Food Safety Risk Assessment in Developing Countries: Principles, Challenges, and Applications provides a comprehensive examination of risk analysis frameworks applied to food safety in low- and middle-income settings. The book explores the scientific foundations of risk assessment, including hazard identification, hazard characterization, exposure assessment, and risk characterization, while emphasizing their practical implementation within resource-limited environments.
It highlights the growing burden of foodborne diseases associated with biological, chemical, and environmental hazards, particularly in contexts where regulatory systems, surveillance capacity, and laboratory infrastructure remain underdeveloped. Key microbial risks such as Escherichia coli, Salmonella spp., and Campylobacter spp. are discussed alongside chemical contaminants including mycotoxins, pesticide residues, and heavy metals. The book underscores how factors such as informal food markets, inadequate sanitation, climate variability, and weak enforcement mechanisms exacerbate risk exposure across the food supply chain.
In addition to theoretical foundations, the book presents applied methodologies for both qualitative and quantitative risk assessment, incorporating case studies and real-world scenarios relevant to developing countries. It further discusses the integration of risk assessment into risk management and risk communication processes, highlighting the importance of evidence-based policymaking and multi-sectoral coordination.
By combining scientific rigor with practical relevance, this work aims to support researchers, regulators, and public health professionals in strengthening food safety systems and reducing the incidence of foodborne illness. It contributes to advancing risk-based approaches as a cornerstone of modern food safety governance in developing regions.
Preface
Food safety remains one of the most pressing public health challenges of the 21st century. Each year, an estimated 600 million people — nearly 1 in 10 individuals globally — fall ill after consuming contaminated food, resulting in approximately 420,000 deaths and the loss of 33 million healthy life years (disability-adjusted life years, DALYs). These staggering figures underscore the urgent need for robust, science-based approaches to identifying, evaluating, and managing the hazards that threaten the safety of ourfood supply.
Risk assessment has emerged as the cornerstone of modern food safety management. By systematically examining the nature of hazards, characterizing the populations likely to be exposed, and quantifying the probability and severity of adverse health outcomes, risk assessment provides a rational, transparent, and reproducible basis for decision-making. Yet the field is rapidly evolving. Traditional frameworks designed to evaluate single agents — a single pathogen, a single pesticide residue — are increasingly challenged by the complexity of real-world food systems where multiple hazards coexist, interact, and change in response to climate shifts, novel production technologies, and global trade dynamics.
This reference book was conceived to address that complexity head-on. It brings together the core principles and latest methodologies for assessing microbiological hazards, chemical contaminants, and the emerging hazards that are reshaping the risk landscape for both regulators and industry. The text is organized to serve a dual audience: practitioners who need practical guidance on applying risk assessment tools in their day-to-day work, and researchers who seek a rigorous theoretical foundation from which to advance the science.
Part I (Chapters 1-3) establishes the conceptual and regulatory foundations of food safety risk assessment. Part II (Chapters 4-6) delves into microbiological hazards, from classical foodborne pathogens to the growing concern of antimicrobial resistance. Part III (Chapters 7-9) covers chemical risk assessment, including contaminants, pesticides, food additives, and the unique challenges posed by mixtures. Part IV (Chapters 10-12) addresses emerging hazards — engineered nanomaterials, novel proteins, climate-driven changes in mycotoxin profiles, and cyberbiosecurity risks in smart food systems. Part V (Chapters 13-15) presents integrated risk management strategies, regulatory case studies, and future perspectives.
Throughout, emphasis is placed on quantitative methods, uncertainty analysis, and the communication of risk to diverse audiences. Worked examples, data tables, and decision frameworks are provided to facilitate application. A comprehensive glossary and curated reference list are included to support further exploration.
The authors are indebted to colleagues across academia, regulatory agencies, and the food industry whose insights, critiques, and contributions have shaped this work. In particular, we acknowledge the foundational contributions of the Codex Alimentarius Commission, the World Health Organization/Food and Agriculture Organization (WHO/FAO), the U.S. Food and Drug Administration (FDA), the European Food Safety Authority (EFSA), and the many national competent authorities whose published risk assessments and guidelines constitute the bedrock of the field.
We hope this volume will serve as both a reliable reference and a stimulus for continued innovation in food safety science. Protecting the food supply is ultimately a shared responsibility — one that demands the best available science, transparent communication, and unwavering commitment to public health.
Chapter 1: Foundations of Food Safety Risk Assessment
1.1 Historical Development and Conceptual Evolution
The discipline of food safety risk assessment has its roots in the broader field of toxicology, which has sought for centuries to understand the relationship between exposure to harmful substances and adverse health outcomes. The Latin aphorism attributed to Paracelsus — 'dosis sola facit venenum' (the dose alone makes the poison) — encapsulates the dose-response relationship that remains central to all risk assessment today. However, the formalization of risk assessment as a structured, reproducible analytical process is a comparatively recent achievement, emerging primarily from developments in environmental and industrial health regulation during the latter half of the 20th century.
The pivotal 1983 report by the U.S. National Research Council (NRC), 'Risk Assessment in the Federal Government: Managing the Process,' established a four-step paradigm — hazard identification, dose-response assessment, exposure assessment, and risk characterization — that continues to define the field. Although originally conceived for chemical hazards, this framework was subsequently adapted for biological agents, physical hazards, and, most recently, for the complex, multi-hazard scenarios that characterize contemporary food systems.
The Codex Alimentarius Commission formalized food-specific risk analysis principles in 1995, distinguishing three interlocking components: risk assessment (the science-based process of estimating risk), risk management (the policy process of weighing and selecting options to control risk), and risk communication (the exchange of information and opinions among risk assessors, managers, and stakeholders). This tripartite model remains the international gold standard for food safety governance.
1.2 The Four-Step Risk Assessment Paradigm
1.2.1 Hazardidentification
Hazard identification is the qualitative step in which agents — biological, chemical, physical, or radiological — capable of causing adverse health effects are identified. For microbiological hazards, this involves cataloguing pathogenic microorganisms and their toxins. For chemical hazards, it encompasses contaminants, pesticide residues, veterinary drug residues, food additives, and naturally occurring toxins. The hazard identification process relies on epidemiological data, outbreak investigation reports, animal bioassay data, in vitro studies, and mechanistic information on biological plausibility.
A critical distinction in hazard identification is between hazard — the intrinsic capability of an agent to cause harm — and risk — the probability that harm will actually occur under specific conditions of exposure. All toxic substances are hazards; whether they pose a meaningful risk depends on the likelihood and magnitude of exposure. This distinction is often lost in public discourse but is fundamental to proportionate risk management.
1.2.2 Hazard Characterization (Dose-Response Assessment)
Hazard characterization examines the quantitative relationship between the dose (amount of hazard ingested or inhaled) and the probability or severity of adverse health effects. For chemical agents, this step typically involves deriving a reference dose (RfD) or acceptable daily intake (ADI) — a dose below which no appreciable health risk is expected. For carcinogens, a cancer slope factor or benchmark dose may be derived.
For microbiological hazards, dose-response models are more complex because pathogens are capable of reproducing within the host. Three classes of dose-response models are widely used: the exponential model, the beta-Poisson model, and the log-normal model. These models differ in their assumptions about host-pathogen interactions and vary in their predictions at low doses — a region of particular regulatory significance where extrapolation uncertainty is greatest.
Key Concept: Threshold vs. Non-threshold Agents
Chemical hazards are generally classified as threshold agents (below a certain dose, no adverse effect is observed) or non-threshold agents (any dose carries some finite risk, as with genotoxic carcinogens). Biological hazards are typically modeled as non-threshold because theoretically a single viable pathogen, under the right conditions, can initiate infection. This distinction profoundly influences how acceptable intake levels are derived and communicated.
1.2.3 Exposure Assessment
Exposure assessment estimates the likely intake of the hazard by specific populations through specific pathways. In food safety, the primary exposure pathway is dietary: the product of the concentration ofthe hazard in food and the quantity of food consumed. Exposure assessments must account for variability in consumption patterns across demographic groups (infants, elderly, pregnant women, immunocompromised individuals), seasonal and regional variation in dietary habits, and the range of concentrations typically found in implicated foods.
Modern exposure assessment increasingly uses probabilistic methods — particularly Monte Carlo simulation — to characterize the full distribution of likely exposures rather than relying solely on deterministic point estimates. This approach better captures the true heterogeneity of human dietary behavior and food contamination levels, and provides risk managers with information about both central tendencies and extreme high-exposure scenarios.
1.2.4 Risk Characterization
Risk characterization integrates the outputs of hazard identification, hazard characterization, and exposure assessment into an overall estimate of risk to the population of interest. For chemical hazards, risk characterization typically involves calculating a hazard quotient (HQ = estimated daily intake / reference dose) or, for carcinogens, an incremental lifetime cancer risk (ILCR). For microbiological hazards, risk is expressed as the probability of infection or illness per serving, per year, or per lifetime, often disaggregated by age group and health status.
A comprehensive risk characterization also includes a transparent and thorough uncertainty analysis, documenting the key sources of uncertainty and variability in the assessment and their likely impact on the risk estimate. Sensitivity analysis — examining how changes in key input parameters affect the final risk estimate — helps prioritize data collection efforts and identify the assumptions most in need offurther investigation.
1.3 Variability and Uncertainty in Risk Assessment
A fundamental challenge in food safety risk assessment is the distinction between variability and uncertainty. Variability reflects true heterogeneity in the world — not all people eat the same amount of a given food, not all batches of a product have the same level of contamination. Variability cannot be reduced by collecting more data; it can only be better characterized. Uncertainty, by contrast, reflects a lack of knowledge — about the true dose-response relationship, about the concentration of a contaminant in a poorly sampled food commodity, about the fraction of illness truly attributable to a specific food. Uncertainty can, in principle, be reduced by additional research.
The failure to distinguish variability from uncertainty can lead to significant misinterpretation of risk estimates. For example, reporting a single 'worst-case' exposure estimate conflates high-end variability (the very highest consumers) with uncertainty (we don't know the true average) in a way that may produce grossly misleading risk estimates. Modern practice recommends using twodimensional (2D) Monte Carlo simulation to separately propagate uncertainty and variability through the risk assessment model.
Illustrations are not included in the reading sample
Chapter 2: International Regulatory Frameworks and Standards
2.1 The Codex Alimentarius System
The Codex Alimentarius Commission (CAC), established jointly by the Food and Agriculture Organization (FAO) and the World Health Organization (WHO) in 1963, provides the preeminent international forum for the development of food standards, guidelines, and codes of practice. Codex standards serve a dual function: they protect consumer health and ensure fair practices in food trade. Under the World Trade Organization (WTO) Sanitary and Phytosanitary (SPS) Agreement, Codex standards are recognized as the reference point for international food safety requirements, providing a significant legal and commercial impetus for national adoption.
Within the Codex system, risk assessment is the exclusive province of two joint FAO/WHO expert bodies: the Joint FAO/WHO Expert Committee on Food Additives (JECFA), which evaluates food additives, contaminants, and veterinary drug residues; and the Joint FAO/WHO Meetings on Pesticide Residues (JMPR), which evaluates pesticide residues. For microbiological hazards, FAO/WHO convenes ad hoc expert consultations and Meetings on Microbiological Risk Assessment (JEMRA) to provide scientific advice.
