Characterization of Women's Health Benefit Claims Made on Cannabis E-Commerce Platforms: A Retrospective Market Surveillance Study.

For this retrospective market surveillance study, CDP listings from e-commerce platforms Leafly and Weedmaps and all duplicate product listings were collected based on the “listing ID” field. Leafly and Weedmaps were used for this study as they represent two of the largest and most popular cannabis websites that connect users to CDP sales listings and U.S. dispensaries, with an estimated 8 million and 17.4 million monthly active users reported in 2022 for Leafly and Weedmaps, respectively. 30 To conduct content analysis of product characteristics derived from product listings, we utilized a framework established by Nali et al. that similarly identified product classifications (e.g., RoA, flavors, etc.) from product listing data on Leafly and Weedmaps. 31 We used this framework as, to our knowledge, there are no other published code books of both Leafly and Weedmaps product listings in the scientific literature. Data collection of CDP sales listings from these platforms utilized custom data mining programs and occurred from September–November 2023. Data collection was obtained by identifying potential women’s health benefit claims, data mining was conducted on the entire corpus of Leafly and Weedmaps product listings using 217 keywords derived from published literature on cannabis-related health and medical benefits for women (see Table 1 ). These keywords were also informed by standardized medical terminology derived from MedDRA (v28.0). Custom data collection programs were developed in Python programming language (Version 3.8) and Selenium (Version 4.4.3), with data from Weedmaps also requiring development of a program for reverse geocoding (Version 1.5.1) and a data request package (Version 2.27.1) as the platform is a web-based map application. Beautiful soup (Version 4.6.0) was used to decode the HTML content sent back to our data mining programs, which then extracted all available product information from listings. The data sources used for this study, Leafly and Weedmaps, do not differentiate whether the products captured are for medical or adult-use (non-medical). Additionally, individual product descriptions lack specificity (see eAppendix for characteristics of the data source). Once data was filtered for study keywords, two coders (First and Second Authors) with experience in detection and characterization of health product claims, inductively coded the filtered dataset to determine whether a product listing included at least one benefit claim for a general health or specific disease or condition. Manual annotation of data was required as there were no known publicly available data sets that could be used to accurately classify health claims from product listing information using other automated approaches (e.g., supervised machine learning, natural language processing, other data mining approaches). Hierarchical content coding of parent categories, sub-parent categories, and specific individual health benefit claims included one of the four general women’s health areas; a sub-class for a specific health, disease, or medical topic; and a final individual health benefit claim code for the specific claim mentioned in the product description. This hierarchical coding framework was developed based on previously published literature addressing distinct stages of the menstrual cycle (Menstruation, Follicular Phase, Luteal Phase), maternal health (Postpartum), and the menopausal transition phase (Perimenopause, menopause). 32 – 34 This framework was also informed by a review article by Corsi et al. that grouped cannabis effects on female reproductive health into categories of fertility, pregnancy outcomes, and menopause (see eAppendix for additional information on coding schema). 35 Additionally, all user-generated product review posts for CDPs discussing at least one health benefit claim were thematically coded for relevance (i.e., whether the review addressed a health benefit claim topic) and sentiment (e.g., agreement, disagreement, neutral) towards an advertised health claim. To address any discrepancy in coding, authors (First and Second) reviewed and conferred on the correct classification with the senior author (Last Author) to achieve consensus on classification. Coders reviewed each product listing independently and achieving a high intercoder reliability (κ=.95) for inclusion and class of health claims. IRB approval was not required for this study as all information collected was from the public domain. Any user-identifiable information from publicly available user-generated product review data was aggregated and removed from study results.

