IMPACT OF REAL-LIFE ENVIRONMENTAL EXPOSURES ON REPRODUCTION: Phthalates disrupt female reproductive health: a call for enhanced investigation into mixtures

Searches were conducted in PubMed and Google Scholar and contained the following exposure, subject, and outcome search terms in combination: phthalate, phthalates, phthalate mixture, phthalate exposure, phthalate mixture exposure, reproduction, human, women, women’s reproductive health, HPO, HPO axis, puberty, puberty onset, puberty timing, menarche, adrenarche, pubarche, thelarche, cyclicity, menstrual cyclicity, ovary, ovarian function, folliculogenesis, steroidogenesis, polycystic ovary syndrome, polycystic ovary, polycystic, primary ovarian insufficiency, menopause, reproductive aging, ovulation, uterus, uterine function, endometriosis, leiomyomas, uterine leiomyomas, fibroids, uterine fibroids, in vitro fertilization (IVF), IVF outcomes, fertility, fecundity, fecundability, pregnancy, pregnancy outcomes, pregnancy length, pregnancy loss, miscarriage, spontaneous abortion, fetal growth, fetal development, preterm birth, gestational age, birth weight, birth length, and anogenital distance. The included epidemiological studies must have investigated associations between phthalate levels and the reproductive outcomes listed above. Very minimal to no exclusion criteria based on statistical analysis, study design, cohort geographical location, cohort size, etc., were applied, but rather any limitations based on these parameters are detailed in the sections below. The included in vivo and in vitro basic sciences studies must have investigated phthalate mixture exposures on the reproductive outcomes listed above. Exclusion criteria included studies only investigating single phthalate exposures. Studies were not excluded if they did not report a significant outcome, and these null associations/effects are reported within the sections below. However, references were excluded if the search terms yielded studies not investigating reproductive outcomes (e.g. leiomyomas outside of the uterus, adrenal steroidogenesis, polycystic kidneys, etc.). An emphasis was placed on studies published within the last ten years, and those beyond this threshold were included if the outcomes studied only had limited references. Seventy-five references met the criteria of investigating the associations of phthalate exposure with impaired reproductive health outcomes in women, while eleven references met the criteria for the basic science studies that explicitly used phthalate mixture exposures when investigating the effects on female reproductive health. The female reproductive system incorporates multiple organ systems that communicate via cyclic hormonal regulation to maintain fertility and reproductive health. Specifically, hormonal regulation of the female reproductive system is facilitated by the HPO axis ( Wilson et al, 1993 ). The hypothalamus produces gonadotropin-releasing hormone (GnRH), which acts on the anterior pituitary to induce secretion of follicle-stimulating hormone (FSH) and luteinizing hormone (LH). FSH acts on the ovary to stimulate follicular growth and the production of estradiol throughout the follicular phase of the menstrual cycle. Once a critical threshold of estradiol is met, the anterior pituitary produces a surge of LH via positive feedback. This rapid and dramatic increase in LH results in various molecular changes in the dominant follicle resulting in ovulation. Ovulation prompts the entry into the next stage of the menstrual cycle, the luteal phase. The remnant follicular structure is differentiated into the corpus luteum, which predominantly produces progesterone under tonic LH stimulation. A comprehensive list of the studies investigating the effects of phthalates on the HPO axis can be found in Supplementary Table 1 , and a summary of the effects are presented in Figure 1 . The HPO axis plays a critical role in the initiation of puberty in girls. Around the time of puberty, GnRH secretion increases in both frequency and amplitude ( Lalwani et al, 2003 ). The anterior pituitary responds to the increase in GnRH by increasing the expression of GnRH receptors. This results in increased sensitivity to GnRH, stimulating the anterior pituitary to secrete higher levels of FSH and LH, ultimately resulting in the establishment of cyclicity ( Lalwani et al, 2003 ). This entry into regular menstrual cycles, or puberty, is defined by menarche (first menstruation), adrenarche (activation of the adrenal gland), pubarche (pubic hair development), and thelarche (breast development). While the HPO axis is not the sole contributor to pubertal onset, it develops to facilitate the multi-organ endocrine communication required for female fertility and cyclicity. Phthalate mixture studies on pubertal timing are limited; however, studies that investigated the effects of single phthalates have reported higher urinary levels of MEP, MBzP, MEHHP, and MEOHP in girls with precocious puberty than age-matched controls ( Chen et al, 2013 ; Srilanchakon et al, 2017 ). In contrast, another study reported an association of higher levels of DEHP metabolites (∑DEHP) in urine with delayed pubertal development ( Kasper-Sonnenberg et al, 2017 ), while another study found no association of phthalate metabolites, either individually or in mixtures, on pubertal timing ( Mouritsen et al, 2013 ). Other studies have investigated the effects on puberty by observing the timing of its individual components: menarche, adrenarche, pubarche, or thelarche. One study investigated the joint associations of urinary phthalate metabolites in a mixture (MBzP, MCiOP, MECPP, MEHHP, MEHP, MEOHP, MiNP, MHiNP, and MOiNP) and found a significant association with delayed pubarche. Specifically, the girls with the highest exposure to this phthalate mixture were delayed by approximately 0.69 years compared to the lowest exposure group ( Frederiksen et al, 2012 ). Higher concentrations of MEHHP, MMP, and MEOHP in urine were associated with earlier menarche in some studies ( Binder et al, 2018 ; Zhang et al, 2015 ), while MEP and ∑DEHP were associated with later menarche ( Berger et al, 2018 ; Binder et al, 2018 ; Kasper-Sonnenberg et al, 2017 ). However, another study found no association of urinary or serum concentration of phthalate metabolites with age at menarche ( Buttke et al, 2012 ). Additionally, higher levels of phthalate diesters and metabolites in serum (DBP, DEP, DEHP, and MEHP), urine (MCiDP, MEHHP, MEOHP, and MEHP), and gestational urine (MBzP and ∑DEHP) were associated with earlier and more rapid thelarche ( Cathey et al, 2020 ; Colón et al, 2000 ; Kasper-Sonnenberg et al, 2017 ; Zhang et al, 2015 ), while higher urinary levels of MnBP, MBzP, MEP, and ∑DEHP have been associated with later thelarche ( Berger et al, 2018 ; Kasper-Sonnenberg et al, 2017 ). Additionally, higher levels of MCiDP in urine was associated with earlier pubarche ( Kasper-Sonnenberg et al, 2017 ). These results differ from some studies which reported no impact of phthalate metabolite levels on pubarche ( Mouritsen et al, 2013 ) or thelarche ( Frederiksen et al, 2012 ; Mouritsen et al, 2013 ). Finally, higher concentrations of ∑DEHP have been linked to earlier adrenarche in girls ( Freire et al, 2022 ). Due to their ability to impair hormonal signaling, phthalate exposure may cause alterations in menstrual cyclicity. One study found that increasing urinary concentrations of MEHP were associated with shorter bleeding duration, and increasing concentrations of MECPP and ∑DEHP in FF were associated with a shorter menstrual cycle ( Li et al, 2024 ). Additionally, urinary concentrations of MCOP were associated with a shorter luteal phase ( Jukic et al, 2016 ). Collectively, these findings suggest that phthalate exposure may impair the proper function of the HPO axis via disruption of pubertal onset and cyclicity. Consistently, higher levels of phthalate metabolites are detected in the urine of individuals with altered timing of pubertal onset. These conclusions are