Lead and cadmium levels in raw bovine milk and dietary risk assessment in areas near petroleum extraction industries

Overview

This Iranian environmental-exposure survey quantifies total lead (Pb) and cadmium (Cd) in raw cow milk (n = 118), fodder (n = 14), and farm water (n = 8) collected from 15 dairy farms distributed across 14 regions of Khuzestan province, Southwest Iran, between February 2015 and 2016. The sampled regions sit close to oil and gas fields, refineries, pipelines, processing plants, and tanker terminals, with two of the farms additionally located in the former Iran-Iraq war zone of Shirin Ab-Dez where Persian Gulf War-era oil-well burning and ongoing fine-dust transport from Iraq and Saudi Arabia compound the contamination footprint. Samples were dry-ashed at 450 degC, extracted with 6 M HCl, redissolved in 0.1 M HNO3, and analyzed by graphite-furnace AAS on a Varian SpectrAA 600 fitted with a platform graphite tube and deuterium background corrector, following Iranian National Standard INS 9266 and AOAC method 999.11. Method validation: recoveries 95–110 % for Pb and 80–97 % for Cd; standard deviation < 0.010 and RSD < 10 % for both metals; LoD 3 µg/kg Pb and 0.4 µg/kg Cd; LoQ 9 µg/kg Pb and 1.2 µg/kg Cd; duplicate analysis on all samples. Mean ± SE concentrations in milk were Pb 47.0 ± 3.9 µg/kg (range ND–250.0) and Cd 4.7 ± 1.0 µg/kg (range ND–100.0). The Pb mean is 2.35× the European Union 20 µg/kg permissible limit for raw milk (Commission Regulation (EC) 1881/2006), and 82.2 % (97/118) of milk samples exceeded that limit; 11 of 14 regions had a Pb mean above the limit. Fodder Pb was 54.0 ± 6.9 µg/kg (range 30.0–130.0) and fodder Cd was 3.5 ± 1.3 µg/kg (range 3.0–20.0). No Pb or Cd was detected in any of the 8 water samples (both below the respective method LoDs), and the authors conclude that fodder rather than irrigation water is the principal exposure route into the milk chain for these herds, with atmospheric deposition from oilfield combustion, refining, and pipeline operations as the most plausible upstream driver. Dietary exposure for an average Iranian adult (60 kg body weight, 0.14 kg milk/day) was 6.6 µg Pb/day and 0.66 µg Cd/day in the mean scenario (equivalent weekly intake 46.2 µg Pb/week and 4.6 µg Cd/week) — well below the JECFA 1972/1987 PTWI of 25 µg Pb/kg bw/week (1500 µg/week for a 60 kg adult; this PTWI was subsequently withdrawn by JECFA in 2011) and the EFSA 2011/JECFA 2010 cadmium reference (TWI 2.5 µg/kg bw/week, PTMI 25 µg/kg bw/month) — but in the maximum-exposure scenario, weekly intake rose to 244.9 µg Pb and 97.9 µg Cd per week, and the modelled WImax for infants and toddlers (1–5 years) reached or exceeded the corresponding TWIs for Pb and clearly exceeded the TWI for Cd across ages 2–15 years. The authors recommend continuous monitoring of Pb and Cd in oilfield-adjacent dairy production and flag infants and toddlers as the principal population at risk on a body-weight basis.

Key numbers

Table 3 — Overall milk-cohort summary (n = 118 raw cow milk samples, page 311)

“N.D.” = not detected (Pb < 3 µg/kg LoD; Cd < 0.4 µg/kg LoD). ”†” footnote reads “There is not an established limit” for Cd in raw milk per Commission Regulation (EC) No. 1881/2006; the authors note that an EU value of 5–10 µg/kg has been suggested for Cd in infant formula (liquid or powdered) and protein hydrolysates per Commission Regulation (EU) 488/2014, but no Codex Alimentarius MRL for Cd in raw milk exists.

