Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jun 24.
Published in final edited form as: Health Educ Res. 2026 Mar 31;41(3):cyag013. doi: 10.1093/her/cyag013

Effects of textual warnings and warning symbols on cannabis advertisements on product perceptions among US young adults

Yuxian Cui 1,*, Shriya Thakkar 1, Rishika Chakraborty 2, Morgan B Speer 1, Cassidy R LoParco 1, Hermela Amdeta 1, Darby Mihelich 1, Diane Mukundwa Ndisebuye 1, Patricia A Cavazos-Rehg 3, Carla J Berg 1,4,*
PMCID: PMC13287986  NIHMSID: NIHMS2187218  PMID: 42054015

Abstract

States with legalized non-medical cannabis require different warnings on packages and advertisements; yet, studies examining effects of textual warnings and different warning symbols on young adults’ perceptions are limited. In 2024, US young adults ages 18–34 (n = 3 579) completed an online survey-based experiment using a 2 × 4 factorial design (textual warning: no, yes; warning symbol: none, leaf, THC, leaf + THC), then reported on 5 perception outcomes: addictiveness, harm, cautiousness, appeal, and interest. Multivariable regressions assessed these conditions and their interactions in relation to outcomes, overall and by past-month cannabis use status. Among participants without past-month use, the THC symbol (versus none) was associated with higher perceived addictiveness (B = 0.32, SE = 0.12), harm (B = 0.23, SE = 0.12), and cautiousness (B = 0.31, SE = 0.13); the textual warning was associated with greater appeal (B = 0.18, SE = 0.09). No effects were found among participants reporting cannabis use. An interaction between textual warning × leaf symbol on perceived harm (B = 0.35, SE = 0.18) indicated that exposure to both the textual warning and leaf symbol decreased perceived harm; exposure to only one (i.e. not together) increased perceived harm. Regulatory bodies must understand effects of different warning requirements, consider requiring the THC symbol, assess potential confusion of the leaf symbol, and explore warnings targeting different subgroups.

Introduction

As of September 2025, 24 states and Washington, DC have legalized non-medical (i.e. ‘recreational’) cannabis [1]. The past decade has also marked increases in cannabis use, particularly in young adults [2–4]. From 2013 to 2022, cannabis use prevalence among young adults ages 18–25 and 26–34 increased from 19.47% and 12.54% to 23.03% and 18.94%, respectively [4]. Along with these regulatory changes, cannabis retail has evolved to offer increasingly diverse products [5–7], although combustible cannabis (e.g. flower, bud, dried herb) has been the most prominent mode of use in the US [8, 9].

Cannabis advertising, ranging from point-of-sale promotions and social media content to product packaging, has been widely used to enhance the appeal of cannabis products, communicate purported ‘benefits,’ and shape pro-cannabis attitudes [10–14]. Such marketing strategies frequently highlight health-related claims, product effects, and other persuasive messages intended to influence consumer decision-making [10–14].

Many states with legalized non-medical cannabis require cannabis products and/or advertisements to include textual warnings and warning symbols indicating that the product contains tetrahydrocannabinol (THC) or cannabis [15]. Notably, states have adopted different sets of textual health warnings on products and advertisements [16]. For example, many states, such as Washington, require textual warnings communicating risks of impaired concentration (e.g. ‘Cannabis can impair concentration, coordination, and judgment. Do not operate a vehicle or machinery under the influence of this drug’) and use among youth (e.g. ‘For use only by adults 21 and older. Keep out of the reach of children’) [16]. Additionally, many states have adopted warning symbols featuring the cannabis leaf or large block letters indicating THC [16].

Given that states differ in their required warnings, it is crucial to understand the effects of different approaches to guide future regulations. Textual warnings have generally been shown to be effective in reducing product appeal [17], and warning symbols may effectively communicate risks about product content and harm, especially for certain populations (e.g. younger or those with low education), which has been documented in US- and Canada-based studies of the THC and leaf symbols [17] as well as in the tobacco literature examining effects of such imagery and pictorial warnings [18]. However, the effectiveness of required warnings, particularly warning symbols, might not be as intended and might not be the same across subgroups. For example, while one qualitative study among individuals who used cannabis suggested that the cannabis leaf warning symbol effectively indicated that products contained cannabis [19], other studies have shown that many individuals who did not use cannabis perceived the meaning of Colorado’s required THC symbol to be unclear [20] or that some individuals do not recognize the cannabis leaf as a symbol representing cannabis [21].

The literature regarding efforts to convey warnings on cannabis products and advertisements is nascent, with several considerable gaps. First, limited research has assessed the effects of different warning symbols. Second, the effects of different warning symbols in the context of textual warnings have been understudied; e.g. it is plausible that the warning symbols may reinforce or undermine the textual warnings. Finally, there is relatively limited research assessing how individuals or subgroups of individuals (e.g. those who use cannabis) differ in how they perceive different health warnings on advertisements.

Investigating these issues is timely given the expansion of cannabis retail and marketing and the rapid evolution of the cannabis regulatory context [22]. Thus, this study examined the impact of different strategies to communicate warnings on advertisements for cannabis flower. We conducted an online survey-based experiment using a 2 × 4 factorial design, specifically randomizing young adults to 1 of 2 textual warning conditions (no textual warning, textual warning) alongside 1 of 4 warning symbols (none, leaf, THC, leaf + THC) [16, 23]. We assessed the effects of the warning conditions on 5 outcomes—perceived addictiveness, harm, cautiousness, appeal, and interest—factors that influence use among young adults [24–26]. We also assessed interactions between textual warning condition and warning symbol condition, as well as between each warning condition and past-month cannabis use status.

