Prenatal phthalate exposure reduction through an integrated intervention strategy

Pregnancy represents a sensitive susceptibility window to phthalate esters (PAEs). In this study, we develop an intervention strategy for reducing the exposure of pregnant women to phthalates. Thirty-five pregnant women, who initially underwent maternity examination, were recruited from an ongoing longitudinal prospective prenatal cohort study. The intervention strategy integrates diet, lifestyle, and environmental factors. Participants were encouraged to modify their behaviors and habits according to the intervention strategy at three different periods. Urine samples were collected from the participants after antenatal examination every month, for 8 months, to measure ten PAE metabolites. Mono-(2-ethyl-5-hydroxyhexyl) phthalate (MEHHP), mono-n-butyl phthalate (MnBP), and mono-(2-ethyl-5-oxohexyl) phthalate (MEOHP) declined significantly after the 1st intervention, while mono-isobutyl phthalate (MiBP) and mono-methyl phthalate (MMP) noticeably decreased after the 2nd intervention. The sum of the molar concentrations of MEHP, MEHHP, MEOHP, and MECPP reduced by 20 to 40% during subsequent intervention. In addition, the sum of the molar concentrations of MEP, MnBP, MMP, and MiBP as well as the sum of the molar concentrations of the ten metabolites also reduced. Our findings suggest that intervention through written recommendations can effectively reduce the burden of phthalates during pregnancy.


Introduction
P h t h a l a t e e s t e r s ( P A E s ) , t h e d i e s t e r s o f 1 , 2benzenedicarboxylic acid, represent a class of synthetic chemicals widely employed in manufacturing processes (Wang et al. 2016). In general, high molecular weight phthalates (HMWPs; ≥250 Da), such as di-2-ethylhexyl phthalate (DEHP), benzyl butyl phthalate (BzBP), butyl benzyl phthalate (BBP), and di-n-octyl phthalate (DnOP), are primarily utilized as softeners in the production of polyvinyl chloride (PVC). Conversely, low molecular weight phthalates (LMWPs), such as dimethyl phthalate (DMP), diethyl phthalate (DEP), and di-n-butyl phthalate (DBP), are common components of personal care products and pharmaceuticals (Koch et al. 2013).
Considering that phthalate plasticizers are not chemically bound to PVC, they can be released into the environment by leaching, evaporation, and abrasion or through the application of phthalate-containing personal care products (Johns et al. 2015). Therefore, phthalate can be detected not only in consumer products but also in water, soil, sediment, food, and indoor air (Wu et al. 2018a). As a result, humans are exposed to these chemicals through ingestion, inhalation, and dermal exposure during their lifetime, including intrauterine development (Radke et al. 2019). Several studies from around the world reported detectable levels of phthalate metabolites in urine from pregnant women and further associated these metabolites with impaired neurodevelopment, altered genital development, and respiratory problems in infants (Ferguson et al. 2015;Minatoya et al. 2017;Qian et al. 2019). Thus, preventing exposure to phthalates in daily life is important, especially for pregnant women, who are more susceptible to endocrine-disrupting chemicals.
Intervention studies for reducing exposure to PAEs are available. Rudel et al. indicated that limiting the consumption of food prepared or stored in plastic containers reduced phthalates in urine (Rudel et al. 2011). In addition, Sathyanarayana et al. demonstrated that bisphenol A and DEHP can be reduced through dietary intervention . Harley et al. suggested that limited use of personal care products can reduce the urine concentrations of MnBP and MiBP in adolescent girls (Harley et al. 2016). However, these studies were either simple (focused on diet), short-term (<2 weeks), or involved few participants (4 families or less). Moreover, the participants in these studies were either children or adults, and thus, the results may inadequately reflect pregnancy. Therefore, a study aimed at reducing the exposure of pregnant women to PAEs is urgently needed.
In this study, we developed an intervention strategy to reduce the exposure of pregnant women to phthalates. Considering that eliminating these chemicals is not feasible (Barrett et al. 2015), the interventions were executed through written recommendations. The concentrations of phthalate metabolites in the participants' urine samples were measured monthly during pregnancy, and the participants received written recommendations in the 1st, 2nd, and 3rd trimesters. We hypothesized that intervention through written recommendations effectively reduces the exposure of pregnant women to PAEs.

