Heavy Metal Contamination in Yogurt from Lebanon: Evaluating Lead (Pb) and Cadmium (Cd) Concentrations Across Multiple Regions

This study measured Pb and Cd in 165 traditionally produced cow’s-milk yogurt samples collected across 11 regions of Lebanon (Nabatiyeh, South Area, Chouf, Metn, Kesserwan, Jbeil, Batroun, North Area, Akkar, Bekaa, Baalbeck-Hermel; n=15 per region) using graphite furnace atomic absorption spectrometry (GFAAS) on a Shimadzu AA-6800. Microwave-acid digestion with HNO3/H2O2 was used; accuracy was validated against BCR-679 white-cabbage certified reference material (94–111% recovery for Cd) with LODs of 0.004 mg/kg (Pb) and 0.0007 mg/kg (Cd). All concentrations are wet-weight on the as-consumed yogurt. The dataset is the first known regional survey of Pb and Cd in Lebanese yogurt and provides regional means, exceedance fractions against EU 2023/915 limits (0.020 mg/kg Pb; 0.005 mg/kg Cd), and estimated dietary intake (EDI) for the adult population.

Key numbers

All values mg/kg (wet weight, as-consumed yogurt). n=165 samples (15 per region). Method: GFAAS. Regulatory benchmark: EU 2023/915 — 0.020 mg/kg Pb and 0.005 mg/kg Cd in raw milk and dairy products.

Per-region mean concentrations, Pb (Table 2, raw data):

Per-region mean concentrations, Cd (Table 2, raw data):

Samples above EU 2023/915 regulatory limits (Table 4):

The abstract and Conclusions report exceedance rates of 10.9% (Pb) and 14.5% (Cd) and list per-region “highest detected” concentrations of Pb 0.118 mg/kg (Baalbeck-Hermel), 0.125 mg/kg (North Area), 0.115 mg/kg (South Lebanon) and Cd 0.094 mg/kg (North Area), 0.076 mg/kg (Baalbeck-Hermel), 0.042 mg/kg (Akkar). These abstract values disagree with Section 3.2 and Table 4 in the same paper, which give 15.15% Pb and 12.12% Cd exceedance and which support the regional means tabulated above. See ## Verification notes below.

Estimated daily intake (Table 7):

Cd mean adult daily intake from yogurt: 0.46 µg/day = 0.006 µg/kg b.w./day (1.89% of EFSA-derived TDI of 24.35 µg/day for a 73.8 kg adult, corresponding to the 2.5 µg/kg b.w./week EFSA TWI from 2009). Per-region Cd EDI range 0.003–0.012 µg/kg b.w./day (Metn <LOQ; Bekaa highest at 0.012).

Pb mean adult daily intake from yogurt: 0.94 µg/day = 0.013 µg/kg b.w./day. The paper compares against the EFSA CONTAM Panel BMDL01 of 0.50 µg/kg b.w./day (developmental neurotoxicity), 1.50 µg/kg b.w./day (cardiovascular), and 0.63 µg/kg b.w./day (kidney). Per-region Pb EDI range 0.004–0.020 µg/kg b.w./day (Metn lowest at 0.004; Baalbeck-Hermel and South Lebanon tied highest at 0.020). The authors note this 0.94 µg/day yogurt-only Pb intake is approximately ten-fold higher than the 0.08 µg/day mean Pb intake from all dairy reported by Nasreddine et al. 2010 for the Lebanese adult population.

Statistical caveat: one-way ANOVA reported p=0.876 for Pb and p=0.853 for Cd between regional means, and post hoc Games–Howell 95% CIs for every regional pair contained zero. The authors interpret this as no statistically significant regional difference between means despite the clear differences in exceedance fractions and the order-of-magnitude spread in per-region maximums. This is consistent with the very wide within-region 95% CIs (most spanning negative values, indicating heavy right-skew with a few high-end samples).

Modelled adult parameters used in EDI: body weight 73.8 kg (Hoteit et al.); mean yogurt consumption 68.3 g/day (Nasreddine et al.). EDI for children and adolescents was not estimated due to absence of pediatric consumption data for Lebanon.

