Risk variables for the development of obesity and type 2 diabetes
van der Meer, Tom
DOI:
10.33612/diss.170143787
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from
it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2021
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
van der Meer, T. (2021). Risk variables for the development of obesity and type 2 diabetes. University of
Groningen. https://doi.org/10.33612/diss.170143787
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Chapter five
Endocrine disrupting chemicals during diet-induced
weight loss – a post-hoc analysis of the LOWER study
Thomas P. van der Meer, Chris H.L. Thio, Martijn van Faassen, André P. van Beek, Harold Snieder, Frank N.R. van Berkum, Ido P. Kema, Konstantinos C. Makris,
Bruce H.R. Wolffenbuttel, Jana V. van Vliet-Ostaptchouk
Abstract
Introduction
The link between exposure to endocrine disrupting chemicals (EDCs) and the rapid increase in prevalence of obesity has recently been suggested. However, the magnitude and health impact of EDC exposure in at-risk populations remain largely unclear. In this study, we investigated the effect of a dietary intervention driven reduction in adipose tissue on the magnitude of urinary EDC exposure and mobilization, and whether higher EDC exposure leads to impaired weight loss in obese individuals.
Methods
In this post-hoc analysis of the Lifestyle, OverWeight, Energy Restriction (LOWER) study from the Netherlands, 218 subjects were included. Five parabens, three bisphenols and thirteen metabolites of eight phthalates were measured in 24-hour urine using LC-MS/MS, before and after three-months of a calory-restricted weight reduction intervention program. Associations between adiposity-related traits and EDCs were tested using multivariable linear regression and linear mixed effects models. A multiple testing correction based on the false discovery rate (FDR) was applied.
Results
After the 3-month intervention, urinary paraben and bisphenol excretions remained similar. Excretions of mono-butyl phthalates and most high-molecular-weight phthalates decreased, whereas mono-ethyl phthalate increased (all FDR<0.05). A reduction in adipose tissue was not associated with higher urinary EDC excretions. Higher baseline EDC excretions were associated with higher post-intervention body-mass index (methyl-, propylparaben), waist circumference (propylparaben, mono-n-butyl phthalate, mono-benzyl phthalate), and body fat percentage (mono-ethyl phthalate, mono-benzyl phthalate). Associations between parabens and body-mass index, and mono-benzyl phthalate and waist circumference and body fat percentage remained after multiple testing correction (all FDR<0.05).
Conclusions
In a study of obese participants, we observed a reduction in most phthalates after a weight reduction intervention. A reduction in adipose tissue may not lead to mobilization and successively to higher urinary EDC excretions. Higher baseline paraben and phthalate exposures were associated with reduced weight loss, suggesting obesogenic properties.
5
Introduction
Obesity is a rapidly growing global problem. Comorbidities of this multifactorial chronic disease include (pre-)diabetes, cardiovascular diseases and musculoskeletal disorders (1). Underlying causes include a surplus of energy due to excessive caloric intake and reduced physical activity, and changes in the metabolism of this energy. Energy metabolism is closely controlled by the endocrine system, involving complex hormonal pathways. Endocrine disrupting chemicals (EDCs) are environmental factors that are proven to possess hormonal activity (2). As such, they can potentially disrupt energy metabolism and may thereby exert obesogenic effects (3,4). Some EDCs are labelled as non-persistent due to half-life times of less than 24 hour (24h) (5–7). These chemicals are metabolised quickly, partly via the first-pass effect of the liver and via the kidneys in the urine. EDCs are applied in a wide variety of consumer products that are used on a daily basis (e.g. plastics, personal care products), and EDCs have been detected ubiquitously in populations across the globe (8–11). Thus, despite their non-persistence, EDCs may exert obesogenic effects that contribute to the global burden of obesity due to their wide-spread and daily use.
One class of EDCs are phenols, which include parabens and bisphenols. Parabens possess estrogenic and anti-androgenic properties (12,13), and have been shown to cause adverse effects on lipid metabolism through the activation of peroxisome-proliferator-activated receptors (PPARs) (14,15). Bisphenol A (BPA) possesses estrogenic properties, and induces adiposity in mice (16). Recently introduced analogues for BPA (i.e. bisphenol F, BPF, and bisphenol S, BPS) have endocrine disrupting potential similar to BPA (17). Another class of EDCs are phthalates, which are known for their anti-androgenic properties, and have also been shown to activate PPARs (18). In humans, exposure to high concentrations of phenols and phthalates has been associated with obesity and adiposity-related traits (19,20,29,21–28). To date, the obesogenic effects of EDCs have mainly been studied in healthy populations, whereas the degree and impact of EDC exposure in obese individuals remain largely unknown. Specifically, the examination of EDCs during a diet intervention in obese individuals may yield insights into exposure to EDCs, their metabolism, and their potential obesogenic effects.
Due to the different properties and application of the EDCs of interest, routes of exposure differ. Dietary intake is the main route of exposure for BPA and high molecular weight (HMW) phthalates (30,31). Exposure to parabens and low molecular weight (LMW) phthalates is believed to occur largely through non-food products (e.g. personal care products) (7,30). Therefore, we hypothesized that restricted dietary intake leads to lower excretions of urinary bisphenol and HMW-phthalates but does not affect excretions of urinary paraben and LMW-phthalates.
A calory restriction diet is often initiated to achieve weight loss through reduction of adipose tissue mass. More persistent EDCs have been shown to be stored in adipose tissue and are released in response to weight loss (32–34). Parabens and bisphenols both show lipophilic properties due to their positive values (1.65 – 3.56) for the log of
the octanol–water partition coefficient (Kow), and have been widely detected in human adipose tissue (35–37). The lipophilic property of phthalates increases with the length of their alcohol chain and thus their molecular weight, making HMW-phthalates (molecular weight: ≥250 g/mol) more prone to be stored in adipose tissue (log(Kow): 4.73 to 8.83) than LMW-phthalates (molecular weight: < 250 g/mol; log(Kow): 1.61 to 4.45). Yet both groups have been detected in adipose tissue (38,39). The molecular weight and log(Kow) for each EDC can be found in Supplementary table 1. Yet, for most of these EDCs, the exact metabolic rates in humans are unknown.
We hypothesized that a reduction in adipose tissue due to dietary intervention will lead to a release of EDCs from this tissue, resulting in higher urinary excretions of lipophilic EDCs.
Thus far, phenols and phthalates have been described to exert obesogenic effects over time in healthy populations (25–27). We hypothesized that higher pre-intervention exposure to EDCs could impair weight loss response in the caloric restriction intervention.
In this study, we aimed to test the three hypotheses described above. To this end, we performed a post-hoc analysis of obese Dutch individuals that participated in a 3-month diet induced weight loss intervention program. We assessed exposure to EDCs (i.e. phenols and phthalates), as measured in 24h urine voids, before (baseline) and after the 3-month intervention. Furthermore, we examined whether a reduction in adipose tissue resulted in higher urinary EDC excretion. Finally, we investigated associations between baseline EDC exposure and weight loss during the intervention.
Materials and Methods
Study Population
Subjects were enrolled in the Lifestyle, OverWeight, Energy Restriction (LOWER) study (clinicaltrials.gov, NCT00862953), an open label, randomized treatment intervention of which details of the study design and eligibility criteria have been previously reported (35). In short, adults with a body mass index (BMI) above 27 kg/m2 were randomized to
one of four energy-restricted diet groups differing in protein and/or carbohydrate content: normal-protein normal-carbohydrate (NPNC), normal-protein low-carbohydrate (NPLC), high-protein normal-carbohydrate (HPNC), and high-protein low-carbohydrate (HPLC). Diet groups were formed based on the following criteria: age, BMI, bodyweight, Baecke’s score for physical activity and Three Factor Eating Questionnaire. The intervention started with a two-week run-in period with 100% energy intake, followed by a three-month energy intake restriction to 33% of the age- and sex-based energy requirements. Required energy intake was based on the Harris-Benedict equation (36). The primary outcome was bodyweight loss after three months. For this post-hoc analysis, subjects with available 24h urine samples at both the end of the two-week run-in period (baseline) and at the end of the 3-month energy restriction period (follow-up) were included. Diabetes status was determined via questionnaire. Due to a lack of regulatory function in glucose
5
homeostasis, subjects with Type 1 Diabetes were excluded. The Medical Ethics committee of the University Medical Centre Groningen approved the study, and all participants gave written informed consent.