Table 1 — Region-level Pb summary, milk samples (page 310)

”†” footnote in the source reads: “Mean Pb levels of these areas were above permissible limits (20 µg/kg).” The Mianrood mean is exactly 20.00 µg/kg in the table; the authors mark 11 of the 14 regions with the † symbol (Daste Lale, Tanour Boland, Golkhane, Cham Golak, Al-Mahdi, Shirin Ab-Dez, Hamzeh, Bande-Bal, Pirouzi, Safiabad, Pishineh) and three regions without (Kolouli mean 15.00, Behrouzi mean 16.70, Mianrood mean 20.00). The Hamzeh mean (114.20) is the highest regional mean and is roughly twice the next-highest (Shirin Ab-Dez 63.14). The Hamzeh maximum (250.0 µg/kg) is the highest individual-sample Pb concentration in the survey.

Table 2 — Region-level Cd summary, milk samples (page 310)

The Tanour Boland mean (26.75 µg/kg) is the single regional outlier for Cd and sits roughly an order of magnitude above the next-highest regional means (Hamzeh 3.14, Behrouzi 2.52, Bande-Bal 2.50, Cham Golak 2.17, Golkhane 2.00, Kolouli 2.00). The Tanour Boland Cd maximum (100.0 µg/kg) is the highest individual-sample Cd concentration in the survey and is the single sample that defines the cohort maximum reported in Table 3. The text on page 311 reads: “However, for Cd, only the average level of Tanour Boland region was well above the average value of the rest of the sampled areas (Table 1)” — this text appears to cite Table 1 (Pb) when the cited content is Table 2 (Cd); the wiki page treats this as a paper-internal table-reference typo, noted in Verification notes.

Fodder and water samples (text, page 311)

• Fodder (n = 14): Pb 54.0 ± 6.9 µg/kg, range 30.0–130.0; Cd 3.5 ± 1.3 µg/kg, range 3.0–20.0.
• Water (n = 8): No Pb or Cd was detected in any of the 8 water samples (both below the respective method LoDs of 3 µg/kg Pb and 0.4 µg/kg Cd).

The paper does not report region-level fodder or water breakouts; only the pooled means and ranges are given.

Method validation (text, page 309)

• Multi-element standard solution (Merck): Pb at 0.2, 1, 10, 50, 100 µg/kg; Cd at 0.5, 1.0, 1.5 µg/kg.
• Calibration-curve concentration range: Pb 5–30 µg/kg; Cd 0.5–1.5 µg/kg.
• Recoveries: Pb 95–110 %; Cd 80–97 %.
• Standard deviation (SD) of recoveries < 0.010 (both metals).
• Relative standard deviation (RSD) < 10 % (both metals).
• LoD: Pb 3 µg/kg; Cd 0.4 µg/kg.
• LoQ: Pb 9 µg/kg; Cd 1.2 µg/kg.

Exposure-assessment outputs (Iranian adult anchor 60 kg / 0.14 kg milk/day; text page 311 + abstract page 308)

The abstract (page 308) states: “The intake of an average Iranian adult (25 years, 60 kg b.w., 0.14 kg milk/day) would be 6.6 µg Pb and 0.66 µg Cd/day (WI of 46.2 and 4.6 µg, respectively), well below the risk values proposed by some international organizations, even in the maximum exposure scenario.” The 6.6 µg Pb/day and 0.66 µg Cd/day are the daily intake for the 60-kg adult; the WI of 46.2 µg Pb/week and 4.6 µg Cd/week are the same exposure expressed weekly (= daily × 7, with the arithmetic 6.6 × 7 = 46.2 and 0.66 × 7 = 4.62 ≈ 4.6, matching the abstract). The per-body-weight daily figures (0.11 µg Pb/kg bw/day and 0.011 µg Cd/kg bw/day) divide the daily intake by the 60 kg body weight and are reported in §3.3 page 311. The authors compute exposure per the formula in §2.5 page 311: Exposure to contaminant = (Contaminant occurrence µg/kg × milk consumption kg/day) / average body weight kg. The WImax values (244.9 µg Pb/week, 97.9 µg Cd/week) reflect the cohort-maximum milk concentration (Pb 250.0, Cd 100.0 µg/kg) in the same model. Both WImean and WImax for the 60-kg adult sit well below the reference values cited (1500 µg Pb/week and 348 µg Cd/week, the latter the JECFA 2011 PTMI of 25 µg/kg bw/month converted at 60 kg / 30.4 d × 7 = 345 µg/week, rounded by the authors to 348). However, in the maximum-exposure scenario at infant/toddler/young-adolescent body weights (Fig. 2), the modelled WI for Pb approaches the corresponding TWI in infants and toddlers and the WI for Cd clearly exceeds the corresponding TWI across ages 2–15 years.

