Heavy Metal Index
Ask the IndexAboutIndex Pro

Chemicals in Cosmetics Used by Washington Residents — Report to the Legislature Pursuant to ESSB 5693 (2022) Section 302 (56)

Cite this pageVersion history

Overview

This is a Washington State Department of Ecology legislative report prepared pursuant to ESSB 5693 (2022) Section 302(56), which appropriated $266,000 for Ecology — in consultation with the Department of Health and community/social-justice organisations — to identify cosmetic products marketed to or used by people of colour (adults and children) and to test those products for potentially harmful chemicals. The report covers Phase 1 of the work: laboratory analysis of 50 low-cost cosmetic products purchased from Puget Sound retail chains in July 2022 (with heavy-metal analysis on 20 of them: 10 lipsticks plus 10 powder foundations) at the Washington Department of Ecology Manchester Environmental Laboratory using EPA Method 6020B (ICP-MS) for lead (Pb)/cadmium (Cd)/arsenic (As) and EPA Method 8270E-SIM for total formaldehyde. The report also reviews company self-reported chemical-use data submitted under the Washington Children’s Safe Products Act (CSPA) and the California Safe Cosmetics Act, and summarises 100+ prior product-testing studies in Appendix C. Phase 2 (ortho-phthalates and asbestos) was anticipated for June 2023 and is not in this report.

Key numbers

Sampling and design

  • n = 50 unique low-cost cosmetic products purchased July 2022 from four Puget Sound retail chains: 33 from three Walmart locations, 4 from a Target, 2 from a Fred Meyer, and 11 from a Dollar Tree (p. 59, Appendix B Methods, Product Collection). No online purchases. Some products purchased in multiples to meet lab weight requirements; full product list with retailer, internal Ecology ID, product name, and category in Table 8 (p. 60-61).
  • Heavy-metal panel (Pb, Cd, As): n = 20 — 10 lipsticks and 10 powder foundations. Foundation selection prioritised darker shades.
  • Formaldehyde panel: n = 40 — 10 hair styling gels/creams, 10 leave-in hair conditioners, 10 body lotions/creams, and the same 10 powder foundations. The 30 hair/lotion products were selected because they listed the formaldehyde-releasing preservative DMDM hydantoin as an ingredient.

Methods (Appendix B p. 62)

  • Sample preparation: EPA Method 3052 (microwave-assisted acid digestion), modified to omit hydrofluoric acid. Analysis: EPA Method 6020B (inductively coupled plasma–mass spectrometry, ICP-MS) for Pb, Cd, and As. All analysis at Manchester Environmental Laboratory (MEL), Washington Department of Ecology.
  • Formaldehyde: EPA Method 8315A–PREP (extraction, performance-based modifications including reduced sample size, addition of surrogates) followed by EPA Method 8270E-SIM for “total formaldehyde” (defined in footnote 20 as free formaldehyde plus formaldehyde released from formaldehyde-releasing preservatives during analysis). All samples extracted in two batches.
  • Heavy-metal QA/QC: Stage 3 data validation. No analytes detected above method reporting limits in two method blanks per batch. Laboratory control sample (LCS) recoveries across all analytes and batches 99–104 %; matrix-spike recoveries 99–106 %; matrix-spike RPDs 1–5 %; LCS RPDs 0.08–2 %. All acceptance criteria met; no data qualifiers added on QA basis. Reporting limit baseline 1 ppm for Pb, Cd, and As; raised on two samples that required smaller digestion volumes to avoid overheating: WM-50-20 lipstick raised to 4.63 ppm (Pb/Cd not detected above that limit), and WM-51-9 cream-to-powder foundation raised to 2.43 ppm (Pb/Cd not detected above that limit).
  • Formaldehyde QA/QC: Level 4 data validation; LCS recoveries 97–103 %; surrogate recoveries within 50–150 % acceptable range. Two samples qualified as estimates: WM-51-9 powder foundation was insoluble in extraction fluid (qualified UJ, not detected at estimated reporting limit) and DT-32-10 leave-in conditioner dissolved in the toluene extraction layer (qualified J, estimate, likely low-biased). One leave-in conditioner (WM-51-3) was leaking during storage and qualified J. Matrix-spike RPDs 7 % and 15 %.