Methods

Study design

This study analysed survey data from a study of 4 031 young adults across all 50 states and DC who participated in the Cannabis Regulation, Marketing & Appeal (CARMA) study, a longitudinal study examining non-medical cannabis laws, marketing, and use (approved by the George Washington University Institutional Review Board) [27]. This parent study, launched in 2023 (baseline), involves 5 waves of online survey data collection, conducted approximately 6 months apart. The current study analyzes experimental data collected during the Wave 3 (W3; fall 2024) survey (detailed further below).

Participants and recruitment

In June–November 2024, advertisements targeting eligible young adults (English-speaking US residents ages 18–34, i.e. the age group with the highest cannabis use prevalence [8]) were posted on Facebook. After clicking on ads, individuals were messaged via chatbot on Facebook Messenger, which ensured each individual had a verified Facebook account and precluded individuals from interacting with the chatbot more than once. The chatbot provided an abbreviated study overview, assessed key factors (age, state of residence, race, ethnicity, sex, past-month cannabis use), and provided individuals deemed preliminarily eligible a unique link to the study webpage that expired after a single use, thus prohibiting individuals to share the link or use bots to submit multiple surveys. Once potential participants clicked the individual link and arrived at the study webpage, formal consent was obtained, eligibility was confirmed, and the baseline survey was administered. Participants were told that the study required providing valid email addresses and phone numbers and confirming their participation by clicking a link in an email sent 7 days post-baseline survey (allowing research staff to verify contact information and check survey responses for logical responses, etc.). After participants confirmed their participation, they received their incentive ($10 Amazon e-gift card). Purposive, quota-based recruitment was used to ensure representation of key subgroups (i.e. ~ 50% past-month cannabis use, ~ 50% males and females, ~ 40% racial/ethnic minorities). Recruitment also sought representation across all states and regions, as well as ~ 50% in states with legal non-medical cannabis.

Overall, 6 908 individuals completed chatbot pre-screening, 6 128 (88.7%) were preliminarily eligible, 5 827 (95.6%) visited the study webpage, 5 672 (97.3%) were consented and eligible, 4 385 (77.3%) completed the survey and were sent confirmation emails, and 4 031 (91.9%) confirmed participation and were officially enrolled in the study. Current analyses focused on participants who completed the W3 survey (n = 3 588, 89.0% retention); the analytic sample included with complete data on all measures used in these analyses (n = 3 579).

Experimental conditions: advertising messages and warnings

Participants were randomized to 1 of 8 advertisements based on a 2 × 4 factorial design—with 2 textual warning conditions and 4 warning symbol conditions (Table 1). The experimental conditions were based on the most common required text warnings and warning symbols across states with legal non-medical cannabis retail [16, 23]. The 2 textual warning conditions were: (i) no textual warning (control) and (ii) textual warning (‘WARNING: CANNABIS CAN IMPAIR CONCENTRATION, COORDINATION, AND JUDGMENT. FOR USE ONLY BY ADULTS 21 YEARS OF AGE OR OLDER. KEEP OUT OF THE REACH OF CHILDREN’). Given that most states that had warning symbol requirements used symbols integrating the cannabis leaf, block letters indicating ‘THC,’ or both (Supplementary Fig. 1) [23], we chose the following 4 warning symbol conditions: (i) no symbol; (ii) leaf symbol; (iii) THC symbol; and (iv) leaf + THC symbol. The advertisement was adapted from real-world examples and intentionally chosen to represent common advertisement designs (i.e. largely featuring the product itself and a brief branded headline) and to represent a brand unfamiliar to most participants (i.e. based in Alaska, where only 4 participants resided) to minimize potential brand-specific influences (e.g. pre-existing positive attitudes) [28, 29].

Table 1.

Experimental conditions: cannabis advertisements featuring no textual versus textual warning and with no, leaf, THC, or leaf + THC warning symbols.

Textual warning condition
Warning symbol condition No textual warning Textual warning (‘Warning: cannabis can impair concentration, coordination, and judgement. For use only by adults 21 years of age or older. Keep out of the reach of children.’)
No symbol

graphic file with name nihms-2187218-f0001.jpg

graphic file with name nihms-2187218-f0002.jpg

Leaf symbol

graphic file with name nihms-2187218-f0003.jpg

graphic file with name nihms-2187218-f0004.jpg

THC symbol

graphic file with name nihms-2187218-f0005.jpg

graphic file with name nihms-2187218-f0006.jpg

Leaf & THC symbol

graphic file with name nihms-2187218-f0007.jpg

graphic file with name nihms-2187218-f0008.jpg

Open in a new tab

Measures

Product perceptions

While viewing the advertisement, participants were asked, ‘Based on this ad: (i) How addictive do you think these products are? (ii) How harmful to your health do you think these products are? (iii) How cautious would you be if you were to try or use these products? (iv) How appealing do you think these products are? (v) How interested are you in trying or using these products?’ (0 = not at all to 7 = extremely).

Cannabis use

Prior to assessing the advertisements, participants were asked, ‘In the past 30 days, how many days did you use marijuana?’ [8] (instructions stipulated to exclude the use of derived cannabis products, e.g. delta-8 THC). Responses were operationalized as no versus yes past-month use. Participants reporting past-month cannabis use also were asked to indicate the type of product they used most often, which was recoded as: (i) edibles, (ii) bud/dried herbs, (iii) vaporized (cannabis oils or liquids for vaping), and (iv) others (cannabis oils or liquids taken orally, tinctures, concentrates, hash or kief, and topicals).