Regulatory references invoked by the authors (text pages 310–311)

• EU 1881/2006, amended by 2015/1005: maximum Pb level in raw milk, heat-treated milk, and milk intended for milk-based products = 20 µg/kg w.w. (wet weight).
• EU 488/2014: maximum Cd levels in infant formulas of 5–10 µg/kg (liquid vs powdered formulae and protein hydrolysates).
• Codex Alimentarius Commission 2011: does not report a permissible Cd level in raw cow milk.
• JECFA 1972: PTWI for Pb 25 µg/kg bw/week (established for adults).
• JECFA 1987: PTWI for Pb 25 µg/kg bw/week extended to all age groups (infants and children added).
• JECFA 2011 (CONTAM/74th meeting): PTWI for Pb 25 µg/kg bw/week deemed inadequate to protect against neurodevelopmental effects and withdrew; “no PTWI for Pb could be established that would be considered health-protective.”
• JECFA 1988: PTWI for Cd 7 µg/kg bw/week.
• JECFA 2010 (73rd meeting) / CONTAM 2011: PTMI for Cd 25 µg/kg bw/month (≈ 5.8 µg/kg bw/week), replacing the 1988 PTWI.
• EFSA 2011/2012: TWI for Cd 2.5 µg/kg bw/week (lower than the JECFA PTMI on the basis of more conservative dose-response analysis).

Methods (brief)

Study design. Convenience sample of 15 dairy farms distributed across 14 regions of Khuzestan province, Southwest Iran (Fig. 1; sampling regions: Daste Lale, Tanour Boland, Golkhane, Cham Golak, Al-Mahdi, Shirin Ab-Dez, Hamzeh, Kolouli, Bande-Bal, Pirouzi, Behrouzi, Safiabad, Mianrood, Pishineh). Most farms sit in the Dezful area; the two Shirin Ab-Dez farms (A and B in Fig. 1) sit at the northern fringe of the province in the current or former Iran-Iraq war zone. The sampled farms were both industrial dairy operations and traditional smallholdings. Sampling represented “the most representative livestock in the investigated area.” Sampling period: February 2015 through 2016.

Sample collection. Immediately after milking, 200 mL of raw cow milk was collected into pre-acid-washed sterile screw-topped bottles (10 % HNO3 soak for 24 h; deionized-water rinse for 48 h; oven-dried). The same procedure was used for the 14 fodder samples and the 8 water samples. All samples were stored at –80 degC until analysis. Total study corpus: 118 milk + 14 fodder + 8 water = 140 samples.

Sample preparation. Per Iranian National Standard INS 9266 and AOAC official method 999.11 (Jorhem). Samples were dried and ashed at 450 degC under a gradual temperature increase. 6 M HCl (1+1) was added and the solution was evaporated to dryness. The residue was dissolved in 0.1 M HNO3, and Pb and Cd were measured by graphite-furnace atomic absorption spectrometry. A blank digestion solution was prepared in parallel for each batch.

Instrumentation. Varian SpectrAA 600 graphite-furnace atomic absorption spectrometer equipped with a platform graphite tube and a deuterium background corrector.

