Signes-Pastor et al. 2017 - infant urinary arsenic and rice-based weaning foods

This PLOS ONE study measured urinary arsenic species in 79 infants from Belfast before weaning, followed 11 infants with paired pre- and post-weaning urine samples, and measured arsenic species in 73 rice-based products purchased in Belfast in February 2016. The food analyses used IC-ICP-MS speciation and directly measured inorganic arsenic (iAs), DMA, and MMA in baby rice, rice crackers/cakes, and rice cereals. The study found higher urinary methylated arsenic species in formula-fed infants before weaning, higher urinary arsenic metabolites after rice-product weaning in the paired subset, and frequent exceedance of the source-cited EU 0.10 mg/kg iAs comparator for rice destined for foods for infants and young children.

Key numbers

Infant urine before weaning (Table 1; median with min-max, microgram/L):

Formula-fed infants had higher log MMA, log DMA, and log tAs than breastfed or mixed-fed infants (p<0.001, p=0.001, and p=0.001, respectively). The authors report formula-fed median urinary MMA, DMA, and tAs as 6.7-fold, 2.2-fold, and 1.9-fold higher than breastfed infants.

Paired infant urine before and after weaning (Table 2; n=11; median with min-max, microgram/L):

Post-weaning urine had higher log MMA, log DMA, and log tAs than paired pre-weaning urine (p=0.013, p=0.005, and p=0.008). Median post-weaning MMA, DMA, and tAs were 7.2-fold, 9.1-fold, and 4.8-fold higher than paired pre-weaning values. Post-weaning iAs trended higher (1.6-fold; p=0.067). Mothers reported that 91% of the paired-subset infants consumed rice or rice-based products as part of the weaning diet.

Rice-based product iAs in 2016 (Table 3; median with min-max, mg/kg):

The authors also compare these 2016 values with a 2014 pre-regulation dataset from their earlier work: baby rice n=29, median iAs 0.121 mg/kg, 58% above 0.10 mg/kg; rice crackers n=36, median iAs 0.127 mg/kg, 80% above 0.10 mg/kg; rice cereals n=53, median iAs 0.076 mg/kg, 18% above 0.10 mg/kg. Across all 73 rice-based products analyzed in 2016, 56% exceeded 0.10 mg/kg and the median iAs was 0.104 mg/kg. Inorganic arsenic represented a median 74% of summed arsenic species across the rice-based products, ranging from 46% to 86%.

Belfast tap-water samples were low in arsenic species: iAs median 0.042 microgram/L (0.039-0.047), DMA 0.029 microgram/L (0.027-0.031), and TMAO 0.053 microgram/L (0.050-0.057).

Methods (brief)

Pre-weaning urine samples were collected from 79 Belfast infants born in 2015 and classified by pre-weaning feeding mode: breastfed (n=20), formula fed (n=32), or mixed feeding (n=27). An 11-infant subset had paired pre- and post-weaning urine samples; the post-weaning samples were collected at median age 7.7 months. Urine was collected with pre-tested cotton pads, squeezed from the pads, aliquoted, frozen at or below -20°C, centrifuged, treated with hydrogen peroxide to convert arsenite to arsenate, and analyzed without dilution adjustment.

The food arm purchased 73 rice-based products in Belfast in February 2016: baby rice (n=13), rice crackers/cakes (n=29), and rice cereals (n=31), covering nine popular commercial brands or manufacturers. This source page reports only aggregate product-group data and does not reproduce brand-coded results. Samples were freeze-dried, powdered, digested in 1% nitric acid using a CEM MARS 6 microwave system, centrifuged, treated with hydrogen peroxide, and analyzed by ion chromatography coupled to ICP-MS (Thermo Scientific IC5000 with Thermo AS7/AG7 column/guard column interfaced to Thermo iCAP Q ICP-MS in collision-cell mode).

Arsenic species measured were DMA, iAs, MMA, tetramethylarsonium, and arsenobetaine. Reported quality-control recoveries were ClinChek urine CRM: iAs 113 +/- 2%, MMA 95 +/- 2%, DMA 102 +/- 2%, AsB 95 +/- 2% (n=9); NIST 1568b rice flour: iAs 99 +/- 1%, MMA 126 +/- 2%, DMA 101 +/- 3% (n=4). LODs were 0.005 microgram/L for urine and 0.003 mg/kg for rice-based products, calculated from DMA calibration. Statistical analyses used R 3.2.3, ANOVA with Tukey’s range test for feeding-mode comparisons, paired t-test for pre/post-weaning comparisons, natural-log transformed arsenic species, and 1/2 LOD substitution for values below LOD.

Implications

Standards work: This source is direct iAs evidence for baby-cereals-dry-rice-based and rice-snacks-crackers, with HPLC/IC-ICP-MS speciation and explicit comparison to the EU 0.10 mg/kg historical rice-for-infant-food comparator. It also provides exposure-biomarker context showing that rice-product weaning coincided with higher urinary MMA, DMA, and tAs in the paired infant subset.

Courses: Useful case study for explaining why rice-containing infant products require inorganic-arsenic speciation rather than total arsenic alone, and why product labels and use-context matter when rice products not explicitly marketed for infants are nevertheless consumed by infants and young children.

App: Supports a high-priority iAs flag for rice-based infant cereals and rice snacks/crackers when used during weaning. The data are product-as-sold concentrations, paired with infant urinary biomarker evidence.

Microbiome: Not addressed in this study.

Wiki pages this source may touch

• arsenic-inorganic
• rice
• infant-rice-cereal
• rice-cakes
• baby-cereals-dry-rice-based
• rice-snacks-crackers
• eu-2015-1006-iAs-rice
• eu2023-contaminants-maximum-levels
• fda-iAs-rice-cereal-100ppb

Verification notes

• Merge-enhanced 2026-05-18 from the gap-driven auto-fetch PDF. The fetched PDF hash matches the existing raw study PDF hash (686b2a50b1780790c23e21fe27b7d7dbae553e71fd68328bdfceecbbb24dda0c), confirming duplicate identity rather than a new source.
• Replaced the prior approximate concentration ranges with exact Table 1, Table 2, and Table 3 values from the PDF.
• Corrected source scope: 2016 rice-product samples were purchased in Belfast, Northern Ireland, not Spain; the source compares against the EU inorganic-arsenic rice standard.
• Removed the invalid regulations/eu-regulation-2015-1006-iAs-infant-cereal link and created the historical regulation page eu-2015-1006-iAs-rice.
• Strict brand firewall: the source’s Table 3 uses brand-coded groupings, but this page reports only aggregate product-group results and does not reproduce brand-coded rows or brand rankings.
• Formula feeding is represented as urinary biomarker context only. The paper did not measure formula-product arsenic concentrations, so formula products are not routed in frontmatter.

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.