supported by strong statistical significance (p=0.001–0.05), but these studies contain relatively small cohort sizes (n=41–73). Studies that investigated cyclicity consistently reported shortened menstrual cycles with strong statistical significance (p=0.01–0.05) and with larger cohort sizes (n=221–441) ( Supplementary Table 1 ). Conclusive association of phthalate exposure to altered timing of puberty or adulthood cyclicity quickly becomes more complicated, and studies presented conflicting results. At present, only three studies incorporated mixture analyses when investigating the effects of phthalate exposure on the HPO axis. Therefore, mixture studies are severely lacking and should be a focus of future studies, as this more accurately reflects human exposure. For individual phthalates, studies have reported associations between higher levels of phthalates and altered pubertal timing, while some studies reported no associations. Of the studies that reported a significant association, cohort size ranged from n=200–725, while null findings were reported in studies with less participants, n=84–400. The difference in cohort size could be a contributor to the discrepancy within the literature. Additionally, both studies that reported no significant association between phthalate exposure and pubertal onset defined significance at p<0.05. Of the ten studies that found significant associations, two defined significance at p<0.1 ( Supplementary Table 1 ), which reflects less stringency and potentially weak associations within these two studies. To limit conflicting findings and ensure reliability of conclusions, future studies should incorporate larger population sizes and utilize robust statistical tests and stringent standards for significance. The ovary facilitates fertility through ovulation, which is dependent upon the appropriate development and maturation of ovarian follicles via folliculogenesis (Fletcher et al., 2022). Folliculogenesis is the irreversible development of immature primordial follicles to preovulatory follicles and occurs simultaneously with ovarian steroidogenesis, which is the production of sex steroid hormones by maturing antral follicles ( Hannon and Flaws, 2015 ). These processes are regulated, in part, by FSH and LH, as well as anti-müllerian hormone (AMH), which is produced by ovarian follicles and is a marker of the ovarian reserve ( Bedenk et al, 2020 ). Both folliculogenesis and the production of critical sex steroid hormones like estradiol and progesterone are essential for female fertility; therefore, it is important that these processes remain undisturbed to preserve female reproductive health. Studies in rodents have shown that phthalate exposures dysregulate the female reproductive system by disrupting folliculogenesis and steroidogenesis, but much less is known in women ( Hannon and Flaws, 2015 ). The effects of phthalates on ovarian function and associated diseases in women are described below and summarized in Supplementary Table 1 and Figure 1 . Given that phthalates are EDCs and that ovarian processes are hormonally regulated, studies have measured phthalate levels and investigated their associations with defects in folliculogenesis and steroidogenesis in women. One study found that the joint associations of urinary phthalate metabolites in a mixture (MEP, MBP, MiBP, MBzP, MEHP, MEHHP, MEOHP, and MECPP) were not significantly correlated with altered antral follicle count (AFC; an assessment of folliculogenesis) or diminished ovarian reserve outcomes (defined by an AFC 10 UI/L, or primary infertility) ( Génard-Walton et al, 2023 ). While that study reported no observable effects, additional studies found that higher levels of MBP, MEOHP, and jointly associated phthalate metabolites in a mixture (MMP, MEP, MBP, MBzP, MEHP, MEHHP, MEOHP, and MOP) in urine and MEHHP in serum were associated with higher AFC ( Hu et al, 2024 ; Li et al, 2022 ; Yao et al, 2023b ). Other studies reported the opposite effect where increased levels of phthalate metabolites in FF (MiBP, MnBP, and MiNP) and urine (MEHP, MEHHP, MEOHP, MECPP and ΣDEHP) were associated with decreased AFC, and urinary levels of MMP, MiBP, MEHP, and MECPP were significantly higher in women with decreased ovarian reserve compared to the control group ( Beck et al, 2024 ; Hu et al, 2024 ; Messerlian et al, 2016 ; Parikh et al, 2024 ). Because ovarian dysfunction can manifest in both increased and decreased AFC, these findings suggest potential disruptions in follicle development or premature depletion of the ovarian reserve associated with phthalate exposure. Additionally, studies show that higher urinary levels of MMP were positively associated with FSH levels, and higher levels of phthalates in urine (MECPP, MEHHP, MEOHP, MMP, and MiBP), FF (MBzP, MiDP), and serum (MiBP) were negatively associated with AMH levels ( Ding et al, 2023 ; Du et al, 2019 ; Hu et al, 2024 ; Li et al, 2022 ; Parikh et al, 2024 ). These findings further suggest that phthalate exposure may disrupt ovarian folliculogenesis by altering the levels of its hormonal regulators, FSH and AMH. Regarding ovarian steroidogenesis, one study reported a dose-response inverse association between higher FF phthalate metabolite levels and lower levels of estradiol (MMP), progesterone (MMP, MEP, and MEHP), and testosterone (MMP) while higher levels of other phthalate metabolites were associated with higher levels of estradiol (MEHHP, MBP), progesterone (MEHHP), and testosterone (MEHHP) ( Du et al, 2019 ). Another study found significant associations between higher urinary levels of MCOP and MnBP and lower levels of testosterone ( Ding et al, 2023 ). Interestingly, while higher serum levels of MEHP and MEOHP were negatively associated with estradiol levels, higher levels of MCMHP were positively associated with estradiol levels ( Li et al, 2022 ). These findings demonstrate a correlation between phthalate exposure and disrupted sex steroid hormone production in women. Primary ovarian insufficiency (POI) is classified as abnormal or diminished ovarian function prior to the age of 40, in which women exhibit decreased levels of estrogen and AMH, increased levels of FSH, and irregular menstrual cycles ( Sopiarz & Sparzak, 2024 ). Two studies showed that higher levels of serum MBP ( Özel et al, 2019 ) and urinary MiBP ( Cao et al, 2020 ) were significantly associated with increased odds of POI. Additionally, a study found that women with elevated urinary levels of MEHHP and MEOHP reported a mean age at last menstrual cycle over 3 years earlier than women with lower levels, associating phthalate exposure with earlier age of menopause ( Grindler et al, 2015 ). Although studies investigating phthalates and prevalence of POI in women are limited, these findings suggest that phthalate exposure may increase the odds of POI and are supported by studies that associate phthalate exposure with higher FSH and lower AFC and AMH ( Beck et al, 2024 ; Ding et al, 2023 ; Du et al, 2019 ; Messerlian et al, 2016 ; Parikh et al, 2024 ). Polycystic ovary syndrome (PCOS) is characterized by irregular menstrual cycles, hyperandrogenism, and polycystic ovaries (enlarged ovaries with presence of fluid-filled cysts) ( Zhang et al, 2023a ). When analyzed as a mixture (MEP, MBP, MiBP, MECPP, MEOHP, MEHHP, MEHP, and MBzP) using the molar sum of phthalate metabolites, a study found a positive association with higher urinary levels of the mixture and prevalence of PCOS ( Zhang et al, 2023a ). The same study and another have shown that there were significantly increased levels of phthalates in FF (DEHP) and in urine (MBzP, MEHP, and ΣDEHP) of women with PCOS compared to women without PCOS ( Jin et al, 2019 ; Zhang et al, 2023a ). Another study reported that increased urinary levels of ΣDEHP or MEOHP and MECPP individually increased the odds of PCOS by 40.5%, 41%, and 38%, respectively ( Al-Saleh, 2022 ). The same study also reported that the odds of PCOS decreased by 44% with higher levels of %MEHP (the ratio of MEHP to ΣDEHP) ( Al-Saleh, 