The United States (US) legal cannabis industry is estimated to grow to $50.7 billion by 2028 as legalization of medical and non-medical adult cannabis use expands across US states [1]. Cannabis, commonly known as marijuana, refers to the dried flowers, leaves, stems and seeds of the cannabis plant [2]. Importantly, the US Federal Control Substance Act (CSA) definition of marijuana excludes products that meet the legal definition of hemp [cannabis plants and derivatives containing no more than 0.3% Δ−9 tetrahydrocannabinol (THC) on a dry weight basis per the US Agriculture Improvement Act of 2018].

The US National Survey on Drug Use and Health reported in 2023 that 61.8 million individuals age 12 years or older reported using marijuana (including cannabinoid chemical substances and THC, the psychoactive components of cannabis), making it the most commonly used psychoactive substance other than alcohol [2, 3]. Increasing cannabis use is driven by multiple factors, for instance, widespread product marketing online and in physical spaces, with advertisement exposure associated with higher odds of marijuana use, heavier use and use of novel products with higher THC concentrations [4–7].

Direct-to-consumer cannabis online marketplaces, that aggregate products sold through physical and online retailers, facilitate the sale of naturally occurring cannabinoids and cannabis-derived product (CDP) [generally defined as products with compounds occurring naturally in the Cannabis sativa L. plant, including cannabidiol (CBD) and THC], which are the focus of this study [8]. These platforms provide consumers with access to a diverse selection of CDPs helping consumers find cannabis vendors/dispensaries and facilitating online shopping options (e.g. in some states online ordering via delivery, pickup). Two large cannabis marketplaces are Leafly (a website with more than 100 million visitors per year) and Weedmaps (a web and mobile-based platform with 17.4 million monthly active users in 2022) [9–11]. CDP listings on these marketplaces can also include descriptions that products treat certain health ailments, conditions and even specific diseases [12].

Prior research has examined different online sources purporting health benefits of CDPs (e.g. Internet articles, social media posts) and also on websites operated by cannabis dispensaries [13–15]. These studies found a number of health- or disease-related marketing claims including for treatment of pain, spasms, anxiety, depression, insomnia and nausea/vomiting to name a few [13]. Importantly, marketing of unsubstantiated health claims can introduce unique public health, social justice and consumer protection issues, including targeted marketing toward certain minorities that could lead to unequal distribution of health impacts of cannabis use [13, 16]. Prior studies specifically conducting analysis on Leafly and Weedmaps have characterized product marketing practices in specific US states and Canada, use of age verification and product pricing data [16–22]. One study of Weedmap found a high percentage of retailers in Colorado (61%) and Washington (44%) that made health benefit claims for CDPs [16].

Despite an increasing number of studies examining the potential health benefits of CDPs on conditions such as pain management, sleep disorders and harm reduction for specific substance use behavior [23–28], there are several known health harms and adverse events associated with acute consumption and chronic cannabis use (e.g. gastrointestinal, cardiovascular, cognition, mood and anxiety disorders and specific conditions such as cannabis hyperemesis syndrome). [29–33] Hence, to better understand the potential consumer and public health implications of direct-to-consumer marketing of health benefit claims, we conducted an exploratory observational study of Leafly and Weedmaps with the aim of detecting and characterizing CDP health benefit claims sold in the United States.

This exploratory study was conducted in four phases: (1) data mining of all cannabis product listings (including extracted and synthesized CDPs and naturally occurring and grown cannabis products) located in the United States on cannabis online marketplaces, Leafly and Weedmaps (chosen as they are two large cannabis websites); (2) keyword filtering of product listings using a set of health- and medical-related terms to identify the health benefits listed in product descriptions; (3) manual annotation and content analysis to conduct in-depth characterization of advertised health benefits; and (4) examined health benefit claims to better understand any statistical differences by examining product characteristics and location of product purchases. See Figure 1 for summary.

This study was exempt from institutional review board approval in accordance with the Common Rule, as all data were publicly available and any user-generated data did not include individually identifiable information, and results are presented in the aggregate. The study followed the Standards for Reporting Qualitative Research (SRQR) reporting guideline.

A total of 1 463 510 CDP listings were collected on both Leafly and Weedmaps. After the removal of duplicate product listings with the same product ID, 624805 (42.7% of the original dataset) unique CDP remained; specifically, 50 951 (8.1%) from Leafly and 573 854 (91.9%) from Weedmaps. After keyword filtering and manual annotation of all unique CDP on Leafly and Weedmaps, 26 669 CDP (4.3% of total unique CDP listings; 2.0% of total Leafly listings; and 4.5% of total Weedmaps listing) were linked to at least one health benefit claim, comprised of 3.7% (n = 998) originating from Leafly listings and 96.2% (n = 25 671) from Weedmaps listings.

Across all Leafly listings reviewed with at least one claim (n = 998), we observed a mean of 3.4 health benefit claims per CDP listing. Of these 3.1% (n = 31) included some form of health warning or disclosure in their product information. For CDPs with at least one health claim on Weedmaps, a mean of 3.6 health benefit claims per CDP listing, with 0.8% (n = 196) having a health warning or disclosure were identified (see Table 1 for examples). Further, although cannabinoid compounds associated with a product were not standardized in CDP listing data, a summary of the breakdown per product listing observed and number of health benefit claims is provided in the Supporting information.

