Environmental Pollution 316 (2023) 120566
Available online 2 November 20220269-7491/© 2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Cross-sectional associations between exposure to per- and polyfluoroalkyl 
substances and body mass index among European teenagers in the 
HBM4EU aligned studies☆ 
Tessa Schillemans a,*, Nina Iszatt b, Sylvie Remy c, Greet Schoeters c,d, Mariana F. Fernandez e,f, 
Shereen Cynthia D’Cruz g, Anteneh Desalegn h, Line S. Haug h, Sanna Lignell i, 
Anna Karin Lindroos i, Lucia Fabelova j, Lubica Palkovicova Murinova j, Tina Kosjek k, 
Ziga Tkalec k, Catherine Gabriel l,m, Denis Sarigiannis l,m,n, Susana Pedraza-Díaz o, 
Marta Esteban-Lopez o, Argelia Casta~no o, Loïc Rambaud p, Margaux Riou p, Sara Pauwels q, 
Nik Vanlarebeke r, Marike Kolossa-Gehring s, Nina Vogel s, Maria Uhl t, Eva Govarts c, 
Agneta Åkesson a 
a Unit of Cardiovascular and Nutritional Epidemiology, Institute of Environmental Medicine, Karolinska Institutet, Sweden 
b Division of Climate and Environmental Health, Norwegian Institute of Public Health, Norway 
c VITO Health, Flemish Institute for Technological Research (VITO), Mol, Belgium 
d Department of Biomedical Sciences, University of Antwerp, Antwerp, Belgium 
e Centre for Biomedical Research (CIBM) and School of Medicine, University of Granada, Granada, Spain 
f Spanish Consortium for Research on Epidemiology and Public Health (CIBERESP), Madrid, Spain 
g Univ Rennes, EHESP, Inserm, Irset (Institut de Recherche en Sante, Environnement et Travail), Rennes, France 
h Division of Food Safety, Norwegian Institute of Public Health, Norway 
i Swedish Food Agency, Uppsala, Sweden 
j Department of Environmental Medicine, Faculty of Public Health, Slovak Medical University, Bratislava, Slovakia 
k Department of Environmental Sciences, Jozef Stefan Institute, Ljubljana, Slovenia 
l Environmental Engineering Laboratory, Department of Chemical Engineering, Aristotle University of Thessaloniki, 54124, Thessaloniki, Greece 
m HERACLES Research Center on the Exposome and Health, Center for Interdisciplinary Research and Innovation, Balkan Center, Bldg. B, 10th Km Thessaloniki-Thermi 
Road, 57001, Greece 
n Environmental Health Engineering, Institute of Advanced Study, Palazzo Del Broletto - Piazza Della Vittoria 15, 27100, Pavia, Italy 
o National Centre for Environmental Health, Instituto de Salud Carlos III, Madrid, Spain 
p Department of Environmental and Occupational Health, Sante Publique France, Saint-Maurice, France 
q Department of Public Health and Primary Care, KU, Leuven, Belgium 
r Department of Analytical and Environmental Chemistry, Free University of Brussels, Belgium 
s German Environment Agency, Umweltbundesamt (UBA), Berlin, Germany 
t Environment Agency Austria, Vienna, Austria   
A R T I C L E  I N F O   
Keywords: 
per-and polyfluoroalkyl substances 
Body mass index 
Teenagers 
HBM4EU 
A B S T R A C T   
Per- and polyfluoroalkyl substances (PFAS) are widespread pollutants that may impact youth adiposity patterns. 
We investigated cross-sectional associations between PFAS and body mass index (BMI) in teenagers/adolescents 
across nine European countries within the Human Biomonitoring for Europe (HBM4EU) initiative. We used data 
from 1957 teenagers (12–18 yrs) that were part of the HBM4EU aligned studies, consisting of nine HBM studies 
(NEBII, Norway; Riksmaten Adolescents 2016–17, Sweden; PCB cohort (follow-up), Slovakia; SLO CRP, Slovenia; 
CROME, Greece; BEA, Spain; ESTEBAN, France; FLEHS IV, Belgium; GerES V-sub, Germany). Twelve PFAS were 
measured in blood, whilst weight and height were measured by field nurse/physician or self-reported in ques-
tionnaires. We assessed associations between PFAS and age- and sex-adjusted BMI z-scores using linear and 
logistic regression adjusted for potential confounders. Random-effects meta-analysis and mixed effects models 
were used to pool studies. We assessed mixture effects using molar sums of exposure biomarkers with 
☆ This paper has been recommended for acceptance by Payam Dadvand. 
* Corresponding author. Institute of Environmental Medicine, Karolinska Institutet, Box 210, SE-171 77 Stockholm Visiting Address: Nobels vag 13, Solna, Sweden. 
E-mail address: tessa.schillemans@ki.se (T. Schillemans).  
Contents lists available at ScienceDirect 
Environmental Pollution 
journal homepage: www.elsevier.com/locate/envpol 
https://doi.org/10.1016/j.envpol.2022.120566 
Received 17 June 2022; Received in revised form 28 October 2022; Accepted 29 October 2022   
Environmental Pollution 316 (2023) 120566
2
toxicological/structural similarities and quantile g-computation. In all studies, the highest concentrations of 
PFAS were PFOS (medians ranging from 1.34 to 2.79 μg/L). There was a tendency for negative associations with 
BMI z-scores for all PFAS (except for PFHxS and PFHpS), which was borderline significant for the molar sum of 
[PFOA and PFNA] and significant for single PFOA [β-coefficient (95% CI) per interquartile range fold change 
  0.06 (  0.17, 0.00) and   0.08 (  0.15,   0.01), respectively]. Mixture assessment indicated similar negative 
associations of the total mixture of [PFOA, PFNA, PFHxS and PFOS] with BMI z-score, but not all compounds 
showed associations in the same direction: whilst [PFOA, PFNA and PFOS] were negatively associated, [PFHxS] 
associated positively with BMI z-score. Our results indicated a tendency for associations of relatively low PFAS 
concentrations with lower BMI in European teenagers. More prospective research is needed to investigate this 
potential relationship and its implications for health later in life.   
1. Introduction 
Childhood and adolescent adiposity patterns are risk factors for 
obesity, hypertension, type 2 diabetes, atherosclerosis and cardiovas-
cular disease in adulthood (Reilly and Kelly, 2011). The prevalence of 
obesity worldwide in children, adolescents and adults has increased over 
the past decades and this trend will likely continue (Ng et al., 2014), 
painting a dangerous picture for future population health. Youth 
adiposity patterns may be impacted by exposure to obesogenic chem-
icals (Grün and Blumberg, 2009), e.g. per- and polyfluoroalkyl sub-
stances (PFAS). PFAS are a human-made group of chemicals that are 
extremely widespread, environmentally persistent and almost omni-
present in humans with generally long half-lives of 3–5 years (Lau et al., 
2007) (2–5 years for PFOS and PFOA and 4–5 years for shorter chain 
PFHxS and PFHpS (Li et al., 2022)). Generally, long-chain, sulfonated 
and linear isomer PFAS have slower excretion rates than shorter-chain, 
carboxylated and major branched isomer PFAS (Zhang et al., 2013). 
Experimental studies indicate that PFAS induces perturbations in 
pathways relevant to metabolism, e.g. peroxisome proliferator activated 
receptor (PPAR) activation (Bijland et al., 2011; Vanden Heuvel et al., 
2006). In addition, PFAS exposure associates with unfavorable lipid 
profile changes as reviewed by the European Food Safety Agency (EFSA 
et al., 2018) as well as with metabolic dysfunction (Margolis and Sant, 
2021). Other potential mechanisms for an obesogenic effect of PFAS 
exposure in youth include endocrine disruption (Du et al., 2013), dys-
lipidemia (Bijland et al., 2011) and inflammation (Takacs and Abbott, 
2007). However, epidemiological studies investigating associations be-
tween prenatal PFAS exposures and youth adiposity patterns are 
inconclusive, reporting positive (Braun et al., 2016; Chen et al., 2019; 
Lauritzen et al., 2018; Liu et al., 2020; Mora et al., 2017), negative 
(Braun et al., 2021; Hartman et al., 2017; Starling et al., 2019; Starling 
et al., 2017) and null (Bloom et al., 2021; Manzano-Salgado et al., 2017; 
Martinsson et al., 2020) associations. Similarly, cross-sectional studies 
with postnatal PFAS exposure during childhood or adolescence are 
equivocal, reporting positive (Averina et al., 2021; Geiger et al., 2021), 
negative (Fassler et al., 2019; Thomsen et al., 2021) and null (Averina 
et al., 2021; Thomsen et al., 2021) associations. One longitudinal study 
indicated childhood perfluorooctane sulfonic acid (PFOS) exposure 
associated with higher adolescent body mass index (BMI), but adoles-
cent PFOS exposure did not associate with higher BMI in adulthood 
(Domazet et al., 2016). In addition to these inconsistent findings, only 
few studies have looked at the effect of mixtures (Janis et al., 2021; 
Vrijheid et al., 2020), whilst these findings could shed light on previous 
inconsistent findings and provide more insight in specific compound 
effects. 
Improved knowledge of preventable risk factors, including obeso-
genic chemicals, is imperative to improve population health and reduce 
disease risk later in life. Therefore, within the Horizon 2020 project 
‘HBM4EU’, the Human Biomonitoring for Europe initiative (see www. 
hbm4eu.eu for details (Ganzleben et al., 2017)), we aimed to investi-
gate associations of single PFAS exposures and PFAS mixtures with BMI 
in teenagers/adolescents (12–18 yrs old; hereafter referred to as teen-
agers). We used data from nine studies from different countries in 
Europe, resulting in a large study population (n  1957) with quality 
assured PFAS measurements. 
2. Methods 
2.1. Study population 
The study population was drawn from the ‘HBM4EU aligned studies’ 
(Gilles et al., 2021; Gilles et al., 2022). The HBM4EU aligned studies are 
a survey aimed at collecting HBM samples and data from European HBM 
studies to derive current internal exposure data for the European pop-
ulation across a wide geographic spread. For the present study, nine 
studies targeting teenagers (12–18 years) had available blood samples 
used for PFAS measurement: NEBII (Norwegian Environmental Biobank 
II; a sub study of The Norwegian Mother, Father and Child Cohort Study 
(MoBa) (Magnus et al., 2016); Norway), Riksmaten adolescents 
2016–17 (Sweden), PCB cohort (follow-up) (Endocrine disruptors and 
health in children and teenagers in Slovakia; Slovakia), SLO CRP 
(Exposure of children and adolescents to selected chemicals through 
their habitat environment; Slovenia), CROME (Cross-Mediterranean 
Environment and Health Network; Greece), BEA (Biomonitorizacion en 
Adolescentes; Spain), ESTEBAN (Etude de sante sur l’environnement, la 
biosurveillance, l’activite physique et la nutrition; France), FLEHS IV 
(Flemish Environment and Health Study IV; Belgium), GerES V-sub 
(German Environmental Survey, 2014–2017 unweighted subsample; 
Germany). All studies were national or regional cross-sectional pop-
ulation-based studies, except the longitudinal NEBII and PCB cohort 
(follow-up). In all studies, parents (and for some studies additionally the 
teenagers) have signed informed consent. 
