Islam et al. 2023 — Heavy metals in commercial spices, Noakhali, Bangladesh

This medRxiv preprint reports Pb, Cd, Cr, Cu, and Fe concentrations in 19 spice samples (15 non-branded, 4 branded) purchased from Sonapur and Maijdee marketplaces in the coastal Noakhali District of Bangladesh, paired with household-level ingestion-rate data from a 180-household survey across six clusters. Concentrations were measured by flame atomic absorption spectrometry (PerkinElmer PinAAcle 900H) after wet acid digestion. The authors compute Estimated Daily Intake (EDI), Total Hazard Quotient (THQ), Total Target Hazard Quotient (TTHQ), and cancer risk (CR) per sample using USEPA and WHO reference doses. The two HMI-relevant findings are: (a) chromium loading is the dominant non-carcinogenic risk driver across most powdered spices, with TTHQ exceeding 1 in ~37% of samples (coriander leaf, green chili, ginger, and all four chili-powder and turmeric-powder samples — branded and non-branded — drive this); (b) carcinogenic risk for Cr and Cd is above the 1×10⁻⁴ USEPA benchmark in coriander leaf, both turmeric-powder samples, both chili-powder samples, green chili, masala mix, and ginger. Cr is reported as total Cr (no Cr-VI speciation); As, Hg, Ni, Al, Sn are not measured. The source is a preprint and has not undergone peer review at the cited version.

Key numbers

Metals measured: Pb, Cd, Cr (total Cr, not speciated), Cu, Fe. Concentrations in mg/kg (= ppm), as reported by the authors (basis: dry/as-sold ground spice; method: PerkinElmer PinAAcle 900H AAS in triplicate after wet digestion in 1:4 HNO₃:H₂SO₄ at 130°C; R²=0.9897–0.9983; LOD = 0.01 ppm/mL).

Sample-level concentrations cannot be tabulated exhaustively from the source as supplied — Figure 2(a–e) provides per-sample bars and abstract-level summary statistics are the principal reported numeric synthesis. Reported abstract-level highs and category narrative below.

Highest concentration per metal (abstract; mean ± SD, mg/kg):

Other narrative concentrations stated in Results:

• Pb: cardamom 15.47 ± 1.93 highest, cinnamon (S-12) 10.05 ± 1.22 second.
• Cd: coriander leaf 1.65 ± 0.01 highest, non-branded turmeric powder (S-2) 1.49 ± 0.02, branded turmeric powder (S-4) 1.43 ± 0.01.
• Cr (total): bay leaf 31.99 ± 3.97 highest, coriander leaf 27.58 ± 1.50, cardamom 25.19 ± 2.26.
• Cu: dried chili 18.84 ± 1.97 highest, non-branded chili powder (S-1) 15.36 ± 2.28, branded cumin (S-16) 13.81 ± 2.38.
• Fe: text reports black pepper as highest at 9.29 ± 1.72; same Results paragraph then names green chili (S-5) 35.46 ± 3.94 and branded chili powder (S-3) 34.02 ± 3.18 — both larger than the “highest” figure. Figure 2(e) shows black pepper as the lowest, not the highest, Fe bar. See Verification notes.

Estimated Daily Intake (EDI) and risk metrics (Table 2, Table 3):

• EDI order across spices: Fe > Cr > Cu > Pb > Cd.
• Reference doses used (mg/kg-day): Pb 0.0035, Cd 0.0005, Cr 0.0003, Cu 0.04, Fe 0.007.
• TTHQ > 1 (potential non-carcinogenic health risk): S-1 non-branded chili powder (4.59), S-2 non-branded turmeric powder (3.51), S-3 branded chili powder (2.72), S-4 branded turmeric powder (1.70), S-5 green chili (5.55), S-11 ginger (2.21), S-17 coriander leaf (8.33). Cr THQ is the dominant driver in each.
• Cancer risk (CR > 1×10⁻⁴) for Cr and/or Cd in the eight samples explicitly enumerated by the authors (Results, page 14, just before Figure 3): “both branded and unbranded turmeric and chili powder, green chili, masala mix, ginger, and coriander leaf” → S-1, S-2, S-3, S-4, S-5, S-8 (masala mix), S-11 (ginger), S-17 (coriander leaf). Figure 3 visual inspection is consistent with this list and does not add other samples to the >1E-04 set.