Methods (brief)

Yogurt samples (15 per region × 11 regions = 165) were collected from traditional dairy farms producing yogurt on-site from regional cow’s milk without preservatives, stabilizers, or flavoring agents. Samples were transported refrigerated and analyzed without delay.

Sample preparation followed a method previously validated by Kassouf et al. (2013) on chickpea matrices: three replicates of 2 g fresh homogenized yogurt (no freeze-drying) were weighed into a modified PTFE-TFM microwave bomb vessel; 7 mL HNO3 (≥65%) and 1 mL H2O2 (30%) were added; digestion was carried out in an Anton Paar Multiwave 3000 microwave digester. After cooling, the clear digest was transferred to 50 mL polypropylene tubes, diluted to volume with ultrapure water, and filtered through 4 µm PTFE syringe filters (Whatman, Maidstone, UK).

Metal quantification used graphite furnace atomic absorption spectrometry (GFAAS) on a Shimadzu AA-6800 with ASC-6100 autosampler and WizAArd data acquisition software. Wavelengths: 283.3 nm (Pb) and 228.8 nm (Cd); lamp current 10 mA. External calibration curves used at least five concentration levels prepared fresh from Sigma-Aldrich and Fluka stocks (Pb 1000 ± 4 mg/L stock) and Merck Certipur (Cd 1000 mg/L stock).

Method validation used BCR-679 white-cabbage certified reference material (Sigma Aldrich, Geel, Belgium) for Cd accuracy (94–111% recovery); Pb accuracy was previously validated by Kassouf et al. (2013) on chickpea samples. LOD/LOQ were estimated from blank signals as average + 3SD (LOD) and average + 10SD (LOQ): Pb LOD 0.004 mg/kg, Pb LOQ 0.012 mg/kg; Cd LOD 0.0007 mg/kg, Cd LOQ 0.0021 mg/kg. Precision (triplicate spiked sample analyses) gave RSD<15%.

Statistical analysis used IBM SPSS Statistics V26: one-sample two-sided t-tests against the EU 2023/915 regulatory limits; one-way ANOVA with Games–Howell post hoc test for between-region comparisons; bootstrap with 1000 resampled samples for an alternative CI estimate. Significance threshold p≤0.05.

EDI was calculated as EDI = C × MS / b.w. where C is the mean concentration in mg/kg wet weight, MS is daily yogurt consumption (68.3 g/day, Nasreddine et al.), and b.w. is adult body weight (73.8 kg, Hoteit et al.). EDI was computed only for analytes above LOQ.

Implications

Certification: This is the first known regional survey of Pb and Cd in Lebanese yogurt. Mean Pb across all regions (0.0137 mg/kg) sits below the EU 0.020 mg/kg limit on average, but the 15.15% sample-level exceedance rate (Table 4) indicates that a Lebanon-sourced dairy supply cannot be presumed compliant without sample-level testing. The right-skewed distribution — wide CIs spanning zero in most regions — means the per-region mean understates the upper-tail risk that drives certification gating. Mean Cd across all regions (0.00675 mg/kg) exceeds the EU 0.005 mg/kg limit on the mean alone, and 12.12% of samples exceeded the limit at the sample level. Both numbers argue for sample-level rather than producer-level acceptance criteria when sourcing traditional Lebanese dairy.

Geographic pattern by mean concentration: Pb highest in South Lebanon, Baalbeck-Hermel, and Bekaa; Cd highest in Bekaa, North Area, and Baalbeck-Hermel. The Bekaa Valley appears in the top three for both analytes; this is Lebanon’s primary agricultural region with documented soil and water Pb/Cd contamination from fertilizer overuse and industrial waste (Darwish et al., Moussa et al. cited).

Courses: Demonstrates that traditional, additive-free, single-source dairy production from a regionally contaminated milk supply does not protect the finished product from carry-over of Pb and Cd, since the heavy metal load enters via the animal feed and water chain rather than via processing additives. Useful case study for “clean processing does not equal clean product.”

App: Contributes Lebanon-anchored cow’s milk yogurt occurrence data for the yogurt contamination profile. Per-region means and exceedance fractions are usable as a regional-supply-chain reference for the Eastern Mediterranean.

