JECFA 2007 - Aluminium PTWI and methylmercury subgroup guidance
JECFA’s 67th report is a primary international committee source for two metal-relevant guidance values. For aluminium, the committee withdrew prior ADIs/PTWI values and established a PTWI of 1 mg/kg bw/week as elemental Al for all aluminium compounds in food, including additives, after concluding that reproductive and developmental nervous-system effects occurred at lower doses than the older 7 mg/kg bw/week PTWI had assumed. For methylmercury, the committee confirmed the 2003 PTWI of 1.6 ug/kg bw/week for women who are or may become pregnant and for children up to about age 17, while clarifying that adults outside those groups could tolerate intakes up to about twice that level without adult neurotoxicity concern. The methylmercury section also concluded that fish maximum-level exclusions alone are often a weak population exposure-reduction tool compared with targeted species advice for higher-risk subgroups.
Key numbers
Aluminium health-based guidance and toxicology basis:
| Parameter | Value | Notes |
|---|---|---|
| New aluminium PTWI | 1 mg/kg bw/week as Al | Applies to all aluminium compounds in food, including food additives |
| Previous aluminium PTWI | 0-7 mg/kg bw/week | Withdrawn by this report |
| Prior temporary ADI for aluminium salts added to food | 0-0.6 mg/kg bw/day as Al | Withdrawn with prior aluminium ADIs |
| Lower end of dietary-study LOEL range | 50 mg/kg bw/day as Al | Used as point of departure |
| Dietary-study LOEL range | 50-75 mg/kg bw/day as Al | From mouse, rat, and dog dietary studies |
| Main uncertainty factor | 100 | Inter- and intraspecies differences |
| Additional uncertainty factor | 3 | Database limitations, including limited NOELs and long-term endpoint data |
| Beagle dog sodium aluminium phosphate basic NOEL | 27 mg/kg bw/day as Al | 702 mg/kg diet, 6-month study |
| Beagle dog high dose with male effects | 75 mg/kg bw/day as Al | 1922 mg/kg diet in males; females at 80 mg/kg bw/day had no effects |
| Reproductive/developmental LOEL range for soluble aluminium compounds | 13-200 mg/kg bw/day as Al | Aluminium nitrate studies; no NOELs identified |
| Neurobehavioral oral-dose effect range | 50-200 mg/kg bw/day as Al | More soluble compounds; one study also indicated 10 mg/kg bw/day NOEL |
| Absorption from most oral aluminium studies | less than 1% | Citrate can increase absorption; silicates/phosphate may reduce absorption |
Aluminium exposure and additive-use context:
| Source or commodity | Value | Notes |
|---|---|---|
| Most foods | less than 5 mg/kg Al | General dietary background statement |
| Natural dietary aluminium exposure | 1-10 mg/day per person | Up to 0.16 mg/kg bw/day for a 60-kg adult |
| Typical untreated water near pH 7 | 1-50 ug/L Al | Acidic water can reach 1000 ug/L |
| Treated drinking-water exposure at 2 L/day | up to 0.4 mg/day | Up to 0.007 mg/kg bw/day |
| Food-contact material migration exposure | up to 7 mg/day | Conservative mean estimate; 0.1 mg/kg bw/day |
| Processed cheese with SALP provision | up to 35,000 mg/kg as Al | Codex GSFA provision summarized in Table 5 |
| Flour with SALP provision | up to 45,000 mg/kg as Al | Codex GSFA provision summarized in Table 5 |
| Bakery products with aluminium ammonium sulfate | up to 10,000 mg/kg | Reporting basis not always specified |
| Salt with sodium/calcium aluminium silicate | up to 20,000 mg/kg | Anti-caking provisions |
| Salt with aluminium silicate | up to 10,000 mg/kg | Anti-caking provision |
| USA processed cheese example | 300 mg/kg Al | Higher-Al processed-food examples submitted to the committee |
| USA home-made cornbread example | 400 mg/kg Al | Attributed to Al-containing leavening agents |
| USA muffins example | 130 mg/kg Al | Higher-Al processed-food examples |
| USA baking powder example | 2300 mg/kg Al | Higher-Al processed-food examples |
| USA table salt example | 164 mg/kg Al | Higher-Al processed-food examples |
| Germany biscuits example | 22 mg/kg Al | Highest processed-food values reported for Germany |
| Germany soft cheese example | 8-16 mg/kg Al | Highest processed-food values reported for Germany |
| UK 2000 total diet miscellaneous cereals | 19 mg/kg Al | Highest mean concentration food group in that survey |
| China 1992-1993 total diet cereal products | 50 mg/kg Al | Attributed to Al-containing leavening agents |
| Adult duplicate-diet mean exposure range | 3-13 mg/day | Highest single reported value: 100 mg/day |
| Adult model-diet / market-basket mean exposure range | 2 to more than 40 mg/day | France at low end; China above 40 mg/day |
| UK young children mean exposure | 0.16 mg/kg bw/day | Ages 1.5-4.5 years |
| USA 2-year-old children mean exposure | about 0.5 mg/kg bw/day | Based on 12-kg body weight |
| China children mean exposure | about 1 mg/kg bw/day | Ages 2-7 and 8-12 years |
| UK young children high-exposure estimate | 0.33 mg/kg bw/day | Ages 1.5-4.5 years |
| Human/cow milk aluminium | less than 0.05 mg/L | Negligible in the committee exposure assessment |
| Cow-milk-based formula ready-to-drink aluminium | 0.01-0.4 mg/L | Reconstituted / ready-to-drink product |