State non-medical cannabis law and sociodemographics

Participants reported their state of residence (coded as legalized versus not legalized non-medical cannabis as of January 2025) [30], as well as age, birth sex, sexual orientation, race, ethnicity, and education level.

Data analysis

First, descriptive and bivariate analyses were conducted to characterize participants overall and by past-month cannabis use status. Second, reports of perceived addictiveness, harm, cautiousness, appeal, and interest were examined overall and by the textual warning condition and warning symbol condition to which they were assigned—among all participants, those reporting past-month cannabis use, and no past-month use. Then, 5 multivariable regression analyses were conducted to assess textual warning and warning symbol condition in relation to each outcome (perceived addictiveness, harm, cautiousness, appeal, and interest). Models were built by first entering state non-medical cannabis legalization, sociodemographics, and past-month cannabis use (block 1), main effects of the textual warning and warning symbol conditions (block 2), and interaction effects among the textual warning and warning symbol conditions (block 3). Similarly, these 3 blocks were used to assess interactions between past-month use status and each warning condition. We also conducted similar analyses stratified by past-month use status. All analyses were conducted SPSS.v27, and significance was set at P < .05.

Results

Participant characteristics

Shown in Table 2, the sample (n = 3 581) was 26.35 (SD = 4.82) years old on average, 60.3% female, 27.6% sexual minority, 12.8% Black, 14.0% Asian, 6.6% other race(s), 18.9% Hispanic, and 54.5% < bachelor’s degree educated. Overall, 40.8% reported past-month cannabis use, with 14.38 (SD = 11.57) days used in the past 30 days on average and 55.9% using bud/dried herb most often. Participants reporting past-month use (versus no use) were older and more likely sexual minority, Black, Hispanic, and < bachelor’s degree educated (P’s < .001).

Table 2.

Participant characteristics overall and by past-month cannabis use status.

Past-month cannabis use
Total
n = 3579
(100.0%)		 No
n = 2118
(59.2%)		 Yes
n = 1461
(40.8%)
Variables M (SD)
or n (%)		 M (SD)
or n (%)		 M (SD)
or n (%)		 P
Legalized non-medical cannabis, n (%) Sociodemographics 2054 (57.4) 1 159 (54.7) 895 (61.3) < .001
Age, M (SD) 26.35 (4.82) 25.99 (4.90) 26.86 (4.66) < .001
Female, n (%)a 2 160 (60.3) 1 268 (69.9) 892 (61.5) .534
Sexual minority, n (%)a 990 (27.6) 469 (22.1) 521 (36.1) < .001
Race, n (%)a < .001
 White 2 251 (62.9) 1 340 (63.3) 911 (64.8)
 Black 457 (12.8) 195 (9.2) 262 (18.6)
 Asian 503 (14.0) 387 (18.3) 116 (8.3)
 Otherb 237 (6.6) 121 (5.7) 116 (8.3)
Hispanic, n (%)a 677 (18.9) 359 (16.9) 318 (22.1) < .001
Education <Bachelor’s degree, n (%) 1951 (54.5) 1 002 (47.3) 949 (65.0) < .001
Cannabis use characteristics
Number of days useda - - 14.38 (11.57) < .001
Form of cannabis used most oftena < .001
 Edibles - - 284 (19.5)
 Bud/dried herb - - 812 (55.9)
 Vaporized - - 215 (14.8)
 Otherc - - 142 (9.8)
Open in a new tab

Notes: M = Mean; SD=Standard deviation.

a

Refuse, Unsure or Don’t know: Sex (n = 24), Sex Orientation (n = 79), Race (n = 125), Ethnicity (n = 53), Forms of cannabis used most often among people with past-month cannabis use (n = 8).

b

Other race includes American Indians or Alaska Natives, Native Hawaiian or Pacific Islanders and Multiracial.

c

Other types of cannabis used most often include cannabis oils or liquids taken orally, tinctures, concentrates, hash or kief, and topicals.

Effects of warning message and symbol condition on perceptions

Supplementary Table 1 shows bivariate results assessing associations between each of the conditions (no textual versus textual warning; no, leaf, THC, and leaf + THC warning symbols), which showed no associations between condition and any of the perception outcomes. Table 3 shows results from the multivariable linear regression models, which showed no effect of the textual warning on any of the perception outcomes. Additionally, while the leaf and leaf + THC warning symbols showed no effects on the perception outcomes, the THC warning symbol (versus no symbol) was associated with higher perceived addictiveness (B = 0.20, SE = 0.09), harm (B = 0.16, SE = 0.08), and cautiousness (B = 0.20, SE = 0.96). However, stratified analyses stratified by past-month cannabis use status showed that these effects were driven by participants reporting no past-month cannabis use. Specifically, among participants reporting no past-month use, the THC symbol was associated with higher perceived addictiveness (B = 0.32, SE = 0.12), harm (B = 0.23, SE = 0.12), and cautiousness (B = 0.31, SE = 0.13), and the textual warning was associated with greater appeal (B = 0.18, SE = 0.09); these effects were not found among participants reporting past-month cannabis use (see Table 3 footnote). Additionally, among those reporting no past-month use, exposure to the textual warning was associated with greater appeal (B = 0.18, SE = 0.09).

Table 3.

Linear regression results of experiment exposing participants to cannabis advertisements with no textual versus textual warnings and with no, leaf, THC, or leaf + THC warning symbols in relation to perceived addictiveness, harm, cautiousness, appeal and interest.