Reagents and calibration. All reagents were purchased from Merck KGaA Laboratories (Darmstadt, Germany). Multi-element Pb + Cd standard solutions (Merck) prepared at Pb 0.2, 1, 10, 50, and 100 µg/kg and Cd 0.5, 1.0, and 1.5 µg/kg. Calibration-curve concentration range: Pb 5–30 µg/kg; Cd 0.5–1.5 µg/kg. The paper does not name a certified reference material (CRM); accuracy is anchored on spike-recovery validation alone.

Method validation. Precision expressed as recovery close to 100 % (95–110 % Pb; 80–97 % Cd) with SD of recoveries < 0.010 and RSD < 10 % for both metals. LoD 3 µg/kg Pb and 0.4 µg/kg Cd; LoQ 9 µg/kg Pb and 1.2 µg/kg Cd. Duplicate analysis was performed for all samples.

Speciation. Total Pb and total Cd — no inorganic-vs-organic fractionation; no organolead species; no arsenic (As)/mercury (Hg) panel.

Reporting basis. Concentrations on a wet-weight (w.w.) basis throughout (the EU regulatory limit of 20 µg/kg Pb in raw milk is wet-weight; the authors compare directly without basis conversion, indicating w.w. for the milk samples). Fodder concentrations are reported in the same units (µg/kg) without explicit wet-vs-dry annotation; the paper’s §2.2 sampling description and §2.3 preparation paragraph do not specify whether fodder samples were dried before ashing beyond the standard 450 degC ashing step. Water concentrations are µg/L equivalent (no Pb or Cd detected above the method LoDs).

Statistics. SPSS (IBM) version 20 plus R statistical environment v3.4.2. Results reported as mean ± SE. ANOVA with Tukey post-hoc multiple comparison was used to compare Pb and Cd levels across sampling locations (milk, water, and fodder), with the EU permissible limit (20 µg/kg w.w.) as the reference threshold; the per-region table of Tukey contrasts is not reproduced in the paper, only the overall ANOVA framing.

Exposure-model parameters. Adult anchor: 25-year-old Iranian, 60 kg body weight, 0.14 kg milk/day per-capita consumption. Mean-scenario contaminant concentration = cohort-mean milk Pb 47.0 and Cd 4.7 µg/kg; maximum-scenario = cohort-maximum milk Pb 250.0 and Cd 100.0 µg/kg. Age-segmented WI curves (Fig. 2) were drawn for ages 2–90 years using “average body weight and milk/dairy consumption at different ages (from toddlers to seniors)” anchored on six Iranian dietary-pattern studies (Ghassemi 2002; Haghdoost 2008; Abdollahi 2014; Esfarjani 2015; Singh 2015; the Iran-population age-specific milk-consumption series implicit in those references).

Tolerable weekly intakes. As reported in §2.4 and §3.3: Pb 1500 µg/week and Cd 348 µg/week as the 60-kg-adult-equivalent JECFA 1972 PTWI (Pb) and JECFA 2011 PTMI (Cd) the paper uses as comparators. The paper explicitly notes that the Pb PTWI of 25 µg/kg bw/week was withdrawn in 2011 because no PTWI for lead can be considered health-protective (JECFA 73rd meeting, 2011), and presents the comparison as the authors’ chosen reference, not as the current JECFA position.

Implications

This source contributes Iranian Khuzestan-province oilfield-adjacent raw bovine milk + fodder + farm-water occurrence data for total Pb and Cd from 15 dairy farms across 14 regions sampled between February 2015 and 2016. Its principal contributions to the wiki evidence pool:

