Impact of Bisphenol A and its alternatives on oocyte health: a scoping review.

Our food comes in contact with a multitude of materials during production, packaging, distribution, storage, and preparation. Food contact materials, such as food packaging in supermarkets, kitchenware, and reusable food and drink containers, are necessary to maintain food quality and safety. However, these materials can pose a significant contamination risk ( Barnes et al. , 2006 ; Han et al. , 2018 ), whereby active chemical components within contact materials migrate into foods via diffusion ( Barnes et al. , 2006 ; Muncke et al. , 2014 ). It is therefore imperative that these food contact materials are appropriately tested and regulated by manufacturers and government food safety bodies to ensure safety. Of growing concern, due to their established ability to impact human health, are endocrine disrupting chemicals in food packaging, such as Bisphenol A (BPA) and its chemical analogues ( Vom Saal and Vandenberg, 2021 ). BPA or 2,2-bis(4-hydroxyphenyl)propane is an industrial chemical that is widely used for the polymerization of plastics including polycarbonate reusable containers and kitchenware, epoxy resins lining metal cans, and some disposable polyvinyl chloride packaging. BPA is also used in non-food-related materials such as in thermal paper receipts, eyewear, and children’s toys ( López-Cervantes and Paseiro-Losada, 2003 ; Metz, 2016 ; Almeida et al. , 2018 ; Vilarinho et al. , 2019 ). Used by manufacturers for over 70 years, BPA is well known for its diverse endocrine disrupting properties and ability to migrate into foods ( Metz, 2016 ; Almeida et al. , 2018 ; Vilarinho et al. , 2019 ). The release of BPA as an active endocrine disruptor and its subsequent migration from plastic into foods is accelerated by exposure to common stressors including heating, microwave radiation, ultraviolet radiation, and repeated use ( Yang et al. , 2011 ; Bertoli et al. , 2015 ). As such, >90% of human BPA exposure occurs through the oral route, with absorption occurring in the gastrointestinal tract ( Vandenberg et al. , 2007 ). BPA is also a prevalent environmental contaminant, rampant within landfill and dispersed throughout ecosystems via microplastics, presenting further opportunities for contamination of food sources ( Liu et al. , 2019 ; Vom Saal and Vandenberg, 2021 ; Makowska et al. , 2022 ). Ultimately, BPA causes an extensive range of adverse health effects in animal models including metabolic, cardiac, hepatic, neurological, reproductive, and developmental pathologies. Further, higher levels of urinary BPA are correlated with many of the same pathologies in humans, evidenced in over 100 observational studies ( Vandenberg et al. , 2013a ; Rancière et al. , 2015 ; Ejaredar et al. , 2017 ; Vom Saal and Vandenberg, 2021 ). Within the past few decades, many BPA alternatives such as Bisphenol S (BPS), Bisphenol F (BPF), Bisphenol AF (BPAF), Bisphenol B (BPB), and Fluorene-9-Bisphenol (BHPF) have risen in popularity in response to consumer-driven BPA health concerns. These chemical analogues are consumed under a ‘BPA-free’ label carrying an assumption of safety, however the emerging literature on these compounds refutes these claims ( Moon, 2019 ). Illustrative of this, the popular substitute BPS has higher oral availability than BPA, providing greater opportunity for exposure ( Khmiri et al. , 2020 ). Consequently, there is a rapidly growing body of evidence in both animal and human studies showing that BPA alternatives can induce similar adverse health effects to BPA ( Rochester and Bolden, 2015 ; Pelch et al. , 2019 ). The mechanisms through which BPA exerts effects are pleiotropic and cell/tissue type specific, with their capacity to bind to estrogen receptors and other hormone receptors, disrupting hormone synthesis and epigenetic regulation ( Wetherill et al. , 2007 ; Acconcia et al. , 2015 ). In studies where the activity of different BPA analogues has been compared to BPA, many similar or more potent effects have been demonstrated. For example, BPAF and BPB have stronger agonistic effects on human estrogen receptor β than that elicited by BPA ( Kojima et al. , 2019 ). Importantly, it is well-documented that bisphenols can have a non-monotonic dose response, that is, a non-linear relationship between dose and effect, complicating the assessment of resultant adverse health effects ( Vandenberg, 2014 ; Eladak et al. , 2015 ). Causing even greater complications are the combined effects of bisphenols with each other or other endocrine disrupting chemicals, which is most reflective of real-life exposure, and results in further alterations of the dose-response dynamics ( Hamid et al. , 2021 ). Female fertility and specifically oocyte health is one aspect of human health that is particularly vulnerable to endocrine disrupting chemicals such as BPA and other bisphenols ( Hunt et al. , 2003 ). This is attributed to the finite and long-lived nature of oocytes, remaining arrested in meiosis within the ovary for many decades ( Telfer and McLaughlin, 2007 ; Inoue et al. , 2008 ; Peters et al. , 2020 ). Over this protracted period of arrest, the pre-ovulatory oocyte is susceptible to a variety of exposures from lifestyle or environmental sources that can compromise oocyte quality and function ( Muncke et al. , 2014 ; Peters et al. , 2020 ). Indeed, resultant oocyte damage can impair the oocyte’s ability to mature and fertilize optimally, impacting fertility with long-term consequences for embryo development, fetal health, pregnancy success and, ultimately, the health of future generations ( Krisher, 2013 ). In 2008, the US Food and Drug Administration (FDA) concluded the BPA no observed adverse effect level (NOAEL) as 5 mg/kg bw/day based on the findings of two multigenerational rodent studies funded by the plastic industry ( Tyl et al. , 2002 , 2008 ; FDA BPA Joint Emerging Science Working Group, 2014b ). Importantly, the NOAEL informs the tolerable daily intake (TDI), that is, the estimated amount of a contaminant set by food safety authorities that can be consumed over a lifetime without an appreciable health risk. For BPA, a TDI of 0.05 mg/kg bw/day was adopted in the USA, Australia, and New Zealand in 2010, joining Japan, South Korea, and the European Union with the same TDI regulations ( Food Standards Australia New Zealand, 2010 ; Almeida et al. , 2018 ). During this same period, there was a consumer-driven ban of BPA from baby bottles and infant formula packaging across most developed countries ( Food Standards Australia New Zealand, 2010 ; European Commission, 2011 ; Food and Drug Administration, 2012 ). More recently, in April 2023, the European Food Safety Authority (European FSA) established a newly lowered TDI of 0.2 ng/kg bw/day due to health concerns over low-level dietary BPA exposure. France has also banned BPA in almost all food contact materials since 2015 ( Eladak et al. , 2015 ; Lambré et al. , 2023 ). There is limited information available regarding the safe exposure limits of BPA alternatives worldwide. This is despite a growing number of these alternatives being identified as reproductive and developmental toxicants and suspected endocrine disruptors ( NICNAS, 2015 ; Kojima et al. , 2019 ; EFSA et al. , 2020 ). For BPS, a popular alternative to BPA, the European FSA claims BPS does not affect reproductive performance at extremely high doses up to 180 mg/kg bw/day in rodents ( EFSA et al. , 2020 ). Similarly, both the Australian National Industrial Chemicals Notification and Assessment Scheme and United States Environmental Protection Agency (US EPA) identifies the reproductive NOAEL for BPS reproductive toxicity at high doses of 60 mg/kg bw/day ( EPA. 2014 ; NICNAS, 2019 ). In line with these high NOAELs, BPA alternatives are minimally regulated in food packaging globally. Some regulations in place for popular alternatives like BPS include migration limits from packaging into food at 0.05 mg/kg in the European Union, and efforts to limit use in children’s products in three US states ( EPA. 2014 ; EFSA et al. , 2020 ). Concerningly, many of the current guidelines for ‘safe’ BPA exposure to food do not consider the effects of BPA on oocyte health and fertility. The aforementioned US FDA-led studies reported minimal primordial follicle count data and nothing at all to assess oocyte health. Furthermore, the use of ‘safer’ alternatives to BPA, including BPS, in food packaging is increasing, with little toxicological investigation or tailored restrictions in place. This lack of regulation and independent safety testing, especially given the history of BPA, is extremely concerning. Herein, we used a scoping review to synthesize the peer-reviewed literature assessing the effects of BPA and BPA alternatives on oocyte health parameters. Through this, we aimed to ascertain what is known within the field to date, to highlight what information is lacking or insufficient, and to determine whether the current restrictions in place for BPA and its alternatives are appropriate for protecting oocyte health. Ultimately, this scoping review highlights oocyte health as an important measure of fertility for toxicological studies, informs new or modified guidelines for both currently unregulated and regulated substances, and finally, provides best practice recommendations for future studies in this field.

