Food contact materials and human health: a new challenge for epidemiological research

As ubiquitous as particulate air pollution (or more), but until recently with a much lower profile, food contact materials (FCMs) have long posed a silent challenge to researchers concerned with human health, nutrition and the environment. FCMs are articles used in packaging, food storage, processing or preparation equipment that come directly into contact with human foods. Most often FCMs are made of plastic or have a synthetic material in direct contact with the foodstuff—for example, as can coating, laminate in beverage cartons or the closures of glass jars. Importantly, most FCMs are not inert. Chemicals contained in the FCM, such as monomers, additives, processing aids or reaction by-products, can diffuse into foods.6 ,7 Known as migration, this chemical diffusion is accelerated by increased temperatures and depends on storage time, chemical properties of the FCM and the foodstuff, as well as on the physical characteristics of the FCM (pore size, thickness and surface area).6 ,7 Some, but not all FCM migrants, are regulated—for example, as indirect food additives (in the USA).

FCMs are a significant source of chemical food contamination,8 ,9 although legally they are not considered as contaminants. As a result, humans consuming packaged or processed foods are chronically exposed to synthetic chemicals at low levels throughout their lives,10 including the most sensitive periods of development. These facts may be of relevance to scientists interested in the developmental origins of health and disease hypothesis (DOHaD), life-course effects of in utero and childhood environmental exposures, plasticity, epigenetics and related processes.4 ,11–18 Thus, FCMs are a new exposure source in the sense that they have received little attention so far in studies concerned with human health effects. Their integration into epidemiological and non-epidemiological research is highly relevant. The dearth of epidemiological publications on FCMs is surely not justified on scientific grounds.

Lifelong, low-dose exposure to FCMs is of concern for several reasons. First, acknowledged toxicants are legally used in FCMs in Europe, the USA and other regions (notably, China). In the USA, several types of asbestos are authorised as indirect food additives for use in rubber.19 Formaldehyde, another known carcinogen, is widely present at low levels in plastic bottles made of polyethylene terephthalate20; formaldehyde also migrates from melamine formaldehyde tableware.21 Considering how widely beverages are consumed from polyethylene terephthalate soda bottles, this may amount to a significant, yet unrecognized, exposure of the population.

Second, numerous controversially discussed chemicals are present in FCMs. Several of these are endocrine-disrupting chemicals (EDCs).22–24 For example the EDCs nonylphenol, bisphenol A, tributyltin, triclosan and several different phthalates25–28 are legally and intentionally used in FCMs in Europe or the USA. Whereas the science for some of these substances is being debated and policy makers struggle to satisfy needs of stakeholders, consumers remain exposed to these chemicals daily, mostly unknowingly.

Third, the total number of known chemical substances intentionally used in FCMs exceeds 400029 ,30; in addition, FCMs also contain an unknown number of polymerisation by-products, impurities and breakdown compounds7 ,31—collectively known as non-intentionally added substances (NIAS). Improvements in analytical chemistry have led to the constant reduction of detection limits, thereby disclosing the migration of NIAS into food.7 ,32 ,33

Given the low levels of toxicants generally found in foods, the difficulty of analysing chemicals in a complex food matrix and the considerable effort required for development of an analytical method, it is not surprising that little is known about most NIAS. Especially, their toxicological hazards often remain unknown, while both industry and regulators are struggling to ensure safety of marketed products using exposure assessment and chemical risk assessment concepts on unknown compounds.34 FCMs are another relevant source of widespread exposure to chemical mixtures.

At the same time, chemical risk assessment is being challenged by several recent scientific findings dealing with chemical toxicity:

EDCs mimic hormones’ property to affect biological systems at low doses, thus causing subtle changes that may lead to adverse effects at later stages in life.26 ,27 ,35 ,36 Research on the DOHaD has shown the fragility of early-life stages to chemical exposures.14 ,17 ,37 ,38 A consequence of such exposure in the womb might be chronic disease later in life. Furthermore, the observed effects may follow non-monotonic dose–response curves, thereby defying current practice of testing at high doses to extrapolate to the low doses of actual exposure.35 What is more, EDC-induced physiological changes are not among the topics being considered by common toxicology, which casts serious doubts on the adequacy of chemical regulatory procedures.36 ,39 ,40 We therefore propose to call EDC effects neotoxic, thereby capturing their unique properties, mechanisms of action and effects, as well as the obligation to think outside traditional mechanistic and risk assessment paradigms when examining their chemical risk. Accordingly, neotoxicants are synthetic chemicals that cause adverse effects through mechanisms different from those commonly tested by traditional toxicology and which have been introduced into the anthroposphere through industrialisation and weak global regulation.