• Raw bovine milk Pb occurrence in an oilfield-adjacent industrial setting (Khuzestan, Iran): cohort mean 47.0 ± 3.9 µg/kg w.w., range ND–250.0 µg/kg w.w., n = 118. 82.2 % of samples (97/118) exceeded the EU 1881/2006 raw-milk Pb maximum of 20 µg/kg w.w., and 11 of 14 regions had a mean above the limit (text page 311). The Hamzeh region mean (114.20 µg/kg w.w.) is the highest regional mean and is approximately twice the next-highest (Shirin Ab-Dez 63.14 µg/kg w.w.). The cohort mean is roughly an order of magnitude above the 2.4 ± 1.4 µg/kg Iranian Shoosh-Khuzestan Pb mean reported by Tajkarimi et al. (2008) in §4 (the only prior Southwest-Iran milk-Pb survey the authors cite) and substantially above the 0.029 ± 0.022 mg/kg (29 ± 22 µg/kg) raw-milk Pb mean reported in chirinos2023-milk-lead-cadmium-arsenic-peru — supporting the hypothesis that oil-and-gas extraction-and-processing infrastructure adds a measurable Pb burden to dairy systems above legacy-mining or background Iranian rural baselines.
• Raw bovine milk Cd occurrence in the same context: cohort mean 4.7 ± 1.0 µg/kg w.w., range ND–100.0 µg/kg w.w., n = 118. The Tanour Boland region (mean 26.75 ± 6.78 µg/kg w.w., max 100.0) is an order-of-magnitude single-region outlier and drives the cohort maximum; the other 13 regions sit at means between 0.33 (Pirouzi) and 3.14 (Hamzeh) µg/kg w.w. The Codex Alimentarius does not set a Cd MRL for raw cow milk; the EU 488/2014 infant-formula Cd limits (5–10 µg/kg) are the closest regulatory anchor, and the cohort mean for milk-Cd sits just below the lower bound of that range, while the Tanour Boland regional mean and the maximum-scenario cohort value both clearly exceed it. The cohort mean is several-fold above the 0.3 ± 0.3 µg/kg Cd mean reported by Najarnezhad and Akbarabadi (2013) for Khorasan-Razavi (North-East Iran), supporting the same atmospheric-deposition / oilfield-emission interpretation as the Pb signal.
• Fodder pathway as principal exposure route: fodder mean Pb 54.0 ± 6.9 µg/kg (range 30.0–130.0) and Cd 3.5 ± 1.3 µg/kg (range 3.0–20.0). Water at the same farms tested ND for both Pb and Cd (n = 8, both below method LoDs of 3 and 0.4 µg/kg respectively). The authors interpret this as evidence that irrigation water is not the proximal driver of milk contamination in this corpus, leaving fodder accumulation (driven by atmospheric deposition from oilfield combustion, oil-spill-related soil contamination, and trans-boundary dust transport from Iraq and Saudi Arabia) as the principal pathway. This pathway framing is consistent with the Bakary et al. 2015 framework the authors cite for “simultaneous evaluation of metals in water and fodder” as an environmental-contamination indicator.
• Adult Iranian exposure burden via raw cow milk: mean-scenario daily intake 6.6 µg Pb/day and 0.66 µg Cd/day per 60 kg adult (= weekly intake 46.2 µg Pb/week and 4.6 µg Cd/week; = per-body-weight 0.11 µg Pb/kg bw/day and 0.011 µg Cd/kg bw/day). Both weekly figures sit well below the JECFA 1972/2011-comparator reference points (1500 µg Pb/week; 348 µg Cd/week as the 60-kg adult equivalent PTMI). The cohort-maximum-scenario WI is 244.9 µg Pb/week and 97.9 µg Cd/week — still below the adult reference points but a factor of 5.3 (Pb) and 21 (Cd) above the mean-scenario WI. Critically, when the same model is applied to lower-body-weight infant, toddler, and young-adolescent ages (1–15 years), the modelled WImax for Pb reaches or exceeds the corresponding TWI in infants and toddlers (per Fig. 2 top panel), and the WImax for Cd clearly exceeds the corresponding EFSA 2.5 µg/kg bw/week TWI across ages 2–15 years (per Fig. 2 bottom panel). These age-stratified exposure outputs anchor the body-weight-normalized “infants and toddlers consume 2–3 times more food than adults relative to their body weight” framing in the authors’ conclusion.
• Methodological reference point: GF-AAS for total Pb and total Cd on Varian SpectrAA 600 with platform graphite tube and deuterium background corrector; dry-ashing digestion at 450 degC followed by 6 M HCl extraction and 0.1 M HNO3 dissolution per Iranian National Standard INS 9266 + AOAC method 999.11. Spike-recovery validation across 95–110 % (Pb) and 80–97 % (Cd); no CRM-anchored accuracy validation. This is the principal reason for the B-tier (rather than A-tier) evidence grading: no certified reference material was used to anchor analytical accuracy, no inter-laboratory cross-validation is reported, and the LoQ for Cd (1.2 µg/kg) sits within the same order of magnitude as the cohort-mean Cd value (4.7 µg/kg), which means a meaningful fraction of the lower-tail individual-sample values sit close to or below the LoQ.
• Geographic scope: 15 farms across 14 regions of Khuzestan province (Daste Lale, Tanour Boland, Golkhane, Cham Golak, Al-Mahdi, Shirin Ab-Dez, Hamzeh, Kolouli, Bande-Bal, Pirouzi, Behrouzi, Safiabad, Mianrood, Pishineh). Most farms cluster in the Dezful area; the two Shirin Ab-Dez farms sit at the Iran-Iraq border. Per-region n ranges from 3 (Al-Mahdi, Kolouli, Bande-Bal, Pirouzi, Safiabad) to 21 (Shirin Ab-Dez) milk samples. The per-region Pb and Cd means and SEs are usable for downstream geographic-variance characterization on the milk-and-dairy ingredient page.

