Tavakkoli & Khanjani 2016 — Systematic review of human cadmium biomonitoring in Iran

This narrative systematic review aggregates 18 Iranian biomonitoring studies (of 25 retrieved) reporting cadmium concentrations in human matrices — blood, urine, breast milk, hair, saliva, amniotic fluid, umbilical cord blood, and deciduous teeth — across general, occupational, and clinical populations in Iran, with sampling years spanning roughly 2003 through 2015. The authors report no formal meta-analysis because of population heterogeneity; the review’s value is corpus-level: it surfaces the Iranian biomonitoring evidence base for Cd, in one place, with consistent attribution. Three headline patterns emerge from the assembled corpus: (1) Cd concentrations are systematically higher in industrial workers than in general-population referents (Yazd electroplating glazers post-shift urine 10.85 ± 12.88 µg/g creatinine vs office-worker referents 3.07 ± 7.24 µg/g; Borujen natural-gas-pipeline back welders 1.72 ± 0.65 µg/g creatinine vs unexposed controls 0.04 ± 0.07 µg/g); (2) Cd concentrations are higher in smokers than non-smokers in general-population samples (Farzin et al. blood Cd in male smokers 2.18 ± 0.89 µg/L vs male non-smokers 1.74 ± 0.66 µg/L; Ebrahim et al. amniotic-fluid Cd in pregnant smokers vs non-smokers, with a smoking-thiocyanate correlation r=0.882); and (3) Cd concentrations are elevated in residents of industrial regions relative to non-industrial regional referents (Tadayon et al. hair Cd Tehran 1.76 ± 1.09 µg/g vs Tekab 0.35 ± 0.56 µg/g; Norouzi et al. breast-milk Cd Lenjan/Mobarakeh-steel-region 2.31 µg/L vs Konjedjan rural-referent 0.772 µg/L). The review does not aggregate food-matrix Cd data and does not propose Iranian exposure or dietary-intake thresholds; the JECFA/FAO 25 µg/kg body weight per month provisional tolerable monthly intake is cited but not compared to per-study figures.

Key numbers

Search and inclusion (review prose, p. 458):

Iranian biomonitoring corpus aggregated by the review (Table 1, pp. 459-460; all values transcribed as published by the review, which transcribes them from the constituent studies):

Reference values cited by the review (Introduction, p. 457):

Comparative international Cd-in-hair values cited from constituent study #6 (Savabieasfahani 2010, p. 461):

Occupational signal magnitudes drawn from Discussion (pp. 461-462):

Two pregnancy-outcome correlations the review highlights:

• Ebrahim & Ashtarinejad 2015 (ref 17): amniotic-fluid Cd correlation with maternal smoking biomarker (plasma thiocyanate) r = 0.882; pre-eclampsia prevalence in smokers 26% vs non-smokers 10.5% (p < 0.05).
• Nejadchehrazin & Moghadam Banaem (ref 25): maternal-blood Cd correlation with umbilical-cord-blood Cd r = 0.734 (the review highlights this as evidence of transplacental transfer).

Methods (brief)

This is a systematic review of the published literature rather than an original measurement study. The methods section describes only the search-and-inclusion procedure; analytical methodology is reported in each constituent study and not re-described in detail by the review.

Database search. The authors searched both Iranian databases (Scientific Information Database (SID), Magiran, IranMedex) and international databases (PubMed, Web of Knowledge, Google Scholar, Science Direct). Search terms included “cadmium poisoning,” “cadmium toxicity,” “cadmium measuring,” “cadmium exposure,” “occupational exposure to cadmium,” “environmental exposure to cadmium,” and “heavy metals” in English and Persian translations. The search included all studies conducted in Iran and indexed in the named databases through March 15, 2016.

Inclusion criteria. Original studies investigating occupational or environmental exposure to Cd in Iran that measured Cd levels in human tissues (blood, hair, saliva, milk, urine) were eligible. Studies done outside Iran and studies on non-human samples were excluded. The review does not state inclusion/exclusion at the level of study-design type (cohort, cross-sectional, case-control); both general-population descriptive surveys and case-control clinical studies are included.

