Review of the Existing Evidence for Sex-Specific Relationships between Prenatal Phthalate Exposure and Children’s Neurodevelopment

Phthalates are well-known, ubiquitous environmental contaminants influencing children’s health and their neurodevelopment. However, results of the previously conducted studies are not entirely conclusive. The aim of this review is to present the current state of knowledge with respect to the association between the prenatal phthalate exposure and sex-specific child neurodevelopmental outcomes. A systematic search of the literature was carried out to identify the studies that analyse the sex-specific association between prenatal exposure to phthalates and cognitive, psychomotor outcomes and behavioural and emotional problems. The search was conducted in May 2021, and it was limited to the papers published in English between January 2015 and April 2021. The following databases were used: PubMed, Scopus and Elsevier. The selection process was carried out by two independent authors according to the inclusion criteria. Of a total of 7542 records, 17 epidemiological studies met the inclusion criteria with regards to phthalate exposure and sex-specific differences in child neurobehavioural development. The review shows no clear pattern of association between maternal exposure to phthalates during pregnancy and offspring neurodevelopment. No clearly pronounced sex specific effects, except for BBzP exposure and decreased motor ablates among girls, have been indicated. Inconsistences in the results, as well as unsolved issues related to the interpretation of the results in the context of the exposure level, outcomes, confounders, and biological plausibility highlight the necessity for further research in the field.


Introduction
Child cognitive, psychomotor and behavioural development is determined by genetic as well as environmental and lifestyle-related factors, and by gen-environment interactions. Among the variety of environmental contaminants, phthalates are one the most frequently investigated ones. Phthalates are endocrine disrupting chemicals widely used as solvents, plasticizers, stabilizers, and additives in a number of products, including personal care products, food packaging, toys, building, constructions materials, and in medications [1]. General population is exposed to these chemicals via various exposure sources and routes with ingestion being the primary one [2,3]. Phthalates are rapidly metabolized in the body (with half-life of 3-18 h) and excreted in urine, which is a common, reliable, and non-invasive matrix for assessing phthalate exposure [4][5][6].
The existing up-to-date studies postulate that exposure to phthalates poses a health hazard with regard to reproductive function, non-communicable diseases including respiratory diseases, allergies, cardiometabolic diseases and neurodevelopmental disorders [7][8][9].
Furthermore, researchers stress that children experience higher exposure levels to most phthalates and consequently, present more pronounced health problems related to exposure in comparison to adults. This is due to their sensitive periods of exposure including prenatal and early childhood, their development, immaturity, diet, metabolism, and age-specific behaviours [10][11][12][13][14][15][16][17]. Phthalates cross the blood-placenta barrier [18] and thus, adversely affect the developing brain. Moreover, a continuous exposure in infancy and early childhood may contribute to long-term health consequences. The vast number of studies in this field suggest that exposure during these sensitive periods might be associated with behavioural, cognitive, and psychomotor consequences in children [14,[19][20][21][22][23][24][25].
Results of the existing studies are not fully conclusive and consistent because of methodological discrepancies such as: study design, different exposure period (prenatal/infancy/childhood), phthalates and their metabolites considered, exposure level, use of a variety of tests for measuring neurodevelopmental outcomes and lack of appropriate control for confounding variables [14,[19][20][21][22][23][24]26,27]. Additionally, some authors highlighted the importance of altered sex-specific differences in infant and child neurodevelopment [19,24,28]. There are several reviews summarizing the existing evidence on this topic [19,24,25,[28][29][30]. Ejaredar et al., (2015), in a systematic review of the literature including studies published until year 2014, reported that boys were more prone to adverse behavioural and cognitive effects [19]. The second review covering 14 birth cohorts (papers published by May 2018) has pointed out that exposure to some phthalates had adverse effects on neurobehavioural development in children, with an increased likelihood of behavioural problems in boys [24]. Radke et al. (2020) based on the systematic review and meta-analysis (covering papers published by March 2019) concluded that overall, there is no clear pattern of association between prenatal phthalate exposure and child neurodevelopment, including cognition, motor effect and behaviour. Only motor effect of BBP in girls has been indicated with moderate evidence [30].
This review builds on the most recent studies regarding the association between prenatal phthalate exposure and children's neurodevelopment with taking their sex into consideration. This review covers a notable number of new publications on this topic since the last review, which was published by Radke at al. (2020) [30]. Moreover, the number of evaluated phthalates/phthalate metabolites as well as neurodevelopmental outcomes is more extensive.

