y Mejor Reduciendo Tóxicos y Contaminantes Ambientales. Bar- Changes in serum dioxin and PCB levels in residents around a municipal waste incin- celona, Grijalbo - Penguin Random House https://www.megustaleer.com/libros/vive- erator in Bilbao. Spain. J. Environ. Res. 156, 738–746. ms-y-mejor/MES-083079. Porta, M., Zumeta, E., 2002. Implementing the Stockholm treaty on POPs [editorial]. Occup. Environ. Med. 59, 651–652. https://bit.ly/2VBv6Q4. 14