2.2 Regional Regulatory Architectures
2.2.1 European Union — EFSA and the General Food Law
The European Union's food safety regulatory system underwent fundamental reform following a series of high-profile food safety crises in the late 1990s, most notably bovine spongiform encephalopathy (BSE, or 'mad cow disease') and dioxin contamination. The resulting White Paper on Food Safety (2000) and the General Food Law Regulation (EC) No 178/2002 established a new paradigm in which risk assessment is functionally and institutionally separated from risk management. The European Food Safety Authority (EFSA), established in 2002 and headquartered in Parma, Italy, serves as the independent scientific body responsible for risk assessment, while risk management decisions remain with the European Commission, the European Parliament, and the Council.
EFSA's scientific work is organized through a series of panels covering different aspects of the food chain, including the Panel on Biological Hazards (BIOHAZ), the Panel on Contaminants in the Food Chain (CONTAM), the Panel on Plant Protection Products and their Residues (PPR), and several others. EFSA has developed a rigorous and transparent methodology for scientific opinions, including systematic review protocols, exposure assessment tools (notably the EFSA Comprehensive European Food Consumption Database), and standardized uncertainty analysis frameworks.
2.2.2 United States — FDA, USDA, and EPA
In the United States, food safety oversight is distributed across multiple federal agencies with overlapping but distinct mandates. The Food and Drug Administration (FDA), under the authority of the Federal Food, Drug, and Cosmetic Act (FD&C Act) and, more recently, the Food Safety Modernization Act (FSMA, 2011), regulates approximately 80% of the food supply, including all domestically produced and imported food except most meat, poultry, and egg products. The Food Safety and Inspection Service (FSIS) of the United States Department of Agriculture (USDA) has primary jurisdiction over meat, poultry, and processed egg products. The Environmental Protection Agency (EPA) regulates pesticide use and sets tolerance levels for pesticide residues in food under the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) and the Food Quality Protection Act (FQPA).
The passage of FSMA in 2011 represented the most sweeping reform of U.S. food safety law since 1938, shifting the regulatory paradigm from reactive (responding to outbreaks) to preventive (anticipating and preventing hazards). FSMA's seven foundational rules — including the Preventive Controls for Human Food rule and the Foreign Supplier Verification Programs rule — are explicitly risk-based, requiring food businesses to conduct hazard analyses and implement risk-based preventive controls proportionate to the severity and likelihood of identified hazards.
2.3 Risk-Based Metrics: FSOs, POs, and Performance Criteria
A significant advance in the practical application of food safety risk assessment has been the development of quantitative risk-based metrics that link microbiological performance along the food chain to public health outcomes. The Food Safety Objective (FSO), introduced by the International Commission on Microbiological Specifications for Foods (ICMSF), expresses the maximum frequency or concentration of a hazard at the moment of consumption that provides or contributes to the appropriate level of protection (ALOP). The FSO provides a tangible target that food businesses, regulators, and international trading partners can work toward, even when they employ different technologies or process designs to achieve it.
Complementing the FSO are Performance Objectives (POs), which specify the required maximum frequency or concentration of a hazard at a specific point in the food chain (upstream of consumption) that contributes to achieving the FSO. Process Criteria define the control parameters (time, temperature, pH, water activity) at a specific step that are necessary to achieve the PO. Together, these metrics create a coherent, risk-based hierarchy from the farm through to the table, enabling risk managers to specify public health goals in outcome-based terms while granting food businesses flexibility in how they achieve them.
Illustrations are not included in the reading sample
Chapter 3: Microbiological Hazards — Pathogens and Toxins
3.1 Classification and Overview of Foodborne Pathogens
Microbiological hazards in food encompass a diverse array of agents, including bacteria, viruses, parasites, and fungi (principally through mycotoxin production). The global burden of foodborne disease is dominated by diarrheal agents, with norovirus, Campylobacter spp., non-typhoidal Salmonella, and diarrheagenic Escherichia coli collectively accounting for the majority of cases. However, the most severe outcomes — in terms of mortality, long-term sequelae, and the potential for large-scale outbreaks — are associated with a relatively small number of agents that warrant particular attention in risk assessment.
The World Health Organization's Foodborne Disease Burden Epidemiology Reference Group (FERG) published landmark estimates in 2015 attributing 31 foodborne hazards to approximately 600 million illnesses and 420,000 deaths annually worldwide. Norovirus was the single largest cause of foodborne illness (125 million cases), while non-typhoidal Salmonella was the leading bacterial cause of mortality (59,000 deaths). These estimates, though subject to significant uncertainty due to the under-ascertainment of foodborne illness in both developed and developing countries, provide an essential foundation for risk prioritization.
3.2 Bacteria of Major Public Health Significance
3.2.1 Salmonella
Salmonella is a genus of Gram-negative, facultatively anaerobic rod-shaped bacteria within the family Enterobacteriaceae. With over 2,600 serovars identified to date, Salmonella exhibits extraordinary antigenic diversity that complicates surveillance, source attribution, and vaccine development. Non- typhoidal Salmonella (NTS), primarily serovars Enteritidis and Typhimurium, causes an estimated 93.8 million illnesses and 155,000 deaths annually from foodborne transmission, making it the leading bacterial cause offoodborne gastroenteritis globally.
Salmonella Enteritidis has a particular predilection for colonizing the reproductive tissues of laying hens, enabling transovarial contamination of eggs — a transmission route that complicates control because the pathogen is contained within the intact shell. Risk assessments for Salmonella in eggs and poultry products have been among the most extensively developed in the food safety literature, providing foundational models for QMRA methodology. Key control points include vaccination of laying flocks, competitive exclusion programs, refrigeration of shell eggs, and consumer cooking practices.
Salmonella Typhi and Paratyphi, the causative agents oftyphoid and paratyphoid fever respectively, differ fundamentally from NTS in their pathogenesis: they are exclusively human pathogens that cause systemic, invasive disease characterized by sustained fever, bacteremia, and, in severe or untreated cases, intestinal perforation and death. Typhoid fever remains a major public health burden in South and Southeast Asia, sub-Saharan Africa, and parts of Latin America, with an estimated 1120 million cases annually. Food and water contaminated with human feces are the primary transmission vehicles.
3.2.2 Campylobacter
Campylobacter jejuni and C. coli are the most commonly reported bacterial causes of foodborne gastroenteritis in many high-income countries. In the European Union, campylobacteriosis has consistently been the most frequently reported zoonosis since at least 2005, with over 200,000 confirmed human cases annually — a figure that vastly underestimates the true burden due to underreporting. In the United States, CDC estimates approximately 1.5 million Campylobacter infections peryear, resulting in approximately 120 deaths.
The primary reservoir of Campylobacter is the gastrointestinal tract of poultry, cattle, and other domestic and wild animals. Contamination of poultry carcasses during slaughter is the principal route by which the pathogen enters the food chain. Risk assessments for Campylobacter in broiler chickens have identified the slaughter process — particularly scalding, defeathering, and evisceration — as critical intervention points, while refrigeration and consumer cross-contamination behavior are major determinants of human exposure.
A particularly concerning sequela of Campylobacter infection is Guillain-Barre syndrome (GBS), an acute inflammatory demyelinating polyneuropathy that can result in temporary or permanent paralysis. Approximately 1 in 1,000 Campylobacter infections is estimated to trigger GBS, making Campylobacter the single most common antecedent infection identifiable in GBS cases globally. The DALY burden of Campylobacter-attributable GBS significantly increases the overall impact of this pathogen beyond the acute gastroenteritis episode.
3.2.3 Listeria monocytogenes
Listeria monocytogenes is a Gram-positive, non-spore-forming, psychrotrophic bacterium capable of growth at refrigeration temperatures (as low as 0-4°C), in high-salt environments, and across a wide pH range. These remarkable survival and growth characteristics make it a particularly challenging hazard in ready-to-eat (RTE) foods that receive no further cooking step before consumption. Despite relatively low incidence — approximately 1,600 cases per year in the United States — listeriosis has an exceptionally high case-fatality rate, estimated at 20-30% in high-risk populations including pregnant women, neonates, elderly individuals, and immunocompromised patients.
The FDA's 2003 quantitative risk assessment of Listeria monocytogenes in ready-to-eat foods remains one of the most comprehensive QMRA documents ever produced, evaluating 23 different RTE food categories and providing risk rankings that have directly informed regulatory policy. The assessment introduced the concept of exponential growth during retail storage and home refrigeration as a critical determinant of consumer exposure, highlighting the importance of maintaining cold chain integrity and minimizing product shelf life as risk reduction measures.
3.2.4 Pathogenic Escherichia coli
Escherichia coli is a ubiquitous commensal inhabitant of the mammalian intestine, but several pathotypes possess virulence factors that enable them to cause intestinal disease ranging from mild diarrhea to life-threatening illness. The pathotypes of greatest food safety significance are: enterohemorrhagic E. coli (EHEC), also known as Shiga toxin-producing E. coli (STEC); enterotoxigenic E. coli (ETEC); and enteroaggregative E. coli (EAEC).
STEC 0157:H7 and non-0157 STEC serogroups (026, 045, 0103, 0111, 0121, 0145) are responsible for severe hemorrhagic colitis and, in approximately 5-10% of cases, hemolytic uremic syndrome (HUS) — a potentially fatal condition characterized by microangiopathic hemolytic anemia, thrombocytopenia, and acute renal failure. Children under five years of age are at particularly high risk of developing HUS following STEC 0157:H7 infection. The infective dose is very low (estimated <100 organisms), and cattle are the primary reservoir, making ground beef, raw milk, and fresh produce contaminated with bovine feces the leading vehicles.
3.3 Viruses in Food
3.3.1 Norovirus
Human norovirus (HuNoV), a member of the Caliciviridae family, is the dominant cause of non- bacterial acute gastroenteritis worldwide, responsible for an estimated 685 million cases and 200,000 deaths annually, with the vast majority of deaths occurring in children under 5 in low-income countries. Norovirus is characterized by its extreme environmental stability, very low infectious dose (estimated 18-1000 virus particles), lack ofa robust cell culture system for routine propagation, and the absence of durable immunity after infection — all features that complicate both risk assessment and control.
Primary transmission routes include person-to-person contact (fecal-oral), consumption of contaminated bivalve mollusks (oysters being particularly high risk due to bioaccumulation of virions from sewage-contaminated waters), fresh produce irrigated with contaminated water, and ready-to- eat foods handled by infected food workers. The inability to culture HuNoV in standard cell culture systems has historically limited dose-response modeling, although recent breakthroughs using human intestinal enteroids and gnotobiotic pig models are beginning to fill this critical data gap.
3.3.2 Hepatitis A Virus
Hepatitis A virus (HAV), a positive-sense single-stranded RNA virus classified within the genus Hepatovirus of the Picornaviridae family, is the causative agent of hepatitis A — an acute, usually self-limiting liver disease. Unlike other hepatitis viruses, HAV is not associated with chronic infection, but severe cases can result in fulminant hepatic failure, particularly in older adults and individuals with pre-existing liver disease. HAV is highly stable in the environment and food, resistant to many disinfectants, and capable ofsurviving on contaminated produce forweeks.
Food-associated outbreaks of hepatitis A have been linked to a diverse range of commodities, including frozen berries, shellfish, green onions, salads, and sandwiches prepared by infected food handlers. The largest documented foodborne HAV outbreak in the United States occurred in 2003 in Monaca, Pennsylvania, linked to green onions, resulting in 601 confirmed cases and 3 deaths. Vaccination is the most effective preventive measure, and HAV vaccines are now routinely recommended forfood handlers in many countries.