Materials and Methods
The PRISMA (preferred reporting items for systematic reviews and meta-analyses) was employed to guide this review [35].
A search of PubMed, Scopus, and Elsevier databases was performed in order to identify the studies which analyze the association between maternal exposure to phthalates during pregnancy and neurodevelopmental outcomes in children including their behaviour, cognitive and psychomotor development. The search was executed in May 2021. The search strategy was based on the following combinations of keywords: phthalate exposure, child exposure, child neurodevelopment, sex-specific behavior, cognition, psychomotor development.
The search was restricted to articles: (1) covering human studies; (2) published in English in peer reviewed journals since 2015; (3) evaluating phthalate metabolite concentrations (for 8 phthalates: di(2-ethylhexyl) phthalate (DEHP), di-n-octyl phthalate (DnOP), di-iso-nonyl phthalate (DiNP), di-methyl phthalate (DMP), di-ethyl phthalate (DEP), diiso-butyl phthalate (DiBP), di-n-butyl phthalate (DnBP), and butyl-benzyl phthalate (BBzP) (Table S1) measured in maternal urine collected at least once during pregnancy; (4) assessing the following neurodevelopmental outcomes: (a) cognition (including general cognition, in-telligence quotient (IQ), mental development index (MDI), language and memory abilities), (b) psychomotor skills (including summary scale of psychomotor developmental index (PDI) or fine and gross motor subscales) and (c) behaviour (including: externalizing and attention problems such as attention-deficit/hyperactivity disorder, oppositional-defiant disorder, and conduct disorder/problems; internalizing problems such as depression and anxiety), measured by means of accepted and validated tools; (5) presenting sex-specific relationships between prenatal phthalate exposure and children's neurodevelopmental outcomes. The selection process was carried out independently by two authors according to the inclusion criteria. In the first step, the titles and abstracts were analyzed and then, the full text was reviewed. In total, 7542 articles were found, and they were all checked for eligibility. The majority of publications were excluded because of the following reasons: assessment of postnatal child exposure only, phthalate metabolite concentration measurements based on samples other than urine (e.g., breast milk, amniotic fluid, blood) and lack of sex-specific association between prenatal phthalate exposure and child neurodevelopment. Finally, 17 publications met eligibility criteria and have been included in the present review [36][37][38][39][40][41][42][43][44][45][46][47][48][49][50][51][52]. The number of publications (n = 17) does not represent the number of cohorts (n = 11) as in some cases there are several analyses and publications from the same study. From all the papers, the following information was extracted: study design and population (cohort name, study location, years of recruitment, sample size, mothers' and children's ages), timing of assessment of exposure, phthalates/phthalate' metabolites and their concentrations, tools used for neurodevelopment assessment, confounders (Table S2) and the results of analyses. To assess the association between exposures and outcomes, the concept of defining the population, exposure, comparator, and outcomes (PECO) as described by Radke et al. (2020) was considered [30]. For the study evaluation we followed the methodology developed by Radke et al. (2020) (we have kept the already assigned assessment for the same studies and completed the assessment for the studies not included in the review by Radke at al., 2020) ( Table S3). The details of the methodology have been provided elsewhere [30]. Briefly, the evaluation was performed for exposure, outcome, selection of the participants, confounding factors and analysis. For each domain it was possible to obtain a rating as: good, adequate, deficient, and critically deficient. Finally, based on the assessment given for each domain, an overall study classification was conducted with the following possible categories: high, medium, low or uninformative. This assessment was performed based on reviewer expert judgments (high was given to the studies that received good ranking for most domains, medium-to those which received adequate or good ranking across most domains, with the impact of any identified limitations not being judged as severe) [30]. Of 11 studies, 3 were classified as high and 7 as medium. One study was assessed as low ) and was not considered in the further analyses and interpretations [36]. Table 1 presents characteristics of the epidemiological studies that met the inclusion criteria applied in the current review. The data included in the review came from 11 prospective cohorts conducted in America (6 studies), Asia (3 studies) and Europe (2 studies). The pregnant women who participated in the studies were between 16 and 46 years of age. The sample size of the included studies varied between 86 and 518 participants. Concentration of phthalate metabolites was determined in the urine samples according to the inclusion criteria. Table 2 shows concentrations of the phthalate metabolites in maternal urine collected during pregnancy or at delivery. Many of the studies analyzed the samples collected during the third trimester of pregnancy, while the data obtained from the HOME, INMA, CHAMACOS and Chinese cohorts was based on several time point assessments. The assessment of neurodevelopment was carried out with the participation of children aged from 27 weeks to 16 years. All the studies described above were adjusted for potential confounders; however, the number of confounding variables varied among the studies (Table S2). The following covariates were among the most frequently evaluated: child age, child sex, maternal age, socioeconomic status and marital status. Table 3 presents the results of the studies analyzing the association between prenatal phthalate exposure and child cognition and language abilities. Four studies have evaluated the association between exposure to the sum of DEHP metabolites and child cognitive development. Only in the analysis performed by Hyland et al. (2019) sex-specific difference were observed, with a positive association among girls β: 1.6 (0.0, 3.3) and a negative one among boys β: −1.7 (−3.8, 0.3) (p for interaction = 0.01) [37]. Other studies do not show statistically significant and consistent results. When the DEHP metabolites were analyzed separately, there were not any statistically significant results for boys and for girls. Moreover sex-differences in the association were not noted. The impact of prenatal DEHP exposure on child language abilities was analysed in the study by Olsen et al. (2018), scored as of medium confidence [38]. That study indicated lower scores in language development with an increased prenatal phthalate exposure among boys and no associations among girls (with an opposite pattern observed).