2022 ). These findings suggest that phthalate exposure may be associated with increased prevalence of PCOS in women but further investigation is needed. Taken together, the summarized studies show that phthalate exposure is associated with altered ovarian folliculogenesis and steroidogenesis and may contribute to the pathogenesis or progression of diseases like POI and PCOS. However, it is evident that there are inconsistencies in the literature depending on the study population, phthalates measured, and strength of statistical analyses. For example, while phthalate exposure was associated with altered AFC, all studies that reported increased AFC occurred in cohorts of women in China, while studies that reported decreased AFC occurred in cohorts of women in the United States, India, and Denmark. Additionally, although a study that investigated exposure to phthalate mixtures reported no association with AFC in a cohort of women in the United States, both studies of single phthalate exposure in cohorts in the United States and phthalate mixture exposure in cohorts in China reported decreased and increased AFC, respectively. Although studies across multiple populations remained consistent in reporting lower AMH levels associated with phthalate exposure and higher phthalate levels in women with POI and PCOS, studies from cohorts in China reported both increased and decreased sex steroid hormone levels with phthalate exposure depending on the compound. Despite these inter-study differences, only 1 out of 15 studies reported no association between phthalate exposure and ovarian outcomes, and the remainder of the studies reported associations with stringent statistical significance (p<0.01–0.05). Additionally, the remainder of the studies reported these significant associations in multiple cohorts ranging from 50–30,000 women, except two studies which included cohorts of 20–30 women. The majority of the existing studies demonstrated strong associations with large cohort sizes, but it is important to note that three studies reported significant associations using less stringent statistical significance and with wide confidence intervals (CIs) ( Supplementary Table 1 ). The inconsistencies in the reported results suggest that phthalate exposure may impact ovarian function differently depending on population demographics and susceptibility to toxicity, the specific phthalate compound(s), and whether single phthalate and/or phthalate mixtures were investigated. Additionally, the summarized studies investigating the effects of phthalate exposure on ovarian function and disease are mostly limited to single phthalate analyses with only three studies incorporating mixture analyses. Thus, future studies that prioritize phthalate mixture analyses, incorporate strong standard for statistical significance, and account for the inconsistencies mentioned previously are needed. The uterus is the site of implantation of the embryo and is responsible for embryonic/fetal development and maintenance of pregnancy ( Taylor & Gomel, 2008 ). The endometrium undergoes regular cycles of proliferation, secretion, and degeneration throughout reproductive life. These ongoing changes are regulated by ovarian sex steroid hormones, uterine inflammatory mediators, and immune cells ( Park & Yang, 2011 ). Thus, disruptions in this interconnected signaling network can have major negative effects on fertility and uterine function. Since uterine function is partially controlled by ovarian hormones, it is important to note that disruptions in uterine function could be due to direct action of phthalates on the uterus, or an indirect effect of phthalates on the ovary. A comprehensive list of the studies investigating the effects of phthalates on uterine function and disease are described below and summarized in Supplementary Table 1 and Figure 1 . Endometriosis is a disease characterized by the growth of endometrial cells outside of the uterus leading to debilitating pain and infertility ( Bulun, 2009 ). As a chronic estrogen-dependent inflammatory disease, symptoms can be improved using estrogen blockers, but surgical and medical management becomes less effective with subsequent treatments ( Bulun et al, 2019 ). While studies that investigated the association between phthalate mixtures and endometriosis are limited, levels of DEHP, DnBP, BzBP, DnOP, and MEHP in serum were higher in women with endometriosis than in women without ( Cobellis et al, 2003 ; Kim et al, 2011 ; Nazir et al, 2018 ; Reddy et al, 2006 ). Other studies found that higher urinary levels of MBP, MCMHP, MECPP, MEHP, MEHHP, and MEOHP were associated with an increased risk of endometriosis ( Buck Louis et al, 2013 ; Cai et al, 2019 ). However, some studies found that urinary MEHP was inversely associated with endometriosis ( Upson et al, 2013 ), and women diagnosed with endometriosis had lower urinary levels of MEHP when compared to healthy controls ( Weuve et al, 2010 ). Other studies reported no association of phthalate metabolites individually or in a mixture with risk of endometriosis ( Itoh et al, 2009 ; Moreira Fernandez et al, 2019 ; Upson et al, 2013 ). Uterine leiomyomas are non-malignant tumors of the endometrium and myometrium ( Ryan et al, 2005 ). Although most cases are asymptomatic, abnormal uterine bleeding, pelvic pain, and uterine dysfunction can occur ( Stewart, 2001 ). Uterine leiomyomas are the most common benign neoplasm in reproductive aged women ( Stewart, 2015 ), with the likelihood of development increasing with age until menopause ( Wise & Laughlin-Tommaso, 2016 ). Additionally, women with a previous diagnosis of endometriosis or cervical polyps are at higher risk of developing leiomyomas ( Ryan et al, 2005 ). Studies have reported that a mixture of urinary anti-androgenic phthalates (ΣAA), metabolites known to disrupt androgen signaling in animal and epidemiological studies ( Hannon & Flaws, 2015 ; Marie et al, 2015 ; Marsee et al, 2006 ), was associated with increased risk of prior leiomyoma diagnosis and increased uterine volume (a risk factor of leiomyoma development) ( Pacyga et al, 2022 ; Zota et al, 2019 ). In contrast, a meta-analysis of nine studies found no significance of total phthalate metabolites on risk of leiomyoma development, despite finding a significant association with ∑DEHP ( Fu et al, 2017 ). Of individual phthalate metabolites, increasing urinary concentrations of MBzP, MEHP, MBP, and ∑DEHP were associated with an increased risk of leiomyoma development or increased uterine volume ( Fruh et al, 2021 ; Pacyga et al, 2022 ; Zhang et al, 2023b ; Zota et al, 2019 ). Other studies have stated that women with leiomyomas had significantly higher urinary levels of ∑DEHP, MEHP, MMP, MECPP, and MEOHP than women without ( Huang et al, 2010 ; Kim et al, 2017 ; Kim et al, 2016 ; Lee et al, 2020 ). Alternatively, one study reported that urinary MMP levels were lower in women with leiomyomas than women without and found no association of total phthalate exposure with odds of leiomyomas ( Pollack et al, 2015 ). Uterine function can be negatively affected by exposure to EDCs like phthalates. Unfortunately, there is no consensus as to whether phthalate exposure increases patient risk of uterine diseases, such as endometriosis and uterine leiomyomas. It is clear that there is a lack of studies investigating phthalate mixture effects on uterine diseases, with only one mixture study on endometriosis and two mixture studies on leiomyomas. Future studies should incorporate mixture analyses to better capture human exposure risks. For individual phthalates, many studies reported both higher levels of phthalates in diseased individuals and increased risk of disease development. However, there are some that reported lower levels of certain phthalates and no association with risk of disease. Discrepancies could be attributed to variable standards of statistical significance (p=0.001–0.05) ( Supplementary Table 1 ) and cohort size (n=30–6579) between the studies. Of note, the largest cohort sizes