Inorganic arsenic in rice-based products for infants and young children

Overview

UK-market survey speciating inorganic arsenic (Asi), DMA, and MMA in 179 commercial rice-based products sold for weaning and toddler-age consumption (baby rice, rice cereals, rice crackers), compared statistically against the US FDA 2014 published rice/rice-product survey for the same three product categories. Asi median across the UK baby-rice subset was 121 µg/kg dry weight, with 14% of samples exceeding the JECFA-proposed 200 µg/kg maximum for polished rice; rice crackers from the UK market showed a significantly higher Asi median (111 µg/kg) than the US-market rice cracker subset (79 µg/kg, p=0.011). Organically-produced samples, which were predominantly whole-grain, showed significantly higher Asi than non-organic milled samples within each product category. The paper closes by arguing that regulatory limits specifically for infant rice-based foods are urgently needed because the modelled per-serving Asi exposures fall within the EFSA BMDL01 risk range for the youngest infant population.

Key numbers

Present-study UK samples (Table 1, mg/kg dry weight):

Baby rice (n=29, 5 manufacturer codes M001–M005):

ΣAs range across manufacturers: 0.063 – 0.334 mg/kg
Asi percentage of ΣAs range: 51.4 – 84.6 %
Asi concentration range: 0.056 – 0.268 mg/kg
14% of baby rice samples were above the JECFA-proposed Asi maximum for polished rice (0.200 mg/kg)
Highest-Asi manufacturer (single brand) had median 0.190 mg/kg (p<0.001 vs other brands in the set)
DMA range: 0.030 – 0.123 mg/kg; MMA range: 0.001 – 0.003 mg/kg

Rice cereals (n=53, 6 manufacturer codes M006–M011):

ΣAs range: 0.042 – 0.396 mg/kg
Asi percentage: 14.2 – 89.6 %
Asi range: 0.008 – 0.323 mg/kg
2% of rice cereal samples above the JECFA Asi maximum
One manufacturer (code M009) had a higher Asi median (0.234 mg/kg, p<0.001) — up to 1.6× the JECFA maximum
DMA range: 0.005 – 0.082 mg/kg; MMA range: 0.001 – 0.003 mg/kg

Rice crackers (n=97, 15 manufacturer codes — M001, M005, M006, M009, M011, M012, M013, M014, M015, M016, M017, M018, M019, M020, M021):

ΣAs range: 0.019 – 0.246 mg/kg
Asi percentage: 71.1 – 100 % (rice crackers showed the highest Asi proportion of the three product types)
Asi range: 0.019 – 0.212 mg/kg
1% of rice cracker samples above the JECFA Asi maximum
Highest-Asi brand (single manufacturer) median 0.205 mg/kg (p<0.001)
DMA range: 0.001 – 0.172 mg/kg

Organic vs non-organic stratification (Table 2, mg/kg dry weight, present study):

Product category	Production type	Type of rice	n	Asi median (range)
Baby rice	Non-organic	Milled	11	0.082 (0.056 – 0.128)
Baby rice	Organic	Milled	9	0.107 (0.024 – 0.120)
Baby rice	Organic	Whole grain	9	0.190 (0.117 – 0.268)
Rice cereals	Non-organic	Milled	48	0.070 (0.008 – 0.188)
Rice cereals	Organic	Whole grain	5	0.162 (0.103 – 0.323)
Rice crackers	Non-organic	Milled	23	0.071 (0.021 – 0.159)
Rice crackers	Non-organic	Whole grain	2	0.081 (0.018 – 0.211)
Rice crackers	Organic	Whole grain	52	0.121 (0.048 – 0.195)

Within-category Asi differences for organic vs non-organic were significant for baby rice (p<0.001), rice cereals (p<0.001), and rice crackers (p<0.001). The organic samples were predominantly whole-grain rice; the elevated Asi is attributed to higher Asi at the pericarp/aleurone (bran) layer relative to milled (polished) rice.