Prenatal Exposure to Phthalates and Child Cognition and Language Abilities
Three studies present the association between prenatal BBzP exposure and child cognition; however, none of them indicates statistically significant associations neither in boys nor in girls. Some potentially sex specific associations between BBzP and neurocognitive development among boys and girls was postulated by Doherty et al.  [39,40]. Ipapo et al. (2017) indicated sex specific associations between BBzP and visual recognition memory at the age of 27 weeks (p for interaction 0.01) [41].
There is no apparent consistency between the studies regarding the impact of prenatal DnBP exposure on child cognitive development. There is some suggestion of the effect modification by sex, so that girls may be more susceptible to cognitive effects of DnBP exposure than boys (girls: β: −2.78 (−5.03, −0.54); boys: β: 1.71 (0.08, 3.34); p for interaction 0.02) [39]. In the study by Ipapo et al. (2017) sex specific associations between DnBP and visual recognition memory at the age of 27 weeks was proposed (p for interaction 0.01) [41].
Looking at the studies evaluating the impact of DEP exposure on child cognitive and language development the results are not consistent, with no sex-specific interaction presented.
Doherty et al. (2017) indicated a statistically significant negative association between prenatal MCPP exposure and child MDI among girls and an opposite pattern, although not statistically significant, among boys (p for interaction 0.007) [39]. There was no statistically significant association between prenatal exposure to DiNP and child language abilities in the study by Olsen et al. (2018) [38].
Summing up, the results for prenatal phthalate exposure and child cognition do not represent a significant pattern of association and sex-specific differences. Table 4 presents the results of the studies analyzing the association between prenatal phthalate exposure and child psychomotor development including fine and gross motor abilities. Four cohorts (one high and three medium confidence studies) present the sex specific analyses for the association between prenatal DEHP exposure (considered as the sum or individual metabolites) with one study presenting sex-specific differences of such an association (girls: β: −1.38 (−5.30, 2.54, boys: β: 3.24 (0.70, 5.78); p for interaction 0.05 [40]. However, no clears pattern of the association can be found based on the research.  WISC-IV-at age 7-10, 5 y; ENI-at age 9 y; NEPSYII -at age 12 y;