Variables Addictiveness Harm Cautiousness Appeal Interest
B (SE) B (SE) B (SE) B (SE) B (SE)
Past-month cannabis use (Ref: nonuse) * −0.85 (0.07)* −1.05 (0.07)* −1.32 (0.08)* 1.40 (0.07)* 2.06 (0.07)*
Legalized non-medical cannabis (ref: illegal) Sociodemographics 0.10 (0.07) 0.19 (0.06)* −0.05 (0.07) −0.04 (0.07) −0.13 (0.07)*
Age −0.04 (0.01)* −0.04 (0.01)* −0.04 (0.07)* 0.03 (0.01)* 0.04 (0.01)*
Male (Ref: female) −0.22 (0.07)* −0.23 (0.07)* −0.22 (0.07)* −0.04 (0.07) 0.10 (0.07)
Sexual minority (Ref: heterosexual) −0.46 (0.08)* −0.40 (0.07)* −0.29 (0.08)* 0.02 (0.08) 0.02 (0.07)
Race (Ref: White)
 Black 0.71 (0.10)* 0.41 (0.10)* 0.20 (0.11) 0.39 (0.10)* 0.39 (0.10)*
 Asian 0.39 (0.11)* 0.60 (0.10)* 0.34 (1.11)* 0.13 (0.11) −0.10 (0.10)
 Other 0.06 (0.11) −0.06 (0.10) 0.00 (0.12) 0.18 (0.11) 0.18 (0.11)
Hispanic (Ref: no) 0.26 (0.09)* 0.24 (0.09)* 0.06 (0.10) −0.01 (0.09) −0.14 (0.09)
Education <bachelor’s degree (Ref: >bachelor’s) −0.36 (0.07)* −0.37 (0.07)* −0.26 (0.08)* 0.07 (0.07) 0.16 (0.07)*
Experimental conditions
Textual warning (Ref: no textual warning) Textual warning 0.06 (0.07) 0.04 (0.06) −0.03 (0.07) 0.09 (0.07) 0.05 (0.06)
Warning symbol (Ref: no warning symbol)
 Leaf symbol 0.08 (0.10) 0.09 (0.09) 0.15 (0.10) −0.04 (0.10) 0.06 (0.09)
 THC symbol 0.20 (0.09)* 0.17 (0.08)* 0.20 (0.10)* −0.03 (0.09) 0.00 (0.09)
 Leaf + THC symbol 0.04 (0.09) 0.01 (0.09) 0.09 (0.10) 0.04 (0.09) 0.05 (0.09)
 Adjusted R-square .094 .136 .121 .127 .256
Open in a new tab

Notes: B = Unstandardized Beta; SE = Standard Error; 95%CI = 95% Confidence Interval; Ref = Referent group. See Table 1 for warning conditions. Statistically significant:

*

P < .05. No significant interactions between past-month use status and warning conditions or between textual and symbol warning conditions, except textual warning × leaf symbol on perceived harm (B = 0.35, SE = 0.18, p = .049). In analyses stratified by past-month use status, the associations between THC warning symbol and perceived addictiveness, harm, and cautiousness were found among participants reporting no past-month use (B = 0.32, SE = 0.12, P = .008; B = 0.23, SE = 0.12, P = .048; B = 0.31, SE = 0.13, P = .017) but not among those reporting past-month use. Additionally, among those reporting no past-month use, exposure to the textual warning was associated with greater appeal (B = 0.18, SE = 0.09, P = .045).

Furthermore, there was a significant interaction between textual warning × leaf warning symbol condition on perceived harm (B = 0.35, SE = 0.18; Fig. 1), such that exposure to both the textual warning and the leaf symbol resulted in lower perceived harm, while exposure to either the textual warning or the leaf symbol (i.e. not in combination) resulted in higher perceived harm. No interactions between textual warning × symbols were found in stratified analyses.

Figure 1.

Figure 1.

Open in a new tab

Interaction between textual warning and leaf warning symbol conditions on perceived harm. Note: In analyses stratified by textual warning condition (controlling for past-month use, legal status, and sociodemographics), the lower perceived harm resulting from the textual warning with leaf symbol versus without leaf symbol was significant (P = .047), while the higher perceived harm resulting from no textual warning with leaf (M = 3.70) versus without leaf (M = 3.63) was not. In analyses stratified by leaf warning versus no leaf warning, the differences by textual warning condition were also not significantly different in either adjusted model.

Sociodemographic correlates of outcomes

Shown in Table 3, factors associated with lower perceived addictiveness included past-month cannabis use and being older, male (versus female), sexual minority (versus heterosexual), White (versus Black and Asian), non-Hispanic (versus Hispanic), and < bachelor’s degree educated (versus ≥ bachelor’s degree); factors associated with lower perceived harm included past-month use, living in non-legalized states, being male, sexual minority, White (versus Black and Asian), non-Hispanic, and < bachelor’s degree educated; and factors associated with lower cautiousness were past-month use and being older, male, sexual minority, and < bachelor’s degree educated. Factors associated with higher appeal included past-month use and being older and Black (versus White), and factors related to higher interest were past-month use, living in non-legalized states, and being older, Black (versus White), and < bachelor’s degree educated.

Discussion

In this sample of US young adults, there were key differences in the effects of warnings symbols and textual warnings on those reporting past-month cannabis use versus no use. Notably, only one warning symbol—the THC symbol—was shown to have the anticipated effects on perceived addictiveness, harm, and cautiousness; these effects only held for young adults reporting no past-month cannabis use (not those reporting cannabis use). One potential explanation, as suggested by prior research [31], is that those with previous knowledge about or experience with about a chemical (i.e. THC in this study) may perceive lower risk or less need to exercise caution, given their existing knowledge and experiences, and thus are less sensitive to such warning cues [31]. For those unfamiliar with THC, the THC warning symbol may have a unique effect in communicating product ingredients, potentially leading to increased perceived risk and need for caution [32].

