On December 5, 2013 the peer-reviewed journal Packaging Technology and Science published an article online on “Detection and Identification of Oestrogen-Active Substances in Plastic Food Packaging Migrates” by Christian Kirchnawy and colleagues from the Austrian Research Institute for Chemistry and Technology in Vienna. Reviewing the literature, the researchers state that endocrine disrupting chemicals (EDCs) are linked to a row of symptoms like precocious  puberty in females, diminished sperm quality in male, obesity as well as high incidences of various types of cancers, observable in both mammals and humans. For the present study the scientist tested samples of plastic food contact materials (FCMs) for migration of estrogen active chemicals. Food was said to be a preeminent exposure source, where FCMs, raw materials and production process alike are considered as sources of EDCs. In contrast to most studies cited in the present work, which tested migration in FCMs that were in contact with food stuff prior to testing, Kirchnawy and colleagues only tested unexposed materials, either granulate or finished packaging received directly from manufacturer or retailers. The aim was to exclude contamination of samples with substances absorbed from food stuff.

The migration experiments were carried out in accordance with the European Commission Regulation Number 10/2011 and EN 1186 at 60°C for 10 days in either 10-95% ethanol or ultrapure water with five grams of each granulate and 100cm2 of each composite films (CF). The tested FCM samples included polyethylene terephthalate (PET), polypropylene (PP), polyethylene (PE), polystyrene (PS) and CF. Migration extract analysis was performed via a yeast estrogen screen (YES), which was preceded by a concentration step applying solid phase extraction (SPE) and dilution in the solvent dimethyl sulfoxide (DMSO). The constituents of positive samples were subsequently identified by gas chromatography-mass spectrometry (GC-MS) and high-performance liquid chromatography-mass spectrometry (HPLC-MS).

None of the PET samples, of which some included high density polyethylene (HDPE) used for screw caps, displayed measurable estrogen activity at the limit of detection (LOD) of 0.2 ng estradiol equivalents (EEQ) per liter. Contrary to other recently published studies, Kirchnawy and colleagues reasoned that PET bottles are no significant contributor to potential ECDs in mineral water. Out of 7 PP, 8 PE and 9 PS samples only one of each was tested positively in the YES bioassay at levels of 1.9 ng EEQ/l, 0.7 ng EEQ/l and 2.1 ng EEQ/l, respectively. For composite films, 4 out of 9 samples displayed estrogen activity; the highest having a level of 59 ng EEQ/l.

Scanning positive samples of migrates from the YES bioassay for estrogen active chemicals, the researchers found in the CF sample with the highest activity the potentially estrogen active plasticizer diethylhexyl adipate. Its estrogen activity could, however, not be verified in the bioassay. The same holds true for the second CF sample. Here the ultra violet absorber bumetrizole (2-(2’-hydroxy-3’-tert-butyl-5-methylphenyl)-5-chlorobenzotriazole) was measured, but its estrogen activity could not be confirmed, either. For both samples the estrogen activity remains unclear. In a further positive CF sample the antioxidant butylated hydroxytoluene (BHT) was identified, whose estrogen activity is described in the literature. The fourth positive sample was left uncomment concerning its active agents. For the positive PS sample GC-MS identification yielded three distinct styrene dimers. For two of them, 1,3-diphenylpropane and 2,4-diphenylcyclobutane, results are controversial. Some studies showed that they are estrogen active only in metabolized forms, whereas in others also the un-metabolized forms displayed estrogen activity. For the third styrene dimer, the diels-alder dimer 1,2-dihydro-3-phenylnaphtalene, no information about estrogen activity is available. In the positive PP sample the estrogen active plasticizer dibutylphthalate was found and its estrogen activity was confirmed in the YES bioassay. Furthermore, degradation products of the antioxidants Irganox 1010 were found among migrants: 3,5-di-tert-butyl-4-hydroxybenzaldehyde, 3,5-di-tert-butyl-4-hydroxyacetophenone, 7,9-di-tert-butyl-1-oxaspiro(4,5)deca-6,9-diene-2,8-dione and benzenepropanoic acid, 3,5-bis(1,1-dimethylethyl)-4-hydroxy-,methyl ester. Even though their chemical structure might imply estrogen activity, no data is available for these chemicals. Screening of PE migration samples showed the presence of 2,4-di-cumylphenol, which the researchers assumed to be a breakdown product of the antioxidant Alkanox 28 (bis(2,4-dicumylphenyl)-pentaerythritol-diphosphite) and which was tested positively in YES. A further degradation product of 2,4-di-cumylphenol, 4-cumylphenol, showed even stronger estrogen binding.

Kirchnawy and colleagues rate their presented procedure combining migrant extraction, SPE and in vitro testing as suitable for assessing endocrine activity of plastic FCMs, being both sensitive and showing good reproducibility. As some migration extracts inhibited the YES bioassay, those had to be diluted for testing in the bioassay, resulting in additional spiking experiments with estradiol for each sample in order to prevent false negative results. Limiting the explanatory power of the procedures for some migranting chemicals, however, was the lack of analytical standard purchasable as a reference, which hindered the testing of their estrogen activity.

In summary, Kirchnawy and colleagues state that their procedure was able to show that packaging can be a source of estrogen-active substances. Their measurement yielded lower concentrations compared with studies measuring the estrogenic activity of bottled mineral water. Interestingly, none of the PET samples, including recycling flakes, showed estrogen activity. This may mean that PET bottles may not be the source of high estrogen activity measured in mineral water, as presented in recent studies. An alternative possibility is that the applied test system was not sensitive or selective enough to identify estrogenic activity from PET migrates.


Kirchnawy, C. et al. (2013). “Detection and Identification of Oestrogen-Active Substances in Plastic Food Packaging Migrates.” Packaging Technology and Science (published online December 5, 2013).