Endocrine-disrupting chemicals: Mainstream recognition of health effects and implications for the practicing internist.

Plastics are a crucial source of many EDCs, including phthalates, bisphenols, and per- and polyfluoroalkyl substances. In February 2022, the United Nations Environment Assembly announced plans to negotiate an international, legally binding instrument to end plastic pollution. Recent public and media attention to microplastics detected in both wildlife and humans 123 has coalesced the policy community around initiatives to reduce the use of single-use plastics. The arguments for this action have often focused on the visible microplastics, despite stronger evidence of human health effects induced by chemicals used as additives in plastics (e.g., phthalates and bisphenols). At the same time, there are substantial gaps in our scientific understanding of the relationships between detectable microplastics in biological specimens and the observed health effects of additive chemicals. The dialogue around the “circular economy” and pivoting to reusable plastics as a solution to the plastic health crisis fails to address concerns about contamination of reused plastics containing EDCs. There is a need to transition to other materials (e.g., glass and stainless steel). While some suggest bioplastics as potential alternatives, early studies suggest potential endocrine activity that may pose similar risk to synthetic, fossil fuel-based plastics. There also remain gaps in information about the health effects of high-density polyethylene and other plastics that do not use phthalates or bisphenols. The global plastics treaty offers a profound opportunity; however, active engagement from the medical community is required to ensure that limiting the adverse health effects of plastics and their constituent chemicals is prioritized.

In animals, the distance from the anus to the genital tubercle has been identified as a sensitive biomarker of androgen action. 41 , 42 In males, this anogenital distance (AGD) is typically measured from the center of the anus to the anterior base of the penis (anopenile distance), or from the center of the anus to the posterior base of the scrotum (anoscrotal distance). During a “masculine programming window” 43 corresponding to weeks 8–14 of gestation in humans, perineal growth and caudal migration of the tubercle in rodents relies upon androgen function, and can be disrupted by antiandrogens. 44 Shortened AGD in boys due to antiandrogenic phthalate exposure is among the best described effects on reproductive development with independent observations from two birth cohorts, The Infant Development and Environment Study (TIDES) 45 and the Swedish Environmental Longitudinal, Mother and Child, Asthma and Allergy (SELMA) study, 46 which showed decreased anogenital distance in boys born to mothers with higher urinary levels of antiandrogenic phthalates in the first trimester of pregnancy. Minipuberty in infancy can also be influential in shaping AGD, 47 , 48 and studies suggest that AGD in animals tracks thereafter to adulthood. 49 Shortened adult AGD is strongly predictive of sperm count, with each 1cm increase in AGD associated with a 4.3 million increase in sperm density per mL, strongly predicting reproductive potential. 50 A study of young adults from the general population confirmed these associations (with a doubling of risk for subfertile sperm count in the lowest 10 percentile of AGD), noting the absence of an association of AGD with contemporaneously measured sex hormones or testicular size. Together, these studies suggest a strong but intermediate link of prenatal phthalate exposure with reductions in sperm count, with AGD as an intermediate biomarker. 51 This is consistent with the concept of the testicular dysgenesis syndrome originally described by Skakkebaek, 52 in which cryptorchidism, hypospadias, and testicular cancer can arise out of decreased production of testosterone and the neohormone insulin-like factor 3 during a susceptible window of male reproductive development. 53 However, multiple studies suggest that later-life exposure to EDCs can impair sperm concentration, motility, and morphology. These studies are generally cross-sectional and do not contain additional information on in utero and early life exposures. Despite these limitations, phthalates have been reported to have negative associations with one or more semen quality parameters, 15 with similar disruptions also reported for BPA, 54 BPS, 55 organophosphate pesticides, 56 – 58 and PFAS. 59 , 60 A modest number of studies have also identified increased risks for reproductive cancer. Occupational exposure to persistent pesticides has been associated with prostate cancer. 61 , 62 Maternal levels of brominated flame retardants have been associated with testicular cancer in their sons. 16 , 63 While further study is needed, PFAS exposure has been linked to kidney and testicular cancer as well as potentially prostate cancer. 64