EDC measurements
Urine samples were collected between 2008 and 2010. All urine voids were collected over 24h, after which they were pooled, homogenized and an aliquot was taken and stored at -80 °C. A total of 21 EDCs was analyzed at baseline and 3 months follow-up using two offline isotope dilution liquid chromatography tandem mass spectrometry (LC-MS/MS) methods of which full analytical methods have been described previously (37). The first method included methyl, ethyl, propyl, n-butyl and benzyl paraben (MeP, EtP, PrP, n-BuP, BzP, respectively) and three bisphenols (BPA, BPF, BPS). The second method determined the mono-esters of four LMW-phthalates: methyl, ethyl, iso-butyl, and n-butyl phthalate (MMP, MEP, MiBP, MnBP), and the mono-esters of eight HMW-phthalates: 2-ethylhexyl, n-hexyl, 2-ethyl-5-hydroxyhexyl, 2-ethyl-5-oxohexyl, 2-ethyl-5-carboxypentyl, benzyl, iso-nonyl, hydroxy-iso-iso-nonyl, iso-decyl phthalate (MnHP, MEHHP, MEOHP, MECPP, MBzP, MiNP, MHiNP, MiDP, respectively). A limit of detection (LOD) was determined at which the signal could be separated from noise and was calculated as “3.3*S0 / b”, where S0 is the standard deviation of the response and b the slope of the calibration curve. The limit of quantification (LOQ) was set where the imprecision was ≤ 20% and the signal to noise ratio was > 10 on six days. The LOD, LOQ and recovery have been described previously (37). For completeness, this information can be found in Supplementary table 1a and 1b.
As the LOQ indicates a threshold at which an analyte can be precisely measured, only those EDCs of which the measured concentration was above the LOQ in at least 60% of the samples at the timepoint relevant for that analysis were eligible for analyses. For concentrations between LOD and LOQ, values generated by the LC-MS/MS system were used. The concentrations below LOD were imputed with their respective LOD divided by the square root of two (“LOD/√2”). The resulting imputed urinary EDC concentrations (ng/ ml) were multiplied by the volume of the in total collected urine (ml) to calculate absolute EDC excretion over 24h (ng/24h). As the EDCs of interest are rapidly metabolized and excreted from the human body, absolute 24h EDC excretion values were used as a proxy for daily EDC exposure. We transformed the 24h excretion values of EDCs by their natural logarithm (lnEDC) to approximate a normal distribution.
Anthropometry
Extensive information on anthropometric measurements are reported elsewhere (35). In short, waist circumference, weight and body composition were available for baseline and follow-up. BMI was calculated as “weight (kg) / (height (m)* height (m))”. Body composition was assessed using the deuterium (2H
2O) dilution technique, after which body fat was
Statistical analysis
Population characteristics
Participant characteristics were expressed as mean (standard deviation). Categorical variables were shown as number (percentage). Differences between baseline and follow-up were calculated using a paired t-test.
EDC exposure before and after a three-month diet-based weight loss
intervention
The proportion of EDCs detected was expressed as percentage and excretions as median [25th quartile, 75th quartile]. Differences between lnEDC at baseline and follow-up were
tested using linear mixed-effect models, implemented in the lme4 package (38). In these models, lnEDC was used as outcome, with fixed effects for sex and age and a random effect for intercept: “lmer(lnEDC ~ time + age + sex + (1 | ID), data)”. Difference in lnEDC between diet groups was assessed using Analysis of covariance (ANCOVA), adjusting for age, sex, and diabetes status.
The effect of adipose tissue loss on urinary EDC excretion
Whether a loss of adipose tissue was associated with higher urinary EDC excretions was investigated using linear mixed-effect models, with fixed effects for age, sex, diabetes status and diet, and a random effect for intercept. As a proxy for adipose tissue loss, BMI, waist circumference and BF% were transformed into z-scores (i.e. adjusted to the mean, scaled by the standard deviation). Here, the association of change in adipose tissue with change in urinary EDC excretion over time was assessed by introducing an interaction term of EDC with time. With this interaction term, we examined whether change in body fat over time was associated with change in EDCs: “lmer(lnEDC ~ adiposity-related trait * time + (age + sex + diabetes + diet) * time + (1 | ID), data)”.
Associations between adiposity-related changes and urinary EDC
excretions at baseline
In order to investigate the obesogenic effects of EDCs during a dynamic setting of weight loss, effect sizes between EDC excretions at baseline and adiposity-related traits at follow-up were tested using multivariable linear regression models, additionally adjusting for the baseline level of the respective trait: “adiposity-related trait (follow-up) ~ lnEDC (baseline) + age + sex + diabetes + diet + adiposity-related trait (baseline)”.
Multiple testing adjustment and sensitivity analysis
Due to the exploratory nature of this study, we report both raw p-values, as well as adjust for multiple testing using Benjamini and Hochberg False Discovery Rate (FDR) for all individual metabolites of respective groups (i.e. phenols, phthalates). Associations were considered significant at a two-sided FDR < 0.05. Given that tests were not independent
5
due to the correlation between EDC compounds, we consider this FDR cut-off conservative. In order to ascertain the robustness of our findings, we performed a sensitivity analysis by repeating the analyses after excluding subjects with T2D. All analyses were performed in R software version 3.5.3 (39).
Results
Population characteristics
In total, 219 out of 247 subjects had available 24h urine at both time points. One subject was diagnosed with Type 1 Diabetes and was therefore excluded, resulting in a total of 218 subjects. Characteristics of the study population at baseline and follow-up are presented in Table 1. After three months, all adiposity-related traits had significantly decreased (all p-values < 0.0001).
EDC exposure before and after a three-month diet-based weight loss
intervention
At baseline, all parabens were detected above LOD in ≥ 89% of the samples, except for BzP (2%). Bisphenols were detected in 97, 37 and 22% of the samples (BPA, BPF, BPS, respectively), whereas the presence of all phthalate metabolites, but MMP and MnHP (46, 84%, respectively) was ubiquitous (nearly 100%) (Table 2). MiNP, MHiNP and MiDP were not detected in any of the samples, and therefore not presented. Median excretions were similar within and between individuals at baseline and follow-up for all parabens and bisphenols. LMW-phthalate excretions were similar for MMP, whereas higher MEP
Table 1. Anthropometric characteristics of the LOWER population (n=218) at baseline and at 3-month follow-up.
Baseline Follow-up p-value
Sex = Male, n (%) 32 (15)
Age 52 (11)
BMI 36.6 (5.6) 32.3 (5.1) <0.0001
Waist circumference 109 (13) 98.0 (12) <0.0001
Body Fat Percentage 45.3 (6.3) 41.4 (6.6) <0.0001
T2D, n (%) 19 (9) Diet type, n (%) HPLC 54 (25) HPNC 53 (24) LPLC 54 (25) LPNC 57 (26)
Data expressed as mean (standard deviation) or number (proportion, %). Differences between baseline and follow-up were calculated using a paired t-test. Abbreviations: BMI, body-mass index; T2D, type 2 diabetes; HPLC, high-protein carbohydrate; HPNC, high-protein normal-carbohydrate; LPLC, low-protein low-carbohydrate; LPNC, low-low-protein normal carbohydrate; n, number.