were achieved by two meta-analyses reexamining existing studies, both of which reported at least one phthalate metabolite to be significantly associated with risk of disease. One included eight studies on endometriosis, n=57–1107 ( Cai et al, 2019 ), and the other included nine studies on leiomyomas, n=57–3003 ( Fu et al, 2017 ). Although it is difficult to increase cohort size in clinical research, meta-analyses present an opportunity for larger analyses with existing data and should be considered for future studies. Despite the need for larger sample sizes, population demographics could contribute to discrepancies within the literature. For example, one study found an overall increase in odds of endometriosis with higher levels of urinary MEHHP, but the association was only significant within the Asian population once patients from Asia and the United States were analyzed separately ( Cai et al, 2019 ). This highlights the importance of demographic analysis within future studies, as certain populations may have different sensitivity to phthalate toxicities. Further, conflicting results can be attributed to the matrix used to assess exposure. For example, while serum levels of MEHP were higher in women with endometriosis, urinary levels were reported to be lower ( Kim et al, 2011 ; Weuve et al, 2010 ). As such, the matrix used to assess phthalate exposure should be considered when interpreting the results of future studies. In vitro fertilization (IVF) is an assisted reproductive technology in which mature oocytes are fertilized outside of the body via intracytoplasmic sperm injection (ICSI), grown in embryonic culture, and transferred back into the uterus of the female patient ( Choe & Shanks, 2024 ). The IVF process is complex and requires multiple checkpoints for appropriate developmental changes following sperm injection. These changes include the combination of genetic material from the oocyte and sperm, resulting in a fertilized egg that undergoes cell division and further embryonic development to the blastocyst stage ( Colaco & Sakkas, 2018 ; Georgadaki et al, 2016 ). After screening for chromosomal and genetic abnormalities, quality embryos (typically in the blastocyst stage) can be transferred back into the patient, whereby uterine implantation and pregnancy hopefully occur. The development of IVF has provided an avenue in which epidemiological associations between exposure levels to reproductive toxicants like phthalates and IVF checkpoints/endpoints can be evaluated ( Land et al, 2022 ). These associations can be expanded to incorporate the reproductive status of the patients to include a population of subfertile women (female patients with various reproductive disorders) and seemingly fertile women (couples experiencing male-factor infertility, egg donors/surrogates, and female same-sex couples that do not have an infertility diagnosis) ( Land et al, 2022 ). Although limited, there are studies investigating the associations of phthalate exposure and IVF outcomes, which are described below and are summarized in Supplementary Table 2 and Figure 2 . While these studies primarily analyzed associations based on the levels of individual phthalates, a few have incorporated mixture analyses. In one study, FF levels of phthalate metabolites were evaluated, and a phthalate mixture (MMP, MEP, MBP, MBzP, MEHP, MEHHP, and MEOHP) specific to this cohort of patients was derived. This mixture was compared to the levels of multiple cytokines that are thought to serve as biomarkers of oocyte quality, and the results indicated that this specific phthalate mixture had a positive effect on TNF-α ( Wang et al, 2023 ). The authors hypothesized that phthalate-induced increases in TNF-α could impair oocyte quality because other studies have shown that poor quality oocytes were contained in follicles that had significantly higher TNF-α levels in FF ( Lee et al, 2000 ). This hypothesis was supported by their other study using a different phthalate mixture (MEP, MiBP, MBP, MBzP, MEHP, MEHHP, MEOHP, and MECPP) relevant to that specific cohort of IVF patients. This mixture was associated with lower total oocyte yield and decreased yield of mature eggs capable of ICSI/fertilization (meiosis II eggs) ( Yao et al, 2023a ). Additional studies analyzing single phthalate levels reported associations with negative IVF outcomes. Specifically, multiple studies reported that when phthalate metabolites levels were individually elevated in urine (ΣDEHP, MEHP, MEHHP, MEOHP, MECPP, MCNP, MCOP, MBP, and MEP) and FF (ΣDEHP, MEHP, MEHHP, and MBzP), there were significantly less total and mature oocytes collected during IVF egg retrieval ( Hauser et al, 2016 ; Machtinger et al, 2018 ; Yao et al, 2023a ). Decreased fertilization odds have also been associated with increased urinary MBP and MEP levels ( Deng et al, 2020 ). In addition to MBP and MEP, elevated urinary levels of MEHHP, MEOHP, MECPP, ΣDEHP, MCNP, and MCOP were also associated with a lower number of fertilized eggs ( Hauser et al, 2016 ; Machtinger et al, 2018 ). Another study reported an association with a lower number of fertilized eggs and elevated MEHP, MEHHP, and MBzP FF levels ( Yao et al, 2023a ). Increased urinary concentrations of certain phthalate metabolites have also been attributed to a lower number of quality embryos (MEHHP, MEOHP, MECPP, ΣDEHP, MHiBP, MBP, and MiBP), decreased odds of quality blastocysts (MEP and MMP), and lower probability of embryo implantation (DEHP-factor: MEHP, MEHHP, MEOHP, and MECPP) ( Deng et al, 2020 ; Machtinger et al, 2018 ; Mínguez-Alarcón et al, 2019 ). Elevated phthalate levels have conflicting associations with probability of clinical pregnancy in women undergoing IVF. Studies have reported a lower likelihood of pregnancy with elevated urinary ΣDEHP, MEHHP, MEOHP, DEHP-factor, and MBP levels ( Begum et al, 2021 ; Hauser et al, 2016 ; Mínguez-Alarcón et al, 2019 ), while another study reported an increased likelihood of pregnancy with elevated urinary levels of MEHP alone ( Al-Saleh et al, 2019 ). This is concerning because a singular DEHP metabolite (MEHP) can have opposite findings individually compared to when it is combined with other DEHP metabolites (ΣDEHP). Finally, elevated urinary levels of phthalate metabolites were associated with reduced probability of live birth (ΣDEHP, DEHP-factor, and MHxP) and higher probability of unsuccessful live birth (MEHP and MEP) following IVF ( Al-Saleh et al, 2019 ; Begum et al, 2021 ; Hauser et al, 2016 ; Mínguez-Alarcón et al, 2019 ). Findings from the cited studies were mostly consistent despite varying cohort sizes, population demographics, and analytical methodology. For example, studies conducted in both North America ( Hauser et al, 2016 ) and Asia ( Machtinger et al, 2018 ; Yao et al, 2023a ) with cohorts ranging from 136–641 women reported lower yield of total oocytes and mature eggs during the IVF retrieval process in addition to a lower number of fertilized eggs with elevated urinary/FF levels of phthalates, but only studies from cohorts in Asia reported a lower number of quality embryos. Additionally, a larger cohort of 663 women in Asia reported lower fertilization odds and lower odds of having quality blastocysts when certain phthalates are elevated in urine ( Deng et al, 2020 ). Multiple studies from cohorts of 56–420 women in North America ( Begum et al, 2021 ; Hauser et al, 2016 ; Mínguez-Alarcón et al, 2019 ) reported associations of elevated phthalate levels and a lower likelihood/probability of pregnancy and live birth following IVF. However, a cohort of 599 couples undergoing IVF/ICSI in Asia ( Al-Saleh et al, 2019 ) reported significant associations of elevated phthalate levels and a higher likelihood of pregnancy, although this study also reported a higher probability of unsuccessful live birth with elevated phthalate levels—consistent with the North American cohorts ( Begum et al, 2021 ; Hauser et al, 2016 ; Mínguez-Alarcón et al, 2019 ). These conflicting findings could be attributed to differences in