Buck Louis et al. described an analogous ovarian dysgenesis syndrome in 2007. 65 Similar to the testicular dysgenesis syndrome, disruptions of the developing reproductive tract—including the ovary, fallopian tubes, and uterus—occur in utero ; however, external phenotypes are not readily appreciated. The clinical manifestations are equally significant even if their documentation is delayed. Epidemiologic evidence of EDC effects on female reproduction has chiefly been based on measures of adult exposure, and data have somewhat lagged behind studies of male reproductive effects. Phthalates 66 and PFAS 67 have been associated with endometriosis, and case-control studies have implicated PFAS and bisphenols with increased risk for polycystic ovarian syndrome. 68 – 75 The effects of BPA are perhaps not surprising in light of concordant laboratory evidence of its ovarian toxicity. 76 Multiple studies have also shown EDCs to contribute to breast cancer risk. In the Child Health and Development Studies, prenatal exposure to N-ethyl-perfluorooctane sulfonamidoacetic acid was associated with breast cancer in daughters. 77 Other studies have revealed associations of adult PFAS exposure with newly incident breast cancers. 78 , 79 Pesticide exposure studies have also revealed increases in cancer risk, but none are based on biological specimens; evidence for phthalates is similarly limited. 15

While beneficial, individual actions have limited power to address systemic threats to health; however, current policy interventions also remain inadequate. We refer the reader to a recent review that speaks to severe gaps in the regulation of chemicals in commerce and efforts to limit endocrine-disruptive effects. 13 In brief, chemicals are not regulated for their endocrine-disrupting effects, and efforts are needed to improve: Testing and identification of chemicals that disrupt endocrine systems, Evaluating exposures, Limiting exposures through regulations, and Establishing an International Agency for Research on EDCs Currently available and validated tests do not cover all endocrine modes of action. For example, US regulations require testing for estrogen agonist activity only for pesticides and drinking water contaminants. Even for estrogen and androgen signaling, the validated tests (e.g., uterotrophic assay) work best for endogenous hormones and are too insensitive for identifying EDCs. Fortunately, sensitive assays exist to test a broad number of nuclear receptors, and assays to examine receptor expression, hormone transport, hormone synthesis, and epigenetic alterations should soon be validated for inclusion in regulatory requirements. Recognizing the need to push a large suite of chemicals through a new, more scientifically up-to-date panel of tests, we suggest in vitro screening using high-throughput platforms followed by in vivo testing using zebrafish or mammalian systems to verify the absence of adverse effects using approaches that account for the integrated, multisystem nature of endocrine physiology. At the very least, such approaches should be applied prospectively to avoid the addition of new adverse exposures. Critically, we also need to regulate chemicals by class, rather than one at a time, as regrettable substitutes have emerged time and again (e.g, diisononylphthalate as a replacement for di-2-ethylhexylphthalate, and BPS for BPA). Regulatory policies across the world use exposure and related risk to judge whether to limit chemicals of concern. However, this approach has extreme limitations, and the disease burden linked to EDCs in adults reveals the failure of a risk-based regulatory paradigm. First, current risk-based paradigms do not account for the non-linear and non-monotonic exposure-response relationships common to many EDCs. 121 Indeed, the risk-based paradigm is largely based on extrapolating from no-adverse effect levels in animals to humans, which cannot be done when non-linear or non-monotonic relationships apply. There is a substantial lag from identifying new exposures to completing studies of human health effects, especially for disease outcomes with longer latencies such as diabetes or cancer. While some risk-based approaches attempt to account for age-related vulnerability, they falsely presume that the population sensitivity can be quantified a priori . Moreover, enhanced sensitivity of individuals to EDCs based on underlying disease states or their treatments remains unaddressed. 122 In a hazard-based regulatory environment, chemicals identified as EDCs would simply be removed from use without regard to the exposure required to achieve an adverse effect. There are precedents for a hazard-based regulatory paradigm, especially in Europe for carcinogens; persistent, bioaccumulative and toxic chemicals; and pesticides and biocides which are EDCs. If we persist with a risk-based approach, we need broader and stronger human biomonitoring platforms, particularly to address gaps in exposure assessment in low- and middle-income countries. Such biomonitoring data will also inform educational campaigns about safe and simple steps to limit exposure. We also suggest the establishment of a new international agency, or a broadening of the International Agency for Research on Cancer (IARC)’s scientific charge, to include endocrine disruption. Established in 1965, IARC was tasked with evaluating the evidence of carcinogenesis due to environmental hazards. Monographs describe three streams of evidence (mechanistic, animal, and epidemiological studies). An autonomous body such as the IARC can bring together diverse experts for international collaborative reports on EDCs, and further support the post-2020 process of the Strategic Alliance for International Chemicals Management.

