Jafari et al. 2018 — Iran systematic review of heavy metals in Iranian-grown vs imported rice (10 studies, 2000-2017) with adult EWI/PTWI comparison

This Data in Brief paper is a systematic-review companion data article that pools 10 primary Iranian rice-contamination studies retrieved from Persian and English databases over a 2000-2017 search window. The authors recomputed weighted-mean concentrations (mg/kg dry weight) of eight elements — Cd, Pb, total As, total Cr, Zn, Ni, Cu, Co — separately for Iranian-grown and imported rice brands sold in Iran, then computed estimated weekly intakes (EWI) at 770 g/week per-capita rice consumption and 60 kg adult body weight, comparing each EWI against the FAO/WHO PTWI. Across both rice categories, every EWI was below its PTWI. The pooled Iranian-grown vs imported means (mg/kg dry weight) were Cd 0.16 vs 0.13, Pb 0.196 vs 0.55, tAs 0.046 vs 0.057, Cr 0.29 vs 0.61, Zn 26.13 vs 3.46, Ni 0.22 vs 0.76, Cu 16 vs 2.08, and Co 0.29 vs 0.29; significant differences (P<0.001) were reported between Iranian-grown and imported rice for all elements except Co (P=0.99) and Zn (P=0.99). Imported rice was higher in Pb, tAs, Cr, and Ni; Iranian-grown rice was higher in Zn and Cu. The paper is a pooled summary, not original primary measurements, and the constituent-study labels in Tables 1-2 are partly mislabelled relative to the reference list (see Verification notes); the pooled means in Table 3 are the operative result. Evidence tier B.

Key numbers

Table 1 — Iranian-grown rice constituent studies (each row is one primary study; values mg/kg, presumed dry weight per the paper’s pooling basis)

The paper’s Tables 1-2 column alignment is partially scrambled in the published version; per-row metal identities here are reconstructed from the column header order (Location, Sample size, Cd, Pb, As, Cr, Zn, Ni, Cu, Co) and verified against Table 3 pooled-mean sample-size sums (e.g., Iranian-grown Zn N=144 = 45 [Tehran-2015] + 99 [Lorestan-2010] ✓; Iranian-grown Cu N=166 = 45 + 99 + (22 unattributed); Iranian-grown Cd N=295 = 45+30+7+99+15+99 = 295 ✓).

Table 2 — Imported rice constituent studies (each row is one primary study; values mg/kg presumed dry weight)

Table 3 — Pooled mean concentrations (Iranian-grown vs imported; mg/kg, presumed dry weight per source pooling basis)

The “DF” column label in the source is ambiguous; values appear to be a mean-difference statistic (Iranian-grown minus imported, except Zn and Cu which run the other direction), not degrees of freedom. The paper does not define “DF” in text. The P-value column is reproduced as printed; P < 0.001 is reported on both rows of each pair where the test was significant.

Table 4 — Permissible-value comparison (Iranian-grown and imported EWI vs PTWI; 770 g/week rice consumption per capita; 60 kg adult body weight)

EWI formula (Eq. 1): EWI = C × WC / BW, where C = mean heavy-metal content (mg/kg dry weight), WC = weekly rice consumption (770 g/week per capita), BW = 60 kg (Iranian adult average). The paper does not state how the per-capita consumption figure was derived.

Search-strategy yield (Fig. 1, p. 455)

• Initial hits across databases: Google Scholar 10 (keywords) + 3 (abstract); Magiran 32 (keywords); ScienceDirect 2 (abstract & keyword); PubMed 11 (title & abstract); IranDoc 72 (all fields). Total 130.
• Excluded for irrelevance after title screening: 89 → 41 remaining.
• Excluded for being theses/conference works: 20 → 21 remaining.
• Excluded for no full-text access, insufficient sample size, or methodology problems: 11 → 10 selected for analysis.