Detailed information about each study, the study selection process 
and data homogenization within the HBM4EU has been described pre-
viously (Gilles et al., 2021; Gilles et al., 2022). In brief, recommenda-
tions for selecting participants were: 1) to have lived at least 5 years in 
the catchment area of the data collection, not be hospitalized or insti-
tutionalized and be between 12 and 19 years of age; 2) to have 
completed a questionnaire and have available serum or plasma samples 
with appropriate sample matrix and volume. Stratification of the par-
ticipants into mutually exclusive subgroups was applied in chronolog-
ical order and with a specified proportion: sex (50%  2% of each sex), 
degree of urbanization according to Eurostat categorization into cities, 
towns/suburbs, rural area (at least 10% of each of the three levels); 
household educational level based on the International Standard Clas-
sification of Education (ISCED, 2011) [low (ISCED 0–2), medium (ISCED 
3–4) and high (ISCED 5–8)] (at least 10% of each of the three levels); 
sampling season (approximately equal distribution over all seasons); age 
(all ages from the original data collection were present in the selection). 
Subsequently, if applicable, participants were randomly selected while 
adhering to the subgroup population proportions. This selection process 
resulted in a final study population of 1957 teenagers with PFAS 
measurements. 
2.2. Chemical analysis of PFAS 
The harmonization of chemical analysis has also been outlined by 
Gilles et al. (2021). In brief, analyses of six studies were performed in 
T. Schillemans et al.                                                                                                                                                                                                                            
Environmental Pollution 316 (2023) 120566
3
laboratories that successfully passed the HBM4EU quality assurance 
quality control (QA/QC) programme (Esteban Lopez et al., 2021; Nübler 
et al., 2022). Data for two studies, ESTEBAN and GerES V-sub, were 
generated before HBM4EU QA/QC programme, but were deemed 
comparable in quality (as determined by evaluation of analytical 
methods and proficiency tests) and were approved posterior. Data for 
Riksmaten Adolescents 2016-17 has been generated outside of the 
HBM4EU and the laboratory presented PFAS in ng/g (μg/kg), which was 
reported to HBM4EU in μg/L assuming that 1 ml blood serum equals 1 g 
blood serum. PFAS concentrations (linear form or sum of all isomers 
including linear and branched forms) were measured in serum in all 
studies, except for NEBII and GerES V-sub, which were measured in 
plasma. Liquid chromatography-tandem mass spectrometry was used in 
all studies, except in NEBII and Riksmaten Adolescents 2016–17 where 
ultraperformance liquid chromatography-tandem mass spectrometry 
was used (Supplemental Table 1). Twelve PFAS were assessed: per-
fluoropentanoic acid (PFPeA), perfluorohexanoic acid (PFHxA), per-
fluoroheptanoic acid (PFHpA), perfluorooctanoate (PFOA), 
perfluorononanoic acid (PFNA), perfluorodecanoic acid (PFDA), per-
fluoroundecanoic acid (PFUnDA), perfluorododecanoic acid (PFDoDA), 
perfluorobutane sulfonic acid (PFBS), perfluorohexane sulfonic acid 
(PFHxS), perfluoroheptanane sulfonic acid (PFHpS) and PFOS. 
PFAS values below the limit of quantification (LOQ) were imputed 
using single random imputation from a truncated lognormal distribu-
tion. We included PFAS measurements within a study if at least 70% of 
the values for a single PFAS were LOQ, to be able to accurately assess 
tertiles and continuous. PFDoDA and PFBS were below this threshold in 
all studies and not used for further analysis. Supplemental Table 1 
provides details of the PFAS measurements for the individual studies, 
including number of observations with PFAS measurements, LOQs, and 
% <LOQ for each PFAS. 
2.3. Assessment of outcome and covariates 
Height and weight were either measured by field nurse/physician 
and/or self-reported (in 3 studies: CROME, BEA and NEBII), 45 partic-
ipants were excluded from the analyses due to missing values (from 
NEBII, PCB cohort follow-up and BEA). A BMI z-score was calculated 
standardized for age and sex according to growth charts from the World 
Health Organization and categorized according to the International 
Obesity Task Force classification, using the ‘zanthro’ package in Stata 
(Vidmar et al., 2013). Self-reported co-variates included highest 
educational level of the household (ISCED scale: low/medium vs high 
education), consumption of several dietary components (fish, meat, 
milk, egg and fastfood; frequency of consumption: <1/week vs 1/week 
vs > 1/week), breastfeeding (in months, available in NEBII, Riksmaten 
Adolescents 2016–17, PCB cohort (follow-up), CROME, FLEHS IV and 
GerES V-sub) and birthweight (in grams, available in NEBII, PCB cohort 
(follow-up), FLEHS IV and GerES V-sub). Missing information on cova-
riates (<20% within each study) was imputed using multiple imputation 
chained equations (20 imputations). 
2.4. Statistical analyses 
For the exposure, the individual PFAS concentrations or molar sums 
were assessed as a continuous variable per interquartile range (IQR) 
increment in log transformed PFAS, as well as categorized into PFAS 
tertiles (all according to the study-specific distribution) to relax the 
linearity assumption. Molar sums were created for i) the most abundant 
PFAS (as also assessed together by EFSA due to similarities in animal 
effects, toxicokinetics and human blood concentrations (EFSA et al., 
2020)) [PFHxS, PFOS, PFOA and PFNA], ii) those with sulfonate func-
tional group [PFHxS and PFOS] and iii) those with carboxyl functional 
group [PFOA and PFNA]. 
Pairwise Spearman rank correlation coefficients were calculated to 
describe correlations between PFAS. Cross-sectional associations 
between PFAS concentrations and continuous BMI z-scores were 
assessed using linear regression models, adjusted for potential con-
founders. Pooled results from all studies, using linear mixed effects 
models, are presented as β-coefficients with corresponding 95% confi-
dence intervals (CI). Random-effects meta-analysis was used to visualize 
individual study results contributing to the overall result and potential 
heterogeneity between studies for IQR increment in log transformed 
PFAS. Additionally, pooled results from all studies, using logistic mixed 
effects models with categorized BMI (binary: normal versus overweight 
including obese or normal versus obese, cutoff points are equivalent to 
adult BMI of <25, 25–29.99, 30 kg/m2) (Cole et al., 2000; Vidmar 
et al., 2013) are presented as odds ratios (OR) with corresponding 95% 
CI. 
Potential confounders were selected based on directed acyclic graphs 
(DAGs), and availability of data in the studies (Supplemental Fig. 1) 
(Textor et al., 2017). Model 1 was unadjusted whilst Model 2 was 
adjusted for highest level of education in the household (2 categories) 
and fish consumption (3 categories). Other dietary components such as 
meat (missing in Riksmaten Adolescents, 2016–17), milk (missing in 
Riksmaten Adolescents, 2016–17), egg (missing in Riksmaten Adoles-
cents, 2016–17 and BEA) and fastfood (missing in Riksmaten Adoles-
cents, 2016–17 and FLEHS IV) consumption were tested, but these did 
not impact the estimates and were removed from the final models. Other 
potential confounders of degree of urbanization and sampling season 
also did not impact estimates and were not considered further. We 
explored potential confounding by breastfeeding or birthweight, by 
additionally including these variables in sensitivity analyses for the 
studies that had these data available (six or four, respectively). Studies 
with self-reported BMI were excluded in sensitivity analyses and this did 
not impact the estimates. Furthermore, potential modification by sex 
using stratified analysis for males and females was also explored. 
Quantile G-computation was used for mixture assessment using the 
‘qgcomp’ package in R (version 3.6.1), this is a relatively new method to 
estimate the effect of an exposure mixture without assuming directional 
homogeneity of the individual compounds (Keil et al., 2020). It trans-
forms exposures into quantized versions and fits a linear model which 
estimates the change in the outcome expected for a one-unit change in 
all exposures (corresponding to the sum of all regression coefficients of 
the exposures). The weights of each exposure are calculated by dividing 
the coefficient for each exposure by the sum of all exposure coefficients. 
We used data from 7 studies with [PFHxS, PFOS, PFOA and PFNA] 
available (PFHxS and PFNA were excluded in BEA and GerES V-sub, 
respectively, due to 70% LOQ), with 500 bootstraps. For the mixture 
assessment, the 7 studies were pooled and ‘study centre’ was used as a 
covariate in the model. As the multiple imputation method was not 
compatible with the quantile G-computation, we used a missing indi-
cator category for missing values. Both the use of simple pooling instead 
of mixed effects models and the use of missing indicator category instead 
of multiple imputation did not impact the estimates in the main analysis. 
All other statistical analyses were performed using the statistical soft-
ware STATA (version 15.1) (Stata Corp LP, College Station, TX, USA) 
and using the ‘metan’ package for the meta-analysis (Harris et al., 2008). 
P-values were calculated based on 2-sided tests and the cut-off for sta-
tistical significance was set at 0.05. 
3. Results 
3.1. Population description 
Population characteristics are described in Table 1 according to the 
distributions within each study. SLO CRP and CROME are the smallest 
studies with <100 observations. All studies were sampled between 2014 
and 2021, had approximately equal numbers of males and females and 
participants between 12 and 18 years old. BMI was the lowest for par-
ticipants of the NEBII study, which also had participants with the lowest 
ages (Table 1). The highest educational level of the household was 
T. Schillemans et al.                                                                                                                                                                                                                            
Environmental Pollution 316 (2023) 120566
4
relatively equally distributed in each study except for NEBII, which had 
mainly participants from highly educated households, and PCB cohort 
(follow-up), which had mainly participants from low/medium educated 
households (Table 1). Fish consumption differed amongst studies, with a 
low consumption in the PCB cohort (follow-up) from Slovakia, whereas 
a more modest consumption was seen in the Western studies and a 
Table 1 
Population characteristics (sampling years 2014–2021) for each of the nine HBM4EU aligned studies in teenagers (12–18 years).   
