Clean Label Project 2024-25 — Heavy metals in US protein powders (Protein Study 2.0)

Summary

This is a Clean Label Project category insights report (Protein Study 2.0) presenting a market-wide screen of heavy metals and bisphenols in 160 protein powder products drawn from 70 top-selling US brands, representing approximately 83% of the US protein supplement market. The report aggregates 35,862 individual analyses across six industrial-chemical panels run at Ellipse Analytics, with heavy metals (As, Cd, Pb, Hg) measured by ICP-MS and bisphenols (BPA, BPS) and pesticides measured by LC-MS/MS. Findings are presented as cohort-level exceedance percentages against the California Proposition 65 Safe Harbor reference for lead (Pb MADL = 0.5 µg/day), not as per-product or per-serving concentration tables. The headline pattern: 47% of all products exceeded at least one federal or state regulatory level for lead and 21% exceeded twice the Prop 65 Pb reference; plant-based powders had the highest exceedance rate at 77%, certified-organic powders ran at 79% (with 41% exceeding twice the Prop 65 reference), and chocolate-flavored powders at 65% (29% over 2× Prop 65). Whey-based and collagen-based powders ran lower, at 28% and 26% respectively, and the report’s own narrative recommends “whey or collagen based protein powders that are not chocolate flavored” as the lowest-lead pattern in the cohort. Cadmium is described in multiplicative terms: organic averaged twice the cadmium of non-organic, plant-based averaged five times the cadmium of whey, and chocolate-flavored averaged 110× the cadmium of vanilla-flavored. Bisphenols are essentially gone from the cohort — BPA or BPS was detected in only 3 of 160 products, against 55% of products in Clean Label Project’s earlier 2018 protein study. The report is non-peer-reviewed industry/advocacy material, presents only cohort-level percentages without per-product concentration tables or summary statistics on the underlying distributions, and is best treated as a B-tier exceedance-pattern anchor for the US protein-powder market that complements (rather than replaces) the per-product datasets in bandara2020-protein-powder-heavy-metals and cleanlabel2020-collagen-heavy-metals.

Key numbers

Source reports cohort-level exceedance percentages against the California Proposition 65 Safe Harbor reference for lead, not per-product concentrations or summary statistics. All percentages below are the proportion of products in each sub-cohort whose lead loading exceeded the stated Prop 65 reference. Denominators for sub-cohort splits (whey, plant-based, collagen, organic, chocolate) are not disclosed.

Cohort-level lead exceedance against California Prop 65 (pages 3, 5):

Cross-cohort multiplicative findings (page 2 executive summary and page 6 narrative):

• Certified-organic products averaged three times the lead and twice the cadmium of non-organic products in the cohort.
• Plant-based powders averaged three times the lead and five times the cadmium of whey-based powders.
• Chocolate-flavored powders averaged four times the lead and 110 times the cadmium of vanilla-flavored powders.

Bisphenols (page 6):

• BPA or BPS detected in 3 of 160 products in this 2024-25 study, versus 55% of products in Clean Label Project’s prior 2018 protein-powder study. Source attributes the change to packaging-supply-chain reformulation.

Context for the comparison ladders. The source frames all heavy-metal exceedances as multiples of the California Proposition 65 Safe Harbor reference for lead. The OEHHA Proposition 65 Pb MADL (oral reproductive endpoint) that those percentages back-derive against is 0.5 µg/day; the 2× Prop 65 multiplier therefore corresponds to a 1.0 µg Pb/day exposure on the source’s serving-size basis. The Pb MADL numeric is imported here from OEHHA’s published Safe Harbor list, not from the report’s own text. The report does not state which serving size or exposure assumption it used to translate measured concentrations into daily Pb exposures, and does not name the Prop 65 reference values for arsenic, cadmium, or mercury. No per-product µg/serving values, no concentration ranges, no detection rates, and no central-tendency statistics (mean, median, percentile values) are disclosed in the public report — only the exceedance percentages above and the multiplicative cross-cohort statements.

