Bisphenol A (BPA) exposure in New Zealand: a basis for discussion

Abstract

This article aims to provide a broad overview of current debates around the safety of bisphenol A (BPA), a commonly used chemical in consumer products, and to argue for greater engagement by the New Zealand government and public with these debates. BPA acts as an endocrine disrupting chemical (EDC), a relatively new category of harmful chemicals that pose a challenge to conventional toxicological assessments. There are ongoing debates surrounding BPA's potential to have harmful effects on human health at current doses, and how to evaluate the reliability of conflicting studies. Based on this uncertainty, a number of countries have taken a precautionary approach and regulated BPA. The debate has also moved into the public arena, with strong public opposition to BPA in countries such as the United States of America. The New Zealand government has maintained its position that BPA is safe without fully considering the challenges it may pose as an EDC or additive effects with other EDCs. Based on a preliminary investigation, there does not currently appear to be a high level of public awareness or opposition to BPA in New Zealand. The authors argue that the New Zealand government and the public could benefit from engaging with current debates on BPA's safety.

Keywords:

• bisphenol A; BPA
• chemical safety and regulation
• endocrine disrupting chemicals
• media
• New Zealand
• politics of science
• social activism

Introduction

Bisphenol A (BPA) is a high-production chemical that has received increasing attention over the past two decades as a potential hazard to human health. Internationally, there have been both scientific and political struggles around defining BPA's safety. Despite the concerns raised about BPA, there appears to have been a limited engagement with these debates by either the New Zealand government or public. The goal of this article is not to make conclusions on these debates but to argue for engagement with them. The focus on arguments against BPA's safety is due to the authors’ opinion that this research needs to be engaged with, and not their assertion that they are completely valid (indeed, the authors are not qualified to make this assertion). This article will give a broad overview of key scientific concepts and debates including the low-dose hypothesis and the cocktail effect. The plethora of contradictory studies has made it difficult to define what regulators can rely on as ‘good science’. New Zealand regulators have relied heavily on studies conducted under good laboratory practice (GLP) guidelines without a critical evaluation of the guidelines’ appropriateness to evaluate BPA's safety. Precautionary approaches have been used in other countries to deal with this uncertainty, while New Zealand has maintained a relatively conservative approach. Overseas, the debate over BPA's safety has moved from the science arena into the public arena, with widespread campaigns for regulation in countries such as the United States of America and Canada. However, BPA has received comparatively little media, public or political attention within New Zealand. Particularly contentious is the role of the public in decision making. While New Zealand regulators have argued against public opinion influencing regulation instead of science, this article will discuss how public campaigns against BPA in the USA were partially led by scientists. Increased levels of discussion about BPA by New Zealand scientists, policymakers, politicians and the public would be beneficial and this article aims to contribute to these discussions.

This article will not systematically review all the literature regarding BPA. Rather it offers a qualitative review of a range of perspectives on key areas of contention. These perspectives are gained from a survey of key international studies on BPA and case studies of the BPA debate. A preliminary investigation of public exposure to discussions about BPA in New Zealand was carried out using search engines to identify New Zealand media articles, blog posts and websites relating to BPA. To investigate opposition and support for BPA in New Zealand, organisations with an interest in BPA were identified and profiled. This included consumer activist groups, health advocacy groups and political parties, as well as some individuals with public presence. Publicly available statements by these groups have been read and analysed. The authors have been in email communication with some of these individuals and organisations for further information, but no in-depth interviews have been conducted at this stage. The description of the New Zealand government's position is based on published statements relating to BPA, as well as a review of their policies and regulatory framework for dealing with chemical hazards. This article does not provide a comprehensive description of BPA in New Zealand but rather aims to discuss key areas of contention with a view to initiate further discussion.

