Co-contamination of aquatic food animals by microplastics, heavy metals, and pesticides: bioaccumulation, human health risk, and role of artificial intelligence.

Aquatic food animals provide essential proteins, omega-3 fatty acids, and micronutrients to billions of people worldwide, yet their safety is increasingly challenged by the co-occurrence of microplastics (MPs), heavy metals (HMs), and pesticides in aquatic ecosystems. Unlike previous reviews that examine these contaminant classes separately, this review presents an integrated framework linking contamination sources, bioaccumulation, trophic transfer, human health risks, food security, and regulatory responses. MPs can act as physicochemical vectors by adsorbing HMs and pesticides and releasing them following gastrointestinal desorption (the Trojan horse mechanism), although the importance of this pathway depends on polymer characteristics, weathering, gut conditions, and ambient contaminant concentrations. Co-exposure may produce synergistic, additive, or antagonistic effects depending on contaminant combinations, exposure conditions, and biological endpoints, highlighting that synergism should not be assumed universally. Human health risk assessment indices are evaluated alongside international regulatory guidance, with emphasis on their limitations as screening-level tools that do not account for contaminant interactions and are sensitive to exposure assumptions. The implications of co-contamination for all four FAO pillars of food security are examined through regional case studies of seafood-dependent populations. Machine learning and explainable AI applications for contaminant detection, environmental monitoring, and bioaccumulation prediction are evaluated and compared across four application domains, distinguishing proof-of-concept studies from regulatory-ready applications. Finally, key research priorities, including mixture toxicology, standardized nanoplastic analysis, long-term biomonitoring in data-sparse regions, and robust ML model validation, are identified to support future research and evidence-based seafood safety governance.