Permissible limits cited by the authors (Mg/Kg/Day):

• Pb 0.1 (USEPA), Cd 0.1 (USEPA), Cr 0.1 / 0.25 (USEPA / WHO), Cu 50 (Muhseen et al.), Fe 300 (WHO).
• The authors state “none of the heavy metals exceeded the maximum permissible limits (MPL)” while simultaneously reporting TTHQ and CR exceedances; the MPL statement refers to per-day intake against the listed MPL columns, not per-kg concentration against food-safety limits.

Methods (brief)

Cross-sectional design. Nineteen commercially-available spice samples (50–125 g commercial sizes) were collected from Sonapur and Maijdee marketplaces in Noakhali (n=15 non-branded, n=4 branded). Spice-consumption frequency was collected from 180 households across six clusters via structured questionnaire; ingestion rate (IR) for the EDI calculation was derived from household daily/weekly usage. Wet digestion used 1:4 HNO₃:H₂SO₄ heated to 130°C for 1 h with 2 mL H₂O₂ added until digest cleared, then diluted to 50 mL with deionized water. Analysis: PerkinElmer PinAAcle 900H AAS, dual-beam optics with continuum-source double-beam background correction, deuterium arc lamp; triplicate measurement; linearity R²=0.9897–0.9983; LOD = 0.01 ppm/mL.

Health-risk calculations: EDI = (C_metal × IR) / BW; THQ = EDI / RfD; TTHQ = ΣTHQ; cancer risk = EDI × CSF. RfDs from USEPA (Pb, Cd, Cr, Fe) and Muhseen et al. (Cu); WHO MPL for Cu and Fe. The cancer risk computation reports CR for Cr and Cd (and a flat near-zero CR line for Pb — labeled “CR of Pd” in Figure 3, which the wiki reads as a typo for Pb).

Method limitations: Cr is reported as total Cr; the cancer-risk calculation assumes the Cr-VI cancer slope factor without speciation evidence, so the CR for Cr is an upper-bound assuming the worst-case speciation. No certified reference material is mentioned. No Hg, As, Ni, Al, or Sn measurement. Sample size per spice category is n=1 with triplicate analytical replicates — the standard deviations reflect instrumental replication, not biological/lot variation. The IR-from-household-survey approach overweights local consumption patterns and is not transferable outside Noakhali.

Implications

Certification: Treat this preprint as B-tier occurrence context for South Asian informal-market spice supply. The Cd and Cr levels in chili powder, turmeric powder, ginger, coriander leaf, bay leaf, and cardamom are consistent with the broader literature documenting Cd and Pb hot-spots in unmonitored Bangladeshi/South-Asian spice supply. Do not pull single-sample maxima into HMI threshold math: n=1 per spice category and the absence of certified reference material put this study outside A-tier occurrence anchors. The cancer-risk values for Cr that assume Cr-VI without speciation are an upper bound and should be flagged as such if cited downstream.

Courses: Useful for the contamination-pathway module — coastal-area soil contamination, informal-market lack of routine regulatory sampling, and turmeric/chili adulteration history are explicitly cited by the authors. Pair with the Cowell et al. 2017 ground-turmeric-Pb-in-US-recalls reference (cited in this paper as ref 13) and the Rahman 2020 / Akter 2016 prior Bangladesh turmeric-Pb studies for a regional thread.

App: Contributes a Bangladesh-informal-market scenario for spices as a Cd-, total-Cr-, and Pb-relevant ingredient category. The contamination signal is strong for chili powder, turmeric powder, coriander leaf, ginger, and bay leaf; individual-sample maxima should not be used as typical-occurrence anchors.

