Mineral oil hydrocarbons in foods: is the data reliable ?

The contamination of foods with mineral oil hydrocarbons (MOH) is a serious concern, requiring in most cases tedious mitigation measures that span across the whole food supply chain. A major issue today is the significant variability of the results generated by laboratories. This study was therefore designed to achieve a deeper insight into the analytical procedures used by commercial laboratories, identifying possible gaps and suggesting improvements that will enhance the reliability of the MOH data, an important prerequisite for risk assessment. In total six different food matrices, i.e. infant formula (IF), cocoa butter, cocoa powder, biscuits, fruit-based baby food containing biscuit and roast and ground coffee were subjected to comparative inter-laboratory studies, as well as one vegetable oil analysed within the frame of a professionally conducted proficiency test. The results indicate that on some matrices with possibly low amounts of MOH contamination, the current methodologies cannot reliably conclude whether or not a food sample is indeed contaminated with mineral oils (<10 mg/kg food). Urgently needed are: (i) an aligned and fully validated sample preparation strategy tested on a range of different food matrices; (ii) a confirmation of positive flame ionisation detection (FID) results by confirmatory methods such as mass spectrometry ? in line with the CEN Standard and the Joint Research Centre (JRC) Guidance Document, (iii) a more detailed root-cause analysis in the reports of laboratories through the use of mineral oil markers, and (iv) a fully validated official method for the concerned foods with a limit of application <10 mg/kg food.