Study population
The study population (n = 35) of this intervention is part of a cohort study performed in the city of Ezhou. This ongoing longitudinal birth cohort study enrolled 676 women from November 2018 to November 2019, of whom 568 were followed until delivery to live singletons. Pregnant women were invited to participate in the study during their first antenatal examinations (<8 weeks of gestation) at the Ezhou Medical & Healthcare Center for Women and Children in November 2018. The women were eligible for enrollment if the following criteria were satisfied: (1) resident in Ezhou city, (2) willing to adhere to the intervention procedure, (3) ready to participate in monthly antenatal examinations and donate urine samples, and (4) planned to deliver in the Healthcare Center.
This study was approved by the Ethics Committee of Hubei University of Chinese Medicine. Written informed consent was obtained from all participants.

Intervention strategy
Considering previous studies, we developed a strategy incorporating three components presented in Table 1. Owing to diet being a major exposure medium to HMWPs (Pacyga et al. 2019), the participants were reminded to restrict consumption of pufferfish, as well as canned and microwave food, and healthy eating was encouraged (e.g., organic food, folic acid supplements, and vegetarian diets). Hsieh and his colleagues poted that phthalate can stay on the skin for a long duration (Hsieh et al. 2019). Therefore, the intervention strategy also considered lifestyle, such as avoiding touching decorated materials and plastic floors, restricting use of personal care products (e.g., body lotion, cosmetics, perfumes, and hair sprays), and limiting storage of prepared food in plastic containers including bags and cans. Owing to the ubiquity of phthalate contaminants in indoor dust and air inside vehicles (Zhang et al. 2014), the participants were encouraged to avoid second-hand smoke, minimize transportation by cars, and exercise adequately.
The present study lasted until the ninth month of pregnancy. The participants were advised to alter their behaviors according to the intervention strategy in the 1st (first antenatal examination), 2nd (between the 4th and 5th months of pregnancy), and 3rd (between the 7th and 8th months of pregnancy) trimesters. During the first visit, a counselor discussed the negative effects of PAEs on health with the participants and asked them to follow written recommendations developed by our team. In subsequent interventions, the women provided feedback from self-monitoring records, and they were encouraged to continue following the intervention strategy and avoid PAE-containing products. A flowchart highlighting the components of the study is shown in Fig. 1.
The metabolites were analyzed in the School of Laboratory Medicine, Hubei University of Chinese Medicine by liquid chromatography-mass spectrometry (LC-MS/MS; Agilent, USA), according to the method for urine phthalates measurement described in Specht et al. (2015). This method was also validated in our previous study (Wu et al. 2018b). The stored urine samples were thawed, and 200 μL of each sample was vortexed, sonicated for 5 min, and buffered using ammonium acetate. The samples were then spiked with 40 μL of a labeled isotope mixture (500 ng/mL), followed by addition of 5 μL of β-glucuronidase to eliminate glucuronic acid. After incubation, the urine samples were diluted using 1 mL of a phosphate buffer (0.14 M NaH 2 PO 4 in 0.85% phosphoric acid) before loading onto a solid-phase extraction cartridge. After sequential equilibration using 1 mL of acetonitrile (ACN), 1 mL of H 2 O, and 1 mL of the phosphate buffer, the solutions were transferred to glass vials before analyses by LC-MS/MS. Good separation was achieved for all analyses, with retention times on the column varying between 6.68 and 27.20 min. The calibration curve covered the range 0.100-200 ng/mL, and each batch of samples analyzed involved blanks and quality control (QC) samples. The intra-and inter-day relative standard deviations (RSD) were below 11.7% and 13.2%, respectively.
The metabolite concentrations were corrected for the urine dilution using specific gravity (SG) as follows: where Pc is the SG-adjusted urine concentration (ng/mL), P is the measured metabolite concentration, SG is the specific gravity of the urine sample, and SG M is the median SG of the samples for the studied population (Upson et al. 2013). The SG of each sample was measured using a handheld refractometer (PAL10-S; Atago, Tokyo, Japan) at room temperature. For PAE concentrations below the limit of detection (LOD), a value equal to LOD/ √ 2 was imputed.