Endocrine disrupting chemicals

BPA is part of a fairly new class of harmful agents known as endocrine disrupting chemicals (EDC). EDC is a relatively new term, first proposed at the Wingspread conference in 1991 (Markey et al. 2003). It refers to chemicals, both artificial and natural, that are able to mimic hormones. This interferes with the natural endocrine system and can result in developmental changes in the body. BPA falls into this category as it mimics oestrogen, in particular 17β-estradiol (Wetherill et al. 2007). Oestrogen-mimicking compounds are referred to as xenoestrogens, and naturally occurring xenoestrogens, such as those found in soy, are referred to as phytoestrogens. EDCs are differentiated from other toxins as they interact with cellular hormone receptors. This interference with the endocrine system gives them the potential to cause developmental changes, even when present at very low concentrations. One implication of this is that the relationship between dose and effect is not necessarily monotonic, that is to say negative health effects may not increase proportionally with the dose. There is also significant debate over whether traditional toxicological testing regimes are suitable to deal with EDCs, as evaluating EDC safety may require the use of hormonal instead of traditional endpoints, as well as including the possibility of non-monotonic dose-effect relationships into toxicological models. It is these distinctions that make the debate around BPA's safety, as well as that of other EDCs, so problematic.

Sources of exposure

BPA is a high-production chemical that can be found in a range of consumer products leading to widespread exposure. BPA molecules are used to form polycarbonate plastics that are widely used for drinking bottles and water coolers (Geens et al. 2012a) (Fig. 1). BPA has been used for baby bottles, but many New Zealand suppliers have voluntarily phased these out (NZPA 2010a). BPA is polymerised to form the epoxy resins used to line most tin cans and the metal lids of glass jars (Geens et al. 2012a). BPA leaches in low quantities into food and beverages stored in both polycarbonate plastic and epoxy resin. In Spain, BPA has been detected in stretch films for food contact applications, although this has not been studied in New Zealand (López-Cervantes & Paseiro-Losada 2003).

Figure 1 A, Bisphenol A molecule; and B, as part of a polycarbonate chain.

BPA is also present in a range of goods not used for food storage. BPA molecules are used as a contrasting agent in some thermal paper, primarily used in receipts. These BPA molecules, which are not polymerised, can be transferred on to the skin and pass into the blood stream (Demierre et al. 2012). In studies in other developed countries, BPA has been detected in 44%–98% of receipts (Geens et al. 2012b), but no data are available for New Zealand. BPA can be found in many dental sealants in high concentrations, as it is used as a precursor for the principle sealant material (Kloukos et al. 2013). While other sources of BPA have been recorded, such as its presence in household dust and drinking water, only the major sources have been identified here.

Levels of exposure

Currently there have been two major studies assessing BPA exposure levels in New Zealand (Thomson et al. 2003; Thomson & Grounds 2005). The 2003 study evaluated total xenoestrogen exposure in New Zealand and approximated the relative significance of BPA. At the time of the 2003 study, no data were available on BPA exposure in New Zealand. A worst case scenario was used as the model, putting BPA's relative contribution to xenoestrogen exposure at 34% of the total xenoestrogen load. The 2005 study assessed the BPA content of 80 tinned foods, finding BPA in all except for soft drinks. From this study, it was estimated that dietary exposure to BPA for adult New Zealanders is between 0.008–0.29 μg/kg bodyweight/day, with BPA making up 7% of total exposure to xenoestrogens. However, no research to date has been conducted on exposure to non-dietary sources such as thermal paper in New Zealand. Established models of BPA transfer through human skin from other developed countries provide a worst case exposure scenario of 0.16 μg/kg bodyweight/day from receipts for occupationally exposed individuals such as cashiers (Demierre et al. 2012). Assuming New Zealand exposure to BPA from thermal paper is equivalent, adding this figure to existing data would give worst case exposure levels of up to 0.45 μg/kg bodyweight/day. BPA from dental sealants migrates into saliva at relatively high concentrations for up to 24 hours after application. An average total dose is 13 µg with worst case doses at 30 mg—a high level of acute exposure (Kloukos et al. 2013). It is important to note that while BPA is a small portion of total xenoestrogen exposure, it may have a disproportionate effect on foetal development, as will be discussed further on.

Currently, there have been no biomonitoring studies for BPA conducted in New Zealand. Biomonitoring studies test for BPA in the body fluids, most commonly urine, of the general population. A large number of biomonitoring studies in countries including the USA, Spain, the Netherlands, Germany, Austria, Korea and Japan have shown widespread BPA exposure, with all studies showing detectable levels of BPA in 75%–100% of the sampled population (Vandenberg et al. 2012a). Based on the ubiquity of BPA found in Thomson & Grounds’ (2005) study, and in comparison with studies carried out in other developed countries, it is reasonable to assume that BPA exposure is widespread in New Zealand. The level of exposure is far below the safety limit of 50 μg/kg bodyweight/day proposed by the USA's Food and Drug Administration (FDA) and the World Health Organization (WHO) and used by the New Zealand government, but within the range of low-dose endocrine-disrupting effects that have been proposed.