population demographics and susceptibility to phthalate toxicity, especially at specific timepoints of the IVF process/pregnancy. While many of these studies provided strong statistical evidence (p<0.01–0.05) with large cohort sizes, it is important to note that some studies reported significant associations using less stringent criteria, suggesting that the results may be less robust ( Supplementary Table 2 ). It is also important to note that only two of the cited studies utilized statistical modeling to evaluate potential mixture effects. Despite the discrepancies noted above, these findings link increased phthalate levels with negative outcomes in many IVF processes including egg retrieval, fertilization success, embryonic development, implantation rate, pregnancy, and live birth. These negative outcomes are observed even when these patients receive supraphysiological doses of hormones to stimulate multiple follicles to develop to maturity and ovulate, thus providing the clinic with multiple chances of fertilization. After successful fertilization, the hormonal profiles of these patients are also strategically controlled via exogenous hormone administration to create the most suitable uterine environment for successful embryo implantation, hopefully leading to a healthy pregnancy and live birth. Despite this, these data suggest that phthalates may be targeting oocytes, the fertilization process, embryonic development, and the appropriate uterine environment needed for pregnancy—all of which cause defects that can contribute to the prevalence of infertility. The ability to conceive, maintain a pregnancy, and successfully give birth is regulated by many complex biological processes that are dependent upon appropriate endocrine function. Issues that arise from conception until childbirth could be attributed to various genetic or environmental factors, including exposure to phthalates. Elevated phthalate levels have been associated with altered pregnancy outcomes, which are described below and are summarized in Supplementary Table 2 and Figure 3 . Previous studies have reported that plasma concentrations of phthalates (MBzP and MEHHP) were elevated in infertile women when compared to fertile women, suggesting that exposure may impact fertility status ( Pednekar et al, 2018 ). One study also reported that urinary levels of ΣDEHP and MnBP were significantly higher in a non-conception cycle versus a conception cycle from the same woman ( Jukic et al, 2016 ). This is supported by another study that reported associations of elevated urinary levels of MBP, MEHP, and MBzP and lower fecundability—the probability of achieving pregnancy within a single menstrual cycle ( Nobles et al, 2023 ). Additionally, elevated urinary levels of MEP were associated with decreased fecundity—the biological ability to reproduce over the reproductive lifespan when actively trying to conceive ( Thomsen et al, 2017 ). Most notably, there are no reports of a phthalate mixture effect on fecundability or fecundity in humans, highlighting the need for further studies. Early pregnancy loss, miscarriage, or spontaneous abortion are often used interchangeably in the literature, and they are defined as pregnancy loss occurring before 20 weeks of gestation ( Radke et al, 2019 ). Recurrent pregnancy loss refers to two or more clinically recognized pregnancies that failed before 20–24 weeks gestation ( Dimitriadis et al, 2020 ). Studies that investigated individual phthalates suggest that exposure could contribute to the incidence of pregnancy loss, while no studies have reported a phthalate mixture effect on pregnancy loss in humans. Elevated urinary ΣDEHP and ΣDBP metabolites were associated with lower odds of early pregnancy loss and elevated risk of recurrent pregnancy loss, respectively ( Jukic et al, 2016 ; Liao et al, 2018 ). One of these studies also reported that elevated urinary levels of an individual DEHP metabolite (MEOHP) were associated with lower odds of early pregnancy loss ( Jukic et al, 2016 ), however another study reported that pregnancy loss was increased among women with elevated urinary levels of a different DEHP metabolite (MEHP) ( Toft et al, 2012 ). Additionally, elevated urinary levels of MMP were associated with an increased risk of miscarriage ( He et al, 2021 ), but a more recent study with a larger sample population reported no consistent associations of elevated urinary levels of individual or jointly associated phthalates with pregnancy loss when evaluating 20 phthalate metabolites originating from 14 different parent phthalate compounds ( Nobles et al, 2023 ). Fetal growth data are direct assessments of the developing fetus, but they are also indirect assessments of the maternal environment in which a fetus develops. Thus, they can be indicative of a woman’s ability to properly carry and grow a fetus. Fetal growth is commonly assessed via femur length (FL), head circumference (HC), abdominal circumference (AC), and estimated fetal weight (EFW) ( Casas et al, 2016 ). Multiple parameters measuring fetal growth (HC, AC, FL, EFW) were decreased with elevated urinary levels of ΣDEHP and MECPP ( Ferguson et al, 2016 ). While one study reported increased FL (MBzP) ( Casas et al, 2016 ), most studies primarily reported associations of elevated urinary phthalate levels with decreased HC (MECPP, MEHOP, MEP, MnBP, and MBzP) ( Casas et al, 2016 ; Ferguson et al, 2016 ; Smarr et al, 2015 ) or increased HC (MCPP and ΣDBP) ( Watkins et al, 2016 ). After delivery, additional data can be collected to assess appropriate fetal development of the newborn, including calculation of gestational age (GA), birth weight and length, and the anogenital index (AGI). Multiple studies have reported decreased gestational age at birth with elevated urinary levels of phthalic acid (PA; the backbone of all parent phthalate compounds), ΣDEHP, ΣDiNP, ΣDnOP, ΣDBP, MCINP, MNP, MMP, and MOP ( Smarr et al, 2015 ; Trasande et al, 2024 ; Watkins et al, 2016 ). However, elevated preconception urinary levels of MEHP were associated with an increased gestational age at birth ( Smarr et al, 2015 ), while a different study reported that the absence of MEHP in cord blood was associated with an increased gestational age at delivery ( Latini et al, 2003 ). Elevated maternal exposure to various phthalates has also been associated with altered birth weight and length in newborns. Specifically, elevated urinary phthalate levels were associated with decreases (MCMHP, MEHP, MMP, MOP, ΣDBP, PA, ΣDEHP, ΣDiNP, ΣDnOP, and MCINP) and increases (MCPP, ΣDBP, and MBzP) in newborn birth weight, in addition to decreases (MMP, PA, ΣDEHP, ΣDiNP, and ΣDnOP) and increases (MCPP, MBzP, and ΣDBP) in birth length ( Casas et al, 2016 ; Smarr et al, 2015 ; Trasande et al, 2024 ; Watkins et al, 2016 ). Finally, elevated prenatal urinary levels of phthalates were individually (MBzP, MBP, MEP, and MiBP) and jointly associated with higher odds of smaller AGI in male infants, suggesting that prenatal exposure to individual and mixtures of phthalates can adversely affect human male reproductive development ( Swan et al, 2005 ). The length of pregnancy averages 40 weeks of gestation, and preterm birth occurs before 37 weeks gestation ( Radke et al, 2019 ). Preterm birth can be further categorized into placental or spontaneous preterm birth. Placental preterm birth is typically suspected by obstetricians as it frequently occurs after signs of intrauterine growth restriction and pre-eclampsia ( Khandre et al, 2022 ). Spontaneous preterm births occur following preterm labor induced by an early membrane rupture ( Khandre et al, 2022 ). Preterm birth is concerning because research suggests that women with babies born preterm have more physical and mental health issues post-partum, while infants that were born preterm are more likely to experience further developmental and health problems ( Adams-Chapman et al, 2018 ; Henderson et al, 2016 ; Humberg et al, 2020 ; Miller et al, 2016 ). Growing literature suggests that elevated phthalate