3.4 Parasites in Food
Foodborne parasites comprise a phylogenetically diverse group of protozoa and helminths that cause significant morbidity and mortality, particularly in low- and middle-income countries where sanitation infrastructure is limited. The FAO/WHO identified 24 foodborne parasites of global significance in a 2014 expert consultation, ranking them by their overall public health impact using a multi-criteria tool that considered disease burden, geographic distribution, food system linkages, and amenability to control.
Taenia solium (pork tapeworm) was ranked first overall, primarily because of the devastating neurological consequences of cysticercosis — infection of human tissues, including the brain (neurocysticercosis), with larval T. solium. Echinococcus granulosus and E. multilocularis (causative agents of cystic and alveolar echinococcosis, respectively), Toxoplasma gondii, and Cryptosporidium spp. were also ranked among the highest priority foodborne parasites. Cryptosporidium parvum and C. hominis are of particular concern in water and fresh produce, given their extreme resistance to chlorination — the primary disinfection treatment used in municipal water supplies worldwide.
Illustrations are not included in the reading sample
Chapter 4: Quantitative Microbiological Risk Assessment (QMRA)
4.1 Principles and Applications of QMRA
Quantitative Microbiological Risk Assessment (QMRA) applies the four-step risk assessment paradigm to foodborne pathogens with the explicit goal of producing quantitative estimates of infection or illness risk. Unlike the qualitative or semi-quantitative approaches that dominated food microbiology prior to the 1990s, QMRA employs mathematical models of pathogen behavior throughout the farm-to-fork continuum and probabilistic techniques to characterize variability and uncertainty in the risk estimate. The result is a transparent, auditable document that can be used to support regulatory decision-making, prioritize interventions, set quantitative food safety targets, and evaluate the health impact of proposed changes in food production practice or consumer behavior.
The power of QMRA lies in its ability to integrate diverse data streams — microbiological monitoring data, growth and inactivation kinetics, consumption survey data, dose-response relationships — into a coherent model that predicts the probability of illness across the full range of realistic exposure scenarios. Sensitivity analysis within a QMRA framework can identify the 'critical nodes' in the food chain where risk is most sensitive to changes in pathogen levels or control measures, directly informing the allocation of control resources.
4.2 Exposure Assessment Models for QMRA
4.2.1 Farm-to-Fork Modeling Approach
A complete farm-to-fork QMRA model tracks the fate of a pathogen from its point of entry into the food system (typically the primary production stage) through all subsequent processing, storage, distribution, retail, and consumer handling steps until the moment of consumption. At each stage, the model accounts for processes that can either increase (growth) or decrease (inactivation, removal) the pathogen load, as well as cross-contamination events that can spread the pathogen to previously uncontaminated food or surfaces.
The prevalence of contamination (proportion of units contaminated) and concentration of the pathogen (number of organisms per unit) are tracked separately throughout the model, since interventions may differentially affect one or both parameters. For example, heat treatment during cooking may dramatically reduce pathogen concentration but has no direct effect on the prevalence of contamination before cooking (though it reduces post-cooking survival to negligible levels if performed correctly).
4.2.2 Predictive Microbiology Integration
Predictive microbiology — the development and use of mathematical models to describe microbial growth, survival, and death as a function of environmental parameters — is an essential tool in QMRA exposure assessment. Primary models describe growth kinetics (e.g., modified Gompertz, logistic, Baranyi-Roberts models) or inactivation kinetics (e.g., first-order, Weibull models) as a function of time. Secondary models describe how the parameters of primary models (e.g., maximum growth rate, lag phase duration) depend on environmental variables such as temperature, pH, water activity, and antimicrobial concentration.
Major repositories of validated predictive microbiology models include the FDA's Pathogen Modeling Program (PMP), ComBase (a collaborative database of microbial responses hosted by USDA and the University of Tasmania), and the EFSA PathogenCombinations database. These tools are invaluable for parameterizing QMRA models when experimental data on pathogen behavior in specific food matrices are unavailable, though users must exercise caution when extrapolating predictions beyond the conditions underwhich models were validated.
4.3 Dose-Response Models for Foodborne Pathogens
The dose-response relationship quantifies the probability of adverse health outcome (typically, probability of infection or probability of illness) as a function of the ingested dose of pathogen. For food safety purposes, dose-response data typically derive from human volunteer feeding studies, epidemiological investigations of foodborne outbreaks, or, in some cases, animal models (with appropriate interspecies extrapolation).
Three dose-response models are most commonly used in QMRA:
• The exponential model: P(infection | dose d)=1 - exp(-r x d), where r is the probability that a single organism initiates infection. This model assumes that each organism acts independently and that the host-pathogen interaction follows a Poisson process.
• The beta-Poisson model: P(infection | dose d)=1 - (1+ d/p)A(-a), where a and p are fitted parameters. This model allows for heterogeneity in host susceptibility and pathogen infectivity, often providing better fit to empirical data across a wider dose range.
• The log-normal model: Used less frequently but appropriate when dose-response data suggest a sigmoidal relationship with a threshold-like region at very low doses. Subject to criticism for implying a true threshold dose.
A critical issue in dose-response modeling for QMRA is the distinction between probability of infection and probability of illness. Not all infected individuals develop clinical symptoms; the fraction that does — the morbidity ratio — depends on the pathogen, the inoculum size, the host's immune status, and host genetic factors. Similarly, only a fraction of ill individuals seek medical attention, and a fraction of those will be correctly diagnosed and reported in surveillance systems. The product of these ratios — from true infection to reported case — is the 'reporting ratio,' which is essential for converting model-predicted illness estimates into expected reported cases for comparison with surveillance data.
4.4 Uncertainty Analysis in QMRA
Uncertainty analysis is arguably the most important — and most frequently neglected — component of a rigorous QMRA. Key sources of uncertainty include: parameter uncertainty (imprecision in estimated model parameters due to limited data); model uncertainty (the possibility that the mathematical structure ofthe model fails to accurately represent the real biological system); scenario uncertainty (incompleteness or errors in the description of the food production and consumption scenarios being modeled); and decision uncertainty (ambiguity about the appropriate policy question to address).
Parameter uncertainty is most commonly addressed through Monte Carlo simulation, in which uncertain parameters are represented as probability distributions (rather than point estimates) and the model is run thousands of times with parameter values sampled from these distributions. The resulting distribution of risk estimates characterizes the uncertainty in the final answer. Second-order (two-dimensional) Monte Carlo simulation, which separately propagates variability distributions (representing real-world heterogeneity) and uncertainty distributions (representing knowledge limitations), provides the most rigorous characterization but is computationally intensive and requires a clear conceptual framework for categorizing each model parameter.
QMRA Best Practice: Sensitivity Analysis
A sensitivity analysis identifies which input parameters most strongly influence the model output (risk estimate). The most common method is Spearman rank correlation between each input parameter and the output variable across Monte Carlo iterations. Parameters with high Spearman correlation coefficients are the primary drivers of risk and should receive priority attention in risk management and data collection strategies. Tornado diagrams provide an intuitive visual representation of sensitivity analysis results.
Chapter 5: Chemical Hazards in Food
5.1 Categories of Chemical Contaminants
Chemical hazards in food encompass a broad spectrum of agents introduced at various stages of the food chain — from agricultural production through processing, packaging, and storage. Unlike microbiological hazards, which are living organisms capable of replication and adaptation, chemical hazards are defined by their molecular structure and physicochemical properties, which determine their fate in food systems, their bioavailability and metabolism in the human body, and the mechanisms through which they exert adverse health effects.
The major categories of chemical hazards in food can be classified as follows: (1) environmental contaminants — persistent organic pollutants (POPs), heavy metals, dioxins, and polychlorinated biphenyls (PCBs) that enter the food chain from contaminated soil, water, or air; (2) agricultural chemicals — pesticide residues and veterinary drug residues that enter food as a result of their intentional use in crop protection and animal husbandry; (3) process-induced contaminants — chemicals formed during food processing as a result of heat treatment, fermentation, or other transformations (e.g., acrylamide, furans, heterocyclic amines, nitrosamines); (4) naturally occurring toxins — mycotoxins, marine biotoxins, phycotoxins, and plant toxins present in certain food commodities; and (5) food contact material migrants — chemical substances that transfer from packaging materials, processing equipment, or storage containers into food.
5.2 Heavy Metals and Metalloids
5.2.1 Lead
Lead (Pb) is a naturally occurring heavy metal that has been extensively redistributed in the environment through industrial activities including mining, smelting, and the historical use of leaded gasoline and paint. Dietary exposure to lead is a major public health concern globally, particularly for children, in whom even low-level lead exposure is associated with irreversible neurodevelopmental toxicity including reduced IQ, behavioral disorders, and impaired academic achievement. There is no established threshold dose for lead neurotoxicity in children; the current evidence supports a linear dose-response relationship at environmentally relevant exposures.
Primary dietary sources of lead vary geographically but commonly include root vegetables grown in contaminated soil, drinking water distributed through lead pipes or lead-soldered plumbing, traditional ceramics glazed with lead-containing materials, certain herbal supplements, and some traditional medicinal preparations. The FDA's Total Diet Study (TDS) provides ongoing monitoring of dietary lead exposure in the U.S. population, while EFSA conducts similar total diet studies across European member states.
5.2.2 Cadmium
Cadmium (Cd) is a non-essential trace element that accumulates primarily in the kidneys, where chronic exposure can cause tubular nephropathy — an irreversible form of kidney damage manifesting as proteinuria, glucosuria, and, in severe cases, Fanconi syndrome and osteomalacia. Cadmium has a biological half-life in the kidney of 15-30 years, meaning that health effects may manifest decades after the primary exposure.
The primary dietary sources of cadmium exposure in non-smoking populations are cereals and cereal products (particularly durum wheat), vegetables (particularly leafy vegetables, root vegetables, and potatoes), and, forindividuals with high consumption, offal (particularly kidney), shellfish (particularly mollusks), and cocoa products. Tobacco smoking is also a major non-dietary source of cadmium exposure. The EFSA Panel on Contaminants established a tolerable weekly intake (TWI) of 2.5 pg/kg body weight per week for cadmium, noting that dietary exposure in some segments of the European population — particularly vegetarians, subsistence farmers in contaminated areas, and regular shellfish consumers — may approach or exceed this level.
5.2.3 Mercury
Mercury occurs in the environment in three primary forms: elemental mercury (Hg0), inorganic mercury compounds (principally Hg2+ salts), and organic mercury compounds, of which methylmercury (MeHg) is by far the most toxicologically significant form in food safety. Methylmercury is formed from inorganic mercury through microbial methylation in aquatic sediments and bioaccumulates through aquatic food chains, reaching highest concentrations in large, long-lived predatory fish and marine mammals. Dietary exposure to methylmercury is therefore dominated by fish and seafood consumption, with tuna, swordfish, shark, king mackerel, and tilefish posing the greatest risk.
Methylmercury is a potent neurotoxin that crosses both the blood-brain barrier and the placenta with ease, making the developing fetal brain exquisitely vulnerable. The Minamata disease disaster in Japan (1956) and the Grassy Narrows contamination in Canada provided devastating evidence of the consequences of high-level methylmercury exposure from fish consumption. However, fish are also an important source of omega-3 fatty acids, iodine, and other nutrients critical for fetal brain development, creating a genuine risk-benefit tension for pregnant women that has been the subject of extensive risk communication research.
