High-heat pet food processing generates acrylamide and AGEs, review finds

Every day, millions of dogs and cats eat the same bowl of kibble or canned food, meal after meal, for years. A review published in Frontiers in Veterinary Science argues that this repetitive feeding pattern deserves closer scrutiny, because the high-temperature processing that makes traditional pet food shelf-stable also generates Maillard reaction products and acrylamide, compounds associated in humans and rodents with diabetes, cardiovascular disease, neurodegeneration, and certain cancers. The review, led by Andreas Hopf and Fariba Dehghani of the University of Sydney together with Matthew Muir and Alex England, concludes that while these compounds are absorbed and measurable in the serum and urine of companion animals, adverse health outcomes attributable to them have not been demonstrated in dogs or cats. No regulatory benchmarks for acrylamide exist in pet foods anywhere, and no benchmarks for advanced glycation end products exist in any jurisdiction for any food.

The chemistry begins with the Maillard reaction, a non-enzymatic browning process between reducing sugars and protein-derived amino groups. It proceeds through three broadly defined stages. In the early stage, reducing sugars bind to the amino group of lysine to form Amadori compounds such as fructoselysine, whose hydrolysis product furosine serves as an indicator of early-stage activity. Intermediate-stage reactions generate reactive carbonyl compounds including glyoxal, methylglyoxal, and 3-deoxyglucosone, along with 5-hydroxymethyl-2-furfural. In the advanced stage, these intermediates form stable, irreversible end products known as advanced glycation end products, or AGEs. The most abundantly studied are Nε-carboxymethyllysine and Nε-carboxyethyllysine, with CML often used as a proxy marker for total dietary AGE exposure. More than twenty distinct Maillard reaction products have been identified to date, although quantitative data in pet foods remain limited.

Acrylamide follows a related but distinct pathway, forming from asparagine and reducing sugars such as fructose or glucose at temperatures above roughly 120 °C. Because of this separate route, it is classified as a contaminant rather than a Maillard reaction product. The International Agency for Research on Cancer classifies acrylamide as a Group 2A probable human carcinogen and a recognized neurotoxin, and government agencies in the European Union, the United States, Australia, and New Zealand monitor its content in human food. In rodent carcinogenicity studies, chronic oral administration induces tumors of the thyroid, mammary glands, and testicular mesothelium. Once ingested, acrylamide is rapidly absorbed and converted by cytochrome P450 enzymes to glycidamide, which forms DNA adducts and is considered responsible for the compound’s genotoxic effect. The European Food Safety Authority has concluded that no safe level of acrylamide intake can be defined.

Traditional pet food manufacturing exposes ingredients to temperatures ranging from approximately 90 °C to over 200 °C. Extrusion, the most common technique for dry foods, mixes powdered ingredients with water and subjects the mixture to high temperature and pressure, typically between 80 and 200 °C, to achieve starch gelatinization and microbiological stability before the product is cut, dehydrated with hot air, and coated. Retorting, used for canned and wet formats, seals blended ingredients in containers and applies pressurized heat at 121 °C or higher for sterilization. Pelleting operates at lower temperatures, typically 65 to 90 °C, but many pre-processed ingredients entering the process, particularly rendered animal co-products such as meat meal, have already accumulated thermal exposure and carry Maillard reaction products and acrylamide into the finished product. The review emphasizes that the complete processing history of a pet food, not only the final formulation step, must be considered in product design.

Measured concentrations vary substantially across formats. Wet and canned pet foods consistently contain the highest AGE concentrations across studies, while the ranking of mildly processed and extruded dry foods varies by marker. In 26 commercial cat foods, total AGE concentrations in mildly cooked and extruded dry diets did not differ, and CML was lowest in extruded dry diets. Acrylamide was detected primarily in extruded dry food, with reported values ranging from 13 to 358 μg/kg across studies, while it was not detected in mildly processed foods. Estimated daily intakes are elevated in companion animals relative to humans. Adult dogs consuming wet diets are estimated to ingest 1.0 mg/kg metabolic body weight per day of CML and 50 mg/kg of HMF, with cats on wet diets estimated at 0.8 and 20 mg/kg respectively. Estimated daily acrylamide intake for dogs fed dry food is approximately 590 ng/kg body weight per day, roughly four times higher than that of adult humans, although this comparison is expressed on a body weight rather than metabolic body weight basis.

The review contextualizes these figures against human foods, noting that pet food values are reported on a dry matter basis while human food values are reported on a fresh weight basis, so the two are not directly comparable and pet food values appear inherently higher due to water removal. Baby food serves as a particularly relevant comparison because it represents a thermally processed, single-source diet fed over an extended period. Meta-analysis of acrylamide in baby food reported mean concentrations as high as 156 μg/kg, exceeding the European Union benchmark of 40 μg/kg for cereal-based baby foods established under Regulation (EU) 2017/2158. The European Commission has set acrylamide benchmark levels for several human food categories under that regulation, yet no equivalent standards exist for pet foods in any jurisdiction.