exposure during gestation could contribute to the incidence of preterm birth. One study reported that elevated urinary levels of MBP, MHBP, MiBP, and MHiBP were significantly associated with shorter gestation ( Ferguson et al., 2019 ). Elevated urinary levels of certain phthalate metabolites were associated with higher odds of uncategorized preterm birth (PA, ΣDEHP, ΣDiNP, ΣDnOP, MBP, MCINP, and MCPP) and spontaneous preterm birth (MBP, MEP, MiBP, and MCPP) ( Trasande et al, 2024 ; Wang et al, 2022 ). Additionally, some studies suggest that the stage of pregnancy in which phthalate exposures are elevated might be a contributing factor to preterm birth risk. For example, one study reported that elevated urinary levels of MECPP during early pregnancy were associated with increased odds of placental preterm birth, but elevated levels of MECPP during late pregnancy were associated with increased odds of spontaneous preterm birth ( Ferguson et al, 2014 ). When urinary levels were elevated around mid-pregnancy (23 weeks), MBP, MiBP, and MHiBP were associated with increased odds of uncategorized preterm birth, while only MiBP and MHiBP were associated with increased odds of spontaneous preterm birth ( Ferguson et al, 2019 ). During late pregnancy, elevated urinary phthalate levels were associated with increased odds of uncategorized preterm birth (MBP) and spontaneous preterm birth (ΣDEHP, MECPP, MBP and MBzP) ( Ferguson et al, 2014 ; Meeker et al, 2009 ). Taken together, the summarized studies show that phthalate exposure is associated with altered pregnancy outcomes and that phthalate-induced effects are somewhat inconsistent depending on the specific phthalate measured, the time frame in which phthalate levels were elevated, and/or fetal sex. Some of these differences may be due to population demographics, cohort size, and/or strength of the association. For example, in cohorts of similar size, a study conducted in North America reported lower odds of early pregnancy loss with elevated phthalate levels ( Jukic et al, 2016 ), while a higher occurrence of pregnancy loss was reported from a study conducted in Europe ( Toft et al, 2012 ). In contrast, a much larger cohort in North America had no consistent associations of elevated urinary levels of individual or jointly associated phthalates with pregnancy loss ( Nobles et al, 2023 ). Most studies reported decreases in parameters measuring fetal growth with elevated phthalate levels, and studies that reported conflicting findings in some of these growth parameters (FL and HC) had wider CIs or smaller cohort sizes when compared to the other studies measuring the same parameters ( Supplementary Table 2 ). Differences in phthalate-induced effects on gestational age may be attributed to when exposure to that phthalate was elevated (preconception versus across pregnancy), the fetal sex, and/or the matrix in which phthalates were measured (maternal urine versus fetal cord blood). Conflicting associations with birth weight and length were dependent on the specific phthalate measured, fetal sex, when exposure was elevated (first trimester versus across pregnancy versus at delivery), and/or strength of association (wide CIs, Supplementary Table 2 ). Finally, the window of elevated exposure to specific phthalates contributed to the conflicting data that reported higher odds/risk of different types of preterm birth. Despite these differences, these data suggest that elevated phthalate exposure throughout every trimester of pregnancy, and even prior to conception, may contribute to disrupted outcomes surrounding pregnancy including: initial fertility status, pregnancy loss, fetal growth, length of pregnancy, and birthing complications. Thus, there is a critical need for further investigation into the toxicities that relevant phthalate exposures may elicit both before and during pregnancy in women. While the epidemiological studies report possible effects of phthalate exposures on female reproductive health, these types of studies are unable to determine the mechanisms of action, and unknown mechanisms are a major limitation in the literature. The basic science studies that are briefly summarized below help to bridge this gap in knowledge by providing direct exposure effects and mechanisms by which phthalate mixtures impair reproductive health outcomes. Altered sex steroid hormone levels are proposed as a mechanism by which phthalate mixture exposures disrupt cyclicity directly in adult mice and in exposed progeny in adulthood in vivo ( Adam et al, 2021 ; Patiño-García et al, 2018 ; Zhou et al, 2017b ). Inhibited progesterone receptor signaling and prostaglandin production are proposed mechanisms by which exposure to a phthalate mixture impairs ovulatory outcomes in primary human granulosa cells from women undergoing IVF and in mouse follicles with decreased ovulation rates in vitro ( Hannon et al, 2023 ; Land et al, 2021 ). Both in vivo and in vitro studies have shown that phthalate mixture exposures decrease follicle numbers, inhibit antral follicle growth, decrease estradiol production, and accelerate reproductive aging in mice ( Brehm & Flaws, 2021 ; Brehm et al, 2020 ; Meling et al, 2020 ; Safar et al, 2023 ). Decreased progesterone is a proposed mechanism by which mixture exposure increases uterine luminal epithelial cell proliferation and the fibrosis response in mice ( Li et al, 2020 ). Perhaps the above cited defects and mechanisms all contribute to observations in additional studies that report impaired fertility outcomes following phthalate mixture exposure in mice. Specifically, mixture exposure impaired fertility-related indices (increased time to pregnancy and reduced mating index, pregnancy rate, fertility index, and gestational index) and breeding capacity (failure to mate, failure to become pregnant after successful mating, loss of pregnancy, and dystocia) ( Zhou et al, 2017a ; b ). Importantly, several of the cited studies investigated prenatal exposure and reported defects across generations of offspring suggesting that phthalate mixture exposure may induce multigenerational and transgenerational effects. Continued efforts are needed to further elucidate the mechanisms of phthalate toxicity and to enhance translation to human health. Future basic science studies should directly expose primary human cells/tissues to environmentally relevant phthalate mixtures. These mixtures should be comprised of the phthalate metabolites that are present in women’s urine as an estimate of daily human exposure ( Koch & Calafat, 2009 ) or, more physiologically relevant, in tissue samples such as FF as a measure of what reaches the ovary ( Du et al, 2016 ; Yao et al, 2020 ). Additionally, the doses of the mixtures should include concentrations of the individual phthalates that are present in the urine and/or tissue samples. This approach provides relevancy by exposing samples obtained from women to mixtures that contain environmentally and physiologically relevant levels of the phthalates that women, or their tissues, are exposed to daily. The same approach should be conducted in in vivo animal studies to elucidate systemic effects and mechanisms of phthalate toxicity, and these studies should prioritize dietary and dermal exposures as these are the most common routes of exposure ( Hannon & Flaws, 2015 ). A crucial challenge remains in extrapolating the basic science findings in this section to the epidemiological ones reported above. This is largely due to the complications in translating in vivo and in vitro doses from basic science studies to the urinary/serum/FF measurements and confounding variables in epidemiology studies. However, both types of research can build upon each other—epidemiology can inform basic science studies to investigate a mechanism for a particular health defect, and basic science findings can guide epidemiologists to investigate prevalence and progression of reproductive disorders in women. Ultimately, the goal is to translate findings from these studies to improve women’s health.