NEBII Riksmaten 
adolescents 
2016-17 
PCB cohort 
follow-up 
SLO CRP CROME BEA ESTEBAN FLEHS IV GerES V-sub Pooled 
cohorts 
Characteristics 
Country 
(Region) 
Norway 
(North) 
Sweden (North) Slovakia 
(East) 
SIovenia 
(South) 
Greece 
(South) 
Spain 
(South) 
France 
(West) 
Belgium 
(West) 
Germany 
(West) 
NA 
n obs 177 300 292 94 52 299 143 300 300 1957 
Sampling Year, 
range 
2016–2017 2016–2017 2019–2020 2018 2020–2021 2017–2018 2014–2016 2018 2014–2017 2014–2021 
Sex, 
% female 
57 50 57 45 44 52 57 50 50 52 
Age (yrs), mean 
(SD) 
12.3 (0.5) 14.8 (1.6) 15.7 (0.6) 13.8 (0.8) 14.4 (1.8) 14.8 (0.8)a 14.2 (1.6) 14.5 (0.6) 14.5 (1.7) 14.5 (1.5) 
BMI, mean (SD) 18.5 (2.5)b 21.3 (3.5) 22.3 (4.7)b 21.2 (4.8) 21.8 (3.8) 21.2 (3.2)b 20.2 (4.1) 21.0 (3.6) 20.7 (3.7) 21.0 (3.9) 
ISCED house., % 
Low/medium 6 42 80 57 33 43 53 39 43 46 
High 85 58 14 43 67 52 47 61 57 52 
Missing 8 0 6 0 0 5 0 0 0 2 
Fish cons., % 
<1/wk 3 2 98 32 12 4 24 0 24 23 
1/wk 12 51 2 61 44 12 64 71 72 42 
>1/wk 69 37 0 7 44 84 5 10 3 29 
Missing 16 10 0 0 0 0 8 19 0 7 
Breastfed 
(mths), mean 
(SD) 
11.6 (4.7) 6.4 (3.7) 7.1 (8.3)  6.9 (5.5)   3.4 (5.0) 7.8 (6.2) NA 
Missing, % 0.2 4 4 100 10 100 100 4 16  
Birthweight 
(gr), mean 
(SD) 
3631 (501)  3368 (520)     3377 
(568) 
3394 (556) NA 
Missing, % 0.01 100 1 100 100 100 100 4 3  
Abbreviations: International Standard Classifiction of Education (ISCED) of the household. 
a There were 4 participants with missing age values within the BEA study. 
b There were 31, 1 and 13 participants with missing BMI values within the NEBII, PCB cohort follow-up and BEA studies, respectively. 
Table 2 
PFAS concentrations (sampling years 2014–2021) for each of the nine HBM4EU aligned studies in teenagers (12–18 years).   
NEBII Riksmaten 
adolescents 
2016–17 
PCB cohort 
follow-up 
SLO CRP CROME BEA ESTEBAN FLEHS IV GerES V-sub 
PFAS in μg/La [median (25–75th percentiles)] 
PFPeA NA NA 47% <LOQ 0.10 
(0.07–0.12) 
70% <LOQ 70% <LOQ 70% <LOQ 70% <LOQ 70% <LOQ 
PFHxA NA 70% <LOQ 0.07 (0.06–0.10) 70% <LOQ 0.14 
(0.11–0.16) 
70% <LOQ 70% <LOQ 70% <LOQ 70% <LOQ 
PFHpA 0.07 
(0.05–0.10) 
70% <LOQ 0.03 (0.02–0.05) 48% <LOQ 70% <LOQ 70% <LOQ 70% <LOQ 70% <LOQ 70% <LOQ 
PFOA 1.28 
(1.05–1.57) 
1.15 (0.89–1.51) 0.71 (0.48–0.96) 0.86 
(0.74–1.06) 
0.88 
(0.74–1.25) 
0.66 
(0.52–0.80) 
1.47 
(1.22–1.80) 
1.10 
(0.88–1.40) 
1.25 
(0.81–1.80) 
PFNA 0.44 
(0.32–0.62) 
0.38 (0.27–0.51) 0.17 (0.09–0.26) 0.25 
(0.19–0.31) 
0.41 
(0.30–0.56) 
0.28 
(0.21–0.38) 
0.54 
(0.43–0.71) 
0.32 
(0.23–0.44) 
70% <LOQ 
PFDA 0.13 
(0.10–0.18) 
41% <LOQ 32% <LOQ 0.14 
(0.10–0.19) 
0.17 
(0.14–0.26) 
70% <LOQ 39% <LOQ 52% <LOQ 70% <LOQ 
PFUnDA 0.09 
(0.06–0.15) 
53% <LOQ 58% <LOQ 0.06 
(0.04–0.08) 
0.04 
(0.01–0.08) 
70% <LOQ 0.11 
(0.07–0.15) 
70% <LOQ 70% <LOQ 
PFHxS 0.48 
(0.37–0.62) 
0.39 (0.26–0.58) 0.29 (0.21–0.44) 0.23 
(0.19–0.30) 
0.28 
(0.21–0.40) 
70% <LOQ 0.68 
(0.53–1.03) 
0.49 
(0.37–0.66) 
0.39 
(0.24–0.53) 
PFHpS 59% <LOQ NA 0.03 (0.02–0.05) 0.03 
(0.02–0.05) 
0.05 
(0.04–0.07) 
70% <LOQ 70% <LOQ 70% <LOQ NA 
PFOS 2.79 
(2.18–3.68) 
2.68 (1.97–4.08) 1.37 (0.84–2.47) 1.65 
(1.16–2.71) 
2.11 
(1.57–3.31) 
1.34 
(0.93–1.84) 
2.01 
(1.51–3.18) 
2.20 
(1.55–3.40) 
2.60 
(1.95–3.48) 
Note: Some studies did not assess specific PFAS compounds (indicated as NA) or had >30% <LOQ in which case it was not included in the analyses. 
Abbreviations: perfluoropentanoic acid (PFPeA), perfluorohexanoic acid (PFHxA), perfluoroheptanoic acid (PFHpA), perfluorooctanoate (PFOA), perfluorononanoic 
acid (PFNA), perfluorodecanoic acid (PFDA), perfluoroundecanoic acid (PFUnDA), perfluorododecanoic acid (PFDoDA), perfluorobutane sulfonic acid (PFBS), per-
fluorohexane sulfonic acid (PFHxS), perfluoroheptanane sulfonic acid (PFHpS) and perfluorooctane sulfonic acid (PFOS). 
a In Riksmaten adolescents 2016–17 the laboratory presented PFAS in ng/g (μg/kg), which was reported to HBM4EU in μg/L assuming that 1 ml blood serum equals 
1 g blood serum. 
T. Schillemans et al.                                                                                                                                                                                                                            
Environmental Pollution 316 (2023) 120566
5
higher consumption in the Northern and Southern studies (particularly 
the studies from Norway and Spain) (Table 1). 
LOQs differed between studies and not all studies could be used for 
all PFAS analyses: on average five PFAS could be included per study. 
PFOA, PFNA, PFHxS and PFOS were 70% LOQ in almost all or all 
studies, but for the other PFAS less studies could be included (PFPeA, 
PFHxA and PFHpA were 70% LOQ in only one or two studies and 
PFDA, PFUnDA and PFHpS in only three or four studies) (Table 2 and 
Supplemental Table 1). We observed generally higher PFAS concentra-
tions in the North and West regions of Europe. The highest median 
concentrations were 1.47, 0.54, 2.79 and 0.68 μg/L for the most prev-
alent PFAS (PFOA, PFNA, PFOS and PFHxS, respectively) (Table 2). The 
strongest correlations were seen for PFOA, PFNA, PFHxS, PFHpS and 
PFOS (0.4 < r < 0.9) (Supplemental Table 2). 
3.2. Associations of individual PFAS compounds and molar sums with 
BMI 
Significant negative associations were found in the pooled analyses: 
the molar sums of [PFHxS, PFOS, PFOA and PFNA], [PFHxS and PFOS] 
and [PFOA and PFNA], as well as individual compounds of PFPeA, 
PFHpA, PFOA, PFNA and PFOS associated with lower BMI z-scores 
(Table 3). Only PFHxS and PFHpS showed non-significant opposite re-
sults with a tendency for positive associations (Table 3, Fig. 1, Supple-
mental Fig. 2). The pooled results from the meta-analyses were similar, 
indicating a significant 0.08 (  0.15,   0.01) lower BMI z-score per 
interquartile range increase in PFOA (Figs. 1 and 2, Supplemental 
Fig. 2). Except for CROME and PCB cohort (follow-up), there were either 
null- or negative associations in the individual studies. Additional 
adjustment for breastfeeding or birthweight did not materially impact 
the estimates for PFAS and BMI associations within individual studies 
(Supplemental Table 3). We did find stronger negative associations with 
BMI for males for most of the PFAS in the pooled analyses (with esti-
mates of   0.12 (  0.22,   0.03) for males and 0.02 (  0.06, 0.11) for 
females for the molar sum of [PFHxS, PFOS, PFOA and PFNA]; Sup-
plemental Table 4). Although the same inverse tendency was observed 
in the pooled logistic regression models between the PFAS and over-
weight/obese (~20% were overweight or obese), the association was 
only significant for PFpeA and PFHpA (Table 4). Similar associations 
were seen for PFAS and obesity (~5% were obese) (data not shown). 
Table 3 
Associations of tertiles and interquartile range-increment of log transformed PFAS with BMI z-score (age and sex standardized) in the pooled HBM4EU aligned studies 
(sampling years 2014–2021).  
