Miguel Angel Alvarez-Mon1,2,3
Carla Ojeda3
Francisco Lara-Abelenda1,4Ángel Asunsolo del Barco2,5,6
Oscar Fraile-Martínez1,2
Cielo García-Montero1,2*Sonia Fernández-Rojo3
Javier Quintero3,7
Miguel Angel Ortega1,2
Melchor Alvarez-Mon1,2,8Fernando Mora3,7- 1Department of Medicine and Medical Specialities, Faculty of Medicine and Health Sciences, University of Alcala, Alcala de Henares, Spain
- 2Ramón y Cajal Institute of Sanitary Research (IRYCIS), Madrid, Spain
- 3Department of Psychiatry and Mental Health, Hospital Universitario Infanta Leonor, Madrid, Spain
- 4Departamento Teoria de la Señal y Comunicaciones y Sistemas Telemáticos y Computación, Escuela Tecnica Superior de Ingenieria de Telecomunicación, Universidad Rey Juan Carlos, Fuenlabrada, Spain
- 5Department of Surgery, Medical and Social Sciences, Faculty of Medicine and Health Sciences, University of Alcalá, Alcala de Henares, Spain
- 6Department of Epidemiology and Biostatistics, Graduate School of Public Health and Health Policy, University of New York, New York, NY, United States
- 7Department of Legal and Psychiatry, Complutense University, Madrid, Spain
- 8Service of Internal Medicine and Immune System Diseases-Rheumatology, University Hospital Príncipe de Asturias, (CIBEREHD), Alcalá de Henares, Spain
Background and aims: Consultations and admissions for pathologies related to cannabis use are growing ininterrumpedly. The lack of public awareness of the risks can have a negative impact on our society, as well as on new policy proposals. In response, we set out to investigate social media posts about cannabis to better understand the online environment in this regard.
Design, setting, and measures: The study will analyze a dataset of tweets posted between 2018 and 2022, written in Spanish, that include the keyword cannabis, marihuana, or hachis. A total of 68,673 tweets were included in our study. A subset of 500 posts for each keyword was manually analyzed by a researcher to establish a codebook. Subsequently, Machine Learning techniques were employed to analyze the remaining 67,173 comments using the established codebook. Finally, 32,646 of the remaining tweets were excluded as they contained information unrelated to the objectives of this study.
Findings: Our research reveals a pronounced Twitter user engagement with cannabis, primarily centered on its regulatory and health dimensions. In more detail, 73.2% of the analyzed tweets were in favor and only 3.5% of the population expressed against its regulation, whereas only 20.4% of the tweets discussed the negative effects of cannabis on physical or mental health. Additionally, 30.1% of the tweets are in favor of the therapeutic use of cannabis, while 69.9% of tweets manifest neutral or against therapeutic use. Our findings also show significant differences on these topics depending on the user type and between consumers versus non-consumers.
Conclusion: This analysis of tweets about cannabis provides information on experiences and opinions related to its use. Therefore, the perspectives of Twitter users constitute valuable input that can help improve physicians’ knowledge about cannabis and their communication with patients about its dangerousness.
1 Introduction
Cannabis, also known by the names of hashish and marijuana is the most commonly consumed illegal drug in the world (1). According to the World Health Organisation (WHO), about 147 million people (2.5% of the world population) consume cannabis annually, compared with 0.2% consuming cocaine and 0.2% consuming opiates (2). The most notable growth in cannabis abuse since the 1960s has been in developed countries, although developing countries are also increasing cannabis consumption (3, 4). According to the National Drug Plan of the Spanish government, 28.6% of the population between the ages of 14 and 18 have tried cannabis at least once in their lifetime and 22.2% have tried it in the last 12 months (5). In the population aged 15–64, cannabis is the most commonly consumed illegal drug in Spain and almost 41% of the surveyed have tried it at least once in their lifetime (6). Its consumption is more frequent among males and significantly decreases as age increases. Cannabis consumption rates in Spain have shown an increasing trend since 2013. On a European level, Spain ranks third in cannabis consumption, following France and Denmark (7, 8). Outside of Europe, other developed countries like the USA have also seen an increase in the number of consumers (9), while Hispanic American countries like Uruguay and Chile have also observed an increased consumption of cannabis over time, peaking among those aged 20–24 and increasing across all age groups, with period effects indicating notably higher prevalence in recent years, especially for women (10). Interestingly, other works have shown differences in Spanish-speaking people depending on their residence. A survey of 549 Spanish-speakers individuals, including 294 residing in the USA show that despite recreational use of cannabis was the most common topic, those living in the USA were more likely to consume daily cannabis smoked or vaporized, also reporting a similar interest with recreational and medical use pattern (11). Therefore, based on these data is of great importance to deepen on the perception of cannabis in Spanish-speaking people, considering specific differences across topics and regions.
Currently, there is a great controversy surrounding the effects and potential applications of the Cannabis sativa plant. On one hand, empirical evidence shows the multiple harms of its acute and chronic consumption when used recreationally, mainly through inhalation (12, 13). On the other hand, the approval of medical cannabis uses for certain conditions in an increasing number of places around the world has led to reforms in the regulatory laws concerning this substance, which in turn has had an impact on its consumption and perceived risk by the general population. In this regard, despite a considerable amount of studies examining cannabis consumption in all its various forms, often these research findings are not adequately synthesized, translated, or communicated to policymakers, healthcare providers, state health officials, and, in general, the entire population (14). In this context, not only cannabis consumption but also legalization issues represents a global subject of debate.
Multiple studies have been conducted on different social media like Twitter in order to understand the public opinion of a particular social concern, as it can be the uses, experiences and opinion around cannabis consumption (15–17). This platform has some particular advantages in comparison to surveys and other study designs, as it can be the perceived safety and facility of users to tweet about honest experiences without feeling judged or anonymity, observe the interactions between users, the accessibility and broad access to information, as well as the possibility to offer peer/social/emotional support, public health monitoring, and potential to influence health policy (18, 19). In this study, we aimed to investigate the public opinion regarding cannabis and its consumption in the context of the Spanish government considering its legalization. We set the following objectives: (1) Determine the main topic of Twitter publications, their generated interest, and their scientific adequacy; (2) understand the users’ perception of the health risks associated with cannabis consumption; (3) characterize the user types that are more supportive or against the legalization, considering therapeutic or recreational use; (4) considering geolocalization and cultural data to understand regional differences in the analyzed tweets.