Authors’ summary risk conclusions (§“Human Risk assessment and exposure to toxic metals”, pp. 455-459)

• Cd: EWI (Iranian-grown 2.05; imported 1.69 µg/kg BW per week) below PTWI 7 µg/kg BW per week — no health risk identified.
• Pb: significant Iranian-grown vs imported difference (P<0.001); both EWIs below PTWI 25 µg/kg BW per week.
• tAs: both EWIs below the Iranian standard and the WHO guideline. The paper notes that cooked rice may increase the heavy-metal content to 1.642 mg/kg, citing prior literature (no constituent-study primary source).
• Cr: significant Iranian-grown vs imported difference; both EWIs below PTWI 23.3 µg/kg BW per week.
• Ni: both EWIs (Iranian 2.82; imported 9.75 µg/kg BW per week) below PTWI 35 µg/kg BW per week.
• The authors conclude: “the rice types consumed in Iran have no health hazard for consumers.”

Methods (brief)

Study design. Systematic review and re-analysis of primary studies reporting heavy-metal concentrations in Iranian-grown or imported rice sold in Iran, with re-aggregated weighted means by metal and rice category and an EWI/PTWI comparison.

Search strategy. Internal (Persian) databases SID (sid.ir), Magiran (Magiran.com), IranDoc (irandoc.ac.ir), and Iran Medex (iranmedex.com); international databases Google Scholar, PubMed, Web of Science, and ScienceDirect. Search terms “heavy metals”, “rice”, “Iran” (and Persian equivalents) in title, abstract, and keywords. Search window: publications 2000-2017.

Inclusion screening. Title screen excluded 89/130 for irrelevance; abstract/full-text screen excluded 20 theses and conference abstracts and 11 for insufficient sample size or no full-text access. Final n=10 studies retained for analysis.

Aggregation. Per metal, per rice category (Iranian-grown vs imported), the paper computes a pooled mean and standard deviation. The pooled-SD formula is not explicit in the text; the per-metal sample-size N is summed across constituent studies that measured that metal in that rice category. Pooled means agree with sample-weighted constituent means to two decimal places where checkable.

Reporting basis. The paper’s EWI equation (Eq. 1) labels C as “mean heavy-metal content (mg/kg dry weight)”; the constituent studies’ reporting bases are not explicit in the table. Downstream pooling should treat Table 3 values as dry-weight rice unless contradicted by a constituent study’s original record.

Risk-assessment parameters. WC (weekly rice consumption) = 770 g/week per capita; BW (Iranian adult average) = 60 kg; EWI compared against FAO/WHO PTWI values (Cd 7, Pb 25, tAs 15, Cr 23.3, Zn 420, Ni 35, Cu 500 µg/kg BW per week; Co MTDI 700 µg/kg BW per week, no PTWI established).

Speciation. Total Pb, total Cd, total As, total Cr, total Zn, total Ni, total Cu, total Co only; no inorganic-As fractionation despite rice being the dominant dietary iAs pathway. No mercury panel.

Instrumentation. Not described — the paper is a literature aggregation; analytical methods are whatever the constituent studies used and are not reproduced here.

Statistics. Not explicit; P-values reported in Table 3 with “DF” column whose meaning is not defined in text (appears to be a between-group mean difference, not degrees of freedom).

Implications

This source contributes a pooled Iran-market summary of heavy-metal concentrations in Iranian-grown vs imported rice over 2010-2016, plus an adult EWI/PTWI comparison. Its principal contributions to the wiki evidence pool:

• Pooled rice Pb occurrence (dry weight): Iranian-grown 0.196 mg/kg (N=295), imported 0.55 mg/kg (N=320). Imported rice is markedly higher; both pools include several constituent studies already separately ingested in the wiki (e.g., Naseri 2015 Shiraz). When pooling with other Iranian-rice Pb sources downstream, this source should not double-count with already-ingested constituents — its value is its pooled-mean summary, not its underlying primary data.
• Pooled rice Cd occurrence (dry weight): Iranian-grown 0.16 mg/kg (N=295), imported 0.13 mg/kg (N=160). Iranian-grown slightly higher than imported, with P<0.001 between groups.
• Pooled rice tAs occurrence (dry weight): Iranian-grown 0.046 mg/kg (N=7), imported 0.057 mg/kg (N=44). The tAs N is small and the speciation is total As, not iAs — the value should not be combined with iAs-speciated rice survey data without speciation correction.
• Pooled rice Cr occurrence (total Cr, dry weight): Iranian-grown 0.29 mg/kg (N=45), imported 0.61 mg/kg (N=195). Significant difference (P<0.001). Speciation is total Cr, not Cr-VI.
• Pooled rice Ni occurrence (dry weight): Iranian-grown 0.22 mg/kg (N=195), imported 0.76 mg/kg (N=15). Imported rice is markedly higher but the imported-rice Ni N is only 15 samples from a single constituent study (Shiraz-2015 [5]), limiting confidence.
• Pooled rice Zn, Cu, Co occurrence (dry weight): non-HMTc analytes, retained as context. Iranian-grown rice has much higher Zn and Cu pools than imported (26.13 vs 3.46 mg/kg Zn; 16 vs 2.08 mg/kg Cu), but each pool reflects only 1-2 constituent studies and the large Iranian-vs-imported gap likely reflects which studies measured these elements (Lorestan-2010 and Tehran-2015 for Iranian-grown Zn/Cu; All-2016 for imported), not a true ecological difference.
• Adult exposure parameter set: Iranian adult per-capita rice consumption of 770 g/week (~110 g/day, ~40 kg/year) at 60 kg body weight. The 770 g/week figure is not sourced in the paper; comparable but lower than the 322 g/day rice figure in Heshmati et al. 2020 (which uses ISIRI 2010 No. 12968 = 322 g/day = 2254 g/week, ~3x higher), suggesting Jafari et al. used a different consumption reference or a different basis (e.g., dry rice as sold vs cooked rice). Downstream synthesis should not adopt the 770 g/week figure as the canonical Iranian rice consumption value without resolving this discrepancy.
• EWI/PTWI comparison: all 7 EWI values (Cd, Pb, tAs, Cr, Zn, Ni, Cu) for both rice categories are below their respective PTWIs at the 770 g/week / 60 kg parameterisation. The paper concludes no health hazard; this conclusion is contingent on the 770 g/week consumption figure and the dry-weight pooling basis, both of which are uncertain.

This is a B-tier source: peer-reviewed in Data in Brief (CC BY 4.0 open access), but a literature aggregation rather than original primary measurements, with documented paper-internal labelling inconsistencies (constituent-study labels in Tables 1-2 partly mismatched with the reference list), an undefined “DF” column in Table 3, an unsourced 770 g/week consumption figure, and ambiguous wet-vs-dry-weight reporting basis. Useful as a single-row summary of pre-2017 Iranian rice surveillance state for pooling with other Iranian and regional rice meta-analyses (e.g., Huang et al. 2022 China Pb-Cd-rice meta), but its underlying constituent studies should be tracked separately to avoid double-counting in downstream synthesis.

Wiki pages this source may touch

• cadmium
• lead
• arsenic-total
• chromium
• nickel
• zinc
• copper
• cobalt
• rice
• rice-bulk-grain
• jecfa-cadmium-ptmi
• jecfa-lead-ptwi-withdrawn