The format of a scoping review was selected for this study in recognition that scoping reviews provide a powerful platform from which to synthesize literature from a range of study designs incorporating diverse data such as different exposures and different measures of oocyte health ( Munn et al. , 2018 ). This scoping review adheres to the PRISMA extension for scoping reviews ( Tricco et al. , 2018 ); no protocol is registered. To initially be eligible for inclusion in this review, a study had to assess the effects of a toxicant resulting from food processing or packaging on parameters related to oocyte health. Such oocyte health parameters included any direct measure of oocyte health including morphology, spindle alignment, cellular markers of health, e.g. apoptosis or autophagy, oxidative stress, inflammation, chromatin modification, capacity to mature in vitro , etc. Studies utilizing oocytes of mammalian origin, either using animal models or human samples, from all years were included. These studies included in vitro , in vivo , and human observational assessments with acute or chronic exposures of contaminant or toxicant chemicals. In addition, in vivo studies that assessed multigenerational or transgenerational effects resulting from prenatal, perinatal, or postnatal exposures were included. Studies focussing on any valid toxicant/food contaminant that is recognized as a possible risk (as determined by the US FDA, Food Standards Australia New Zealand, European FSA, or otherwise) within the western diet were included. Studies assessing toxicant impacts on non-mammalian species such as fish and insects were excluded. Studies in languages other than English were excluded. Systematic/scoping and narrative reviews, case studies, editorials, conference abstracts, and grey literature (non-academic publishing) were excluded. Four databases (Medline, Embase, Scopus, and Web of science) were searched using relevant search terms on the 23rd of February 2022. The search terms were constructed to identify a broad range of studies assessing the effects of food contaminants and toxicants on oocyte health. These were designed through consulting the US FDA, Food Standard Australia New Zealand, and European FSA legislations on food chemical safety, to target the most commonly recognized substances within the modern western diet. Search terms and operators were modified according to the database requirements. Additional sources found within reference sections of key studies or government review documents were manually examined by A.E.P to determine if they met the inclusion criteria for this review. An example of the search in Medline includes: (oocyte.mp. or Oocytes/or pre-ovulatory.mp. or germinal vesicle.mp.) AND (toxicant*.mp. or Food Contamination/or food contamin*.mp. or (food* adj3 packag*).mp. or Food Packaging/or food contact material.mp. or food processing.mp. or Plasticizers/or plastici? er*.mp. or Phthalic Acids/or phthalate.mp. or Microplastics/or microplastic*.mp. or (leachate* adj3 plastic*).mp. or bisphenol.mp. or Endocrine Disruptors/or Styrenes/or styrene*.mp. or plastic*.mp. or Plastics/or (plastic* adj3 additive*).mp. or Vinyl Chloride/or acrylamide*.mp. or chloropropanol*.mp. or 3-MCPD.mp. or glycidyl ester*.mp. or 4-methylimidazole*.mp. or ethyl carbamate*.mp. or Furans/or furan*.mp. or Polycyclic Aromatic Hydrocarbons/or heterocyclic aromatic amine.mp. or Nitrosamines/or nitrosamine*.mp.). A second search was conducted on the 1st of August 2023 to capture additional literature added to the four databases since the previous search. As this search was conducted after bisphenols were chosen as the focus of this study, the search terms were modified to only capture studies assessing the impacts of bisphenols on oocyte health. This search was also expanded to include any studies that may not have included the term ‘oocyte’ but did examine ovarian follicles in line with follicle counts being a key study parameter recorded in this review. An example of the second search in Medline includes: ((Ovary/or ovar*.mp.) AND (Ovarian Follicle/or follic*.mp.) or oocyte.mp. or Oocytes/or oocytes.mp. or pre-ovulatory.mp. or germinal vesicle.mp.) AND bisphenol*.mp. limit to dt = 2022023-20230801. After the removal of duplicates, two reviewers (A.E.P. and E.A.F.) independently conducted title and abstract screening of the search results in Covidence using the defined inclusion criteria. Any discrepancies in screening results were resolved through discussion. Studies that passed title and abstract screening were then obtained and subject to full-text review by A.E.P, and exclusions were confirmed by an independent reviewer (J.M.S.). Full texts that met this inclusion criteria and included the assessment of BPA or BPA alternatives were chosen as the focus of this study and proceeded to the data extraction stage. Data were extracted by A.E.P. who collected the following information: author/s, year of publication, bisphenol studied, study design, dose and administration of bisphenol, detected bisphenol concentration of participant samples (for human studies), test species and age, location (for human studies), sample size, primary aim, the standardized oocyte health parameters measured, and their outcomes. These standardized oocyte health parameters were a set of five parameters chosen based on the US EPA guidelines for reproductive toxicity and standard observable/morphological measures of oocyte health used within Australian IVF clinics, and internationally applicable ( EPA, 1996 ; Rienzi et al. , 2012 ). These parameters were also the most consistently reported across the included studies regardless of their primary aim. These include follicle counts, oocyte yield, oocyte meiotic capacity (ability for the oocyte to progress through meiosis up until MII arrest), morphology of the oocyte and cumulus cells, and oocyte meiotic spindle integrity. Other cellular indicators of oocyte/ovarian health that were measured in a proportion of included studies, such as oxidative stress, apoptosis, or epigenetic changes, were not data extracted as part this review. Although study quality is typically not assessed in scoping reviews ( Tricco et al. , 2018 ), aspects of study quality pertaining to toxicological study design were extracted, including the number of bisphenol concentrations/doses assessed in each study, the reporting of animal numbers within a dosing regime, and whether bisphenol delivery was oral. Extracted data was reviewed for completeness and accuracy by J.M.S. The results of this review are reported as a narrative synthesis.

Searching across four databases and additional sources initially yielded 5651 studies, with an additional 429 identified in the second ‘bisphenol only’ search ( Fig. 1 ). Following the removal of duplicates, 3147 studies remained, with title and abstract screening as per the inclusion criteria reducing the total eligible for full text review to 335. To be included in this review, a study had to assess the effects of a toxicant resulting from food processing or packaging on mammalian oocyte health parameters. At full text review, 149 studies were excluded for not appropriately addressing the criteria, with most common reasons including: no published full text associated with abstract, not assessing parameters directly related to the oocyte or follicle, or the toxicant assessed was not found to be associated with food processing or packaging. Four additional studies found in US FDA review documents were manually added at full text review. At the conclusion of full text assessment, 166 studies (along with 20 from the second search) were suitable for inclusion in this review ( Fig. 1 ). PRISMA flowchart for the selection of studies assessing dietary toxicants and bisphenols on oocyte health and female fertility. The 166 studies that initially met our inclusion criteria addressed a wide variety of toxicants arising from food processing and packaging, grouped accordingly in Fig. 2 . More than half of these studies (52.4%, 87) investigated the impact of BPA and/or alternatives of BPA on oocyte health. To enable a more comprehensive assessment of these studies, the focus of this review is on these 87 studies, plus the 20 additional studies from the second search. These studies underwent data extraction with their study design characteristics and aims being detailed in Table 1 . Pie chart presenting 166 studies included after initial full-text review, categorized by the type of toxicant they assessed. Studies that included more than one toxicant were presented in a separate category. Colour key corresponds to each category, and the number of studies in each category is represented on the chart. Summary of key features of 107 included studies. In vitro : GVs treated with 50, 100, 200, or 400 µM for 14 h In vivo : Delivered via gastric cannula in peanut oil at 2, 10, or 50 mg/kg bw/day for 12 days In vitro : PND0 ovaries treated with 10, 50, or 100 μM for 3 days In vivo : Delivered intraperitoneally to newborn mice in saline solution at 2 or 10 mg/kg bw/day for 3 days In vitro : ≥3 replicates per experiment, ovary numbers not provided In vivo : Animal numbers per treatment group not provided, ≥3 replicates per experiment In vitro : GV oocytes treated with 50, 100, 200, 400, 800 ng/ml, or 10 µg/ml for 16 h In vivo : Delivered orally in corn oil to 22 day old F1 females at 20, 40, or 100 ng/g bw/day for 7 days In vitro : ≥3 experimental replicates, 4–12 mice (∼150–450 oocytes) per treatment group in total In vivo : 8–20 animals per treatment group Human observational: median (Q1, Q3): 0.66 (0.29, 1.14) µg/l (Cr-adjusted) In vivo : Delivered orally in peanut oil to 6–8 weeks old mice at 1, 10, or 100 µg/kg bw/day for 2 weeks Human observational: normogonadotropic infertile patients undergoing IVF/ICSI, 20–40 years old, China In vivo : Kunming