A total of 1,463,510 unique CDP listings were collected from Leafly and Weedmaps in the U.S. After keyword filtering, 3,325 unique CDP listings (0.2%) were extracted for further manual review specifically, [404 (12.2%) from Leafly and 2921 (87.9%) from Weedmaps]. These CDP listings originated from 214 distinct cannabis dispensaries in 18 states (Alaska, Arizona, California, Colorado, Illinois, Maine, Maryland, Massachusetts, Michigan, Missouri, Montana, Nevada, New Jersey, New Mexico, New York, Oregon, Washington, and Virginia), with an average of 1.70 listings per dispensary where only Weedmaps data fields could be extracted. Following manual annotation and content coding, 408 CDP listings were identified [44 (10.8%) from Leafly and 364 (89.22%) from Weedmaps] linked with at least one relevant health claim in their respective product descriptions, resulting in a total of 465 health benefit claims that mapped to five major women’s health areas. We note that a single product listing could contain multiple health benefit claims in the product description text, which resulted in a higher number of claims than individual product listings. Among the product listings with health benefit claims, five RoAs were identified, including products that could be administered through topical/transdermal application (e.g., dermatological agents, transdermal products, n=134), soluble oral (e.g., transmucosal, digestive, n=85), those inhaled [e.g., combustible cannabis products (CCP), cannabis electronic delivery systems (CEDS), n=81], multisystem products (i.e., flower buds, concentrates, n=86), and non-consumable items (e.g., suppositories, etc., n=22). Of these product listings with health benefit claims, 14.7% included a product with a characterizing flavor (e.g., banana, chocolate, mint). Specific terpenes (a class of compounds found in many plants, including cannabis, some of which are associated with flavor attributes and have been studied for potential health benefits), such as β-caryophyllene and δ-limonene, were also detected and attributed to specific women’s and reproductive health benefit claims in 65 CDP listings (see Table 2 for examples). Cannabinoid content included in product listing descriptions was highly unstructured and non-standardized. From this data a total of 356 cannabinoid compound descriptors were identified across unique CDP listings. The most frequently reported compounds included delta-9-tetrahydrocannabinol (THC, n=199) cannabidiol (CBD, n=118), cannabigerol (CBG, n=21), cannabinol (CBN, n=9), tetrahydrocannabivarin (THCV, n=7), and cannabichromene (CBC, n=2). Health benefit claims encompassed a variety of areas, including reproductive conditions, different menstrual cycle phases, menopause, and other diagnosable conditions, such as osteoporosis. Using our content coding approach, five major women’s health areas emerged: menstrual cycle (57.20%, n=266), fertility (13.98%, n=65), women’s health conditions (13.12%, n=61), menopausal transition (15.05%, n=70), and maternal health (0.65%, n=3). Among these major women’s health areas, there were 10 different sub-classes of specific health, disease, or medical topics, with the highest number of health benefit claims mapped to the luteal phase of the menstrual cycle (n=137), followed by menstruation (n=120) and menopause (n=67). Derived from these health or medical condition sub-codes were 33 unique claim types. The top four detected were related to premenstrual syndrome (n=131), menstrual cramps (n=76), libido (n=58), and osteoporosis (n=39) (see Table 3 ). For CDP listings with claims, on average 1.25 individual claims were present in their product descriptions. Our review for the presence of health warnings or disclaimer language among the 408 analyzed CDP listings found only 27 (6.6%) with any type of warning or disclaimer language. These health warnings and disclaimers included statements indicating the FDA has not evaluated this product for safety or efficacy and has not approved statements made by sellers/manufacturers of CDP listings (n=11); this product is not intended to diagnose, treat, cure, or prevent any disease – if you are pregnant or take pharmaceutical drugs, please contact your physician before use (n=14); or take extra caution to ensure edibles are out of the reach of children (n=2). Additionally, a total of 179 user-generated reviews of CDP listings with at least one health benefit claim were retrieved and reviewed. Of these, (7.82%, n=14) had thematic relevance to the purported health benefit claim(s) of the CDP. Of these relevant user-generated reviews, 12 (85.71%) were in agreement with the CDP’s health benefit claim (e.g., positive sentiment, such as users self-reporting relief from menstrual aches, cramps and pain), 2 (14.29%) were in disagreement (i.e., negative sentiment, such as product did not improve sex life), and none expressed neutral sentiments. Many of these users expressed positive sentiment towards products that provided relief toward menstrual symptoms, anxiety relief, and enhanced sexual experience.

