Summary
Gidal and colleagues purchased 202 commercially available CBD products from US online and brand-website sources in October to December 2021. Products covered four formats: 100 tinctures, 48 gummies, 34 vape products, and 20 topicals. The study assessed label accuracy for cannabinoid content and product-type claims (full spectrum, broad spectrum, CBD isolate), and measured four heavy metals (lead, arsenic, cadmium, mercury), 15 residual solvents, and 232 pesticides. Lead was the most prevalent heavy metal, detected in 42 of 202 products; mercury was not detected in any sample. Three percent of heavy-metal results violated at least one regulatory threshold.
Key numbers
All heavy metal concentrations are reported as parts per billion (ppb) in the product as tested. The paper does not specify wet or dry weight; CBD products are typically liquid or semi-solid matrices tested on an as-received basis.
Across all 202 products (all formats combined):
- Lead detected in 42 of 202 products. Mean 916.21 ppb (SD 4,222.29); median not detected; range not detected to 27,150 ppb. Five products (3 percent) violated Proposition 65 Safe Harbor and CDPHE thresholds: four tinctures and one gummy (Results, Heavy metal contamination section).
- Arsenic detected in 6 of 202 products. Mean 1.06 ppb (SD 5.97); median not detected; range not detected to 57.40 ppb (Results, Heavy metal contamination).
- Cadmium detected in 4 of 202 products. Mean 0.89 ppb (SD 5.91); median not detected; range not detected to 56.40 ppb (Results, Heavy metal contamination).
- Mercury: not detected in any of 202 products (Results, Heavy metal contamination).
- Across all four metals, detections occurred 52 times across 44 of 202 products. The combined across-all-metals mean is stated as 192.24 ppb (SD 1,943.91); median not detected; range not detected to 27,150 ppb. This mean is dominated by the lead distribution and should not be read as a per-metal value.
Gummy-specific findings (n = 48 gummies):
- One gummy violated a lead regulatory threshold (Proposition 65 or CDPHE). The paper does not report a lead concentration for the violating gummy separately, nor does it report the number of gummies with detectable lead, the mean, or the range for the gummy subgroup. Per-format breakdowns for heavy metals are not tabulated; the threshold-violation count (four tinctures, one gummy) is the only gummy-specific heavy-metal number in the main text (Results, Heavy metal contamination; Discussion).
- Label accuracy for gummy CBD potency: 56 percent of gummy products (27 of 48) contained less than 90 percent of labeled CBD; 8 percent (4 of 48) exceeded 110 percent; 35 percent (17 of 48) were within the 90-110 percent range (Table 2).
Regulatory benchmarks used:
- Proposition 65 (California OEHHA) no-significant-risk levels and maximum allowable dose levels, applied as micrograms per serving when serving size was stated.
- Colorado Department of Public Health and Environment (CDPHE) industrial hemp regulations.
- US Pharmacopeia (USP) elemental impurity safe harbor limits (USP general chapter 232 and 2232, and USP 39).
- Benchmarks are applied as stated in the paper; this page does not independently verify those threshold values.
Contamination source discussion:
The paper attributes heavy metal contamination to uptake from soil via fertilizer before harvest, cross-contamination during processing, and post-processing adulteration (Dryburgh et al. 2018, cited in Discussion). Cannabis is a documented accumulator of lead, cadmium, and other metals from contaminated growing media. The paper notes that 25 percent of all products accounted for 93 percent of total contamination across all contaminants, suggesting heterogeneous product quality in this market.
Methods (brief)
Products were purchased from US online and brand-website retail between October and December 2021. A single sample of each product was tested. All testing was conducted at an ISO 17025-accredited laboratory (Ellipse Analytics, Denver, Colorado). Heavy metals (cadmium, arsenic, mercury, lead) were measured by inductively coupled plasma mass spectrometry (ICP-MS; NexION 350, PerkinElmer) after homogenization to uniform texture. Standards were prepared in the same diluent as the sample matrix. ICP-MS data were processed with Syngistix software. Supplementary material is cited for further homogenization and standard-preparation details, but those supplement pages are not part of this deposit.
The paper does not distinguish total arsenic from inorganic arsenic; it reports “arsenic” without speciation. The study does not report limits of detection or limits of quantification for the heavy-metal assay in the main text (supplementary tables cited). Certified reference materials and spike recovery data are described as present in supplementary material but are not in the main article. Quality control parameters are stated to include analyte limits of quantification, details in Supplementary Material (Methods, Heavy metals analysis).
The paper benchmarks against Proposition 65, CDPHE, and USP safe harbor limits. It does not apply EU or Codex maximum levels. It does not distinguish organic from inorganic arsenic, and it does not distinguish total mercury from methylmercury.
Evidence Fitness
This source supports modeled or limited evidence. The main text provides overall detection rates and means for all 202 products combined, the range of lead concentrations across positive detects, and threshold-violation counts by product format. It does not provide per-format (gummy, tincture, vape, topical) means, medians, or distributions for any heavy metal. The gummy-specific heavy-metal data are limited to a single threshold-violation count (one of 48 gummies violated a lead threshold). The source was funded by Jazz Pharmaceuticals and several authors are employees or paid consultants of that firm; the funder participated in design, data collection, analysis, and writing. This conflict should be weighed when interpreting conclusions.
The convenience sampling design (US online retail, single sample per product) limits generalizability. Single-sample-per-product testing means lot-to-lot variability is unknown. The study does not provide per-gummy heavy-metal concentrations, so the evidence cannot support contamination-profile estimates for the cannabis-edibles page. It supports the qualitative finding that lead contamination occurs in US CBD gummy products at detectable and occasionally threshold-violating concentrations.
Implications
Certification: The finding that one of 48 CBD gummies tested had a lead level above Proposition 65 or CDPHE thresholds is consistent with the classification of cannabis-derived edibles as a category warranting lead monitoring. The paper does not provide sufficient per-format data to set a contamination-profile estimate for the cannabis-edibles row.
Courses: Useful for illustrating the market-quality heterogeneity of unregulated CBD products and the rationale for systematic testing.
App: This source does not support a populated contamination_profile block for the cannabis-edibles page. It supports noting that lead contamination at violating levels has been detected in at least one US CBD gummy product in this survey.
Microbiome: No microbiome content.
Provenance notes
Full open-access article available under CC BY at https://doi.org/10.3389/fphar.2024.1335441. Obtained from PMC open-access archive (PMCID PMC10982813, PMID 38562466). No retraction or correction found in Europe PMC or Crossref update/relation fields as of 2026-11-04. The study was funded by Jazz Pharmaceuticals, with the funder involved in design, analysis, and writing; this is disclosed in the article. The study analyzes CBD products that are not FDA-approved or regulated as conventional foods.
Related evidence
- Edibles (gummies, baked goods, beverages)
- Tinctures and sublingual oils
- Topicals (creams, balms, salves)
- Vape cartridges and hardware
Update history
No substantive edit history is available in this build. The full commit record is available in git.