Heavy-metal results — lipsticks (Table 13, p. 67; ppm)

Ten lipsticks tested. Nine of ten lipsticks had Pb, Cd, and As all below the 1 ppm reporting limit (one lipstick required a raised 4.63 ppm reporting limit and was also <reporting limit). One lipstick — a bronze-shade product purchased at Walmart — had detectable lead at 1.08 ppm Pb; cadmium was <1 ppm and arsenic was <0.862 ppm in that sample. Cadmium was not detected in any of the 10 lipsticks. Arsenic was not detected in any of the 10 lipsticks (reporting limit 0.862–0.996 ppm across the set; 4.63 ppm on the one raised-limit sample).

Heavy-metal results — powder foundations (Table 14, p. 68; ppm)

Ten powder foundations tested, with selection biased toward darker shades. Two products had Pb above the 1 ppm reporting limit; one of those also had As above the 1 ppm reporting limit:

  • A dark-shade pressed powder foundation purchased at Walmart: Pb 5.55 ppm and As 2.15 ppm; Cd <1 ppm.
  • A second dark-shade pressed powder foundation purchased at Walmart: Pb 1.52 ppm; Cd <1 ppm, As <0.936 ppm.

The remaining eight foundations had Pb, Cd, and As all below the 1 ppm reporting limit (one required a raised 2.43 ppm reporting limit and was also <reporting limit; another required a 1 ppm reporting limit across all three analytes and was also <reporting limit). Cadmium was not detected in any of the 10 powder foundations.

Heavy-metal summary across n = 20

  • Pb detected above 1 ppm reporting limit in 3 of 20 products (1 lipstick at 1.08 ppm; 2 powder foundations at 1.52 and 5.55 ppm). Maximum observed: 5.55 ppm in a dark-shade powder foundation.
  • As detected above 1 ppm reporting limit in 1 of 20 products (1 powder foundation at 2.15 ppm). The same dark-shade product accounted for both the maximum Pb and the only As detection.
  • Cd detected: 0 of 20 products above the 1 ppm reporting limit.
  • In the ten powder foundations, lead was above the 1 ppm reporting limit in 2/10, arsenic in 1/10, and cadmium in 0/10. Across the twenty cosmetics, arsenic was above 1 ppm in 1/20 and cadmium in 0/20. Formaldehyde was detectable in 7/10 body lotions.
  • All 20 samples below the FDA 2016 recommended industry-maximum guidance of Pb ≤10 ppm in finished cosmetic lip products and externally-applied cosmetics (p. 27). The single Pb 5.55 ppm dark-shade powder foundation would exceed the Germany BVL technically-avoidable-impurity guidance value Pb ≤5.0 mg/kg for makeup powder/rouge/eye shadow/eye liner (BVL 2017; p. 27), and the same product’s As 2.15 ppm would exceed BVL As ≤0.5 mg/kg for general cosmetics and BVL As ≤2.5 mg/kg for theatre/carnival make-up.

Formaldehyde results — hair styling gels and creams (Table 9, p. 64; ppm total formaldehyde)

All 10 hair styling gels and creams had detectable formaldehyde. Range: 254 to 1660 ppm. Maximum 1660 ppm in an extra-hold conditioning styling gel purchased at Walmart. Median ≈ 463 ppm. Of the 10 products, 7 exceeded 400 ppm and all 10 exceeded the 200 ppm SCCS sensitisation threshold (SCCS, 2021).

Formaldehyde results — leave-in hair conditioners (Table 10, p. 65; ppm)

9 of 10 leave-in conditioners had detectable formaldehyde. Range 39.2 (J, estimate, low-biased due to toluene-layer dissolution) to 654 ppm. One product (WM-51-2, a 14-in-1 leave-in conditioner purchased at Walmart) had no detectable formaldehyde at 8.71 uranium (U) despite listing DMDM hydantoin as an ingredient. One children’s-marketed product — a watermelon-scented spray detangler purchased at a discount retailer (DT-32-8) — tested at 214 ppm total formaldehyde, above the 200 ppm SCCS sensitisation threshold.

Formaldehyde results — body lotions and creams (Table 11, p. 65; ppm)

7 of 10 body lotions and creams had detectable formaldehyde. Range 202 to 603 ppm. Three lotions had no detectable formaldehyde (all from the same discount retailer) despite listing DMDM hydantoin as an ingredient. Maximum 603 ppm in a coconut-oil moisturizing cream.

Formaldehyde results — powder foundations (Table 12, p. 66; ppm)

0 of 10 powder foundations had detectable formaldehyde. Reporting limits 8.27–9.59 U across the set; one sample (WM-51-9) was qualified UJ because it was insoluble in the extraction fluid.