Data extraction. The authors state that all articles were reviewed by two people separately and that first author, year of data collection, population, sample size, mean and standard deviation of Cd levels, and location were extracted from each article.

Statistical analysis. The authors decided not to merge data in a meta-analysis because of population heterogeneity (industrial workers, pregnant women, children, clinical patient groups, general-population volunteers; biological matrices ranging from amniotic fluid to deciduous teeth). The review reports per-study summary statistics as published; no pooled estimates, no random-effects models, no heterogeneity (I²) calculations, no risk-of-bias scoring, and no formal PRISMA-style flow diagram are presented.

Constituent-study analytical methods (not re-described by the review). The 18 constituent studies span multiple analytical platforms: atomic absorption spectrometry (graphite furnace and flame variants), electrothermal AAS, and ICP-OES. Sample preparation, LOD/LOQ, and quality-control details are reported in the constituent papers and not aggregated by this review.

Implications

This source is a narrative systematic review aggregating Iranian human biomonitoring evidence on cadmium. It contributes a corpus-level characterisation of Iranian Cd exposure across general, occupational, and clinical populations; it does not contribute original measurement data, dietary-intake estimates, or food-matrix concentration values.

• Corpus pointer for Iranian Cd biomonitoring literature 2003-2015. The 18 constituent studies span Tehran, Isfahan, Yazd, Hamadan, Birjand (South Khorasan), Varamin, Borujen, Lenjan, Konjedjan, Zarrin Shahr, and Tekab (West Azarbaijan). They cover blood, urine, breast milk, hair, saliva, amniotic fluid, umbilical cord blood, and deciduous teeth. Downstream wiki synthesis on Iranian Cd biomonitoring should treat this paper as a pointer to the constituent studies rather than as a standalone quantitative source — the review’s per-study figures should be cross-checked against the original papers when those papers are ingested individually.
• Occupational-vs-environmental signal in Iranian biomonitoring. Urinary Cd in industrial cohorts (Yazd electroplating glazers, Borujen natural-gas-pipeline welders) runs roughly 3-40× higher than in unexposed referents drawn from the same studies. This is consistent with the broader occupational-Cd biomonitoring literature, but provides Iran-specific occupancy quantitatively for any HMTc work that needs Iranian-population referent ranges. The Yazd glazer post-shift figure (10.85 µg/g creatinine) sits in the same magnitude band as the ICdA-cited 2-50 ng/m³ “acceptable” occupational-air range when converted assuming standard pulmonary-uptake and creatinine-excretion kinetics.
• Smoking and Cd in Iranian populations. Three constituent studies (Farzin et al. blood Cd by smoking status; Ebrahim et al. amniotic-fluid Cd by smoking biomarker; the review’s secondary references to Hallen et al., Grasmick et al., Elinder et al., Ellis et al., Zenzes et al., and Mannino et al.) corroborate the smoker-vs-non-smoker Cd gap. Magnitudes are modest in blood (~1.25-1.3× for male smokers in Farzin’s Tehran cohort) but the smoking-thiocyanate correlation r = 0.882 in amniotic-fluid Cd (Ebrahim) is a strong gestational-exposure signal worth tagging for the cadmium page’s smoking/pregnancy section.
• Regional industrial-area signal (Lenjan, Mobarakeh-steel region). Norouzi et al.’s Lenjan-vs-Konjedjan breast-milk Cd comparison (2.31 vs 0.772 µg/L, ≈3× elevated in the Mobarakeh-steel-region cohort) provides Iran-specific environmental-exposure quantification near a known industrial source. This is one of the few breast-milk Cd values in the corpus that includes a within-study geographic referent.
• Headroom-only-on-average breast-milk Cd values. Three breast-milk Cd values in the corpus (Nazarpour 5 ± 6.9 µg/mL, Goudarzi 1.92 ± 1.04 µg/mL, Norouzi-Lenjan 2.31 µg/L, Rahimi 2.44 µg/L) — see anomaly catalogue below for unit-label concerns on Nazarpour and Goudarzi — sit at or above the WHO breast-milk-Cd recommendation tier of approximately 1 µg/L. The review notes only that the Varamin figure “was even higher than the international standard” without naming the standard. Downstream synthesis on breastmilk (cadmium sub-block) should treat these Iranian per-study values as candidate-pool contributors with the unit-attribution caveats below.
• Transplacental-transfer signal. Across the three studies in the corpus that paired maternal-blood Cd with umbilical-cord-blood Cd (Dahaghin maternal 2.3 vs umbilical 0.34 µg/L; Nejadchehrazi maternal 0.31 vs umbilical 0.29 µg/dL with r = 0.734; Vigeh maternal 0.50-0.54 µg/L), the cord/maternal ratio varies from ~0.15 (Dahaghin) to ~0.94 (Nejadchehrazi). The review treats Nejadchehrazi’s high correlation as evidence of placental Cd transfer; Dahaghin’s near-zero cord Cd in the presence of elevated maternal Cd is treated as evidence of placental barrier function. Both readings are present in the review without resolution.
• What this paper does not provide. This review reports human-biomarker concentrations; it does not report Iranian dietary-intake estimates, food-matrix Cd concentrations, or estimated weekly intake (EWI) figures for Iranian foods. It does not propose HMTc threshold values, does not compare biomarker concentrations to international biomarker reference values (e.g., German HBM-I/HBM-II, US CDC NHANES geometric means), and does not perform meta-analytic pooling. For Iranian food-matrix Cd values, see the parallel rice-specific review sharafatichaleshtori2016-iran-rice-heavy-metals-review.
• Review-level limitations for downstream pooling. No formal risk-of-bias assessment. No PRISMA flow diagram. No I² heterogeneity statistics. Population mixing across vastly different exposure regimes (industrial workers, general-population volunteers, clinical patient groups, pregnant women) precludes pooled estimation; the authors explicitly state this is why no meta-analysis was attempted. The narrative-review category limits this paper to corpus-pointer status for the wiki’s purposes; per-study values should be re-verified against the constituent papers when those are independently ingested.