2 Methods
2.1 Search and collection of tweets
This analysis focused on tweets related to cannabis posted on the social media platform Twitter. We included tweets that met the following criteria: (a) Public tweets; (b) Containing readable text in Spanish; (c) Using any of the keywords “cannabis,” “marihuana,” or “hachis” anywhere in the tweet; (d) Having received at least 10 retweets; (e) Published between January 1, 2018, and April 30, 2022.
We used the Tweet Binder tool to collect the tweets, which has been widely used in previous research and provides access to 100% of public tweets (20, 21). In addition to the tweet text, this tool provides the count of retweets and likes for each tweet, as well as the date of publication, a link to the tweet for contextual viewing, and user description. The number of retweets and likes received by each tweet was used as an indicator of the interest generated among users for the corresponding content (22–24).
The search resulted in a total of 247,156 tweets collected, out of which 178,483 were excluded as they were written in a language other than Spanish or contained too little text. Out of the remaining 68,673 tweets, a researcher analyzed 500 posts for each keyword (500 cannabis tweets, 500 marihuana tweets, and 500 hachís tweets) and established a codebook to analyze the remaining 67,173 comments using Machine Learning. Finally, 32,646 of the remaining tweets were excluded as they contained information unrelated to the objectives of this study or were written in a way that their meaning was uncertain.
2.2 Identification of thematic categories and creation of a codebook
The authors employed an inductive-deductive mixed approach to develop a codebook for classifying the content of the tweets based on key thematic categories. Deductively, they used categories from previous research that have also analyzed content posted on social media (25, 26). Inductively, they explored an initial subset of 1,500 tweets (from a small manually classified subset) to identify possible new themes and refine the codebook. Two researchers coded these 1,500 tweets, discussing any discrepancies with the research team and reaching a final consensus on coding. Once the final codebook was agreed upon, the machine coded the remaining 67,173 tweets, of which 34,528 could be analyzed, as the rest were unclassifiable.
The tweets were classified as classifiable or unclassifiable. A tweet was considered unclassifiable if its content was purely political, if the information was irrelevant to the objectives of the current work, or if it was a joke, uncertain or insufficient content. Among the tweets considered classifiable, it was determined whether the content was medical or non-medical, with these categories being mutually exclusive. In turn, the medical tweets were classified according to the clinical area of interest mentioned or discussed in the tweet text: (1) Health risks; (2) Discussion of therapeutic or medicinal use; (3) Content on preventive measures; (4) Sentiment regarding consumption; (5) Type of consumption. In the medical content tweets, it was evaluated whether they addressed the legalization debate or were related to recreational use of cannabis. Lastly, the users were classified into three categories: (1) General Twitter users and healthcare professionals (psychiatrists or doctors from other specialties, psychologists, nurses, clinical researchers, etc.); (2) Media and governmental (health institution) and non-governmental organizations (pro-cannabis associations); and (3) Public figures (politicians). In cases where tweets with nearly identical content were found, they were classified in the same way as the first tweet encountered. The classification criteria and examples of tweets are shown in Table 1.
Table 1. Examples of tweets.
2.3 Ethical considerations
This study has been conducted in accordance with the ethical research principles outlined in the Declaration of Helsinki (seventh revision, 2013) and has been approved by the ethics committee of the Complutense University. In any case, it did not directly involve human subjects nor included any interventions. Furthermore, we have taken care not to directly disclose any usernames in this study and have avoided citing information that could identify specific individuals.
2.4 Machine learning classifier
Recent technological advances have led to the emergence of artificial intelligence (AI), which can process and analyze data (27). Machine Learning (ML) is a branch of AI that focuses on extracting knowledge from data using computational models. Deep Learning (DL), a subset of ML, employs neural networks inspired by the human brain to process information (28). Neural networks have various applications, including weather prediction (29) or object recognition (30). Besides, Natural Language Processing (NLP) extensively utilizes neural networks to analyze text, recreate conversations, or and extract key ideas (31). In this project, a pretrained network called BERTWEET, trained on 850 million English tweets, was used to classify cannabis-related tweets (32).
Before implementing BERTWEET, the database underwent preprocessing. Non-English tweets were translated to English using Google Translator, and the tweets were normalized by removing special characters, splitting negative contractions, and eliminating repetitions. Since BERTWEET was not initially trained for the specific categories, fine-tuning was performed. The manually classified tweets were randomly divided into an 70% training subset and a 30% testing subset. The training subset was used to fine-tune the network, and the testing subset validated its performance. Additionally, the training set contains imbalanced categories, with varying counts across different options. To solve that, we employed the easy data augmentation (EDA) pipeline (33) to generate additional tweets, ensuring balanced representation across categories. EDA creates new data by substituting words with synonyms, randomly deleting words, and swapping word positions.
Using the training, set we trained a separate model for each category and calculated the F1-score on the test set. The models achieved the following F1-scores across categories: classifiable/non-classifiable (0.76), user (0.75), medical or non-medical (0.88), legalization (0.82), health risks (0.79), discussion of therapeutic or medicinal use (0.88), content on preventive measures (0.5), sentiment regarding consumption (0.7), and type of consumption (0.91). Due to the low predictive quality, we excluded the model for “content on preventive measures” from our analysis. However, the other categories achieved satisfactory F1-scores, indicating strong model performance.
Finally, emotion analysis was conducted using a pretrained neural network called emotion-english-distilroberta-base (34), capable of detecting six basic Elkman’s emotions (35) and neutral sentiment. This network, previously used in research studies, was applied to the preclassified tweets. This methodology has been correctly validated, already (36, 37).
2.5 Statistical analysis
First, a descriptive study was conducted for all collected variables. For categorical variables, absolute and relative frequencies were calculated. For quantitative variables, the normality of the distribution was initially assessed using the Kolmogorov–Smirnov test and graphical representations. Since the variables did not follow a normal distribution, the median and interquartile range were used to summarize the results. Second, bivariate analyses were performed to answer the research questions. For the cross-tabulation of categorical variables, the Pearson’s chi-squared test or Fisher’s exact test was used, as appropriate. In cases where a quantitative variable was compared with a categorical variable, the Mann–Whitney U test (for bivariate comparisons) or the Kruskal–Wallis test (for three or more categories) was employed. No p-value adjustment was performed, with a significance level of <0.05 considered statistically significant. The obtained p-values were reported in all cases. The analysis of results was conducted using IBM SPSS Statistics version 27.
Overall in Figure 1 we provide a flow chart of the methodology used in this study.
Figure 1. Flow chart of the methodology followed in this work.
3 Results
3.1 General description of analyzed tweets
Table 2 provide a global description of the analyzed tweets. Firstly, we observed that tweets containing medical information represented a 47.1% of the total tweets whereas those that do not include medical information represented the remaining 52.9%. In parallel, our study reports that most tweets came from general users (anonymous counts) of Twitter (61.6%), followed by media (20.3%), although up to a 10.6% of the tweets were not identified.