Prenatal Exposure to Phthalate and Child Psychomotor Development
SRS-2 at age 14                    [39].
For prenatal DiBP exposure and child motor functions among both girls and boys a negative pattern of association, although not significant in all the analyses, except for general and gross functions among girls in Balalian et al. (2019) [42], was observed. No sex differences were noted.
Even though Balalian et al. (2019) [42] observed significant sex differences for the association between MEP exposure and general as well as gross motor functions (p for interaction 0.02 and 0.03 respectively) with the negative impact in girls, the other studies do not present consistent results [42].
For other phthalates and their metabolites, the results were based on single studies so a conclusion about their impact on child motor functions cannot be drown.
All in all, the existing studies indicate stronger negative associations between prenatal phthalate exposure, especially for BBzP, in girls, but additional evidence is needed.

Prenatal Exposure to Phthalate and Child Behavioural and Emotional Problems
Many analyses focus on the evaluation of the association between prenatal phthalate exposure and child behavioural and emotional outcomes, with a variety of dimensions being assessed (Table 5). Among them internalizing and externalizing problems are the ones which are most frequently analysed. Table 5. Sex-specific differences between phthalates exposure and child behavioural and emotional problems.      Seven analyses have focused on sex specific associations between prenatal exposure to DEHP or its specific metabolites and behavioural, and emotional outcomes in offspring. The results were statistically significant or consistent across the studies neither for internalizing nor for externalysing problems, with no sex specific differences noted.
Similarly, for the exposure to other phthalates (including DMP, DEP, DnBP, DiNP) the associations with behavioural and emotional problems were consistent neither in girls nor in boys. For DiBP, England-Mason et al. (2020) (assessed as a medium confidence study) presented a significant positive association with internalizing problems in boys; however, sex interaction was not observed (p for interaction 0.7) [43].
Summarizing, there is lack of consistency between the studies and the lack of sex differences for the impact of prenatal phthalate exposure on child behaviour.
In addition to the results presenting sex stratified analyses for the association between prenatal phthalate exposure and child neurodevelopmental outcomes, some studies only check if the effect estimate modification by sex exists. None of them indicates sex specific interaction [53][54][55][56][57].