Wiki pages this source may touch

• yogurt
• milk-and-dairy
• lead
• cadmium
• eu-2023-915-cadmium
• eu2023-contaminants-maximum-levels
• efsa-cadmium-twi
• efsa-lead-contam-2010

Verification notes

2026-05-19 (Claude Code, merge-enhance from prior 2026-05-14 page). The prior revision had ## Key numbers thinned to the abstract’s three “highest detected” values per metal and explicitly noted “Exact mean and median values by region are in the published tables but not reproduced in extracted text.” This revision reads Tables 2, 4, 5, 6, and 7 in full and reproduces the per-region means, exceedance counts, ANOVA results, and EDI values. Methods section is expanded from one sentence to a full reproducibility block (GFAAS Shimadzu AA-6800, microwave digestion Anton Paar Multiwave 3000, BCR-679 CRM, LODs/LOQs, statistical software). Reference materials and instrument vendor/model are named per the Part 12 brand-firewall scientific-method exception. Frontmatter updated: advanced to today; cite_key, raw_handle, raw_path, license preserved. Legacy heading ## Wiki pages updated on ingest renamed to ## Wiki pages this source may touch per current Part 6 template. ingredients/milk updated to ingredients/milk-and-dairy to match the live taxonomy slug. Added regulations/efsa-cadmium-twi and regulations/efsa-lead-contam-2010 to the wiki-pages-touched list since Section 3.4 anchors the EDI risk characterization against EFSA’s 2009 TWI for Cd and EFSA CONTAM Panel 2010 BMDL01 values for Pb.

Paper-internal contradiction flagged for downstream synthesis. The abstract and Conclusions report Pb exceedance of 10.9% and Cd exceedance of 14.5%, but Section 3.2 and Table 4 report Pb exceedance of 15.15% (25/165) and Cd exceedance of 12.12% (20/165). These two summaries are mutually inconsistent and the abstract figures do not correspond to an integer count of the 165 samples (10.9% × 165 ≈ 18.0; 14.5% × 165 ≈ 23.9, neither matching Table 4’s 25 and 20). Table 4 is reported as authoritative on this page because it is paired with raw sample counts and per-region breakdowns. The abstract’s “highest detected” per-region values (Pb 0.118, 0.125, 0.115 mg/kg; Cd 0.094, 0.076, 0.042 mg/kg) are not reproduced in any table in the body of the paper and appear to be per-region maximum sample values; they are reported here verbatim with this provenance caveat.

Brand firewall. No brand names appear in the source paper outside Methods (instrument and reagent vendors), which are retained per the locked 2026-05-17 scientific-method exception. No brand-by-brand or producer-by-producer ranking exists in the source.

Wiki/HMTc firewall. No HMTc threshold proposals, consumer risk advisories, or cross-source synthesis claims are made on this page. The ## Implications section limits itself to what this single paper contributes to threshold work (sample-level vs producer-level acceptance, the right-skew structure of the distribution) and does not propose certification values.

Products field. products: [] is retained because no yogurt product slug exists in the current taxonomy. The matrices field captures the product form (yogurt, dairy); creation of a yogurt product page is a Step 0 Lock decision and is not made here.

Audit subagent (2026-05-19) flagged that [[regulations/eu-2023-915-lead-infant-young-child-foods]] was the wrong EU 2023/915 sub-slug for this paper because the source measures Pb in raw-milk-derived dairy (EU 2023/915 dairy table, 0.020 mg/kg limit), not the infant/young-child foods Pb sub-table. Verified against the source (paper benchmarks against “Commission Regulation (EU) 2023/915 … 0.020 mg/kg and 0.005 mg/kg of lead and Cd” for “raw milk and dairy products”, p.3) and against the regulation page (wiki/regulations/eu2023-contaminants-maximum-levels.md is the umbrella slug covering all metals/all matrices in EU 2023/915, including the dairy Pb limit). Finding correct — replaced [[regulations/eu-2023-915-lead-infant-young-child-foods]] with [[regulations/eu2023-contaminants-maximum-levels]].

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.