Table 2. Distribution of tweets according to the shared content.
We also observed that 17.3% of tweets express positive emotions around cannabis, whereas the remaining 82.7% included neutral or negative emotions. Excluding neutral opinions, the most frequently observed emotion was fear (Figure 2). Finally, only 2.5% openly discussed personal cannabis consumption, while 97.5% do not.
Figure 2. Distribution of tweets according to the expressed emotion around cannabis content.
3.2 A significant percentage of tweets express support for the legalization of cannabis, overlooking both the potential harms of this substance and its therapeutic use
Regarding legalization, we observed that 73.2% of the analyzed tweets were in favor and only a 3.5% of the population expressed against its regulation (Table 2). The remaining 23.3% of tweets were neutral around this question. In terms of health-related aspects of cannabis, we observed that only 20.4% of the tweets discussed negative effects of cannabis for physical or mental health. The remaining 79.6% do not discuss or mention any danger associated with its consumption. In paralell, 30.1% of the tweets are in favor of the therapeutic use of cannabis, while 69.9% of tweets manifests neutral or against therapeutic use (Table 2).
When we deepen on legalization and health issues related to cannabis consumption, our study reveals that the most active users on the platform are those advocating for the legalization and unrestricted consumption of cannabis (Table 3). Tweets originating from self-proclaimed cannabis consumers demonstrated the highest reach, as evidenced by a substantial number of retweets and likes. The predominant content in these tweets was characterized by strong support for cannabis legalization, along with the promotion of purported benefits associated with its use.
Table 3. Distribution of tweet categories based on the number of retweets and likes received.
The majority of individuals opposed to cannabis legalization firmly believe that it should not be considered as a therapeutic option (Table 4). Conversely, among those in favor of legalization, only 34.6% believe in its potential bodily benefits, while 65.5% do not, suggesting a prevalent recreational intent. Additionally, both proponents and opponents of legalization exhibit skepticism regarding the potential adverse effects on physical and mental health, with 93.8 and 95%, respectively, expressing disbelief in such risks.
Table 4. Distribution of tweets regarding the therapeutic use of cannabis and the health risks associated with its consumption in subjects with neutral views, in favor and against cannabis legalization.
3.3 A comparative analysis reveals diverse backing for legalization, varied stances on health risks and medicinal use according to the user type
As shown in Table 5, from the three identified user types, over 75% do not reference the risks associated with cannabis or firmly believe that such harm to health does not exist. Regarding legalization, we observe that in all three user types, the highest percentages (74.3, 73.3, and 60.9% respectively) indicate that they are in favor of cannabis being regulated by the state.
Table 5. Distribution of tweets according to the type of user.
Regarding medicinal cannabis, all three user types (68.7, 63.2, and 90.8%) are against the use of cannabis for therapeutic purposes. This suggests that, considering the previous observation that the majority support its legalization, it is likely for recreational purposes.
All three user types have a neutral or negative sentiment toward cannabis consumption, with over 79.9% expressing such sentiment.
3.4 Differing perspectives on risks, legalization, medicinal use, and sentiments among consumers and non-consumers
Table 6 shows a comparative between consumers and non-consumers. It is noticeable that non consumers are less likely to discuss risks and harms of cannabis consumption in comparison to consumers (20.2% versus 27.7%) whereas non-consumers advocate for cannabis legalization more frequently than consumers do (75% versus 4.7%). In both groups, they believe that cannabis could be used as a medical alternative to treat various conditions. It is worth noting that this is much more treated in non-consumers (30.8% versus 4%). Regarding the emotions toward consumption, almost 84% of non-consumers manifest negative or neutral views, with only 16.3% expressing positive feelings in comparison to up to a 55.2% of consumers. It is of note that non consumers are more likely to exhibit neutral emotions that consumers (43.7% versus 32.6%, Figure 3). When distinguishing the analyzed emotions we observed that appart from neutral sentiments, non-consumers more commonly manifest fear (24.1%), anger (12.8%) and happiness (8.4%), whereas consumers more commonly exhibit anger (18.9%), disgust (17.3%) and happiness (13.4%) (Figure 3).
Table 6. Distribution of tweets classified according to the type of consumer.
Figure 3. Emotion analysis in non-consumers (blue) versus consumers (red).
3.5 Geolocalization data reveals that most tweets written in Spanish discussing cannabis came from Spain, Chile, Argentina, Mexico and Colombia
Out of the total analyzed tweets, 73.3% had geolocation data available. Of the localized tweets, 78.54% originated from five countries: Spain (19.13%), Chile (17.48%), Argentina (16.99%), Mexico (14.19%), and Colombia (10.75%). Other regions geolocalized in our study was United States, Uruguay, Ecuador, Paraguay, Peru and Venezuela (Figure 4).
Figure 4. Distribution of tweets by country.
4 Discussion
In this study, we have observed that Twitter users show a great interest in cannabis, focusing mainly on regulatory aspects of the substance as well as health-related aspects. The most active users are those who are in favor of its legalization and unrestricted consumption. Tweets written by users who claimed to consume cannabis have had the highest reach, meaning they received a greater number of retweets and likes, and their content mainly supported the legalization and promoted the alleged benefits of cannabis. On the other hand, healthcare professionals did not have a strong presence in driving Twitter conversations regarding cannabis, nor did medical institutions or government organizations.
Firstly, we observed that 73.2% of the posts we analyzed are in favor of cannabis legalization. According to previous works, the main reasons explaining the public support of cannabis legalization was to state that this drug is less dangerous than other substances and has significant medical benefits, also considering criminal justice reform and the potential for tax revenue as potential benefits of legalization (38). Besides, in the same study the authors also found that harms associated with cannabis use were the most commonly reasons for opposing legalization. Likewise legalizing cannabis could contribute to a lower perception of risks and reduce the fear in the population, potentially encouraging increased consumption (39). Indeed, available literature reports that this has occurred in certain regions like the United States, where there appears to be a ~ 20% average increase in cannabis use frequency attributable to recreational legalization (40). According to the NCBI, cannabis consumption has increased among adults over 21 years of age in countries like Canada and the United States that have passed laws allowing its medical use, leading to higher rates of daily use, misuse, and dependence (41, 42). Additionally, the number of adult men seeking treatment for cannabis-related disorders has increased more in these countries. Howsoerver, compelling evidence supports that public engagement with information about medical cannabis in the internet and social media are one of the main mechanisms by which medical cannabis legalization is associated with cannabis legalization (43). In this sense, our study seems to be in line with these claims, and Twitter could be a platform reflecting the complex social environment surrounding cannabis legalization and medical use.