PFAS (n 
studies) 
n observations Median (25-75th percentiles, nmol/L or μg/L)  Model 1 – β (95% CI) Model 2 – β (95% CI) 
T1 T2 T3 T1 T2 T3 Total T2 T3 IQR 
(25th 
  75th) 
T2 T3 IQR 
(25th – 
75th) 
ƩPFAS 
(7) 
446 444 436 6.14 
(4.27–7.50) 
9.58 
(7.17–10.93) 
14.90 
(12.51–19.24) 
9.24 
(6.49–12.9) 
¡0.19 
(-0.33, 
-0.04) 
¡0.19 
(-0.33, 
-0.04) 
¡0.07 
(-0.13, 
-0.01) 
¡0.17 
(-0.31, 
-0.02) 
¡0.16 
(-0.31, 
-0.02) 
  0.03 
(  0.10, 
0.03) 
ƩsPFAS 
(8) 
545 545 536 3.57 
(2.53–4.45) 
5.92 
(4.83–6.70) 
9.88 
(8.27–13.53) 
5.78 
(4.00–8.28) 
¡0.19 
(-0.32, 
-0.06) 
¡0.19 
(-0.32, 
-0.06) 
¡0.08 
(-0.13, 
-0.03) 
¡0.16 
(-0.29, 
-0.03) 
¡0.16 
(-0.29, 
-0.03) 
  0.05 
(  0.11, 
0.01) 
ƩcPFAS 
(8) 
539 540 533 2.11 
(1.50–2.70) 
3.17 
(2.33–3.72) 
4.47 
(3.29–5.67) 
3.04 
(2.24–4.14) 
  0.11 
(  0.24, 
0.02) 
¡0.18 
(-0.31, 
-0.05) 
¡0.09 
(-0.14, 
-0.04) 
  0.09 
(  0.22, 
0.04) 
¡0.15 
(-0.28, 
-0.02) 
¡0.07 
(-0.13, 
-0.01) 
PFPeA 
(1) 
35 30 29 0.05 
(0.01–0.07) 
0.10 
(0.09–0.11) 
0.14 
(0.13–0.17) 
0.10 
(0.07–0.12) 
  0.15 
(  0.74, 
0.44) 
¡0.64 
(-1.24, 
-0.05) 
  0.05 
(  0.20, 
0.11) 
  0.15 
(  0.73, 
0.42) 
¡0.67 
(-1.25, 
-0.10) 
  0.05 
(  0.20, 
0.11) 
PFHxA 
(2) 
129 108 106 0.05 
(0.04–0.06) 
0.08 
(0.07–0.09) 
0.12 
(0.10–0.15) 
0.09 
(0.06–0.11) 
  0.20 
(  0.50, 
0.10) 
  0.08 
(  0.39, 
0.22) 
  0.08 
(  0.21, 
0.06) 
  0.19 
(  0.49, 
0.11) 
  0.08 
(  0.38, 
0.22) 
  0.07 
(  0.20, 
0.07) 
PFHpA 
(2) 
147 168 122 0.02 
(0.02–0.04) 
0.04 
(0.03–0.06) 
0.10 
(0.07–0.12) 
0.05 
(0.03–0.08) 
¡0.36 
(-0.61, 
-0.11) 
¡0.54 
(-0.81, 
-0.27) 
¡0.19 
(-0.30, 
-0.07) 
¡0.36 
(-0.60, 
-0.11) 
¡0.51 
(-0.78, 
-0.24) 
¡0.19 
(-0.30, 
-0.07) 
PFOA (9) 646 639 627 0.65 
(0.46–0.86) 
1.08 
(0.77–1.24) 
1.55 
(1.13–1.96) 
0.99 
(0.71–1.38) 
¡0.15 
(-0.26, 
-0.03) 
¡0.20 
(-0.32, 
-0.08) 
¡0.09 
(-0.15, 
-0.04) 
¡0.13 
(-0.24, 
-0.01) 
¡0.18 
(-0.29, 
-0.06) 
¡0.08 
(-0.13, 
-0.02) 
PFNA (8) 549 542 521 0.19 
(0.12–0.26) 
0.32 
(0.26–0.39) 
0.53 
(0.41–0.71) 
0.32 
(0.22–0.47) 
¡0.15 
(-0.28, 
-0.02) 
¡0.18 
(-0.31, 
-0.04) 
¡0.09 
(-0.15, 
-0.03) 
¡0.13 
(-0.26, 
-0.00) 
¡0.14 
(-0.27, 
-0.00) 
¡0.07 
(-0.13, 
-0.01) 
PFDA (3) 111 88 93 0.09 
(0.08–0.11) 
0.15 
(0.13–0.16) 
0.23 
(0.19–0.30) 
0.14 
(0.10–0.20) 
  0.12 
(  0.42, 
0.17) 
  0.16 
(  0.45, 
0.14) 
  0.03 
(  0.19, 
0.14) 
  0.07 
(  0.37, 
0.23) 
  0.13 
(  0.42, 
0.17) 
  0.02 
(  0.19, 
0.14) 
PFUnDA 
(4) 
155 144 136 0.04 
(0.03–0.06) 
0.09 
(0.07–0.11) 
0.15 
(0.13–0.21) 
0.08 
(0.05–0.13) 
  0.25 
(  0.50, 
0.01) 
  0.25 
(  0.51, 
0.01) 
  0.12 
(  0.25, 
0.00) 
  0.21 
(  0.47, 
0.05) 
  0.19 
(  0.46, 
0.08) 
  0.09 
(  0.22, 
0.04) 
PFHxS 
(8) 
558 540 528 0.23 
(0.17–0.31) 
0.41 
(0.32–0.48) 
0.71 
(0.56–1.01) 
0.4 
(0.26–0.59) 
  0.01 
(  0.14, 
0.12) 
0.01 
(  0.12, 
0.14) 
0.00 
(  0.05, 
0.06) 
0.00 
(  0.13, 
0.13) 
0.02 
(  0.11, 
0.15) 
0.01 
(  0.04, 
0.07) 
PFHpS 
(3) 
203 100 134 0.02 
(0.00–0.02) 
0.04 
(0.03–0.04) 
0.08 
(0.06–0.12) 
0.03 
(0.02–0.06) 
0.11 
(  0.17, 
0.40) 
0.17 
(  0.10, 
0.43) 
0.03 
(  0.06, 
0.11) 
0.10 
(  0.18, 
0.38) 
0.17 
(  0.09, 
0.43) 
0.04 
(  0.04, 
0.13) 
PFOS (9) 652 629 631 1.25 
(0.85–1.70) 
2.22 
(1.55–2.63) 
3.84 
(3.03–5.39) 
2.1 
(1.39–3.13) 
¡0.18 
(-0.30, 
-0.06) 
¡0.19 
(-0.31, 
-0.07) 
¡0.07 
(-0.12, 
-0.01) 
¡0.15 
(-0.27, 
-0.03) 
¡0.15 
(-0.27, 
-0.03) 
  0.04 
(  0.10, 
0.02) 
Note: Beta1 coefficients from linear regression mixed effects models (fitted with a random intercept for study) per study specific IQR increase (difference between 25th 
to 75th percentile) in log transformed PFAS or per tertile (T1 reference). Model 1 was unadjusted and model 2 was adjusted for highest educational level in the 
household and fish consumption. Significant results are indicated in bold. Abbreviations: ƩPFAS, molar weight sum including PFHxS, PFOS, PFOA and PFNA; ƩsPFAS, 
molar weight sum including PFHxS and PFOS; ƩcPFAS, molar weight sum including PFOA and PFNA; IQR, interquartile range; T, tertile. 
T. Schillemans et al.                                                                                                                                                                                                                            
Environmental Pollution 316 (2023) 120566
6
3.3. Associations of PFAS mixture with BMI 
Mixture assessment using g-computation indicated the same ten-
dency for negative associations, although not significant, of the mixture 
of the most abundant PFAS [PFHxS, PFOS, PFOA and PFNA] with BMI z- 
score with estimates of   0.05 (  0.13, 0.03) per one quartile increase in 
mixture (Fig. 3a and c). In line with the individual exposure linear 
regression models, PFHxS showed opposite associations with BMI z- 
scores as compared to PFOS, PFOA and PFNA (Figs. 1c and 3b). Thus, to 
separate these associations, we included in the same linear regression 
model a molar sum of [PFOS, PFOA and PFNA] adjusting for PFHxS. In 
agreement with the previous results, we observed significant negative 
and positive associations, respectively (Fig. 3c). 
4. Discussion 
In the present cross-sectional study, we found overall negative as-
sociations between relatively low PFAS (PFPeA, PFHpA, PFOA, PFNA 
and PFOS) blood concentrations and BMI z-scores in teenagers in nine 
studies across Europe (the HBM4EU aligned studies). The same inverse 
tendency was observed for PFAS concentrations with risk of overweight/ 
obesity. Mixture assessment indicated similar negative associations for 
PFOS, PFOA and PFNA, but a positive association for PFHxS. 
4.1. Epidemiological evidence 
Although the PFAS levels observed in the present study are consid-
erably lower, our findings are in line with another recent cross-sectional 
study among highly exposed Italian adolescents and children (Canova 
et al., 2021), and another among young US girls (Fassler et al., 2019); 
both showing negative associations between single PFAS and BMI. 
Negative associations between PFOS and body fat have also been 
described in Danish boys (Thomsen et al., 2021). Our findings are 
similar to those of others using mixture approaches: a longitudinal study 
indicated a negative association of PFAS mixture exposure during 
childhood with less accrual of lean mass in early adolescence, mainly 
driven by PFOA (Janis et al., 2021) and an exposome-wide study in 
children placed PFOA among the chemicals associated with lower BMI 
(Vrijheid et al., 2020). 
However, there are also cross-sectional studies during adolescence 
that found a positive association between PFOA and overweight/obesity 
in US teenagers (Geiger et al., 2021) and between PFHpS and PFHxS and 
obesity in Norwegian teenagers (Averina et al., 2021), which was not the 
case for the Norwegian teenagers in this study. Although, PFHpS and 
PFHxS both had positive estimates in the present study as well and for 
PFHxS this association became significant when adjusting for the sum of 
PFOA, PFNA and PFOS. Meanwhile, others mostly reported null asso-
ciations between PFAS and BMI or body fat measures (Averina et al., 
2021; Thomsen et al., 2021). Similarly, we found null associations for 
PFHxA, PFDA, PFUnDA, PFHxS and PFHpS potentially due to high 
Fig. 1. Associations between PFAS and BMI z-score (age and sex standardized) using linear regression meta-analysis models combining the HBM4EU aligned studies 
in teenagers. Associations are shown per study specific IQR increase (difference between 25th to 75th percentile) in log transformed PFAS and are adjusted according 
to model 2 (highest level of household education and fish consumption). a) PFOA, b) PFNA, c) PFHxS and d) PFOS. 
T. Schillemans et al.                                                                                                                                                                                                                            
Environmental Pollution 316 (2023) 120566
7
correlations between PFAS with opposite effects (as indicated by our 
mixture analysis results: PFHxS versus PFOA, PFNA and PFOS). There are 
only a few prospective studies, one European multicenter study indi-
cated that childhood PFOS exposure was associated with higher BMI 
during adolescence, while adolescent PFOS exposure did not associate 
with higher BMI during young adulthood (Domazet et al., 2016). 