Formaldehyde summary across n = 40

  • 26 of 40 tested products (65 %) had detectable total formaldehyde.
  • 24 of the 26 detect products exceeded 200 ppm, the SCCS-cited threshold above which formaldehyde in cosmetics can cause allergic reactions in sensitised individuals.
  • All 26 detects were in hair-styling/leave-in-conditioner/body-lotion products that listed DMDM hydantoin as an ingredient. None of the 10 powder foundations had detectable formaldehyde.

Self-reported chemical-use data — substances reported as intentional ingredients (Table 3, p. 32)

Ecology compiled the Washington Children’s Safe Products Act (CSPA) High Priority Chemicals Database System and the California Safe Cosmetics Program (CDPH) database, downloaded October 2022. Reportable substances likely intentionally added to cosmetics, with the number of companies reporting at least one such product:

  • Synthetic fragrances — 296 companies (most widely-reported substance across all 18 product categories tracked).
  • Ortho-phthalates — 28 companies (reported across 12 product categories including complexion, conditioner, eye products, hair/scalp, lip products, nail products, perfume, shampoo, skin products, wash/cleanser, wipes, and “Products for children: Yes”).
  • Parabens — 22 companies.
  • Cyclosiloxanes — 10 companies.
  • Styrene — 9 companies.
  • Toluene — 6 companies.
  • Formaldehyde and formaldehyde-releasing chemicals — 6 companies (in conditioner, lotion, hair/scalp, lip products, wipes; “Feminine care products: Yes”).
  • Alkylphenol ethoxylates — 4 companies.
  • Alkylphenols — 2 companies.
  • Ethylene glycol — 2 companies.
  • Lead or lead acetate — 1 company (reported in eye products; “Products for children: No”).
  • Bisphenol A — 1 company.
  • PFAS — 1 company.
  • p-phenylenediamine — 1 company (reported in hair color).
  • Asbestos (reported as talc containing asbestiform fibers) — 1 company (in eye products).
  • Coal tar compounds — 2 companies.
  • Ethyl acrylate, benzalkonium chloride — 1 company each.

Self-reported chemical-use data — substances reported as possible contaminants (Table 4, p. 34)

Substances likely present as contaminants rather than intentional ingredients (function reported as “contaminant” or “other”):

  • Formaldehyde and formaldehyde-releasing chemicals — 20 companies (most-reported contaminant; in body art, conditioner, eye products, hair color, hair/scalp, lotion, nail products, shampoo, skin products, wipes; “Children’s products: No”).
  • 1,4-dioxane — 16 companies (in conditioner, eye products, hair/scalp, lotion, shampoo, skin products, wash/cleanser, wipes; “Children’s products: Yes; Feminine care products: Yes”).
  • Cyclosiloxanes — 13 companies.
  • Coal tar compounds — 11 companies (in eye products, lotion, skin products; “Children’s products: No”).
  • Styrene — 9 companies.
  • Toluene — 7 companies.
  • Ethylene glycol — 7 companies (in body art, lotion, perfume, skin products, wash/cleanser; “Children’s products: Yes”).
  • Ortho-phthalates — 6 companies.
  • Lead or lead acetate — 6 companies (in lotion, perfume, skin products, sun protection, shampoo).
  • Arsenic or arsenic compounds — 4 companies (in body art, conditioner, eye products, hair/scalp, lip products, nail products, skin products, wash/cleanser; “Children’s products: Yes”).
  • Cadmium or cadmium compounds — 3 companies (in body art, eye products, hair/scalp, lotion, skin products, sun protection; “Children’s products: Yes”).
  • Asbestos — 2 companies (in complexion products, eye products, lip products).
  • Mercury or mercury compounds — 2 companies (in skin products, sun protection; “Children’s products: Yes”).
  • Bisphenol A — 1 company.
  • 2-ethylhexyl acrylate — 2 companies.
  • PFAS — 2 companies (in eye products, hair/scalp, lotion; “Children’s products: No”).
  • Ethyl acrylate — 4 companies.
  • o-phenylenediamine — 1 company.
  • Synthetic fragrances — 18 companies.

Synthetic fragrances dominated the intentional-ingredient list (296 companies); formaldehyde dominated the contaminant list (20 companies). Companies are not required to report on m-phenylenediamine, aluminium salts, sodium laurel/laureth sulfate, methylisothiazolinone, methylchloroisothiazolinone, triclosan, or hydroquinone, so absences here do not mean absence in market.