Wiki pages this source may touch

• cadmium

Verification notes

• Source type and evidence tier. Peer-reviewed narrative systematic review in Reviews on Environmental Health (De Gruyter), aggregating 18 prior peer-reviewed Iranian biomonitoring studies. Graded B-tier: peer-reviewed methodology with an explicit search-and-inclusion procedure (database list, search terms, date cutoff, inclusion criteria, two-person data extraction), but conveys secondary (review-summarised) values rather than original measurement data and applies no formal meta-analytic pooling or risk-of-bias scoring. Per-study figures pulled from Table 1 should be verified against the constituent papers when those papers are ingested individually.
• Identity checks (all three clean). DOI 10.1515/reveh-2016-0042 grep against wiki/sources/ returned no hits. raw_handle: MFK_tavakkoli2016 grep returned no hits. ls wiki/sources/ | grep -i tavakkoli2016 returned no hits. No collision with existing pages.
• Frontmatter discipline. All metal, matrix, and jurisdiction slugs verified against the 2026-05-18 taxonomy snapshot for metals (Cd) and against the matrices vocabulary used by other biomonitoring source pages (urine, blood, breast-milk, hair, saliva, cord-blood, biomonitoring). Two matrix terms (amniotic-fluid, deciduous-teeth, umbilical-cord-blood, occupational-exposure) are newly introduced by this page; umbilical-cord-blood is already in use on at least one prior source page (per matrices-vocabulary grep, 2026-06-01). The other three are novel but follow the same bare-string controlled-vocabulary pattern. Flagging here so the matrices vocabulary list in docs/gpt-collaboration/system-prompt.md can absorb them on next refresh. ingredients: [] and products: [] are intentional — this is a human-biomonitoring corpus, not a food-matrix study, and the wiki does not route biomonitoring sources to ingredient or product pages.
• Speciation discipline (Part 14). All 18 constituent studies measured total elemental cadmium (Cd) by AAS, electrothermal AAS, or ICP-OES. Cadmium does not have an inorganic-vs-organic speciation distinction analogous to iAs/tAs or MeHg/tHg; the elemental form is the appropriate biomarker. The wiki page uses Cd (not a speciation-modified variant) accordingly.
• Brand firewall (Part 12). The review aggregates per-population and per-cohort biomarker concentrations across 18 constituent studies. No commercial brand names appear in the review’s tables, prose, or references with respect to sampled products (the Mobarakeh steel works named in Norouzi et al. (28) is an industrial-facility identifier, not a consumer brand, and is the subject of the environmental-exposure measurement rather than a commercial-product ranking — analogous to Part 12 Exception 1’s regulatory-event-subject framing, except here it is an industrial-source-attribution framing rather than a regulatory-event framing; the Mobarakeh attribution is reported as the review and the constituent study publish it). Scientific-method vendor identifiers (AAS instruments) are not named at the brand level by the review itself. No Part 12 redactions were required because no brand-attributed contamination values were present to redact.
• Wiki/HMTc firewall (Part 2). The review reports what 18 constituent studies found in Iranian biomonitoring; it does not propose HMTc threshold values, does not advocate for regulatory action beyond a general recommendation that industries handling Cd be located at a safe distance from residential areas, and does not target consumer-audience messaging. The wiki page mirrors this register: it reports the review’s numerical findings (per-study biomarker concentrations, occupational-vs-environmental gradients, smoking signals, transplacental-transfer signals) without synthesising across other wiki pages or proposing certification levers. Cross-paper synthesis is left to the Part 9 synthesis workflow.