Verification notes

• Cite-key choice. jafari2018-iran-rice-meta follows the descriptive-suffix convention (first author, year, region, topic short-slug); meta signals systematic-review / pooled-summary rather than primary measurement, consistent with adjacent cite-keys (e.g., huang2022-pb-cd-rice-china-meta).
• Evidence tier B. Peer-reviewed in Data in Brief (Elsevier, CC BY 4.0). Tier B not A because: (a) the source is a literature aggregation of 10 primary studies rather than original primary measurements; (b) the constituent-study labels in Tables 1-2 are partly mismatched with the reference list (see paper-internal inconsistencies below); (c) the “DF” column in Table 3 is undefined in text; (d) the EWI parameterisation uses 770 g/week per-capita rice consumption with no cited source, conflicting with the more commonly cited ISIRI 2010 No. 12968 figure of 322 g/day (used by Heshmati et al. 2020); (e) the dry-vs-wet weight reporting basis is implicit, not stated per row, despite the EWI equation labelling C as mg/kg dry weight.
• Paper-internal inconsistency #1 — Constituent-study labels vs reference numbering. Tables 1-2 use compact “Location-Year” labels for each constituent study (e.g., “Shiraz-2010”) but the bracketed reference numbers in those rows do not always correspond to a study matching that location and year. Specific cases identified:
  • Table 1 row “Kashan-2015” cites [1] = Naseri et al. 2015 — but Naseri 2015 is “Concentration of some heavy metals in rice types available in Shiraz market” (Food Chem. 175:243-248), a Shiraz-market study, not Kashan-2015.
  • Table 1 row “Shiraz-2010” cites [5] = Adibi et al. 2014 — but Adibi 2014 is the Kermanshah washing-and-soaking study (J. Kermanshah Univ. Med. Sci. 17:628-636), not a 2010 Shiraz survey.
  • Table 1 row “Kermanshah-2013” cites [3] = Rabbani et al. 2015 — but Rabbani 2015 is the Kashan shopping-centre study (Int. Arch. Health Sci. 2:25-29), not a 2013 Kermanshah survey.
  • Table 2 row “All-2016” cites [8] = Falahi et al. 2010 — but Falahi 2010 is in the 2010 issue (Food Addit. Contam. A 3:80-83), not 2016. The “All-2016” label appears to reference Yousefi et al. 2016 (ref [9]) which is the actual 2016 imported-rice study.
  • These appear to be label-vs-citation transcription errors by the authors. The pooled means in Table 3 are still arithmetically consistent with the row values printed in Tables 1-2 (sample-weighted sums verified for Iranian-grown Cd N=295=45+30+7+99+15+99; imported Pb N=320=100+81+60+30+14+35), so the pooling math holds even though the row labels are unreliable. Downstream synthesis should treat Table 3 as the canonical result and ignore the Tables 1-2 row labels.
• Paper-internal inconsistency #2 — DF column undefined. Table 3 has a column header “DF” with numeric values per metal (e.g., Cd 0.03, Pb -0.354, tAs -0.011, Cr -0.32, Zn 22.67, Ni -0.54, Cu 13.92). The text never defines this column. The values are not degrees of freedom (which would be integers); they appear to be a between-group mean-difference statistic (Iranian-grown mean minus imported mean for Cd 0.16-0.13=0.03 ✓ and for Pb 0.196-0.55=-0.354 ✓), but the sign convention flips for some metals (Zn 26.13-3.46=22.67 ✓ but Cu 16-2.08=13.92 ✓ also matches), so the consistent interpretation is “Iranian-grown minus imported mean.” This is documented for downstream awareness; the values are not reused as pooled SD estimates or test statistics.
• Paper-internal inconsistency #3 — Tables 1-2 column alignment in published PDF. The published Tables 1-2 have partially-scrambled column alignment in the typeset version (some ”—” en-dashes are missing, causing values to shift columns visually). The metal-to-column mapping in the Key numbers tables above is reconstructed from the column header order and verified against the Table 3 pooled-sample-size sums (e.g., Iranian-grown Zn N=144 = Tehran-2015 N=45 + Lorestan-2010 N=99 ✓). The reconstruction is robust to the printing artefact but readers consulting the original PDF should not rely on visual column alignment alone.
• Paper-internal inconsistency #4 — Consumption figure unsourced. §1.2 (p. 455) states “the weekly rice consumption for each person (g week⁻¹) (770 g per capita per week)” with no citation. The contemporary Iranian official figure (ISIRI Standard No. 12968, 2010, used by Heshmati et al. 2020) is 322 g/day, equivalent to 2254 g/week — roughly 3× the Jafari et al. value. Possible explanations include a different rice-form basis (raw dry rice as sold vs. cooked rice), but the paper does not disambiguate. The 770 g/week figure should not be carried forward into wiki synthesis as Iran’s canonical rice consumption without resolving this gap.
