Clarity in the face of confusion: New studies tip the scales on bisphenol A (BPA)

By now, bisphenol A (BPA) has become a ‘household name’; many consumers are aware of this compound and the various products where it is found. BPA is used in a wide range of goods including food and beverage containers, medical and sports equipment, other plastic products, thermal papers, and cosmetics, among others (Geens et al., 2012). Reference populations from around the world have shown that exposures to this compound are widespread in all age groups, indicating that leaching from consumer products is occurring (Vandenberg, 2011).

BPA is widely acknowledged to be an endocrine disrupting chemical (EDC), broadly defined as compounds that interfere with one or more hormone actions (Zoeller et al., 2012). In vitro and in vivo studies have shown that BPA can bind to and activate nuclear estrogen receptors (ER) α and β, membrane ERs, the G-coupled protein receptor GPER, thyroid hormone receptor, androgen receptor, and estrogen related receptor (ERR)γ (Vandenberg et al., 2009). Thus, studies that examine the effects of BPA solely from the perspective of nuclear ER signaling are too limited to appropriately evaluate this promiscuous compound.

Hundreds of studies published to date have demonstrated that low doses of BPA, e.g. doses below the toxicological NOAEL of 50 mg/kg body weight/day, alter a range of endpoints in laboratory animals following controlled exposures (Richter et al., 2007; Vandenberg et al., 2013b). In fact, many of these studies have shown effects of BPA at doses below the US Environmental Protection Agency (EPA)’s reference dose of 50 µg/kg body weight/day. Many different groups of scientific experts have concluded that BPA can disrupt development of multiple organs in the body including those found in the male and female reproductive tracts, the prostate gland, the mammary gland, tissues and organs involved in metabolism, and the brain; that BPA alters neurobehaviors; and that BPA can sensitize tissues like the mammary and prostate glands to hormonal and carcinogen challenges (Diamanti-Kandarakis et al., 2009; Gore et al., 2015; Peretz et al., 2014; Richter et al., 2007; Seachrist et al., 2016; Vandenberg et al., 2013b).

The picture from human studies is equally concerning: more than 100 epidemiology studies suggest associations between BPA exposures and a range of conditions and diseases including metabolic syndrome, infertility, and severity of asthma (Rochester, 2013). Building a case for causal relationships between BPA and human disease will take significant time and resources, especially because exposures are so ubiquitous. Decade-long longitudinal studies that adequately measure developmental exposures are needed to recapitulate what has been experimentally tested (and observed) in laboratory animals.

Considering the strength of the evidence linking BPA to harmful health outcomes, why is there so much controversy surrounding it? A widely circulated answer is that studies have not always replicated others’ adverse findings, with some reports refuting adverse results entirely while others produce disparate or non-identical findings. Another contributing influence is a lack of consensus on what constitutes “adverse”; with some claiming that overt signs of toxicity must be identified whereas others consider developmental disruptions, interruption of measures of homeostasis, or other endpoints that are predictors of disease as qualifiers (Woodruff et al., 2008). However, one factor that cannot be ignored is the role of ‘manufactured doubt’, a concept that was originally invented by representatives from the tobacco industry to generate scientific debate about issues that were relatively well settled solely for the purpose of shaping public opinion and delaying regulatory action (Michaels, 2006). In the case of tobacco, the industry argued that the science that linked smoking to lung cancer was ‘unsettled’ and that ‘consensus’ had not been reached, long after their own scientists knew the true extent of the causal relationship between smoking and cancer (among other outcomes). It has been suggested that the chemical industry, as well as trade manufacturing groups, have used similar tactics to keep harmful EDCs including BPA on the market, long after scientific inquiry has identified their potential to cause harm (Bergman et al., 2015; Oreskes et al., 2015).

Another element that has contributed to the controversy surrounding BPA is the general lack of regulatory action against this chemical by agencies in the USA (e.g. the US EPA and US Food and Drug Administration, FDA) or Europe (e.g. the European Food Safety Authority, EFSA). Why have these agencies failed to act, in spite of hundreds of studies showing that BPA can induce adverse and/or disruptive outcomes in exposed laboratory animals? Unfortunately, the vast majority of studies on BPA conducted to date have been ignored by these regulatory agencies because of requirements that studies meet particular criteria for the results to be considered in risk assessments.

