Kicińska & Kowalczyk 2025 — Heavy metals (Cr, Fe, Mn, Ni, Zn) in online-market cosmetics with QRA-based health risk
This Scientific Reports paper measured Cr (total), Fe, Mn, Ni, Zn by ICP-MS in 23 cosmetic products (foundations, blushes, lipsticks, face creams, face masks, eye shadows) purchased from the online consumer market and computed margin of safety (MoS), hazard quotient (HQ), hazard index (HI), lifetime cancer risk (LCR) and quantitative risk assessment (QRA) per US EPA guidelines. Mean concentrations ranked Fe (6,407) > Zn (190) > Mn (88) > Cr (24) > Ni (8) mg/kg across the 23-sample set. The authors concluded that the calculated MoS, HQ and HI values for all cosmetic samples indicated the products do not present a risk to human health under modeled use, but reported elevated maxima for Cr (416 mg/kg in a blush) and Ni (95 mg/kg in the same blush) that exceed values reported in literature for those product categories.
Key numbers
- Per-product-category ranges and mean ± SD (Table 2, p. 5; mg/kg):
- Foundation (n=4): Cr 2.11-4.52 (mean 3.32 ± 0.8); Fe 5137-8046 (6636 ± 1271); Mn 4.2-10.6 (7.0 ± 2.4); Ni 1.62-2.77 (2.21 ± 0.42); Zn 0.1-927.7 (234.3 ± 400.4).
- Face blush (n=4): Cr 7.18-416.29 (110.37 ± 176.6); Fe 538-16,563 (8941 ± 5681); Mn 24.9-40.4 (29.3 ± 6.4); Ni 3.79-95.66 (27.33 ± 39.44); Zn 1.7-13.7 (5.1 ± 4.9).
- Lipstick (n=4): Cr 5.13-7.26 (5.86 ± 0.8); Fe 30-5280 (1391 ± 2246); Mn 0.5-7.8 (3.0 ± 2.8); Ni 3.12-4.34 (3.73 ± 0.43); Zn 2.1-2.9 (2.6 ± 0.3).
- Face mask (n=4): Cr 1.73-34.86 (11.99 ± 13.4); Fe 15-18,758 (5136 ± 7896); Mn 0.2-80.2 (21.3 ± 34.0); Ni 0.91-24.62 (7.33 ± 10.00); Zn 2.3-29.4 (11.0 ± 11.0).
- Face cream (n=4): Cr 3.58-5.03 (4.39 ± 0.5); Fe 16-22 (19 ± 2); Mn 0.2-0.9 (0.5 ± 0.3); Ni 2.15-2.75 (2.48 ± 0.22); Zn 0.9-1.5 (1.2 ± 0.2).
- Eye shadow (n=3): Cr 6.29-9.96 (8.03 ± 3.5); Fe 6052-24,882 (16,318 ± 9763); Mn 86.4-1211.9 (465.4 ± 499.6); Ni 3.48-7.97 (5.43 ± 2.78); Zn 5.3-2649.0 (887.2 ± 1145.2).
- Pooled-mean ranking across all n=23 (Results, p. 5; Abstract): Fe 6407 > Zn 190 > Mn 88 > Cr 24 > Ni 8 mg/kg.
- Highest individual values (Results, p. 5; Fig. 1): Cr 416 mg/kg in a blush (B/Mb, Polish manufacturer); Ni 95 mg/kg in the same B/Mb blush; Fe 24,882 mg/kg in an eye shadow (S/MR, USA manufacturer); Mn 1212 mg/kg in eye shadow (S/E, German manufacturer); Zn 2649 mg/kg in eye shadow (S/E, German manufacturer).
- Reference doses used in HQ/MoS calculations (Eq. 3, p. 4): Cr-III 1.5; Cr-VI 0.3; Fe 70; Mn 14; Ni 2; Zn 30 mg/kg per day (RfD); uncertainty factor (UF) 100; modifying factor (MF) 1.
- Cancer slope factor used for LCR (Eq. 6, p. 4): Cr 0.5; Ni 0.91 (mg/kg day)⁻¹. HT25 for Cr-VI and Ni: 0.04 and 0.047 mg/kg per day (Eq. 7, p. 4).
- Bioavailability scenarios modeled: BF = 50% and 100%; body weight 60 kg (adult); retention factor 1 for leave-on / 0.1 for masks; application frequency 2/day for foundations, creams, eye shadows, blushes / 1/day for masks; quantity applied 0.057 g/cm² (lipsticks), 0.51 (foundations, creams, masks), 0.02 (eye shadows, blushes); surface areas 4.8 cm² (lipsticks), 563 (foundations, creams, masks), 24 (eye shadows), 24 (blushes).
- SED ranges (Table 4, p. 7; µg/kg BW per day): BF = 100%: Cr 5.34E-05 to 4.20E-02; Fe 1.27E-02 to 6.35E+01; Mn 1.70E-05 to 2.38E-02; Ni 2.77E-05 to 6.63E-02; Zn 2.38E-05 to 2.24E+00. The estimated SED for Fe exceeded its recommended daily intake reference (12.5-18 mg/day); SEDs for Cr, Ni, Mn, Zn were below TDI/recommended-daily-intake references.