mice, collected at ∼8–10 weeks old Human observational: 106 participants, blood samples and two urine samples collected per cycle In vivo : six animals per treatment group Abbreviations: BADGE, Bisphenol A diglycidyl ether; BHPF, Fluorene-9-bisphenol; BPA, Bisphenol A; BPAF, Bisphenol AF; BPB, Bisphenol B; BPF, Bisphenol F; BPS, Bisphenol S; bw, body weight; COC, cumulus-oocyte complex; Cr, creatinine; DES, diethylstilbestrol; DMSO, dimethylsulfoxide; DPC, days post coitum; EARTH study, The Environment and Reproductive Health study; GV, germinal vesicle; ICSI, intracytoplasmic sperm injection; IVF, in vitro fertilisation; IVM, in vitro maturation; LOAEL, lowest observed adverse effect level; MI, metaphase I; MII, metaphase II; NOAEL, no observed adverse effect level; PCOS, polycystic ovarian syndrome; PND, postnatal day; SG, specific gravity; TDI, tolerable daily intake. Out of all 107 studies, there were 86 (80.4%) that investigated BPA and 30 (28.0%) that investigated BPA alternatives including: BPS (18), BPAF (6), BPF (5), BHPF (3), and BPB (2) ( Fig. 3 ). Nine of these studies examined BPA alongside one or more BPA alternatives, and two studies examined two BPA alternatives in combination ( Nevoral et al. , 2021 ). There was also one study that assessed BPAF alongside Bisphenol A diglycidyl ether (BADGE) ( Fig. 3 ), the resulting basic monomer of the process used to create BPA-based epoxy resins ( Abdallah et al. , 2023 ; Yue et al. , 2023b ). These studies were published between 1997 to 2023 for BPA, and 2016 to 2023 (year the most recent search was conducted) for the BPA alternatives. A modified Venn diagram representing 107 studies for final inclusion in this review after the second search and the bisphenols they assessed. Each coloured circle represents the number of studies in this review (not to scale) assessing a bisphenol, or multiple bisphenols (overlapping areas): Bisphenol A (BPA, 86), Bisphenol S (BPS, 18), Bisphenol F (BPF, 5), Bisphenol AF (BPAF, 6), Bisphenol B (BPB, 2), and Bisphenol A diglycidyl ether (BADGE, 1). The total number of studies assessing each bisphenol are provided in brackets. Key results of these 107 studies were recorded and summarized, based on the effect seen, in Fig. 4 for BPA and Fig. 5 for the BPA alternatives. Five standardized parameters of oocyte health were developed for assessment in this review based on the US EPA guidelines for reproductive toxicity and standard observable/morphological measures of oocyte health used within Australian IVF clinics, and internationally applicable ( EPA, 1996 ; Rienzi et al. , 2012 ). These included follicle counts (quantitative data), yield of oocytes collected, meiotic capacity, morphology of the oocyte or cumulus-oocyte-complex (COC), and meiotic spindle characteristics. Effects in these categories were considered adverse in this review if they significantly differed from the control group. Seven studies included in this review did not measure any of these parameters and so their findings were not included in Figs 4 and 5 . Overall, excluding these studies, 85% (67/80 or 83.8% for BPA and 26/29 or 89.7% for BPA alternatives) of studies that assessed at least one of five parameters, documented an adverse effect. Studies assessing the effect of BPA on oocyte health grouped according to study type and the oocyte health parameter that was assessed within each study. Studies that found an adverse effect in an oocyte health parameter are in orange shaded cells, and studies that found no effect are in blue shaded cells, with increasing intensity of colour representing a higher number of studies. In vivo studies indicated with an asterisk utilized an indirect exposure model, i.e. in utero or breastmilk exposure. Please note that the presence of a study in the blue shaded cells indicates no adverse effect only for the relevant outcome, not the whole study. Many studies may have reported an adverse effect for a different outcome in the figure or, alternatively, a different outcome that was not covered in this review. Studies that were listed as having no adverse effect above but which reported other adverse effects of BPA on the oocyte, ovary or female fertility outside the scope of this review include Campen et al. (2017) , Fernández et al . (2010) , Moore-Ambriz et al. (2015) , Lebachelier de la Riviere et al. (2023) , Bloom et al. (2011) , Chavarro et al. (2016) , Herez et al. (2022) , Lin et al. (2021) , Mínguez-Alarcón et al. (2016) , Shen et al. (2020) , and Yenigül et al. (2021) . Studies assessing the effect of BPA alternatives on oocyte health grouped according to study type and the oocyte health parameter that was assessed within each study. Studies that found an adverse effect in an oocyte health parameter are in orange shaded cells and studies that found no effect are in blue shaded cells. In vivo studies indicated with an asterisk utilized an indirect exposure model, i.e. in utero or breastmilk exposure. Please note that the presence of a study in the blue shaded cells indicates no adverse effect only for the relevant outcome, not the whole study. Many studies may have reported an adverse effect for a different outcome in the figure or, alternatively, a different outcome that was not covered in this review. Studies that were listed as having no adverse effect above but which reported other adverse effects of BPA on the oocyte, ovary or female fertility outside the scope of this review include Nguyen et al. (2022) , Sabry et al. (2021a ), Vignault et al. (2022) , Desmarchais et al. (2022) , Nevoral et al. (2018) , Nevoral et al. (2021) , Prokešová et al. (2020) , and Lebachelier de la Riviere et al. (2023) . Abbreviations: BHPF, fluorene-9-bisphenol; BPAF, bisphenol AF; BPB, bisphenol B; BPF, bisphenol F; BPS, bisphenol S. Of the studies assessing BPA, 37.2% (32/86) utilized in vitro methods. Five studies did not include any assessable parameters aligned to this study. The concentrations of BPA employed within these studies varied vastly, and importantly, 11 studies employed treatment concentrations at or below the in vitro equivalent of the US FDA lowest observed adverse effect level (LOAEL) for BPA of 50 mg/kg bw/day ( Vandenberg et al. , 2013a ). The extrapolation of this in vivo LOAEL to an equivalent in vitro treatment concentration presents some difficulty ( Ziv-Gal et al. , 2013 ) due to the vastly different nature of these two exposure models, however the LOAEL equivalent is estimated to be 50 ng/ml or ∼219 nM ( Wetherill et al. , 2007 ). Overall, of the 27 in vitro studies that included at least one assessable parameter, 96.3% (26) reported at least one adverse effect in response to BPA treatment, with seven (25.9%) studies reporting adverse effects below the estimated in vitro LOAEL. In contrast to BPA, there is currently no widely established LOAEL or in vitro equivalent for any BPA alternative included in this review. Of the studies captured in this review assessing BPA alternatives, 18/30 (60%) employed in vitro treatment models. One study did not include any assessable parameters aligned to this study, and of those remaining, 16/17 (94.1%) reported an adverse outcome. The BPA studies most commonly exposed COCs, isolated follicles, or whole ovaries, with only five studies exposing denuded oocytes (cumulus cells removed) to BPA ( Mohri and Yoshida, 2005 ; Eichenlaub-Ritter et al. , 2008 ; Yang et al. , 2020 ; Pan et al. , 2021 ; Ding et al. , 2022 ). These biological samples were obtained from mice (53.1%), cows (18.9%), pigs (9.4%), humans (9.4%), rats (6.3%), and sheep in one instance. Interestingly, whether denuded, within a COC, or remaining within an isolated follicle, oocytes exhibited similar adverse effects of premature meiotic arrest and deformations in spindle morphology when treated with BPA in vitro , even when exposed to doses below the LOAEL in vitro equivalent ( Lenie et al. , 2008 ; Acuña-Hernández et al. , 2018 ; Yang et al. , 2020 ; Pan et al. , 2021 ). Studies assessing BPA alternatives most frequently treated GV oocytes, MII oocytes, or COCs from mice (44.4%), cows (33.3%), pigs (16.7%), or sheep in one instance, with varying concentrations of BPA alternative. The reporting of methodological characteristics pertaining to sample size amongst in vitro studies was inconsistent. Eight (18.6%, 8/43) studies clearly reported their total utilized animal/human specimen numbers for the entirety of the study and across all experiments and/or experimental replicates. Amongst the remaining majority of studies that did not provide complete sample size information, there was variable reporting of experimental replicates. These studies described replicates as either independent (25.6%, 11/43), biological (20.9%, 9/43), or not specified (16.3%, 7/43), rendering specific biological sample size and the repetition of experiments open to reader interpretation in many cases. Eight studies (18.6%, 8/43) did not disclose any information about the replicates undertaken to generate their data, and 14% (6/43) of studies did not specify the oocyte or ovary numbers used. For the BPA studies, 50% (16) included three or more concentrations whilst only five studies included five or more. For studies assessing BPA alternatives, 72.2% (13) included at least three concentrations and two studies included at least five. Additionally, the age or reproductive status of utilized animals was not specified in 25.6% (11/43) of in vitro studies. The most reported feature of oocyte health within the studies evaluating BPA and BPA alternatives in vitro was meiotic progression in 20 and 12 studies, respectively. There were 18/20 (90.0%) and 10/12 (83.3%; including BPS, BPAF, BHPF, BPF, and BPB) studies which reported adverse effects ( Figs 4 and 5 ). All the studies assessing meiotic progression examined the in vitro maturation of pre-ovulatory oocytes except for