Opinion of the French Agency for Food, Environmental and Occupational Health & Safety on the safety of baby diapers (revised opinion, Request No 2017-SA-0019)

Overview

This revised ANSES opinion is the French food, environmental and occupational health agency’s collective expert appraisal of the safety of single-use baby diapers, prompted by January 2017 popular-press reporting of pesticides, dioxins, furans, PAHs and VOCs in diapers (60 Millions de Consommateurs, 2017). The CES on “Assessment of chemical risks of consumer items and products” conducted a quantitative health risk assessment (QHRA) of chemical exposures via dermal contact for children aged 0-36 months, using analytical results from 23 diaper products tested by SCL (DGCCRF Joint Laboratory Service) and INC (French National Consumer Institute) in 2016, 2017 and 2018. The QHRA concluded that a health risk could not be ruled out for two fragrances (Lyral® / hydroxyisohexyl 3-cyclohexene carboxaldehyde and Lilial® / butylphenyl methylpropional), certain PAHs, PCB-126, and the sums of dioxins, furans and DL-PCBs. Heavy metals were among the substance classes screened for by the INC, but they are not listed among the substances quantified or detected at least once in single-use baby diapers sold in France, are not included in the QHRA, and are not the subject of any of the opinion’s recommendations. This page is ingested as out-of-core-scope regulatory context for the diapers and components product page — it carries the agency-level null finding for heavy metals in French-market diapers and the population-frame exposure parameters (diaper weight, diaper changes per day, body weight, dermal absorption assumptions) that may be reusable for HMI’s future heavy-metals dermal exposure modelling in this product category, alongside the parallel non-metal findings (PAHs, dioxins/furans/DL-PCBs, fragrances) that contextualise the broader chemical-contamination picture in the same product class.

Key numbers

Substance screening panels (p. 4)

• INC screening panel (3 published test sets, 2016/2017/2018; published in 60 Millions de Consommateurs): pesticides, PAHs, dioxins and furans, fragrances and volatile organic compounds (VOCs), heavy metals, nonylphenol, octylphenol, nonylphenol monoethoxylates.
• SCL screening panel (DGCCRF-commissioned, 2017 and 2018): pesticides, PAHs, dioxins, furans and DL-PCBs (“dioxin-like” polychlorinated biphenyls), phthalates, organotins, VOCs, fragrances, azoic dyes.
• Number of products tested: 23 single-use baby diaper products sold on the French market, including best-sellers, retailers’ own-brand products, and “eco-friendly” products.

Substances quantified or detected at least once (p. 4-5)

In shredded whole diapers (solvent extraction):

• VOCs: naphthalene, styrene, toluene, dichlorobenzenes, p-isopropyltoluene, xylenes, chlorobenzene
• Pesticides: hexachlorobenzene, quintozene and its metabolite pentachloroaniline, glyphosate and its metabolite AMPA
• Formaldehyde
• Dioxins, furans, DL-PCBs
• Fragrances: benzyl alcohol, benzyl salicylate, coumarin, hydroxyisohexyl 3-cyclohexene carboxaldehyde (Lyral®), butylphenyl methylpropional (Lilial®), limonene, linalool, alpha-isomethyl ionone

In shredded diaper parts (defined as a component considered separately — elastic bands, inner layer, absorbent pad, etc.; footnote 3, p. 5):

• Dioxins, furans: in outer layer, inner layer and other parts (not the absorbent core)
• PAHs in the elastics: benzo[b]fluoranthene, benzo[a]anthracene, indeno[1,2,3-c,d]pyrene, benzo[g,h,i]perylene

In migration tests with whole and shredded whole diapers using a urine simulant (Colon et al. 2015 simulant composition: urea, creatinine, ammonium citrate, NaCl, KCl, KHSO₄, MgSO₄, KH₂PO₄, KHCO₃ in water; footnote 4, p. 5):

• Dioxins, furans, DL-PCBs
• PAHs
• Formaldehyde

Heavy metals: not listed among substances quantified or detected at least once. Although INC’s screening panel included heavy metals, no metal is named in the detected/quantified inventory on p. 4-5, no metal appears in the QHRA hazard-identification, dose-response, or risk-characterisation sections (p. 5-12), and no metal is the subject of any of the agency’s recommendations (p. 12-14). For dioxins, furans, and DL-PCBs, the opinion proposes a maximum near the limit of quantification, and names the lowest LQ used in the appraisal as around 0.02 ng/kg. That figure is not a health threshold.