• Paper-internal inconsistency #5 — Body weight figure. §1.2 (p. 455) reports “the average of Iranian grown body weight is 60 kg.” This phrasing is grammatically garbled in the original (“Iranian grown body weight” appears to be a typesetting error for “Iranian adult body weight”); the value 60 kg differs from the 70 kg used by Heshmati et al. 2020 for the same population. Documented for downstream cross-source EWI/EDI harmonisation.
• Frontmatter discipline. All slugs verified against the 2026-05-18 taxonomy snapshot. ingredients/rice exists. products/rice-bulk-grain exists. metals use abbreviation vocabulary (Cd, Pb, tAs, Cr, Zn, Ni, Cu, Co) — tAs not iAs because the paper does not speciate arsenic; Cr not Cr-VI because total Cr only. matrices: [rice] uses the broadest matches vocabulary entry per the system prompt’s matrices guidance. jurisdictions: [IR] for Iran (the imports come from Thailand, India, and Pakistan per the paper, but the jurisdiction of analysis is Iranian-market rice).
• Sample_n convention. sample_n: 10 reflects the systematic-review unit-of-analysis (number of constituent primary studies), not the sum of underlying composite samples; the per-metal pooled sample-sizes are documented in sample_population and in Table 3.
• Speciation. Total As only — the paper does not separate iAs from organic species, so tAs is used. Total Cr only — no Cr(VI) speciation. No mercury panel at all. These limits are non-trivial for rice (where iAs is the dominant toxic species) and should be flagged when this source is considered for downstream iAs synthesis.
• Basis. The EWI equation (Eq. 1) labels C as “mg/kg dry weight”; the constituent-study reporting bases are not made explicit per row. Downstream synthesis should treat the Table 3 pooled means as dry-weight rice unless a constituent study record contradicts.
• Brand firewall (Part 12). The paper itself states “Iranian and imported rice brands” in the abstract but the constituent studies and Table 3 aggregate by rice category (Iranian-grown vs imported), not by brand name. No brand-level contamination values appear in this source page. The paper references that imported rice is sourced from Thailand, India, and Pakistan — these are country-of-origin labels, not brand names, and are retained.
• Wiki/HMTc firewall (Part 2). No HMTc threshold proposals, no consumer advisories, no synthesis claims about other Iranian rice surveys beyond what the paper itself draws. The paper’s own conclusion that “rice types consumed in Iran have no health hazard for consumers” is reported in Key numbers as the authors’ summary, with explicit framing that it is the paper’s conclusion contingent on the 770 g/week and 60 kg parameterisation — not the wiki’s view.
• Double-counting risk. Several constituent studies in this systematic review are independently ingested elsewhere in the wiki (Naseri 2015 Shiraz, Falahi 2010, possibly others). This is the principal downstream-routing concern: this source’s value is its pooled-mean summary, not its underlying primary data. Synthesis sessions should treat Table 3 as one row in the Iranian-rice pool and not also count each constituent study row separately when computing weighted means.
• Regulatory citations. The paper cites the FAO/WHO PTWI values for Cd (7), Pb (25), tAs (15), Cr (23.3), Zn (420), Ni (35), Cu (500), and the MTDI for Co (700) in µg/kg BW per week. JECFA Cd PTMI and the withdrawn JECFA Pb PTWI are routed via regulations/jecfa-cadmium-ptmi and regulations/jecfa-lead-ptwi-withdrawn. The paper also references “ISRI (No: 12968)” (Iranian Standard for Industrial Research) as a footnote in Table 4 without specifying which value it provides; ISIRI 12968 is the same Iranian per-capita food consumption standard cited by Heshmati et al. 2020.
• Data integrity. All Table 1 (7 rows × 10 columns), Table 2 (6 rows × 10 columns), Table 3 (16 rows × 7 columns), and Table 4 (8 rows × 5 columns) cells transcribed from the original PDF (pages 456-458). Pooled N sums verified for Cd (Iranian 295 = 45+30+7+99+15+99 ✓; imported 160 = 81+30+14+35 ✓), Pb (Iranian 295 = 45+30+7+99+15+99 ✓; imported 320 = 100+81+60+30+14+35 ✓), and Zn (Iranian 144 = 45+99 ✓; imported 100 = 100 ✓). The paper-internal label-vs-reference mismatches in Tables 1-2 are documented above and do not affect the pooled values.

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.