This study detected 408 unique CDP listings on Leafly and Weedmaps, which included a variety of products, such as edibles, CEDS, and topicals. These products claimed health benefits across the female reproductive life cycle, including for menstruation, postpartum maternal health, and menopausal transition (perimenopause and menopause), despite lack of clinical research consensus supporting such claims. Specifically, this study detected products that promote purported benefits for health conditions commonly experienced among women (e.g., menopause, menstrual symptoms), and products that could appeal to women by advertising specific health benefits (e.g., depression, nausea) that can be indirectly linked to areas specific to women-centric health concerns (e.g., post-partum depression, morning sickness). While limited studies report reductions in menstrual-related symptoms following administration of oral CBD or CBD-infused tampons, many products identified in this study for menstrual symptoms are meant for topical administration and may not produce desired effects. Women undergoing the menopausal transition, especially those with multiple medical conditions and/or severe symptomatology, have also demonstrated openness to trying cannabinoid products. However, epidemiological studies provide evidence that cannabis use (specifically products containing THC and often consumed via inhalation) among women may be associated with cardiovascular, female reproductive health (e.g., disruption of hormonal and menstrual cycles), and depression and anxiety risks. 24 , 36 – 39 This study also identified a small percentage of CDP listings with health benefit claims that also included characterizing flavors (e.g., fruit and coffee/espresso flavors) that may appeal to specific consumer demographics, such as younger women. 40 – 42 Hence, more research is needed to understand potential health benefits and risks of cannabis-derived products for women’s health matters and variability by product features (e.g., product formulation, cannabinoid content, characterizing flavor, etc.). Additionally, we found that only 6.6% of online product listings included disclaimers for women who are pregnant or breastfeeding to consult with a physician before use, a concerning finding given known health risks associated with cannabis use among this specific population of women. 40 , 43 , 44 Previous studies addressing cannabis use advice-seeking among women have identified barriers that occur in the healthcare environment and could be impacted by exposure to commercial health benefit claims. These include stigma surrounding cannabis use, healthcare providers’ (HCPs’) lack of knowledge to advise patients or patients’ perception of such, HCPs’ avoidance or reluctance about advising on cannabis use, and different state laws prohibiting or allowing HCPs to authorize medical cannabis use. 5 , 43 , 45 – 50 Surveys have also found that women with a history of cannabis use may be more likely to continue use during pregnancy, leading to calls for more preventive approaches to address prenatal cannabis use. 51 , 52 Hence, future cannabis advice should take into consideration the potential impact on cannabis uptake and use behavior pertaining to various claims for treating women’s health concerns that require further evidence to substantiate. 52 , 53

Study findings suggest that health information communicated through product marketing may be confusing to women consumers, as purported health claims are not supported by a strong body of published literature. 41 , 42 This underscores the need for more comprehensive regulations that can better address existing and emerging health benefit claims and misleading marketing practices targeted towards women. The implications of this study’s findings are also comparable to recommendations from the 2024 National Academies of Sciences, Engineering, and Medicine report on cannabis and public health, which specifically called for conducting research and developing best practices for marketing restrictions that could address women’s health topics. 54 Findings also support the need to regulate cannabis e-commerce platforms given the national digital reach of their marketing content.

This study has several limitations, including limited generalizability to the broader cannabis retail marketplace, as not all licensed dispensaries in the U.S. list or sell products on Leafly and Weedmaps and they do not differentiate products as medical or adult-use. The time frame of the study conducted in late 2023 further limits generalizability to changes that occur with new and emerging CDPs and marketing tactics and messaging. Additionally, data collected for this study consisted primarily of cannabis products that did not represent themselves as used in the context of a state medical cannabis program or dispensed with a doctor’s authorization. Furthermore, we did not control, normalize, or otherwise adjust for vendors that may have had multiple product listings due to the relatively low average of listings per vendor that met our study inclusion criteria and due to our focus on understanding the entirety of types and characteristics of claims made. Future studies should explore potential differences in how medical cannabis dispensaries, adult-use dispensaries, and dispensaries that serve both types of customers market products in the context of health benefit claims. The filtered keywords used to detect promoted health benefits in product descriptions may not be comprehensive as we used general women’s health terms and standardized medical terminology from MedDRA for purposes of keyword filtering and some fertility-related terms can also apply to both women’s and men’s health and reproductive topics. Future studies should include a broader set of terms, including those used more colloquially for health benefits, in conjunction with targeted market surveillance for claims that may be new or novel. Additionally, we were unable to conduct content analysis of health benefit claims based on cannabinoid content (e.g., CBD, delta-8 THC, THC concentrations, etc.) due to lack of standardization of product descriptions and inconsistency of terminology across product listings observed. Lack of standardization of CDP descriptors has also been highlighted as a challenge in another market surveillance study and should be improved in future studies or leverage different commercial cannabis datasets (e.g., point-of-sales data). 31 Further, given the small sample size of consumer reviews, further studies should also conduct statistical analysis and explore consumer perceptions and attitudes using larger datasets to provide more comprehensive insights into cannabis user perceptions and attitudes towards health claims and their purported benefits. Finally, our hierarchical classification framework, while designed to provide a life-course perspective on female reproductive health, has inherent limitations. The distribution of health claims was heavily skewed toward menstruation cycle topics, which may limit the interpretive balance across women’s health categories. While our framework was informed by both existing literature and inductive analysis, we acknowledge that alternative classification approaches may yield different insights. These limitations are important to consider when interpreting trends across the coded health claims and should be adopted in future studies with larger datasets.