One criterion that is heavily favored is the use of Good Laboratory Practices (GLP), a set of rules for recordkeeping and study conduct (Agrawal et al., 2009). GLP was instituted by regulatory agencies after the discovery of serious cases of scientific fraud; it was implemented to prevent widespread fraud that affected the integrity of data collected by groups with a financial stake in the safety of the compounds being examined (Myers et al., 2009a). Because most academic laboratories cannot conduct studies according to GLP due to the high cost and regulatory hurdles associated with GLP practice, the majority of academic studies have not been included in chemical risk assessments. Another important criterion used in the selection of studies to be used by regulatory agencies is adherence to test guidelines which dictate the design of experiments and assessment of endpoints (Myers et al., 2009b). OECD and EPA guidelines exist to test compounds to determine if they are reproductive toxicants, developmental toxicants, carcinogens, skin irritants, etc. The endpoints examined in these guideline studies are overt signs of toxicity (e.g. mortality, loss of body weight, organ weight, litter sex ratio, etc.) rather than indicators of disease (Vandenberg et al., 2013a). Further, due to their lack of advanced inquiry and limited endpoints, many academic studies do not follow test guidelines, focusing instead on more state-of-the-art endpoints that better map to diseases of concern (Myers et al., 2009b).

Thus, while the scientific consensus clearly indicates that BPA is an EDC, that BPA alters development at low, environmentally relevant doses, and that BPA is likely affecting humans at current levels of exposure, very few regulatory actions have been taken to protect public health or reduce human exposures. Absent studies that follow GLP, examine guideline endpoints, and include low doses, the landscape for BPA regulation seems unlikely to change.

In this issue of Andrology, two new studies meet exactly this demand. Studies carried out at the National Food Institute of Denmark have used a robust study design with large sample sizes, controlled oral exposures, a range of doses, and endpoints that assess effects of BPA on sexual maturation, weights and histopathology of male and female reproductive organs, regularity of oestrus cyclicity, sperm counts, mammary gland growth parameters, and neurobehaviors (Hass et al., 2016; Mandrup et al., 2016). Importantly, these studies show that BPA exposures during development can affect a range of endpoints in exposed rats including sperm count, body weight, sweet preference, spatial learning ability, male mammary gland outgrowth, and female mammary intraductal hyperplasias.

There are a few take-home messages from these two new studies. First, some effects are seen at low doses of BPA, but not at higher doses. These may be examples of non-monotonic dose responses, which are common for hormones and EDCs, and can be explained by a number of endocrine-mediated mechanisms including receptor down-regulation, receptor competition, and the overlap of competing monotonic curves (Vandenberg et al., 2012). The examples provided by Hass et al. (2016) and Mandrup et al. (2016) offer additional evidence that non-montonicity is a typical feature of BPA studies and can be observed in standard guideline endpoints. Second, although the experimental design, doses selected, and measurement tools differ somewhat from prior studies, the overall picture is the same: low doses of BPA can disrupt hormone-sensitive endpoints when exposures occur during sensitive periods of development. Finally, endocrinologists will likely find the sexually dimorphic effects of BPA to be both expected and interesting. As has been demonstrated in other studies, Hass and colleagues (2016) show that BPA can masculinize (or defeminize) female behaviors, but no significant effects are observed in males. These sex-specific responses point to the likelihood that BPA disrupts neurobehaviors via ER-mediated effects in specific regions of the brain (Frye et al., 2012).

Hass and colleagues conclude by stating that their findings suggest that EFSA’s “temporary tolerable daily intake of 4 µg/kg bw/day is not sufficiently protective” of humans in the general population. Mandrup et al. similarly state that “highly exposed humans may not be sufficiently protected.” We hope that their words do not fall on deaf ears. Regulatory agencies have two new studies that meet the criteria on their ‘wish lists’ for inclusion in chemical risk assessments. These studies clearly ‘tip the scales’ and indicate a need for new regulations, including the setting of new reference doses (in the USA) and tolerable daily intake doses (in Europe). Furthermore, these studies reiterate what has been known in the scientific community for many years: that low doses of BPA can significantly alter a range of hormone sensitive endpoints.

While regulators around the world have been slow to act, scientific communities have taken a stand against EDCs like BPA (Bergman et al., 2013a; Bergman et al., 2013b; Gore et al., 2013; vom Saal et al., 2007). And consumers have similarly acted, calling for BPA-free products. Unfortunately, many of these products include other bisphenols, which are poorly studied but have been sufficiently examined to raise concern about their endocrine disrupting properties (Rochester and Bolden, 2015; Vandenberg et al., 2015). We sincerely hope that this movement toward BPA replacements will not lead to another 20 years of regulatory inaction while the public continues to be exposed.