However, these proposed limits do not account for multiple exposures to other xenoestrogens, so a consideration of the ‘cocktail effect’ is necessary. The cocktail effect refers to the wide range of chemicals individuals are exposed to on a regular basis and how this can lead to additive or synergic effects that may be harmful to human health (Shaw 2014). Additive effects describe the cumulative impact of different chemicals that have a similar effect: in this case those chemicals that mimic oestrogen. Commonly used safety limits of total oestrogen load that the body can be exposed to are used to set the safety limit of each xenoestrogen. However, as xenoestrogens display similar effects, then their total effect will be greater than that assessed for each xenoestrogen. In the particular case of BPA, the different sources of BPA all contribute to total exposure yet they are regulated individually. The ubiquitous nature and multiple sources of BPA and other EDCs suggest a more holistic regulatory approach needs to be taken.

Human health impacts

BPA has been linked to a range of negative health impacts in humans including breast, vaginal, prostate and testicular cancer, decreasing sperm quality and fertility, early sexual maturation of females, malformed genitalia, obesity, coronary heart disease, weakened immune systems and neurobehavioral issues (den Hond & Schoeters 2006; vom Saal & Welshons 2006; Keri et al. 2007; Richter et al. 2007; Vandenberg et al. 2007; Wetherill et al. 2007; Soto et al. 2008; Melzer et al. 2010). Although these studies present strong evidence of negative health impacts from BPA, these claims are contested and other studies have not found definitive links between BPA exposure at current levels and these health endpoints.

Many of the health concerns related to BPA are due to developmental changes and the ‘foetal origins of disease’ hypothesis. As the development of the foetus occurs under strict hormonal control, changes in the endocrine system may lead to developmental changes in the foetus predisposing it to health concerns later in life (Shaw et al. 2009). For example foetal organisational changes to tissue structure that lead to the malfunction of sex hormone targets may be carcinogenic, predisposing individuals to breast, vaginal, prostate and testicular cancer. This link has been reported in animal studies and the mechanism investigated through in vitro testing (Murray et al. 2007; Soto et al. 2008). As noted in the 2005 Thomson & Grounds (2005) study, BPA presents only a small portion of total xenoestrogens to which the New Zealand population is exposed (7%). However, there is a possibility that BPA passes through the placenta far more readily than phytoestrogens. The placenta removes 17β-estradiol and structurally similar compounds. However, xenoestrogens with different structures, such as BPA, may not be removed—BPA mimics oestrogen through the spatial arrangement of its OH groups and not because it has a steroid nucleus (Shaw et al. 2009). A foetus may be exposed to levels of BPA sufficient to cause developmental changes predisposing it to health concerns such as breast cancer later in life (Soto et al. 2008).

BPA's potential effect on sperm quality is of particular concern for New Zealand as the decrease in sperm concentration nationally is one of the highest in the world. A study published in 2008 provides evidence indicating that the New Zealand sperm count has been decreasing at 2.5% per annum over the past 20 years, compared with 1%–1.5% in other countries (Carlsen et al. 1992; Swan et al. 2000; Shine et al. 2008). While sperm concentrations remain above the ‘danger mark’, the rapid decrease gives cause for concern. New Zealand toxicologist Professor Ian Shaw has suggested that low doses of BPA and other xenoestrogens are one of the likely causes for this decrease (Shaw et al. 2008).

No robust epidemiological data are available to evaluate long-term effects relating to foetal exposure. As BPA exposure is near ubiquitous in developed countries and any negative health effects may be the result of exposure decades earlier, it is very difficult to build reliable epidemiological data. Cross-sectional studies comparing BPA exposure with current health effects have been conducted, with a minority of these showing a positive correlation (see Braun et al. 2009; Newbold et al. 2009; Melzer et al. 2010; Clayton et al. 2011; Li et al. 2011); however, these results are only preliminary. This difficulty also problematises regulatory evaluations. For example, Food Standards Australia New Zealand (FSANZ) carefully monitors the results of any regulatory decision undertaken, and in the case of BPA claims that it has been used safely for many years. However, claims such as these are near impossible to make, and because of these problems, risk assessments and evaluation cannot reliably use epidemiological data.