5.3 Persistent Organic Pollutants (POPs)
Persistent organic pollutants (POPs) are a group of organic chemicals that resist environmental degradation through chemical, biological, and photolytic processes; accumulate in the food chain through bioaccumulation and biomagnification; and can have adverse effects on human health and the environment. The Stockholm Convention on Persistent Organic Pollutants (2001) established an international framework for eliminating or restricting the production and use of the 'dirty dozen' initial POPs, with subsequent amendments adding additional substances. The list now includes chlorinated pesticides (DDT, chlordane, aldrin, dieldrin, endrin, heptachlor, mirex, toxaphene, endosulfan), industrial chemicals (PCBs, hexachlorobenzene), and combustion by-products (dioxins and furans).
Polychlorinated dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans (PCDFs), and dioxin-like polychlorinated biphenyls (dl-PCBs) are assessed collectively as 'dioxins and dioxin-like compounds' using toxic equivalency factors (TEFs) that weight each congener's potency relative to the most toxic form, 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). The WHO's World Health Organization Toxic Equivalency Quotient (WHO-TEQ) system allows the complex mixture of these compounds in food to be expressed as a single number representing the equivalent toxicity of pure TCDD, facilitating comparison with health-based guidance values.
5.4 Mycotoxins
Mycotoxins are toxic secondary metabolites produced by filamentous fungi (molds) that can contaminate a wide range of agricultural commodities before harvest (field fungi such as Fusarium spp.) or during storage (storage fungi such as Aspergillus and Penicillium spp.). Over 400 mycotoxins have been described, but regulatory attention and risk assessment effort have focused primarily on aflatoxins, ochratoxin A (OTA), fumonisins, trichothecenes (including deoxynivalenol, DON; zearalenone, ZEN; and T-2/HT-2 toxins), and patulin.
Aflatoxins, produced primarily by Aspergillus flavus and A. parasiticus, are among the most potent naturally occurring carcinogens known. Aflatoxin B1 (AFB1) is the most toxic of the four major aflatoxins (B1, B2, G1, G2) and is classified as a Group 1 human carcinogen by the International Agency for Research on Cancer (IARC) on the basis of epidemiological evidence linking dietary exposure to hepatocellular carcinoma, particularly in populations with concurrent hepatitis B virus infection (which acts synergistically with AFB1 to increase cancer risk dramatically). Aflatoxin M1 (AFM1), a hydroxylated metabolite of AFB1, is excreted in the milk of dairy animals that consume contaminated feed, making contamination of cow's milk and infant formula an important food safety and public health concern.
Illustrations are not included in the reading sample
Chapter 6: Chemical Risk Assessment Methodology
6.1 Toxicokinetics and the ADME Paradigm
Chemical risk assessment is grounded in an understanding of toxicokinetics — the study of how the body handles a chemical substance over time. The ADME framework (Absorption, Distribution, Metabolism, Excretion) provides the conceptual structure for describing the journey of a chemical from the moment of ingestion to its ultimate elimination from the body. Toxicokinetic data are essential for understanding the relationship between external exposure (dietary intake) and internal dose (the concentration of the parent compound or its active metabolite at the target organ), which is the more directly relevant determinant oftoxicological effect.
Absorption describes the processes by which a chemical enters the systemic circulation from the gastrointestinal tract following oral ingestion. Absorption efficiency — the fraction of ingested chemical that reaches systemic circulation — is the key determinant of oral bioavailability and varies widely across chemical classes and food matrices. For example, inorganic arsenic is readily absorbed (~90-95% bioavailability), while many mycotoxins are partially detoxified or poorly absorbed in the gut, reducing systemic exposure relative to dietary intake.
Distribution describes the movement of absorbed chemical through the body to tissues and organs. Many chemicals exhibit preferential distribution to specific tissues due to binding to tissue proteins, sequestration in lipid-rich compartments (lipophilic POPs), or active transport mechanisms. Methylmercury, for example, is avidly transported across the blood-brain barrier by amino acid carrier proteins, explaining its selective neurotoxicity despite being distributed throughout the body.
Metabolism — primarily hepatic biotransformation — is a critical determinant of both the duration of chemical action and the toxicological consequences of exposure. Phase I reactions (principally cytochrome P450-mediated oxidations) can either activate relatively non-toxic parent compounds to more reactive, toxic metabolites (bioactivation — e.g., AFB1 epoxidation to the hepatocarcinogenic 8,9-epoxide) or detoxify reactive compounds to less harmful intermediates. Phase II reactions (conjugation with glucuronate, sulfate, glutathione, etc.) generally increase water solubility and facilitate urinary or biliary excretion.
Excretion is the final phase of chemical disposition, with urinary excretion being the primary route for water-soluble compounds and their metabolites, and biliary excretion (in feces) more important for large, lipophilic molecules. Some metals (lead, mercury, arsenic) are partially excreted through hair, nails, and sweat, providing opportunities for biomonitoring of chronic exposure using accessible biological matrices.
6.2 Establishing Health-Based Guidance Values
6.2.1 Acceptable Daily Intake (ADI) and Tolerable Intakes
Health-based guidance values (HBGVs) represent the maximum daily or weekly intake ofa chemical substance that, on the basis of current knowledge, is considered to be without appreciable risk of adverse health effects when ingested over a lifetime. For food additives and pesticide residues subject to deliberate use — where a 'no observed adverse effect level' (NOAEL) or 'lowest observed adverse effect level' (LOAEL) can be clearly established from toxicological studies — the primary HBGV is the Acceptable Daily Intake (ADI), expressed in mg of substance per kg body weight per day.
The ADI is derived by dividing the NOAEL from the most appropriate toxicological study by a composite uncertainty factor (UF), which accounts for interspecies extrapolation (typically a 10-fold factor for extrapolating from animal to human), intraspecies variability (a 10-fold factor for variability within the human population), and sometimes additional factors for database deficiencies, subchronic-to-chronic extrapolation, or special susceptibility concerns. The resulting 100-fold composite uncertainty factor, applied to a NOAEL from a good-quality animal study, is the traditional basis ofthe ADI, though modern practice increasingly uses the Benchmark Dose (BMD) approach as the point of departure.
6.2.2 Benchmark Dose Approach
The Benchmark Dose (BMD) approach has emerged as a preferred alternative to the NOAEL/LOAEL approach for deriving points of departure for HBGVs. Rather than using a single data point (the highest dose with no observed effect), the BMD method fits a dose-response model to the full dataset from a toxicological study and uses the model to estimate the dose corresponding to a predefined level of response — typically a 5% or 10% increase in the incidence of an adverse effect over background (BMR05 or BMR10). The lower 95% confidence bound on this estimate, the BMDL, is used as the point of departure.
The BMD approach offers several advantages over the NOAEL: it uses all available dose-response data rather than a single point, accounts for the statistical power of the study in the confidence interval calculation, is not limited to doses actually tested in the study, and provides a consistent and objective basis for comparison across studies and substances. EFSA and many national regulatory agencies now routinely use the BMD approach as the default for deriving HBGVs, supported by dedicated software tools (e.g., EFSA's PROAST software, the U.S. EPA's BMDS).
6.3 Non-threshold Carcinogens: Linear Low-Dose Extrapolation
For genotoxic carcinogens — substances capable of directly damaging DNA and initiating carcinogenesis — it is generally assumed that no safe threshold dose exists: any exposure, however small, carries some finite incremental risk of cancer. This non-threshold assumption, while scientifically contested for some agents, is adopted as a default in most regulatory frameworks as a conservative (health-protective) assumption in the face of uncertainty.
For non-threshold carcinogens, health-based guidance values take different forms. The Margin of Exposure (MoE) approach, recommended by EFSA, calculates the ratio of the BMD01 (the dose corresponding to a 1 % increase in tumor incidence) to the estimated dietary exposure. A MoE greater than 10,000 for genotoxic carcinogens is generally considered to indicate a low priority for risk management, while lower MoEs trigger more urgent regulatory attention. The U.S. EPA uses cancer slope factors to calculate the Incremental Lifetime Cancer Risk (ILCR) — the probability of developing cancer over a 70-year lifetime as a result of a given daily dietary exposure.
6.4 Chemical Mixture Risk Assessment
A major limitation of traditional chemical risk assessment is its focus on individual substances in isolation, whereas human dietary exposure invariably involves simultaneous exposure to multiple chemicals. Mixture risk assessment addresses the question: do multiple chemicals, each present at concentrations below their individual HBGVs, combine to produce a health risk?
The two primary models of mixture toxicity are: dose addition (also known as concentration addition), which assumes that chemicals in a mixture act through the same mechanism and can be treated as dilutions of each other; and independent action (response addition), which assumes that chemicals act through different mechanisms and produce statistically independent biological responses. For most regulatory purposes, dose addition is adopted as the default for chemicals with similar mechanisms (e.g., all organophosphate pesticides acting via cholinesterase inhibition, assessed as a group using relative potency factors), while independent action is applied to groups with dissimilar mechanisms.
The Hazard Index (HI) approach — summing the hazard quotients of individual chemicals — is the simplest implementation of dose addition for mixtures. An HI > 1 indicates that the cumulative exposure may exceed health-based guidance values. More sophisticated approaches, such as the Cumulative Assessment Group (CAG) methodology developed by EFSA, group chemicals by common mechanism and derive group-level dietary exposure estimates for comparison with group- level reference values.
Chapter 7: Pesticide Residues and Veterinary Drug Residues
7.1 Pesticide Risk Assessment Framework
Pesticide residues in food arise from the intentional use of plant protection products in agricultural production. The regulatory framework governing pesticide residues is designed to ensure that the benefits of pest control — increased crop yields, reduced post-harvest losses, improved food quality — are realized while protecting consumers from unacceptable dietary exposure to pesticide residues. The key instrument in this framework is the maximum residue level (MRL) — the highest legally permitted concentration of a pesticide residue in or on food or feed.
MRLs are derived from Good Agricultural Practice (GAP) trials that document the residue levels resulting from pesticide application according to the registered label conditions. The MRL is set to reflect the highest residue level likely to occur when the pesticide is used in accordance with GAP. MRLs are not per se health-based standards; rather, they are compliance tools reflecting agricultural practice, with the health-based check performed separately through dietary risk assessment. A pesticide MRL is considered acceptable from a health perspective only if the resulting dietary exposure does not exceed the ADI.
7.2 The Role ofJMPR in International MRL Setting
The Joint FAO/WHO Meeting on Pesticide Residues (JMPR) is the international scientific body responsible for evaluating the safety of pesticide residues in food and recommending MRLs to the Codex Committee on Pesticide Residues (CCPR). JMPR evaluations cover two complementary aspects: toxicological evaluation (establishing an ADI and, for acute dietary risk, an Acute Reference Dose, ARfD) and residue evaluation (reviewing field trial data and processing studies to recommend MRLs and estimate dietary exposure).
The JMPR follows a structured dietary risk assessment approach that distinguishes between chronic dietary risk (comparing long-term average dietary intake with the ADI) and acute dietary risk (comparing high-percentile single-day dietary intake with the ARfD). Acute dietary risk assessment is particularly important for pesticides with rapid neurotoxic or endocrine-disrupting effects at high doses, where a single meal with a high residue level could theoretically exceed the ARfD. The International Estimated Daily Intake (IEDI) and International Estimated Short-Term Intake (IESTI) are the JMPR's standardized tools for these assessments.