Ta bl e 2. C o m p ar is o ns b et w ee n u ri na ry e xc re tio ns o f m ea su re d E D C s a t b as el in e a nd 3 -m o nt h f o llo w -u p ( µg /2 4h ). N>LOD (%) N>LOQ (%) median [25th quartile; 75th quartile] FDR-value Baseline Follow-up Baseline Follow-up Baseline Follow-up Parabens* Methyl paraben MeP 218 (100) 217 (100) 215 (99) 212 (97) 51.3 [11.8; 148] 61.5 [16.3; 150] 0.5111 Ethyl paraben EtP 212 (97) 203 (93) 143 (66) 149 (68) 2.41 [0.78; 13.3] 4.81 [1.05; 17.4] 0.1415 Pr opyl paraben PrP 218 (100) 216 (99) 150 (69) 152 (70) 8.66 [1.54; 32.7] 8.81 [1.62; 38.1] 0.7205 Benzyl paraben BzP 5 (2) 4 (2) 0 (0) 0 (0) 0.96 [0.32; 4.89] 1.07 [0.37; 5.19] 0.5367 n-Butyl paraben n-BuP 195 (89) 196 (90) 72 (33) 74 (34) 0.44 [0.19; 0.72] 0.35 [0.32; 0.46] 0.7979 Bisphenols** Bisphenol A B PA 212 (97) 204 (94) 132 (61) 132 (61) 3.51 [2.05; 6.05] 4.06 [2.01; 8.11] 0.5367 Bisphenol F BPF 80 (37) 66 (30) 25 (11) 19 (9) 1.60 [0.76; 4.34] 1.51 [0.91; 4.16] 0.7205 Bisphenol S BPS 49 (22) 49 (22) 10 (5) 16 (7) 0.36 [0.19; 0.85] 0.57 [0.22; 2.08] 0.5367 Low-Molecular -W eight phthalates* Mono-methyl phthalate MMP 100 (46) 103 (47) 15 (7) 16 (7) 1.90 [1.18; 3.22] 1.87 [1.32; 2.97] 0.8718 Mono-ethyl phthalate MEP 218 (100) 217 (100) 218 (100) 217 (100) 217 [75.8; 588] 388 [107; 1146] <0.0001 Mono-iso-butyl phthalate MiBP 217 (100) 217 (100) 217 (100) 217 (100) 55.4 [36.9; 91.6] 48.0 [29.5; 78.7] <0.0001 Mono-n-butyl phthalate MnBP 217 (100) 217 (100) 217 (100) 217 (100) 45.6 [31.5; 66.5] 38.4 [25.3; 62.5] <0.0001 High-Molecular -W eight phthalates** Mono-(2-ethylhexyl) phthalate MEHP 218 (100) 217 (100) 136 (62) 91 (42) 5.44 [3.45; 8.98] 3.57 [1.98; 6.46] <0.0001 Mono-n-hexyl phthalate MnHP 184 (84) 155 (71) 9 (4) 8 (4) 0.36 [0.26; 0.52] 0.36 [0.27; 0.50] 0.4345 Mono-(2-ethyl-5-hydr oxyhexyl) phthalate MEHHP 218 (100) 217 (100) 218 (100) 217 (100) 26.1 [18.4; 40.8] 20.1 [11.7; 27.6] <0.0001 Mono-(2-ethyl-5-oxohexyl) phthalate MEOHP 217 (100) 216 (99) 217 (100) 216 (99) 17.7 [12.4; 27.0] 13.1 [7.94; 19.4] <0.0001 Mono-(2-ethyl-5-carboxypentyl) phthalate MECPP 218 (100) 217 (100) 218 (100) 217 (100) 27.8 [18.9; 43.7] 21.1 [12.6; 32.9] <0.0001 Mono-benzyl phthalate MBzP 216 (99) 212 (97) 216 (99) 212 (97) 20.0 [10.4; 36.2] 13.8 [7.81; 30.0] <0.0001 Dif fer
ences between time-points wer
e investigated for volume-adjusted, log-transfor
med 24h EDC excr
etions, using linear mixed ef
fect models adjusting for sex and
age. FDR-values wer
e calculated by corr
ecting
p
-values for multiple testing (Benjamini-Hochber
g False Discovery Rate < 0.05) and expr
essed bold when significant.
Abbr
eviations: LOD, Limit of Detection; LOQ, Limit of Quantification; Mono-iso-nonyl phthalate, Mono-hydr
oxy-iso-nonyl phthalate, and Mono-iso-decyl phthalate
did not exceed
the LOD in any of the samples, and ar e ther efor
e not displayed. Main r
oute of exposur
e: *non-food pr
oducts; **food-pr
5
and lower MiBP and MnBP excretions were detected at follow-up (FDR = 0.9; all other FDR < 0.0001). Regarding HMW-phthalates, all metabolite excretions but MnHP (FDR = 0.4), were lower at follow-up (all FDR < 0.0001). Medians and percentiles of EDC excretions including samples < LOD are shown in Supplementary Table 2. No significant differences in EDC excretions either at baseline or at follow-up were observed between the different dietary groups (Supplementary Table 3a, 3b).
The effect of adipose tissue loss on urinary EDC excretion
Differences in adiposity-related traits within and between individuals during the intervention period did not result in a change in EDC excretions given that none of the tested associations were statistically significant (Figure 1a-c and Supplementary table 4).
Associations between adiposity-related changes and urinary EDC
excretions at baseline
To assess whether high pre-intervention EDC exposure impairs the response to a diet-induced weight loss intervention, associations between urinary EDC excretions at baseline and adiposity-related traits at follow-up were assessed (Figure 2a-c). In the context of significant decreases in adiposity-related traits in the overall study population (Table 1), some EDCs were associated with an impaired reduction in these traits. Higher baseline excretions of the phenols MeP and PrP were associated with an impaired reduction in adiposity-related traits at follow-up (BMI: MeP, PrP; waist circumference: PrP, all p ≤ 0.028). Regarding phthalates, higher excretions at baseline were associated with impaired reduction in waist circumference (MnBP, p = 0.031; MBzP, p = 0.005) and BF% (MEP, p = 0.044; MBzP, p = 0.005). After adjusting for multiple testing, associations between MeP and PrP, and BMI, and MBzP and waist circumference and BF% remained significant (FDR < 0.05). Effect sizes, confidence-intervals, p- and FDR-values are reported in Supplementary table 5. As sensitivity analysis, we repeated the analysis after excluding all subjects with T2D at baseline. Except from the association between BF% and MEP (p = 0.3), all associations retained their significance (Supplementary table 6).
Discussion
In an obese population, urinary excretion of phenols and phthalates was assessed before and after a 3-month dietary weight loss intervention program. We investigated the effects of dietary intervention on exposure to these EDCs, and whether a reduction of adipose tissue leads to mobilization and successively an increase in urinary excretions. Furthermore, we assessed the obesogenic effects of these chemicals in a setting of weight loss.
Effects of a dietary weight loss intervention on urinary EDC excretion
Main routes of exposure to phenols and phthalates include inhalation, ingestion and transdermal exposure, and largely depend on the foodstuff or consumer products in which
Figure 1. Associations between change in adiposity-related traits within and between individuals and change in EDC excretions during a diet-induced weight loss program. Regression coefficients with 95% Confidence Intervals of associations for A. Body-mass index, B. Waist circumference and C. Body fat percentage and endocrine disrupting chemicals (EDCs). Formula: “lmer(lnEDC ~
adiposity-related trait * time + (age + sex + diabetes + diet) * time + (1 | ID), data)”. EDC levels were imputed,
volume-adjusted, and natural log-transformed. Linear mixed effect models with fixed effects for age, sex, diabetes status and diet, and random effects for intercept and time. An interaction term between adiposity-related trait and time was introduced in the models. Beta coefficients represent effects per
e-fold increase per month in EDC parameter per month. For full names of compounds, see table 2. For
exact estimates, standard errors and raw p-values, see supplementary table 3. Red: parabens; orange: bisphenols; green: Low Molecular Weight-phthalates; blue: High Molecular Weight-phthalates.
5
Figure 2. Associations between baseline EDC excretions and adiposity-related traits after a diet-induced weight loss program. Regression coefficients with 95% Confidence Intervals derived from ordinary least squares linear regression for A. Body-mass index, B. Waist circumference and C. Body fat percentage and endocrine disrupting chemicals (EDCs). Formula: “adiposity-related trait
(follow-up) ~ lnEDC (baseline) + age + sex + diabetes + diet + adiposity-related trait (baseline)”. EDC levels
were imputed, volume-adjusted, and natural log-transformed. Models were corrected for age, sex, diabetes status, diet, and respective trait at baseline. Regression coefficients represent effects per
e-fold increase in EDC parameter over 3 months. Positive regression coefficients indicate impairment
of reduction of the respective trait. For full names of compounds, see table 2. For exact estimates, standard errors and raw p-values, see supplementary table 4. Red: parabens; orange: bisphenol A; green: Low Molecular Weight-phthalates; blue: High Molecular Weight-phthalates. *Benjamini-Hochberg False Discovery Rate < 0.05.
these EDCs are contained. Uses of BPA and HMW-phthalates include food packaging and water bottles (40,41), and urinary excretions of these chemicals have been shown to rapidly decrease after a 48h period of fasting (30,31). On the other hand, parabens and LMW-phthalates are mainly used in personal care and cosmetic products (42), and their main route of exposure is believed to be mainly transdermal (7). Consistent with this transdermal route of exposure, urinary LMW-phthalate excretions have been shown to remain stable during a period of fasting (30). In the present study, caloric intake was restricted to 33% of the daily required consumption. As exposure to bisphenols and HMW-phthalates is believed to be mainly through food products, we expected urinary excretions to decrease over time. Given the transdermal route of exposure of parabens and LMW-phthalates, and their lack of response to food deprivation, we hypothesized that urinary excretions of parabens and LMW-phthalates remain stable during the intervention.