Interestingly, analyses among all participants indicated that exposure to the combination of the textual warning and leaf symbol was associated with reduced perceived harm (albeit a small effect). This may echo previous tobacco and cannabis research that indicated that the leaf symbol may be perceived as an advertising element [21, 33], rather than a warning, particularly when other branding elements are presented [34] or when a textual warning is already displayed. This is concerning, especially for young adults with less knowledge about the risks of cannabis, including disproportionately-impacted subgroups [35], as they may be more likely to misidentify the cannabis leaf and misperceive the risks of using, which may lead to unintentional consumption or overconsumption [21]. Given that this experimental study could not explicitly shed light on the mechanisms underlying this finding, qualitative research is needed to understand how young adults interpret or understand leaf warning symbols on packaging and advertisements.

Another important finding was that, among those who did not use cannabis, exposure to the textual warning was associated with increased product appeal, contrary to previous research [17]. There are several possible reasons for this. One is that the content of the warning was not compelling (i.e. cannabis can impair certain functions, adult-use only, keep from children), as these may be warnings already well understood— and the adult-use and child warnings are common across several products outside of cannabis, thus normalizing risks among those who do not use cannabis [36]. Another is that the simple and health-related ad content was more appealing and novel to those who did not use cannabis [37], and thus drew more attention than the long warning statement with 3 phrases. Relatedly, heuristic processing may have played a role [38, 39], such that, rather than processing the risks included in the warning, individuals interpreted the presence of a warning as communicating regulatory oversight or legitimacy. This phenomenon was identified among individuals reporting no tobacco use who were exposed to e-cigarette warnings and generally failed to fully process risk information, especially when multiple warnings or messages were presented [40]. Given that those who do not use cannabis—or do not use frequently—may initiate use or become more frequent users in response to advertised effects of cannabis (i.e. psychological effects) rather than for fun or curiosity [41], future studies are needed to understand why textual warnings may increase appeal among those who do not use cannabis.

Several findings regarding sociodemographic differences are important to note. Echoing previous research indicating that sexual minorities, males, Black, and < bachelor’s degree educated individuals are more likely to use cannabis or use more cannabis [42–50], this study found that these subgroups reported lower risk perceptions and greater appeal. Although previous research suggests that those ages 18–25 are more likely to use cannabis [4], this study found that being younger was associated with greater risk perceptions and lower appeal. Additionally, being non-Hispanic was associated with lower perceived addictiveness and harm, although cannabis use rates among this subgroup are generally lower than other racial/ethnic subgroups [49, 51]. Notably, while most research indicates recreational cannabis legalization is associated with past-month cannabis use [52], findings from this study indicated that people living in legalized states reported greater perceived harm and lower interest in response to the warnings. This might highlight the roles and opportunities of anti-cannabis efforts (e.g. health warnings, educational campaigns) following legalization [53], but invite future research to understand proactive factors that lead to decisions to use despite warnings. Together, these findings suggest that future studies are needed to replicate these findings and assess specific mechanisms underpinning the effects of different combinations of warning elements among subgroups of young people.

Current findings have implications for research and practice. First, given that we found that a certain combination of warning elements has interactive effects, it is important to examine other warning element combinations (e.g. colours, shapes) [54] to understand the broader range of potential interactions and their mechanisms, as in other cannabis and tobacco research [55–57]. Second, the fact that THC symbols were effective but showed differential impacts by use status highlights the need to investigate potential mechanisms to inform the development of interventions to reduce such risks and disparities. Finally, regulatory, advocacy, and educational efforts must educate consumers, particularly young people, about the effects and risks of cannabis products. Within the current regulatory context where states have different regulations for advertising and warnings, this study may underscore a critical need for enhanced policies on warning (e.g. use of THC symbols), as well as for future studies investigating the mechanisms that maximize the impact of a specific warning combination against advertising for a particular subgroup [34, 58–60].

Limitations

This study is limited in generalizability, given social-media based recruitment and purposive sampling of ~ 50% young adults reporting past-month cannabis use. While our recruitment process was efficient and integrated several fraud prevention features, we cannot rule out fraud (although we have maintained high retention rates and communication records with participants), and this approach also involved limitations in terms of reach to those with digital access barriers, limited our ability to respond in real-time to questions potential participants may have had, and may have introduced additional sampling biases [61]. To mitigate these concerns, we implemented several verification measures, including unique digital ID checks and phone/email confirmation, to reduce the likelihood of fraudulent participation. Additionally, data were cross-sectional and self-reported, precluding causal inference and preventing examination of temporal relationships between warning conditions and actual use patterns.

Thus, future research using longitudinal data, more representative samples, and more objective behavioural measures is needed. Also, as previously noted, this study only assessed the effects of certain warning elements, chosen based on the prominence of such advertising themes and the number of states requiring certain warnings [16] and the literature regarding the effectiveness of certain warnings [17]. However, additional research is needed to assess a broader range of warnings as well as their combinations on different types of ads.

Conclusions

Among those who did not use cannabis, it was found exposure to the THC warning symbol was associated with higher risk perceptions. While the exposure to the leaf symbol or the textual warning increased perceived harm among all participants, the exposure to the combination reduced perceived harm. Furthermore, certain sociodemographic subgroups perceived more risk and/or lower appeal than their counterparts, highlighting the need to investigate the differential effects of these warning elements and perceptions that influence young adults’ use-related decisions and disparities.

Supplementary Material

suppl materials

Supplementary data is available at HEAL online.