Exposure parameters used in the QHRA (Table 1, p. 9)

Two scenarios (worst-case and refined), six age strata (0-6, 6-12, 13-18, 19-24, 25-30, 31-36 months):

The most realistic scenario is scenario 2.2 (whole-diaper extraction with urine simulant), because the capacity to extract substances from diaper to urine is measured rather than modelled (no skin-transfer T modelling needed) and only substances actually appearing in pressed urine output are quantified (no reflux ratio R needed).

QHRA findings — Scenario 2.2 (whole diaper + urine simulant, SCL 2018) (Annex 3, p. 20-28)

This is the most realistic scenario per the CES (p. 8). HQ > 1 (threshold effects) or IER > 10⁻⁶ (no-threshold carcinogenic effects) findings for any 0-36 month age class:

• Benzo[a]pyrene: HQ 25.1-66.3 (all ages exceed); IER 1.42e-3 (0-6 mo), 1.82e-3 (6-12), 2.30e-3 (13-18), 2.99e-3 (19-24, maximum), 2.91e-3 (25-30), 2.47e-3 (31-36) — all ages above the 10⁻⁶ acceptable-risk threshold (p. 23)
• Dibenzo[a,h]anthracene: HQ 19.3-51 (all ages exceed); IER 1.09e-3 to 2.30e-3 (all ages above 10⁻⁶) (p. 23)
• Benzo[k]fluoranthene: HQ 2.28-6.03 (all ages exceed); IER 1.29e-4 to 2.72e-4 (all ages above 10⁻⁶) (p. 22)
• Benzo[j]fluoranthene: HQ 2.28-6.03 (all ages exceed); IER 1.29e-4 to 2.72e-4 (p. 22)
• Benzo[b]fluoranthene: HQ 2.36-6.24 (all ages exceed); IER 1.34e-4 to 2.82e-4 (p. 21)
• Cyclopenta[c,d]pyrene: HQ 1.93-5.10 (all ages exceed); IER 1.09e-4 to 2.30e-4 (p. 21)
• Chrysene: HQ 20.22 for 13-18 months only (other ages 0.15-0.41, below 1); IER 8.75e-6 to 1.84e-5 across all ages (above 10⁻⁶) (p. 22)
• PCB-126: HQ 1.57-4.16 (all ages exceed) (p. 27)
• Sum of DL-PCBs (quantified): HQ 1.69-4.46 (all ages exceed) (p. 28)
• Sum of dioxins + furans + DL-PCBs: HQ 1.73-4.58 (all ages exceed) (p. 28)
• Benzo[g,h,i]perylene, 5-methylchrysene, benzo[e]pyrene — Scenario 2.2 HQ < 1 across all ages but IER > 10⁻⁶ for at least one age (per text p. 11 listing the 10 PAHs with IER exceedances).

Scenario 2.2 substances with HQ < 1 and IER ≤ 10⁻⁶ across all ages (no exceedance under the realistic scenario):

• Formaldehyde: HQ 0.34-0.9 across age strata (p. 20).
• Sum of dioxins and furans (quantified): HQ 0.23-0.62 (p. 27).
• Specific dioxin/furan congeners measured in Scenario 2.2 (1,2,3,6,7,8-HxCDD, 1,2,3,4,6,7,8-HpCDD, 2,3,7,8-TCDF, 2,3,4,7,8-PeCDF, 1,2,3,4,7,8-HxCDF, 1,2,3,4,6,7,8-HpCDF, OCDF, 2,3,4,6,7,8-HxCDF) all HQ in the 10⁻³ to 10⁻⁴ range (p. 26-27).

QHRA findings — Scenario 1 (solvent extraction from shredded whole diapers, INC 2017/2018 + SCL 2017) (Annex 3)

Solvent extraction is a worst-case-of-bioavailability scenario; the CES treats Scenario 2.2 as more realistic. Scenario 1 findings flagged:

• Hydroxyisohexyl 3-cyclohexene carboxaldehyde (Lyral®): MOEref/MOE 1.72 (0-6 mo, red); 1.10 (6-12 mo, red); 0.93 (13-18 mo); 0.91 (19-24 mo); 0.82 (25-30 mo); 0.65 (31-36 mo) (p. 25). Two youngest age classes exceed the reference margin.
• Butylphenyl methylpropional (Lilial®): MOEref/MOE 1.72 (0-6 mo, red); 1.10 (6-12 mo, red); 0.93 (13-18 mo); 0.91 (19-24 mo); 0.82 (25-30 mo); 0.65 (31-36 mo) (p. 25). Same exceedance pattern as Lyral®.
• Hexachlorobenzene (pesticide): HQ 3.71e-2 to 9.82e-2 across age strata (all below 1); IER 8.84e-8 (0-6 mo, green) through 2.00e-7 (31-36 mo, orange) — orange in older age classes (between 10⁻⁷ and 10⁻⁶) (p. 20).
• Sum of DL-PCBs (Scenario 1): HQ 8.05e-2 to 0.21 across age strata (0-6 mo orange) (p. 28).
• Sum of dioxins + furans + DL-PCBs (Scenario 1): HQ 0.11 to 0.29 (p. 28).
• Benzo[a]pyrene Scenario 1 (diaper parts shredded, elastics): IER 6.14e-11 to 1.07e-9 (all green) (p. 21).
• VOCs: 1,2,3-trichlorobenzene HQ 4.06e-2 to 0.11 (0-6 mo orange, others green); 1,2,4-trichlorobenzene HQ 9.00e-3 to 2.38e-2 (green) with IER 6.13e-8 to 1.46e-7 (orange in older age classes); 1,3,5-trimethylbenzene HQ 1.56e-2 to 4.13e-2 (all green) (p. 23). None exceed HQ = 1.
• Other fragrances (Scenario 1): Coumarin HQ 0.33-0.86 (all orange); Limonen [sic, Limonene] HQ 0.33-0.86 (all orange); Lyral/Lilial above; alpha-isomethyl ionone MOEref/MOE 6.49e-3 to 0.17 (green to orange); benzyl alcohol HQ 6.49e-3 to 1.72e-2 (green); benzyl salicylate MOEref/MOE 6.49e-3 to 0.17 (orange in 0-6 / 6-12 mo); linalol [sic, linalool] HQ 6.49e-3 to 1.72e-2 (green) (p. 24-25).
• Formaldehyde (Scenario 1): HQ 9.71e-3 to 2.57e-2 across age strata (all green, no exceedance) (p. 20). The much higher Scenario 2.2 HQ (0.34-0.9) is reported in the Scenario 2.2 section above.

QHRA findings — substances explicitly stated to have no exceedance (p. 13, Agency conclusions)

ANSES’s conclusion explicitly lists substances where no exceedance was found across the QHRA:

• Other fragrances (benzyl alcohol, coumarin, limonene, linalool, alpha-isomethyl ionone — none of these substances triggered HQ > 1; coumarin and limonene have orange-band HQs in Scenario 1 for older age classes but none above 1)
• Prohibited pesticides (the majority found are EU-prohibited: lindane and quintozene since 2000, hexachlorobenzene since 2004)
• Glyphosate or its metabolite AMPA
• VOCs (none exceed HQ = 1)
• Formaldehyde (no exceedance — but see Scenario 2.2 HQ 0.34-0.9, orange in younger age classes)

Population frame (p. 3)

• 90% of EU families have used single-use diapers since the 1990s (EDANA 2011)

• 95% of French babies have worn disposable diapers for ~20 years (Group’Hygiène 2015)

• Estimated 3800-4800 disposable diapers used per child before toilet training (varies with toilet-training age 2.5-3 years)
• Diaper-dermatitis prevalence 7-50% (varies with country and hygiene practices), most commonly between 9 and 12 months of age

Methods (brief)