Regulatory anchors and policy context cited

  • FDA 2016 industry guidance — recommended maximum Pb ≤10 ppm in finished cosmetic lip products and externally-applied cosmetics (based on 2007–2013 FDA testing showing 99 % of products tested contained ≤10 ppm Pb).
  • EU Cosmetics Regulation (EC) 1223/2009 — heavy metals in EU-marketed cosmetics permitted only at concentrations unavoidable under good manufacturing practice.
  • Germany BVL 2017 “technically-avoidable” non-regulatory guidance values for impurity Pb/Cd/As/mercury (Hg)/Sb in cosmetics (decreased in 2017; set such that 90 % of tested cosmetics are expected to pass):
    • As ≤2.5 ppm in theatre/fan/carnival make-up; ≤0.5 ppm in all other cosmetics.
    • Cd ≤0.1 ppm in all cosmetics.
    • Pb ≤5.0 ppm in makeup powder, rouge, eye shadow, eye liner; ≤2.0 ppm for all other cosmetics.
  • Washington State Children’s Safe Products Act (CSPA) — manufacturers of children’s products sold in WA must report any product containing one or more Chemicals of High Concern to Children (Ecology, 2017).
  • California Safe Cosmetics Act + Cosmetic Fragrance and Flavor Ingredient Right to Know Act — cosmetics manufacturers must report products containing reportable harmful ingredients (including fragrance ingredients as of January 2022) to CDPH (CDPH, 2022; CA HSC 111791.5).
  • WA ESSB 5703 (2022) — the underlying state legislative authority for the Safer Products for Washington program from which the SB 5703 chemical list is drawn (Appendix A summarises hazard endpoints for the SB 5703 chemicals).
  • Washington ESSB 5693 (2022) Section 302(56) — the legislative directive funding the Phase 1 testing reported here.

Evidence Fitness

This is an A-tier state-government laboratory product-testing report executed by an EPA-recognised laboratory (Manchester Environmental Laboratory, MEL) under a published Quality Assurance Project Plan (Ecology, 2022a), using EPA Methods 3052 and 6020B for metals and EPA Methods 8315A-PREP and 8270E-SIM for total formaldehyde. Data validation was Stage 3 for metals and Level 4 for formaldehyde. All quality control criteria were met; data are usable for all purposes as reported. Evidence the report supports:

  • Direct-evidence occurrence data for Pb, Cd, and total As (no inorganic-As speciation) in 10 lipsticks and 10 powder foundations from low-cost Puget Sound retail channels in July 2022, with darker foundation shades over-sampled. The 5.55 ppm Pb plus 2.15 ppm As co-occurrence in one dark-shade powder foundation is the principal heavy-metal signal in the dataset.
  • Direct-evidence occurrence data for total formaldehyde (free + DMDM-hydantoin-released) in 10 hair styling gels/creams, 10 leave-in hair conditioners, 10 body lotions/creams, and 10 powder foundations; including a children’s-marketed leave-in spray detangler at 214 ppm.
  • Reproducible methods documentation (sample preparation per ESSB 5693 Phase 1 QAPP, EPA Method 3052 digestion modified to omit HF, EPA Method 6020B ICP-MS analysis at MEL; EPA 8315A-PREP and 8270E-SIM for total formaldehyde).
  • Compiled snapshot (October 2022) of company self-reported intentional-ingredient and contaminant chemical-use data for cosmetics sold in Washington and California, indexed by product category — useful as a coarse signal of which chemical/product-category pairs the cosmetics industry itself reports as plausibly present.
  • Comprehensive cross-reference of the regulatory landscape (FDA 2016 guidance, EU 1223/2009, Germany BVL technically-avoidable limits, Washington CSPA, California Safe Cosmetics Act) as of January 2023.

The report does NOT support:

  • Percentile-distribution claims by product category (n = 10 per heavy-metal subcategory is below the threshold for distribution-shape claims; OPSS 2023 makes the same n-too-small caveat for its 9/10 per group).
  • Inorganic-vs-total speciation of arsenic (the EPA 6020B ICP-MS method reports total elemental As only).
  • Methylmercury-vs-total speciation of mercury (Hg was NOT analysed in this Phase 1; the SB 5703 substance list includes mercury but no mercury panel was completed; Phase 2 may add this — not in this report).
  • Asbestos or ortho-phthalate occurrence (Phase 2 deliverables, anticipated June 2023, NOT included in this Phase 1 report).
  • Cancer-risk or cumulative-exposure modelling — the report describes “compounded burdens” qualitatively but does not produce numerical exposure estimates.

Methods (brief)

Product collection (Appendix B, Methods p. 59). In July 2022, Ecology staff purchased 50 unique low-cost cosmetic products from four Puget Sound area retail chain stores (Walmart 33, Target 4, Fred Meyer 2, Dollar Tree 11), guided by selection criteria in the project Quality Assurance Project Plan (Ecology, 2022a) Appendix A. Some products purchased in multiples to meet lab weight requirements. No online purchases.