Brieño-Enríquez et al. (2011) , who cultured fetal oocytes. The studies disclosing adverse effects detailed reduced polar body extrusion rates and variability in stage of premature arrest or atresia. Three studies in total reported no adverse effects on meiotic progression in COCs from cows and rats for BPA and BPS at doses up to 50 µM, and 0.05 mg/ml or 200 µM for BPF ( Campen et al. , 2017 ; 2018 ; Nguyen et al. , 2022 ). Overall, spindle integrity was a frequently reported parameter across the in vitro studies (12 for BPA, 11 for BPA alternatives) with all reporting adverse effects including chromosome misalignment or altered spindle dimensions due to BPA or BPA alternative exposure. All five BPA alternatives included in this review were represented in this category. Campen et al. (2018) study, covering both BPA and BPS, reported extensive spindle abnormalities in oocytes exposed to almost all concentrations of both chemicals tested; these concentrations were as low as 1 fM (10 −15 molar) and well below the NOAEL for both bisphenols. Eleven in vitro studies assessed morphology and all 11 recorded adverse morphological outcomes resulting from BPA treatment of COCs or follicles. These commonly reported reduced cumulus cell expansion and, in some cases, darkened granulosa/cumulus cells, however oocyte morphology was rarely described. Five out of seven BPA alternative studies assessing morphology reported adverse effects in COC expansion and darkened cumulus cells for BPF, BPS, and BHPF, and/or cytoplasmic oocyte abnormalities such as granular appearance for BHPF. All four studies that measured follicle numbers in cultured ovaries identified adverse effects of BPA treatment through delayed germ cell nest breakdown causing reduced primordial follicles in neonatal mouse and rat ovaries ( Zhang et al. , 2014 ; Zhao et al. , 2014 ; Zhou et al. , 2015 ; Ganesan and Keating, 2016 ). In contrast, the only study assessing BPS on mouse germ cell nest breakdown demonstrated the opposite effect whereby BPS treated ovaries in culture contained reduced numbers of germ cell cysts and increased numbers of follicles compared to the control group ( Liu et al. , 2021 ). There were 35 studies assessing BPA in vivo included in this review; 22 (62.9%) of these were carried out in mice, 10 (28.6%) in rats, and 1 each in macaques, gerbils, and sheep. As mentioned earlier, the LOAEL for BPA is 50 mg/kg bw/day; this dose informs the subsequent no observed adverse effect level (NOAEL) of 5 mg/kg bw/day, and finally the TDI, another 100 times lower at 0.05 mg/kg bw/day ( FDA BPA Joint Emerging Science Working Group, 2014b ). Of the 35 studies in this review, 30 (85.7%) in vivo BPA studies reported an adverse effect in at least one of the five assessable parameters. Each of these studies documented effects at or below the LOAEL, with a further 71.4% below the NOAEL, and 40% reporting effects at or below the TDI. There were 13 studies which measured the in vivo impacts of BPA alternatives including BPS alone (eight studies), BPS alongside BPF (one study), BPAF alongside BPA, BPB, or BADGE (one study each), and BHPF (one study). Of these 13 studies, 12 (92.3%) observed at least 1 adverse outcome, and all 9 BPS studies that documented adverse effects recorded these below the European FSA reproductive toxicity limit of 180 mg/kg bw/day ( EFSA et al. , 2020 ). The 12 in vivo models that resulted in adverse outcomes were conducted in mice or, in one case, hamsters. The only study that reported no adverse effects delivered very low levels of BPS at 4 or 50 µg/kg bw/day in the diet of primiparous sheep for at least 3 months and reported no changes in follicle counts, oocyte yield, and meiotic capacity ( Desmarchais et al. , 2022 ). Only one in vivo study compared BPA with an alternative (BPAF) and the same adverse effect of increased atretic follicles was recorded across all BPA and BPAF treated groups regardless of dose ( Jones et al. , 2018 ). We identified 16 studies (14 for BPA, 1 for BPS, 1 for BPAF and BADGE) that delivered bisphenols to mothers during gestation (perinatally), and subsequently measured the effects of in utero exposure on offspring fertility. In six cases, BPA or a combination of BPS and BPF was delivered to mothers during lactation to assess indirect exposure through breastmilk to neonates and their reproductive health outcomes. Of all the perinatal exposure studies, 92.9% (13/14) that assessed a standardized parameter of offspring oocyte health reported an adverse effect regardless of the timing, dose (at the LOAEL, NOAEL, or TDI), and duration of exposure during gestation. These adverse effects most commonly culminated in numerous changes to follicle numbers in F1 offspring (11 studies), specifically reducing primordial follicle count in seven of these studies. There were also four studies that reported effects on the meiotic spindle including spindle aberrations, chromosome misalignment, and aneuploidy in F1 generation oocytes. Similar reductions in primordial follicle count or perturbance of ovarian cyst breakdown were further observed in four instances, in the F2 generation or skipping to the F3 generation resulting from F0 exposure to low-dose BPS, BPA alone, or combined with a high fat diet or other EDCs found in plastic ( Nilsson et al. , 2012 ; Berger et al. , 2016 ; Zhang et al. , 2020b ; Huang et al. , 2022 ). Twelve of the 47 in vivo studies (25.5%) did not include sufficient details regarding the number of animals used in each treatment group within their dosing regimen. Oral delivery of BPA/BPA alternatives was employed in 68.1% (32/47) of all studies, with three or more doses delivered in 38.3% (18/47) of cases. Extending this approach to study parameters, six studies could not be analyzed for their results due to the lack of any comparable endpoints. Changes in follicle number was the most reported outcome in response to in vivo BPA and BPA alternative exposure, with 22/24 (91.7%) and 9/11 (81.1%) studies reporting an adverse effect, respectively. For BPA, 13 studies observed decreased primordial follicle numbers, indicating a depleted ovarian reserve resulting from direct or in utero exposure. Alongside this, there were two reports of increased primordial follicle numbers as a result of either mouse neonatal exposure or adult gerbil exposure followed by ageing to 18 months ( Karavan and Pepling, 2012 ; Ruiz et al. , 2023 ). In three cases, a reduction in primordial follicle numbers was identified in either the F2 or F3 generation of rats or mice from F0 mothers exposed during gestation and/or lactation ( Nilsson et al. , 2012 ; Berger et al. , 2016 ; Huang et al. , 2022 ). However, in two of these studies, the effects of BPA were observed in combination with other plastic derived endocrine disrupting chemicals or a high-fat diet in the F0 generation ( Nilsson et al. , 2012 ; Huang et al. , 2022 ). Across the total 22 studies recording changes in follicle number, 9 reported increased atretic follicle numbers in BPA-treated groups, and 4 documented increased numbers of multi-oocyte follicles, however, the quantity of other follicle types were altered in an inconsistent manner. In contrast, only two BPA alternative studies recorded a change in primordial follicle counts. These studies presented opposing changes of decreased primordial follicle numbers resulting from BPAF exposure of adult mice ( Yue et al. , 2023b ) or increased primordial follicle numbers in F1 offspring of BPS exposed pregnant mice for four days mid-gestation, however, these effects were not maintained in the F2 offspring ( Zhang et al. , 2020b ). Notably, five studies observed an increase in atretic follicle numbers due to oral delivery of BPS, BHPF, BPAF, or BPB to sexually mature mice or hamsters, or in utero F1 exposed mouse offspring. Similar to BPA, the BPA alternative studies reported a number of varying observations in quantities of other follicle stages. The integrity of oocyte spindles was found to be compromised in six out of seven in vivo BPA studies, and in all five in vivo BPA alternative studies (four BPS, one BPF, one BPAF + BADGE). BPA exposures below the LOAEL in a variety of models led to increased rates of aberrant spindles and/or misaligned chromosomes in three studies ( Hunt et al. , 2003 ; Muhlhauser et al. , 2009 ; Zhang et al. , 2017 ), as well as abnormal MI spindles ( Chao et al. , 2012 ), aneuploidy ( Susiarjo et al. , 2007 ), and premature centrosome separation ( Pacchierotti et al. , 2008 ). Spindle alignment was impacted in all studies by low dose oral BPS and BPF exposure in mice for durations of between 4 days and 4 weeks. Three studies also documented chromosomal abnormalities (misalignment, aneuploidy) resulting from BPS, BPAF, or BADGE exposure in F1 offspring of exposed mothers or in neonates feeding from exposed mothers ( Zhang et al. , 2020b ; Nevoral et al. , 2021 ; Abdallah et al. , 2023 ). Two out of four studies investigating in vitro meiotic progression of pre-ovulatory oocytes retrieved from young mice exposed to BPA orally or hypodermically revealed reduced germinal vesicle breakdown rates under the NOAEL ( Chao et al. , 2012 ) or reduced polar body extrusion rates under the LOAEL ( Zhang et al. , 2017 ). Likewise, Zhang et al. (2020) observed decreased rates of both germinal vesicle breakdown and polar body extrusion in in vitro matured oocytes retrieved from F1 mice exposed to low-dose BPS in utero . Three other studies assessed the meiotic progression of oocytes retrieved from sheep or mice orally exposed to BPS and BPF at or below the TDI (0.05 mg/kg bw/day). These three studies reported no changes in the in vitro maturation rates, regardless of the exposure model (acute, chronic, or indirect neonatal exposure through breast milk) or the presence or absence of cumulus cells ( Prokešová et al. , 2020 ; Nevoral et al. , 2021 ; Desmarchais