In 2023, the Substance Abuse and Mental Health Services Administration (SAMHSA) reported that 19.5 million women aged 12 and older had used marijuana in the previous month, reflecting an overall rise in its use among the general population. 1 An increasing number of women in the U.S., where availability and legality vary by state, 2 are using various types of cannabis-derived products (CDPs), generally defined as products with compounds occurring naturally in the Cannabis sativa L. plant, including cannabidiol (CBD) and delta-9 tetrahydrocannabinol (THC). 3 Recent findings from the 2023 Monitoring the Future Survey found that 19- to 30-year-old females had a higher prevalence of past-year cannabis use compared to male respondents, reflecting a shift in consumption that has prompted the cannabis industry to invest in women as a key consumer demographic through rebranding and new products. 4 Studies have reported that women who report use of cannabis for medical purposes are more likely than men to report using cannabis to treat symptoms (e.g., alleviate anxiety, nausea, anorexia, etc.) and have also described women’s preferences for specific cannabis RoAs (e.g., preference for edibles). 5 – 7 Others have discussed medical marijuana use benefits for menopause and for female sexual function. 8 – 11 Additionally, certain at-risk populations, such as pregnant women living with HIV, have increasingly reported use of medical and adult-use cannabis during pregnancy, including in states that have legalized adult use (non-medical). 12 – 14 Generally, though there are more studies examining sex differences (e.g., the biological variations between males and females) of medical marijuana use and product preference, there are far fewer examining cannabis use behavior and therapeutic benefit perception of women associated with adult-use despite growing commercial availability and legalization. 6 In relation to possible benefits and risks associated with cannabis consumption, studies have examined potential health benefits of cannabis on pain management, sleep disorders, and harm reduction for specific substance use behavior. 15 – 20 However, known health harms and adverse events associated with adult-use acute and chronic cannabis use without medical benefit have been reported (e.g., gastrointestinal, cardiovascular, cognition, mood and anxiety disorders). 21 – 24 This includes a growing number of emergency room cases of Cannabis Hyperemesis Syndrome (CHS), occurring alongside a rise in legalization of adult-use cannabis and increases in THC concentration. 25 Cannabis exposure can have health implications across the life course for women, specifically for fertility, pregnancy, and menopause. 26 Additionally, increasing use, evolving risk profiles due to emergence of new and highly concentrated THC products, changing state-level policies, and growing commercial availability of adult-use (and hemp) products has created a diverse CDP marketplace. A good proportion of this marketplace occurs online through large cannabis e-commerce platforms such as Dutchie, Leafly, Weedmaps, Jane, and other online cannabis product brands and websites that connect consumers directly to dispensaries in various states. Several published reports describe accounts of dispensaries and online sellers marketing adult-use cannabis products with unverified health claims to treat specific diseases or health conditions. 27 Specifically, prior studies have examined health benefit claims of CDPs occurring on the Internet (e.g., from Internet articles, search engine results, social media posts) and on websites operated by cannabis dispensaries and found several health or disease-related claims, including for the treatment of pain, spasms, anxiety, depression, insomnia, and nausea/vomiting. 15 , 28 , 29 Despite growing research regarding sex differences in cannabis use and growing numbers of purported health benefit claims made for CDPs, no study to our knowledge has examined the characteristics of health benefit claims made on cannabis e-commerce websites specific to women’s reproductive health issues. In order to fill this research gap, the objective of this study was to characterize CDP listings that include health claims and their associated consumer reviews and health warning disclaimers on two major e-commerce platforms with a focus on nonmedical adult-use (i.e., adult-use) cannabis products given the data set utilized. The aim of this study is also to generate market surveillance insights that can inform future cannabis regulatory science.