Wiki pages updated on ingest

• spices
• herbs-and-spices
• turmeric
• cinnamon
• black-pepper
• spices
• condiments-general
• lead
• cadmium
• chromium

Verification notes

• 2026-06-02 Claude Code fresh ingest from raw/manual-fetch/Kimi_Agent_Download Corruption Issue/condiments2_papers/04_Chocolate_Seasonings_Salt/02_Chowdhury_2023_heavy_metals_spices_Bangladesh.pdf (KADC corruption-folder drain). PDF reads cleanly end-to-end (text, tables, figures all extractable); no corruption observed in this file.
• The folder filename credits “Chowdhury” but the paper’s first author is Md Shahedul Islam (Akibul Islam Chowdhury is second author). Cite-key follows HMI convention: first-author surname + year → islam2023-noakhali-spices-bangladesh. The raw_handle preserves the filename token “chowdhury” because that is the on-disk artifact.
• Paper-internal inconsistency on Fe: the abstract and Results both name Black Pepper as the spice with the “greatest” Fe concentration at 9.29 ± 1.71 mg/kg, but the same Results paragraph immediately lists green chili (35.46) and branded chili powder (34.02) as “followed by” values — both larger than the supposed maximum. Figure 2(e) shows Black Pepper (S-19) as the shortest Fe bar (~10 mg/kg), with green chili and chili-powder bars at ~35 mg/kg. The wiki page records the values as the paper states them (with the inconsistency noted) and does not silently correct: this is “obvious mirroring pattern” in the sense that the labeling error is self-evident from the same paragraph. Downstream synthesis should use Fe with care from this source.
• Brand firewall (Part 12): the paper uses “branded” vs “non-branded” as structural categories without naming brand names. No brand attribution is reproduced. This is consistent with HMI brand-firewall posture.
• Speciation: Cr is total Cr (the paper does not differentiate Cr(III) from Cr(VI)); the authors apply a Cr cancer-slope factor that presumes the Cr-VI cancer profile, which makes the reported cancer risk an upper bound rather than a measured risk. HMI synthesis must not treat the paper’s CR(Cr) values as evidence for Cr-VI exposure.
• The paper labels the cancer-risk line for Pb as “CR of Pd” in Figure 3 — read as a typographical error for Pb. The Pb cancer-risk line in Figure 3 sits near zero across all samples.
• Ingredients other than turmeric, cinnamon, and black pepper (cardamom, coriander, ginger, cumin, fenugreek, clove, bay-leaf, chili-pepper, masala-mix) are routed to the umbrella [[ingredients/spices]] and [[ingredients/herbs-and-spices]] slugs per the existing taxonomy; per-spice ingredient pages are not created here (auto-stub freq-≥2 handles that downstream).
• Evidence tier: B. Preprint, not peer-reviewed at time of upload to medRxiv. Sample size n=1 per spice category with triplicate analytical replication. No CRM reported. Single-region (Noakhali) sampling limits transferability.
• Source SD-precision disagreement (carried through verbatim in Key numbers): the abstract reports Cd in coriander leaf as 1.65 ± 0.011 and Fe in black pepper as 9.29 ± 1.71; the Results paragraph reports the same data points as 1.65 ± 0.01 and 9.29 ± 1.72. The wiki uses the Results-paragraph precision (± 0.01, ± 1.72) in the Highest-concentration table for consistency. The abstract figures (± 0.011, ± 1.71) reflect the source’s own internal SD-rounding mismatch, not a wiki transcription error.
• Audit subagent (general-purpose fresh context, 2026-06-02) flagged an ❌ on the original cancer-risk sample list, which had included S-10 (coriander powder) alongside the eight samples the authors explicitly enumerate. Verified against the page-14 narrative (“CR of Cr and Cd for both branded and unbranded turmeric and chili powder, green chili, masala mix, ginger, and coriander leaf were higher than 1E-04”) and Figure 3 (S-10 Cr and Cd CR values both ~5–8×10⁻⁵, below the 1×10⁻⁴ benchmark). S-10 removed from the list; corrected to the eight samples the authors enumerate.

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.