| Soy-based formula ready-to-drink aluminium | 0.4-6 mg/L | Reconstituted / ready-to-drink product |
| Infant exposure, soy-based formula high value | up to 1 mg/kg bw/day | 3-month, 6-kg infant consuming 1 L/day |
| Infant exposure, milk-based formula high value | 0.06 mg/kg bw/day | Same infant-consumption assumption |
| Adult mean exposure from overall diet including additives | 14-280 mg/week | Table 6; 20-500% of PTWI for a 60-kg adult |
Methylmercury guidance and fish-risk-management context:
| Parameter | Value | Notes |
|---|---|---|
| Confirmed methylmercury PTWI | 1.6 ug/kg bw/week | Established in 2003; confirmed here |
| Prior methylmercury PTWI | 3.3 ug/kg bw/week | Clarified as withdrawn in 2003 |
| Maternal hair BMDL/NOEL anchor | 14 ug/g mercury | Faroe Islands and Seychelles neurodevelopmental cohorts |
| Maternal blood equivalent | 56 ug/L methylmercury | Converted from maternal hair concentration |
| Daily intake equivalent before uncertainty factor | 1.5 ug/kg bw/day | Used to derive the PTWI |
| Uncertainty factor for PTWI derivation | 6.4 | Existing 2003 derivation |
| Adults outside pregnancy/childbearing concern | up to about 2x PTWI | Committee considered this not a neurotoxicity risk for adults |
| Women who are or may become pregnant | do not exceed PTWI | To protect embryo/fetus |
| Infants and children up to about age 17 | no higher safe intake identified | Committee could not identify a level above the PTWI without developmental neurotoxicity concern |
| Current Codex guideline level, predatory fish | 1.0 mg/kg methylmercury | Evaluated for impact on exposure and risk |
| Current Codex guideline level, non-predatory fish | 0.5 mg/kg methylmercury | Evaluated for impact on exposure and risk |
| Effect of excluding above-guideline high-Hg species | 30-100% lower mean in those species | France, UK, and USA submissions; required removing most samples of those species |
| French sensitivity noted by committee | 0.5 mg/kg cutoff for all fish | Needed before women exceeding PTWI fell significantly; children 3-10 still not significantly reduced |
Methods (brief)
For aluminium, JECFA began from IPCS and UK COT assessments, reviewed published animal and human evidence, and assessed exposure from natural food content, drinking-water, food-contact materials, food additives, infant formulae, and non-dietary sources. No original toxicological data on aluminium-containing food additives were submitted. The committee considered dietary studies more relevant than gavage studies, used the lower end of the 50-75 mg/kg bw/day dietary LOEL range as the point of departure, applied uncertainty factors of 100 and 3, and expressed the guidance value as a PTWI because of potential bioaccumulation.
For methylmercury, the committee did not recalculate a new PTWI. It reviewed new toxicokinetic, toxicological, epidemiological, infant/child, adult, cardiovascular, and fish-risk-management evidence since the 61st meeting, then clarified which subgroups remained covered by the 1.6 ug/kg bw/week PTWI and how guideline levels for fish affect exposure.
Implications
The aluminium section is routeable as toxicological and regulatory-context evidence for aluminium-containing food additives, baked/cereal products, salt anti-caking uses, processed cheese/cocoa applications, drinking-water, food-contact materials, and infant formula exposure. It should not be treated as a new market survey for HMTc product percentiles, but its commodity examples and exposure ranges can support page context and audit decisions about aluminium as an additive-driven exposure source.
The methylmercury section strengthens the distinction between a sensitive-subgroup PTWI and adult neurotoxicity guidance. It also supports seafood pages by documenting why broad fish maximum levels may have limited population impact when the market is dominated by low-mercury species, while species-specific advice can still matter for consumers who preferentially eat high-mercury fish.
Wiki pages this source may touch
- aluminum
- mercury
- mercury-methyl
- bread
- cornbread
- processed-american-cheese
- cocoa
- salt
- soy-based-infant-formula
- milk-based-infant-formula
- fish
- seafood
- shellfish
- bread-and-baked-goods
- flour-non-rice
- salt
- infant-formula-powder-soy
- infant-formula-powder-dairy
- infant-formula-rtf-liquid-soy-based
- infant-formula-rtf-dairy
- fish-marine-predatory
- fish-marine-non-predatory
- seafood
- shellfish
Verification notes
- Read the full 104-page PDF text, with close checks on the title/citation pages, contents, section 4.1 on aluminium, Table 5 aluminium additive provisions, Table 6 dietary exposure ranges, section 4.3 on methylmercury, section 6 recommendations, and Annex 2 contaminant-summary entries.
- The PDF provides no DOI and carries WHO/FAO all-rights-reserved copyright language. The WHO publication page identifies ISBN 978-92-4-120940-3 and WHO Technical Report Series 940.
WHO_TRS_940_eng 2.pdfis byte-identical to the canonicalWHO_TRS_940_eng.pdffile (sha256782f8a3326e26b4aada1c3837fa90f83f54ba77e1159f6c7e5e0653068a84195) and is tracked as a duplicate rather than a second source.
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.