Phthalates are synthetic chemicals comprised of phthalic acid with varying lengths of alkyl chains and are used as plasticizers and solvents in various common consumer goods ( Heudorf et al, 2007 ). Such goods include food and beverage packaging, building materials in homes and workspaces, medical equipment, and personal care/cosmetic products; thus, daily human exposure occurs via oral ingestion, inhalation, and dermal contact ( Heudorf et al, 2007 ; Hogberg et al, 2008 ). Upon absorption, the alkyl diester parent compounds are rapidly metabolized to monoester metabolites and additional metabolic derivatives, and these metabolites are distributed to various organs ( Koch & Calafat, 2009 ). A list of the parent compounds and their metabolites that are discussed in this review are provided in Table 1 . Human exposure to a mixture of phthalates is ubiquitous, and this is a major public health concern because phthalates are known endocrine-disrupting chemicals (EDCs) ( Gore et al, 2015 ; Land et al, 2022 ). As EDCs, phthalates target organ systems regulated by hormones via disruption of normal endocrine signaling and function. Measurable levels of phthalate metabolites are detected in human urine and blood ( Hogberg et al, 2008 ), amniotic fluid ( Silva et al, 2004 ), umbilical cord blood ( Latini et al, 2003 ), breast milk ( Hogberg et al, 2008 ), and ovarian follicular fluid (FF) ( Du et al, 2019 ; Du et al, 2016 ; Krotz et al, 2012 ), suggesting that phthalates can target reproductive and developmental health. Appropriate endocrine function is paramount for reproductive health, and endocrine-related disruptions commonly manifest into various reproductive diseases and contribute to the prevalence of infertility ( Gore et al, 2015 ; Land et al, 2022 ). Notably, phthalate exposure has been linked to disruptions in endocrine signaling through the hypothalamus-pituitary-ovarian (HPO) axis ( Gore et al, 2015 ), ovarian function ( Hannon & Flaws, 2015 ; Land et al, 2022 ), uterine function ( Caserta et al, 2021 ; Martínez-Ibarra et al, 2024 ), pregnancy ( Filardi et al, 2020 ), and fetal development ( Latini et al, 2006 ). Commonalities in these cited reviews that are also limitations in the literature include: 1) there are limited studies investigating the specific mechanisms by which phthalates disrupt reproductive and developmental health; 2) the known effects are primarily from studies investigating single phthalate exposures, even though humans are exposed to a mixture of phthalates daily; and 3) the majority of studies use animal models, so the direct translational effects on women’s health are largely unknown. To address these limitations, recent epidemiological studies have investigated the impact of phthalate mixture exposures on clinical reproductive health outcomes in women. Outcomes measured in women include effects on endocrine function (HPO axis), reproductive organ function and disease (ovary and uterus), and fertility endpoints ( in vitro fertilization and pregnancy outcomes). The primary focus of this review is to present recent literature outlining phthalate mixture effects on women’s reproductive health, as mixtures represent real life exposure to these ubiquitous EDCs. As these data are limited, epidemiological studies using single phthalate analyses will supplement mixture studies. Basic science studies investigating phthalate mixture effects on female reproductive function are briefly summarized in their own section to supplement the clinical studies in women and to discuss potential mechanisms of toxicity following exposure to phthalate mixtures. While epidemiological studies using mixture analyses and single phthalate analyses are both presented, discrepancies may arise when interpreting single phthalate effects versus mixture effects. Depending on the model used and analysis performed, a single phthalate analyzed individually may have conflicting findings when that same phthalate is analyzed within a mixture. Ultimately, chemicals within a mixture may produce a different toxic effect in the body than the individual chemical alone. For instance, chemicals in a mixture may act similar to each other and exert an additive toxic effect, act independent of each other and exert separate effects, or interact with each other to exert either a synergistic effect (greater than additive) or an antagonistic effect ( Bloch et al, 2023 ; Mourikes & Flaws, 2021 ). Given the complexity of how chemicals in a mixture may interact in the body and that mixtures reflect true human exposures, it is critical to incorporate phthalate mixture exposures and analyses in toxicology, epidemiology, and risk assessment studies.

Clinical literature compiled in this review further expands what has already been extensively documented in animal studies—that phthalates are endocrine-disrupting chemicals that target female reproductive health. While incorporating phthalate mixtures in study designs has not been done historically, recent analyses have shown that increased phthalate exposure, when analyzed as a mixture, is associated with delayed pubarche, higher AFC, the prevalence of PCOS and uterine leiomyomas, lower oocyte yield and increased biomarkers of poor oocyte quality in the IVF setting, and decreased AGI in male newborns. When analyzed individually or as the sum of the metabolites of an individual phthalate, increased phthalate exposure is associated with disruptions in the HPO axis, ovarian function, uterine function, IVF outcomes, and pregnancy outcomes. More specifically, increased phthalate exposure is associated with infertility, altered pubertal timing, disrupted menstrual cyclicity, increased odds of POI and PCOS, decreased successful pregnancies and live births in the IVF setting, increased risk of endometriosis and leiomyomas, and increased risk of preterm birth. In basic science studies, disrupted hormone production and signaling are suggested mechanisms by which phthalate mixture exposures impair cyclicity, ovarian function, uterine function, and fertility and pregnancy outcomes. Discrepancies in the literature are present, which unfortunately is a common occurrence when reviewing the impact of phthalates on human health. As cited in the sections above, increased levels of the same phthalate may have positive, negative, or null associations with female reproductive dysfunction depending on the study, and the strength of these associations can be weak to moderate. Thus, making conclusive statements on the impact of phthalate exposure on women’s reproductive health remains a challenge. These discrepancies can be attributed to the size of the cohort, geographical location of the cohort, timing of patient enrollment, sample matrix (urine versus blood versus FF), sample collection frequency (a singular sample versus the average of multiple collections), timing of sample collection on a given day (morning versus evening), timing of sample collection during a life-stage event (preconception versus early pregnancy versus late pregnancy), analytical methods used for measurements, and design of the study, among other potential variables. To address these discrepancies, careful attention must be given to matrix selection for the outcome measured in future studies. Urine, over serum, has long been considered the best matrix to measure total daily exposure levels to phthalates ( Koch & Calafat, 2009 ). This is because phthalates are non-persistent, are rapidly metabolized and excreted, and, as such, urinary levels are higher than serum levels and more indicative of daily exposure ( Koch & Calafat, 2009 ). Therefore, interpreting studies using serum measurements of phthalates/phthalate metabolites is difficult as this matrix may not correlate to a woman’s true exposure profile. Additionally, correlations between urinary and tissue (specifically FF) levels of phthalates are weak to moderate, suggesting