7.3 Endocrine-Disrupting Chemicals (EDCs) — Special Considerations
Endocrine-disrupting chemicals (EDCs) present unique challenges to conventional dose-responsebased risk assessment because of features of their toxicology that violate the standard assumptions underlying the derivation of HBGVs. Three characteristics of EDC toxicology are most problematic:
• Non-monotonic dose-response relationships: Unlike the monotonically increasing doseresponse relationships assumed in traditional toxicology, EDCs — acting through receptor- mediated mechanisms — can produce inverse or biphasic dose-response curves where low doses produce greater effects than higher doses. This can render the standard NOAEL/LOAEL approach profoundly misleading.
• Critical windows of exposure: Endocrine disruption during critical developmental periods (embryonic development, puberty) can produce irreversible health effects at doses that are without effect in adult animals or during non-critical exposure windows. Risk assessments based on adult animal studies may therefore severely underestimate risk to the developing organism.
• Low-dose effects and mixture interactions: Some EDCs produce significant biological effects at doses below conventional NOAELs, and mixtures of EDCs — each at sub-threshold concentrations — can produce additive or synergistic effects through dose addition at shared receptor targets.
These challenges have prompted calls for fundamental reform of the regulatory approach to EDCs, including the development of EDC-specific testing protocols, the adoption of non-monotonic doseresponse models in regulatory risk assessment, and the application of the precautionary principle in cases of scientific uncertainty. The debate over whether EDCs should be subject to hazard-based (classification) regulation — prohibiting use regardless of exposure — or risk-based regulation — permitting use when exposure is below a safe threshold — remains one of the most contentious issues in contemporary food safety policy.
7.4 Veterinary Drug Residues
Veterinary medicines are used in food-producing animals for therapeutic purposes (treating disease), prophylactic purposes (preventing disease), metaphylactic purposes (treating a group of animals when some are infected), and, in some countries, as growth promoters. Residues of veterinary drugs — the parent compound or its metabolites — can persist in edible animal tissues (muscle, liver, kidney, fat), eggs, milk, and honey after treatment.
The JECFA is responsible for evaluating the safety of veterinary drug residues and recommending Maximum Residue Limits (MRLs) for veterinary drugs in animal-derived foods. JECFA derives ADIs for veterinary drug residues using the same methodology applied to food additives, and then estimates dietary exposure using standard food consumption patterns and assumed residue concentrations at various stages of the food chain.
A particular concern in the context of veterinary drug residues is the use of antimicrobial agents in food-producing animals and its contribution to the selection and spread of antimicrobial-resistant bacteria — a topic addressed in detail in Chapter 9 of this volume. Beyond antimicrobial resistance, other classes of veterinary drugs of food safety significance include hormonal growth promoters (banned in the EU but permitted in other major producing countries), non-steroidal anti-inflammatory drugs (NSAIDs), and antiparasitic agents.
Chapter 8: Allergens and Food Intolerances as Risk Hazards
8.1 Immunology of Food Allergy
Food allergy is an adverse health effect arising from a specific immune response that occurs reproducibly on exposure to a given food. IgE-mediated food allergy — the most clinically significant and best understood form — involves sensitization of mast cells and basophils by allergen-specific IgE antibodies produced during an initial exposure phase, followed by release of histamine and other inflammatory mediators upon subsequent exposure to the trigger food. Clinical manifestations range from mild oral allergy syndrome (itching, tingling, or swelling of the lips, mouth, tongue, and throat) through urticaria, angioedema, gastrointestinal symptoms, and rhinitis, to severe systemic anaphylaxis, which can be life-threatening without prompt administration of epinephrine.
The 'Big Nine' major food allergens — milk, eggs, fish, crustacean shellfish, tree nuts, peanuts, wheat, soybean, and sesame — account for the vast majority of IgE-mediated food allergic reactions in Western populations and form the basis of mandatory food allergen labeling requirements in most countries. However, additional allergens of regional significance (e.g., buckwheat in Japan and Korea, celery and mustard in Europe, lupin in Mediterranean countries) are increasingly recognized as important public health hazards in specific demographic contexts.
8.2 Allergen Threshold Doses and VITAL
A fundamental challenge in allergen risk assessment is the determination of threshold doses — the minimum amount of allergen that can elicit an adverse reaction in sensitized individuals. Threshold data are essential for establishing precautionary allergen labeling thresholds, setting allergen cleaning validation standards in food manufacturing facilities, and determining the significance of adventitious cross-contact during food production.
The VITAL (Voluntary Incidental Trace Allergen Labeling) program, developed by the Allergen Bureau of Australia and New Zealand and subsequently adopted in various forms by regulatory bodies and industry in multiple countries, provides a science-based framework for setting allergen reference doses. VITAL reference doses are derived from double-blind, placebo-controlled food challenge (DBPCFC) studies that provide individual threshold data, which are then modeled with log-normal or other statistical distributions to identify doses that would be expected to trigger reactions in no more than a specified fraction (typically 1%) ofthe allergic population.
8.3 Celiac Disease and Gluten Risk Assessment
Celiac disease is an autoimmune disorder triggered by gluten — a complex mixture of storage proteins found in wheat, barley, rye, and related cereals — in genetically predisposed individuals carrying HLA-DQ2 or HLA-DQ8 alleles. Unlike classic IgE-mediated food allergy, celiac disease involves an adaptive T-cell immune response to deamidated gliadin peptides presented by HLA- DQ2/DQ8 molecules, resulting in villous atrophy, crypt hyperplasia, and malabsorption in the small intestine. Prevalence is estimated at approximately 1% of the general population in Western countries, though the majority of cases remain undiagnosed.
Risk assessment for gluten in celiac disease differs from allergen risk assessment in several important respects: the response is not acute and anaphylactic but rather chronic and cumulative; repeated small doses can cause ongoing mucosal damage even in the absence of overt gastrointestinal symptoms; and the endpoint of concern is histological (mucosal damage) rather than subjective (symptoms). Available evidence suggests that the threshold for mucosal damage in celiac patients varies widely among individuals, but that daily gluten intake below 10 mg is unlikely to cause significant intestinal damage in the majority of celiac patients, forming the scientific basis for the Codex Alimentarius standard defining 'gluten-free' foods as those containing less than 20 ppm of gluten.
Chapter 9: Antimicrobial Resistance — An Emerging Food Safety Hazard
9.1 The AMR Crisis and the Food-Human Interface
Antimicrobial resistance (AMR) is recognized by the World Health Organization as one ofthe greatest global public health threats of the 21st century. The emergence, selection, and dissemination of bacteria resistant to antibiotics used in human medicine — driven in part by the extensive use of antimicrobials in food-producing animals — represents a fundamental challenge to the assumptions underlying modern healthcare. If current trends continue, drug-resistant infections are projected to cause 10 million deaths annually by 2050 — surpassing cancer as the leading cause of mortality globally.
The food-human interface is a major pathway for the transmission of antimicrobial-resistant bacteria (ARB) and antimicrobial resistance genes (ARGs) to humans. Food-producing animals treated with antimicrobials — or living in environments contaminated with antimicrobial residues from animal waste — are subject to strong selective pressure that enriches the commensal and pathogen microflora for resistant variants. Resistant bacteria and their genetic resistance determinants can then enter the human food chain through: consumption of meat, poultry, fish, and other animal products; contamination of plant-based foods with animal fecal matter through irrigation, composting, or direct contact; and indirect transmission through the environment (soil, water, air).
9.2 Risk Assessment Framework for AMR in Food
The Codex Alimentarius Commission adopted the Code of Practice to Minimize and Contain Antimicrobial Resistance (CAC/RCP 61-2005), and the Codex Ad Hoc Intergovernmental Task Force on Antimicrobial Resistance (TFAMR) has been at the forefront of developing international guidance on AMR risk assessment and management. The WHO's Critically Important Antimicrobials for Human Medicine (CIA) list — which classifies antimicrobials used in medicine by their importance for treating serious infections when no alternatives exist — provides the basis for prioritizing risk management measures for antimicrobials used in animals.
Quantitative risk assessment for AMR in the food chain involves several analytically distinct but interconnected steps: estimation of the prevalence and concentration of resistant bacteria in foodproducing animals; modeling of the survival, transmission, and amplification of resistant organisms through slaughter, processing, and the food chain; estimation of human exposure through dietary consumption; dose-response modeling for gastrointestinal infection by the resistant organism; and
estimation of the additional probability that treatment with a critically important antimicrobial will fail due to resistance. This last step — quantifying treatment failure probability as the ultimate adverse health outcome — is conceptually distinct from conventional food safety risk assessment and requires epidemiological data on treatment patterns, clinical cure rates, and the relationship between in vitro resistance and clinical outcome.
9.3 Integrated Surveillance Systems
Effective AMR risk management in the food chain requires integrated surveillance systems that can track the emergence and spread of resistance in both animal and human populations, ideally linking AMR data to specific food commodities and consumption patterns through source attribution analysis. Several major surveillance initiatives have made significant progress in this direction:
• The U.S. National Antimicrobial Resistance Monitoring System (NARMS), operated jointly by CDC, FDA, and USDA, monitors AMR in enteric bacteria isolated from humans, retail meats, and food-producing animals at slaughter.
• The European Antimicrobial Resistance Surveillance Network (EARS-Net) and the European Food SafetyAuthority's Joint Inter-AgencyAntimicrobial Consumption and Resistance Analysis (JIACRA) reports provide integrated human and animal AMR surveillance across EU member states.
• The Global Antimicrobial Resistance and Use Surveillance System (GLASS), launched by WHO in 2015, provides a standardized approach for country-level AMR surveillance to support global policy-making.
Despite these advances, major data gaps persist, particularly in low- and middle-income countries where both AMR burden and antimicrobial use in agriculture are highest. Strengthening AMR surveillance infrastructure in these settings, and linking food safety and human health data through a One Health analytical framework, remains a global priority.
Chapter 10: Emerging Hazards in Food Systems
10.1 Defining and Identifying Emerging Hazards
An emerging hazard in food safety is defined as a newly identified biological, chemical, or physical agent or condition (including dietary practices or food technologies) whose occurrence in food represents a potential threat to human health that has not been previously recognized, or whose risk is known to be increasing. The identification of emerging hazards requires continuous horizon scanning — a systematic, prospective process of surveying the broader scientific literature, epidemiological surveillance data, trade statistics, climate and land use change data, and technological innovation reports to detect signals that might indicate the development of new food safety threats.
Emerging hazards can arise through multiple pathways: the emergence of new or mutant pathogens with altered virulence or host range (e.g., novel norovirus variants, emerging STEC serogroups); the introduction of new food ingredients, production methods, or technologies that have not been fully safety-assessed (e.g., novel protein sources, nanomaterials, gene-edited crops); the migration of previously geographically restricted hazards to new regions through climate change (e.g., expansion of aflatoxin-producing molds into previously cool temperate regions of Europe); and the emergence of resistance to existing control measures (e.g., the development of heat-tolerant Salmonella variants capable of surviving pasteurization).