Methods (brief)

Clean Label Project procured 160 protein powder products from 70 brands sourced from Nielsen and Amazon best-seller listings, supplemented with top products from the natural and organic marketplace, and forwarded them to Ellipse Analytics, an independent analytical chemistry laboratory. Six industrial-chemical panels were run, including heavy metals and bisphenols. Heavy metals (arsenic, cadmium, lead, and mercury) were measured by Inductively Coupled Plasma Mass Spectrometry (ICP-MS). Bisphenols (BPA, BPS) and pesticides were measured by Liquid Chromatography–Tandem Mass Spectrometry (LC-MS/MS). The report cites 35,862 individual analyses across 258 chemicals/heavy metals as the underlying dataset volume. No metal speciation is described; arsenic and mercury are total elemental measurements. The report does not disclose level of quantification, sample-prep digestion method, lab-blank or spike-recovery protocols, replicate-sample structure, or per-product analytical detail. Serving-size denominators used to convert concentrations into per-serving Pb exposures (the basis for all Prop 65 exceedance percentages) are not stated.

Implications

• This is the Index’s market-screen exceedance-pattern anchor for the US protein-powder category in 2024-25, complementing the per-product peer-reviewed risk assessment in bandara2020-protein-powder-heavy-metals (which re-analyzed the prior Clean Label Project 2018 cohort) and the per-serving collagen-supplement market screen in cleanlabel2020-collagen-heavy-metals. All three support the supplements-protein-collagen-powders product page.
• Lead is the load-bearing analyte in this source. 47% of the 160-product cohort exceeded the California Prop 65 Pb MADL on a per-serving basis, and 21% exceeded twice that reference. The exceedance pattern is sharply stratified by sub-cohort: plant-based (77%), organic (79%), and chocolate-flavored (65%) sit at the top; whey (28%) and collagen (26%) sit at the bottom.
• The plant-based vs whey gradient, the organic vs non-organic gradient, and the chocolate vs vanilla gradient all point in the same direction (plant + organic + chocolate = higher Pb and Cd), consistent with soil-uptake and cocoa-input mechanisms reported elsewhere in the cocoa and plant-based-protein literature.
• Cadmium is described only multiplicatively here (organic 2× non-organic, plant-based 5× whey, chocolate 110× vanilla), without underlying concentrations. The chocolate-vs-vanilla cadmium ratio in particular is large and should be treated as cohort-level signal rather than as a per-product quantitative finding until corroborated by sources reporting actual concentrations.
• Bisphenol contamination has effectively disappeared from this product category between Clean Label Project’s 2018 cohort (55% of products with detectable BPA/BPS) and the 2024-25 cohort (3 of 160 products); this is one of the few categories in HMI where industry-wide packaging reformulation has produced a measurable, near-complete drop in a contaminant of concern.
• Evidence fitness caveat: non-peer-reviewed industry/advocacy report, accredited-lab ICP-MS but no per-product values, no concentration distribution, no central-tendency statistics, no detection-rate denominators for sub-cohort splits, no QA/QC disclosure, no serving-size or exposure-conversion methodology disclosed. Treat as a B-tier exceedance-pattern anchor for the 2024-25 US protein-powder market. Do not synthesize per-product concentration claims from this source; the underlying per-product data is held by Clean Label Project and not published in the report.

Wiki pages this source may touch

• supplements-protein-collagen-powders — 2024-25 market-screen lead exceedance pattern for the US protein-powder category; bisphenol near-elimination signal vs the 2018 cohort.
• lead — 47% of US protein powders exceed Prop 65 Pb MADL per-serving in 2024-25, with plant-based and organic and chocolate-flavored sub-cohorts running 65–79% exceedance.
• cadmium — multiplicative cohort findings: organic 2× non-organic, plant-based 5× whey, chocolate 110× vanilla; no per-product values.
• arsenic — included in the ICP-MS panel; no quantitative findings reported.
• mercury — included in the ICP-MS panel; no quantitative findings reported.