Discussion
Data collected in this review shows no clear pattern of association between maternal exposure to phthalates during pregnancy and offspring neurodevelopment, with no clearly pronounced sex specific effects, except for BBzP exposure and decreased motor ablates among girls.
Such inconsistencies in the existing studies are also pointed out in the previously published reviews. A systematic review covering the studies published until 2014 has indicated the association between prenatal exposure to phthalates and adverse cognitive and behavioural outcomes in children, including lower IQ, attention problems and hyperactivity as well as poorer social communication [19]. Radke et al. have underlined that for each phthalate/outcome combination, there was slight or indeterminate evidence of an association, with the exception of motor effects for BBP, for which moderate evidence has been noted [30]. Minatoya and Kishi have concluded that prenatal exposure to phthalates may contribute to neurobehavioural outcomes in children, but the evidence is still limited [29]. Another review, the one by Palanza et al. (2021) underlined that animal and human studies demonstrate the evidence that EDCs exposure during critical periods of development affects sex differences in emotional and cognitive behaviours [28]. It has been also concluded that results are more heterogeneous when social, sexual, and parental behaviours are considered.
Differences in the results between the studies may be due to several reasons. Different phthalates metabolites were analyzed at various time periods, where most of studies collected one prenatal urinary sample during different trimesters of pregnancy. This might have resulted in an imprecise assessment of prenatal phthalate exposure because they have relatively short biological half-lives. Moreover, some phthalates are relatively structurally similar and moderately correlated with one another (e.g., DBP and DIBP), while others differ considerably in terms of structure, sources of exposure as well as represent low correlations (DEP and DEHP) [30]. In addition, concentrations of phthalate metabolites varied between the countries. Additionally, different tools were used to assess children's cognitive, psychomotor, behavioural, and emotional development and different covariates were considered in the performed analyses.
The proposed mechanism linking prenatal phthalate exposure to the adverse effects on children's neurobehavioural development although still not fully established, assumes that phthalates can have an adverse effect on the systems and processes that are crucial for the development of the fetal brain and nervous system [27,58]. Phthalates and their metabolites can interact with sex-hormones, either in the pregnant woman or in the developing fetus [59,60]. The rodent studies have pointed out that the females are more emotionally vulnerable, while males are more cognitively susceptible, specifically in spatial memory [61,62].
It is established that action of certain metabolites of phthalates can result in an abnormal production and functioning of specific thyroid hormones [31,[63][64][65]. Thyroid hormone homeostasis is crucial for fetal and infant brain development and in consequence may impact children's cognitive and motor abilities [27]. Morgenstern et al., (2017) reported inverse and sex-specific associations between some metabolites of phthalates determined in early childhood and thyroid function in preschool children [66]. These results support the assumption that lower thyroid hormones mediate the association between exposure to phthalates in early life and child cognitive outcomes. Furthermore, phthalates may disrupt dopaminergic neurons [34], which are associated with the progression of motor skills, calcium signaling and lipid metabolism, all of which play important roles in various neurodevelopment processes [67]. Moreover, phthalate environmental exposure has a considerable influence on the brain. This fact is invariably supported by data that presents damage to the hippocampal structural and functional plasticity [68]. The hippocampus has a prominent role in learning and memory [69]. In one study, developmental exposure to certain phthalates has resulted in a reduced number of neurons and synapses in the medial prefrontal cortex [70].
The consequences of maternal exposure to phthalates for child development depend on both genetic and environmental factors, as well as their interactions, which determine individual susceptibility to their effects. Thus, identification of such factors and their influence on the occurrence of children's neurobehavioural and neurodevelopmental disorders is of high scientific, public health and clinical relevance, consequently making it a priority public health target. Further prospective studies are required to consider sex as a statistical variable and to elucidate sex differences in susceptibility with respect to environmental factors that children are exposed to.

Conclusions
The results exhibited in this review are not entirely conclusive, with the sex specific effect not clearly pronounced. The considerable limitations in the existing literature, which are mostly related to misclassification in the exposure assessment, different tools used to assess children's neurodevelopment and different covariates considered in the performed analyses need to be highlighted. Studies addressing those limitations are crucial and might help to understand which phthalates are important factors determining child cognitive, psychomotor, behavioural and emotional development. Such knowledge can be translated into effective public health actions to protect this vulnerable population.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/ijerph182413013/s1, Table S1: selected phthalates and their respective metabolites (primary and secondary); Table S2: confounders adjusted for the model's study; Table S3: evaluation of the studies (based on methodology developed by Radke et al. 2020).
Author Contributions: A.J. collected, verified and assessed the research papers, organized, and prepared the tables, wrote the original draft of the manuscript; L.N. co-drafted the paper and performed language editing, D.K. performed a literature search, verified and assessed the collected papers and co-drafted the paper; K.P. supervised the work and co-drafted the paper. All authors have read and agreed to the published version of the manuscript.
Funding: This work was supported by the European Union's H2020 Programme under Grant Agreement Number 825884 (Synergies for Cohorts in Health: Integrating the Role of all Stakeholders).

Conflicts of Interest:
The authors declare no conflict of interest.