10.2 Engineered Nanomaterials (ENMs) in Food
Engineered nanomaterials — materials with at least one dimension in the range of 1-100 nm — are finding increasing application in the food sector as food additives (e.g., silicon dioxide, titanium dioxide as anti-caking agents and whiteners), food packaging materials (e.g., silver nanoparticles for antimicrobial packaging, nanocomposites for barrier properties), and precision delivery systems for bioactive compounds, flavors, and nutrients. The unique physicochemical properties of nanomaterials — high surface-area-to-volume ratio, altered reactivity, and size-dependent optical and electrical properties — that make them technologically attractive may also confer distinct biological and toxicological properties compared to their bulk counterparts of the same chemical composition.
The risk assessment of ENMs presents fundamental challenges that have not been fully resolved by existing regulatory frameworks. Standard toxicological testing methods were designed for soluble chemicals or relatively insoluble particles in the micron size range; the unique behavior of nanoparticles — including their tendency to agglomerate in biological media, their interactions with macromolecules (the 'protein corona'), and their potential to translocate across epithelial barriers and accumulate in tissues — requires adaptation of existing methods and development of new assays. EFSA published a guidance document on the risk assessment of nanomaterials in 2018, acknowledging significant uncertainties and recommending a tiered testing strategy, but the scientific basis for comprehensive ENM risk assessment remains a work in progress.
Particular concern has been raised about titanium dioxide (E171) as a food additive. EFSA's 2021 opinion concluded that TiO2 could no longer be considered safe as a food additive, primarily because of genotoxicity concerns associated with nano-sized particles, leading to a European ban on E171 that took effect in 2022. This case illustrates how ENM-specific risk assessment concerns can lead to dramatic regulatory action, and underscores the importance of proactive pre-market safety assessment for nanotechnology applications in food.
10.3 Novel Protein Sources
Growing global population, resource constraints, and environmental sustainability imperatives are driving rapid development and commercialization of novel protein sources as alternatives to conventional meat and dairy. These include insect-based proteins, single-cell proteins derived from bacteria, yeast, or algae, plant-based meat analogs, and cell-cultivated (also called cultured or lab- grown) meat. Each of these categories presents distinctive food safety considerations that require adapted risk assessment approaches.
Insects represent a significant source of protein, fat, and micronutrients and are consumed by approximately 2 billion people worldwide in many traditional cultures. However, their use as food and feed ingredients in Western markets — facilitated by regulatory approvals in the EU under Novel Food Regulation (EU) 2015/2283 — raises safety questions including: the potential for microbiological contamination (particularly Salmonella, Listeria, and enteric viruses) if substrates are not properly controlled; the presence of chitin as an allergen and digestibility modifier; the potential for insect allergies in consumers with existing crustacean or house dust mite allergies (due to cross-reactive tropomyosin proteins); and the accumulation of chemical contaminants (heavy metals, mycotoxins, pesticides) from the rearing substrate.
Cell-cultivated meat presents perhaps the most novel risk profile, given that it has no history of safe use and its production process — in vitro cultivation of animal muscle and fat cells in bioreactor systems using cell culture media containing growth factors — is fundamentally different from any existing food production system. Key safety questions include: the safety of scaffolding materials and bioreactor components; the microbiological safety of the production environment (potential for contamination with pathogens not subject to the same environmental controls as conventional slaughter); the potential allergenicity of cell culture medium components transferred to the final product; and the safety of any genetic modifications introduced into the cell lines.
10.4 Climate Change and Food Safety
Climate change is increasingly recognized as a threat multiplier for food safety — altering the environmental conditions that determine the geographic distribution, seasonal occurrence, and behavior of foodborne hazards. A comprehensive 2010 WHO report identified several pathways through which climate change affects food safety: changes in the geographic range and intensity of mycotoxin-producing molds; altered occurrence and distribution of harmful algal blooms (HABs) producing marine biotoxins; shifts in the epidemiology of foodborne pathogens due to temperature effects on growth, survival, and the ecology of animal reservoirs; and indirect effects through impacts on water availability, food production systems, and trade patterns.
The northward expansion of aflatoxigenic Aspergillus species in Europe, driven by rising summer temperatures and increased drought stress in maize crops, represents one of the most extensively documented climate-food safety interactions. Predictive models suggest that aflatoxin contamination in European maize could increase substantially by 2030 under intermediate warming scenarios, creating significant regulatory and public health challenges in regions where aflatoxin has historically been a minor concern. Similar dynamics are expected for fumonisin and other Fusarium mycotoxins in small grain cereals.
Marine biotoxin events — toxic phytoplankton blooms that contaminate shellfish and fish with paralytic shellfish toxins (PSTs), amnesic shellfish toxins (ASTs), diarrhetic shellfish toxins (DSTs), and other marine biotoxins — are strongly influenced by sea surface temperature and salinity, both of which are affected by climate change. Increased frequency, geographic expansion, and intensity of harmful algal blooms are projected under most climate scenarios, with corresponding increases in the frequency ofshellfish harvesting closures and human poisoning incidents.
Chapter 11: Food Safety Risk Management Frameworks
11.1 The Interface Between Risk Assessment and Risk Management
The Codex Alimentarius model of risk analysis explicitly separates risk assessment (a science-based activity) from risk management (a policy process that weighs and selects options to control identified risks). This functional separation is designed to protect the scientific integrity of the risk assessment from political and commercial pressures, while ensuring that risk management decisions are transparently informed by the best available science. In practice, the boundary between risk assessment and risk management is not always sharp: risk managers necessarily make value judgments in commissioning risk assessments (what question to ask, what endpoint to use, what level of protection to target), and risk assessors inevitably make choices with policy implications (what default assumptions to apply, how to treat uncertainty).
Effective risk management follows a structured process: initial risk management activities (identifying and characterizing the risk management problem, preliminary risk profile); commissioning of a risk assessment (if needed) and evaluation of the risk assessment output; identification and selection of risk management options; implementation of the chosen option; monitoring and review of the effectiveness of the implemented measure. This iterative process ensures that risk management decisions are regularly revisited in light of new scientific information, changes in the food system, or evidence that existing measures are not achieving their stated objectives.
11.2 HACCP and Preventive Controls
Hazard Analysis and Critical Control Points (HACCP) is a systematic preventive approach to food safety that addresses physical, chemical, and biological hazards through analysis and control of biological, chemical, and physical hazards from raw material production, procurement, and handling, to manufacturing, distribution, and consumption of the finished product. HACCP was developed by NASA and the Pillsbury Company in the 1960s for the production of safe food for space missions, and has since become the internationally recognized standard for food safety management, mandated by law in many jurisdictions and required by food safety certification schemes (e.g., SQF, BRC, FSSC 22000).
The seven HACCP principles are:
1. Conduct a hazard analysis — identify all potential biological, chemical, and physical hazards that could occur in the food system and determine which are significant.
2. Determine Critical Control Points (CCPs) — identify points, steps, or procedures in the process where controls can be applied to prevent, eliminate, or reduce food safety hazards to an acceptable level.
3. Establish critical limits — specify the maximum and minimum values to which biological, chemical, or physical parameters must be controlled to prevent, eliminate, or reduce the occurrence of identified hazards.
4. Establish monitoring procedures — describe the measurements or observations used to assess whether a CCP is under control.
5. Establish corrective actions — define actions to be taken when monitoring indicates a deviation from an established critical limit.
6. Establish verification procedures — confirm that the HACCP system is working effectively, through activities such as additional testing, record review, and periodic revalidation.
7. Establish record-keeping and documentation — maintain records that demonstrate the HACCP system is functioning effectively and that critical limits have been met.
11.3 Risk-Benefit Analysis in Food Safety Decision-Making
Food safety risk management increasingly recognizes that foods are complex systems in which hazards coexist with nutritional benefits, and that risk management decisions must account for potential trade-offs between reducing hazard-related risk and diminishing beneficial nutritional components. This tension is most starkly illustrated by the methylmercury-omega-3 fatty acid tradeoff in fish consumption: fish are a major source of methylmercury (hazard) but also the principal dietary source of long-chain omega-3 fatty acids EPA and DHA (benefit), which are essential for fetal neurodevelopment. Excessive restriction offish consumption advice for pregnant women may reduce methylmercury exposure at the cost of inadequate omega-3 intake — potentially resulting in net harm rather than net benefit.
Risk-benefit analysis (RBA) for foods requires the definition of common metrics that allow comparison of harm and benefit effects of the same dietary pattern or food. Several approaches have been proposed: DALYs (disability-adjusted life years) and QALYs (quality-adjusted life years) provide a common health metric that can be applied to both disease burden (harm) and health outcomes (benefit); the Marginal Safety Index relates the magnitude ofthe risk change to the magnitude ofthe benefit change; and integrated nutritional risk models attempt to simultaneously capture both the hazard and nutritional profiles offoods within a single quantitative framework.
11.4 Precautionary Principle and Uncertainty Management
The precautionary principle — operationalized in EU food safety law through Article 7 of Regulation (EC) No 178/2002 — provides that where, following an assessment of available information, the possibility of harmful effects on health is identified but scientific uncertainty persists, provisional risk management measures may be adopted, pending further scientific information for a more comprehensive risk assessment. Precautionary measures must be proportionate, non-discriminatory, and provisional; they must be reviewed within a reasonable time in the light of developments in scientific knowledge.
The precautionary principle has been a source of significant trade disputes, particularly between the European Union and majoragricultural exporting nations. Critics argue that its broad application can serve as a disguised form of trade protectionism, restricting imports of products (e.g., genetically modified organisms, hormone-treated beef, chlorine-washed poultry) that are assessed as safe under the risk assessment frameworks of the exporting country but are politically contentious in the importing country. Defenders argue that the principle appropriately reflects the asymmetric consequences of Type I (false negative — declaring something safe when it is not) versus Type II (false positive — restricting something safe) errors in food safety regulation, given the potentially irreversible consequences ofwidespread harm.
Chapter 12: Risk Communication and Stakeholder Engagement
12.1 Principles of Effective Risk Communication
Risk communication — the interactive exchange of information and opinions about risk among risk assessors, risk managers, and the public — is the third pillar of the Codex risk analysis framework. Effective risk communication goes far beyond the transmission of technical information; it encompasses the building of trust, the acknowledgment of uncertainty, the recognition of stakeholder values and concerns, and the facilitation of informed decision-making by individuals, communities, and policymakers. Poor risk communication — characterized by inconsistency, apparent concealment of information, failure to acknowledge uncertainty, or condescension toward public concerns — can severely undermine confidence in food safety systems and lead to consumer responses (food avoidance, excessive anxiety, failure to take appropriate precautions) that are themselves harmful to public health.
Research in risk perception has consistently demonstrated that the public's assessment of food safety risks does not conform to a simple expected-value model in which risk magnitude (probability x severity) drives concern. Psychological factors including dread (the fear of uncontrollable, catastrophic outcomes), novelty (unfamiliarity with the hazard), and lack of personal control powerfully amplify risk perception beyond what purely actuarial risk levels would predict. Conversely, familiar, voluntary risks — such as the risks associated with alcohol consumption — are typically underestimated relative to their actual magnitude. Risk communicators must be sensitive to these psychological dynamics and avoid the temptation to dismiss or invalidate public risk perceptions as 'irrational.'
12.2 Communicating Uncertainty to Non-Expert Audiences
Perhaps the greatest challenge in food safety risk communication is conveying the inherent uncertainty of risk estimates to audiences who may prefer — and expect — definitive answers. Scientists are trained to view uncertainty as an honest and essential feature of knowledge; many members of the public, however, interpret expressions of uncertainty as evidence of ignorance, incompetence, or concealment. This mismatch in expectations can lead communicators to avoid explicit acknowledgment of uncertainty, resulting in overconfident statements that subsequently undermine credibility when the science evolves.