It was also surprising for us that up to 79.6% of tweets do not perceive the risk of consuming cannabis, stating that it is not associated with any harmful health risks. The perception of cannabis as harmless often stems from its “natural” origin; however, this does not imply physiological safety, as cannabis exerts significant psychoactive effects, primarily from THC, which acts quickly on the brain (44, 45). This is particularly concerning among youth, especially those with coexisting mental or substance abuse disorders (46–49). Cannabis use is linked with a range of acute and chronic adverse effects, including hyperemesis syndrome, anxiety, and long-term neurocognitive, cardiovascular, and respiratory issues, with risks escalating with earlier and more frequent use (50, 51). Importantly, the connection between cannabis use and psychosis is well-documented but often underrepresented in both public discourse and policy. Evidence shows that high-potency cannabis use significantly increases the risk of developing psychotic disorders, especially in vulnerable populations. For instance, the EU-GEI multicenter case–control study found that differences in cannabis use patterns across European cities were significantly associated with variations in the incidence of psychotic disorders (7). Furthermore, early cannabis use has been shown to act as a modifiable environmental risk factor for psychosis onset, alongside genetic and social variables (52) Moreover, cannabis use in individuals with early psychosis is associated with increased risk of relapse and hospitalization, although recent studies indicate that treatment with long-acting injectable antipsychotics, such as aripiprazole, may help reduce these risks and improve quality of life (53, 54).
Likewise, cannabis consumption over time can lead to a situation of dependency named cannabis use disorders (CUD) (55, 56). CUD commonly co-occurs with other mental health disorders, increasing risks of self-harm, overdose, and mortality among youth with mood disorders (41). The risk of developing this type of addiction seems to be greater in individuals aged between 13 to 18 years old (57). Additionally, cannabis use disorder is a common comorbidity and a risk marker for self-harm, all-cause mortality, unintentional overdose, and homicide among young people with mood disorders (58). While no specific treatment for CUD exists, symptom-targeted medication, psychotherapy, and psychoeducation are recommended, especially for adolescents. Furthermore, the use of unregulated or unpurified cannabis extracts presents additional health concerns. These products often lack standardized labeling and may inaccurately report the concentrations of key cannabinoids such as THC, CBD, (59, 60). This can mislead consumers about the potency and potential effects of the extract. In some cases, unregulated extracts may contain other bioactive compounds like terpenes, flavonoids, or alkaloids, which can alter the pharmacological profile of the product, potentially enhancing psychoactive effects or increasing the risk of adverse reactions. Without appropriate oversight and quality control, such formulations pose unpredictable health risks, particularly when consumed by vulnerable populations (60). Addressing the low-risk perception and social allure of cannabis use is key to improving health outcomes, as smoked cannabis poses notable acute and chronic health risks (8).
On the other hand, the medicinal use and health benefits of cannabis were also considered in an important percentage of tweets. Surprisingly, our results show that despite most tweets being in favor of cannabis legalization and perceiving a low risk of its use, almost 70% of tweets did not consider or rejected the therapeutic use of cannabis. It is important to highlight that a growing number of studies have shown multiple benefits from cannabinoids, especially the compound named cannabidiol (CBD), another component found in the Cannabis sativa plant. Initial evidence supports its effects to alleviate insomnia, inflammation, anxiety, depressive symptoms, pain, post-traumatic stress disorder, and so on (61–63), especially in form of CBD and hemp oils. However, the existing literature claims that there is little regulation around these products and studies have found inaccurate labeling of CBD and THC quantities (64). Besides, further clinical research is required, as well as to find adequate doses and applications for each subject and explored condition (65). Currently, the FDA has approved the use of synthetic cannabis-related drug products, mainly two compounds containing dronabinol (a synthetic form of THC) for the treatment of anorexia associated with weight loss in AIDS patients, another product with nabilone (THC) prescribed for the treatment of nausea and vomiting associated with anticancer chemotherapy and the proper CBD, which can be used to manage and treat the seizure disorders Lennox–Gastaut syndrome and Dravet syndrome (66). Also, there is moderate evidence from the use of cannabis and cannabinoids for pain relief in patients with chronic pain and for treating multiple sclerosis (MS)-related spasticity (67). Overall, medical uses and benefits from cannabis, especially in the form of CBD are increasingly being supported, although it is also true that many of these properties have been overclaimed and further regulation is required before widespread use of these components, evidencing the need of contextualizing this complex picture in social media like Twitter.
In our study, only 2.5% openly discuss their consumption, whereas the remaining 97.5% were identified as non-consumers. When compared to non-consumers, consumers were more likely to discuss detrimental physical/mental health effects of cannabis (27.7% versus 20.2%). Conversely, we observed that, unlike consumers, non-consumers tended to show in favor of cannabis legalization (75% versus 4.7%) and its therapeutic use (30.8% versus 4.8%), also exhibiting more commonly negative feelings around cannabis consumption (83.7% versus 44.8%). Previous works conducted on Twitter found an important number of tweets discussing unsustained health benefits derived from cannabis use, especially those tweeted by social bots (17). The authors reported that this type of information might be influencing Twitter users to perceive CBD and cannabis as anticancer and effective treatments for several diseases, whereas the level of evidence regarding their uses remains to be fully investigated in most cases. Thus, it would be reasonable that these types of beliefs and ideas could influence the legal debate around cannabis use. However, it was surprising for us that, of those subjects in favor to legalize cannabis consumption, only 35% knew and supported the therapeutic use of cannabis, suggesting that most tweets in favor of cannabis legalization were related to its recreational use. Also, the fact that an important percentage of non-consumers support cannabis legalization could also indicate that an important part of Twitter users are not knowledgeable about the acute and chronic effects of recreational cannabis consumption, as well as the number of individuals affected by such use and the complex relationships that exist between legalization and the social perception of a drug. The therapeutic effects of cannabinoids are exerted mainly through the modulation of cannabinoid receptors (CB1 and CB2) and the endocannabinoid system (ECS). The ECS system has a negative feedback mechanism and retrograde signaling to maintain physiological balance. Exogenous cannabinoids (like THC and CBD) can enhance or prolong ECS effects, also activating other receptors beyond CB1 and CB2, including transient receptor potential (TRP) channels, such as TRPV1, peroxisome proliferator-activated receptors (PPARs), or other non-cannabinoid targets like GPR55, GPR18, and serotonin receptors, all of which contribute to the broad pharmacological profile of cannabinoids in various therapeutic contexts, including epilepsy, multiple sclerosis, anxiety, and chronic pain (68–71). Understanding these diverse molecular pathways is key to evaluating both the potential medical benefits and associated risks of cannabinoid-based treatments. Because of this, there are authors that claim the need for designating timely social media communications with new cannabis-related information by authoritative institutions that deal with public health in order to help a general public mostly exposed to pro-cannabis content on Twitter (72).
