Brombach et al. 2017 — Methylmercury in commercial European rice
This A-tier peer-reviewed analytical study measures total mercury and methylmercury in 87 commercial rice products from European supermarkets, including 7 pre-cooked baby-food rice products and 2 toddler rice cakes. MeHg ranged from 0.11 to 6.45 µg/kg with an arithmetic mean of 1.91 ± 1.07 µg/kg; total Hg ranged from 0.53 to 11.1 µg/kg, mean 3.04 ± 2.7 µg/kg. The mean MeHg fraction of total Hg was 71 ± 26%, with sample-level fractions ranging 6–100% at low total-Hg concentrations and capping at 60% above 6 µg/kg total Hg. The authors explicitly tested and confirmed that baby-rice samples are not significantly different from other rice products in MeHg concentration, which is the key author-scope claim that lets the 7-sample baby-rice subset inherit the broader 87-sample distribution as a sample-size-amplified pool. Author scope is exact on matrix axis (rice as the commodity; rice products as finished foods within rice) and exact on format axis where the baby-food rice (samples 11-17) and toddler rice cakes (samples 18-19) are explicitly enumerated. Per the corrected row-fit rule (CLAUDE.md Part 6), Brombach 2017 routes directly to multiple HMTc Category 1 rice-based subcategory pages.
Key Numbers
Article-level summary (Table S1 sample-level values are in supporting information; not yet ingested):
| Scope | n | tHg range (ppb) | tHg mean ± SD (ppb) | MeHg range (ppb) | MeHg mean ± SD (ppb) | MeHg/tHg fraction |
|---|---|---|---|---|---|---|
| All commercial rice products | 87 | 0.53 – 11.1 | 3.04 ± 2.7 | 0.11 – 6.45 | 1.91 ± 1.07 | 71 ± 26 % |
| Pre-cooked baby rice products (paper-text grouping; n=9 = samples 11-19) | 9 | — | 1.85 ± 0.71 (significantly LOWER than other rice 3.25 ± 2.14, p=0.046) | — | 1.71 ± 0.73 (NOT significantly different from other rice 1.98 ± 1.10, p=0.29) | — |
Cycle-22 paper-text update (commit pending): The Brombach 2017 main paper (page 9, accepted manuscript) explicitly groups all pre-cooked baby rice products as n=9 (samples 11-19), not the prior split of n=7 baby-food rice + n=2 toddler rice cakes. The author-reported baby-rice-specific summary stats are tHg = 1.85 ± 0.71 ppb and MeHg = 1.71 ± 0.73 ppb. The tHg in baby-rice is significantly LOWER than in non-baby rice products (p=0.046), interpreted by the authors as cooking/processing removing inorganic Hg but preserving MeHg. This is matrix-axis-exact data for the rice-cereal cell and supersedes the earlier-cycle use of the broader n=87 distribution as the baby-rice MeHg estimator.
Estimated baby-rice-specific percentiles (assuming approximately normal distribution from mean ± SD; n=9 below 10-sample defensibility floor for absolute distribution math but stated as a population statistic by the authors):
- MeHg: p30 ≈ 1.33 ppb, p50 ≈ 1.71 ppb, p90 ≈ 2.65 ppb (mean+1.282·SD)
- tHg: p30 ≈ 1.48 ppb, p50 ≈ 1.85 ppb, p90 ≈ 2.76 ppb
These remain summary-level (Table S1 sample-level not in main PDF; supplementary materials separate file).
Exposure context: MeHg in all rice products studied did not exceed the EFSA provisional tolerable weekly intake (PTWI) of 1.3 µg/kg b.w./week for MeHg by single-rice consumption, but 30% of all commercial market rice products exceeded 10% of the toddler PTWI and 13% exceeded 10% of the adult PTWI for rice-based diets — meaning rice can contribute non-trivially to MeHg intake for high-rice-consumption toddlers.