EU (present study UK) vs US FDA 2014 published-survey comparison (Table 3, mg/kg dry weight):

Product category	Market	n	Asi median (range)	DMA median (range)	MMA median (range)	ΣAs median (range)	Total As median (range)	Asi % (range)
Baby rice	US FDA 2014	85	0.114 (0.039 – 0.254)	0.071 (0.015 – 0.204)	0.003 (0.001 – 0.012)	0.215 (0.550 – 0.341) [sic — source typo; total-As bounds 0.060–0.373 indicate likely intended ΣAs range upper bound is ~0.341]	0.200 (0.060 – 0.373)	65.5 (22.0 – 82.9)
Baby rice	UK present study	29	0.121 (0.056 – 0.268)	0.041 (0.003 – 0.123)	0.001 (0.001 – 0.004)	0.175 (0.084 – 0.334)	—	69.8 (51.4 – 84.6)
	p-value		0.235	0.001	0.002	0.028	—	0.676
Rice cereals	US FDA 2014	105	0.091 (0.023 – 0.283)	0.042 (0.007 – 0.493)	0.001 (0.001 – 0.014)	0.135 (0.041 – 0.625)	0.132 (0.050 – 0.723)	63.0 (18.8 – 87.7)
Rice cereals	UK present study	53	0.075 (0.008 – 0.323)	0.037 (0.005 – 0.081)	0.001 (0.001 – 0.004)	0.119 (0.042 – 0.396)	—	70.6 (14.2 – 89.6)
	p-value		0.096	0.004	0.590	0.005	—	0.077
Rice crackers	US FDA 2014	199	0.079 (0.008 – 0.273)	0.034 (0.001 – 0.477)	0.001 (0.001 – 0.021)	0.116 (0.012 – 0.657)	0.121 (0.009 – 1.931)	65.1 (18.4 – 97.6)
Rice crackers	UK present study	97	0.111 (0.018 – 0.201)	0.025 (0.001 – 0.172)	0.001 (0.001 – 0.006)	0.141 (0.019 – 0.328)	—	80.9 (44.8 – 100.0)
	p-value		0.011	<0.001	0.224	0.153	—	<0.001

UK rice crackers had significantly higher Asi than US rice crackers (median 111 vs 79 µg/kg, p=0.011); UK and US baby rice and rice cereals did not differ significantly in Asi. UK samples consistently had a higher Asi proportion of total speciated arsenic (As), attributed by the authors to lower DMA in UK rice (geographic/cultivar effects on As speciation).

Per-serving exposure estimates (Table 4 and Fig. 2, present study; WHO body-weight standards):

Baby rice consumed by 4-month-old infants (recommended serving 15 g w.wt., range 15.0 – 15.0):

Asi intake per serving (UK present study, n=29): median 0.7 µg, range 0.3 – 2.2 µg
Asi intake per serving (US FDA 2014, n=85): median 1.7 µg, range 0.6 – 3.8 µg
3 servings/day at median body weight (6.70 kg) and median Asi/serving = 0.31 µg Asi/kg b.w./day
This could be 1.67- to 3.37-fold higher with 4 servings/day and 3rd-percentile body weight (5.35 kg) or 97th-percentile Asi/serving

Rice cereals consumed by 12-month-old young children (recommended serving 30 g w.wt., range 22.0 – 35.0):

Asi intake per serving (UK present study, n=53): median 2.2 µg, range 0.2 – 9.3 µg
Asi intake per serving (US FDA 2014, n=105): median 2.6 µg, range 0.8 – 11.0 µg
2 servings/day at median body weight (9.30 kg) = 0.47 µg Asi/kg b.w./day
This could be 2.48- to 4.58-fold higher with 4 servings/day and 3rd-percentile body weight (7.50 kg) or 97th-percentile Asi/serving

Rice crackers consumed by 12-month-old young children (recommended serving 30 g w.wt., range 8.0 – 50.0):

Asi intake per serving (UK present study, n=97): median 3.5 µg, range 0.5 – 9.1 µg
Asi intake per serving (US FDA 2014, n=199): median 2.4 µg, range 0.2 – 8.2 µg
1 serving/day at median body weight = 0.38 µg Asi/kg b.w./day
This could be 4.95- to 8.93-fold higher with 4 servings/day at 3rd-percentile body weight or 97th-percentile Asi/serving

All three scenarios fall within the EFSA (2009) BMDL01 range of 0.3 – 8 µg Asi/kg b.w./day, which the authors interpret as evidence that infant rice-product Asi exposure cannot be excluded from the cancer-and-skin-lesion benchmark range.

Analytical performance:

LOD for DMA and Asi: 0.003 mg/kg dry weight (n=6 blanks)
Speciation CRM recoveries (mean ± SE, n=15): DMA 105 ± 3 %, MMA 93.4 ± 5 %, Asi 94.5 ± 2 %
All samples were above LOD for Asi; only a few below LOD for DMA and MMA (½ LOD substituted)

Methods (brief)

UK-market sampling: 179 rice-based products purchased Feb 2014 – Mar 2016 from 36 UK food retailers (15 local shops + 21 supermarkets) — duplicate samples of the same product and brand were always purchased from different retailers. Manufacturers were assigned anonymous codes M001 – M021. Three product categories: baby rice (n=29), rice cereals (n=53), rice crackers (n=97).