Finally, when we considered geolocalization data, most of the tweets (56.1%) were from Spain and Hispanic American countries, with Chile, Argentina, Mexico, and Colombia representing up to 42.4% of the analyzed tweets. Without considering neutral opinions, our sentiment analysis showed that fear and anger were the most common manifestations expressed on Twitter. Previous works conducted in Twitter have evidenced the relevance from considering geographic differences in the sentiment and content of cannabis-related tweets, specially due to the different legislation and/or reported consumption in different regions (73). On the one hand, in Spain, the medical use of cannabis is already legal, but it is regulated and has nothing to do with recreational use (74). However, as previously mentioned, Spain is ranked as the third European country in cannabis consumption after France and Denmark, and the recreational uses of this drug is a growing concern affecting this country, especially among youth and men (75). Thus, Twitter may clearly reflect the associated feelings of worry with this situation. On the other hand, it seems that there is a huge heterogeneity in attitudes toward drug policies in Hispanic American countries. According to the 2014 Annual Survey of the Observatory of Drug Policies and Public Opinion (76), a notable diversity in perspectives pertaining to drug policies is observable across South America countries, indicative of a discernible scrutiny directed toward the prevailing norms in the region. In the case of Mexico, Argentina, Colombia and Chile, an important percentage of surveyed people are in favor of recreational uses of cannabis, whereas Chile, Colombia and Mexico ranked as the first, second and fourth Latin American countries with more people supporting cannabis legalization, with more of the 40% of the people surveyed Despite only 26.4% of people in Argentina were in favor of cannabis legalization, this country obtained the second-highest rate of people who had ever consumed cannabis after Chile (76). To this complex picture, it should also be considered the regulatory framework around medical use of cannabis, as it is also a matter of concern in these regions (77) and the fact that some of these countries like Chile and Colombia are among those with the highest incidence of cannabis use disorder in Latin American (78). Overall, our results agree with the complex background around cannabis consumption, legalization and recreational issues observed in these countries, explaining why the fear and anger are the predominant feelings in our observed tweets.
5 Limitations
This study has some limitations. Firstly, since Twitter users tend to be younger than the general population, it’s possible that our results may not apply to older age ranges. In fact, adolescent patients, although experiencing an exponential growth of psychotic outbreaks and hospital admissions for this reason, develop other diseases such as cancer or COPD at a later age. Secondly, we were unable to examine how clinical characteristics, symptom severity, duration of use, or residual cognitive dysfunction associated with marijuana influenced the content of the social media posts due to a lack of psychiatric evaluation. Thirdly, the coding book and text analysis we used involve a degree of subjectivity. However, this methodology is consistent with previous medical research studies on Twitter and could be applied to various topics by different authors. Fourthly, the list of keywords included generic terms, but tweets that contained spelling errors may have been excluded. Lastly, incorporating studies that analyze cannabis content on platforms like Facebook, Instagram or TikTok could enhance the accuracy of our understanding of public perceptions of these drugs on social media and within the broader population (79–81).
6 Conclusion
Our findings highlight the potential of leveraging social media to better understand the rise of cannabis as a drug of choice in our population. As a preventive measure in a society that is increasingly in favor of approving a law for the free consumption of cannabis, it is important for healthcare professionals and medical and political authorities to intervene by publishing more content about the risks of its consumption and doing so more frequently. These findings should be considered as states consider the legalization of medical and recreational marijuana, as both are associated with increased cannabis consumption and related risks. An increased risk of cannabis use disorder. Twitter can also serve as an additional educational tool to raise awareness among its users about the current status of therapeutic and medical uses of cannabis and, above all, its numerous disadvantages and consumption risks primarily derived from recreational and inhalation use.
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Ethics statement
This study was approved by the Research Ethics Committee of Universidad de Alcalá and is compliant with the ethical principles from the World Medical Association Declaration of Helsinki (7th revision, 2013).
Author contributions
MiA-M: Conceptualization, Funding acquisition, Investigation, Project administration, Resources, Supervision, Visualization, Writing – original draft, Writing – review & editing. CO: Formal analysis, Investigation, Methodology, Writing – original draft. FL-A: Data curation, Methodology, Software, Writing – review & editing. ÁA: Conceptualization, Methodology, Software, Supervision, Writing – review & editing. OF-M: Investigation, Supervision, Writing – original draft, Writing – review & editing. CG-M: Investigation, Supervision, Writing – original draft, Writing – review & editing. SF-R: Conceptualization, Validation, Writing – review & editing. JQ: Conceptualization, Supervision, Validation, Writing – review & editing. MO: Conceptualization, Investigation, Resources, Visualization, Writing – original draft, Writing – review & editing. MeA-M: Conceptualization, Funding acquisition, Project administration, Resources, Supervision, Validation, Writing – review & editing. FM: Investigation, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing.
Funding
The author(s) declare that financial support was received for the research and/or publication of this article. This study (FIS-PI22/00653) was supported by the Instituto de Salud Carlos III (grant no. Estatal de I + D + I 2020–2027).
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
1. Kesner, AJ, and Lovinger, DM. Cannabis use, abuse, and withdrawal: cannabinergic mechanisms, clinical, and preclinical findings. J Neurochem. (2021) 157:1674–96. doi: 10.1111/JNC.15369
2. World Health Organization. Alcohol, Drugs and Addictive Behaviours. (2023). Available online at: https://www.who.int/teams/mental-health-and-substance-use/alcohol-drugs-and-addictive-behaviours/drugs-psychoactive/cannabis (Accessed July 13, 2023).
3. Hall, W, and Degenhardt, L. Prevalence and correlates of cannabis use in developed and developing countries. Curr Opin Psychiatry. (2007) 20:393–7. doi: 10.1097/YCO.0B013E32812144CC
4. Carvalho, AF, Stubbs, B, Vancampfort, D, Kloiber, S, Maes, M, Firth, J, et al. Cannabis use and suicide attempts among 86,254 adolescents aged 12-15 years from 21 low- and middle-income countries. Eur Psychiatry. (2019) 56:8–13. doi: 10.1016/J.EURPSY.2018.10.006
5. Observatorio Español de las Drogas y las Adicciones. Informe 2022. Alcohol, tabaco y drogas ilegales en España. Delegación del Gobierno para el Plan Nacional sobre Drogas. Madrid: Ministerio de Sanidad. (2022) 293 p.