Prenatal exposures are also considered important for adiposity in 
children and teenagers and the evidence for PFAS exposures is similarly 
inconsistent, as recently reviewed by Lee et al. (2021). In general, 
studies investigating prenatal PFAS exposure and associations with 
childhood adiposity indicate more of a negative association with BMI 
during the first years of life (Starling et al., 2019; Starling et al., 2017) 
and a positive association with BMI during childhood (Braun et al., 
2016; Chen et al., 2019; Lauritzen et al., 2018; Mora et al., 2017) and 
adolescence (Liu et al., 2020). However, negative associations between 
prenatal PFAS and childhood and adolescent adiposity (Braun et al., 
2021; Hartman et al., 2017) and overall null associations were also 
found (Andersen et al., 2013; Bloom et al., 2021; Manzano-Salgado 
et al., 2017; Martinsson et al., 2020). A recent systematic review and 
meta-analysis indicated null-associations between prenatal PFOS and 
PFOA with childhood BMI (Stratakis et al., 2022). 
Potential modifications by sex, race-ethnicity, maternal BMI (Bloom 
et al., 2021) and education (Hartman et al., 2017) as well as other 
obesogenic and PFAS chemicals (Starling et al., 2019; Starling et al., 
2017) may explain these divergent results. Furthermore, age and pu-
berty status may impact PFAS’ effects on BMI, potentially as a 
confounder (menarche both affects PFAS levels and BMI) or as mediator 
(PFAS may impact hormone levels and thereby affect puberty status and 
BMI). 
4.2. Molecular evidence 
We found that higher PFAS levels associated with lower BMI. How-
ever, PFAS has been consistently shown to associate with higher 
cholesterol in adults (EFSA et al., 2020) and in adolescents (Averina 
et al., 2021; Canova et al., 2021). Experimental studies indicate that 
PFAS-induced PPARα or PPARγ activation may induce toxicological 
changes in metabolism (Takacs and Abbott, 2007), which could decrease 
BMI (Xie et al., 2002). PPARs are linked to modulating glucose and lipid 
metabolism, adipogenesis, adipocyte differentiation and inflammation. 
Other nuclear receptors relevant for lipid homeostasis may also be 
involved, such as constitutive androstane receptor (CAR) (Rosen et al., 
2017), estrogen receptor (ER) (Rosen et al., 2017), pregnane X receptor 
(PXR) (Bijland et al., 2011), liver X receptor (LXR) (Bijland et al., 2011), 
farsenoid X receptor (FXR) (Bijland et al., 2011) and hepatocyte nuclear 
factor (HNF4α) (Beggs et al., 2016). Other pathways potentially relevant 
for metabolic disruptions include PFAS-related disruptions in the 
enterohepatic circulation and hepatic injury (Roth et al., 2021) and 
PFAS-related changes in hormone levels and sexual maturation via al-
terations of thyroid and/or steroid hormone synthesis and metabolism 
(Lee and Choi, 2017; Zhao et al., 2011). We found stronger negative 
associations between PFAS and BMI z-score in males, potentially related 
to endocrine disruption (Du et al., 2013; Zhou et al., 2017). PFAS has 
been shown to inhibit 11-β hydroxysteroid dehydrogenase 2 which 
Fig. 2. Associations between PFAS and BMI z-score (age and sex standardized) using linear regression meta-analysis models combining the HBM4EU aligned studies. 
Associations are shown per study specific IQR increase (difference between 25th to 75th percentile) in log transformed PFAS and are adjusted according to model 2 
(highest level of household education and fish consumption). a) ƩPFAS [molar weight sum including PFHxS, PFOS, PFOA and PFNA] b) ƩsPFAS [molar weight sum 
including PFHxS and PFOS] c) ƩcPFAS [molar weight sum including PFOA and PFNA]. 
T. Schillemans et al.                                                                                                                                                                                                                            
Environmental Pollution 316 (2023) 120566
8
Table 4 
Associations of tertiles and interquartile range-increment of log transformed PFAS with overweight (age and sex standardized) in the pooled HBM4EU aligned studies 
(sampling years 2014–2021).  
PFAS (n 
studies) 
n observations (overweight/ 
normal) 
Model 1 – OR (95% CI) Model 2 – OR (95% CI) 
T1 T2 T3 T1 
(ref) 
T2 T3 IQR (25th 
  75th) 
T1 
(ref) 
T2 T3 IQR (25th 
  75th) 
ƩPFAS (7) 109/ 
308 
86/317 91/312  0.76 (0.55, 
1.06) 
0.82 (0.59, 
1.13) 
0.84 (0.76, 
0.94)  
0.79 (0.57, 
1.10) 
0.86 (0.62, 
1.19) 
0.95 (0.83, 
1.08) 
ƩsPFAS (8) 128/ 
375 
105/ 
397 
107/ 
387  
0.78 (0.58, 
1.04) 
0.81 (0.60, 
1.08) 
0.82 (0.74, 
0.92)  
0.81 (0.60, 
1.09) 
0.86 (0.63, 
1.15) 
0.91 (0.80, 
1.03) 
ƩcPFAS (8) 129/ 
377 
116/ 
375 
105/ 
386  
0.90 (0.68, 
1.21) 
0.79 (0.59, 
1.06) 
0.85 (0.77, 
0.94)  
0.96 (0.72, 
1.29) 
0.84 (0.62, 
1.13) 
0.92 (0.83, 
1.04) 
PFPeA (1) 14/19 8/21 2/27  0.52 (0.18, 
1.50) 
0.10 (0.02, 
0.49) 
0.86 (0.65, 
1.12)  
0.51 (0.16, 
1.58) 
0.08 (0.01, 
0.43) 
0.85 (0.64, 
1.13) 
PFHxA (2) 36/84 24/75 32/62  0.75 (0.41, 
1.36) 
1.20 (0.68, 
2.15) 
1.01 (0.78, 
1.31)  
0.75 (0.41, 
1.38) 
1.20 (0.67, 
2.15) 
1.02 (0.78, 
1.32) 
PFHpA (2) 48/92 32/118 15/92  0.49 (0.29, 
0.84) 
0.34 (0.17, 
0.65) 
0.73 (0.58, 
0.93)  
0.50 (0.29, 
0.86) 
0.35 (0.18, 
0.67) 
0.74 (0.58, 
0.94) 
PFOA (9) 157/ 
449 
128/ 
451 
119/ 
460  
0.81 (0.62, 
1.06) 
0.74 (0.56, 
0.97) 
0.87 (0.77, 
0.98)  
0.85 (0.65, 
1.11) 
0.77 (0.58, 
1.01) 
0.91 (0.81, 
1.02) 
PFNA (8) 131/ 
384 
115/ 
378 
104/ 
376  
0.89 (0.67, 
1.19) 
0.81 (0.61, 
1.09) 
0.87 (0.77, 
0.99)  
0.94 (0.70, 
1.26) 
0.89 (0.66, 
1.20) 
0.90 (0.80, 
1.02) 
PFDA (3) 22/84 14/67 17/69  0.83 (0.39, 
1.76) 
0.95 (0.46, 
1.94) 
1.19 (0.78, 
1.81)  
0.93 (0.43, 
2.01) 
1.01 (0.48, 
2.10) 
1.24 (0.82, 
1.87) 
PFUnDA (4) 32/115 24/101 22/100  0.87 (0.48, 
1.57) 
0.79 (0.43, 
1.46) 
0.83 (0.64, 
1.07)  
0.93 (0.50, 
1.75) 
0.88 (0.46, 
1.71) 
0.85 (0.62, 
1.15) 
PFHxS (8) 110/ 
406 
122/ 
370 
108/ 
383  
1.22 (0.91, 
1.64) 
1.04 (0.77, 
1.41) 
1.02 (0.91, 
1.16)  
1.24 (0.92, 
1.67) 
1.07 (0.79, 
1.45) 
1.03 (0.92, 
1.15) 
PFHpS (3) 51/135 22/70 43/83  0.83 (0.47, 
1.48) 
1.37 (0.84, 
2.24) 
1.08 (0.92, 
1.27)  
0.81 (0.45, 
1.45) 
1.39 (0.85, 
2.28) 
1.10 (0.93, 
1.29) 
PFOS (9) 155/ 
446 
129/ 
450 
120/ 
464  
0.83 (0.63, 
1.08) 
0.74 (0.57, 
0.98) 
0.83 (0.72, 
0.96)  
0.87 (0.66, 
1.14) 
0.80 (0.61, 
1.06) 
0.91 (0.81, 
1.03) 
Note: Odds ratios (95% confidence intervals) from logistic regression mixed effects models (fitted with a random intercept for study) per study specific IQR increase 
(difference between 25th to 75th percentile) in log transformed PFAS or per tertile (T1 reference). Model 1 was unadjusted and model 2 was adjusted for highest level 
of education in the household and fish consumption. Abbreviations: ƩPFAS, molar weight sum including PFHxS, PFOS, PFOA and PFNA; ƩsPFAS, molar weight sum 
including PFHxS and PFOS; ƩcPFAS, molar weight sum including PFOA and PFNA; IQR, interquartile range; T, tertile. 
Fig. 3. Associations between PFAS and BMI z-score 
(age and sex standardized) using quantile G-compu-
tation using data from seven HBM4EU aligned studies 
(n  1326). Associations are shown per 1-quartile 
increase in mixture PFAS (PFOA, PFNA, PFHxS, 
PFOS) and are adjusted according to model 2 (highest 
level of household education, fish consumption). 
Panel a) shows the association between PFAS mixture 
and BMI z-score, b) shows the weights of each indi-
vidual PFAS contributing to the mixture and c) shows 
the estimates for the G-computation and for the linear 
regression mixed effects model including ƩgPFAS 
[molar weight sum of PFOS, PFOA and PFNA] and 
PFHxS (both mutually adjusted and separated). Ab-
breviations: ƩgPFAS [molar weight sum including 
PFOS, PFOA and PFNA].   
T. Schillemans et al.                                                                                                                                                                                                                            
Environmental Pollution 316 (2023) 120566
9
could lead to the suppression of testosterone production in Leydig cells 
(Zhao et al., 2011). Negative associations between PFAS and testos-
terone in adolescent males have been found before (Zhou et al., 2016) 
and low testosterone could decrease lean mass (Kelly and Jones, 2015). 