After the intervention, all urinary HMW-phthalate excretions but MnHP decreased, corroborating the dietary source of these EDC exposures. Counter to expectation, urinary bisphenol excretions remained stable during the intervention. Two weeks prior to the start of the intervention, subjects were consulted by a dietician to meet an energy intake restriction to 100% of the calculated energy requirement. As the consumption of less calorie-dense food products has been associated with lower BPA concentrations (43), this change in diet prior to the baseline measurement could have influenced BPA excretions. Median pre-intervention urinary BPA excretions were nearly half of those of a general Dutch population (44) (median [25th quartile; 75th quartile]: 3.51 [2.05; 6.05] µg/24h vs
1.65 [1.02; 2.87] µg/24h). This implies that a change in type of food products rather than a quantitative restriction could lead to a decrease in exposure to BPA. This observation is consistent with the results of an intervention study by Rudel et al, which showed that BPA excretions were substantially reduced by avoiding cans, plastic food packaging and plastic cooking utensils (45).
Consistent with our hypothesis, urinary paraben excretions did not differ pre- and post-intervention. Contrary to expectations however, the excretion of LMW-phthalates did change in response to the intervention. Urinary excretion of dibutyl phthalate metabolites were observed to decrease 1.2-fold from baseline, while MEP excretion showed a 1.8-fold increase. At baseline, we detected higher concentrations compared to a general Dutch population (median [25th quartile; 75th quartile]: 217 [75.8; 588] µg/24h vs 83.5 [34; 247]
µg/24h), but similar a Belgian study investigating weight loss in an obese population (104 µg/ml vs 75 µg/ml) (46). Though Dirtu et al observed a decrease in MEP concentrations over three months, the proportion of MEP relative to other phthalates increased. LMW-phthalates have proven to remain stable over a short period of fasting (30) and thus are considered to be unaffected by dietary sources. On the other hand, the sole structural change during the intervention being dietary in combination with the relatively short intervention period (i.e. 3 months) makes changes in other non-dietary sources of exposure such as personal care products unlikely. Moreover, several studies reported direct associations between higher MEP excretion and vegetable consumption (47,48).
5
In the present study, subjects shifted to less calorie dense products such as vegetables to achieve a reduction in caloric intake. This shift to vegetables is a potential explanation of the increase in urinary MEP excretion during the intervention. Given its purported obesogenic properties (27), it is alarming that exposure to MEP almost doubled during the intervention, potentially counteracting weight loss in the longer term.
Effects of adipose tissue loss on 24h urinary EDC excretion
A reduction in dietary intake had the expected results for urinary paraben and HMW-phthalate excretions, as excretions of these compounds reduced during the study period. Unexpectedly, excretion of bisphenols remained stable over time, whereas MEP increased. For persistent EDCs, weight loss has been shown to lead to a mobilization of chemicals, resulting in higher circulating concentrations (32–34). Phenols and phthalates have been widely detected in human adipose tissue (43,49–52). As these EDCs are suspected of obesogenic properties, an increase in exposure due to weight loss could in turn lead to weight gain creating a yo-yo effect. Quirós et al. found associations between higher urinary concentrations of parabens and lower body fat, which they hypothesized was due to an increase in storage capacity and therefore reduced excretion (53). Based on those findings, we hypothesized that a reduction of adipose tissue would lead to the release of stored EDCs and therefore increase their urinary excretions. However, we found no significant associations between a decrease in adiposity-related traits and an increase in urinary EDC excretion. Therefore, we found no indication that a reduction of adipose tissue leads to the mobilization of phenols and phthalates and therewith increase in exposure and excretion.
Obesogenic effects of EDCs during a weight loss program
Consistent with our hypothesis, we observed that higher baseline EDC exposure was associated with an impaired reduction in BMI, waist circumference and BF% in response to a weight loss program. This suggests the existence of obesogenic effects of EDC that impair the response to weight loss through dietary intervention.
Parabens have been described to have obesogenic effects in vitro and in vivo (14,15), although cross-sectional studies in humans yielded inconclusive results (22,44,53). To the best of our knowledge, this is the first prospective study in humans that described obesogenic effects of parabens. Phthalates have been positively associated with adiposity-related traits in large cross-sectional studies (20,21,26). Two studies prospectively investigated the effects of EDCs on adiposity-related traits. Song et al found that higher urinary concentrations of MBzP and the sum of butyl phthalate metabolites were associated with weight gain after ten years in a healthy population (25). A study conducted by Díaz Santana et al found higher concentrations of the metabolites MEP and MnBP to be associated with weight gain after three years in post-menopausal women (27). These findings are consistent with the associations found between higher baseline
excretions of MEP, MnBP, and MBzP and a reduced adipose tissue loss in current study. The association between MEOHP and weight gain reported by Díaz Santana et al could not be reproduced. This could be due to a number of factors, including differences in population characteristics (postmenopausal women vs obese subjects), sample size (n = 660 vs n = 218), biological pathways involved (weight gain versus impaired weight loss) and study design (observational vs experimental).
Some modelling approaches better accommodate non-normally distributed data. Therefore, we repeated our analysis using untransformed EDC concentrations in generalized linear (mixed) models. Yet, generalized models failed to converge and therefore we do not report the results of this procedure.
Strengths and Limitations
To our knowledge, this is the first study to investigate changes in body burden of this range of EDCs during a restrictive diet-based weight loss intervention. Due to the design we were able to obtain valuable insights in the effects of diet on exposure to EDCs, the effects of adipose tissue loss on EDC excretion and the obesogenic effects of EDCs in a setting of active weight loss intervention program in obese individuals. One of the strengths of the study is the availability of body composition information using the deuterium (2H
2O)
dilution technique. This way we were able to calculate the BF% and use this as an accurate marker for adipose tissue mass. A limitation of the study design is the lack of a control group. Although the intervention took a relative short period of time, potential bias in exposure to EDCs due to change in lifestyle factors other than dietary intake cannot be ruled out.
Phenols and phthalates are known to be rapidly metabolized and excreted from the body. Due to their short half-lives (5–7), EDC excretions are prone to variate throughout the day resulting in intra-person variability (54). As 24h urine is strenuous to collect, most studies use spot- or morning-urine samples yielding less precise estimations of daily exposure (55). By using 24h urine in this study, we were able to provide a reliable estimation of the average daily excretion of these EDCs.
A major strength of the present study was that a wide variety of compounds, of two common classes of EDCs, were measured in this study. This facilitated a comprehensive assessment of EDCs and their relation to adiposity-related traits. Most previous studies only focused on a single compound, or class of compounds, and may therefore have missed relevant associations. However, this increased the multiple testing burden, necessitating correction based on the false discovery rate.
Several studies have reported associations between higher urinary BPA concentrations and higher adiposity-related traits (19,25,29). However, in the current study, BPA was not associated with BMI, waist circumference or BF%. This lack of findings might be explained by the fact that we observed much lower urinary BPA excretions at baseline compared to a general population, and therefore the BPA excretions detected in the current study might not be representative for long-term exposure. Although we measured EDC excretion
5
at two different time points, the two-week run-in period prior to the first measurement could have affected exposure to BPA. Introducing urine measurements before and shortly (e.g. 48h) after the start of the run-in period would give more insight in the short-term effects of product restriction on exposure to EDCs.
Implications
In this study, we gained insights in the restrictive dietary intervention program effects on the body burden of phenols and phthalates and their possible obesogenic effects in a dynamic setting of a 3-month weight loss program. Moreover, weight loss interventions might benefit from the reduction of EDCs. We did not find any indication of higher EDC excretion as a result of release from diminishing adipose tissue, and further research is warranted to replicate our findings in animal models.
Conclusion
This study assessed the exposure to common EDCs (i.e. phenols and phthalates), as measured by 24h urine excretions of EDCs, in obese individuals before and after a restrictive diet-induced weight loss program. The body burden of butyl phthalate metabolites and HMW-phthalates decreased over the course of the intervention, while MEP concentrations increased. In contrast, the body burden of BPA and parabens remained unaffected by the dietary caloric restriction intervention program. The documented reduction in adipose tissue did not lead to higher urinary EDC excretion. Higher baseline urinary excretion of parabens and phthalates was associated with impaired loss of adipose tissue, suggesting obesogenic properties in a dynamic setting of weight loss.
