Dose and Label Accuracy in Edible Medical Cannabis Products (2024)

As the use of cannabis (marijuana) for medical purposes has expanded, a variety of edible products for oral consumption has been developed. An estimated 16% to 26% of patients using medical cannabis consume edible products.^1,2 Even though oral consumption lacks the harmful by-products of smoking, difficult dose titration can result in overdosing or underdosing, highlighting the importance of accurate product labeling.

Regulation and quality assurance for edible product cannabinoid content and labeling are generally lacking. We investigated the label accuracy of edible cannabis products.

An Internet directory of dispensaries, with a menu of products available at each, was used to determine purchase locations in San Francisco, California, Los Angeles, California, and Seattle, Washington. A list of dispensaries was generated, with individual businesses randomly selected until 3 were identified in each city that offered at least 1 edible cannabis product from each of 3 common categories (baked goods, beverages, candy or chocolate) with package labels that provided, at minimum, specific Δ⁹-tetrahydrocannabinol (THC) content.

Between August and October 2014, individuals with a physician’s letter (in compliance with state laws) and no history of purchasing edible cannabis products were sent to the preselected dispensaries and instructed to buy as large a variety of products, in terms of type and labeled cannabinoid content, as possible within budget ($400/city). Purchasers, but not dispensary staff, were aware that the products would be evaluated following purchase. None knew the details of the analyses or testing to be completed.

Cannabidiol (CBD) and THC are typically the most concentrated chemical components of cannabis and are believed to primarily drive therapeutic benefit.³ Studies suggest improved clinical benefit and fewer adverse effects with a THC:CBD ratio of 1:1.⁴ Even though other cannabinoids were analyzed, results focus on THC and CBD.

For testing, entire package contents were hom*ogenized (crushed or mixed). Two 1.5-g (solid) or 25-g (liquid) samples of each product were tested via high-performance liquid chromatography, with results averaged and adjusted for total product weight.

When results of duplicate tests differed by more than 10% (cannabinoid heterogeneity), the entire product was analyzed (n = 37). Five randomly selected products in which duplicate testing was within 10% were subject to complete testing; results confirmed the accuracy of the duplicate testing method. Products were considered accurately labeled if the measured THC and CBD content was within 10% of the labeled values, underlabeled if the content was more than 10% above the labeled values, and overlabeled if the content was more than 10% below the labeled values.

A χ² test was used (SPSS version 22; SPSS Inc) to evaluate effects of location on label accuracy. Significance was determined at P < .05 (2-sided).

Of 75 products purchased (47 different brands), 17% were accurately labeled, 23% were underlabeled, and 60% were overlabeled with respect to THC content (Table 1). The greatest likelihood of obtaining underlabeled products was in Los Angeles and overlabeled products in Seattle (χ² = 12.94, P = .01).

Non-THC content was generally low (Table 2). Forty-four products (59%) had detectable levels of CBD; only 13 had CBD content labeled. Four products were underlabeled and 9 were overlabeled for CBD. The median THC:CBD ratio of products with detectable CBD was 36:1, 7 had ratios of less than 10:1, and only 1 had a 1:1 ratio.

Edible cannabis products from 3 major metropolitan areas, though unregulated, failed to meet basic label accuracy standards for pharmaceuticals. Greater than 50% of products evaluated had significantly less cannabinoid content than labeled, with some products containing negligible amounts of THC. Such products may not produce the desired medical benefit.

Other products contained significantly more THC than labeled, placing patients at risk of experiencing adverse effects.^5,6 Because medical cannabis is recommended for specific health conditions, regulation and quality assurance are needed.

A limited number of cities, dispensaries, and products were included. Because no source lists all dispensaries, and many products are not labeled with cannabinoid content, a true random sample was not possible and the results may not be generalizable. However, this study illustrates the variability in label accuracy for edible cannabis products within 2 of the largest medical cannabis markets in the United States.

Section Editor: Jody W. Zylke, MD, Deputy Editor.

Corresponding Author: Ryan Vandrey, PhD, Behavioral Pharmacology Research Unit, Johns Hopkins University School of Medicine, 5510 Nathan Shock Dr, Baltimore, MD 21224 (rvandrey@jhmi.edu).

Author Contributions: Dr Vandrey had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Vandrey, J. Raber, Douglass, Bonn-Miller.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Vandrey, J. Raber, M. Raber, Bonn-Miller.

Critical revision of the manuscript for important intellectual content: Vandrey, J. Raber, Douglass, Miller, Bonn-Miller.

Statistical analysis: Bonn-Miller.

Obtained funding: Vandrey.

Administrative, technical, or material support: All authors.

Study supervision: Vandrey, J. Raber, Douglass.

Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Bonn-Miller reported receiving personal fees and nonfinancial support from Tilray (Division of Privateer Holdings) and travel reimbursem*nt from Americans for Safe Access. No other disclosures were reported.

Funding/Support: This project was funded by Johns Hopkins University School of Medicine except for the cost of analytical testing, which was covered by Werc Shop Laboratories.

Role of the Funder/Sponsor: Neither Johns Hopkins University nor Werc Shop Laboratories played a role in the design or conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or the decision to submit the manuscript for publication.

Correction: This article was corrected on June 25, 2015, to fix the overlabeled and underlabeled descriptors used incorrectly in the text and in Table 1.