Urine concentration change
The urine concentration change (%) was calculated using the following equation: Urine concentration change = [(urine phthalate concentration during a visit − baseline urine phthalate concentration)/(baseline urine phthalate concentration)] × 100.

Statistical analyses
Descriptive statistics were employed to characterize the study population with values expressed as a percentage or using the mean ± standard deviation (SD). The detection frequency, geometric mean (GM), and median SG-adjusted concentrations of urine phthalate metabolites for various antenatal examinations were calculated and used to characterize exposure. The Mann-Whitney U-test was utilized for comparing the significance of the urine concentration differences for each visit and the baseline. The Spearman rank correlation test was employed for assessment of correlations among the urine phthalate metabolite levels, with the significance set at p < 0.05. All statistical analyses were performed using R (version 3.5.3; http://www.r-project.org).

Intervention strategy
The intervention strategy used for the participants involved three periods of written recommendations. This involved a first intervention representing the baseline during the first antenatal examination, a second intervention during the fourth month of pregnancy (2nd trimester), and a third intervention in the seventh month (3rd trimester). At least 77% (27/35) of the participants underwent antenatal examination each month. Finally, 241 urine samples were collected from the baseline to the endpoint.

Distribution and variability of phthalate metabolites
All metabolites, except MOP (<40%), were detected in > 85% of the samples. The distributions of nine phthalate metabolites (excluding MOP), involving three groups, for each visit are presented in Fig. 2

Urinary concentration change
The urine concentration changes associated with the participants of this study are presented in Fig. 3. Most of the phthalate metabolites declined after each visit, except for MBzP and MEHP. Compared with the baseline concentration, MEHHP, MnBP, and MEOHP decreased by more than 20% after the first intervention. Relatedly, MiBP and MMP reduced by almost 30% in the fifth month, after the second intervention. In fact, the ΣLPAEs, ΣDEHP, and Σ 10 PAEs decreased by 20% after the first intervention and reduced by 40% after the final visit.