6. Observatorio Español de las Drogas y las Adicciones. Informe 2022. Alcohol, tabaco y drogas ilegales en España. Delegación del Gobierno para el Plan Nacional sobre Drogas. Madrid: Ministerio de Sanidad. (2023) 270 p.
7. Di Forti, M, Quattrone, D, Freeman, TP, Tripoli, G, Gayer-Anderson, C, Quigley, H, et al. The contribution of cannabis use to variation in the incidence of psychotic disorder across Europe (EU-GEI): a multicentre case-control study. Lancet Psychiatry. (2019) 6:427–36. doi: 10.1016/S2215-0366(19)30048-3
8. Khademi, S, Hallinan, CM, Conway, M, and Bonomo, Y. Using social media data to investigate public perceptions of Cannabis as a medicine: narrative review. J Med Internet Res. (2023) 25:e36667. doi: 10.2196/36667
9. Cerdá, M, Mauro, C, Hamilton, A, Levy, NS, Santaella-Tenorio, J, Hasin, D, et al. Association between recreational marijuana legalization in the United States and changes in marijuana use and Cannabis use disorder from 2008 to 2016. JAMA Psychiatry. (2020) 77:165–71. doi: 10.1001/JAMAPSYCHIATRY.2019.3254
10. Castillo-Carniglia, A, Rivera-Aguirre, A, Calvo, E, Queirolo, R, Keyes, KM, and Cerdá, M. Trends in marijuana use in two Latin-American countries: an age, period, and cohort study. Addiction (Abingdon Engl). (2020) 115:2089. doi: 10.1111/ADD.15058
11. Martin-Willett, R, Garza, EZ, and Bidwell, LC. Cannabis use patterns and related health outcomes among Spanish speakers in the United States and internationally. Yale J Biol Med. (2022) 95:327.
12. Archie, SR, and Cucullo, L. Harmful effects of smoking Cannabis: a cerebrovascular and neurological perspective. Front Pharmacol. (2019) 10:1481. doi: 10.3389/FPHAR.2019.01481
13. Cohen, K, Weizman, A, and Weinstein, A. Positive and negative effects of Cannabis and cannabinoids on health. Clin Pharmacol Ther. (2019) 105:1139–47. doi: 10.1002/CPT.1381
14. National Academies of Sciences E and M, Division H and M, Practice B on PH and PH, Agenda C on the HE of MAER and R. The health effects of cannabis and cannabinoids. Psychiatria. (2017) 15:88–92. doi: 10.17226/24625
15. Shutler, L, Nelson, LS, Portelli, I, Blachford, C, and Perrone, J. Drug use in the Twittersphere: a qualitative contextual analysis of tweets about prescription drugs. J Addict Dis. (2015) 34:303–10. doi: 10.1080/10550887.2015.1074505
16. Lim, TY, Stringfellow, EJ, Stafford, CA, DiGennaro, C, Homer, JB, Wakeland, W, et al. Modeling the evolution of the US opioid crisis for national policy development. Proc Natl Acad Sci USA. (2022) 119:4119. doi: 10.1073/PNAS.2115714119
17. Allem, JP, Escobedo, P, and Dharmapuri, L. Cannabis surveillance with Twitter data: emerging topics and social bots. Am J Public Health. (2020) 110:357–62. doi: 10.2105/AJPH.2019.305461
18. Little, JS, and Romee, R. Tweeting from the bench: Twitter and the physician-scientist benefits and challenges. Curr Hematol Malig Rep. (2020) 15:419–23. doi: 10.1007/S11899-020-00601-5
19. Berry, N, Lobban, F, Belousov, M, Emsley, R, Nenadic, G, and Bucci, S. #WhyWeTweetMH: understanding why people use Twitter to discuss mental health problems. J Med Internet Res. (2017) 19:6173. doi: 10.2196/JMIR.6173
20. Alvarez-Mon, MA, Llavero-Valero, M, Sánchez-Bayona, R, Pereira-Sanchez, V, Vallejo-Valdivielso, M, Monserrat, J, et al. Areas of interest and stigmatic attitudes of the general public in five relevant medical conditions: thematic and quantitative analysis using Twitter. J Med Internet Res. (2019) 21:4110. doi: 10.2196/14110
21. Viguria, I, Alvarez-Mon, MA, Llavero-Valero, M, del Barco, AA, Ortuño, F, and Alvarez-Mon, M. Eating disorder awareness campaigns: thematic and quantitative analysis using Twitter. J Med Internet Res. (2020) 22:7626. doi: 10.2196/17626
22. Alvarez-Mon, MA, del Barco, AA, Lahera, G, Quintero, J, Ferre, F, Pereira-Sanchez, V, et al. Increasing interest of mass communication media and the general public in the distribution of tweets about mental disorders: observational study. J Med Internet Res. (2018) 20:9582. doi: 10.2196/JMIR.9582
23. Alvarez-Mon, MA, Donat-Vargas, C, Santoma-Vilaclara, J, de Anta, L, Goena, J, Sanchez-Bayona, R, et al. Assessment of antipsychotic medications on social media: machine learning study. Front Psych. (2021) 12:7684. doi: 10.3389/FPSYT.2021.737684
24. Alvarez-Mon, MA, Donat-Vargas, C, Llavero-Valero, M, Gea, A, Alvarez-Mon, M, Martinez-Gonzalez, MA, et al. Analysis of media outlets on women’s health: thematic and quantitative analyses using Twitter. Front Public Health. (2021) 9:4284. doi: 10.3389/FPUBH.2021.644284
25. Castillo-Toledo, C, Fraile-Martínez, O, Donat-Vargas, C, Lara-Abelenda, FJ, Ortega, MA, Garcia-Montero, C, et al. Insights from the Twittersphere: a cross-sectional study of public perceptions, usage patterns, and geographical differences of tweets discussing cocaine. Front Psych. (2024) 15:1282026. doi: 10.3389/FPSYT.2024.1282026 /BIBTEX
26. Carabot, F, Fraile-Martínez, O, Donat-Vargas, C, Santoma, J, Garcia-Montero, C, da Costa, MP, et al. Understanding public perceptions and discussions on opioids through Twitter: cross-sectional infodemiology study. J Med Internet Res. (2023) 25:13. doi: 10.2196/50013
27. Helm, JM, Swiergosz, AM, Haeberle, HS, Karnuta, JM, Schaffer, JL, Krebs, VE, et al. Machine learning and artificial intelligence: definitions, applications, and future directions. Curr Rev Musculoskelet Med. (2020) 13:69–76. doi: 10.1007/S12178-020-09600-8
28. Lecun, Y, Bengio, Y, and Hinton, G. Deep learning. Nature. (2015) 521:436–44. doi: 10.1038/nature14539
29. Hongkang, W, Li, L, Yong, W, Fanjia, M, Haihua, W, and Sigrimis, NA. Recurrent neural network model for prediction of microclimate in solar greenhouse. IFAC. (2018) 51:790–5. doi: 10.1016/J.IFACOL.2018.08.099
30. Kousik, NV, Natarajan, Y, Arshath Raja, R, Kallam, S, Patan, R, and Gandomi, AH. Improved salient object detection using hybrid convolution recurrent neural network. Expert Syst Appl. (2021) 166:114064. doi: 10.1016/J.ESWA.2020.114064
31. Deng, L, and Liu, SY. Deep learning in natural language processing. Nat Lang Eng. (2021) 27:373–5. doi: 10.1017/S1351324919000597
32. Quoc Nguyen, D, Thanh, V, and Anh Tuan, N. BERTweet: A pre-trained language model for English Tweets. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations. (2020). p. 9–14, Online. Association for Computational Linguistics.