Effective strategies for communicating uncertainty include: explaining the distinction between variability (the food supply is heterogeneous; not all fish have the same mercury content) and uncertainty (we don't know precisely how much mercury is in this specific fish); using visual aids such as confidence intervals, fan charts, or scenario ranges to convey the spread of plausible outcomes; contextualizing uncertainty relative to other uncertain decisions that people routinely make; and being explicit about what is known, what is unknown, and what is unknowable — while still providing actionable guidance.
12.3 Digital Media, Social Networks, and Food Safety Misinformation
The proliferation of digital media and social networking platforms has fundamentally transformed the food safety information environment. While digital technologies offer unprecedented opportunities for rapid dissemination of food safety warnings, consumer education, and stakeholder engagement, they have also created fertile conditions for the spread of food safety misinformation — false or misleading claims aboutthe safety of specific foods, ingredients, or production methods that can cause significant harm through unnecessary food avoidance, economic damage to food businesses, and erosion of confidence in the food safety system.
Research on the spread of food safety misinformation on social media suggests that false information spreads faster, farther, and more broadly than true information — a phenomenon attributable to the higher emotional resonance, novelty, and shareability of sensational or fear-evoking content. Food safety agencies and communicators face an asymmetric challenge: it takes significantly more effort to correct a false narrative than to create one, and corrections often fail to fully displace prior misinformation even when they reach the same audience. Proactive prebunking — warning audiences in advance about the tactics used in food safety misinformation, and equipping them with critical appraisal skills — has shown promise as a more effective strategy than reactive debunking.
Chapter 13: Integrated Risk Assessment — Case Studies
13.1 Case Study: Listeria monocytogenes in Ready-to-Eat Deli Meats
The FDA/USDA 2003 risk assessment for Listeria monocytogenes in ready-to-eat foods remains one of the most influential QMRA documents ever published, providing a methodological template that has been widely emulated and the scientific foundation for risk-based regulatory policies in both the United States and, through Codex channels, internationally. This case study summarizes the methodology and key findings of that assessment, as well as subsequent regulatory developments.
The risk assessment evaluated 23 RTE food categories classified by their ability to support L. monocytogenes growth (growth-supporting vs. non-growth-supporting) and by the likely level of contamination at retail. The four categories found to pose the greatest risk per serving and per annum were: deli meats — high contamination, growth supporting; frankfurters — not reheated; pate and meat spreads; and smoked seafood. The finding that deli meats posed the greatest absolute risk per annum was particularly significant, given that this category accounts for a substantial proportion of total listeriosis cases based on source attribution data.
Key factors driving risk in the deli meat category included: the potential for significant post-lethality contamination at retail deli slicers; growth to high concentrations during extended refrigerated storage; and high consumption rates in the population. Sensitivity analysis identified temperature control and product shelf life as the parameters most strongly influencing risk — consistent with the recommendation that maintaining strict cold chain integrity and minimizing refrigerated storage time were among the highest-priority control measures.
The risk assessment directly informed the USDA FSIS Listeria Rule (9 CFR Part 430), which established three alternatives for RTE meat and poultry facilities to demonstrate compliance: use of a post-lethality treatment capable of eliminating L. monocytogenes, combined with an antimicrobial agent or process that suppresses growth; use of a post-lethality treatment alone; or use of a sanitation program alone. The rule used a risk-based tiering structure in which facilities using the highest-control alternative (Alternative 1)were subject to the least frequent regulatory testing.
13.2 Case Study: Acrylamide in Starchy Foods
The discovery of acrylamide in heat-processed starchy foods — first reported by Swedish researchers in 2002 — generated immediate global regulatory concern, given that acrylamide is classified as a probable human carcinogen (Group 2A) by IARC on the basis of evidence from animal studies and occupational exposure epidemiology. Acrylamide forms via the Maillard reaction between the amino acid asparagine and reducing sugars (primarily glucose and fructose) during high-temperature cooking (frying, baking, roasting) at temperatures above approximately 120°C. The highest concentrations are found in potato crisps, French fries, biscuits, bread, breakfast cereals, and coffee.
The risk assessment challenge for acrylamide is substantial: dietary exposure is widespread and ubiquitous across all age groups; the primary mechanism of carcinogenicity (genotoxicity via the metabolite glycidamide) is consistent with a non-threshold dose-response; and reducing dietary exposure to zero is not practically feasible without eliminating large categories of culturally significant cooked foods. Regulatory responses have therefore focused on mitigation — reducing acrylamide formation to the lowest reasonably achievable level — ratherthan prohibition.
EFSA's 2015 risk assessment estimated that acrylamide dietary exposure was in the range of concern for neoplastic effects across all age groups in the European population, and that no single food category was responsible for a dominant share of total exposure. Mitigation measures identified in the Codex Code of Practice for the Reduction of Acrylamide in Foods (CAC/RCP 67-2009) include: selection of low-asparagine or low-reducing-sugar potato varieties; adjustment of storage conditions for raw materials; modification of cooking temperatures and times; and use of asparaginase enzyme to enzymatically degrade asparagine in dough before baking.
13.3 Case Study: Aflatoxins in Maize in Sub-Saharan Africa
Aflatoxin contamination of maize in sub-Saharan Africa represents a particularly severe food safety crisis, combining very high contamination levels, a population with high dietary exposure due to reliance on maize as a staple food, high background rates of hepatitis B virus infection (which synergistically amplifies aflatoxin-related liver cancer risk), and severe constraints on analytical capacity and cold chain infrastructure that limit the effectiveness of conventional control measures.
Acute aflatoxicosis outbreaks — caused by consumption of heavily contaminated grain, typically during food shortages when quality control is relaxed — have resulted in mass casualties in Kenya (2004: 317 cases, 125 deaths; 2016: 75 deaths), Tanzania, and other African nations. These outbreaks are distinct from the chronic low-level hepatocellular carcinoma risk that dominates the public health burden in high-income country risk assessments, and they represent a distinct risk management challenge: the immediate, visible humanitarian emergency must be addressed while simultaneously implementing long-term strategies to reduce chronic exposure.
Interventions with demonstrated efficacy include: use of aflasafe — a biological control agent based on competitive exclusion of toxigenic by atoxigenic Aspergillus strains — that significantly reduces pre-harvest aflatoxin contamination; improved drying and storage technologies (hermetic storage bags, improved grain stores) that reduce post-harvest contamination; and the use of aflatoxin biomarkers (urinary AFM1 and serum aflatoxin-albumin adducts) to monitor population exposure and evaluate intervention effectiveness.
Chapter 14: Digital Technologies and the Future of Food Safety Risk Assessment
14.1 Big Data, Artificial Intelligence, and Machine Learning
The digital transformation ofthe food system — encompassing precision agriculture, real-time supply chain monitoring, electronic point-of-sale data, social media outbreak signals, and genomic surveillance of pathogens — is generating data resources of unprecedented scale and richness that offer transformative opportunities for food safety risk assessment. However, realizing these opportunities requires substantial advances in data integration, analytical methodology, and the governance of data ownership and access.
Whole genome sequencing (WGS) of foodborne pathogens has revolutionized outbreak investigation by enabling high-resolution discrimination of pathogen strains and near-real-time identification of epidemiological clusters. PulseNet — the molecular subtyping network coordinated by CDC — has successfully used WGS to identify multi-state outbreak clusters linked to specific food vehicles and contamination sources, leading to numerous successful recalls and attributing illnesses that would previously have gone unrecognized as outbreak-related. The global GenomeTrakr network, coordinated by FDA, now holds over 300,000 pathogen genome sequences from food, environmental, and clinical sources, providing an unprecedented resource for source attribution, antimicrobial resistance profiling, and the identification of hyper-virulent or particularly hazardous clonal lineages.
Machine learning algorithms are increasingly being applied to food safety risk assessment problems, including: predictive modeling of contamination events in supply chains from environmental, weather, and production data; image recognition for rapid identification of visible mold contamination or physical defects; natural language processing of food safety inspection records, consumer complaint databases, and scientific literature to identify emerging hazard signals; and ensemble modeling approaches that integrate outputs from multiple predictive models to improve risk ranking accuracy.
14.2 Blockchain and Supply Chain Transparency
Blockchain technology — a distributed ledger system that provides an immutable, tamper-evident record of transactions — offers promising applications for food supply chain traceability, enabling rapid and reliable identification of contamination sources during food safety incidents. In 2018, an outbreak of E. coli O157:H7 linked to romaine lettuce from the Yuma, Arizona growing region required approximately 7 days to trace the source, during which time a precautionary advisory was issued to avoid all romaine lettuce in the United States — a public health measure with significant economic consequences for the entire sector. Blockchain-based traceability systems, if universally adopted, could potentially compress the traceback time from days to hours or even minutes, enabling more targeted recalls with smaller economic footprint.
Walmart's collaboration with IBM on a blockchain-based food traceability initiative demonstrated that the time required to trace a mango from a store shelf to its farm of origin could be reduced from 6.5 days to 2.2 seconds using blockchain technology. Several major food retailers and manufacturers have since implemented or piloted blockchain traceability systems for high-risk commodities including leafy greens, berries, and ground beef. However, the effectiveness of blockchain traceability depends critically on data quality at the point of entry — 'garbage in, garbage out' — and requires standardization of data capture methods and formats across the supply chain to be fully effective.
14.3 Cyberbiosecurity as an Emerging Dimension
The increasing digitization and connectivity of food production systems — from automated agricultural equipment controlled by GPS and artificial intelligence, to smart processing plants with networked sensors and control systems, to automated logistics and cold chain management — creates novel attack surfaces for malicious actors seeking to compromise food safety. Cyberbiosecurity — the intersection of cybersecurity and food/agricultural biosecurity — is an emerging field that addresses the risk that cyberattacks, including ransomware, targeted manipulation of control systems, or data poisoning attacks, could be used to deliberately introduce food safety hazards or cause food supply disruptions.
Documented examples of cyber threats to food systems include the 2021 ransomware attack on JBS Foods — the world's largest meat processing company — that disrupted operations at multiple U.S., Canadian, and Australian plants for several days; and the 2021 cyberattack on the Oldsmar, Florida water treatment facility in which an attacker remotely modified the concentration of sodium hydroxide (lye) in the drinking water supply. While the Oldsmar attack was detected and remediated before it caused harm, it demonstrated the real potential for cyberattacks on critical food and water infrastructure to have direct public health consequences.
14.4 Harmonization of Risk Assessment Methods Across Jurisdictions
Despite significant advances in international coordination through Codex and bilateral regulatory cooperation agreements, substantial differences persist in risk assessment methodologies, data requirements, and risk management outcomes across regulatory jurisdictions. These differences — arising from variations in underlying risk tolerances, legal frameworks, data availability, and the political economy of food safety regulation — create significant challenges for international food trade and for multinational food companies seeking to navigate diverse regulatory requirements.
Key areas of methodological divergence include: the treatment of non-threshold carcinogens (the MoE approach favored by EFSA versus the ILCR approach used by U.S. EPA); the application of the precautionary principle in the face of scientific uncertainty; the definition and assessment criteria for endocrine-disrupting chemicals; the regulatory treatment of genetically modified organisms; and the risk assessment framework for novel foods and novel protein sources. The development of internationally harmonized risk assessment guidance — building on the Codex principles but going further in specifying methodological defaults — is a priority for the international risk assessment community.
