Kepinska-Pacelik et al. 2023 - Trace elements in nutraceutical dog snacks

Kepinska-Pacelik and colleagues produced four dog-snack prototypes and measured proximate composition, minerals, trace elements, structure, and palatability. This is finished pet-treat evidence for experimental prototypes, not a commercial-market survey. The source’s routeable product facts are dry-matter trace-element concentrations in four snack variants; Cd, Pb, Co, Cr, and Ni were reported as not detected, but no LOD/LOQ was provided for those nondetects.

Key numbers

Snack variants and formulation scope

The four snack variants were:

All variants also included banded cricket meal, dried hemp inflorescences, dry spirulina biomass, wholegrain buckwheat flour, wholegrain spelt flour, and reverse-osmosis water. The source does not report metal concentrations for the individual ingredients; ingredient names should not be routed to human-food ingredient occurrence pages.

Trace elements in finished snacks

Table 5 reports trace elements in mg/100 g dry matter. The table below preserves the native unit and adds mg/kg dry-matter conversions in parentheses.

Mean trace-element concentrations were higher in baked snacks than extruded snacks for Cu, Fe, Mn, and Zn. Extruded means were Cu 0.51 mg/100 g DM, Fe 3.46 mg/100 g DM, Mn 1.37 mg/100 g DM, Zn 2.30 mg/100 g DM, and Mo 0.13 mg/100 g DM. Baked means were Cu 0.62 mg/100 g DM, Fe 5.50 mg/100 g DM, Mn 1.52 mg/100 g DM, Zn 2.51 mg/100 g DM, and Mo 0.12 mg/100 g DM. The authors state that baked snacks had statistically higher trace-element means except for Mo.

Cd, Pb, Co, Cr, and Ni were not detected in any of the four variants. Because the paper does not print detection limits for these nondetects, this page treats them as qualitative nondetects rather than numeric censored concentration values.

Proximate composition and palatability context

Extruded snacks had lower fat and higher carbohydrate than baked snacks. The mean extruded composition was 92.19 g/100 g dry matter, 20.14 g/100 g DM crude protein, 2.35 g/100 g DM crude fat, 2.81 g/100 g DM crude ash, 1.17 g/100 g DM crude fibre, 73.54 g/100 g DM total carbohydrates, and 346.23 kcal/100 g metabolizable energy. The mean baked composition was 93.30 g/100 g dry matter, 21.00 g/100 g DM crude protein, 5.74 g/100 g DM crude fat, 3.02 g/100 g DM crude ash, 1.77 g/100 g DM crude fibre, 69.41 g/100 g DM total carbohydrates, and 364.64 kcal/100 g metabolizable energy.

In the ten-dog two-bowl palatability tests, S1 versus S2 was 50%/50%; S3 versus S4 was 60%/40%; S1 versus S3 was 20%/80%; and S2 versus S4 was 30%/70%. The authors conclude that dogs preferred baked variants and that higher buckwheat content reduced preference.

Methods (brief)

The authors prepared two extruded and two baked dog-snack prototypes and ground samples for proximate and mineral analysis. Proximate composition followed AOAC methods. Ca, K, Mg, Na, Fe, Mn, Zn, and Cu were measured after wet mineralization in nitric and perchloric acids by atomic absorption spectrometry; phosphorus was measured colorimetrically by the Egner-Riehm method. A certified skimmed-milk-powder reference material (ERM-BD151) was used for analytical verification. The Methods section describes analytical procedures only for the eight macromineral and micronutrient elements above plus phosphorus; it does not describe any analytical procedure, instrument, digestion protocol, or limit of detection for the toxic-metal panel (Cd, Pb, Co, Cr, Ni) or for Mo, whose nondetects and detected values are reported in Table 5. The abstract states that mineral and heavy-metal content was analyzed by colorimetry and mass spectrometry, but the methods section provides no mass-spectrometry detail. Palatability was tested with ten adult dogs in a two-bowl design.

Implications

Certification: This source can route to the pet-treats-snacks product row as finished-snack prototype evidence, but it is weak for benchmark-pool admission because the samples were self-produced prototypes rather than market products and the toxic-metal nondetects lack LOD/LOQ. Do not use the nondetects as zeroes.

Courses: Useful for illustrating the difference between nutrient trace-element values and toxic-metal occurrence values in pet-food research. It also shows why process/formulation prototypes should not be silently pooled with commercial market surveys.

App: Surface as context that insect-meal/spirulina pet-treat prototypes measured Cu, Fe, Mn, Zn, and Mo, while Cd, Pb, Co, Cr, and Ni were nondetect without LODs. Avoid ingredient-level claims for cricket meal, hemp, spirulina, buckwheat, spelt, linseed, or guar gum.

Wiki pages this source may touch

• pet-treats-snacks
• lead
• cadmium
• chromium
• nickel
• copper
• iron
• manganese
• zinc
• molybdenum

Verification notes

The PDF has author attribution and DOI 10.3390/app13052806; no DOI conflict was observed. Names in the frontmatter are ASCII transliterations of the byline. The source reports dry-matter concentrations as mg/100 g DM; this page converts to mg/kg DM by multiplying by 10 and preserves the native units beside each conversion. The paper does not report arsenic or mercury in the trace-element table. It reports Cd, Pb, Co, Cr, and Ni as nd but does not provide LOD/LOQ, so downstream routing should keep those as nondetect-without-limit rather than numeric censored rows.

Matrices vocabulary: pet-treat and dog-snack-prototype are bare-string additions outside the human-food-centric matrices list in docs/gpt-collaboration/system-prompt.md. No standard term covers self-produced experimental pet-treat prototypes, so these two are proposed novel matrices flagged here per the system-prompt guidance. Earlier process- and formulation-specific tokens (extruded-dog-snack, baked-dog-snack, insect-meal-formulation, spirulina-formulation) were removed as over-granular for a four-prototype study without per-formulation metal contrasts.

Audit subagent (2026-06-02) flagged two ⚠️ concerns: matrices over-granularity (applied — collapsed to pet-treat plus dog-snack-prototype and flagged as novel-vocabulary proposals) and incomplete Methods statement on the toxic-metal analytical gap (applied — Methods now states explicitly that the paper describes no analytical procedure or LOD for Cd/Pb/Co/Cr/Ni or Mo). Numerical fidelity, brand firewall, and HMTc firewall checks were clean.

Page history

The five most recent substantive edits to this page. The full version history lives in git; when DOI minting comes online (see schema docs), each entry below will also link to a version-pinned DataCite DOI.