Funding

This work was supported by the US National Institute on Drug Abuse (R01DA054751, MPIs: Berg, Cavazos-Rehg) and the National Science Foundation (#2415890, Co-PIs Cui & Berg).

Footnotes

Conflicts of interest

None declared.

Ethics approval

This study was approved by the George Washington University Institutional Review Board (NCR224124). All participants provided informed consent.

Declaration of generative AI and AI-assisted technologies in the writing process

AI was not used in preparing this manuscript.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.

References

  • 1.Elliott DJ, Marissa Y. Where is weed legal? A guide to marijuana legalization. U.S. News. 2025. Available from: https://www.usnews.com/news/best-states/articles/where-is-marijuana-legal-a-guide-to-marijuana-legalization [Google Scholar]
  • 2.Health, N.I.o. Marijuana and Hallucinogen Use among Young Adults Reached all-Time High in 2021. Washington: Federal Information & News Dispatch, LLC, 2022. [Google Scholar]
  • 3.Youth Risk Behavior Surveillance System (YRBSS). High School YRBS: 2019 Results. Atlanta, G A: Centers for Disease Control and Prevention; 2021. Available at: https://www.cdc.gov/healthyyouth/data/yrbs/index.htm (23 March 2026, date last accessed). [Google Scholar]
  • 4.Mattingly DT, Richardson MK, Hart JL. Prevalence of and trends in current cannabis use among US youth and adults, 2013–2022. Drug and Alcohol Dependence Reports 2024;12:100253. 10.1016/j.dadr.2024.100253 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Rup J, Goodman S, Hammond D. Cannabis advertising, promotion and branding: differences in consumer exposure between ‘legal’and ‘illegal’markets in Canada and the US. Prev Med 2020;133:106013. 10.1016/j.ypmed.2020.106013 [DOI] [PubMed] [Google Scholar]
  • 6.Birdoes H Association between cannabis retail outlet availability and sources of cannabis acquisition among young adults in Washington State, 2015–2021. Seattle (WA): University of Washington; 2025. Available from: https://digital.lib.washington.edu/researchworks/items/77cd0f2f-3722-4de8-947d-38ac2b474987 [Google Scholar]
  • 7.Hyatt AS, Overhage L, Lê Cook B. Use of tobacco and cannabis following state-level cannabis legalization. JAMA Netw Open 2025;8:e2520093–3. 10.1001/jamanetworkopen.2025.20093 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Substance Abuse and Mental Health Services Administration. Key Substance Use and Mental Health Indicators in the United States: Results from the 2023 National Survey on Drug Use and Health. Rockville, MD: Center for Behavioral Health Statistics and Quality, Substance Abuse and Mental Health Services Administration. Retrieved from: https://www.samhsa.gov/data/release/2023-national-survey-drug-use-and-health-nsduh-releases#detailed-tables, 2024. [Google Scholar]
  • 9.Inman A, Cservenka A. Cannabis flower, concentrates, and edibles: a narrative review comparing prevalence of use, methods of consumption, and cannabis use disorder outcomes. J Addict Dis 2024;44:3–13. 10.1080/10550887.2024.2418225 [DOI] [PubMed] [Google Scholar]
  • 10.Cohn AM, Alexander AC, Ehlke SJ et al. Seeing is believing: how cannabis marketing exposure is associated with cannabis use attitudes and behavior in a permissive medical cannabis policy environment. Am J Addict 2023;32:333–42. 10.1111/ajad.13390 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Whitehill JM, Trangenstein PJ, Jenkins MC et al. Exposure to cannabis Marketing in Social and Traditional Media and past-year use among adolescents in states with legal retail cannabis. J Adolesc Health 2020;66:247–54. 10.1016/j.jadohealth.2019.08.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Willoughby JF, Hust SJT, Li J et al. Exposure to pro and anti-cannabis social media messages and teens’ and college students’ intentions to use cannabis. Health Commun 2024;39:183–94. 10.1080/10410236.2022.2162707 [DOI] [PubMed] [Google Scholar]
  • 13.Han B, Shi Y. Associations of recreational cannabis dispensaries’ availability, storefront signage and health benefit signs with cannabis use: findings from a representative adult sample in California. United States Addiction 2023;118:1270–9. 10.1111/add.16132 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Kowitt SD, Yockey RA, Lee JGL et al. The impact of cannabis packaging characteristics on perceptions and intentions. Am J Prev Med 2022;63:751–9. 10.1016/j.amepre.2022.04.030 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Schauer GL. Cannabis policy in the United States: implications for public health. J Natl Cancer Inst Monogr 2021;2021:39–52. 10.1093/jncimonographs/lgab016 [DOI] [PubMed] [Google Scholar]
  • 16.Prescription Drug Abuse Policy System (PDAPS). Retail Marijuana Laws dataset. 2024. Available at: https://pdaps.org/datasets/retail-marijuana-laws (23 March 2026, date last accessed).
  • 17.Massey ZB, Hammond D, Froeliger B. A systematic review of cannabis health warning research. Prev Med Rep 2024;37:102573. 10.1016/j.pmedr.2023.102573 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Orenstein DG, Glantz SA. Regulating cannabis manufacturing: applying public health best practices from tobacco control. J Psychoactive Drugs 2018;50:19–32. 10.1080/02791072.2017.1422816 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Ranney LM, Clark SA, Meek CJ et al. Perceptions of cannabis warnings and recommendations for improvement: a qualitative study with people who use cannabis from the United States. BMC Public Health 2025;25:2363. 10.1186/s12889-025-23518-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Kosa KM, Giombi KC, Rains CB et al. Consumer use and understanding of labelling information on edible marijuana products sold for recreational use in the states of Colorado and Washington. Int J Drug Policy 2017;43:57–66. 10.1016/j.drugpo.2017.01.006 [DOI] [PubMed] [Google Scholar]