33. Wei, J, and Zou, K. EDA: Easy Data Augmentation Techniques for Boosting Performance on Text Classification Tasks. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). Hong Kong, China: Association for Computational Linguistics. (2019) pp. 6382–6388.
34. Hartmann, J.. Emotion English DistilRoBERTa-base. (2022). Available online at: https://huggingface.co/j-hartmann/emotion-english-distilroberta-base (Accessed July 4, 2023).
35. Ekman, P. Basic Emotions. In Handbook of Cognition and Emotion. (T Dalgleish and MJ Power, editors.) (1999) doi: 10.1002/0470013494.ch3
36. de Anta, L, Alvarez-Mon, MA, Ortega, MA, Salazar, C, Donat-Vargas, C, Santoma-Vilaclara, J, et al. Areas of interest and social consideration of antidepressants on English tweets: a natural language processing classification study. J Pers Med. (2022) 12:155. doi: 10.3390/JPM12020155/S1
37. Butt, S, Sharma, S, Sharma, R, Sidorov, G, and Gelbukh, A. What goes on inside rumour and non-rumour tweets and their reactions: a psycholinguistic analyses. Comput Human Behav. (2022) 135:107345. doi: 10.1016/J.CHB.2022.107345
38. Resko, S, Ellis, J, Early, TJ, Szechy, KA, Rodriguez, B, and Agius, E. Understanding public attitudes toward Cannabis legalization: qualitative findings from a statewide survey. Subst Use Misuse. (2019) 54:1247–59. doi: 10.1080/10826084.2018.1543327
39. Lopez-Quintero, C, Cobos, JPde los, Hasin, DS, Okuda, M, Wang, S, Grant, BF, et al. Probability and predictors of transition from first use to dependence on nicotine, alcohol, cannabis, and cocaine: results of the National Epidemiologic Survey on alcohol and related conditions (NESARC). Drug Alcohol Depend (2011) 115:120–130. doi: 10.1016/J.DRUGALCDEP.2010.11.004
40. Zellers, SM, Ross, JM, Saunders, GRB, Ellingson, JM, Anderson, JE, Corley, RP, et al. Impacts of recreational cannabis legalization on cannabis use: a longitudinal discordant twin study. Addiction. (2023) 118:110–8. doi: 10.1111/ADD.16016
41. Hasin, DS, Saha, TD, Kerridge, BT, Goldstein, RB, Chou, SP, Zhang, H, et al. Prevalence of marijuana use disorders in the United States between 2001-2002 and 2012-2013. JAMA Psychiatr. (2015) 72:1235–42. doi: 10.1001/JAMAPSYCHIATRY.2015.1858
42. Najafizada, M, Rahman, A, Donnan, J, Dong, Z, and Bishop, L. Analyzing sentiments and themes on cannabis in Canada using 2018 to 2020 Twitter data. J Cannabis Res. (2022) 4:1–14. doi: 10.1186/S42238-022-00132-1, /TABLES/8
43. Lewis, N, and Sznitman, SR. Engagement with medical cannabis information from online and mass media sources: is it related to medical cannabis attitudes and support for legalization? Int J Drug Policy. (2019) 73:219–27. doi: 10.1016/J.DRUGPO.2019.01.005
44. Freeman, TP, Morgan, CJA, Hindocha, C, Schafer, G, Das, RK, and Curran, HV. Just say “know”: how do cannabinoid concentrations influence users’ estimates of cannabis potency and the amount they roll in joints? Addiction. (2014) 109:1686–94. doi: 10.1111/ADD.12634
45. Fontanella, CA, Steelesmith, DL, Brock, G, Bridge, JA, Campo, JV, and Fristad, MA. Association of cannabis use with self-harm and mortality risk among youths with mood disorders. JAMA Pediatr. (2021) 175:377–84. doi: 10.1001/JAMAPEDIATRICS.2020.5494
46. Han, B, Compton, WM, Blanco, C, and Jones, CM. Time since first cannabis use and 12-month prevalence of cannabis use disorder among youth and emerging adults in the United States. Addiction. (2019) 114:698–707. doi: 10.1111/ADD.14511
47. Weinberger, AH, Zhu, J, Lee, J, Anastasiou, E, Copeland, J, and Goodwin, RD. Cannabis use among youth in the United States, 2004-2016: faster rate of increase among youth with depression. Drug Alcohol Depend. (2020) 209:107894. doi: 10.1016/J.DRUGALCDEP.2020.107894
48. Hall, WD. Cannabis use and the mental health of young people. Aust N Z J Psychiatry. (2006) 40:105–13. doi: 10.1080/J.1440-1614.2006.01756.X
49. Urits, I, Gress, K, Charipova, K, Li, N, Berger, AA, Cornett, EM, et al. Cannabis use and its association with psychological disorders. Psychopharmacol Bull. (2020) 50:56.