• Expert appraisal framework. French Standard NF X 50-110 “Quality in Expert Appraisals — General Requirements of Competence” (May 2003). Conducted by the CES on “Assessment of chemical risks of consumer items and products” (CES 1, May 2016 - August 2017; subsequently CES 2). Methodological and scientific aspects presented to the CESs between May 2016 and November 2018; collective expert appraisal adopted at the CES meeting of 15 November 2018. Director General sign-off by Dr Roger Genet, 17 January 2019. Conflict-of-interest declarations public on the ANSES website.
• Stakeholder hearings (April-May 2017). UFC (Federal Union of Consumers), companies and trade federations (Love & Green, Procter & Gamble, FCD = Federation of Trade and Retail Companies, SNITEM = National Association of the Medical Technology Industry, Group’Hygiène = French grouping of manufacturers of single-use products for hygiene/health/wiping, EDANA = European Disposables And Nonwovens Association), and the public body French National Consumer Institute (INC).
• Test data sources. INC tests 2016, 2017, 2018 (published in 60 Millions de Consommateurs plus confidential supplements); SCL tests 2017 and 2018 commissioned by DGCCRF; Group’Hygiène 2018 confidential tests. 23 diaper products including best-sellers, retailers’ own-brand, and “eco-friendly”. Three test protocols: (i) solvent extraction from shredded whole diapers or shredded diaper parts; (ii) extraction with urine simulant from shredded whole diapers (SCL 2017); (iii) extraction with urine simulant from whole diapers (SCL 2018).
• Urine simulant composition (Colon et al. 2015; footnote 4, p. 5). Urea, creatinine, ammonium citrate, NaCl, KCl, KHSO₄, MgSO₄, KH₂PO₄, KHCO₃ in water.
• QHRA framework. NRC 1983 four-step framework (hazard identification, dose-response, exposure assessment, risk characterisation). Hazard identification consulted CLP Regulation harmonised classifications (EC No 1272/2008), IARC carcinogenicity classifications, and endocrine-disruption databases (European Commission BKH 2000, 2002; DHI 2007; US EPA; Illinois EPA; TEDX; SIN List). TRVs prioritised by chronic-duration, repeated and/or long-term low-to-moderate-dose criteria, focused initially on dermal route but extended to oral route where dermal TRVs were unavailable (no dermal TRVs were found for any substance in this appraisal). For PAHs and dioxins/furans/DL-PCBs, only TRVs for the reference congener (benzo[a]pyrene for PAHs; 2,3,7,8-TCDD for dioxins/furans/DL-PCBs) were identified; other congeners assessed by toxic equivalency factors (TEFs).
• Daily exposure dose (DED) formulae (p. 7-8).
  • Scenario 1 (solvent extraction): DED = (C_shredded-material × W × F × T × Abs) / BW
  • Scenario 2.1 (urine simulant in shredded diaper): DED = (C_shredded-material-simulant × W × F × R × Abs) / BW
  • Scenario 2.2 (urine simulant in whole diaper): DED = (C_diaper-simulant × W × F × Abs) / BW
• Risk characterisation. HQ = DED / TRV (threshold effects, traffic-light thresholds < 0.1, 0.1-1, > 1); IER for no-threshold (carcinogenic) effects (< 10⁻⁷, 10⁻⁷-10⁻⁶, > 10⁻⁶ acceptable risk threshold); MOE = critical dose / DED where no TRV available.
• Worst-case-first approach. Worst-case scenario applied to all substances first; refined scenario applied only where worst-case exceeded TRV.

Implications

• Certification (HMTc): This opinion provides no heavy-metal occurrence data and contributes nothing directly to the HMTc 10-analyte panel (Pb, total arsenic (tAs), cadmium (Cd), methylmercury (MeHg), total mercury (tHg), inorganic arsenic (iAs), nickel (Ni), aluminum (Al), hexavalent chromium (Cr-VI), Sn). For Category 9 Row 7 (diapers and diaper components), it is one of the rare national-agency-level data points and carries the agency-level null finding for heavy metals in French-market diapers — useful as regulatory-context evidence that a competent national authority screened heavy metals in this product category and did not surface any metal as a risk-relevant finding requiring further QHRA. The opinion’s population-frame parameters (24-45 g per diaper across age strata; 4.7-7.98 diaper changes per day across age strata; 3.9-11.4 kg body weight across age strata; 7% skin transfer per Odio et al. 2000; 1.32% reflux ratio per Dey et al. 2016; 100% dermal absorption retained because diaper dermatitis cannot be excluded) are directly reusable in any future HMI dermal-exposure model for metals in disposable diapers — the same DED framework applies regardless of analyte class.
• Courses: Useful as a reference for any HMTc Cat 9 module covering national-agency regulatory positioning on diaper safety, the structure of a NRC-1983-style QHRA applied to a dermal-contact infant product, and the worst-case → refined two-stage approach to triaging substances by exceedance.
• App: Not directly applicable to the heavy-metals consumer app. Useful as background for a future app explainer on what the French national agency has and has not assessed in the diaper category.
• Microbiome: Not addressed.

Related evidence

• Diapers and diaper components
• Occurrence of tin in disposable baby diapers
• Determination of Selected Harmful Substances in Baby Diapers Available on the South African Market
• Infants' Dermal Exposure to Phthalates from Disposable Baby Diapers and Its Association with DNA Oxidative Damage

Update history

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