Expert panels organized by the Endocrine Society conservatively estimated the costs of adult diseases ( Table 1 ) in the European Union attributable to EDCs. 89 , 90 Teresa Attina et al. expanded these to the US in 2016, 91 and Julia Malits et al. expanded these to Canada in 2022. 92 Trasande et al. have also estimated cardiovascular mortality due to phthalates, with over 90,000 deaths annually and at least $39 billion/year in lost economic productivity in the US alone. 40 Vladislav Obsekov et al. estimated PFAS costs in the United States. 93 Moreover, minoritized populations disproportionately account for EDC-associated healthcare costs. Though non-Hispanic Blacks and Mexican Americans comprise 12.6% and 13.5% of the US population, they bear 16.5% and 14.6% of the disease burden due to EDCs, respectively. 94 Critically, these are conservative estimates for several reasons, including: they are limited to a subset of chemicals in plastic materials that contribute to disease and disability; they are limited to a subset of diseases due to the few chemicals studied; and the cost estimates represent a subset of the entire costs due to the disease studied. Despite these limitations, EDC-associated healthcare costs are substantial.

There are now more than 50 chemical “obesogens.” 18 A fungicide used in marine paints and now a known contaminant of many plastic containers, tributyltin is arguably the first and prototypical obesogen with a well-characterized mechanism of action—activation of peroxisome proliferator-activated receptor-γ signaling—through which it promotes adipogenesis and increases adipose mass. 19 Furthermore, in experimental models, tributyltin’s metabolic effects have been shown to be epigenetically transmitted across multiple unexposed generations. 20 Adult exposures to synthetic chemicals are also known to increase weight gain and contribute to diabetes independent of well-known risks such as caloric excess and physical inactivity. Serum levels of per- and polyfluoroalkyl substances (PFAS) used in nonstick cooking materials and oil- and water-resistant clothing were shown to augment weight regain after a successful weight loss intervention, possibly due to reductions in resting metabolic rate. 21 In the Diabetes Prevention Program trial, total PFAS was associated with increased weight gain in the control group. 22 Studies from multiple countries have also identified increases in incident diabetes in association with antecedent measures of serum PFAS. 23 – 25 Bisphenols used in aluminum can linings and thermal paper receipts such as bisphenol A (BPA) and one of its emerging replacements, bisphenol S (BPS), have been associated with incident diabetes in a nested case-control study within a large representative French cohort. 26 The cardiovascular consequences of these exposures are also substantial. BPA is known to reduce levels of the cardioprotective adipokine adiponectin 27 and promote oxidative stress, 28 – 31 a major mechanism underlying atherosclerosis and cardiovascular events. In human studies, BPA has been associated with greater carotid intimal media thickness, 32 severity of coronary artery disease on angiography, 33 and reduced heart rate variability. 34 Linking national representative surveys from Americans to death records, Bao et al. identified substantially increased risk for all-cause and cardiovascular, but not cancer, mortality among those with higher urinary BPA levels. 35 Phthalates are EDCs known to antagonize testosterone action, 36 and concerns have been raised about consequential cardiovascular risks given that low testosterone is a marker for or predictor of mortality. 37 Furthermore, phthalates are potent inducers of oxidative stress, 38 , 39 and thus may augment cardiovascular disease risk in both sexes. Indeed, linkage of urinary phthalate levels with death records revealed increases in cardiovascular mortality, independent of the increase observed with phthalates, in both sexes, suggesting approximately 90,761–107,283 preventable deaths annually in the US. 40