In humans and rodents, dietary AGEs exert pathological effects through two primary mechanisms. First, AGEs bind to receptors on cell membranes, most notably the receptor for advanced glycation end products, RAGE, activating signaling pathways that trigger oxidative stress and chronic inflammation. Second, AGEs directly cross-link tissue proteins, altering their structural and functional properties. In diabetes mellitus, these mechanisms impair insulin signaling and promote pancreatic β-cell dysfunction, creating a self-reinforcing cycle of inflammation and insulin resistance. In non-obese diabetic mice, feeding an AGE-restricted diet reduced disease incidence to less than 15 percent, with protective effects persisting across two subsequent generations. In cardiovascular disease, cross-linking of collagen and elastin in blood vessel walls leads to arterial stiffness and cardiac fibrosis, while AGE modification of plasma lipoproteins increases their atherogenic potential. Healthy rats fed a Western-style diet for 18 weeks developed hyperglycaemia, elevated AGE levels, and cardiac impairment. In neurodegeneration, rodent studies show that AGEs promote phosphorylation of Tau protein, and a clinical study of older adults without dementia reported that higher dietary AGE intake was associated with faster decline in memory.

The companion animal evidence base consists mainly of biomarker and observational data. In observational studies of privately owned animals, dogs fed traditionally processed dry food excreted urinary CML, CEL, and LAL at 2.03, 2.14, and 3.0 times higher concentrations respectively than raw-fed dogs, with cats showing a similar pattern. A controlled four-week Latin-square crossover trial in healthy adult dogs fed four diets differing in thermal processing found total plasma AGEs highest in dogs fed traditionally processed wet food and lowest in those fed mildly processed food, at medians of 9.99 versus 2.50 nM/50 μl. A 12-month longitudinal study in geriatric dogs reported lower serum CML and pyrraline concentrations and reduced AGE-associated metabolomic signatures in dogs fed mildly processed food compared with kibble. However, the review cautions that in both studies the diets differed in composition as well as in processing intensity, so the observed biomarker differences cannot be attributed to processing alone. Carbohydrate content, which is higher in dry formats, contributes to Maillard reaction formation during manufacture and also influences postprandial glycaemic responses.

Acrylamide exposure in companion animals has also been measured directly. In a cross-sectional environmental exposure study, the geometric mean urinary concentration of AAMA, the primary acrylamide metabolite, was 36.6 ng/ml in dogs and 11.1 ng/ml in cats, with 77 percent of dogs and 50 percent of cats exceeding the hazard quotient threshold of 1 for non-carcinogenic effects, and all animals exceeding the screening-level cancer risk benchmark applied by the authors. Dogs are known to be susceptible to acrylamide-induced neuropathy, with oral administration at 7 mg/kg per day producing first clinical signs after 20 to 28 days. A case of acute neurotoxicity in Labrador puppies that developed ataxia and convulsions after consuming burnt maize porridge was suggested to be consistent with acrylamide toxicity, although the compound’s content in the food was never established. The conversion rate of acrylamide to glycidamide remains unknown for dogs, and the catalytic efficiency of canine cytochrome P450 2E1 toward acrylamide has not been determined. The estimated daily canine intake of 590 ng/kg is well below the rodent no-observed-adverse-effect level for neurotoxicity of 0.2 to 0.5 mg/kg per day, but the review notes this comparison does not account for the cumulative effect of sustained daily exposure over years, nor for potential species differences in metabolism and susceptibility.

The authors argue that the more relevant comparison may be long-term cumulative burden rather than single-dose toxicity thresholds, since pets may consume the same thermally processed food regularly for years, exceeding the duration of existing experimental models. Where exposure cannot be eliminated, they suggest the As Low As Reasonably Achievable principle, widely applied in food contaminant regulation, could provide a practical framework for setting manufacturing targets in the absence of established safe thresholds. Priority research needs include prospective cohort studies linking dietary Maillard reaction product exposure to clinical outcomes, species-specific biomarkers, and long-term controlled feeding trials, which would provide the strongest evidence but remain costly and logistically challenging. Dogs with canine cognitive dysfunction may serve as translational models to investigate whether dietary AGE restriction can mitigate neurodegeneration. The review also stresses that mildly processed diets carry trade-offs, including a smaller microbiological safety margin, dependence on cold-chain distribution, and limits on affordability and accessibility, so lower Maillard reaction product content is one consideration among several rather than a basis for recommending a dietary category. Microplastics have recently been reported in commercial dry pet foods in European and South American markets, meaning a single processed staple may carry several unrelated contaminant classes under the same pattern of chronic exposure. Given the plausible shared pathophysiology and the absence of species-specific safety data, the authors conclude that a precautionary approach to pet food safety standards and veterinary nutrition guidance may be warranted.

Subject of Research: Dietary acrylamide and advanced glycation end products in processed pet food and their potential implications for dog and cat health

Article Title: Dietary acrylamide and AGE products in processed pet food: a review of the evidence and implications for safety thresholds in dogs and cats

Article References: Hopf, A., Muir, M., England, A., & Dehghani, F. (2026). Dietary acrylamide and AGE products in processed pet food: a review of the evidence and implications for safety thresholds in dogs and cats. Frontiers in Veterinary Science, 13, Article 1970122. https://doi.org/10.3389/fvets.2026.1970122

Image Credits: AI Generated

DOI: 10.3389/fvets.2026.1970122

Keywords: acrylamide, advanced glycation end products, Maillard reaction, pet food, dogs, cats, thermal processing, food safety, chronic disease, kibble, regulatory benchmarks, companion animals