Low-dose hypothesis

The potential for EDCs to exhibit effects at low doses brings into doubt many current safety testing regimes that have conventionally been used to set safety limits. A key principle of modern toxicology is the Paracelsus principle (that the dose makes the poison), which states that potentially harmful chemicals can be used safely below certain levels (Myers et al. 2009a). While it is widely accepted that BPA can be harmful, the real question is, at what dose is it harmful? Safety limits have been set at 10–50 μg/kg bodyweight/day by groups such as the USA's FDA and WHO, far above the estimated exposure levels of 0.008–0.45 µg/kg bodyweight/day in New Zealand. However, a large number of studies have indicated effects on hormonal endpoints at far lower doses. ‘Low dose’ is a vague term, but the USA's National Toxicology Program's (NTP) definition is useful: ‘Biological changes that occur in the range of human exposures or at doses lower than those typically used in the standard testing paradigm of the U.S. Environmental Protection Agency for evaluating reproductive and developmental toxicology’ (Melnick et al. 2002, p. 1). Whether or not BPA has harmful effects at low doses remains controversial, but the principle has been well established and accepted by the NTP for other chemicals and even forms of oestrogen itself (Melnick et al. 2002; Galvao et al. 2014; Matsui et al. 2014). These low-dose effects show that traditional toxicological testing regimes based on the Paracelsus principle are inadequate for dealing with EDCs.

The traditional safety testing protocols for toxins assume monotonic dose response curves. Large doses are administered to laboratory animals and gradually decreased until no effect on the endpoint in question is observed (Andrade et al. 2006). This is referred to as the ‘no observed adverse effect level' (NOAEL). A safety limit is then set below this, usually 100-fold lower. However, this method is problematic when applied to EDCs, which often display non-monotonic dose response curves (NMDRC). NMDRCs describe situations where the effect on a particular endpoint does not increase proportionally with the dose. For example, as has been suggested for BPA, adverse effects may be observed at low doses but not at moderate doses. There is reasonably strong evidence that this applies to BPA. BPA is accepted as having effects on conventional toxicological endpoints at higher doses. Currently, FSANZ assumes that BPA is safe at lower doses. However, there is a growing body of evidence showing adverse effects on hormonal endpoints from very low doses of BPA. Since this effect was first observed from animal studies there has been evidence from in vitro studies showing biochemical mechanisms for this. One of the most important of these mechanisms is the ‘spare receptor hypothesis’ (Vandenberg et al. 2012b). Hormones bind receptors on cell membranes sending intracellular signals. Evidence from in vitro studies indicate that the maximal response occurs when 1%–50% of the available receptors are bound (Welshons et al. 2003). Increasing the dosage beyond this point floods the system and results in a lower response. As BPA mimics 17β-estradiol, small doses may have a greater effect than larger doses leading to an NMDRC. While the low-dose hypothesis remains controversial for BPA, in vitro research and animal studies have shown its plausibility and likely effects, and it has been generally accepted for a range of other chemicals and 17β-estradiol.

Regulatory approaches

Faced with conflicts over the validity of science, regulators generally have two broad approaches to utilise: ‘weight of evidence’ and the ‘precautionary principle’. These are very broadly defined but cover fundamental differences in the approach to uncertainty. Weight of evidence approaches use regulation when a clear link has been established between a chemical and its effect (Edge & Eyles 2013). This approach can be vulnerable to ‘doubt production'—scientific research that confuses the issue—and this can subvert strong evidence resulting in a lack of regulation (Michaels 2008). The precautionary principle differs in that it does not consider uncertainty a justification for inaction (Gillespie 2011), although there is a range of ways in which this can be implemented. This approach allows regulators to take action based on reasonable evidence of harm, minimising use of the chemical until more robust evidence is available.