- MoS values (Table 5, p. 8; BF=100%; selected): Foundation Cr 9.41E+03, Fe 1.10E+03, Mn 2.10E+05, Ni 9.46E+04, Zn 1.34E+04. All MoS values for all groups exceeded the WHO MoS ≥ 100 safety threshold.
- HQ values (Table 6, p. 8; BF=100%; “Own study”): Highest HQ in foundations — Cr 1.06E-02, Fe 9.07E-02, Mn 4.74E-04, Ni 1.04E-03, Zn 7.48E-03. All HQ values < 1.
- HI values (Table 7, p. 9; BF=100%; “Own study”): Foundation 1.10E-01; Face blushes 8.18E-04; Lipsticks 3.79E-05; Face creams 1.55E-02; Face masks 5.69E-03; Eyeshadow 5.21E-04. All HI < 1.
- LCR values for Cr-VI and Ni (Table 8, p. 9; BF=100%; “Own study”): Cr-VI Foundation 6.37E-05, Face blushes 3.53E-06, Lipsticks 1.07E-07, Face creams 1.15E-05, Face masks 8.40E-05, Eyeshadow 2.57E-07; Ni Foundation 2.32E-05, Face blushes 4.80E-07, Lipsticks 3.74E-08, Face creams 3.85E-05, Face masks 2.61E-05, Eyeshadow 9.55E-08. All within acceptable range 1.0E-06 to 1.0E-04.
- QRA for nickel (Table 12, p. 10; SAF 300, NESIL 401.07 µg/cm²/day, AEL 1.34 µg/cm²): AEL/CEL ratios in “Own study” — Foundation 549.2, Lipsticks 0.07, Face creams 478.6, Eyeshadow 0.12. Authors concluded (p. 10) that use of these products does not pose a risk of skin sensitization; their narrative states AEL > CEL “for all” foundations, creams, eye shadows and lipsticks, which is consistent with Table 12 for foundations and face creams (ratios 549.2 and 478.6) but inconsistent with Table 12 for lipsticks (0.07) and eyeshadows (0.12) where AEL < CEL — see Verification notes.
- EDI/TMDI for lipsticks (Table 10, p. 10; daily lipstick intake assumed 57 mg/kg BW/day): EDI Cr 3.34E-04; Fe 7.93E-02; Mn 1.73E-04; Ni 1.92E-04; Zn 1.49E-04 mg/kg/day. EDIs substantially lower than the corresponding TMDI values.
- Correlation coefficients (Table 3, p. 7): almost-certain correlation (0.9 ≤ r < 1.0) for Cr-Ni (r=0.96) and Mn-Zn (r=0.93), interpreted by the authors as a common mineral source of origin; moderate (paper-defined as 0.3 ≤ r < 0.5) for Fe-Cr (r=0.35) and Fe-Ni (r=0.44). Table 3 thresholds: 0.5 ≤ r < 0.7 high; 0.7 ≤ r < 0.9 very high; 0.9 ≤ r < 1.0 almost-certain.
Methods (brief)
23 cosmetic samples purchased online from various websites (4 foundations, 4 blushes, 4 lipsticks, 4 face creams, 4 face masks, 3 eye shadows; manufacturers in Czech Republic, USA, France, Germany, Poland, Great Britain, Korea, Russia). Samples dried at 105 °C for 2 h to constant weight. Single-step acid mineralization with HCl:HNO₃ (1:3 v/v, aqua regia) at 1:10 solid:solution ratio for 2 h at 130 °C in microwave (SCP Science DigiPREP HT). Pseudo-total concentrations of Cr (total), Fe, Mn, Ni, Zn by ICP-MS (Perkin Elmer Elan DRC-e); DLs Cr 5×10⁻³, Fe 2×10⁻³, Mn 3×10⁻³, Ni 1×10⁻³, Zn 1×10⁻³ mg/dm³. Accuracy 95-109%; NIST 1633b reference material; QA/QC per PN-EN ISO 17025 (blank samples, doubled samples min. 25%, marked samples in each series); analytical work at PCA-accredited lab (AB1050) at AGH University of Krakow. Statistical compilation in Excel 2013 and Statistica ver. 13.3. Health-risk indicators: MoS (= NOAEL/SED), HQ (= SED/RfD), HI (= ΣHQ), LCR (= SED·CSF or SED/HT25), QRA via AEL/CEL ratio per Basketter et al. 2003 / Api et al. 2008 framework. US EPA RfD and CSF values used as authoritative reference. Limitations stated by the authors: brand names not revealed due to legal considerations (only country of manufacturer disclosed); Cr-total measured (not speciated to Cr-VI); foundations, creams, masks applied at 0.51 g/cm² as default; BW = 60 kg adult only (no infant/child scenario).