Chemicals targeting the same site of action are known for their ability to act additively when present in mixtures.41 ,42 Chemical risk assessment practices assume that there is a threshold for exposure to an individual chemical below which the chemical's toxicity is considered unproblematic. In Europe, chemical migration from FCMs into food resulting in levels <10 ppb is not considered toxicologically relevant.43 For some reason, it is also assumed to be clinically irrelevant. However, humans are not exposed to single chemicals in isolation. Especially for FCMs, many different substances migrate, but are not necessarily detected.44 Indeed, several studies, using in vitro assays, have shown that the total toxicity of all migrates from a given FCM cannot be fully explained by the known/identified migrants.45 ,46

Establishing causality between lifelong (and largely invisible) exposure to FCMs and human chronic diseases is challenging for several reasons, including the fact that no reference populations are completely unexposed to FCMs—everybody is exposed to synthetic chemicals from FCMs, usually at low doses. What is more, large interindividual and social differences in internal concentrations of food contact substances may exist in most populations, as is the case for commonly detected environmental contaminants in foods and people.47–49 Progress is thus urgently needed in population-based exposure assessment and biomonitoring of FCMs. It is a major challenge—for epidemiology, toxicology and other health and life sciences—to tease out the cause–effect relationships between food contact chemicals and chronic diseases like cancer, obesity, diabetes and neurological and inflammatory disorders.13 ,16 ,17

Epidemiology can help to improve knowledge on the role of food contact chemicals in diseases of complex aetiology

In the developed world chronic diseases are responsible for around two-thirds of deaths, with about 16% of such deaths occurring before age 60.50 While most chronic, non-communicable diseases are rightly considered ‘diseases of complex aetiology’ (and, therefore, have multiple causes), there is strong evidence linking these disorders with chronic exposure to environmental pollutants.26 ,51 The WHO and the United Nations Environment Programme (UNEP) recently concluded in their ‘State of the Science on Endocrine Disrupting Chemicals 2012’ report that EDCs are a global public health threat.50 EDCs and other neotoxicants are commonly used, or present, in FCMs22–24; their safety for this use has often not been established.52 The direct health consequences of this exposure to neotoxicants via FCMs are unknown. Since most foods are packaged53 and the entire population is likely to be exposed, it is of utmost importance that gaps in knowledge are reliably and rapidly filled.

Unravelling the role of FCMs in the development of chronic disease is of high scientific and public interest. In contrast to other challenges in nutritional and environmental epidemiology, chemical exposures from FCMs offer the benefit of a relatively discrete and measurable route of exposure. Methodological progress is feasible. We propose, specifically, that in addition to using food frequency questionnaires and other dietary assessment methods (dietary intake records, 24 h recalls) and technologies,54 dietary habits should additionally be characterised according to FCMs and supplemented by biomonitoring efforts. Such a task will include analyses of the uses of materials in contact with food throughout the food supply chain (processing, packaging, storage) and food packaging in stores, at home, the workplace and other settings. Furthermore, studies should also measure—through validated instruments and procedures—the frequency of consumer practices, such as storage in freezers, heating foods in plastic dishes and containers, use of plastic films, as well as packaging preferences when buying foods and beverages (eg, higher or lower preference for unpackaged foods, glass, cans and plastic packages). In Europe, for example, the FACET database can support such efforts: this newly established database from the EU-funded research project Flavorings, Additives and Food Contact Materials Exposure Task (FACET) contains levels of food packaging migrants from FCMs and links them with food consumption data.55 ,56 Subsequently, statistical analyses would integrate this type of information with data traditionally used in nutritional, environmental and molecular epidemiology.

Innovative research could also expand knowledge on toxic mechanisms—for example, on oestrogenic, androgenic, thyroid and glucocorticoid effects of chemicals migrating from FCMs; on the homeostasis of glucose and lipid metabolism, energy homeostasis and insulin resistance; on the role of agonists and antagonists of nuclear receptors in modulation of nuclear receptor function and endocrine diseases, including non-nuclear steroid membrane receptors and non-steroid receptors; on metabolic and mitochondrial dysfunction, inflammation, adipogenesis and adipose macrophages.12–14 ,16 ,17 ,26–28 ,35 ,36 ,41 ,42 ,57

Also, given the economic and cultural influences on food consumption, social epidemiology should develop a research agenda on FCMs, health and well-being.

Integrating knowledge about FCM chemical composition and migration into food in epidemiological studies is in our view an opportunity and a duty for epidemiologists. Eventually, such research will strengthen primary prevention policies by reducing chemical exposures resulting from a manageable source. It will also advance basic and applied knowledge on the molecular and physiological mechanisms that link some environmental chemicals and human diseases.