Regulation of BPA internationally has generally involved the precautionary principle. For example, regulation of BPA baby bottles in Canada occurred through the implementation of a highly precautionary chemical management plan (CMP) (Scott 2009). In New Zealand the FSANZ risk management framework (FSANZ 2013) contains precautionary principles, although the term precaution is not explicitly used. This framework encourages the regulation of novel chemicals for which there is scientific uncertainty regarding effects. The fact that precaution has not been implemented in the case of BPA perhaps reflects that the uncertainty comes from scientific disagreement as opposed to a lack of scientific research on the subject, and that a large number of studies have been rejected as invalid. As precaution is generally used as an interim measure until research can be completed (Gillespie 2011), in this case precaution would be used until scientific certainty has been reached. However, defining this point of scientific certainty is difficult, and there is an ongoing debate over how to define what is ‘good science’ and how this can be used reliably.

Politics of science

Deciding what studies are ‘good science’ that can reliably be applied in national policy and legislation is key to defining BPA's safety. The two main methods used to ensure reliability of results are peer review and GLP guidelines. GLP guidelines are a set of quality standards for research laboratories introduced in 1978 (Vogel 2009). These GLP guidelines regulate aspects of experiment design and strict record keeping. This became accepted as the gold standard of research by regulatory agencies that have given more weight to studies complying with GLP (Myers et al. 2009b). This is in contrast to peer review studies that allow greater flexibility in experiment design and operation, with quality assurance provided by review of other experts in the field. GLP is generally used by industrial labs, as the requirements are difficult to meet for public laboratories and can be restrictive in experiment design (Myers et al. 2009b). GLP has been criticised for not necessarily ensuring good science as much as creating records of what was carried out and under what conditions to allow verification, as well as its inadequacy for dealing with EDCs.

The importance of the type of quality assurance used can be seen in a breakdown of research on BPA. As of 2005, the 11 major industry-conducted studies that used GLP have not found evidence of low-dose effects of BPA. This can be compared with 115 peer-reviewed publicly funded studies on low dose effects of BPA, of which 94 found showed some effects, 31 of these significantly below the Environmental Protection Agency's (USA) reference dose of 50 µg/kg bodyweight/day (vom Saal & Hughes 2005). Major reviews have been carried out by the FDA, NTP and WHO, all of which disputed low-dose effects that gave preference to GLP studies. The arguments against BPA's safety were consolidated at the Chapel Hill conference, in which 38 researchers, from a range of relevant fields including toxicology and endocrinology, evaluated evidence relating to BPA and concluded with ‘confidence’ that BPA could cause negative health impacts at low doses (vom Saal et al. 2007). This differed from the FDA, NTP and WHO reviews in that peer-reviewed studies were also used and that GLP studies were not given preference. The FDA, NTP, WHO and Chapel Hill reviews were conducted at a similar time, with the same information available. As the choice of quality assurance system used can have such a major impact on review findings, these systems must be critically analysed.

GLP studies offer larger, easily reproducible studies, but may not be suited for dealing with EDCs. Conversely, peer-reviewed studies are able to better respond to the new challenges EDCs present but may not be as large or as well regulated. The GLP studies conducted have not considered hormonal endpoints and have administered BPA orally (Myers et al. 2009b). However, many of the publicly funded studies used hormonal endpoints, which they considered more appropriate for EDCs, and administered BPA intravenously. The use of non-hormonal endpoints by GLP studies has been criticised (Myers et al. 2009b) and GLP guidelines pre-date research on EDCs. The insistence on orally administered doses by GLP studies may no longer be an appropriate model as the public is exposed to BPA entering the bloodstream from thermal paper. In a clear challenge to GLP, 28 scientists, many of whom were present at Chapel Hill, co-published an article detailing the failings of GLP and the relative benefits of peer review (Myers et al. 2009b). Peer review can also be problematic, as there can be variability in the quality of reviews. However, the fact that a large number of studies showing low-dose effects have been published in reputable journals allays this concern. While GLP offers many benefits, its use as a gold standard for research on EDCs should be treated with caution.