• Reference-value attribution. The ICdA suspended-particulate ranges (rural/urban/industrial), the ICdA occupational range, the ICdA cigarette content figure, the ICdA food-matrix ranges (plant 30-150 ppb; fish/meat 5-40 ppb; organ meats up to 1000 ppb), and the JECFA/FAO 25 µg/kg bw/month PTMI are all reported as the review attributes them. The review cites these via web-archive links to cadmium.org and who.int rather than to JECFA-numbered reports directly, but the figures match the JECFA 2010 evaluation as documented elsewhere in the wiki (jecfa-cadmium-ptmi).
• Paper-internal data anomalies catalogued (four; none rise to stop-condition severity).
  1. Ebrahim 2011-2012 amniotic-fluid non-smoker SD scale. Table 1 row 1 publishes “Non-smokers = 780 ± 0.231 ng/L” while the same row’s smokers value is “13.4 ± 0.91 µg/L.” The mean (780 ng/L = 0.78 µg/L) is plausible in context but the SD scale (“0.231 ng/L”) is far too small to be a real standard deviation when the mean is 780 ng/L; this is almost certainly a typesetting error in the review’s Table 1 where the SD should read “0.231 µg/L” (= 231 ng/L) or “231 ng/L.” The wiki page transcribes the value as the review publishes it and flags the anomaly here. The Ebrahim et al. 2015 Iran J Neonat paper (ref 17) would need to be ingested individually to confirm the intended SD.
  2. Nazarpour and Goudarzi breast-milk unit labels (µg/mL vs µg/L). Table 1 rows 3 and 4 publish “5 ± 6.9 µg/mL” (Nazarpour, Varamin, 100 mothers) and “1.92 ± 1.04 µg/mL” (Goudarzi, Isfahan, 37 mothers). If literal, 5 µg/mL = 5 mg/L = 5,000 µg/L, which would be roughly 5,000× the WHO breast-milk Cd reference of ~1 µg/L and is biologically implausible for human breast milk. The other three breast-milk Cd values in the corpus (Norouzi-Lenjan 2.31 µg/L, Norouzi-Konjedjan 0.772 µg/L, Rahimi 2.44 µg/L) all use µg/L and sit in a 0.5-2.5 µg/L range that is consistent with the published Iranian and international breast-milk Cd literature. The Nazarpour and Goudarzi unit labels are almost certainly typesetting errors where the published values should read µg/L. The wiki page transcribes the values as Table 1 publishes them and flags the unit-label concern here; the Nazarpour 2013 Asian J Chem paper (ref 19) and the Goudarzi 2013 Toxicol Ind Health paper (ref 20) would need to be ingested individually to confirm the intended units. The Implications section’s “headroom-only-on-average breast-milk Cd values” bullet reads the figures under the µg/L interpretation, which is the only one that matches the surrounding literature.
  3. Kargar office-worker urine identical pre- and post-shift values. Table 1 row 7 publishes “Office workers (before their work shift) = 3.07 ± 7.24 µg/g creatinine” and “Office workers (after their work shift) = 3.07 ± 7.24 µg/g creatinine” — bitwise identical mean and SD for the two time points. For office workers without occupational Cd exposure, identical pre- and post-shift biomarker values are biologically plausible (no shift-related Cd exposure should change urinary Cd over an 8-hour interval), but the bitwise identity (down to the SD’s second decimal place) is suspicious for a transcription duplication. The wiki page transcribes the values as Table 1 publishes them and flags the duplication here; the Kargar 2015 Int J Occup Hyg paper (ref 23) would need to be ingested individually to confirm whether the two time points really produced identical statistics or whether the review’s table contains a copy-paste artefact. Note also that the SD (7.24 µg/g creatinine) is more than 2× the mean (3.07 µg/g creatinine), which is unusual for a creatinine-adjusted biomarker in an unexposed cohort.