The contribution is bounded by several limitations: per-region milk-sample counts are highly uneven (3 to 21 samples per region), with five regions at n = 3 producing wide SEs that limit single-region characterisation; the paper does not report region-level fodder or water breakouts despite collecting paired water/fodder samples per farm; sampling year is approximately one full annual cycle (Feb 2015 to 2016) but no seasonal stratification is reported, so seasonal variance in oilfield-driven Pb/Cd deposition cannot be derived from this paper alone; analytical accuracy is anchored on spike recovery alone without a CRM (rice flour SRM 1568b is the closest published CRM for biological-matrix Pb/Cd validation, used in the Naseri 2015 rice-grain paper in the same corpus); the As/Hg panel is absent (the authors note in §4 that “no studies have been carried out on the concentration of heavy metals in milk from livestock living close to or exposed to contamination from oil extraction activities” prior to this work, but the survey itself only quantifies Pb and Cd); the cohort-maximum Cd is a single-sample maximum driven by one Tanour Boland sample and may not generalise to other oilfield-adjacent dairy systems. Pools with other Iranian milk surveys (Najarnezhad and Akbarabadi 2013 Khorasan-Razavi, Najarnezhad et al. 2015 West Azerbaijan, Mostafidi et al. 2016 Iran) and with the broader oilfield-adjacent dairy literature (the discussion’s §4 comparators across Turkey, Croatia, India, Thailand, Italy, South Africa, Spain, Greece, Iran, Romania, Pakistan, France, Sudan, Kosovo, Egypt) to characterise the upper-tail occurrence of Pb and Cd in raw bovine milk adjacent to oil-and-gas infrastructure; does not anchor a standalone clean-baseline characterisation given the explicit selection on oilfield-adjacent farms.

Related evidence

• Lead
• Cadmium
• Milk and Dairy
• Milk and Dairy (HMTc Category 24 master)
• Commission Regulation (EC) No 1881/2006 setting maximum levels for certain contaminants in foodstuffs
• JECFA — Lead Provisional Tolerable Weekly Intake (Withdrawn 2010)
• JECFA — Provisional Tolerable Monthly Intake for Cadmium, 25 µg/kg b.w./month
• EFSA — Tolerable Weekly Intake for Cadmium, 2.5 µg/kg b.w./week
• EFSA — Lead in Food (2010), No-Threshold Position

Update history

No substantive edit history is available in this build. The full commit record is available in git.