Haydous et al. 2025 — Toxic metals in pet food marketed in Lebanon and UAE
This study analyzed 196 commercial pet food stock keeping units (SKUs) collected from Lebanon (n = 75) and the UAE (n = 121) for nine metals by ICP-MS. The matrix split was 81 dry and 115 wet products. Lead and arsenic exceeded regulatory safety thresholds in a subset of samples, raising concerns for animal welfare and indirect human exposure via pet-owner handling and cross-contamination. This is the first systematic survey of toxic metal contamination in pet food marketed in Lebanon and the UAE, and fills a gap in Middle Eastern food-safety surveillance data.
Key numbers
Dry food (n = 81) vs. wet food (n = 115), means (mg/kg):
- Pb: dry 0.981 ± SD, wet 0.421 ± SD; p = 0.039 (dry > wet)
- Cd: dry 0.102, wet 0.296; p = 0.045 (wet > dry — differential solubility/retention)
- As: dry and wet means aggregated across all samples: mean 0.193 ± 0.324, median 0.094 mg/kg (total arsenic)
- tHg: mean 1.213 ± 0.245, median 1.293 mg/kg across all samples (notably elevated; narrow IQR 1.144–1.370 suggests consistent matrix-wide presence)
- Cr (total): mean 2.745 ± 0.923, median 2.779 mg/kg
Overall sample statistics (all 196 SKUs):
- Pb: mean 0.653 ± 0.684 mg/kg, median 0.439 mg/kg; IQR 0.244–0.892 mg/kg; max observed ~5.0 mg/kg (sample 5 dry, Lebanon)
- Cd: mean 0.216 ± 0.219 mg/kg, median 0.149 mg/kg; IQR 0.112–0.155 mg/kg
- As (total): mean 0.193 ± 0.324 mg/kg, median 0.094 mg/kg; IQR 0.025–0.259 mg/kg
Cat food (n = 147) vs. dog food (n = 49): cat food contained higher Cd (0.251 vs. 0.112 mg/kg); dog food had higher Zn and Cu.
LODs: As 0.00005, Cd 0.00003, Pb 0.00004 mg/kg (all well below detection for these samples). Instrument: iCAP Q/RQ ICP-MS (ThermoFisher); microwave digestion with HNO3/H2O2; calibration R ≥ 0.9999.
Country-of-origin pattern: Lebanon samples predominantly dry, sourced from Italy, Czech Republic, Brazil; UAE samples predominantly wet, with 37.2% from Thailand; 65.3% of all SKUs from developed countries.
Methods (brief)
Market screening Fall 2021; all pet shops visited in both countries; 196 distinct SKUs collected in original packaging. Microwave digestion (0.5 g sample, 8 mL 69% HNO3 + 2 mL 30% H2O2, Anton Paar Multiwave ECO); 2 mL HCl added post-digestion; diluted 5× in 3% HNO3. ICP-MS quantified nine metals simultaneously: Cr (52), Co (59), Cu (63), Zn (64), As (75), Mo (98), Cd (114), Hg (200), Pb (208). All measurements in full quantitative mode; isotopic ratios checked for polyatomic interference. Statistical analysis: SPSS V27; Mann-Whitney U or independent t-test for group comparisons; p < 0.05 significance threshold.
Note on arsenic speciation: the study reports total arsenic, not inorganic arsenic. For pet food (fish- and meat-containing matrices), total arsenic may be substantially organic (arsenobetaine, DMA). The authors do not speciate. This limits comparison to EU maximum levels set for total arsenic in pet food feed materials, but does not resolve the inorganic arsenic fraction.
Note on mercury: the Hg values (mean 1.213 mg/kg) appear consistently elevated across samples. The authors attribute this to methylmercury from fish-based ingredients, which is common in wet cat food. The study does not speciate Hg into methylmercury vs. inorganic mercury. All Hg values reported here are total mercury.
Implications
Certification: HMT&C certifies human food and dietary supplements, not pet food. This paper does not directly contribute to product-category thresholds. However, the supply chains overlap substantially: pet food ingredients (fish meal, meat by-products, grain fractions) draw from the same ingredient pools as human dietary supplements and processed foods. Metal levels in pet food provide a useful indirect indicator of contamination burdens in shared ingredient streams.
Courses: Illustrates the global reach of heavy metal contamination in processed food products, the role of moisture content in apparent concentration differences (dry vs. wet), and how the same ingredient origins (Thailand fish-based, European grain-based) drive metal profiles across markets.
App: No human food concentration data. Not applicable to contamination-profile updates for human food ingredients.