et al. , 2022 ). All four in vivo BPA studies that reported on follicle and/or oocyte morphology observed a range of adverse effects. These included cystic dilation of follicles with decreased sparse granulosa cells ( Lin et al. , 2021 ), increased oocyte diameter ( Chao et al. , 2012 ), presence of granules in the oocyte cytoplasm ( Liu et al. , 2017 ), and decreased zona pellucida thickness, oocyte volume, and nuclear volume in antral follicles ( Mansoori et al. , 2013 ). In three out of the four studies, sexually mature mice or rats were exposed to BPA orally. We identified a single study examining the effects of oral 300 µg/kg bw/day BPS on the uterus and ovary of adult mice. This study recorded unclear oocyte structure within follicles, compromised zona pellucida appearance, and atrophy of immature follicles ( Yue et al. , 2023a ). Only one out of six studies recorded a reduced yield of healthy oocytes due to 10 mg/kg bw/day BPA exposure for 42 days in mice ( Hu et al. , 2018 ). Similarly, one out of five BPA alternative studies documented a decrease in mean oocyte yield with increasing BPS dose for five doses up to 100 µg/kg bw/day for 21 days in mice ( Nourian et al. , 2017 ). There were 21 observational studies that assessed BPA levels in patients attending fertility clinics; one of these studies assessed BPS, BPF, and BPAF alongside BPA ( Lebachelier de la Riviere et al. , 2023 ). These studies collected urine, blood, and/or follicular fluid from patients to measure the amount of BPA present and correlated this with patient outcomes. The parameters of interest recorded in these studies were oocyte yield and antral follicle count. There were four studies that examined these correlations in the context of additional lifestyle factors including folate intake, soy intake, plastic container use, and bottled water intake ( Chavarro et al. , 2016 ; Mínguez-Alarcón et al. , 2016 ; Aftabsavad et al. , 2021 ; Yenigül et al. , 2021 ). No studies reported on in vitro oocyte maturation, recorded descriptive morphological data on the oocytes collected, or examined the spindle integrity of MII oocytes. Participant numbers across the 21 studies ranged from 44 to 450, with an age range of 18–46.7 years old. All 10 studies conducted in the US, one Korean, and one Greek study documented a mean or median participant age between 35 and 36 years of age, except for a mean age of 38 for the Greek study participants with tubal factor infertility (mean of total participants not provided) ( Mina et al. , 2022 ). Studies from other countries including China, Iran, Iraq, Poland, France, and Turkey had younger participants with a mean or median age between 27 and 33, however one Iranian study did not provide statistical data on patient age ( Poormoosavi et al. , 2019 ). In 76.2% (16/21) of studies, patient populations were undergoing autologous IVF or ICSI cycles for a variety of infertility factors including female factor, male factor, combined, and unexplained infertility. In addition to these studies, Zhou et al. (2017) only assessed patients with polycystic ovarian syndrome (PCOS), Shen et al. (2020) assessed those with tubal factor infertility, and Mina et al. (2022) studied participants with either of these diagnoses. Two studies only included patients undergoing ICSI that had either an unexpected poor ovarian response ( Aftabsavad et al. , 2021 ) or unexplained infertility ( Yenigül et al. , 2021 ). Eight of the 21 studies were undertaken as part of the Environment and Reproductive Health (EARTH) study, an ongoing prospective cohort study carried out in Massachusetts General Hospital Fertility Center between 2004 and 2012 to evaluate environmental and dietary determinants of fertility. Overall, 52.4% (11/21) of the clinical studies recorded at least one adverse effect. Of the 18 studies that reported on oocyte yield, nine (50%) documented reduced total and/or mature (MII) oocyte yield associated with higher levels of BPA in urine, serum, or follicular fluid. Conversely, eight studies conducted antral follicle counts of participants and three of these recorded a negative association between increasing BPA concentration and antral follicle numbers ( Souter et al. , 2013 ; Zhou et al. , 2017 ), though one of these studies recorded a P value of 0.051 for this parameter ( Aftabsavad et al. , 2021 ). Adverse effects were reported in some of the studies assessing specific subsets of patient populations seeking ART procedures in 6 out of 11 instances, including those undergoing ICSI ( Fujimoto et al. , 2011 ), those with tubal factor infertility ( Shen et al. , 2020 ; Mina et al. , 2022 ), those with poor response to ovarian stimulation ( Aftabsavad et al. , 2021 ), or those with PCOS ( Zhou et al. , 2017 ). The studies that revealed associations between oocyte or follicle parameters with BPA were performed on patient cohorts from China (three studies), USA (three studies), Iran (two studies), Greece (one study), and Poland (one study). There were an equal number of studies (9/18, 50%) that reported no correlation between BPA concentrations measured in patient samples and oocyte yield, and five studies that failed to record a correlation with antral follicle count. These studies were primarily based on patient cohorts from the USA (six studies), alongside four studies from Korea, Iraq, France, and Turkey. Five studies assessed oocyte yield and antral follicle count together; of these, two reported no changes in both parameters ( Bloom et al. , 2011 ; Yenigül et al. , 2021 ), two studies documented an association with oocyte yield but not antral follicle count ( Shen et al. , 2020 ; Lin et al. , 2021 ), and one revealed associations with both parameters ( Aftabsavad et al. , 2021 ). Despite Bloom et al. (2011) reporting no correlations between BPA concentrations and total oocyte yield or follicle counts, another study using the same patient cohort by Fujimoto et al. (2011) reported negative correlations between serum BPA concentration and mature MII oocyte yield in nine Asian women undergoing ICSI. A recent study from France was the first to assess conjugated BPS, BPF, and BPAF (alongside BPA) in follicular fluid and correlated these levels with oocyte yield. This study did not observe a correlation with any bisphenol, however they only examined this by grouping participant samples based on whether bisphenols were detectable or undetectable and not by concentration ( Lebachelier de la Riviere et al. , 2023 ). The levels of detected BPA in patient samples varied greatly from study-to-study, even amongst studies from the same country with sample BPA concentrations below the limit of detection (commonly 0.1–0.4 µg/l for urine) ranging from 0% up to 56.8% ( Kim et al. , 2021 ) and 88.1% ( Lebachelier de la Riviere et al. , 2023 ). Additionally, for the BPA alternatives, 87% (BPS) and 98.6% (BPF) of samples were below the limit of detection, and BPAF was undetectable across all samples ( Lebachelier de la Riviere et al. , 2023 ). The median adjusted urinary BPA concentrations ranged from 0.66 to 2.35 µg/l, and the maximum detected BPA in urine ranged from 10.45 to 79.19 µg/l ( Fig. 6 ). Only three out of five studies that assessed serum/plasma, and two out of seven studies that assessed follicular fluid, reported median BPA concentrations; these were 0.012–2.53 µg/l and 0.063–1.13 µg/l, respectively ( Bloom et al. , 2011 ; Kim et al. , 2021 ; Mina et al. , 2022 ). Box and whisker plot of 15 human observational studies ranked by participant number, displaying urinary, serum, or follicular fluid BPA concentrations of the participants. BPA concentration data presented includes the median, quartile 1, quartile 3, the minimum detected concentration or limit of detection, and maximum detected concentration where available. Orange indicates that the study found an adverse effect associated with BPA concentration, blue indicates the absence of an adverse effect. Six studies did not provide adequate information on the bisphenol concentrations of participant samples and were not included in this figure. Overall, only 7 out of 21 studies (33%) provided a full range of standard statistical parameters on BPA levels in their study population including quartiles or percentiles, means, standard deviations, maximum concentrations, the limit of detection, and the percentage of the population that fell below this limit. A further eight studies provided adequate information missing one or two of these details, and six studies provided very minimal, insufficient data on BPA concentrations across their participants and as such are not included in Fig. 6 .

Taken together, the available in vitro and in vivo evidence demonstrates that BPA has detrimental effects on oocyte health at low levels, that are well below the US FDA LOAEL and NOAEL for oral BPA exposure. This review has highlighted that BPA can impact follicle development, meiotic progression of the oocyte, the morphology of the oocyte or cumulus-oocyte-complex, as well as the integrity of the meiotic spindle, all of which are crucial components capable of contributing to poor oocyte quality and reduced fertility if disrupted. Concerningly, ‘BPA free’ alternatives that carry a public perception of safety can elicit the same spectrum of effects upon oocyte health. There is thus an urgent need for the revision of current guidelines by food safety authorities globally, following in the footsteps of the European FSA, to reduce prescribed safe exposure levels of BPA and introduce clear restrictions for the use of BPA analogues in line with the recommendations we have outlined. Ultimately, this scoping review highlights oocyte health as a fundamentally important endpoint in female reproductive toxicological studies, indicating an important direction for future research into endocrine disrupting chemicals to improve fertility outcomes.