Sample processing (Appendix B p. 61). Lipsticks and powder foundations (solid/semi-solid) were processed into sample jars the day before courier transport to Manchester Environmental Laboratory (MEL). Hair-care and skin-lotion products were sent in original sealed bottles to minimise atmospheric exposure. Seal-intact / seal-unknown / no-seal status was documented. MEL received samples in good condition per chain-of-custody (COC) records.

Laboratory procedures (Appendix B p. 62). Heavy metals: sample prep per EPA Method 3052 (microwave-assisted acid digestion), modified to omit hydrofluoric acid. Analysis per EPA Method 6020B (ICP-MS). Formaldehyde: extraction per a performance-based modification of EPA Method 8315A (“EPA 8315A-PREP,” with reduced sample size, reduced extraction-chemical volumes, and surrogate addition for extraction-efficiency monitoring); analysis per EPA Method 8270E-SIM (single-ion-monitoring GC-MS).

Speciation note. EPA Method 6020B determines total elemental Pb, Cd, and total As; this paper does NOT separate inorganic vs total As (recorded as total arsenic (tAs) in this page’s metals: field). Mercury was not in the Phase 1 panel.

Data validation (Appendix B p. 62-63). Stage 3 validation for metals (Stage 4 was requested in the QAPP but is not customarily performed for the review components used for a metals dataset). Level 4 validation for formaldehyde. For metals: no analytes detected above method reporting limits in two method blanks per batch; LCS recoveries 99–104 %; matrix-spike recoveries 99–106 %; matrix-spike RPDs 1–5 %; LCS RPDs 0.08–2 %. For formaldehyde: LCS recoveries 97–103 %; surrogate recoveries within 50–150 % acceptable range; two batch-internal matrix-spike RPDs of 7 % and 15 %. Reporting-limit qualifications: WM-50-20 lipstick raised to 4.63 ppm for all three metals; WM-51-9 cream-to-powder foundation raised to 2.43 ppm for all three metals; WM-51-9 powder foundation qualified UJ for formaldehyde (insoluble in extraction fluid); DT-32-10 leave-in conditioner qualified J for formaldehyde (toluene-layer dissolution, likely low bias); WM-51-3 leave-in conditioner qualified J for formaldehyde (leaking during storage).

Other research review (Methods p. 19-26). Beyond the laboratory work, the report reviewed (a) the Washington CSPA High Priority Chemicals Database System and the Oregon Toxic Free Kids Act database (both via the Interstate Chemicals Clearinghouse, IC2), downloaded October 2022 — covering the most-recent three years of company reports; (b) the California CDPH cosmetic ingredient database, downloaded October 2022; and (c) a literature review of prior product-testing studies (summarised in Table 15, Appendix C, p. 69–71).

Implications

  • Certification. This report is the highest-quality publicly-available state-government Pb/As occurrence dataset for low-cost adult-market cosmetic powder foundations and lipsticks at the Puget Sound retail tier as of 2023. Direct-evidence Pb 5.55 ppm + As 2.15 ppm co-occurrence in a single dark-shade pressed powder foundation contributes occurrence data for HMTc Cat 13 Row 5 (makeup-foundation-powders-blush). The lipstick data (max Pb 1.08 ppm; 1/10 detect) contributes to a future Cat 13 lip-products row when scaffolded (no current Cat 13 lip-products row is locked; adult lipstick is reserved per the retirement note on products/lipstick.md). The formaldehyde dataset contributes preservative-release occurrence data for Cat 13 Row 4 (body-hand-leave-on-skin-care) and Row 11 (leave-on-hair-preparations).
  • Courses. Useful as a US-state regulatory-program teaching example: the legislative-directive → QAPP → MEL-laboratory → published-report chain documented in this paper is a clean illustration of how a state agency translates a budget appropriation into reproducible product-testing evidence under EPA methods. Also useful as a cosmetics-regulatory-literacy reference: the gap between the FDA 2016 industry-recommended Pb ≤10 ppm and the Germany BVL technically-avoidable Pb ≤5.0 ppm (rouge/eye shadow/eye liner/powder) or Pb ≤2.0 ppm (general cosmetics) shows that “cosmetic Pb limit” varies five-fold across major regulatory jurisdictions.
  • App. Not directly relevant to the food-app contamination-profile workflow; this is a cosmetics paper.

Update history

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

Referenced by