4.3. Strengths and limitations 
The present study has several strengths and limitations. Pooling the 
studies gave a large sample size and a wider range of PFAS concentra-
tions. The harmonization of the sampling protocol and the participation 
in the QA/QC programme for PFAS measurements reduced the potential 
for bias from exposure misclassification. However, studies had different 
procedures (different questionnaires, measured versus self-reported BMI, 
serum versus plasma, linear versus linear and branched PFAS and 
different LOQs for PFAS), which complicated comparisons of PFAS 
levels between data collections. The measurement of multiple PFAS 
allowed us to consider PFAS mixtures. However, for PFAS sums we 
assumed equipotency of the included compounds and this might not 
reflect the reality, thus we also included another mixture approach (g- 
computation) that allowed the weights of each compound to differ. 
Additionally, BMI is vulnerable to outcome misclassification as it cannot 
differentiate between lean and fat body mass and was self-reported for 
some participants. There was some heterogeneity between studies, 
which could potentially be explained by differences in age (puberty 
could impact associations), bias resulting from differences in outcome 
measurements (e.g. self-reported in CROME), residual or unmeasured 
confounding (e.g. fish consumption was lower in PCB cohort-follow up) 
or differences in PFAS levels (e.g. PFHxS is lower in SLO CRP). However, 
exclusions of studies with self-reported BMI and adjustments for main 
potential confounders (according to our DAG; socio-economic status, 
diet, breastfeeding and birthweight) only impact results marginally. 
Although PFAS have long half-lives and thus we assess cumulative 
exposure over the last years, we are still unable to infer causality due to 
the cross-sectional nature of this study. Reverse causality is theoretically 
possible as PFAS is excreted in the bile and urine (EFSA et al., 2018) and 
higher BMI has been associated with enterohepatic circulation (Haeusler 
et al., 2016) and with reduced eGFR (Gunta and Mak, 2013). Higher BMI 
is associated with increased bile acid synthesis so potentially higher BMI 
causes increased PFAS excretion. However, PFAS affecting bile acid 
synthesis is also plausible via decreasing CYP7A1 (Behr et al., 2020). 
Higher BMI associations with reduced eGFR would not explain the 
negative associations found in this study, as higher BMI would then give 
higher PFAS levels. As PFAS is not lipid-soluble, reverse causality due to 
PFAS storage in fat is not likely (Forsthuber et al., 2020). Furthermore, 
we were not able to investigate the potential effects of prenatal PFAS 
exposures and these have been proposed to be more relevant than 
postnatal PFAS exposures for adiposity in childhood (Papadopoulou 
et al., 2021). Additionally, we cannot yet study the potential long-term 
consequences of the association between PFAS and lower BMI in teen-
agers in this population. 
5. Conclusions 
We found evidence for negative cross-sectional associations with 
age- and sex-standardized BMI for 5 of the 10 PFAS compounds (PFPeA, 
PFHpA, PFOA, PFNA and PFOS) in teenagers from nine studies across 
Europe. This indicates that these exposures may impact pathways rele-
vant to metabolism in teenagers. Our results furthermore suggest 
opposite associations for PFOS, PFOA and PFNA versus PFHxS on BMI z- 
scores in teenagers. Although effect estimates were only minor, this 
could still be relevant for population health considering that such a large 
part of the population is exposed to these compounds. Further research 
is needed both on epidemiological and molecular level to understand 
these associations and their implications in teenagers as well as subse-
quently during adulthood. 
Funding 
This publication has been developed under the HMB4EU initiative: 
www.HBM4EU.eu. This project has received funding from the European 
Union’s Horizon 2020 research and innovation program under grant 
agreement No. 733032 HBM4EU. T Schillemans and A Åkesson were 
supported by the Swedish Research Council no 2017-00822. 
Credit author statement 
Conceptualization: Sylvie Remy, Maria Uhl, Eva Govarts. Meth-
odology: Tessa Schillemans, Nina Iszatt, Sylvie Remy, Greet Schoeters, 
Mariana F. Fernandez, Anteneh Desalegn, Eva Govarts, Agneta Åkesson. 
Formal analysis: Tessa Schillemans. Resources: Shereen Cynthia 
D’Cruz, Anteneh Desalegn, Line S. Haug, Sanna Lignell, Anna Karin 
Lindroos, Lucia Fabelova, Lubica Palkovicova Murinova, Tina Kosjek, 
Ziga Tkalec, Catherine Gabriel, Denis Sarigiannis, Susana Pedraza-Díaz, 
Marta Esteban-Lopez, Argelia Casta~no, Loïc Rambaud, Margaux Riou, 
Sara Pauwels, Nik Vanlarebeke, Marike Kolossa-Gehring, Nina Vogel. 
Writing – Original Draft: Tessa Schillemans. Writing - Review & 
Editing: all co-authors. Visualization: Tessa Schillemans. Supervision: 
Agneta Åkesson. Project administration: Sylvie Remy, Maria Uhl, Eva 
Govarts. 
Declaration of competing interest 
The authors declare that they have no known competing financial 
interests or personal relationships that could have appeared to influence 
the work reported in this paper. 
Data availability 
The data that has been used is confidential. 
Acknowledgements 
The Norwegian Institute of Public Health (NIPH) has contributed to 
funding of the Norwegian Environmental Biobank (NEB II). The labo-
ratory measurements have partly been funded by the Research Council 
of Norway through research projects (275903 and 268465). The Nor-
wegian Mother, Father and Child Cohort Study is supported by the 
Norwegian Ministry of Health and Care Services and the Ministry of 
Education and Research. We are grateful to all the participating families 
in Norway who take part in this on-going cohort study. 
Riksmaten adolescents 2016–17 was performed by the Swedish Food 
Agency with financial support from the Swedish Civil Contingencies 
Agency and the Swedish Environmental Protection Agency. 
PCB cohort (follow-up) received additional funding from the Minis-
try of Health of the Slovak Republic, program 07B0103. 
The funding of GerES by the German Federal Ministry for the Envi-
ronment, Nature Conservation, Nuclear Safety and Consumer Protection 
is gratefully acknowledged. 
Appendix A. Supplementary data 
Supplementary data to this article can be found online at https://doi. 
org/10.1016/j.envpol.2022.120566. 
References 
Andersen, C.S., Fei, C., Gamborg, M., Nohr, E.A., Sørensen, T.I., Olsen, J., 2013. Prenatal 
exposures to perfluorinated chemicals and anthropometry at 7 years of age. Am. J. 
Epidemiol. 178, 921–927. 
Averina, M., Brox, J., Huber, S., Furberg, A.S., 2021. Exposure to perfluoroalkyl 
substances (PFAS) and dyslipidemia, hypertension and obesity in adolescents. The 
Fit Futures study. Environ. Res. 195, 110740. 
T. Schillemans et al.                                                                                                                                                                                                                            
Environmental Pollution 316 (2023) 120566
10
Beggs, K.M., McGreal, S.R., McCarthy, A., Gunewardena, S., Lampe, J.N., Lau, C., 
Apte, U., 2016. The role of hepatocyte nuclear factor 4-alpha in perfluorooctanoic 
acid- and perfluorooctanesulfonic acid-induced hepatocellular dysfunction. Toxicol. 
Appl. Pharmacol. 304, 18–29. 
Behr, A.C., Kwiatkowski, A., Ståhlman, M., Schmidt, F.F., Luckert, C., Braeuning, A., 
Buhrke, T., 2020. Impairment of bile acid metabolism by perfluorooctanoic acid 
(PFOA) and perfluorooctanesulfonic acid (PFOS) in human HepaRG hepatoma cells. 
Arch. Toxicol. 94, 1673–1686. 
Bijland, S., Rensen, P.C., Pieterman, E.J., Maas, A.C., van der Hoorn, J.W., van Erk, M.J., 
Havekes, L.M., Willems van Dijk, K., Chang, S.C., Ehresman, D.J., Butenhoff, J.L., 
Princen, H.M., 2011. Perfluoroalkyl sulfonates cause alkyl chain length-dependent 
hepatic steatosis and hypolipidemia mainly by impairing lipoprotein production in 
APOE*3-Leiden CETP mice. Toxicol. Sci. 123, 290–303. 
Bloom, M.S., Commodore, S., Ferguson, P.L., Neelon, B., Pearce, J.L., Baumer, A., 
Newman, R.B., Grobman, W., Tita, A., Roberts, J., Skupski, D., Palomares, K., 
Nageotte, M., Kannan, K., Zhang, C., Wapner, R., Vena, J.E., Hunt, K.J., 2021. 
Association between gestational PFAS exposure and Children’s adiposity in a diverse 
population. Environ. Res. 203, 111820. 
Braun, J.M., Chen, A., Romano, M.E., Calafat, A.M., Webster, G.M., Yolton, K., 
Lanphear, B.P., 2016. Prenatal perfluoroalkyl substance exposure and child adiposity 
at 8 years of age: the HOME study. Obesity 24, 231–237. 
Braun, J.M., Eliot, M., Papandonatos, G.D., Buckley, J.P., Cecil, K.M., Kalkwarf, H.J., 
Chen, A., Eaton, C.B., Kelsey, K., Lanphear, B.P., Yolton, K., 2021. Gestational 
perfluoroalkyl substance exposure and body mass index trajectories over the first 12 
years of life. Int. J. Obes. 45, 25–35. 
Canova, C., Di Nisio, A., Barbieri, G., Russo, F., Fletcher, T., Batzella, E., Dalla Zuanna, T., 
Pitter, G., 2021. PFAS concentrations and cardiometabolic traits in highly exposed 
children and adolescents. Int. J. Environ. Res. Publ. Health 18. 
Chen, Q., Zhang, X., Zhao, Y., Lu, W., Wu, J., Zhao, S., Zhang, J., Huang, L., 2019. 
Prenatal exposure to perfluorobutanesulfonic acid and childhood adiposity: a 
prospective birth cohort study in Shanghai, China. Chemosphere 226, 17–23. 
Cole, T.J., Bellizzi, M.C., Flegal, K.M., Dietz, W.H., 2000. Establishing a standard 
definition for child overweight and obesity worldwide: international survey. BMJ 
320, 1240–1243. 
Domazet, S.L., Grøntved, A., Timmermann, A.G., Nielsen, F., Jensen, T.K., 2016. 
Longitudinal associations of exposure to perfluoroalkylated substances in childhood 
and adolescence and indicators of adiposity and glucose metabolism 6 and 12 Years 
later: the European youth heart study. Diabetes Care 39, 1745–1751. 
Du, G., Hu, J., Huang, H., Qin, Y., Han, X., Wu, D., Song, L., Xia, Y., Wang, X., 2013. 
Perfluorooctane sulfonate (PFOS) affects hormone receptor activity, steroidogenesis, 
and expression of endocrine-related genes in vitro and in vivo 32, 353–360. 