Acknowledgements
We thank Irene van der Kooi-Wijbenga for her contribution to the technical measurements. This work was supported by a Diabetes Funds Junior Fellowship from the Dutch Diabetes Research Foundation (to JVvVO, project no. 2013.81.1673)
Competing interests
The authors declare that they have no competing interests. This work was supported by a Diabetes Funds Junior Fellowship from the Dutch Diabetes Research Foundation (to JVvVO, project no. 2013.81.1673).
Author contributions
TPvdM performed the analysis, interpreted data and wrote the manuscript; CHLT interpreted the data and analyses and wrote the manuscript; MvF coordinated and performed the measurements; APvB acquired data and/or provided study materials; HS contributed to interpretation of the data and analyses; FNRvB acquired data and/ or provided study materials; IPK contributed to interpretation of the data and analyses; KCM contributed to interpretation of the data and analyses; BHRW acquired data and/or provided study materials; and JVvVO conceived, designed and implemented the study, was involved in data acquisition and contributed to writing the manuscript. All authors reviewed and approved the final manuscript.
5
References
1. WHO. Global status report on noncommunicable diseases 2014. World Health. 2014.
2. Newbold RR, Padilla-Banks E, Jefferson WN. Environmental estrogens and obesity. Mol Cell Endocrinol. 2009; 304: 84-89.
3. Heindel JJ, vom Saal FS, Blumberg B, Bovolin P, Calamandrei G, Ceresini G, et al. Parma consensus statement on metabolic disruptors. Env Heal. 2015; 14: 54.
4. Darbre PD. Endocrine Disruptors and Obesity. Curr Obes Rep. 2017; 6: 18–27.
5. Anderson WAC, Castle L, Scotter MJ, Massey RC, Springall C. A biomarker approach to measuring human dietary exposure to certain phthalate diesters. Food Addit Contam. 2001; 18: 1068-1074. 6. Völkel W, Colnot T, Csanády GA, Filser JG, Dekant W. Metabolism and kinetics of bisphenol a in
humans at low doses following oral administration. Chem Res Toxicol. 2002; 15: 1281-1287. 7. Janjua NR, Frederiksen H, Skakkebæk NE, Wulf HC, Andersson AM. Urinary excretion of phthalates
and paraben after repeated whole-body topical application in humans. Int J Androl. 2008; 31: 118-130. 8. Frederiksen H, Jensen TK, Jorgensen N, Kyhl HB, Husby S, Skakkebaek NE, et al. Human urinary excretion of non-persistent environmental chemicals: an overview of Danish data collected between 2006 and 2012. Reproduction. 2014; 147: 555-565.
9. Kasper-Sonnenberg M, Koch HM, Wittsiepe J, Bruning T, Wilhelm M. Phthalate metabolites and bisphenol A in urines from German school-aged children: results of the Duisburg birth cohort and Bochum cohort studies. Int J Hyg Env Heal. 2014; 217: 830-838.
10. Cutanda F, Koch HM, Esteban M, Sanchez J, Angerer J, Castano A. Urinary levels of eight phthalate metabolites and bisphenol A in mother-child pairs from two Spanish locations. Int J Hyg Env Heal. 2015; 218: 47–57.
11. Koch HM, Rüther M, Schütze A, Conrad A, Pälmke C, Apel P, et al. Phthalate metabolites in 24-h urine samples of the German Environmental Specimen Bank (ESB) from 1988 to 2015 and a comparison with US NHANES data from 1999 to 2012. Int J Hyg Environ Health. 2017; 220: 130-141.
12. Satoh K, Nonaka R, Ohyama K, Nagai F. Androgenic and Antiandrogenic Effects of Alkylphenols and Parabens Assessed Using the Reporter Gene Assay with Stably Transfected CHO-K1 Cells (AR-EcoScreen System). J Heal Sci. 2005; 51: 557-568.
13. Watanabe Y, Kojima H, Takeuchi S, Uramaru N, Ohta S, Kitamura S. Comparative study on
transcriptional activity of 17 parabens mediated by estrogen receptor α and β and androgen
receptor. Food Chem Toxicol. 2013; 57: 227-234.
14. Taxvig C, Dreisig K, Boberg J, Nellemann C, Schelde AB, Pedersen D, et al. Differential effects of environmental chemicals and food contaminants on adipogenesis, biomarker release and PPARgamma activation. Mol Cell Endocrinol. 2012; 361: 106-115.
15. Hu P, Kennedy RC, Chen X, Zhang J, Shen CL, Chen J, et al. Differential effects on adiposity and serum marker of bone formation by post-weaning exposure to methylparaben and butylparaben. Environ Sci Pollut Res. 2016; 23: 21957-21968.
16. Yang M, Chen M, Wang J, Xu M, Sun J, Ding L, et al. Bisphenol a promotes adiposity and inflammation in a nonmonotonic dose-response way in 5-week-old male and female C57BL/6J mice fed a low-calorie diet. Endocrinology. 2016; 157: 2333-2345.
17. Rochester JR, Bolden AL. Bisphenol S and F: A Systematic Review and Comparison of the Hormonal Activity of Bisphenol A Substitutes. Env Heal Perspect. 2015; 123: 643-650.
18. Desvergne B, Feige JN, Casals-Casas C. PPAR-mediated activity of phthalates: A link to the obesity epidemic? Mol Cell Endocrinol. 2009; 304: 43-48.
19. Liu B, Lehmler H-J, Sun Y, Xu G, Liu Y, Zong G, et al. Bisphenol A substitutes and obesity in US adults: analysis of a population-based, cross-sectional study. Lancet Planet Heal. 2017; 1:e114-e122. 20. James-Todd TM, Huang T, Seely EW, Saxena AR. The association between phthalates and
metabolic syndrome: the National Health and Nutrition Examination Survey 2001-2010. Env Heal. 2016; 15: 52.
21. Hatch EE, Nelson JW, Qureshi MM, Weinberg J, Moore LL, Singer M, et al. Association of urinary phthalate metabolite concentrations with body mass index and waist circumference: A cross-sectional study of NHANES data, 1999-2002. Environ Health. 2008; 7: 27.
22. Kolatorova L, Sramkova M, Vitku J, Vcelak J, Lischkova O, Starka L, et al. Parabens and their relation to obesity. Physiol Res. 2018; 67: S465-S472.
23. Stahlhut RW, van Wijngaarden E, Dye TD, Cook S, Swan SH. Concentrations of urinary phthalate metabolites are associated with increased waist circumference and insulin resistance in adult U.S. males. Environ Health Perspect. 2007; 115: 876-882.
24. Shankar A, Teppala S, Sabanayagam C. Urinary Bisphenol A Levels and Measures of Obesity: Results from the National Health and Nutrition Examination Survey 2003–2008. ISRN Endocrinol. 2012; 2012: 965243.
25. Song Y, Hauser R, Hu FB, Franke AA, Liu S, Sun Q. Urinary concentrations of bisphenol A and phthalate metabolites and weight change: A prospective investigation in US women. Int J Obes. 2014; 38: 1532-1537.
26. Lind PM, Roos V, Rönn M, Johansson L, Ahlström H, Kullberg J, et al. Serum concentrations of phthalate metabolites are related to abdominal fat distribution two years later in elderly women. Environ Health. 2012; 11: 21.
27. Díaz Santana MV, Hankinson SE, Bigelow C, Sturgeon SR, Zoeller RT, Tinker L, et al. Urinary concentrations of phthalate biomarkers and weight change among postmenopausal women: A prospective cohort study. Environ Health. 2019; 18: 20.
28. Dirinck E, Dirtu AC, Geens T, Covaci A, Van Gaal L, Jorens PG. Urinary phthalate metabolites are associated with insulin resistance in obese subjects. Environ Res. 2015; 137: 419-423.
29. Mouneimne Y, Nasrallah M, Khoueiry-Zgheib N, Nasreddine L, Nakhoul N, Ismail H, et al. Bisphenol A urinary level, its correlates, and association with cardiometabolic risks in Lebanese urban adults. Environ Monit Assess. 2017; 189: 517.
30. Koch HM, Lorber M, Christensen KLY, Pälmke C, Koslitz S, Brüning T. Identifying sources of phthalate exposure with human biomonitoring: Results of a 48h fasting study with urine collection and personal activity patterns. Int J Hyg Environ Health. 2013; 16: 672-681.