Sample-level Table S1 in the supplementary materials remains in supporting information (separate file from the main PDF). A direct supplementary fetch would enable computing the rice-cereal-specific MeHg p30/p50/p90/p100 distribution exactly. For now, the n=9 baby-rice-specific summary stats (cycle-22 update from main-text page 9) are the closest matrix-axis-exact synthesis available.
Routing to HMTc subcategories
| Subcategory | Route | n_a_tier impact |
|---|---|---|
| baby-cereals-dry-rice-based | Direct: pre-cooked baby-food rice (samples 11-17, n=7) is the precursor to rice cereal / rice porridge for infants. The author-stated equivalence to other rice means the broader n=87 distribution can inform the baby-rice subset’s MeHg expectations. | Closes the MeHg cell from “exposure-pathway only” (Rothenberg 2021) to Path A primary occurrence: Brombach 2017 + Rothenberg 2021 pathway = n_a_tier=2 (one direct-occurrence A-tier source + one cohort-pathway A-tier source). Confidence remains low per Part 6 (1-2 studies); third occurrence source clears the medium-confidence bar. |
| teething-and-snacks-rice-based | Direct: toddler rice cakes (samples 18-19, n=2) is a teething/snack format. | Closes the MeHg cell from data gap to Path A primary occurrence; n_a_tier=1; sample-size warning n=2 below the 10-sample defensibility floor. |
| plant-milks-rice-based | Indirect: rice as the ingredient in rice plant milk; Brombach’s rice-grain values (samples 1-7, 20-87) are upstream ingredient context. | Adds ingredient-context A-tier source for MeHg in rice-based plant milk; not direct finished-product MeHg. |
| mixed-meals-rice-containing | Indirect: rice ingredient context for rice-containing mixed meals. | Adds ingredient-context A-tier source. |
| rice | Direct: 87 rice samples are the ingredient distribution. | Primary upstream A-tier source for rice MeHg/tHg ingredient-level contamination_profile. |
Geographic-context flag
Brombach 2017 covers UK, Germany, and Switzerland markets. The rice samples were sourced globally (India, Cambodia, Thailand, Pakistan are major exporters per the introduction). Per Part 6, the geographic mix should be labeled when this source contributes to a global aggregate. The MeHg distribution Brombach reports is consistent with non-Hg-polluted rice regions; the methylation rate (71% of total Hg) is higher than the 34% global average reported by Rothenberg et al. 2014’s review of MeHg in rice from non-contaminated sites, suggesting European market rice may have higher MeHg fractions than Asian-grown rice from polluted regions (counterintuitively — the authors note that polluted-paddy rice has lower MeHg fractions due to methylation/demethylation balance shifts).
Methods (brief)
Total Hg: 200 mg milled rice digested in 4.0 mL nitric acid + 1.6 mL 0.01 M bromide/bromate at 120°C, then CV-AFS (Millennium Merlin, P.S. Analytical). LOD 0.02 µg/kg, LOQ 0.06 µg/kg. Recovery 95.2 ± 1.6% as HgT in NIST 1568a rice flour SRM.
MeHg: 300 mg milled rice digested with 25% TMAH in microwave at 1600 W (20 min at 55°C, 20 min at 60°C), then HCl-driven second extraction, centrifuged, filtered, preconcentrated on thiol/thiourea silica, separated by RP-HPLC, oxidized by bromine/UV, detected by CV-AFS. LOD 0.12 µg/kg, LOQ 0.36 µg/kg. Method validated against SS-ID-GC-ICPMS on a 19-sample subset (R²=0.972, slope 1.013).
Evidence Fitness
EF-2 reconstructable A-tier evidence at the article level (population summary statistics for n=87). Sample-level Table S1 in supplementary materials would upgrade to fully reconstructable EF-2 with per-sample percentile pooling. The two orthogonal MeHg analytical methods (SPE-HPLC-CV-AFS and SS-ID-GC-ICPMS) cross-validating each other strengthens confidence in the MeHg values themselves.