This scoping review was designed to determine the extent of knowledge on BPA and BPA alternatives to date, and the effects of these chemicals on oocyte health. Additionally, we sought to provide insight into current regulations, or lack thereof, for BPA and its analogues established by the US FDA and mirrored across many countries, and the suitability of these for oocyte health and female fertility. Specific endpoints that measure oocyte health and related ovarian parameters such as those documented in this review have been neglected within US FDA guideline studies and in the CLARITY-BPA program, and are generally understudied compared to other endpoints ( Tyl et al. , 2002 , 2008 ; FDA BPA Joint Emerging Science Working Group, 2014b ; Camacho et al. , 2019 ). It is of great importance for human health that the quality of the oocyte (a cell capable of governing reproductive success, pregnancy success, and offspring health) is considered by food safety regulatory bodies worldwide when deciding safe exposure of toxicants within our diet. Importantly, this review highlights only one example of the potential thousands of endocrine disrupting chemicals (EDCs) humans are exposed to through diet or within the environment daily that remain largely untested and unregulated ( Futran Fuhrman et al. , 2015 ). In preparing this review, it was found that 85% (85/100) of studies that assessed at least one of five oocyte health parameters (including follicle counts, oocyte yield, oocyte meiotic capacity, oocyte/follicle morphology, and spindle integrity) recorded an adverse effect. Overall, all bisphenol analogues examined in these studies displayed similar patterns in terms of the impact they exerted on the oocyte in vitro and in vivo . In vitro , the most frequently affected parameters of oocyte health were meiotic progression, spindle integrity, and oocyte/COC morphology. In vivo , resulting from direct or in utero exposure, the most affected parameters were follicle counts followed by spindle integrity for BPA and its alternatives. Adverse outcomes in oocyte spindle integrity, including misaligned chromosomes and abnormal spindle morphology, were conserved between the two exposure models. Patterns in adverse effects that have emerged from the synthesis of data in this review provide insight into the mechanisms and extent of damage BPA can cause to the oocyte. These data also provide concerning evidence that the analogues of BPA can elicit the same detrimental effects, however the field would greatly benefit from further studies of these alternatives incorporating additional endpoint analyses. In this context, we noted that the majority of the in vivo studies that assessed a BPA alternative focused solely on BPS (8/13), so more in vivo studies assessing the range of other BPA alternatives are warranted. The evidence for adverse outcomes associated with BPA exposure gleaned from human observational studies revealed different trends. For instance, the studies that assessed follicle counts were only able to quantify antral follicles via ultrasound, where three out of eight studies reported an associated decrease in antral follicles with BPA exposure. The outcomes of these human studies cannot be directly compared to the other study types in this review where a broader range of follicle types could be quantified, and the most common effect was reduced primordial follicles. Decreased oocyte yield, the most documented parameter in human studies, was associated with increased urinary BPA in half of the studies. The human oocytes retrieved were largely of unreported quality with respect to both meiotic spindle integrity and morphology. Aftabsavad et al. (2021) did report decreased ‘oocyte quality’ associated with higher follicular fluid BPA concentration, however the authors provided no indication of how oocyte quality was assessed in their study design. In contrast, most in vivo animal studies found no association between BPA exposure and oocyte yield, however, five studies that documented unchanged oocyte yield subsequently recorded that these oocytes had increased chromosome abnormalities or spindle alignment issues after BPA, BPS, or BPF exposure. Utilizing discarded human oocytes from patients who provided urine samples would enable assessment of meiotic spindle defects or oocyte morphology providing valuable information on the impacts of bisphenols on human oocyte quality. Alternatively, parameters of embryo development and quality could be used as an indicator of oocyte quality; this was included in a number of human studies ( Mínguez-Alarcón et al. , 2016 ; Radwan et al. , 2020 ; Shen et al. , 2020 ), however, embryo outcomes were not a parameter chosen for assessment in this scoping review. It should therefore be emphasized that the strength of evidence presented by the oocyte-specific parameters herein does not reflect or predict potential adverse embryo or pregnancy outcomes reported in these clinical studies. The in vitro studies considered in this review were primarily mechanistic, being designed with the aim of investigating and understanding how bisphenols alter oocyte or follicle health parameters. As such, the bisphenol concentrations used in these studies are generally not reflective of actual exposure levels in human populations. In vitro , the metabolic processing that alters bisphenol bioactivity is removed ( Wetherill et al. , 2007 ; Thayer et al. , 2015 ) and chronic exposure is simply not possible to imitate, particularly when culturing oocytes. Despite these limitations, seven BPA studies recorded low dose adverse effects below the equivalent in vitro LOAEL of 50 ng/ml. It is clear from the in vitro data summarized in this review that the oocyte is susceptible to damage from bisphenol exposure at very low concentrations in vitro . The mechanistic investigations of in vitro studies, despite not being the focus of this review, give valuable insights into the reasons underlying the profound impacts of bisphenol exposure on oocyte health. Bisphenols can act directly on receptors in an agonistic or antagonistic manner, interfering with receptor signalling pathways and hormone synthesis. It has been demonstrated in multiple instances that BPA and its analogues including BPS, BPB, BPF, and BPAF have comparable or superior estrogenic agonist potencies on human estrogen receptors α and β ( Kojima et al. , 2019 ; Durcik et al. , 2022 ). Many bisphenols also possess similarities in their effects on the androgen receptor, pregnane X receptor, constitutive androstane receptor, and the glucocorticoid receptor ( Kojima et al. , 2019 ). This functional comparability in the actions of BPA and BPA alternatives provides an explanation for the similarities in their effects on the oocyte and follicle and emphasizes the urgency of revising their safety. In addition to these direct impacts on receptors, bisphenols can also have indirect effects resulting in induction of apoptosis, oxidative stress, and inflammation in many components of the female reproductive system including cells within the ovary ( Pan et al. , 2024 ; Wang et al. , 2024 ). Oxidative stress, and in some cases, apoptosis, was induced by BPA or BPA alternative exposure in numerous in vitro studies that recorded standardized oocyte health parameters reviewed herein. These resultant impacts of increased oxidative stress and dysregulated apoptosis on cultured neonatal ovaries ( Zhang et al. , 2014 ; Zhou et al. , 2015 ), pre-antral follicles ( Li et al. , 2022 ), cumulus-oocyte complexes ( Park et al. , 2018 ; Li and Zhao, 2019 ; Jiao et al. , 2020 ), and oocytes ( Ding et al. , 2017 , 2022 ) are key explanatory factors contributing to the understanding of how bisphenols have detrimental impacts on follicle and oocyte development/maturation and on oocyte quality. A characteristic observed across some studies on bisphenols was non-monotonicity; that is, a non-linear relationship between dose and effect, that can often be ‘U’ shaped ( Vandenberg, 2014 ; Eladak et al. , 2015 ). This phenomenon has been observed far beyond EDCs and toxicological contexts, in pharmacological and even biological responses to natural hormones and hormonal drugs ( Calabrese and Baldwin, 2001 ; Vandenberg et al. , 2012 ). Indeed, some in vitro studies reviewed here displayed non-linear dose responses to BPA and its analogues, whereby low concentrations produced more severe adverse effects than that associated with higher concentrations ( Acuña-Hernández et al. , 2018 ; Campen et al. , 2018 ). In the case of non-monotonicity, it can be argued that there is no safe low dose, as low-dose effects can be just as harmful, if not more harmful, than higher doses. Dose response dynamics are further complicated by concurrent exposure to other EDCs that can produce synergistic, additive, or antagonistic effects on various receptors and hormone axes when combined ( Hamid et al. , 2021 ). Compounds with endocrine disrupting properties are greatly abundant in food and daily life, including phytoestrogens such as genistein, other chemical components of plastic such as phthalates, and pharmaceutical drugs such as antibiotics. Taken together, these factors create a challenge for BPA regulation by governing bodies such as the US FDA who hold BPA to the standard of a linear dose-response, leading to the dismissal of published data that demonstrates non-monotonic effects ( Vom Saal and Vandenberg, 2021 ). For change in the regulation of bisphenols to occur, a paradigm shift in the understanding of bisphenol mechanisms led by in vitro studies is required. Accordingly, the performance of additional studies is recommended to increase our understanding of the mechanisms by which bisphenols act on the oocyte alone and in combination with other EDCs to drive non-monotonic dose responses. The endocrine disrupting nature of bisphenols leads to many of their impacts being sex-specific. While this review addresses those impacts on the female reproductive system, more specifically the oocyte and follicle, it should remain of high importance that bisphenols can impact all aspects of human reproduction, including male fertility. Valuable in vitro and in vivo studies have revealed that bisphenols can disturb the function of testicular mitochondria, dysregulate testicular hormone production by affecting Leydig and Sertoli cells, interfere with spermatogenesis, affect sperm leading to reduced embryo cleavage and blastocyst formation, and alter testis morphology ( Adegoke et al. , 2020 ; Cariati et al. , 2020 ; Li et al. , 2020 ; Ryu et al. , 2023 ). These impacts are underpinned by similar underlying mechanisms as those that affect the female reproductive system such as interfering with receptor signalling, increased oxidative stress and apoptosis, and induction of epigenetic modifications. The consideration of the sex-specific impacts of bisphenol exposure together provides added insight into the mechanisms responsible and should be employed in human risk assessments as these effects ultimately culminate in same outcomes of compromised reproduction/fertility and offspring