There are particularly large gaps in the adult epidemiologic literature that would serve to advance our understanding of EDCs and their contribution to adult chronic disease. Large national and international biobank consortia (e.g., All of Us, UK Biobank Cohort) have not measured chemical exposures in sera or urine to evaluate later life vulnerabilities to these conditions. Some notable exceptions include the Nurses’ Study 109 and the Study of Women’s Health Across the Nation, 110 but even in these samples, the researchers found limited statistical power to examine the potential effects. These cohorts have also included more women than men, which may limit their capacity to discern effects that are likely to be sexually dimorphic. Fewer still have provided mechanistic insights by examining hormone levels or leveraging multi-omic approaches to confirm or corroborate observed effects. Further work is needed to empower large-scale studies to fully interrogate the environmental drivers of disease risk in diverse populations.

Healthcare facilities also use many products that increase the risk of EDC exposures. 95 Phthalates, for example, are abundant in polyvinylchloride-based medical devices such as blood bags, nutrition pockets, tubing, umbilical venous catheters or disposable gloves, where they can account for up to 40% of the final product. 96 They are also used to make coatings for oral medications and in flooring. Bisphenols are used in polycarbonate-based medical tubing, hemodialysis equipment, newborn incubators, syringes, and nebulizers. 97 Parabens are used in medications and intravenous catheters for their antimicrobial properties. 98 Exposures are likely the greatest per unit body weight in neonatal intensive care units, where non-invasive respiratory support and feeding tubes were identified as the most significant drivers of phthalate exposure. 99 While modern healthcare has transformed human health and acknowledging that in some instances the removal of select EDCs from medical devices may augment risk, 98 the lack of knowledge of medical care as a vector of exposure undermines core principles of medical ethics. 100 Importantly, given the magnitude of plastic waste generated by modern healthcare systems, addressing EDCs in healthcare is likely to have broader benefits on health. Unfortunately, health care providers know little about EDCs and seldom offer steps to patients to limit exposure. 101 – 104 Given that disparities in EDC exposure are well documented, 105 , 106 collaborative efforts are needed between scientists and healthcare organizations to develop products that improve provider knowledge about EDCs and support the use of safer alternatives in medical devices and other equipment. Healthcare providers are well positioned to communicate safe and simple steps to limit exposures; however, knowledge about environmental exposures remains limited, and health care providers still have modest self-efficacy in managing common exposures and communicating advice for prevention. 101 , 103 , 107 , 108

Patients are potentially exposed to a host of EDCs associated with cardiometabolic, reproductive, and other disorders that contribute substantially to healthcare expenditures. Moreover, differential exposure to these chemicals is a likely but underrecognized driver of health disparities. The internist can play a critical role in addressing this underlying health threat by working with patients on strategies to reduce their exposures; collaborating across the healthcare system to reduce exposures arising from clinical care; and advocating for local, national, and global policies that identify EDCs and remove them from our environment.

Many endocrine disorders are characterized by marked disparities in disease prevalence, complications, and/or mortality—including metabolic, thyroid, and reproductive disorders. 80 – 84 In most instances, Black, Hispanic/Latinx, Indigenous, and low-income communities are disproportionately burdened. While recognition of the social/structural determinants of health (SDoH) has improved our understanding of the origins of these disparities, SDoH discussions often focus on educational attainment, economic opportunity, healthy food and activity ecosystems, and access to health care, 85 without substantial regard for disparities in exposures to EDCs. This is a missed opportunity as minoritized and low-income communities are known to be disproportionately exposed to a variety of toxicants linked to endocrine dysfunction, including polychlorinated biphenyls (PCBs), phthalates, bisphenols, organochlorine pesticides, air pollution, PFAS, toxic metals, and brominated flame retardants (reviewed in ref. 86 ). The drivers of these disparities are myriad but include neighborhood segregation, racialized labor and beauty practices, grandfathering clauses for industrial sites, suburbanization, and urban disinvestment, all of which are rooted in intersectional systems of oppression. 87 , 88 While the bases of these exposure disparities have deep historical roots and are nurtured by current systems of power, unlike genetics, environments are modifiable; thus, improving environmental health has the potential to mitigate endocrine health disparities.