  • 21.Hust SJ, Willoughby JF, Couto L et al. Washington state teens’ perceptions of cannabis-infused product packaging: a qualitative study. J Health Commun 2025;30:238–46. 10.1080/10810730.2025.2514835 [DOI] [PubMed] [Google Scholar]
  • 22.Goodman S, Wadsworth E, Leos-Toro C et al. Prevalence and forms of cannabis use in legal vs. illegal recreational cannabis markets. Int J Drug Policy 2020;76:102658. 10.1016/j.drugpo.2019.102658 [DOI] [PubMed] [Google Scholar]
  • 23.Distru Inc. A state-by-state breakdown of cannabis labeling standards. Distru Blog. 2025. Available at: https://www.distru.com/cannabis-blog/state-by-state-breakdown-of-cannabis-labeling-standards (23 March 2026, date last accessed). [Google Scholar]
  • 24.Romm KF, Wang Y, Ma Y et al. The reciprocal relationships of social norms and risk perceptions to cigarette, e-cigarette, and cannabis use: cross-lagged panel analyses among US young adults in a longitudinal study. Drug Alcohol Depend 2022;238:109570. 10.1016/j.drugalcdep.2022.109570 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Cui Y, LoParco CR, Romm KF et al. How are proand anti-cannabis messaging exposures related to US young adult cannabis use-related factors? Health Educ Res 2024;40:cyae038. 10.1093/her/cyae038 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Wang Y, Romm KF, Edberg MC et al. Two-part models identifying predictors of cigarette, e-cigarette, and cannabis use and change in use over time among young adults in the US. Am J Addict 2024;33:559–68. 10.1111/ajad.13569 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Berg CJ, Romm KF, LoParco CR et al. Young Adults’ experiences with cannabis retailer marketing and related practices: differences among sociodemographic groups and associations with cannabis use-related outcomes. J Racial Ethn Health Disparities 2024;12:2840–52. 10.1007/s40615-024-02092-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Goodman S, Rynard VL, Iraniparast M et al. Influence of package colour, branding and health warnings on appeal and perceived harm of cannabis products among respondents in Canada and the US. Prev Med 2021;153:106788. 10.1016/j.ypmed.2021.106788 [DOI] [PubMed] [Google Scholar]
  • 29.Leos-Toro C, Fong GT, Hammond D. The efficacy of health warnings and package branding on perceptions of cannabis products among youth and young adults. Drug and alcohol review 2021;40:637–46. 10.1111/dar.13240 [DOI] [PubMed] [Google Scholar]
  • 30.DISA Global Solutions. Marijuana legality by state— updated January 1, 2025. 2025. Available at: https://disa.com/marijuana-legality-by-state (5 February 2025, date last accessed).
  • 31.U.S. Environmental Protection Agency, Office of Chemical Safety and Pollution Prevention. The effectiveness of labeling on hazardous chemicals and other products. EPA-HQ-OPPT-2016-0231-0247. Washington, DC: U.S. Environmental Protection Agency; 2016. Available at: https://downloads.regulations.gov/EPA-HQ-OPPT-2016-0231-0247/content.pdf (23 March 2026, date last accessed). [Google Scholar]
  • 32.Ventresca M, Elliott C. Cannabis edibles packaging: communicative objects in a growing market. Int J Drug Policy 2022;103:103645. 10.1016/j.drugpo.2022.103645 [DOI] [PubMed] [Google Scholar]
  • 33.Young WJ, Ganz O, Jeong M et al. Perceptions of game cigarillo packaging among young adult tobacco users: the effect of package color and the “natural leaf” descriptor. Addict Behav 2022;132:107334. 10.1016/j.addbeh.2022.107334 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Goodman S, Leos-Toro C, Hammond D. The impact of plain packaging and health warnings on consumer appeal of cannabis products. Drug Alcohol Depend 2019;205:107633. 10.1016/j.drugalcdep.2019.107633 [DOI] [PubMed] [Google Scholar]
  • 35.Goodman S, Hammond D. Perceptions of the health risks of cannabis: estimates from national surveys in Canada and the United States, 2018–2019. Health Educ Res 2022;37:61–78. 10.1093/her/cyac006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Sillero-Rejon C, Leonards U, Munafò MR et al. Avoidance of tobacco health warnings? An eye-tracking approach Addiction 2021;116:126–38. 10.1111/add.15148 [DOI] [PubMed] [Google Scholar]
  • 37.Cooper M, Shi YY. Packaging of cannabis edibles, health warning recall, and perceptions among Young Adults. JAMA Netw Open 2025;8:e253117–7. 10.1001/jamanetworkopen.2025.3117 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Chaiken S, Maheswaran D. Heuristic processing can bias systematic processing: effects of source credibility, argument ambiguity, and task importance on attitude judgment. J Pers Soc Psychol 1994;66:460–73. 10.1037/0022-3514.66.3.460 [DOI] [PubMed] [Google Scholar]
  • 39.Flynn BS, Worden JK, Bunn JY et al. Evaluation of smoking prevention television messages based on the elaboration likelihood model. Health Educ Res 2011;26:976–87. 10.1093/her/cyr082 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Katz SJ, Erkkinen M, Lindgren B et al. Assessing the impact of conflicting health warning information on intentions to use e-cigarettes-an application of the heuristic-systematic model. J Health Commun 2018;23:874–85. 10.1080/10810730.2018.1533052 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Bray BC, Berglund PA, Evans-Polce RJ et al. A latent transition analysis of self-reported reasons for marijuana use during Young adulthood. Eval Health Prof 2021;44:9–24. 10.1177/0163278720984514 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Parent N, Coulaud PJ, Amirie M et al. Cannabis use and mental health among young sexual and gender minority men: a qualitative study. Int J Drug Policy 2021;91:102980. 10.1016/j.drugpo.2020.102980 [DOI] [PubMed] [Google Scholar]