Chapter 15: Integrated Approaches and Future Perspectives
15.1 One Health Approaches to Food Safety Risk Assessment
The One Health concept— recognizing the fundamental interconnectedness of human, animal, and environmental health — provides an essential integrative framework for contemporary food safety risk assessment. Many of the most significant food safety hazards, including zoonotic pathogens, antimicrobial resistance, and environmental contaminants, originate at the human-animalenvironment interface and cannot be effectively understood or controlled through a siloed, sectorspecific approach. True One Health risk assessment integrates epidemiological data from human, animal, and environmental surveillance systems; models pathogen transmission and hazard dispersal across the full ecological network; and evaluates control interventions at all points in the system, including at the human-animal interface and in the environment.
Implementation of One Health approaches in food safety risk assessment is constrained by institutional, disciplinary, and data-sharing barriers. Veterinary, public health, and environmental agencies often operate in separate regulatory silos with different legislative mandates, different data systems, and different institutional cultures. Overcoming these barriers requires deliberate institutional design — including joint surveillance systems, shared data platforms, integrated risk assessment exercises, and cross-trained workforces — as well as political commitment at the highest levels.
15.2 Precision Nutrition and Personalized RiskAssessment
Advances in nutritional genomics, metabolomics, and the microbiome sciences are beginning to challenge the 'average consumer' models that underlie most current food safety risk assessments. The emerging field of precision nutrition recognizes that individuals vary substantially in their metabolic response to foods and food contaminants due to genetic polymorphisms, epigenetic variation, gut microbiome composition, and other biological factors. For some chemical hazards — notably those that are activated or detoxified by polymorphic enzymes in Phase I and Phase II metabolism — individual variation in susceptibility may span an order of magnitude or more, calling into question the adequacy ofthe standard 10-fold intraspecies uncertaintyfactor.
Personalized risk assessment — the application of individual-level biological data to generate customized risk estimates and dietary recommendations — remains a long-term aspiration rather than a current reality for most food safety hazards. However, in specific contexts — such as the risk of methylmercury neurotoxicity in pregnant women with known polymorphisms in mercury methylation enzymes, or the risk of celiac disease in individuals screened for HLA-DQ2/DQ8 genotype — personalized risk assessment approaches are already clinically and scientifically feasible, and may provide the template for broader application as the costs of genomic and metabolomic profiling continue to fall.
15.3 Climate Resilience and Future-Proofing Food Safety Systems
Building food safety systems that are resilient to the ongoing and projected impacts of climate change requires both adaptation (modifying existing monitoring, surveillance, and control systems to address climate-driven changes in hazard profiles) and mitigation (reducing the contributions of food systems to greenhouse gas emissions that drive climate change). The food industry is a significant contributor to global greenhouse gas emissions, and efforts to improve environmental sustainability — such as reducing food loss and waste, shifting toward lower-emissions protein sources, and optimizing cold chain energy use — may have co-benefits for food safety that should be recognized and quantified in risk-benefit frameworks.
Climate-adaptive food safety monitoring systems must incorporate climate projections into predictive models of hazard occurrence, contamination levels, and control efficacy. Scenario-based risk assessment — evaluating risk under a range of plausible future climate scenarios — can inform both the design of new food safety infrastructure (e.g., enhanced mycotoxin monitoring in regions where climate projections indicate increased aflatoxin risk) and the development of climate-adaptive Good Agricultural Practices. International coordination through FAO, WHO, and Codex will be essential to ensure that climate adaptation in food safety is globally equitable and does not simply shift risk to the most vulnerable countries.
15.4 Strengthening Risk Assessment Capacity in Low- and MiddleIncome Countries
The global burden of foodborne disease falls disproportionately on low- and middle-income countries (LMICs), yet risk assessment capacity — in terms of trained personnel, analytical laboratory infrastructure, epidemiological surveillance systems, and regulatory frameworks — is most limited in precisely these settings. Strengthening risk assessment capacity in LMICs is therefore not only a matter of equity but also a global food safety imperative, since pathogen emergence, AMR, and mycotoxin contamination are global threats that cannot be effectively controlled without robust scientific and regulatory capacity in the regions where they originate.
Capacity-strengthening initiatives must go beyond the provision of training courses and equipment; they must invest in sustainable institutional development, including the establishment of functioning national food safety authorities with genuine legislative mandate and enforcement capacity; national total diet studies to characterize dietary exposure; accredited food testing laboratories capable of meeting international standards; and effective linkage between the scientific risk assessment function and the regulatory risk management function. South-South cooperation — knowledge sharing between countries at similar stages of development — is an increasingly important modality, complementing but not replacing the technical assistance provided by international organizations and bilateral development partners.
15.5 Conclusion: Toward an Integrated, Adaptive Food Safety Risk Assessment Paradigm
The science and practice of food safety risk assessment have advanced dramatically over the past three decades, moving from largely qualitative hazard identification to sophisticated quantitative probabilistic modeling, from single-hazard focus to integrated multi-hazard frameworks, and from reactive outbreak response to proactive preventive systems management. Yet the challenges facing the field have grown in parallel with its capabilities: climate change, AMR, emerging technologies, global supply chains, and digital disruption are creating a food safety landscape of unprecedented complexity.
Meeting these challenges will require not only continued methodological innovation but also institutional reform: stronger international coordination in risk assessment methodology, data sharing, and regulatory harmonization; greater investment in surveillance and monitoring systems, particularly in LMICs; more effective translation of risk assessment science into risk management action; and more genuine, two-way engagement with consumers and other stakeholders in the risk analysis process.
The four-step risk assessment paradigm established in the 1983 NRC report remains a sound and durable framework, but it must be interpreted flexibly and supplemented by new approaches — Bayesian methods, WGS-based surveillance, predictive modeling, risk-benefit analysis — that were not available to its framers. The goal remains constant: to provide the best possible scientific basis for protecting the health of consumers while enabling a safe, nutritious, and sustainable global food supply.
Looking Forward: Key Priorities for the Field
The most important priorities for advancing food safety risk assessment over the coming decade include: (1) development of validated QMRA models for viruses and parasites, where doseresponse data remain severely limited; (2) integration of AMR endpoints into standard food safety risk assessment frameworks; (3) harmonization of nanomaterial risk assessment methodology internationally; (4) development of climate-adaptive risk assessment tools; (5) application of One Health analytical frameworks to complex, multi-hazard food safety problems; and (6) investment in risk assessment capacity in LMICs as a global public good.
Glossary of Key Terms
Illustrations are not included in the reading sample
Selected References and Further Reading
Foundational Texts
Codex Alimentarius Commission. (2007). Working Principles for Risk Analysis for Food Safety for Application by Governments (CAC/GL 62-2007). Rome: FAO/WHO.
EFSA Scientific Committee. (2018). Guidance on the risk assessment of nanomaterials to be applied in the food and feed chain: human and animal health. EFSA Journal, 16(7), e05327.
FAO/WHO. (2009). Risk characterization of microbiological hazards in food: guidelines. Microbiological RiskAssessment Series No. 17. Rome: FAO.
FAO/WHO. (2014). Multicriteria-based ranking for risk management of food-borne parasites. Microbiological Risk Assessment Series No. 23. Rome: FAO.
Food and Drug Administration / Food Safety and Inspection Service. (2003). Quantitative Assessment of Relative Risk to Public Health from Foodborne Listeria monocytogenes Among Selected Categories of Ready-to-Eat Foods. Washington, D.C.: FDA/USDA.
ICMSF. (2002). Microorganisms in Foods 7: Microbiological Testing in Food Safety Management. New York: KluwerAcademic.
Nauta, M. J. (2008). The modular process risk model (MPRM): A structured approach to food chain exposure assessment. In D. W. Schaffner (Ed.), Microbial Risk Analysis of Foods. Washington, D.C.: ASM Press.
National Research Council. (1983). Risk Assessment in the Federal Government: Managing the Process. Washington, D.C.: National Academy Press.
WHO. (2015). WHO estimates of the global burden of foodborne diseases: Foodborne Disease Burden Epidemiology Reference Group 2007-2015. Geneva: World Health Organization.
Microbiological RiskAssessment
Buchanan, R. L., Smith, J. L., & Long, W. (2000). Microbial risk assessment: Dose-response relations and risk characterization. International Journal of Food Microbiology, 58(3), 159-172.
Haas, C. N., Rose, J. B., & Gerba, C. P. (2014). Quantitative Microbial RiskAssessment (2nd ed.). Hoboken, NJ: John Wiley & Sons.
Havelaar, A. H., et al. (2015). World Health Organization global estimates and regional comparisons ofthe burden offoodborne disease in 2010. PLOS Medicine, 12(12), e1001923.
Mead, P. S., et al. (1999). Food-related illness and death in the United States. Emerging Infectious Diseases, 5(5), 607-625.
Chemical RiskAssessment
EFSA Panel on Contaminants in the Food Chain (CONTAM). (2009). Scientific opinion on arsenic in food. EFSAJournal, 7(10), 1351.
EFSA Panel on Contaminants in the Food Chain (CONTAM). (2021). Assessment of the dietary exposure to aflatoxins in the European population. EFSA Journal, 19(3), e06410.
EFSA Scientific Committee. (2012). Scientific opinion on the principles and methods behind EFSA's guidance on uncertaintyanalysisinscientific assessment. EFSAJournal, 16(1), e05122.
FAO/WHO. (2019). JMPR Report: Pesticide residues in food 2018. JMPR Monographs Series. Rome: FAO.
IARC Working Group. (2012). Chemical agents and related occupations. IARC Monographs on the Evaluation ofCarcinogenic Risksto Humans, 100F. Lyon: IARC.
Emerging Hazards
EFSA Scientific Committee. (2021). Statement on the safety of titanium dioxide (E171) as a food additive. EFSAJournal, 19(5), e06585.
FAO/WHO. (2020). Food safety aspects of cell-based food. Meeting Report. Rome: FAO.
Grace, D. (2015). Food safety in low and middle income countries. International Journal of Environmental Research and Public Health, 12(9), 10490-10507.
Interagency Coordination Group on AMR. (2019). No time to wait: Securing the future from drugresistant infections. Reportto the Secretary-General ofthe United Nations. Geneva: WHO.
Mbow, C., et al. (2019). Food Security. In IPCC Special Report on Climate Change and Land. Geneva: IPCC.
Tiede, K., et al. (2008). A review of the environmental fate of nanomaterials in terrestrial ecosystems. Science ofthe Total Environment, 400(1-3), 396-414.
Risk Communication and Management
Covello, V. T., & Sandman, P. M. (2001). Risk communication: Evolution and revolution. In A. Wolbarst (Ed.), Solutions to an Environment in Peril. Baltimore: Johns Hopkins University Press.
EFSA. (2019). Risk communication guidelines for food safety authorities. EFSA Supporting Publications, 16(4), EN-1641.
Slovic, P. (1987). Perception of risk. Science, 236(4799), 280-285. van Asselt, M. B. A., & Vos, E. (2008). Wrestling with uncertain risks: EU regulation of GMOs and the uncertainty paradox. Journal of Risk Research, 11(1-2), 281-300.
[...]
Buy now
- Quote paper
- Alfi Sophian (Author), 2026, Food Safety Risk Assessment. Integrating Microbiological, Chemical, and Emerging Hazards, Munich, GRIN Verlag, https://www.grin.com/document/1708705