An analysis of all CDP listings on Leafly and Weedmaps with health benefit claims (n = 26 669) revealed that 4.9% (n = 1304) of these listings received at least one user-generated product review, from which a total of 6458 individual reviews were collected. Of these reviews, 4.6% (n = 295) included a user mentioning their general sentiment toward the health benefit claim(s) made by that product (e.g. agreeing or disagreeing that the CDP helped address the health issue/claim). Of reviews expressing user sentiment, 82.4% (n = 243) included positive sentiment toward the product with the health benefit claim, 14.6% (n = 43) included negative sentiment and 3.0% (n = 9) included neutral sentiment.

This exploratory study conducted market surveillance of two popular cannabis online marketplaces used by millions of Americans and found 4.47% (n = 26 669) of CDP listings had at least one health benefit claim (with a mean of 3.4 and 3.6 claims per product with at least one health benefit claim for Leafly and Weedmaps, respectively). The descriptions analyzed included purported benefits for serious, complex and life-threatening diseases such as cancer, mental health disorders, autoimmune diseases, neurological disorders and non-communicable diseases that require careful clinical management with a healthcare provider. The distribution of claims was higher in states where medical and non-medical cannabis use was legal (79.8%) compared to states that only allowed medical cannabis use (20.2%). These products encompassed 96 054 distinct individual health benefit claims across all CDPs reviewed, with the majority of user reviews interacting with these products having positive sentiment toward the health claim topic.

Currently, the FDA has only approved one CDP (Epidiolex) and three synthetic cannabis-related drug products (Marinol, Syndros and Cesamet) for limited therapeutic use [31]. Hence, the marketing of health benefits for adult recreational use cannabis on e-commerce platforms may influence consumer attitudes on CDPs as an alternative treatment or disease management method despite lack of FDA approval or competent and reliable scientific evidence to support claims. This may lead to consumers foregoing evidence-based care in favor of experimenting with cannabis, or may complicate existing treatment and care plans, particularly if a patient’s cannabis use behavior is not discussed because of perceived stigma with a care provider [32, 33, 35, 36].

Further, only a small percentage of CDP listings with health benefit claims had product warnings or disclaimers. An overall lack of product warnings or disclaimers was observed despite passage of state laws requiring specified warnings about the safety of cannabis use on packaging and separate research indicating that health warnings are an important strategy to communicate risks to consumers (e.g. risks associated with addiction, impaired driving, worsening mental health, cannabis use disorder, use while pregnant or breastfeeding, etc.) [37]. Additionally, although not formally coded, we generally observed other CDP marketing characteristics coinciding with health benefit claims that included appealing marketing cues, such as attractive product marketing and packaging, flavoring and concurrent offers for sales promotions that may also impact consumer behavior.

Similar to prior studies examining cannabis health benefits made on different online platforms, we observed a large diversity of claims that spanned a host of general health areas, therapeutic topics and specific diseases [16, 17, 20]. We also observed statistically significant differences between RoA and frequency of health benefit claims among CDPs with at least one health benefit claim. Specifically, multi-system RoA CDPs were observed as having a higher frequency of claims compared to certain other RoAs in both the Weedmaps and Leafly datasets. These results need further study, as CDPs that can be consumed in multiple ways may lead to consumers being influenced by health benefit claims in how they use a cannabis product, which can impact how cannabis and its psychoactive components are absorbed in the bloodstream and its intoxicating effects (e.g. respiratory versus digestion). Additional research is needed to better understand how combinations of different health benefit claims (e.g. a product claiming to treat anxiety, cancer and a gastrointestinal disorder), their RoAs and other product characteristics (e.g. specific cannabinoid compounds) have potential to elevate risk, such as through counterindications with existing treatment or negative changes in health behaviors.

To our knowledge, this is the first study to comprehensively characterize CDP-related health benefit claims from two major nationwide online marketplaces. Most US states that allow adult-use have regulations concerning cannabis advertising and marketing, including some that prohibit claims that cannabis has therapeutic or curative benefit [38]. Reflecting ongoing concerns about CDP-associated unsubstantiated health benefits, the FDA and Federal Trade Commission (FTC) have issued warning letters to sellers for marketing unapproved CBD and Δ−8 THC products and/or making claims lacking scientific substantiation, including for treating coronavirus disease 2019, Alzheimer’s disease, cancer and multiple sclerosis [31, 39–41]. The FTC can regulate advertising claims about cannabis products to ensure health benefit claims are supported by scientific evidence, while FDA can take enforcement action against products deemed unapproved when they make claims about treatment of medical conditions or for other therapeutic uses [40, 42]. Hence, findings from this study suggest that proactive online cannabis marketplace monitoring could help enhance enforcement of existing laws and regulations and better ensure correct representation of CDP risks and benefits. It could also help regulators identify higher risk CDPs with more concerning claims, while also identifying the need to enforce mandates on requiring health warnings for online CDP listings. Further, the potential reclassification of marijuana to a Schedule III drug may also impact future CDP commercial marketing practices, necessitating the development of best practices on marketing restrictions, a recommendation also highlighted in the 2024 NASEM report on cannabis policy that should be supported with additional research [30].

SUPPORTING INFORMATION

Additional supporting information can be found online in the Supporting Information section at the end of this article.