that urinary levels may not accurately reflect what is found in the ovary and other reproductive tissues ( Yao et al, 2020 ). Therefore, when specifically measuring ovarian outcomes, it is more advantageous to use FF, as the distribution of phthalate metabolites in FF is rather distinct from measurements in the urine ( Du et al, 2016 ; Yao et al, 2020 ). Weak correlations between urinary phthalate levels and tissue phthalate levels complicate interpretation. For instance, similar associations are found between poor ovarian outcomes and elevated phthalate levels in both urine and FF ( Hauser et al, 2016 ; Yao et al, 2023a ), but poor pregnancy outcomes are only associated with elevated phthalate levels in urine ( Hauser et al, 2016 ) and not in FF ( Yao et al, 2023a ). The null findings using FF as a matrix do not consider that these pregnancy outcomes also rely on proper uterine function, and it is presently unknown if phthalate levels in any matrix correlate to levels that reach the uterus. Additionally, maternal urinary levels of MEHP ( Smarr et al, 2015 ) and fetal cord blood levels of MEHP ( Latini et al, 2003 ) had conflicting associations with gestational age, further suggesting that maternal levels of phthalate exposure may not correlate well with fetal exposure levels. To further address the discrepancies between studies reporting conflicting or null associations, appropriate statistical modeling must be considered, especially when analyzing exposures to mixtures of phthalates. The lack of incorporation of mixture analyses continues to be a major limitation in the literature. This is concerning because humans are exposed to a mixture of phthalates daily, not just individual phthalates. Herein, there are only 13 references that use adequate statistical modeling to investigate phthalate mixture effects in the epidemiological studies while 62 references are limited to single phthalate analyses. Among the complex statistical modeling that is done in epidemiological studies, the novel Bayesian kernel machine regression (BKMR) model has emerged as an approach to better identify mixture effects and the contributions of individual components within the mixture responsible for an adverse outcome ( Bobb et al, 2018 ; Bobb et al, 2015 ). Major strengths of the BKMR approach include the ability to adequately estimate exposure-outcome relationships when the effects are nonlinear and non-additive while still adjusting for covariates/confounders, the identification of potentially responsible components within the mixture through variable selection, and the identification of potentially responsible groups of highly correlated exposures when individual components cannot be identified ( Bobb et al, 2018 ; Bobb et al, 2015 ). Similar to BKMR, hierarchical integrative group least absolute shrinkage and selection operator (Higlasso), selection of nonlinear interactions by a forward stepwise algorithm (SNIF), and factor analysis for interactions (FIN) are recent, novel approaches that incorporate interactions among chemicals within a mixture ( Boss et al, 2021 ; Ferrari & Dunson, 2021 ; Narisetty et al, 2019 ). Use of these approaches has recently grown in toxicological and epidemiological studies, including studies cited above, revealing discoveries and associations that were otherwise unobservable using standard regression approaches ( Bobb et al, 2018 ; Coull et al, 2015 ; Valeri et al, 2017 ; Wang et al, 2023 ; Yao et al, 2023a ). More historical approaches for mixture analyses include weighted quartile sum (WQS) regression, which utilizes a simpler model than those cited above to assess the overall effects of the mixture and the risk contribution of individual components, but the effects of each component must be uniform (all positive or all negative) ( Zhu et al, 2024 ). Ultimately, selection of the appropriate statistical model relies on the approach/goal of the model (total mixture effects versus the contributions of chemical interactions versus prediction of the adverse health outcome and risk stratification), the specific scientific questions and hypotheses, and the variables and outcomes being measured ( Taylor et al, 2016 ; Zhu et al, 2024 ). However, there is no consensus on the best statistical model for mixture analysis given the complexities of study designs, the scientific questions being asked, the data collected, and the selection criteria mentioned above ( Taylor et al, 2016 ). Toxicologists, epidemiologists, and statisticians must continue to collaborate to develop novel statistical models for mixture analyses and to establish the standards for appropriate model selection ( Taylor et al, 2016 ). In conclusion, exposure to phthalates is increasingly being recognized as a public health concern, especially related to women’s reproductive health ( Gore et al, 2015 ; Land et al, 2022 ). Phthalate mixture exposures are associated with impairments in ovarian and uterine function and IVF and developmental outcomes, but studies using mixture analyses are far fewer than those that only analyze single phthalate effects. Thus, there continues to be an urgent need to incorporate real life, environmentally relevant phthalate mixtures in the designs of epidemiology studies (mixture analyses) and toxicology studies (direct exposures). Additionally, it is important to note that humans are not just exposed to phthalates, but rather are exposed daily to a multitude of EDCs, including bisphenols, parabens, per‐ and poly‐fluoroalkyl substances (PFAS), and pesticides, among other chemical classes. In addition to phthalates, these other classes of EDCs are individually associated with negative fertility outcomes ( Land et al., 2022 ), but few studies have investigated combined exposures to multiple classes of EDCs. As mixture analyses are beginning to be conducted on a per chemical class basis, future studies must also incorporate real life, human exposures to multiple different classes of chemicals when examining the impact on women’s reproductive health.

Supplementary Table 1: Overview of studies analyzing associations of phthalate levels and effects on the HPO axis, ovary, uterus, and associated disease states with additional statistical information. Studies measuring the sum of multiple phthalate metabolites originating from a single parent phthalate compound are designated as: ∑parent phthalate (specific metabolites). Studies measuring the sum of anti-androgenic phthalate metabolites are designated as: ∑AA. Polycystic Ovary Syndrome (PCOS); Antral Follicle Count (AFC); Anti-Müllerian Hormone (AMH); Follicle Stimulating Hormone (FSH); Primary Ovarian Insufficiency (POI); In Vitro Fertilization (IVF); Decreased Ovarian Reserve (DOR); Hazard Ratio (HR); Confidence Interval (CI); Mean Shift (MS); Odds Ratio (OR); Quartile (Q); Risk Ratio (RR); Tertile (T); Logistic Regression Coefficient (β); Percent Difference (%D); Mean Change (MC); Percent Change (%C); Percent Increase (%I). Supplementary Table 2: Overview of studies analyzing associations of phthalate levels and IVF or pregnancy outcomes with additional statistical information. Studies measuring the sum of multiple phthalate metabolites originating from a single parent phthalate compound are designated as: ∑parent phthalate (specific metabolites). In Vitro Fertilization (IVF); Intracytoplasmic Sperm Injection (ICSI); Tumor Necrosis Factor-α (TNF-α); Femur Length (FL); Head Circumference (HC); Abdominal Circumference (AC); Estimated Fetal Weight (EFW); Recurrent Pregnancy Loss (RPL); Anogenital Index (AGI); Relative Risk (RR); Confidence Interval (CI); Tertile (T), Odds Ratio (OR); Quartile (Q); Bayesian Kernel Machine Regression (BKMR); Percent Change (%C); Logistic Regression Coefficient (β); Standard Deviation Change in Z-score (SDΔ); Fecundability Ratio (FR); Change in Outcome Associated with an Interquartile Range (IQR) Increase (Δ/IQR).