health. The majority (33/35, 94.3%) of in vivo studies reviewed herein were performed in rodent models. The relevance of rodent models to human risk assessment of BPA exposure has been trivialized by regulatory bodies such as the FDA, maintaining the perspective that rodent toxicity data is not highly relevant to humans. This stance was generated by the findings of two studies performed by Völkel et al. (2002) in a small number of human volunteers, concluding that orally ingested BPA is efficiently converted to its less active form, BPA glucuronide, via first-pass metabolism in a complete manner before it is rapidly excreted in urine resulting in minimal burden to the human body. It was also highlighted that this process occurs rapidly in humans compared to rats, indicative of major species differences and that highly active unconjugated BPA detected in human biological samples is likely due to contamination ( Völkel et al. , 2008 ). These findings and subsequent studies led by the FDA have been used to defend decisions to maintain the current BPA TDI, insist that human unconjugated BPA exposure is negligible, and dismiss low-dose exposure studies using rodent models. There are, however, several human biomonitoring studies that provide opposing evidence, demonstrating that glucuronidation is not a complete process in humans for BPA or its alternatives ( Thayer et al. , 2015 ; Karrer et al. , 2018 ). Indeed, there is also evidence suggesting high similarity in the unconjugated bioavailability of BPA in adult humans, rhesus monkeys, rats, and mice despite differences in the routes of elimination ( Taylor et al. , 2011 ; Thayer et al. , 2015 ). Additionally, there are other important factors that must be considered when modelling human exposure and comparing relevance of model species. One of these factors is developmental maturity, as neonatal glucuronidation capacity, or the ability for the neonate to efficiently metabolize bisphenols, is not yet fully functional, carrying major implications for fetal and neonatal exposure ( Ginsberg and Rice, 2009 ; Karrer et al. , 2018 ). Another key factor is the route of exposure, as different bisphenol exposure routes lead to variations in the way it is metabolized. For example, transdermally absorbed BPA has a longer half-life than orally absorbed BPA, as it is not subject to first-pass metabolism ( Vandenberg et al. , 2013b ; Sasso et al. , 2020 ). Taken together, it is evident that bisphenol pharmacokinetics have many intricacies, and whilst species specific differences occur, these are not significant enough to disregard low-dose rodent exposure models. The in vivo rodent data in this review provides a wealth of valuable information, however like all scientific findings, this should be interpreted alongside data from other species and study designs. Due to the aforementioned metabolic differences in the developing neonate, bisphenol exposure during gestation and neonatal development is of high concern. Our review identified 13 studies that delivered bisphenols to mothers during gestation (perinatally) and reported an adverse effect in a standardized oocyte health parameter in offspring. These adverse effects were overwhelmingly observed as changes to follicle numbers in F1 generation ovaries, commonly reducing primordial follicles, and dysregulating oocyte spindle alignment, indicative of detrimental impacts to follicle formation and oogenesis. Disturbances to the ovary, again, most often reduced primordial follicles, were further observed in the F2 generation or skipping to the F3 generation resulting from F0 exposure to bisphenols alone or in combination with other insults, demonstrating the sensitivity of fetal germ cells to bisphenols and the transgenerational implications of bisphenol exposure on fertility ( Nilsson et al. , 2012 ; Berger et al. , 2016 ; Zhang et al. , 2020b ; Huang et al. , 2022 ). A mechanism through which BPA exposure has been implicated in causing multigenerational and transgenerational effects on hormonal regulation and fertility outcomes is epigenetic modification. Epigenetic modifications are chemical changes to chromatin including DNA methylation, histone modification, and modulation of non-coding RNA. These processes result in the alteration of gene expression without causing changes to the DNA sequence, and importantly, are heritable ( Santangeli et al. , 2017 ). Several studies captured in this review also identified epigenetic changes in oocytes exposed to bisphenols in vitro , in vivo , and in F1 offspring ( in utero ). These included overall methylation changes to DNA, including imprinted genes, and differential methylation of histones ( Trapphoff et al. , 2013 ; Wang et al. , 2016 ; Prokešová et al. , 2020 ; Zhang et al. , 2020a , b ). This evidence of multigenerational and transgenerational effects alongside epigenetic disturbance in the ovary and oocyte could manifest in hugely detrimental long-term impacts on future generations, and urgently warrants further investigation. A large proportion of the reviewed in vivo studies assessed and reported ‘low dose’ effects, that is, effects at doses below the BPA NOAEL of 5 mg/kg bw/day and reflective of human exposure. Equivalently low doses of BPA alternatives were employed yielding similar impacts, despite their supposed low reproductive toxicity reported by regulatory organizations, e.g. European BPS NOAEL is 180 mg/kg bw/day ( EFSA et al. , 2020 ). The US FDA conducted three reviews between 2009 and 2013 in which they evaluated available ‘low dose’ literature assessing outcomes of BPA below the NOAEL of 5 mg/kg bw/day ( FDA BPA Joint Emerging Science Working Group, 2011 , 2013 , 2014a ). Eight in vivo studies included in this scoping review are evaluated in these US FDA reviews ( Adewale et al. , 2009 ; Fernández et al. , 2010 ; Rodriguez et al. , 2010 ; Rivera et al. , 2011 ; Chao et al. , 2012 ; Hunt et al. , 2012 ; Karavan and Pepling, 2012 ; Signorile et al. , 2012 ). Despite all eight studies reporting adverse effects, the US FDA dismissed the results of seven of these studies due to a lack of clarity in the reporting of the experimental design and methods, assessing less than three doses or less than 10 replicates, criticism of experimental design, and/or lack of ‘meaningful’ endpoints. The current review identified an additional 18 studies on BPA that have observed low-dose adverse effects in vivo , many of which have been published in the subsequent 10 years since the release of the latest US FDA review. Alongside these US FDA reviews, the US government initiated a series of studies with independent academic collaborators termed the CLARITY-BPA program to corroborate findings from regulatory guideline studies and academic experimental studies. Beginning in 2012, this was collectively the largest animal study conducted on BPA and its long-term effects ( Camacho et al. , 2019 ). This was monumental for BPA research, and reported adverse effects in various organs, including the uterus and ovary, at doses well under the NOAEL. The evidence from this study drove academic collaborators to recommend tighter restrictions on BPA including a new TDI of 2.5 ng/kg bw/day, a mere 0.005% of the current TDI. This recommendation was controversially rejected by the US FDA based on the assumption that human BPA exposure is negligible ( Vom Saal and Vandenberg, 2021 ). In contrast, the European FSA 2023 review of BPA concluded that it was likely that BPA caused female reproductive toxicity based on the animal evidence, some of which is again included in this review ( Moore-Ambriz et al. , 2015 ; Berger et al. , 2016 ; Santamaría et al. , 2016 ; Lambré et al. , 2023 ). More specifically, they concluded there are likely effects on ovarian histology after developmental and adult exposure to BPA, and effects on follicle counts after adult exposure. These conclusions are echoed in the present review. Based on this European FSA review, a new TDI of 0.2 ng/kg bw/day was established for Europe, with concerns for the health of consumers with average and high exposure to BPA exceeding this TDI ( Lambré et al. , 2023 ). Based on the balance of evidence from in vivo studies considered in this review in tandem with the European FSA conclusions, it appears timely to recommend that the current BPA NOAEL, risk assessment strategies, and risk management strategies should be reconsidered by organizations such as the US FDA. Further supporting this recommendation, bisphenols and more broadly EDCs, pose a number of challenges in the context of traditional risk assessment of chemical contaminants. Some of these challenges include chronic human exposure occurring in combination with other EDCs, the unknown combinatorial effects of EDC mixtures, latent transgenerational effects, non-monotonicity, and effects caused by EDCs not always being categorically adverse. Moreover, the definition of an ‘adverse effect’ is not universally defined or clear within every context and there appears to be little consensus on what endpoints are most suitable in EDC toxicological studies ( Futran Fuhrman et al. , 2015 ). This review identifies strong evidence trends from in vivo animal studies that substantiate the adverse effects of BPA, and emerging evidence for its analogues, on oocyte quality in low dose contexts. Taken together, these factors encourage us to closely evaluate the way in which study results are interpreted, what is classified as an adverse effect for the oocyte, and the very definition of reproductive toxicity in an in vivo context amongst food safety governing bodies internationally. Despite the measurement of urine, serum, and follicular fluid BPA concentrations in the studies included in this review and within the wider population, urine sample data was the most robust, frequently sampled, and reported in detail, so will be the primary focus of the following discussion. Due to the large inconsistencies in reporting of BPA concentrations in follicular fluid in the included studies, and the inability to compare these follicular fluid data to the wider healthy human population, these data will not be discussed. The median concentrations of urinary BPA observed in the included observational studies was 0.66–2.35 µg/l (or µg/g of creatinine), which is comparable to concentrations in general populations across different continents including Asia, North America, and Europe where the median creatinine-adjusted concentration is 1.36–2.41 µg/g ( Zhang et al. , 2011 ; Sarigiannis et al. , 2016 ; Colorado-Yohar et al. , 2021 ). A 2018 study used available global urinary BPA data to back-calculate the estimated daily intake of BPA for adults and children ( Huang et al. , 2018 ). Global average daily intakes of BPA were calculated to be 38.78 ng/kg bw/day in adults and 51.74 ng/kg bw/day in children; exposures that exceed the European FSA recommended TDI of 0.2 ng/kg bw/day ( Lambré et al. , 2023 ). Other studies that have used median human blood/serum concentrations to calculate predicted exposure have however obtained much higher values, up to 500 µg/kg bw/day ( Vandenberg et al. , 2007 ). Once ingested, BPA is metabolized and eliminated from the body via conjugation enzyme systems in the liver and gastrointestinal tract. Once converted to its conjugated form, most commonly BPA glucuronide, BPA is considered less biologically active and hence less of a health risk, however the more biologically active unconjugated or ‘free’ BPA has still been detected, albeit to a lesser extent, in urine and blood/serum ( Vandenberg et al. , 2010 ; Thayer et al. , 2015 ). This detection of unconjugated BPA in urine importantly aligns with literature reporting the incomplete first pass metabolism of BPA in humans, rhesus monkeys, and mice, thus prolonging exposure to the harmful, more biologically active BPA form ( Thayer et al. , 2015 ; Taylor et al. , 2011 ). The studies included in this review primarily reported total BPA except for Lebachelier de la Riviere et al. (2023) who measured conjugated bisphenols in follicular fluid, and Bloom et al. (2011) who analysed unconjugated serum BPA. There are benefits and disadvantages for the use of different bodily fluids to measure BPA; for example, phenols were found to be more stable within blood/serum compared to urine samples during storage, however BPA is more rapidly metabolized and cleared from the human circulation ( Völkel et al. , 2002 ; Vandenberg et al. , 2010 ). Despite their differences, a meta-analysis of BPA concentrations in adult human populations from 14 studies found that equivalent BPA concentrations in urine and blood/serum were similar, including in unadjusted or creatinine-adjusted measures ( Colorado-Yohar et al. , 2021 ). Due to the less invasive nature of collection, urine is the most common sample collected to measure human BPA exposure, and this is reflected in the studies within this review. The most prominent issue that remains for BPA detection in urine is that a single urine sample only reflects short-term exposure. The half-life of BPA in the human body is variably documented, however orally administered BPA can be rapidly absorbed, detected in the blood within 15 min, and excreted via urine within 24 h of exposure in a non-pregnant adult human ( Thayer et al. , 2015 ). Despite its ability to be rapidly excreted, BPA also possesses moderately lipophilic properties and as such can accumulate in human tissues. It has been detected within a variety of human tissues, particularly adipose tissue, placenta, liver, and fetal tissue in addition to fluids such as breast milk and amniotic fluid ( Corrales et al. , 2015 ). Yet regarding urine detection, the short-lived nature of BPA metabolism and excretion means single spot urine samples provide little insight into long-term exposure, particularly given that exposures are also often episodic and vary over time ( Braun et al. , 2011 ). The human studies considered in this review commonly collected one or two urine samples per IVF cycle, detected BPA via isotope dilution high-performance liquid chromatography (HPLC) coupled with tandem mass spectrometry, and normalized urinary concentrations to creatinine or specific gravity (the ratio of the relative density of urine to water ( Muscat et al. , 2011 )) to correct for urine concentration. The geometric mean of urine concentrations was used if there was more than one sample per IVF treatment cycle, generating a single urinary BPA concentration per cycle. This BPA concentration was then statistically correlated with various outcomes of the IVF cycle in a cross-sectional or prospective context. This approach has the inherent limitation that BPA exposure may not be accurately represented owing to variations in exposure between days, time of day, and whether urine was collected during fasting ( Ye et al. , 2008 ), all of which were factors that differed among these studies. Similar concerns regarding timing of urine sampling and using the geometric mean, obscuring episodic data from individual urine samples, were raised by the US FDA when reviewing the Mok-lin et al. (2010) study ( FDA BPA Joint Emerging Science Working Group, 2011 ). In addition, indirect methods of sample preparation involving enzymatic deconjugation of the sample to hydrolyze conjugated BPA into free BPA prior to measurement via HPLC mass spectrometry have been demonstrated to underestimate BPA levels ( Gerona et al. , 2020 ). To ensure the accuracy of BPA measurement in urine samples, direct quantification of BPA in its conjugated and unconjugated form is essential. There were eight studies in this review that formed part of the EARTH prospective cohort study in Massachusetts, constituting almost half of the human clinical studies included. This raises concerns of bias within the findings of this scoping review due to overrepresentation of one demographic and the potential overlap of data and patients between these studies. With the exception of the EARTH cohort studies, studies that found adverse effects versus those that did not were conducted in different countries. However, studies comparing the BPA concentrations in populations across different continents including Asia, North America, and Europe, have demonstrated that BPA exposure is widespread and relatively comparable (median creatinine-adjusted concentration 1.36–2.41 µg/g) ( Zhang et al. , 2011 ; Sarigiannis et al. , 2016 ; Colorado-Yohar et al. , 2021 ). Despite literature documenting the impact of BPA alternatives on human health being much more scarce, emerging data from China, Sweden and Spain has revealed lower median urine concentrations (creatinine adjusted) of around 0.04–0.15 µg/g for BPF and 0.05–0.08 µg/g for BPS, compared to that of BPA ( Derakhshan et al. , 2019 ; Sanchis et al. , 2020 ; Cui et al. , 2021 ). The one study measuring BPA alternatives reviewed herein tested follicular fluid rather than urine and documented average conjugated concentrations of 0.21 µg/l for BPS and 0.13 µg/l for BPF. These concentrations were higher than the average detected BPA concentration (0.075 µg/l); this is likely due to BPA being banned in food contact materials within France, however all bisphenols in this study had a low detection rate amongst participants ( Lebachelier de la Riviere et al. , 2023 ). BPA alternatives require much more characterization, and conclusions about their safety should be approached with caution as their toxicokinetics are not well understood. For example, in a sheep pregnancy model BPS was found to reach higher concentrations than BPA in the maternal circulation and have a longer half-life than BPA in the fetal compartment ( Gingrich et al. , 2019 ). There is a pressing need for future studies in the fertility clinic setting to include a spectrum of bisphenols in their detection protocols to assess the prevalent BPA alternatives in the context of human fertility and oocyte health. Key recommendations based on the findings from this review and elements of the US FDA Redbook 2000 (Toxicological Principles for the Safety Assessment of Food Ingredients) guidelines ( U.S. Food and Drug Administration, 2000; Revised 2007 ) are provided in Table 2 , for the purpose of informing best practice in future studies aimed at assessing the impact of bisphenol on oocyte health and female fertility. We contend that the adoption of these recommendations will help guide more consistent study outcomes, contributing to formulation of a robust and reliable evidence base. It is our hope that such information will provide an evidence-based framework to inform legislative bodies regarding the health implications of the spectrum of bisphenols we encounter, including those linked to low-dose exposures. Key recommendations for future studies assessing the impact of bisphenol exposure on oocyte health and female fertility. In vitro and in vivo studies should employ at least three concentrations/doses or five if assessing non-monotonicity, including equivalent in vitro low doses below 50 ng/ml or in vivo doses below the NOAEL and/or TDI of BPA; the same concentrations/doses should be maintained for BPA alternatives. Animal dosing of bisphenols via oral delivery is preferable if assessing them within a dietary contaminant context, in line with US FDA requirements. In vitro and in vivo : Any BPA alternative such as BPS, BPF, BPB, BPAF etc. being assessed either in isolation or in combination with other bisphenols should be examined alongside BPA as an additional treatment group to characterize and compare the mechanisms of action. BPA can be used to provide a point of reference to build the profile of BPA analogues that are not yet well established within the literature; also generating additional evidence for BPA. Human studies in the fertility clinic setting should aim to incorporate the relevant alternative bisphenols (BPS, BPF, BPB, BPAF, etc.) in their detection protocols alongside BPA. In vitro and in vivo : Studies prioritizing molecular and mechanistic endpoints should pair these with one or more comparable endpoints associated with oocyte or ovarian adverse effects (follicle counts, oocyte yield, oocyte meiotic capacity, oocyte and/or cumulus cell morphology, and oocyte meiotic spindle integrity). Human studies: Spindle assessment or morphological assessment of discarded oocytes from participants should be included where possible alongside current endpoints of oocyte yield and antral follicle count to link to animal study outcomes. ART outcomes including measures of embryo quality, embryo development, and implantation rates are also ideal to report in order to assess the developmental potential of oocytes selected for use in IVF/ICSI. In vitro and in vivo : For study results to be considered by governing bodies, studies must include detailed reporting of methodologies including all aspects of animal use, experimental replicates, cell numbers, and dosing protocols as per US FDA guidelines. These guidelines such as the Redbook 2000 should be clearly defined, updated, and made accessible by the US FDA. Peer review processes for publication of these studies should hold authors to the correct reporting standards of methodologies in this toxicological context. Human studies should report the full range of statistical population parameters for BPA concentration in samples (means, standard deviation, quartiles/percentiles, maximums, the limit of detection, and percent of participants samples that were below the limit of detection, unadjusted and adjusted urinary parameters) for ease of interpretation and comparison between study population. Human observational studies should aim to collect multiple urine samples across different time points using a prospective cohort approach, consider these as separate data points, and where possible, measure unconjugated BPA. To obtain accurate results, it is critical that detection and quantification of BPA and BPA alternatives is conducted using the direct method (not indirect) as detailed in Gerona et al. 2020 ). Abbreviations: BPA, Bisphenol A; BPAF, Bisphenol AF; BPB, Bisphenol B; BPF, Bisphenol F; BPS, Bisphenol S; NOAEL, no observed adverse effect level; TDI, tolerable daily intake; US FDA, United States Food and Drug Administration.