Sample preparation: freeze-dried, then powdered to nominal homogeneity using a Retsch PM100 rotary ball-mill with zirconium-oxide-lined vessel and zirconium-oxide grinding balls. 0.1 g aliquots digested in 10 mL 1 % Aristar nitric acid (overnight rest, then microwave digestion on a CEM MARS 6 in batches of approximately 30 samples — three-stage slow heating program: ramp to 55 °C in 5 min and hold 10 min; ramp to 75 °C in 5 min and hold 10 min; ramp to 95 °C in 5 min and hold 30 min). 1 mL aliquot transferred to 2 mL polypropylene vial; 10 µL analytical-grade H₂O₂ added to convert arsenite to arsenate prior to chromatography.

Speciation analytical chain: Thermo Scientific IC5000 ion chromatography system with a Thermo AS7 2 × 250 mm column and Thermo AG7 2 × 50 mm guard column, ammonium-carbonate gradient mobile phase (A: 20 mM NH₄CO₃; B: 200 mM NH₄CO₃; gradient program detailed in §2.2), interfaced to a Thermo iCAP Q ICP-MS monitoring m/z 75 with He collision-cell gas to suppress argon-chloride interference. Authentic standards run for DMA, Asi (arsenate form), MMA, tetramethylarsonium (TETRA), and arsenobetaine (AB); presence under each chromatographic peak calibrated using a DMA concentration series.

Quality control: each digestion batch included a method blank and NIST 1568b rice flour CRM. Additionally, batches of 30 samples were prepared which include a blank and the rice-flour CRM. As-speciation certified reference material (DMA, Asi, MMA) used for speciation recovery; certified DMA 0.182, MMA 0.012, Asi 0.092 mg/kg; recoveries 105 ± 3 %, 93.4 ± 5 %, 94.5 ± 2 % respectively (n=15). LOD for DMA and Asi: 0.003 mg/kg d.w. (½ LOD substituted for the small number of <LOD DMA/MMA samples in statistical analysis).

Statistics: general linear model and Duncan’s multiple-range test for between-manufacturer differences; IBM SPSS Statistics 21.0. US comparison data: Asi, DMA, MMA, total As, and ΣAs values from the published US FDA 2014 rice-and-rice-products survey (US FDA 2014; 85 baby rice, 105 rice cereals, 199 rice crackers) — these are reanalysed in Table 3 alongside the UK present-study data, not independently sampled by this paper.

Speciation flag: the paper reports both inorganic arsenic (Asi) and total speciated arsenic (ΣAs = Asi + DMA + MMA) by direct chromatographic speciation. Total As (sum of all species including any unspeciated As) is reported only for the US FDA-derived rows in Table 3; the UK present-study rows report ΣAs (speciation sum) rather than acid-digested total As.

Authors disclose Marie Curie Intra-European Fellowship funding (PIEF-GA-2013-622096) within the EU 7th Framework Programme.

Implications

Certification (HMTc): contributes A-tier UK-market occurrence data on inorganic arsenic in three rice-based-product subcategories that are within the HMTc infant-and-child-foods category. The 29 baby-rice and 53 rice-cereal samples both contribute Asi data to the dry-as-sold baby-cereals-dry-rice-based row; the 97 rice-cracker samples contribute Asi data to the teething-and-snacks-rice-based row. Source-stated medians, ranges, and percentile-style information are available; source does not report sample-level Asi values that could be admitted one-by-one into a pooled percentile distribution. The dataset is also directly comparable to the US FDA 2014 rice/rice-products survey, which is on a separate FDA source page; the present paper’s Table 3 is the statistical bridge between the two.

Courses: useful case study for brand QA and regulatory-affairs courses on speciation discipline (Asi vs ΣAs vs total As), on the organic / whole-grain Asi elevation in rice products marketed as “organic” or “wholegrain” (an audience-counterintuitive finding that brands and retailers should be prepared to communicate), and on the per-serving exposure framing for infant populations.

App: contributes Asi occurrence data (UK market) for the rice → baby-cereals-dry-rice-based and rice → teething-and-snacks-rice-based cells of the contamination-likelihood matrix. The 14% exceedance over the JECFA 0.200 mg/kg proposed maximum for baby rice is a notable upper-tail anchor.

Update history

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