In New Zealand, FSANZ gives priority to studies conducted under GLP guidelines and will often use them exclusively in its decision making (FSANZ 2013). FSANZ has published responses to studies raising concerns about BPA (FSANZ 2011). Common criticisms of the studies include the use of non-conventional endpoints and administering BPA doses intravenously, not orally. These statements appear to be based on adhering to GLP protocols, rather than a critical evaluation of the techniques used. As a result, a large number of studies have been excluded from risk analysis reviews, heavily distorting the findings. Unlike many regulatory groups overseas, New Zealand government agencies are not legally bound to use GLP research (IANZ 2015) and so are able to consider other studies. A comprehensive review of health and safety data on BPA must incorporate non-GLP studies and critically evaluate whether GLP guidelines are appropriate for EDC research.

Public response

A number of bans overseas, in particular many of the state-level bans in the USA, appear to have been as much a result of public pressure and activism influencing political activity, as science directly affecting regulatory activity. In the USA, there have been well-coordinated public campaigns against BPA (Lubitow 2013), as well as high levels of media coverage (Brewer & Ley 2011; Kiss 2013). Kiss (2013) showed that there was a significant association between the number of articles on BPA in the media within a state and the likelihood of that state regulating BPA. Many of these state-level bans contradicted the position of federal-level groups such as FDA, NTP and the Environmental Protection Agency which considered BPA safe based on the current scientific studies. Lofstedt (2013) argued this was a result of public responses to perceived rather than objective risk. In New Zealand, FSANZ (2015) has asserted that regulation overseas is the result of public pressure and not science, citing this as the reason it has not needed to implement similar regulation. This separation of public opinion and science is a flawed simplification.

Separating science and public opinion, however, is a false dichotomy and the origins of negative public opinion must be examined. Lubitow's (2013) study on social campaigns for BPA regulation in the USA focused on the relationship between scientists investigating BPA and social movements, showing their collaboration on setting the agenda for the campaign. Lubitow noted that many scientists who were concerned about the health effects of BPA had become politicised in their push for regulatory action. This, in itself, is indicative of flaws in the regulatory model, as scientists felt the need to use political action instead of scientific evidence to achieve regulation. Scientists partnered with environmental health advocacy groups, providing scientific expertise and support for their campaigns (Lubitow 2013). That these campaigns have received significant public support despite the evidence of BPA's safety from the NTP, FDA and the Environmental Protection Agency may indicate a lack of faith in these agencies or that the public prefers a more precautionary approach. Viewed from this perspective, public pressure on regulation was not a move away from science, but an alternative indirect route for bringing low-dose scientific evidence into the regulatory debate in a precautionary fashion.

Public awareness of BPA in New Zealand

BPA has had a relatively low level of exposure in national and local media compared with many countries and states where regulation has been introduced. The main journalist to report on BPA is Tracey Barnett, who published a number of articles on the subject in the New Zealand Herald (see Barnett 2008, 2009, 2010). A small number of other articles which mention BPA in limited depth have appeared in mainstream media (see Griffin 2010; Orr 2012; Moylan 2013). Barnett is a Kiwi-American journalist who was aware of the large presence of BPA in American media and the lack of coverage in New Zealand and her articles are a response to this gap (T. Barnett, pers. comm. 2014). BPA coverage in the New Zealand media is far lower than what has been shown in American studies (Brewer & Ley 2011; Kiss 2013) that quantified the presence of BPA in the media. This is likely to be linked to the lack of public discussion of BPA in New Zealand.

There are a range of groups in New Zealand that have voiced concern over BPA. Sue Chetwin of Consumer NZ has called for a ban of BPA bottles (NZPA 2010b). Despite groups such as ZeroWaste and the Cancer Society indicating their concern (Cancer Society of New Zealand 2012; Parker 2012), BPA has never appeared to have become a priority issue for these organisations. Catherine Delahunty of the Green Party has advocated for banning all EDCs and using safer substitutes instead of BPA (Delahunty 2010, 2013). However, while many of the state-level bans in the USA have been the result of political votes, there is no impetus for this in New Zealand. In the USA, the strength of the campaign against BPA appears to be a result of concerned and vocal scientists partnering with pre-existing environmental health advocacy groups (Lubitow 2013). Some New Zealand scientists, principally toxicologist Professor Ian Shaw, have publicly voiced concern about BPA in the media (Radio NZ 2008; Barton 2010). However, there does not seem to have been the same partnership between scientists and environmental health groups as seen in the USA. Most regulation overseas has been based on a precautionary approach and the impetus to make this decision is largely linked to public awareness and concern. It would appear that often the public has a more precautionary leaning than their respective regulatory agencies. The New Zealand government has not fully engaged with international debates over BPA's safety, and the lack of public discussion and awareness has meant that there is little impetus for this situation to change.