31. Christensen KLY, Lorber M, Koslitz S, Brüning T, Koch HM. The contribution of diet to total bisphenol A body burden in humans: Results of a 48hour fasting study. Environ Int. 2012; 50: 7-14.
5
32. Jandacek RJ, Anderson N, Liu M, Zheng S, Yang Q, Tso P. Effects of yo-yo diet, caloric restriction, and olestra on tissue distribution of hexachlorobenzene. Am J Physiol Gastrointest Liver Physiol. 2005; 288: G292-299.
33. Imbeault P, Tremblay A, Simoneau J-A, Joanisse DR. Weight loss-induced rise in plasma pollutant is associated with reduced skeletal muscle oxidative capacity. Am J Physiol Endocrinol Metab. 2002; 282: E574-579.
34. Malarvannan G, Van Hoorenbeeck K, Deguchtenaere A, Verhulst SL, Dirinck E, Van Gaal L, et al. Dynamics of persistent organic pollutants in obese adolescents during weight loss. Environ Int. 2018; 110: 80-87.
35. Soenen S, Bonomi AG, Lemmens SG, Scholte J, Thijssen MA, van Berkum F, et al. Relatively high-protein or “low-carb” energy-restricted diets for body weight loss and body weight maintenance? Physiol Behav. 2012; 107: 374-380.
36. Harris JA, Benedict FG. A Biometric Study of Human Basal Metabolism. Proc Natl Acad Sci U S A. 1918; 4: 370-373.
37. van der Meer TP, van Faassen M, Frederiksen H, van Beek AP, Wolffenbuttel BHR, Kema IP et al. Development and interlaboratory validation of two fast UPLC-MS/MS methods determining urinary bisphenols, parabens and phthalates. J Anal Toxicol. 2019; 43: 452-464.
38. Bates D, Maechler M, Bolker BSW. Fitting Linear Mixed-Effects Models Using lme4. Journal of Statistical Software. 2015; 67: 1-48.
39. Team RDC. R: A language and environment for statistical computing [Internet]. Vienna, Austria: R Foundation for Statistical Computing; 2017.
40. Burridge E. Bisphenol A: product profile. Eur Chem News. 2003; 78: 14-20.
41. Serrano SE, Braun J, Trasande L, Dills R, Sathyanarayana S. Phthalates and diet: A review of the food monitoring and epidemiology data. Environmental Health. 2014; 13: 43.
42. National Research Council. Phthalates and Cumulative Risk Assessment. Phthalates and Cumulative Risk Assessment: The Tasks Ahead. 2008.
43. Artacho-Cordón F, Arrebola JP, Nielsen O, Hernández P, Skakkebaek NE, Fernández MF, et al. Assumed non-persistent environmental chemicals in human adipose tissue; matrix stability and correlation with levels measured in urine and serum. Env Res. 2017; 156: 120-127.
44. van der Meer TP, van Faassen M, van Beek AP, Snieder H, Kema IP, Wolffenbuttel BHR, et al. Exposure to Endocrine Disrupting Chemicals in the Dutch general population is associated with adiposity-related traits. Sci Rep. 2020; 10: 9311.
45. Rudel RA, Gray JM, Engel CL, Rawsthorne TW, Dodson RE, Ackerman JM, et al. Food packaging and bisphenol A and bis(2-ethyhexyl) phthalate exposure: Findings from a dietary intervention. Environ Health Perspect. 2011; 119: 914-920.
46. Dirtu AC, Geens T, Dirinck E, Malarvannan G, Neels H, Van Gaal L, et al. Phthalate metabolites in obese individuals undergoing weight loss: Urinary levels and estimation of the phthalates daily intake. Environ Int. 2013; 59: 344-353.
47. Trasande L, Sathyanarayana S, Jo Messito M, S. Gross R, Attina TM, Mendelsohn AL. Phthalates and the diets of US children and adolescents. Environ Res. 2013; 126: 84-90.
48. Colacino JA, Harris TR, Schecter A. Dietary intake is associated with phthalate body burden in a nationally representative sample. Environ Health Perspect. 2010; 118: 998-1003.
49. Artacho-Cordón F, Arrebola JP, Nielsen O, Hernández P, Skakkebaek NE, Fernández MF, et al. Assumed non-persistent environmental chemicals in human adipose tissue; matrix stability and correlation with levels measured in urine and serum. Env Res. 2017; 156: 120-127.
50. Wang L, Asimakopoulos AG, Kannan K. Accumulation of 19 environmental phenolic and xenobiotic heterocyclic aromatic compounds in human adipose tissue. Env Int. 2015; 78: 45-50.
51. Mes J, Coffin DE, Campbell DS. Di-n-butyl-and Di-2-ethylhexyl phthalate in human adipose tissue. Bull Environ Contam Toxicol. 1974; 12: 721-725.
52. Mathieu-Denoncourt J, Wallace SJ, de Solla SR, Langlois VS. Influence of Lipophilicity on the Toxicity of Bisphenol A and Phthalates to Aquatic Organisms. Bull Environ Contam Toxicol. 2016; 97: 4-10. 53. Quirós-Alcalá L, Buckley JP, Boyle M. Parabens and measures of adiposity among adults and children
from the U.S. general population: NHANES 2007–2014. Int J Hyg Environ Health. 2018; 221: 652-660. 54. Preau JL, Wong LY, Silva MJ, Needham LL, Calafat AM. Variability over 1 week in the urinary
concentrations of metabolites of diethyl phthalate and di(2-ethylhexyl) phthalate among eight adults: An observational study. Environ Health Perspect. 2010; 118: 1748-1754.
55. Sun Q, Bertrand KA, Franke AA, Rosner B, Curhan GC, Willett WC. Reproducibility of urinary biomarkers in multiple 24-h urine samples. Am J Clin Nutr. 2017; 105: 159–168.
5
Supplementary materials
Su p p lement ar y t ab le 1 . R aw v al ue s ( µg /m l) o f E nd o cr in e D is ru p tin g C he m ic al s a t b as el in e a nd 3 -m o nt h f o llo w -u p, m ea su re d i n 2 4-ho ur u ri ne s am p le s. Raw values (µg/ml)Volume adjusted values (µg/24h)
Baseline Follow-up Baseline Follow-up Phenols Methyl paraben MeP 6.00 [22.9, 70.4] 7.24 [31.2, 69.5] 51.3 [11.8, 148] 61.4 [16.1, 149] Ethyl paraben EtP 0.33 [1.08, 5.58] 0.3 [1.72, 7.65] 2.19 [0.72, 12.7] 3.16 [0.72, 16.4] Pr opyl paraben PrP 0.57 [4.04, 15.2] 0.83 [4.39, 16.4] 8.66 [1.54, 32.7] 8.59 [1.54, 37.2] n-Butyl paraben n-BuP 0.11 [0.32, 1.76] 0.15 [0.44, 2.02] 0.75 [0.23, 3.72] 0.84 [0.27, 4.04] Bisphenol A B PA 1.02 [1.65, 2.87] 0.8 [1.86, 3.96] 3.39 [2.00, 6.02] 3.79 [1.76, 7.61] Bisphenol F BPF <LOD [<LOD, 0.41] <LOD [<LOD, 0.38] <LOD [<LOD, 0.87] <LOD [<LOD, 0.74] Phthalates Mono-methyl phthalate MMP <LOD [<LOD, 0.80] <LOD [<LOD, 0.92] <LOD [<LOD, 1.74] <LOD [<LOD, 1.86] Mono-ethyl phthalate MEP 35.6 [104, 286] 51.3 [163, 610] 217 [75.8, 588] 386 [107, 1146] Mono-iso-butyl phthalate MiBP 17.5 [28.0, 44.1] 13.1 [22.8, 38.1] 55.3 [36.3, 91.4] 48.0 [29.4, 78.3] Mono-n-butyl phthalate MnBP 14.7 [22.7, 35.2] 11.2 [18.6, 28.4] 45.4 [31.4, 66.5] 38.3 [25.3, 62.5] Mono-(2-ethylhexyl) phthalate MEHP 1.44 [2.77, 4.44] 0.98 [1.6, 2.87] 5.44 [3.45, 8.98] 3.53 [1.95, 6.42] Mono-n-hexyl phthalate MnHP 0.10 [0.17, 0.21] <LOD [0.15, 0.2] 0.32 [0.19, 0.48] 0.29 [<LOD, 0.44] Mono-(2-ethyl-5-hydr oxyhexyl) phthalate MEHHP 8.11 [13.3, 21.1] 5.85 [8.6, 13.8] 26.1 [18.4, 40.8] 20.0 [11.6, 27.6] Mono-(2-ethyl-5-oxohexyl) phthalate MEOHP 5.74 [8.67, 14.7] 3.84 [5.65, 9.06] 17.7 [12.4, 27.0] 13.0 [7.78, 19.3] Mono-(2-ethyl-5-carboxypentyl) phthalate MECPP 8.12 [13.9, 22.9] 6.49 [9.49, 15.8] 27.8 [18.9, 43.7] 21.1 [12.6, 32.7] Mono-benzyl phthalate MBzP 4.91 [9.11, 16.5] 3.57 [6.54, 12.7] 19.9 [10.3, 35.9] 13.5 [7.27, 29.7] Abbr
Supplementary table 2a. Concentrations of Endocrine Disrupting Chemicals per diet-group at baseline (median [interquartile range]).