EFSA, C.P., Knutsen, H.K., Alexander, J., Barregård, L., Bignami, M., Brüschweiler, B., 
Ceccatelli, S., et al., 2018. Scientific opinion on the risk to human health related to 
the presence of perfluorooctane sulfonic acid and perfluorooctanoic acid in food. 
EFSA J. 16 (12), 5194 https://doi.org/10.2903/j.efsa.2018.5194. 
Esteban Lopez, M., Goen, T., Mol, H., Nübler, S., Haji-Abbas-Zarrabi, K., Koch, H.M., 
Kasper-Sonnenberg, M., Dvorakova, D., Hajslova, J., Antignac, J.-P., Vaccher, V., 
Elbers, I., Thomsen, C., Vorkamp, K., Pedraza – Díaz, S., Kolossa-Gehring, M., 
Casta~no, A., 2021. The European human biomonitoring platform - design and 
implementation of a laboratory quality assurance/quality control (QA/QC) 
programme for selected priority chemicals. Int. J. Hyg Environ. Health 234, 113740. 
Fassler, C.S., Pinney, S.E., Xie, C., Biro, F.M., Pinney, S.M., 2019. Complex relationships 
between perfluorooctanoate, body mass index, insulin resistance and serum lipids in 
young girls. Environ. Res. 176, 108558. 
Forsthuber, M., Kaiser, A.M., Granitzer, S., Hassl, I., Hengstschlager, M., Stangl, H., 
Gundacker, C., 2020. Albumin is the major carrier protein for PFOS, PFOA, PFHxS, 
PFNA and PFDA in human plasma. Environ. Int. 137, 105324. 
Ganzleben, C., Antignac, J.P., Barouki, R., Casta~no, A., Fiddicke, U., Klanova, J., 
Lebret, E., Olea, N., Sarigiannis, D., Schoeters, G.R., Sepai, O., Tolonen, H., Kolossa- 
Gehring, M., 2017. Human biomonitoring as a tool to support chemicals regulation 
in the European Union. Int. J. Hyg Environ. Health 220, 94–97. 
Geiger, S.D., Yao, P., Vaughn, M.G., Qian, Z., 2021. PFAS exposure and overweight/ 
obesity among children in a nationally representative sample. Chemosphere 268, 
128852. 
Gilles, L., Govarts, E., Rambaud, L., Vogel, N., Casta~no, A., Esteban Lopez, M., Rodriguez 
Martin, L., Koppen, G., Remy, S., Vrijheid, M., Montazeri, P., Birks, L., Sepai, O., 
Stewart, L., Fiddicke, U., Loots, I., Knudsen, L.E., Kolossa-Gehring, M., Schoeters, G., 
2021. HBM4EU combines and harmonises human biomonitoring data across the EU, 
building on existing capacity - the HBM4EU survey. Int. J. Hyg Environ. Health 237, 
113809. 
Gilles, L., Govarts, E., Rodriguez Martin, L., Andersson, A.M., Appenzeller, B.M.R., 
Barbone, F., Casta~no, A., Coertjens, D., Den Hond, E., Dzhedzheia, V., Erzen, I., 
Lopez, M.E., Fabelova, L., Fillol, C., Franken, C., Frederiksen, H., Gabriel, C., 
Haug, L.S., Horvat, M., Halldorsson, T.I., Janasik, B., Holcer, N.J., Kakucs, R., 
Karakitsios, S., Katsonouri, A., Klanova, J., Kold-Jensen, T., Kolossa-Gehring, M., 
Konstantinou, C., Koponen, J., Lignell, S., Lindroos, A.K., Makris, K.C., Mazej, D., 
Morrens, B., Murínova Ľ, P., Namorado, S., Pedraza-Diaz, S., Peisker, J., Probst- 
Hensch, N., Rambaud, L., Rosolen, V., Rucic, E., Rüther, M., Sarigiannis, D., 
Tratnik, J.S., Standaert, A., Stewart, L., Szigeti, T., Thomsen, C., Tolonen, H., 
Eiríksdottir, A., Van Nieuwenhuyse, A., Verheyen, V.J., Vlaanderen, J., Vogel, N., 
Wasowicz, W., Weber, T., Zock, J.P., Sepai, O., Schoeters, G., 2022. Harmonization 
of human biomonitoring studies in Europe: characteristics of the HBM4EU-aligned 
studies participants. Int. J. Environ. Res. Publ. Health 19. 
Grün, F., Blumberg, B., 2009. Minireview: the case for obesogens. Mol. Endocrinol. 23, 
1127–1134. 
Gunta, S.S., Mak, R.H., 2013. Is obesity a risk factor for chronic kidney disease in 
children? Pediatr. Nephrol. 28, 1949–1956. 
Haeusler, R.A., Camastra, S., Nannipieri, M., Astiarraga, B., Castro-Perez, J., Xie, D., 
Wang, L., Chakravarthy, M., Ferrannini, E., 2016. Increased bile acid synthesis and 
impaired bile acid transport in human obesity. J. Clin. Endocrinol. Metab. 101, 
1935–1944. 
Harris, R.J., Deeks, J.J., Altman, D.G., Bradburn, M.J., Harbord, R.M., Sterne, J.A.C., 
2008. Meta: Fixed- and Random-Effects Meta-Analysis 8, 3–28. 
Hartman, T.J., Calafat, A.M., Holmes, A.K., Marcus, M., Northstone, K., Flanders, W.D., 
Kato, K., Taylor, E.V., 2017. Prenatal exposure to perfluoroalkyl substances and body 
fatness in girls. Child. Obes. 13, 222–230. 
Janis, J.A., Rifas-Shiman, S.L., Seshasayee, S.M., Sagiv, S., Calafat, A.M., Gold, D.R., 
Coull, B.A., Rosen, C.J., Oken, E., Fleisch, A.F., 2021. Plasma concentrations of per- 
and polyfluoroalkyl substances and body composition from mid-childhood to early 
adolescence. J. Clin. Endocrinol. Metab. 106, e3760–e3770. 
Keil, A.P., Buckley, J.P., O’Brien, K.M., Ferguson, K.K., Zhao, S., White, A.J., 2020. 
A Quantile-Based g-Computation Approach to Addressing the Effects of Exposure 
Mixtures 128, 047004. 
Kelly, D.M., Jones, T.H., 2015. Testosterone and obesity. Obes. Rev. 16, 581–606. 
Lau, C., Anitole, K., Hodes, C., Lai, D., Pfahles-Hutchens, A., Seed, J., 2007. 
Perfluoroalkyl acids: a review of monitoring and toxicological findings. Toxicol. Sci. 
99, 366–394. 
Lauritzen, H.B., Larose, T.L., Øien, T., Sandanger, T.M., Odland, J., van de Bor, M., 
Jacobsen, G.W., 2018. Prenatal exposure to persistent organic pollutants and child 
overweight/obesity at 5-year follow-up: a prospective cohort study. Environ. Health 
17, 9. 
Lee, J.E., Choi, K., 2017. Perfluoroalkyl substances exposure and thyroid hormones in 
humans: epidemiological observations and implications. Ann Pediatr Endocrinol 
Metab 22, 6–14. 
Lee, Y.J., Jung, H.W., Kim, H.Y., Choi, Y.J., Lee, Y.A., 2021. Early-life exposure to per- 
and poly-fluorinated alkyl substances and growth, adiposity, and puberty in 
children: a systematic review. Front. Endocrinol. 12, 683297. 
Li, Y., Andersson, A., Xu, Y., Pineda, D., Nilsson, C.A., Lindh, C.H., Jakobsson, K., 
Fletcher, T., 2022. Determinants of serum half-lives for linear and branched 
perfluoroalkyl substances after long-term high exposure—a study in Ronneby, 
Sweden. Environ. Int. 163, 107198. 
Liu, Y., Li, N., Papandonatos, G.D., Calafat, A.M., Eaton, C.B., Kelsey, K.T., Chen, A., 
Lanphear, B.P., Cecil, K.M., Kalkwarf, H.J., Yolton, K., Braun, J.M., 2020. Exposure 
to per- and polyfluoroalkyl substances and adiposity at age 12 Years: evaluating 
periods of susceptibility. Environ. Sci. Technol. 54, 16039–16049. 
Magnus, P., Birke, C., Vejrup, K., Haugan, A., Alsaker, E., Daltveit, A.K., Handal, M., 
Haugen, M., Høiseth, G., Knudsen, G.P., Paltiel, L., Schreuder, P., Tambs, K., Vold, L., 
Stoltenberg, C., 2016. Cohort profile update: the Norwegian mother and child cohort 
study (MoBa). Int. J. Epidemiol. 45, 382–388. 
Manzano-Salgado, C.B., Casas, M., Lopez-Espinosa, M.J., Ballester, F., I~niguez, C., 
Martinez, D., Romaguera, D., Fernandez-Barres, S., Santa-Marina, L., 
Basterretxea, M., Schettgen, T., Valvi, D., Vioque, J., Sunyer, J., Vrijheid, M., 2017. 
Prenatal exposure to perfluoroalkyl substances and cardiometabolic risk in children 
from the Spanish INMA birth cohort study. Environ. Health Perspect. 125, 097018. 
Margolis, R., Sant, K.E., 2021. Associations between exposures to perfluoroalkyl 
substances and diabetes, hyperglycemia, or insulin resistance: a scoping review. 
J Xenobiot 11, 115–129. 
Martinsson, M., Nielsen, C., Bjork, J., Rylander, L., Malmqvist, E., Lindh, C., Rignell- 
Hydbom, A., 2020. Intrauterine exposure to perfluorinated compounds and 
overweight at age 4: a case-control study. PLoS One 15, e0230137. 
Mora, A.M., Oken, E., Rifas-Shiman, S.L., Webster, T.F., Gillman, M.W., Calafat, A.M., 
Ye, X., Sagiv, S.K., 2017. Prenatal exposure to perfluoroalkyl substances and 
adiposity in early and mid-childhood. Environ. Health Perspect. 125, 467–473. 
Ng, M., Fleming, T., Robinson, M., Thomson, B., Graetz, N., Margono, C., Mullany, E.C., 
Biryukov, S., Abbafati, C., Abera, S.F., Abraham, J.P., Abu-Rmeileh, N.M., 
Achoki, T., AlBuhairan, F.S., Alemu, Z.A., Alfonso, R., Ali, M.K., Ali, R., Guzman, N. 