Silent Spring was not a text routinely included in the medical school curriculum of the 1960s, when Rachel Carson warned about the consequences of widespread use of dichlorodiphenyltrichloroethane (DDT).[ 1 ] Except for malaria prevention in endemic regions, DDT is rarely used today; however, Carson’s prescient warnings have implications for the present-day internist. Indeed, substantial evidence suggests that pesticides used as DDT replacements contribute to myeloid leukemias and colorectal cancer.[ 2 ] While further studies are needed to better characterize the carcinogenicity of newer pesticides, one longitudinal study has associated greater organic food consumption with reduced cancer risk.[ 3 ] Since Carson’s landmark work, roughly one to three thousand new synthetic chemicals have been approved for industrial use annually. Current estimates suggest 300,000 synthetic chemicals are used to create consumer products,[ 4 ] and many of these chemicals were introduced without testing for safety, particularly for effects on endocrine systems.[ 5 ] In 2009, the Endocrine Society published its first scientific statement describing how a broad suite of actively used chemicals impaired endocrine function with consequences for neurodevelopment, metabolism, reproduction, and carcinogenesis. 6 This was soon followed by a World Health Organization joint report with the United Nations Environment Programme (UNEP) that identified EDCs as a global public health issue. 7 A second Endocrine Society scientific statement followed in 2015, 8 and the Strategic Alliance for International Chemicals Management—the global multi-sectoral and multi-stakeholder policy framework hosted by UNEP—“welcomed” the WHO-UNEP report, acknowledging only disagreement from the chemical and pesticide industry. 9 The Endocrine Society defines EDCs as “exogenous chemical[s], or mixture[s] of chemicals, that interfere…with any aspect of hormone action.” 10 To date, the US Food and Drug Administration has identified more than 1800 chemicals that disrupt at least one of three endocrine pathways (estrogen, androgen, and thyroid). 11 The actual number of EDCs is likely to be much higher, as very few chemicals have been tested for endocrine disruption, and as other receptors and the broader suite of human hormones are considered. We now appreciate that receptor binding is but one path by which hormone action can be interfered. Indeed, a recent analysis identified 10 key characteristics of EDCs, 12 with some mechanisms such as epigenetic effects on the endocrine system only recently coming under interrogation ( Figure 1 ). 13 The early evidence documenting effects of endocrine-disrupting chemicals (EDCs) linked pregnancy-associated exposures with neurodevelopmental disabilities in exposed children. 14 Subsequent studies documented consequences of early life exposures that included obesity and diabetes 15 as well as alterations in male 16 and female 17 reproductive development. While these studies focused upon exposures during early life, substantial evidence has accumulated indicating that the entire lifespan is a window of EDC susceptibility. This perspective focuses on effects that have been identified as a result of adult EDC exposure. Furthermore, it describes safe and simple steps people can take to limit their exposures.