New Zealand regulation

Currently New Zealand has not taken steps to regulate BPA use, although limited voluntary phase-outs of BPA baby and sports bottles have occurred. New Zealand has accepted the safety limit of 50 μg/kg bodyweight/day, as stated by FSANZ and the Ministry for Primary Industries (MPI—into which the former New Zealand Food Safety Authority merged). BPA exposure below this level has been declared safe, and as exposure in New Zealand is indeed lower, no regulation has been implemented. A number of different groups are responsible for evaluating BPA in New Zealand: FSANZ and MPI are responsible for the safety of food in containers; chemicals in plastics are monitored by the Environment Protection Authority; general product safety is the responsibility of Consumer Affairs, part of the Ministry of Business, Innovation and Employment (MBIE); the Ministry of Health monitors dental sealants. Of the groups responsible, FSANZ and MPI have paid BPA the most attention, publishing reports and explanations of their opinions for both scientific and general audiences. Neither the Environmental Protection Authority, Consumer Affairs nor the Ministry of Health have publicly available information on BPA.

The assessment of BPA's safety in New Zealand, and that of xenoestrogens in general, has been rather piecemeal. The safety of BPA is not considered by one single agency, and the additive effects of BPA with other xenoestrogens have not been fully considered. For example, FSANZ has conducted a comprehensive review on the safety of BPA in food and has declared it safe. However, the FSANZ review has not considered studies with non-oral routes of exposure, many of which have indicated negative health impacts at current doses, as these are not relevant to food safety (FSANZ 2011). BPA does enter the body from non-oral pathways, such as dermally from thermal paper. However, thermal paper is the responsibility of the Environmental Protection Authority and not FSANZ, which has not published any reviews of BPA safety. As BPA is one of a number of xenoestrogens, assessing the total oestrogen load the body is exposed to will require all xenoestrogens to be considered (Shaw 2014). Current assessments of BPA do not fully consider the additive impacts of other xenoestrogens. This piecemeal approach to assessing safety, and neglecting to consider additive effects of other xenoestrogens, shows a lack of robustness in the current regulatory framework.

FSANZ has conducted the most comprehensive review of BPA safety in New Zealand. FSANZ is a bi-national agency created to harmonise food standards between Australia and New Zealand. However, New Zealand is not bound to use the same regulations as Australia and is able to apply for exemptions to standards. FSANZ sets food standards, and in New Zealand these are enforced by the MPI. FSANZ uses the minimum regulation necessary to ensure a safe food supply, while simultaneously using precaution to manage potential risk. FSANZ aims to harmonise its food standards with food standards set by foreign governments and international organisations (FSANZ 2013). Despite regulation of BPA overseas, such as BPA being labelled ‘toxic’ by the Canadian CMP (Scott 2009), neither FSANZ nor the MPI have followed this example as they consider these bans to be based on public opinion and not science (FSANZ 2015). This, however, contradicts the evidence provided by Lubitow (2013) which shows that scientists in the USA are heavily involved in activist movements for more stringent regulation of BPA. Most regulation overseas has occurred through using a precautionary approach. Groups such as the FDA and Environmental Protection Agency in the USA have not regulated BPA, however, they have been criticised for not using precaution in their decision-making process (Vogel 2013). FSANZ has a precautionary framework but needs to base its decision on science. By a strict adherence to GLP standards without critically analysing their appropriateness, FSANZ has not considered the growing body of evidence pointing towards harmful effects from BPA exposure. A more balanced review of the evidence would consider peer-reviewed studies in its analysis, and not just GLP studies, with consideration of whether a precautionary principle should be applied.

