Human risk assessment of multiple chemicals using component‐based approaches: A horizontal perspective

This technical report implements the EFSA MIXTOX guidance document for human risk assessment of combined exposure to multiple chemicals(i.e.regulated compounds and contaminants)using component‐based approaches.Alow tier risk assessment methodologyfor regulated compounds and contaminants with non‐cancer effects, is describedusing available harmonised methods, tools and data sources.

For exposure assessment, exposure metrics can be collected or calculated from previous EFSA opinions or monitoring data from available tools(e.g. EFSA pesticide residue information, occurrence data from Member States, EFSA comprehensive consumption database, JECFA and JMPR monographs etc..). Harmonisation of the reporting of exposure metricsis proposed using the upper bound and the lower bound of the 95^th centiles.For hazard assessment,it is assumed that for all compounds reference values as health‐based guidance values (HBGV)are available and represent conservative estimates for regulated products and contaminants with non‐cancer effects.The OpenFoodTox database is proposed as a simple open source tool which provides HBGVsfor acuteand chronic effects asacute reference dose(AfRD) and the acceptable or tolerable daily intake(ADI, TDI) respectively. For risk characterisation, the Hazard Index (HI) is applied as a low tier harmonised approach using the dose addition default assumption for combined toxicity.HIs are derived as the sum of the ratios between hazard and exposure metrics. EFSA's MIXTOX reporting table summarising the findings of the risk assessment is provided. For non‐cancer effects, HIswith values below 1 are interpreted as of low concern whereas HIsabove1 suggest either the need to refine the risk assessment or a risk management consideration.options for refining the approaches include exposure‐criteria,risk‐criteria and grouping based on specific toxicological effects as well as the use of toxicokinetic data and physiologically‐based kinetic models.