Verification notes

• All cohort percentages taken from pages 3 and 5 (graphical “Key Data and Findings” panels). Multiplicative cross-cohort statements taken from the page 2 Executive Summary and the page 6 “Aren’t These Contaminants Regulated?” narrative. Bisphenol finding taken from page 6. Methods text taken from page 7.
• Part 12 brand firewall: the source PDF names eight Clean Label Project Certified brands on page 8 (Genetic Nutrition, Ritual, Puori, Wicked Protein, Gorilla Formula 23, Oziva, Wellbeing Nutrition, NutraBox) as part of CLP’s own certification program promotion. None of these brand names is reproduced on this wiki page. This is not a contamination-by-brand listing — the report itself does not attribute contamination values to individual brands in the published PDF — but Part 12 also bars naming of certified brands in the wiki, and the safer reading is to drop the entire certification-promotion block from the source page. The lab name (Ellipse Analytics) is retained under the Part 12 scientific-method vendor exception; it identifies the analytical service provider, not a sampled consumer product.
• Part 2 wiki/HMTc firewall: the source is heavily framed around advocacy (“a wake-up call,” “stricter safety standards,” “the future for consumer trust and industry change”) and around Clean Label Project’s own Purity Award and Transparency Certification programs. Those framings are not reproduced here; only the analytical occurrence findings and exceedance percentages are recorded. No HMTc threshold proposal is drawn from this source, and no Clean Label Project program endorsement is reproduced.
• Speciation: ICP-MS reports total elemental concentrations; tAs and tHg are correct labels, iAs and MeHg are not claimed. No speciation chemistry is described in the report.
• Units and bases: all heavy-metal exceedances are framed against the California Prop 65 Safe Harbor reference for lead, which is a daily-exposure µg/day reference, not a concentration ppb reference. The source does not state the serving-size denominator or per-product concentration that produces each exceedance count, so the underlying ppb concentrations cannot be back-calculated from this report. The 0.5 µg/day Pb MADL numeric is imported here from OEHHA’s published Safe Harbor list, not from the source’s own text.
• Year selection: report self-titles as “2024-25 Protein Powder Category Report” and also as “Protein Study 2.0.” Cite-key year selected as 2025 to match the publication-distribution year and the source filename (04_CleanLabel_2025_Protein_Powder_Heavy_Metals.pdf); sampling year range recorded as 2024-2025 in frontmatter.
• Cohort denominators: the source discloses total cohort n=160 but does not disclose sub-cohort denominators (how many of the 160 were plant-based, whey, collagen, organic, chocolate, vanilla). Sub-cohort exceedance percentages are therefore on undisclosed denominators and should not be back-converted into product counts.
• Bisphenol comparison: the 2018 reference figure (“55%”) refers to Clean Label Project’s prior 2018 protein-powder investigation, not to the 2020 collagen study at cleanlabel2020-collagen-heavy-metals; the 2020 collagen white paper did not report bisphenols and is not the comparator.
• Scaffold-introducing slugs: matrices labels plant-based-protein, chocolate-flavor, and vanilla-flavor are not yet in docs/gpt-collaboration/taxonomy-snapshot.md. The existing protein-powder, whey-protein, and collagen-supplement labels follow the convention established by bandara2020-protein-powder-heavy-metals and cleanlabel2020-collagen-heavy-metals. New slugs surfaced for the next taxonomy-snapshot regeneration.
• Evidence tier B rationale: non-peer-reviewed industry/advocacy report published as a downloadable PDF on cleanlabelproject.org; no DOI; no per-product data; no concentration distribution; no QA/QC reporting. Analytical chemistry performed by an accredited third-party lab (Ellipse Analytics) using standard methods (ICP-MS for metals, LC-MS/MS for bisphenols), which justifies B over C, but the absence of peer review, of per-product reporting, of statistical detail on the underlying distributions, and of serving-size/exposure-conversion methodology prevents A-tier classification.

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.