In 2022, the National Academies of Science, Engineering, and Medicine published “Guidance on PFAS Exposure, Testing, and Clinical Follow-Up,” the first clinical guidelines on an EDC. 111 These guidelines are a milestone that bring the internist into the world of environmental health. In addition to providing extensive background on PFAS contamination and potential adverse health effects, the guidelines empower clinicians with robust, actionable guidance to identify and mitigate risk. The guidelines are focused upon a subset of potential PFAS effects for which the committee found sufficient evidence for increased risk (i.e., reductions in birthweight, dyslipidemia, kidney cancer, and reduced antibody responses) while also commenting on areas with limited suggestive evidence (i.e., breast cancer, pregnancy-induced hypertension, liver enzyme elevations, testicular cancer, thyroid dysfunction, and ulcerative colitis). Key recommendations of the guidelines include: 4–1: “Clinicians advising patients on PFAS exposure reduction should begin with a conversation aimed at first determining how they might be exposed…and what exposures they are interested in reducing.” 4–2: “If patients are exposed occupationally…clinicians should consult with occupational health and safety professionals…to determine the most feasible ways to reduce that exposure.” 4–3: “Clinicians should advise patients with elevated PFAS in their drinking water that they can filter their water to reduce their exposure.” 4–4: “In areas with known PFAS contamination, clinicians should advise patients that PFAS can be present in fish, wildlife, meat, and dairy products and direct them to local consumption advisories.” 4–5: “Clinicians should direct patients interested in learning more about PFAS to authoritative sources for information on how PFAS exposure occurs and what mitigating actions they can take.” 4–6: “When clinicians are counseling parents of infants of PFAS exposure, they should discuss infant feeding and steps that can be taken to lower sources of PFAS exposure. The benefits of breastfeeding are well known ; the AAP, the AAFP, and the ACOG support and recommend breastfeeding for infants, with rare exceptions. Clinicians should explain that PFAS can pass through breast milk from a mother to her baby. PFAS may also be present in other foods, such as water used to reconstitute formula and infant food, and potentially in packaged formula and baby food. It is not yet clear what types and levels of exposure to PFAS are of concern for child health and development.” 4–7: “Federal environmental health agencies should conduct research to evaluate PFAS transfer to and concentration in breast milk and formula to generate data that can help parents and clinicians make shared, informed decisions about breastfeeding.” The guidelines identify who should be tested, how PFAS levels should be interpreted with regard to health risks, and how those risks should be monitored in adults and children. A cornerstone of all environmental health interventions, exposure reduction strategies are highlighted (included in Table 2 ) as are potential therapeutic approaches such as the use of bile acid sequestrants or phlebotomy, although further evidence is likely needed to advocate for broad use of these approaches. Critically, these guidelines highlight many of the key principles of modern environmental health practice: patient education, clinician engagement, shared decision-making, community engagement, patient/community/clinician advocacy, and system-wide policy development and implementation that focuses on human health. Despite its sole focus on PFAS, the impact of these guidelines is difficult to overstate; however, it is also clear that expansion of clinical guidelines to address other EDCs is urgently needed. In the meantime, an array of intervention studies reveals a path forward for reducing other EDC exposures. Though large-scale intervention studies have not been conducted, small-scale interventions support the feasibility of reducing EDC exposures. Lu et al. reduced organophosphate pesticide metabolites in urine to nondetectable levels through an organic diet intervention. 112 Though concerns about the additional costs associated with organic food are appropriate, a more recent dietary intervention similarly reduced pesticide metabolites in a low-income, agricultural population. 113 In young girls, choosing personal care products labelled to be free of phthalates, parabens, triclosan, and benzophenones reduced personal exposure by 27%–44%. 114 Another dietary intervention study that replaced diets in a small sample of families with fresh foods reduced urinary levels of phthalate metabolites and bisphenols by 53%–56%. 115 Household interventions can also reduce exposure. One study performed a graded comparison of offices, common areas, and classrooms, measuring dust levels of PFAS, polybrominated diphenyl ethers (PBDEs), and organophosphorus esters (OPE), the latter having increasingly replaced PBDEs in electronics and furniture. Rooms with full “healthier” materials had 78% lower dust levels of PFAS, 65% lower OPE levels, and 45% lower PBDE levels than rooms with only partial interventions, adjusted for covariates related to insulation, electronics, and furniture. 116 To be clear, not all studies have achieved expected changes in EDC levels. One study reported an increase in urinary phthalate metabolites due to substantial phthalate contamination in the coriander provided to participants. 117 Use of a BPA risk score based on characteristics of food containers and packaging did not reduce urinary levels. 118 Much more work is needed to formalize clinical guidance for providers to empower them to address patient questions about exposure reduction strategies; however, recognition of exposure sources and burgeoning clinical studies can assist providers in guiding their patients ( Table 2 ). 119 , 120