Implications
- Certification (HMTc): Direct occurrence evidence for Cr (total), Fe, Mn, Ni, Zn in EU/US-market consumer cosmetics. The authors compare their values to prior cosmetic-metal literature in Table 2 and note (p. 6) that Cr and Ni in blushes in this study exceeded the Iwegbue et al. blush values (Iwegbue Cr 12.7 mg/kg, Ni 19.7 mg/kg vs this study Cr mean 110.4 and max 416 mg/kg, Ni mean 27.3 and max 95 mg/kg). The total-Cr measurement does not resolve Cr-VI vs Cr-III speciation; authors used RfD = 0.3 for Cr-VI in their conservative LCR scenario but only total-Cr values were measured, so the LCR scenario should be treated as upper-bound. Useful as occurrence data for the cosmetics-and-personal-care HMTc rows (foundations/powders/blush, face/neck leave-on, lipstick, eye-makeup). Adult-only modeling (BW 60 kg) — no child/infant exposure scenario.
- Courses: Teaching reference for the EU-EPA QRA methodology applied to consumer cosmetics; pairs with EU Cosmetic Regulation 1223/2009 framework and the SCHER 2010 children’s-products methodology for cross-framework comparison.
- App: Indirect relevance (cosmetics not in primary food-ingredient app scope). Useful for any future personal-care app extension that surfaces metal contamination of finished cosmetics.
Wiki pages this source may touch
- chromium
- iron
- manganese
- nickel
- zinc
- makeup-foundation-powders-blush
- face-neck-leave-on-skin-care
- lipstick
- eye-makeup
Verification notes
- 2026-05-17 re-ingest (Claude Opus 4.7): corrected
licensefrom “CC BY 4.0” to “CC BY-NC-ND 4.0” per PDF p. 13 Open Access notice (“Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License”). - 2026-05-17 re-ingest: corrected
productsfrom[face-paint, lipstick, childrens-makeup]to[makeup-foundation-powders-blush, face-neck-leave-on-skin-care, lipstick, eye-makeup]to reflect what the paper actually measured. The paper studied adult cosmetics (foundations, blushes, lipsticks, face creams, face masks, eye shadows; BW = 60 kg adult). It did not measure face paints and did not study children’s makeup specifically. Routing tochildrens-makeupandface-paintwas a pre-Cat-2-Step-0-Lock artifact when those slugs served as catchalls; with the 2026-05-16 Cat 2 Step 0 Lock providing proper rows (makeup-foundation-powders-blushfor Cat 13 Row 5,face-neck-leave-on-skin-carefor Row 3,lipstickfor the existing lipstick page,eye-makeupfor Row 7), routing should reflect actual sample composition. The paper’s value for the children’s-makeup HMTc row is via metal-context (Cr/Ni dermal-sensitization pathway documentation), not direct product-occurrence evidence. - Cr-total measurement only; authors did not speciate Cr-VI vs Cr-III despite using the Cr-VI RfD (0.3) in conservative LCR scenario. Per Part 14, total Cr must not be substituted for Cr-VI; the source’s risk-scenario LCR values should be treated as conservative upper-bound only when reused for synthesis.
- Sample sizes per product category are very small (n=3-4); the maxima are single-sample observations, not population estimates. Per Part 12 brand firewall, the country-of-manufacturer pattern (e.g., highest Cr/Ni in a Polish-manufacturer blush, highest Fe/Mn/Zn in US/German-manufacturer eye shadows) is reported here at country level only; the paper itself withheld brand identities for legal reasons (paper p. 2, “Research materials”).
- Pb, Cd, As, Hg not measured in this study; the analyte panel was limited to Cr, Fe, Mn, Ni, Zn.
matrices: [cosmetics]is added to the controlled vocabulary; “online-market” is sourcing context, not a matrix.ingredients: []is intentional — paper measures finished cosmetics, not raw ingredients (mineral pigments etc.). The routing layer’s malformed-advisory for missingingredientsis expected for cosmetic-finished-product papers.- Audit subagent (2026-05-17) flagged a source-internal inconsistency in the paper’s QRA conclusions: Table 12 reports AEL/CEL ratios of 0.07 (lipsticks) and 0.12 (eyeshadows), which mean AEL < CEL (i.e., consumer exposure exceeds the allowable level under the paper’s own definition on p. 6: “The concentration of a given substance in a product is acceptable when AEL is higher than CEL”). The paper’s narrative on p. 10, however, asserts AEL was higher than CEL for “all samples of eye shadows and lipsticks” and concludes none of the analyzed cosmetics poses a skin-sensitization risk. This is an internal contradiction in the source. For HMTc synthesis purposes, the Table 12 values are the primary data and should be cited directly; the source’s “no risk” conclusion for lipstick and eyeshadow under the QRA framework is not supported by its own Table 12.
- Audit subagent (2026-05-17) flagged a textual defect in Implications (“n=12.7 mg/kg Cr”); corrected to attribute the Iwegbue et al. comparison explicitly with both values restated from Table 2.
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.