Discussion
In this study, we developed an intervention strategy comprising diet and lifestyle habits as well as environmental components. This intervention strategy involved repeated and voluntary self-restraining components aimed at reducing urine phthalate concentrations during pregnancy. The decline in the urinary concentrations of the 10 phthalate metabolites measured implied that our intervention approach efficiently reduced the exposure of the participants to PAEs.
Diet is the primary route for intake of phthalates. Studies conducted in Canada (Pacyga et al. 2019) and the USA (Koch et al. 2013) showed detectable concentrations of phthalates in fast foods such as french fries, hamburgers, and sandwiches. However, other epidemiologic studies indicated that the consumption of more fresh vegetables, fruits, nuts, and fish is associated with low exposure to phthalates (Bai et al. 2015;Romano et al. 2019). Therefore, our diet in the intervention strategy restricted participants from eating fast foods and encouraged them to consume healthy foods. Our results demonstrated that the concentrations of MnBP, MEHHP, and MEOHP significantly declined after the 1st intervention, with the decrease reaching 40% after the 3rd intervention. These findings are consistent with those of Rudel et al. (2011) and Sathyanarayana et al. (2013) studies, in which phthalate metabolites were significantly lowered after the diet of participants was restricted to foods involving limited packaging. However, these studies involved complete diet replacement.
In contrast, in our study, the participants were advised to pursue a self-guided intervention. Therefore, although our study did not involve a complete replacement intervention, the repeated reminders through written recommendations appeared to exhibit similar effectiveness. According to an intervention study by Barrett et al. (2015), the MMP and MiBP concentrations showed no appreciable change across three time periods. In fact, after a 3-day dietary intervention, the mean concentrations of MiBP were 15.9, 24.0, and 23.7 ng/mL for the pre, mid, and post-intervention periods, respectively. The corresponding MMP concentrations for the three periods were 19.4, 24.6, and 27.5 ng/mL. However, a study involving a school for girls in Taiwan reported reduction in MiBP, MEP, and MMP through an intervention strategy . A possible reason for the inconsistency of results from different studies is that most interventions involve a simple dietary change, whereas the underlying intervention strategy requires avoiding foods in plastic containers and using less personal care products. Many studies have also demonstrated that besides the ingesting of contaminated foods, dermal absorption from personal care products is another route for human intake of phthalates (Valvi et al. 2015;Wenzel et al. 2018). In the present study, after the pregnant women decreased their usage of hair dye, shampoo, perfumes, body lotions, and nail polish, the concentrations of MiBP and MMP significantly declined after the 2nd intervention, reaching 50% after the 3rd intervention. These results suggest that some LMWPs could be reduced through our intervention strategy.
However, after the 3rd intervention, MBzP showed no statistically significant changes, a result which may be related to its low concentration (<10 ng/mL) from the baseline to the end. We also observed that MEHP changed sharply, reducing by approximately 30% in the fifth month of antenatal examination and declining by 60% in the seventh antenatal month. These results are similar to those of a 2-week randomized dietary trial in the USA, with MEHP geometric means of 3.9, 4.1, and 4.2 ng/mL for the baseline, mid-intervention, and post-intervention periods ), This response is possibly because of the metabolite characteristics of DEHP. The DEHP can initially be metabolized to MEHP, followed by MEHP metabolizing to the secondary mono-phthalate esters including MEHHP, MECPP, and MEOHP (Minatoya et al. 2017). Notably, in our study, the ΣDEHP, ΣLMWP, and Σ 10 PAEs significantly declined after the 1st and subsequent interventions. This is consistent with a previous study that reported that through a dietary intervention, the GM concentrations of BPA reduced by 66%, while those of DEHP metabolites decreased by 53-56% (Ackerman et al. 2014).
The environment is another factor associated with PAE exposure. According to previous studies, MEOHP and MEHPP were higher among women who do smoke compared to non-smokers (Cantonwine et al. 2014). In the present study, after the 1st intervention, MEOHP and MEHPP declined significantly, and this may be partially attributed to the avoidance of second-hand smoke (Wang et al. 2020). We also encouraged the participants to undertake limited transportation by car. Owing to the elevated internal temperature of the cabin of a car or truck, the vinyl interior trim of some vehicles can deteriorate and decompose, thereby releasing phthalate particles (Rakkestad et al. 2007).
In addition, we recommended adequate exercise, such as walking and yoga, to the pregnant women. According to a study in Australia, slightly higher total phthalate metabolite concentrations were associated with insufficient activity (Bai et al. 2015). Further investigation is warranted to better understand the link between physical activity and low phthalate ester levels biologically.
In this study, an intervention strategy for reducing the PAE concentrations in pregnant women is introduced. This strategy comprises eight components associated with diet, lifestyle, and the environment. This approach closely parallels real life compared to studies focusing on intervention involving just one component. Moreover, repeated intervention and measures are better for evaluating the efficiency of the intervention strategy. However, the present study involves several limitations that require attention. First, the intervention components were not distinguished to clarify their impact on the phthalate reduction efficiency. Considering that some intervention components are clustered, determining the contribution of each component to the phthalate concentrations reduction was not achieved. Second, some findings may be biased because of the relatively small sample size. However, the longitudinal sampling involved in the present study allowed individuals to serve as their controls, thereby avoiding the many sources of confounding information that limit cross-sectional studies. Finally, because the participants in this study are from an economical city in Hubei Province, China, generalizing the study efficiency to other countries and socioeconomic groups may be limited.

Conclusion
The present study demonstrated that intervention through written recommendations can effectively reduce the burden of phthalates on the body. These findings suggest that our intervention strategy is valuable for establishing specific PAE-limiting factors for improving pregnancy and fetal outcomes.