  • 43.Krueger EA, Fish JN, Upchurch DM. Sexual orientation disparities in substance use: investigating social stress mechanisms in a National Sample. Am J Prev Med 2020;58:59–68. 10.1016/j.amepre.2019.08.034 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Mauro PM, Philbin MM, Greene ER et al. Daily cannabis use, cannabis use disorder, and any medical cannabis use among US adults: associations within racial, ethnic, and sexual minoritized identities in a changing policy context. Prev Med Rep 2022;28:101822. 10.1016/j.pmedr.2022.101822 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Schofield K, Cuttler C, Conner BT et al. Pot at the end of the rainbow: Cannabis use among sexual minorities. Cannabis Cannabinoid Res 2023;8:942–8. 10.1089/can.2022.0240 [DOI] [PubMed] [Google Scholar]
  • 46.Dyar C A review of disparities in cannabis use and cannabis use disorder affecting sexual and gender minority populations and evidence for contributing factors. Current Addiction Reports 2022;9:589–97. 10.1007/s40429-022-00452-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Connolly D, Gilchrist G. Prevalence and correlates of substance use among transgender adults: a systematic review. Addict Behav 2020;111:106544. 10.1016/j.addbeh.2020.106544 [DOI] [PubMed] [Google Scholar]
  • 48.Jeffers AM, Glantz S, Byers A et al. Sociodemographic characteristics associated with and prevalence and frequency of cannabis use among Adults in the US. JAMA Netw Open 2021;4:e2136571. 10.1001/jamanetworkopen.2021.36571 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Substance Abuse and Mental Health Services Administration (SAMHSA). Results from the 2021 National Survey on Drug U se and Health: Detailed tables. Rockville, MD: SAMHSA; 2022. Available at: https://www.samhsa.gov/data/sites/default/files/reports/rpt39441/NSDUHDetailedTabs2021/NSDUHDetailedTabs2021/NSDUHDetailedTabsTOC2021.htm (23 March 2026, date last accessed). [Google Scholar]
  • 50.Cuttler C, Mischley LK, Sexton M. Sex differences in cannabis use and effects: a cross-sectional survey of cannabis users. Cannabis Cannabinoid Res 2016;1:166–75. 10.1089/can.2016.0010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Montgomery L, Dixon S, Mantey DS. Racial and ethnic differences in cannabis use and cannabis use disorder: implications for researchers. Current addiction reports 2022;9:14–22. 10.1007/s40429-021-00404-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.O’Grady MA, Iverson MG, Suleiman AO et al. Is legalization of recreational cannabis associated with levels of use and cannabis use disorder among youth in the United States? A rapid systematic review. European Child & Adolescent Psychiatry 2024;33:701–23. 10.1007/s00787-022-01994-9 [DOI] [PubMed] [Google Scholar]
  • 53.Orenstein DG, Glantz SA. Cannabis legalization in state legislatures: public health opportunity and risk. Marquette law review 2020;103:1313. [PMC free article] [PubMed] [Google Scholar]
  • 54.Gantiva C, Illidge-Cortes J, González-Millares D et al. Health warning labels on cannabis products. What is the best design? Int J Drug Policy 2024;126:104355. 10.1016/j.drugpo.2024.104355 [DOI] [PubMed] [Google Scholar]
  • 55.Berg CJ, Duan Z, Wang Y et al. Impact of different health warning label and reduced exposure messages in IQOS ads on perceptions among US and Israeli adults. Prev Med Rep 2023;33:102209. 10.1016/j.pmedr.2023.102209 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Berg CJ, Duan Z, Wang Y et al. Impact of FDA endorsement and modified risk versus exposure messaging in IQOS ads: a randomised factorial experiment among US and Israeli adults. Tob Control 2024;33:e69–77. 10.1136/tc-2022-057639 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Berg CJ, Levine H, LoParco CR et al. Qualitative examination of US and Israeli adults’ perceptions of IQOS advertising messages: modified exposure and risk statements, US FDA endorsement, and health warnings. Nicotine Tob Res 2025;27:1083–91. 10.1093/ntr/ntae266 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.White AE, van Tubbergen C, Raymes B et al. Cannabis-infused edible products in Colorado: food safety and public health implications. Am J Public Health 2020;110:790–5. 10.2105/AJPH.2020.305601 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Cooper M, Shi Y. The impacts of packaging on preferences for cannabis edibles: a discrete choice experiment. Int J Drug Policy 2024;128:104453. [DOI] [PubMed] [Google Scholar]
  • 60.Cooper M, Shi Y. Appeal rating and visual attention associated with youth-appealing cannabis packaging: an eye-tracking experiment. Drug Alcohol Depend 2023;253:110992. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Kooli C Chatbots in education and research: a critical examination of ethical implications and solutions. Sustainability 2023;15:5614. 10.3390/su15075614 [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

suppl materials
NIHMS2187218-supplement-suppl_materials.pdf (285.5KB, pdf)

Data Availability Statement

The data underlying this article will be shared on reasonable request to the corresponding author.

RESOURCES