There has been some voluntary industry regulation of BPA in New Zealand. In 2010, several producers voluntarily phased out BPA baby bottles (NZPA 2010a). In 2008, following regulation of BPA bottles overseas, unsellable stock of these bottles was offered to a New Zealand supplier at an 85% discount (Barnett 2009). The major importer of these, Mark Ward, considered them safe based on information provided by FSANZ. However, after being approached by Tracey Barnett about health concerns he has since decided to stop selling them. Ward has expressed frustration with FSANZ and the lack of a clear indication of their safety (Barnett 2009). Ward has replaced his stock with triton bottles, which do not contain BPA or BPA derivatives and are generally considered safe (Barnett 2009) (Fig. 2). While voluntary recalls can be positive they are not comprehensive and may not respond to areas of high risk but low awareness such as thermal paper. There is also the risk of more dangerous but less well-known plastics, such as the closely related bisphenol S (BPS), being used as a replacement (Barrett 2013). Voluntary recalls remain a constructive way of reducing exposure to BPA, but would benefit from greater regulatory oversight.

Figure 2 A, Bisphenol A; and B, structurally similar bisphenol S.

As there is no requirement for items containing BPA to be labelled as such, it is impractical for concerned consumers to avoid BPA exposure. In New Zealand, tin cans, jar lids, stretch film and thermal paper do not state whether BPA is used. Many baby and drink bottles state that they are ‘BPA free’, however there is not always a description of what material has been used instead. While food itself is rigorously labelled, paradoxically, food containers are not. There have been growing calls for packaging labelling, especially with regard to nanomaterials in packaging (Walker 2014). The ingredients of dental sealants are listed and it is possible to tell if BPA was a precursor to the sealant, in which case residual BPA is very likely to be present. There are dental sealants branded ‘BPA-free’ available in New Zealand (Gunzdental 2010). Improved labelling of packaging would assist concerned consumers to avoid BPA through their purchasing choices.

There are a number of potential problems associated with regulating BPA. Alternative products may not necessarily be safer and BPA does have a number of important applications. BPA-based food containers can reduce the risk of pathogenic bacteria in food (McKenzie 2010). BPA dental sealants do contribute to oral health (Kloukos et al. 2013). An immediate ban on BPA use for these applications could have some negative health impacts. However, regulation can be targeted at high-risk/low-benefit areas. These could include products for babies, as infancy is a crucial phase in development when the risk from BPA is higher. BPA use in thermal paper is relatively high risk due to its non-oral route of administration, but offers no benefit to health outcomes. In lower risk areas, such as for the lining of tin cans, regulation could be implemented slowly, allowing time for safer alternatives to be introduced. A number of products in the USA labelled ‘BPA-free’ have been found to be based on BPS, a very similar compound with potentially worse health effects (Barrett 2013) (Fig. 2). Regulation focused solely on BPA and not the alternatives could be counterproductive. While regulation can have negative impacts, this can be managed by targeted regulation of high-risk/low-benefit applications of BPA and by careful consideration of available alternatives.

Conclusion

Despite extensive research, the safety of BPA remains uncertain. This is highly concerning considering the widespread exposure of the New Zealand population to BPA. Many of the challenges BPA presents are due to it acting as an endocrine disruptor and not conforming to traditional toxicological models. In New Zealand, regulatory agencies have not fully engaged with debates over endocrine disruption at low doses. Studies linking BPA to negative health impacts have been dismissed for not being conducted under GLP guidelines, despite arguments that these guidelines are not appropriate for EDCs. The additive effects of BPA contributing to the total xenoestrogen burden have not been fully considered. Other government agencies responsible for BPA, including the Environmental Protection Authority and Ministry of Health, have not published information on BPA's safety. The assessment of BPA's safety has been conducted in a piecemeal fashion, with several different government agencies responsible for assessing different applications of BPA. Additive effects with other xenoestrogens, or even of BPA in different applications, have not been fully considered. Internationally, bans have been introduced through the use of the precautionary principle and social activist campaigns against it. New Zealand has the ability to apply a precautionary approach but has yet to do so. There has been relatively little public discussion of, or opposition to, BPA, reducing the impetus to reconsider the status quo. As defining the safety of BPA is important to human health outcomes in New Zealand, both the New Zealand government and the public would do well to better engage with these debates.

Acknowledgements

We would like to thank everyone who took the time to comment on the draft of this article.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the Massey University Research Fund [Grant number: RM17259].