A., Ammar, W., Anwari, P., Banerjee, A., Barquera, S., Basu, S., Bennett, D.A., 
Bhutta, Z., Blore, J., Cabral, N., Nonato, I.C., Chang, J.C., Chowdhury, R., 
Courville, K.J., Criqui, M.H., Cundiff, D.K., Dabhadkar, K.C., Dandona, L., Davis, A., 
Dayama, A., Dharmaratne, S.D., Ding, E.L., Durrani, A.M., Esteghamati, A., 
Farzadfar, F., Fay, D.F., Feigin, V.L., Flaxman, A., Forouzanfar, M.H., Goto, A., 
Green, M.A., Gupta, R., Hafezi-Nejad, N., Hankey, G.J., Harewood, H.C., 
Havmoeller, R., Hay, S., Hernandez, L., Husseini, A., Idrisov, B.T., Ikeda, N., 
Islami, F., Jahangir, E., Jassal, S.K., Jee, S.H., Jeffreys, M., Jonas, J.B., 
Kabagambe, E.K., Khalifa, S.E., Kengne, A.P., Khader, Y.S., Khang, Y.H., Kim, D., 
Kimokoti, R.W., Kinge, J.M., Kokubo, Y., Kosen, S., Kwan, G., Lai, T., Leinsalu, M., 
Li, Y., Liang, X., Liu, S., Logroscino, G., Lotufo, P.A., Lu, Y., Ma, J., Mainoo, N.K., 
Mensah, G.A., Merriman, T.R., Mokdad, A.H., Moschandreas, J., Naghavi, M., 
Naheed, A., Nand, D., Narayan, K.M., Nelson, E.L., Neuhouser, M.L., Nisar, M.I., 
Ohkubo, T., Oti, S.O., Pedroza, A., Prabhakaran, D., Roy, N., Sampson, U., Seo, H., 
Sepanlou, S.G., Shibuya, K., Shiri, R., Shiue, I., Singh, G.M., Singh, J.A., Skirbekk, V., 
Stapelberg, N.J., Sturua, L., Sykes, B.L., Tobias, M., Tran, B.X., Trasande, L., 
Toyoshima, H., van de Vijver, S., Vasankari, T.J., Veerman, J.L., Velasquez- 
Melendez, G., Vlassov, V.V., Vollset, S.E., Vos, T., Wang, C., Wang, X., 
Weiderpass, E., Werdecker, A., Wright, J.L., Yang, Y.C., Yatsuya, H., Yoon, J., 
Yoon, S.J., Zhao, Y., Zhou, M., Zhu, S., Lopez, A.D., Murray, C.J., Gakidou, E., 2014. 
Global, regional, and national prevalence of overweight and obesity in children and 
adults during 1980-2013: a systematic analysis for the Global Burden of Disease 
Study 2013. Lancet 384, 766–781. 
Nübler, S., Esteban Lopez, M., Casta~no, A., Mol, H.G.J., Haji-Abbas-Zarrabi, K., 
Schafer, M., Müller, J., Hajslova, J., Dvorakova, D., Antignac, J.-P., Koch, H.M., 
T. Schillemans et al.                                                                                                                                                                                                                            
Environmental Pollution 316 (2023) 120566
11
Haug, L.S., Vorkamp, K., Goen, T., 2022. Interlaboratory Comparison Investigations 
(ICIs) and External Quality Assurance Schemes (EQUASs) for human biomonitoring 
of perfluoroalkyl substances (PFASs) in serum as part of the quality assurance 
programme under HBM4EU. Sci. Total Environ. 847, 157481. 
Papadopoulou, E., Stratakis, N., Basaga~na, X., Brantsæter, A.L., Casas, M., Fossati, S., 
Grazulevicien _e, R., Småstuen Haug, L., Heude, B., Maitre, L., McEachan, R.R.C., 
Robinson, O., Roumeliotaki, T., Sabido, E., Borras, E., Urquiza, J., Vafeiadi, M., 
Zhao, Y., Slama, R., Wright, J., Conti, D.V., Vrijheid, M., Chatzi, L., 2021. Prenatal 
and postnatal exposure to PFAS and cardiometabolic factors and inflammation status 
in children from six European cohorts. Environ. Int. 157, 106853. 
Reilly, J.J., Kelly, J., 2011. Long-term impact of overweight and obesity in childhood and 
adolescence on morbidity and premature mortality in adulthood: systematic review. 
Int. J. Obes. 35, 891–898. 
Rosen, M.B., Das, K.P., Rooney, J., Abbott, B., Lau, C., Corton, J.C., 2017. PPARα- 
independent transcriptional targets of perfluoroalkyl acids revealed by transcript 
profiling. Toxicology 387, 95–107. 
Roth, K., Yang, Z., Agarwal, M., Liu, W., Peng, Z., Long, Z., Birbeck, J., Westrick, J., 
Liu, W., Petriello, M.C., 2021. Exposure to a mixture of legacy, alternative, and 
replacement per- and polyfluoroalkyl substances (PFAS) results in sex-dependent 
modulation of cholesterol metabolism and liver injury. Environ. Int. 157, 106843. 
EFSA, Panel on Contaminants in the Food Chain, Schrenk, D., Bignami, M., Bodin, L., 
Chipman, J.K., del Mazo, J., Grasl-Kraupp, B., Hogstrand, C., Hoogenboom, L., 
Leblanc, J.-C., Nebbia, C.S., Nielsen, E., Ntzani, E., Petersen, A., Sand, S., 
Vleminckx, C., Wallace, H., Barregård, L., Ceccatelli, S., Cravedi, J.-P., 
Halldorsson, T.I., Haug, L.S., Johansson, N., Knutsen, H.K., Rose, M., Roudot, A.-C., 
Van Loveren, H., Vollmer, G., Mackay, K., Riolo, F., Schwerdtle, T., 2020. Risk to 
human health related to the presence of perfluoroalkyl substances in food 18, 
e06223. 
Starling, A.P., Adgate, J.L., Hamman, R.F., Kechris, K., Calafat, A.M., Ye, X., Dabelea, D., 
2017. Perfluoroalkyl substances during pregnancy and offspring weight and 
adiposity at birth: examining mediation by maternal fasting glucose in the healthy 
start study. Environ. Health Perspect. 125, 067016. 
Starling, A.P., Adgate, J.L., Hamman, R.F., Kechris, K., Calafat, A.M., Dabelea, D., 2019. 
Prenatal exposure to per- and polyfluoroalkyl substances and infant growth and 
adiposity: the Healthy Start Study. Environ. Int. 131, 104983. 
Stratakis, N., Rock, S., La Merrill, M.A., Saez, M., Robinson, O., Fecht, D., Vrijheid, M., 
Valvi, D., Conti, D.V., McConnell, R., Chatzi, V.L., 2022. Prenatal exposure to 
persistent organic pollutants and childhood obesity: a systematic review and meta- 
analysis of human studies. Obes. Rev. 23, e13383. 
Takacs, M.L., Abbott, B.D., 2007. Activation of mouse and human peroxisome 
proliferator-activated receptors (alpha, beta/delta, gamma) by perfluorooctanoic 
acid and perfluorooctane sulfonate. Toxicol. Sci. 95, 108–117. 
Textor, J., van der Zander, B., Gilthorpe, M.S., Liskiewicz, M., Ellison, G.T., 2017. Robust 
causal inference using directed acyclic graphs: the R package ‘dagitty. Int. J. 
Epidemiol. 45, 1887–1894. 
Thomsen, M.L., Henriksen, L.S., Tinggaard, J., Nielsen, F., Jensen, T.K., Main, K.M., 
2021. Associations between exposure to perfluoroalkyl substances and body fat 
evaluated by DXA and MRI in 109 adolescent boys. Environ. Health 20, 73. 
Vanden Heuvel, J.P., Thompson, J.T., Frame, S.R., Gillies, P.J., 2006. Differential 
activation of nuclear receptors by perfluorinated fatty acid analogs and natural fatty 
acids: a comparison of human, mouse, and rat peroxisome proliferator-activated 
receptor-alpha, -beta, and -gamma, liver X receptor-beta, and retinoid X receptor- 
alpha. Toxicol. Sci. 92, 476–489. 
Vidmar, S.I., Cole, T.J., Pan, H., 2013. Standardizing anthropometric measures in 
children and adolescents with functions for egen: Update. STATA J. 13, 366–378. 
Vrijheid, M., Fossati, S., Maitre, L., Marquez, S., Roumeliotaki, T., Agier, L., 
Andrusaityte, S., Cadiou, S., Casas, M., de Castro, M., Dedele, A., Donaire- 
Gonzalez, D., Grazuleviciene, R., Haug, L.S., McEachan, R., Meltzer, H.M., 
Papadopouplou, E., Robinson, O., Sakhi, A.K., Siroux, V., Sunyer, J., Schwarze, P.E., 
Tamayo-Uria, I., Urquiza, J., Vafeiadi, M., Valentin, A., Warembourg, C., Wright, J., 
Nieuwenhuijsen, M.J., Thomsen, C., Basaga~na, X., Slama, R., Chatzi, L., 2020. Early- 
life environmental exposures and childhood obesity: an exposome-wide approach. 
Environ. Health Perspect. 128, 67009. 
Xie, Y., Yang, Q., Nelson, B.D., DePierre, J.W., 2002. Characterization of the adipose 
tissue atrophy induced by peroxisome proliferators in mice. Lipids 37, 139–146. 
Zhang, Y., Beesoon, S., Zhu, L., Martin, J.W., 2013. Biomonitoring of perfluoroalkyl acids 
in human urine and estimates of biological half-life. Environ. Sci. Technol. 47, 
10619–10627. 
Zhao, B., Lian, Q., Chu, Y., Hardy, D.O., Li, X.K., Ge, R.S., 2011. The inhibition of human 
and rat 11β-hydroxysteroid dehydrogenase 2 by perfluoroalkylated substances. 
J. Steroid Biochem. Mol. Biol. 125, 143–147. 
Zhou, Y., Hu, L.-W., Qian, Z., Chang, J.-J., King, C., Paul, G., Lin, S., Chen, P.-C., Lee, Y. 
L., Dong, G.-H., 2016. Association of perfluoroalkyl substances exposure with 
reproductive hormone levels in adolescents: by sex status. Environ. Int. 94, 189–195. 
Zhou, Y., Hu, L.-W., Qian, Z., Geiger, S.D., Parrish, K.L., Dharmage, S.C., Campbell, B., 
Roponen, M., Jalava, P., Hirvonen, M.-R., Heinrich, J., Zeng, X.-W., Yang, B.-Y., 
Qin, X.-D., Lee, Y.L., Dong, G.-H., 2017. Interaction effects of polyfluoroalkyl 
substances and sex steroid hormones on asthma among children. Sci. Rep. 7, 899. 
T. Schillemans et al.