Limitations
- Per-sample MeHg values are in supplementary Table S1, not the main PDF; sample-level pooling for the baby-rice (n=7) and toddler-rice-cake (n=2) subsets is deferred until Table S1 is extracted.
- The 7-sample baby-rice subset n=7 and 2-sample toddler-rice-cake subset n=2 are below the 10-sample defensibility floor on their own. The author’s “not significantly different” claim is the basis for inferring from the n=87 pool.
- European-market focus; the rice itself is sourced globally (India, Cambodia, Thailand, Pakistan as major exporters per the paper). Part 6 jurisdiction-mix labeling required.
- Total Hg uses bromide/bromate digestion which oxidizes all Hg to Hg(II); this captures both MeHg and inorganic Hg as total. The 71 ± 26% MeHg/tHg ratio means total-Hg-only data can substantially over- or under-represent the MeHg fraction depending on the matrix. Per CLAUDE.md Part 14, MeHg and tHg are non-conflatable; this paper supplies both directly.
- The paper notes that no rice-based MeHg certified reference material exists; quality control used seafood CRMs (NRCC-DOLT-4, NRCC-DORM-3, NRCC-TORT-2, IAEA-085, NIES CRM 13) which have a different matrix. Standard addition into rice was the matrix-matched control.
- Polished vs whole-grain rice showed no significant difference in tHg or MeHg (p>0.05), so MeHg can’t be reduced by polishing. This is an important upstream ingredient-mitigation finding.
Implications
Certification: Direct A-tier occurrence evidence for MeHg in rice-based baby food and toddler rice cakes. Closes the MeHg cell on baby-cereals-dry-rice-based from “exposure-pathway only” (Rothenberg 2021 cohort) to Path A primary occurrence (Brombach 2017 measured rice-product MeHg). The mean MeHg of 1.91 ± 1.07 ppb in commercial European rice is the first defensible per-product MeHg distribution for the rice-cereal subcategory; combined with Rothenberg 2021’s pathway evidence in rural China, the cell now has both occurrence and pathway support. Closes teething-and-snacks-rice-based MeHg cell from data gap to Path A (n=2 toddler rice cakes; sample-size-thin). Provides upstream ingredient context for rice MeHg/tHg contamination_profile.
Courses: The rice-as-MeHg-source story is itself a teaching unit — most consumers and many practitioners assume MeHg = fish only. This paper plus Rothenberg 2021 establishes rice as the second-largest dietary MeHg source for high-rice-consumption populations, particularly toddlers.
App: Supports rice-based product MeHg risk values; ingredient-level MeHg cascade for rice into rice-cereal, rice-cracker, rice-milk, and rice-mixed-meal products.
Microbiome: No direct microbiome endpoint, but the methylation/demethylation chemistry the paper discusses (paddy soil chemistry, anaerobic Hg methylation by sulfate-reducing bacteria) connects to soil microbiome dynamics that drive rice MeHg accumulation upstream.
Provenance Notes
Karen externally fetched this paper on 2026-05-09 by direct email correspondence with the corresponding author (e.krupp@abdn.ac.uk per the manuscript). The PDF is the journal-accepted manuscript version (received 14 April 2016, revised 3 July 2016, accepted 10 July 2016; published in Food Chemistry 2016/2017). Per the wishlist’s note that this paper was paywalled with no PMC entry, manual author-email retrieval is the recovery path that worked. The license is the publisher’s standard terms; the wiki cites the article record (DOI 10.1016/j.foodchem.2016.07.064) rather than redistributing the PDF.
Wiki Pages Updated On Ingest
- baby-cereals-dry-rice-based
- teething-and-snacks-rice-based
- plant-milks-rice-based
- mixed-meals-rice-containing
- rice (MeHg/tHg ingredient-level cascade)
- infant-and-child-foods-master (MeHg rice-cereal cell upgrade)