  4. Tadayon 2009-2010 hair Cd SD-greater-than-mean. Table 1 row 11 publishes diabetes-cohort hair Cd as “0.192 ± 1.53 µg/g” — SD (1.53) approximately 8× the mean (0.192), which is implausible for a real distribution and suggests either a typo (e.g., SD intended as 0.153) or a heavy-right-tailed distribution where the mean is not a good summary statistic. The healthy-cohort value (0.076 ± 0.6 µg/g) shows a similar but less extreme pattern. The wiki page transcribes the values as Table 1 publishes them and flags the SD-vs-mean anomaly here; the Tadayon 2012 Acad Res Int paper (ref 27) would need to be ingested individually to confirm the intended SD.
• Basis. Concentration values in the review’s Table 1 are reported per the constituent studies’ published units: blood and urine in µg/L, µg/dL, or µg/g creatinine; hair in µg/g or µg/kg; breast milk in µg/L or (with unit-label concern, see anomaly 2) µg/mL; saliva in ng/100 mL; deciduous teeth in µg/g. The review does not normalise units across the corpus. The wiki page preserves the unit as each row publishes it. There is no wet-vs-dry basis question for biomarker matrices analogous to the food-matrix basis question.
• Sampling year range. The earliest constituent study with a stated sampling year is Vigeh 2003-2004 (ref 34); the latest is Ebrahim 2011-2012 (ref 17, with Ebrahim & Ashtarinejad’s 2015 publication year cited in the Discussion as the 2015 publication of those 2011-2012 data). Several constituent studies (Pashmi, Nazarpour, Nejadchehrazi, Tadayon ref 30, Farzin) report “year not reported” in Table 1. The 2003-2015 range in the frontmatter reflects the stated-sampling-year coverage; the actual data-collection envelope may extend slightly outside that range for the year-not-reported entries.
• Data integrity. All values in the Key numbers section transcribed from Table 1 (constituent-study figures, pp. 459-460), the Introduction reference-value paragraph (p. 457), the Discussion-section comparative figures (England, Iraq hair Cd; pre-eclampsia smoker/non-smoker percentages; pp. 461-462), and the Results-section corpus summary (p. 458). The four anomaly catalogue entries above are the only paper-internal inconsistencies observed; none requires the source to be quarantined or held back from ingest. Constituent-study figures should be cross-verified against the original papers if those papers are independently ingested.
• Downstream pooling caveat. Because this is a narrative review aggregating prior values rather than a primary measurement source, the Part 9 synthesis workflow should treat per-study values surfaced by Table 1 as pointers to the constituent paper (the cited study should be ingested individually for authoritative per-study figures), not as independent quantitative observations. The review’s value to the wiki is as (a) a corpus pointer for Iranian Cd biomonitoring literature 2003-2015, (b) a consolidated Iran-specific occupational-vs-environmental gradient, and (c) a smoking-Cd and transplacental-Cd-transfer signal corroborated across multiple Iranian cohorts.

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.