HPLC (n=54) HPNC (n=53) LPLC (n=54) LPNC (n=57) p-value MeP 42.0 [8.43, 134] 45.9 [19.9, 160] 74.1 [17.1, 160] 50.2 [11.8, 138] 1.000 EtP 1.93 [0.62, 12.1] 1.79 [0.53, 20.7] 2.43 [0.79, 10.0] 3.21 [1.04, 12.2] 0.897 PrP 7.64 [1.66, 29.9] 11.2 [0.90, 33.2] 8.17 [2.18, 32.4] 12.2 [1.93, 33.1] 0.838 BPA 2.93 [1.98, 5.67] 3.51 [2.00, 6.05] 3.10 [1.82, 6.14] 3.77 [2.45, 5.76] 0.654 MEP 236 [92.1, 808] 264 [67.7, 705] 205 [68.8, 496.9] 201 [78.4, 489] 0.767 MiBP 48.8 [33.8, 89.3] 58.0 [35.0, 83.6] 60.2 [45.6, 108] 49.5 [36.1, 92.7] 0.126 MnBP 41.6 [29.5, 60.1] 46.2 [30.8, 66.8] 50.0 [35.2, 79.3] 53.8 [35.0, 66.3] 0.677 MEHP 5.35 [2.87, 8.51] 5.68 [3.45, 9.93] 4.99 [3.10, 8.93] 5.24 [3.80, 9.01] 0.839 MEHHP 22.5 [16.5, 42.7] 30.0 [19.7, 44.1] 25.6 [19.0, 37.6] 25.7 [18.0, 40.8] 0.507 MEOHP 16.6 [10.8, 26.7] 18.4 [11.2, 31.1] 18.1 [13.1, 24.8] 16.6 [12.4, 27.8 0.625 MECPP 27.4 [17.7, 44.5] 30.0 [20.1, 50.4] 29.4 [19.2, 39.2] 26.2 [19.0, 42.2] 0.736 MBzP 16.9 [10.7, 36.3] 22.3 [10.3, 34.7] 20.8 [9.99, 35.4] 18.6 [10.7, 35.1] 0.598
Reported excretions are imputed and volume adjusted. Excretions between dietary intervention groups were compared using ANCOVA, adjusting for age, sex and diabetes status. For ancova, excretions were transformed by their natural logarithm. Abbreviations: n, number; HPLC, high-protein low-carbohydrate; HPNC, high-protein normal-carbohydrate; LPLC, low-protein low-carbohydrate; LPNC, low-protein normal carbohydrate. For full names of compounds, see supplementary table 1.
Supplementary table 2b. Concentrations of Endocrine Disrupting Chemicals per diet-group at 3-month follow-up. HPLC (n=54) HPNC (n=53) LPLC (n=54) LPNC (n=57) p-value MeP 58.7 [13.1, 136] 72.5 [10.3, 258] 81.7 [23.2, 173] 73.6 [16.1, 142] 0.786 EtP 4.11 [0.57, 25.9] 3.95 [0.71, 23.7] 5.21 [0.80, 22.3] 3.32 [0.74, 15.0] 0.853 PrP 7.59 [1.38, 33.3] 11.2 [1.34, 51.3] 13.2 [2.58, 26.0] 8.08 [1.81, 36.2] 0.852 BPA 5.27 [2.66, 8.86] 3.65 [2.17, 6.32] 3.17 [2.13, 6.86] 3.39 [1.37, 8.33] 0.437 MEP 405 [156, 1188] 335 [94.1, 1481] 435 [134, 1072] 464 [128, 1319] 0.969 MiBP 47.5 [36.8, 78.6] 45.3 [33.7, 68.0] 56.1 [33.1, 89.0] 48.4 [32.6, 74.8] 0.548 MnBP 43.7 [28.2, 58.2] 43.4 [28.1, 58.9] 45.4 [30.0, 71.2] 39.8 [27.8, 65.8] 0.699 MEHP 3.78 [2.34, 6.10] 4.17 [2.33, 6.11] 3.10 [2.10, 7.03] 4.22 [2.42, 7.72] 0.299 MEHHP 18.9 [13.9, 33.0] 21.3 [14.2, 33.6] 18.0 [13.6, 30.6] 20.0 [13.5, 28.2] 0.658 MEOHP 13.8 [8.73, 20.5] 14.4 [9.87, 19.3] 12.2 [8.90, 19.7] 14.4 [8.95, 20.3] 0.864 MECPP 23.4 [16.3, 34.4] 23.7 [16.1, 36.4] 19.2 [15.3, 31.8] 23.7 [14.1, 32.7] 0.789 MBzP 18.5 [10.6, 33.5] 12.6 [8.07, 26.5] 13.7 [8.03, 29.4] 13.4 [8.00, 27.5] 0.471
Reported excretions are imputed and volume adjusted. Excretions between dietary intervention groups were compared using ANCOVA, adjusting for age, sex and diabetes status. For ancova, excretions were transformed by their natural logarithm. Abbreviations: n, number; HPLC, high-protein low-carbohydrate; HPNC, high-protein normal-carbohydrate; LPLC, low-protein low-carbohydrate; LPNC, low-protein normal carbohydrate. For full names of compounds, see supplementary table 1.
5
Su p p lement ar y t ab le 3 . A ss o ci at io ns b et w ee n c ha ng e i n a d ip o si ty -r el at ed t ra it s w it hi n a nd b et w ee n i nd iv id ua ls a nd c ha ng e i n E D C e xc re tio ns d ur in g a d ie t-in d uc ed w ei g ht l o ss p ro g ra m . Body-mass index (kg/m 2) W aist cir cumfer ence (cm)Body fat per
centage (%) Estimate (CI) p -value q -value Estimate (CI) p -value FDR Estimate (CI) p -value FDR MeP -0.01 [-0.08; 0.06] 0.7695 0.9009 -0.01 [-0.09; 0.06] 0.7243 0.8604 0.00 [-0.09; 0.10] 0.9606 0.9606 EtP -0.05 [-0.14; 0.05] 0.3679 0.7359 -0.05 [-0.16; 0.05] 0.3249 0.6499 -0.05 [-0.18; 0.08] 0.4312 0.6551 PrP -0.01 [-0.10; 0.09] 0.9009 0.9009 -0.01 [-0.11; 0.09] 0.8604 0.8604 0.04 [-0.08; 0.17] 0.4913 0.6551 B PA 0.04 [-0.03; 0.10] 0.2819 0.7359 0.06 [-0.02; 0.13] 0.1360 0.5438 0.07 [-0.02; 0.16] 0.1148 0.4592 MEP 0.04 [-0.05; 0.12] 0.3750 0.6841 -0.05 [-0.14; 0.04] 0.2636 0.7092 0.07 [-0.04; 0.18] 0.2262 0.5343 MiBP 0.03 [-0.01; 0.06] 0.1165 0.4396 0.02 [-0.02; 0.06] 0.2649 0.7092 0.03 [-0.01; 0.08] 0.1691 0.5343 MnBP 0.02 [-0.01; 0.06] 0.1648 0.4396 0.01 [-0.03; 0.05] 0.6260 0.9435 0.04 [-0.01; 0.08] 0.1289 0.5343 MEHP 0.01 [-0.02; 0.05] 0.4612 0.6841 0.02 [-0.02; 0.07] 0.2659 0.7092 0.02 [-0.04; 0.07] 0.5668 0.6477 MEHHP 0.01 [-0.02; 0.04] 0.6841 0.6841 0.00 [-0.03; 0.04] 0.8705 0.9435 0.02 [-0.03; 0.06] 0.4975 0.6477 MEOHP 0.01 [-0.03; 0.04] 0.6127 0.6841 0.00 [-0.04; 0.04] 0.9435 0.9435 0.01 [-0.04; 0.06] 0.7198 0.7198 MECPP 0.01 [-0.02; 0.04] 0.5776 0.6841 0.00 [-0.03; 0.04] 0.8347 0.9435 0.01 [-0.03; 0.06] 0.5349 0.6477 MBzP 0.03 [-0.01; 0.08] 0.1298 0.4396 0.01 [-0.03; 0.06] 0.5760 0.9435 0.03 [-0.03; 0.10] 0.2672 0.5343 For mula: “ lmer(lnEDC ~ adiposity-r
elated trait * time + (age + sex + diabetes + diet) * time + (1 | ID), data)
”. Linear mixed ef
fect models with fixed ef
fects for age, sex,
diabetes status and diet, and random ef fects for inter cept and time. An interaction ter m between adiposity-r elated trait and time was intr oduced in the models. Beta coef ficients r epr esent ef fects per e-fold incr
ease in EDC parameter per month. For full names of EDCs, see supplementary table 1.
