A Monte Carlo-based predictive model to support acrylamide risk mitigation in fast-food French fries

Acrylamide, a probable human carcinogen formed via the Maillard reaction, remains a persistent food safety concern in fried potato products. Despite standard operating procedures (SOPs), field observations reveal substantial variability in fast-food frying operations driven by differences in raw material composition, oil management, temperature, time, and human decision-making. This study developed and validated a Monte Carlo–based predictive model, grounded in multiple linear regression, to quantify acrylamide formation in shoestring French fries using empirical data from selected fast-food chains. Implemented in @RISK software (10,000 iterations, triplicate runs), the model integrates real-world operational variability to produce probabilistic predictions that closely matched measured data (Kolmogorov–Smirnov, p = 1.000). Unlike models derived from controlled laboratory settings, this approach reflects the complexity of commercial environments, enhancing its relevance for risk management. Sensitivity analysis identified glucose content, moisture, frying time, and temperature as the most influential variables. Scenario analysis demonstrated that maintaining frying temperatures below 170 °C and limiting frying time to under 3 min are high-impact, feasible mitigation measures. Additional scenarios covering oil type, thawing practices, and fryer usage highlight operational factors contributing to unexplained variance, offering complementary avenues for risk reduction. The model delivers quantified, scenario-based outcomes across diverse frying practices, enabling regulators and industry stakeholders to assess intervention effectiveness before implementation. While demonstrated for acrylamide in French fries, the framework is adaptable to other chemical hazards and food sectors, providing a versatile, data-driven platform to anticipate risks, optimise mitigation strategies, and strengthen food safety governance globally.

Text 124673.pdf - Published Version Restricted to Repository staff only Download (3MB) Official URL or Download Paper: https://linkinghub.elsevier.com/retrieve/pii/S0956...

Actions (login required)

View Item