Risk Assessment of Food Contaminants: A Systematic Review of Toxicological Parameters, Analytical Models, and Reference Values

Résumé

 Food safety is now a major public health issue due to the wide range of chemical contaminants present along the food chain, including heavy metals, pesticides, mycotoxins, emerging pollutants, and process-related contaminants. This systematic review strictly followed the PRISMA 2020 methodology, relying on three major databases (ScienceDirect, PubMed, Scopus) and applying explicit inclusion and exclusion criteria. Out of 560 references identified, 31 articles were retained after full screening and analysis. The findings reveal a predominance of deterministic approaches: Estimated Daily Intake (EDI) or Average Daily Intake (ADI) were applied in 70.96% of studies, Hazard Quotient (HQ) in 61.29%, and Cancer Risk (CR/ILCR) in 45.16%. Probabilistic approaches also appeared significant, with Monte Carlo Simulation (MCS) used in 48.38% of cases and Margins of Exposure (MoE/MoET) in 32.25%. Advanced methods such as PBPK and QSAR models remain marginal, reported in only 3.22% and 9.67% of publications, respectively. From a toxicological perspective, several studies highlighted concerning risks, such as margins of exposure below 10,000 for aflatoxins or Hazard Quotients above 1 for lead and cadmium in children. Notably, only 6.45% of studies addressed risk assessment related to plastic migrants, underscoring a major gap in the literature. This review demonstrates substantial heterogeneity in the models applied and stresses the need for international harmonisation of toxicological reference values to improve comparability of results across regions and contaminant families. Probabilistic and integrative approaches are presented as a necessary pathway to enhance the robustness and relevance of food risk assessment.