p-values wer
e adjusted for multiple
testing using Benjamini and Hochber
g False Discovery Rate (FDR) for all individual metabolites of r
espective gr
oups (i.e. phenols, phthalates), r
esulting in
q
Su p p lement ar y t ab le 4 . A ss o ci at io ns b et w ee n b as el in e E D C e xc re tio ns a nd a d ip o si ty -r el at ed t ra it s a ft er a d ie t-in d uc ed w ei g ht l o ss p ro g ra m . Body-mass index (kg/m 2) W aist cir cumfer ence (cm)
Body fat per
centage (%) Estimate [CI] p-value q-value Estimate [CI] p-value q-value Estimate [CI] p-value FDR MeP 0.11 (0.02; 0.21) 0.0234 0.0469 0.26 (-0.05; 0.56) 0.1039 0.2078 -0.11 (-0.39; 0.17) 0.4445 0.8890 EtP 0.04 (-0.04; 0.13) 0.3214 0.4285 0.13 (-0.14; 0.40) 0.3341 0.4455 -0.03 (-0.27; 0.21) 0.7992 0.8903 PrP 0.10 (0.02; 0.18) 0.0188 0.0469 0.29 (0.03; 0.54) 0.0280 0.1121 0.02 (-0.21; 0.25) 0.8903 0.8903 B PA -0.02 (-0.18; 0.14) 0.7757 0.7757 0.06 (-0.43; 0.56) 0.8045 0.8045 0.19 (-0.26; 0.64) 0.4049 0.8890 MEP 0.07 (-0.04; 0.18) 0.2115 0.3384 0.25 (-0.09; 0.59) 0.1531 0.3062 0.32 (0.01; 0.62) 0.0438 0.1752 MiBP 0.04 (-0.15; 0.23) 0.6555 0.6555 0.37 (-0.23; 0.97) 0.2312 0.3220 0.28 (-0.26; 0.83) 0.3124 0.3982 MnBP 0.15 (-0.05; 0.36) 0.1456 0.3310 0.61 (0.06; 1.16) 0.0315 0.1258 0.43 (-0.08; 0.93) 0.0994 0.2651 MEHP 0.07 (-0.12; 0.26) 0.4505 0.5148 0.00 (-0.58; 0.58) 0.9995 0.9995 -0.25 (-0.77; 0.27) 0.3484 0.3982 MEHHP 0.21 (-0.02; 0.45) 0.0793 0.3310 0.58 (-0.14; 1.30) 0.1179 0.3062 0.39 (-0.26; 1.05) 0.2421 0.3874 MEOHP 0.11 (-0.10; 0.32) 0.3056 0.4074 0.38 (-0.26; 1.02) 0.2415 0.3220 0.23 (-0.35; 0.81) 0.4411 0.4411 MECPP 0.19 (-0.05; 0.44) 0.1254 0.3310 0.36 (-0.37; 1.09) 0.3377 0.3859 0.50 (-0.16; 1.17) 0.1399 0.2799 MBzP 0.11 (-0.05; 0.27) 0.1655 0.3310 0.68 (0.21; 1.16) 0.0053 0.0424 0.62 (0.19; 1.05) 0.0053 0.0420 Regr ession coef
ficients with 95% Confidence Intervals derived fr
om or
dinary least squar
es linear r
egr
ession. For
mula: “a
diposity-r
elated trait (follow-up) ~ lnEDC
(baseline) + age + sex + diabetes + diet + adiposity-r
elated trait (baseline)”
. EDC levels wer
e imputed, volume-adjusted, and natural log-transfor
med. Models wer
e
corr
ected for age, sex, diabetes status, diet and r
espective trait at baseline. Regr
ession coef ficients r epr esent ef fects per e-fold incr
ease in EDC parameter over 3
months. Positive r
egr
ession coef
ficients indicate impair
ment of r
eduction of
the
respective trait. For full names of EDCs, see supplementary table 1.
p-values wer
e
adjusted for multiple testing using Benjamini and Hochber
g False Discovery Rate (FDR) for all individual metabolites of r
espective gr
oups (i.e. phenols, phthalates),
resulting in
q
-values. Abbr
5
Supplementary table 5. Sensitivity analysis of associations between baseline EDC excretions and adiposity-related traits after a diet-induced weight loss program, excluding all individuals with type 2 diabetes at baseline (n = 19).
Body-mass index (kg/m2) Waist circumference (cm) Body fat percentage (%)
Estimate [CI] p-value Estimate [CI] p-value Estimate [CI] p-value
MeP 0.13 [0.02; 0.23] 0.0170 0.23 [-0.10; 0.56] 0.1804 -0.11 [-0.37; 0.16] 0.4259 EtP 0.04 [-0.05; 0.13] 0.3509 0.13 [-0.16; 0.41] 0.3835 -0.04 [-0.27; 0.18] 0.7082 PrP 0.11 [0.02; 0.19] 0.0123 0.28 [0.01; 0.55] 0.0420 0.05 [-0.17; 0.27] 0.6450 BPA -0.06 [-0.23; 0.11] 0.4947 -0.04 [-0.57; 0.50] 0.8897 0.09 [-0.34; 0.52] 0.6729 MEP 0.07 [-0.05; 0.19] 0.2412 0.24 [-0.12; 0.60] 0.1891 0.16 [-0.13; 0.45] 0.2896* MiBP 0.01 [-0.20; 0.21] 0.9567 0.31 [-0.33; 0.96] 0.3433 0.18 [-0.34; 0.70] 0.4937 MnBP 0.13 [-0.08; 0.34] 0.2107 0.59 [0.02; 1.16] 0.0425 0.44 [-0.02; 0.90] 0.0633 MEHP 0.06 [-0.14; 0.26] 0.5628 -0.01 [-0.62; 0.60] 0.9677 -0.14 [-0.63; 0.34] 0.5606 MEHHP 0.18 [-0.06; 0.43] 0.1473 0.59 [-0.16; 1.34] 0.1267 0.31 [-0.29; 0.92] 0.3124 MEOHP 0.08 [-0.13; 0.30] 0.4555 0.41 [-0.26; 1.07] 0.2303 0.17 [-0.37; 0.71] 0.5437 MECPP 0.15 [-0.10; 0.41] 0.2495 0.38 [-0.39; 1.14] 0.3390 0.38 [-0.24; 1.00] 0.2314 MBzP 0.08 [-0.08; 0.25] 0.3230 0.71 [0.20; 1.23] 0.0071 0.53 [0.12; 0.95] 0.0129
Regression coefficients with 95% Confidence Intervals derived from ordinary least squares linear regression. Formula: “related trait (follow-up) ~ lnEDC (baseline) + age + sex + diabetes + diet +
adiposity-related trait (baseline)”. EDC levels were imputed, volume-adjusted, and natural log-transformed. Models
were corrected for age, sex, diabetes status, diet and respective trait at baseline. Regression coefficients represent effects per e-fold increase in EDC parameter over 3 months. Positive regression coefficients indicate impairment of reduction of the respective trait. For full names of EDCs, see supplementary table 1.
p-values were adjusted for multiple testing using Benjamini and Hochberg False Discovery Rate (FDR) for all
individual metabolites of respective groups (i.e. phenols, phthalates), resulting in q-values. Abbreviations: CI, confidence interval. *The association between MEP and Body Fat Percentage lost significance after the exclusion of type 2 diabetic individuals.