50. Karila, L, Roux, P, Rolland, B, Benyamina, A, Reynaud, M, Aubin, H-J, et al. Acute and long-term effects of cannabis use: a review. Curr Pharm Des. (2014) 20:4112–8. doi: 10.2174/13816128113199990620
51. Oberbarnscheidt, T, and Miller, NS. The impact of Cannabidiol on psychiatric and medical conditions. J Clin Med Res. (2020) 12:393–403. doi: 10.14740/JOCMR4159
52. Arango, C, Dragioti, E, Solmi, M, Cortese, S, Domschke, K, Murray, RM, et al. Risk and protective factors for mental disorders beyond genetics: an evidence-based atlas. World Psychiatry. (2021) 20:417–36. doi: 10.1002/WPS.20894
53. Koiliari, EI, Mouzas, I, Alevizopoulos, G, Lesch, O, Walter, H, and Pasparakis, EL. Long – acting injectable aripiprazole in patients with psychosis is associated with improved quality of life, better general clinical outcome and fewer hospitalizations. Eur Psychiatry. (2024) 67:S729. doi: 10.1192/J.EURPSY.2024.1516
54. Rozin, E, Vanaharam, V, D’Mello, D, Palazzolo, S, and Adams, C. A retrospective study of the role of long-acting injectable antipsychotics in preventing rehospitalization in early psychosis with cannabis use. Addict Behav Rep. (2019) 10:221. doi: 10.1016/j.abrep.2019.100221
55. Rotter, A, Bayerlein, K, Hansbauer, M, Weiland, J, Sperling, W, Kornhuber, J, et al. CB1 and CB2 receptor expression and promoter methylation in patients with cannabis dependence. Eur Addict Res. (2013) 19:13–20. doi: 10.1159/000338642
56. Gorelick, DA, Levin, KH, Copersino, ML, Heishman, SJ, Liu, F, Boggs, DL, et al. Diagnostic criteria for cannabis withdrawal syndrome. Drug Alcohol Depend. (2012) 123:141. doi: 10.1016/J.DRUGALCDEP.2011.11.007
57. Winters, KC, and Lee, CYS. Likelihood of developing an alcohol and cannabis use disorder during youth: association with recent use and age. Drug Alcohol Depend. (2008) 92:239–47. doi: 10.1016/J.DRUGALCDEP.2007.08.005
58. Gobbi, G, Atkin, T, Zytynski, T, Wang, S, Askari, S, Boruff, J, et al. Association of cannabis use in adolescence and risk of depression, anxiety, and suicidality in young adulthood: a systematic review and meta-analysis. JAMA Psychiatr. (2019) 76:426–34. doi: 10.1001/JAMAPSYCHIATRY.2018.4500
59. Johnson, E, Kilgore, M, and Babalonis, S. Label accuracy of unregulated cannabidiol (CBD) products: measured concentration vs. label claim. J Cannabis Res. (2022) 4:28. doi: 10.1186/S42238-022-00140-1
60. Bonn-Miller, MO, Loflin, MJE, Thomas, BF, Marcu, JP, Hyke, T, and Vandrey, R. Labeling accuracy of cannabidiol extracts sold online. JAMA. (2017) 318:1708. doi: 10.1001/JAMA.2017.11909
61. Vickery, AW, and Finch, PM. Cannabis: are there any benefits? Intern Med J. (2020) 50:1326–32. doi: 10.1111/IMJ.15052
62. Sholler, DJ, Schoene, L, and Spindle, TR. Therapeutic efficacy of Cannabidiol (CBD): a review of the evidence from clinical trials and human laboratory studies. Curr Addict Rep. (2020) 7:405–12. doi: 10.1007/S40429-020-00326-8
63. Rapin, L, Gamaoun, R, El Hage, C, Arboleda, MF, and Prosk, E. Cannabidiol use and effectiveness: real-world evidence from a Canadian medical cannabis clinic. J Cannabis Res. (2021) 3:19. doi: 10.1186/S42238-021-00078-W
64. Van Dolah, HJ, Bauer, BA, and Mauck, KF. Clinicians’ guide to cannabidiol and hemp oils. Mayo Clin Proc. (2019) 94:1840–51. doi: 10.1016/J.MAYOCP.2019.01.003
65. Huestis, MA, Solimini, R, Pichini, S, Pacifici, R, Carlier, J, and Busardò, FP. Cannabidiol adverse effects and toxicity. Curr Neuropharmacol. (2019) 17:974–89. doi: 10.2174/1570159X17666190603171901
66. Meissner, H, and Cascella, M. Cannabidiol (CBD) StatPearls. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing. (2025). Available at: https://www.ncbi.nlm.nih.gov/books/NBK556048/
67. National Academies of Sciences E and M, Division H and M, Practice B on PH and PH, Agenda C on the HE of MAER and R. (2017). Therapeutic effects of cannabis and cannabinoids. Treasure Island (FL): StatPearls Publishing. Available at: https://www.ncbi.nlm.nih.gov/books/NBK556048/
68. Copas, G, Amazonas, E, and Brandon, S. The pharmacology of cannabinoids In: S Cital, K Kramer, L Hughston, and J Gaynor, editors. Cannabis therapy in veterinary medicine: a complete guide. Switzerland: Springer (2021). 17–59.
69. Miranda-Cortés, A, Mota-Rojas, D, Crosignani-Outeda, N, Casas-Alvarado, A, Martínez-Burnes, J, Olmos-Hernández, A, et al. The role of cannabinoids in pain modulation in companion animals. Front Vet Sci. (2023) 9:884. doi: 10.3389/FVETS.2022.1050884
70. Miller, LK, and Devi, LA. The highs and lows of cannabinoid receptor expression in disease: mechanisms and their therapeutic implications. Pharmacol Rev. (2011) 63:461–70. doi: 10.1124/PR.110.003491
71. Grotenhermen, F, and Müller-Vahl, K. The therapeutic potential of Cannabis and cannabinoids. Dtsch Arztebl Int. (2012) 109:495–501. doi: 10.3238/ARZTEBL.2012.0495
72. Van Draanen, J, Krishna, T, Tsang, C, and Liu, S. Keeping up with the times: how national public health and governmental organizations communicate about cannabis on Twitter. Subst Abuse Treat Prev Policy. (2019) 14:1–7. doi: 10.1186/S13011-019-0224-3/TABLES/2
73. van Draanen, J, Tao, HD, Gupta, S, and Liu, S. Geographic differences in Cannabis conversations on Twitter: Infodemiology study. JMIR Public Health Surveill. (2020) 6:e18540. doi: 10.2196/18540
74. Rojas-Jara, C, Polanco-Carrasco, R, Cisterna, A, Hernández, V, Miranda, F, Moreno, A, et al. Medicinal use of cannabis: a review of the evidence. Ter Psicol. (2019) 37:166–80. doi: 10.4067/S0718-48082019000200166
75. la Torre, JA, De, E-OFJ, Molina, AJ, Colell, E, Dávila-Batista, V, Moreno-Montero, F, et al. Relationship of problematic cannabis use among youth in Spain with perceived risk, environmental factors and sociodemographic factors. Adicciones. (2021) 33:63–70. doi: 10.20882/ADICCIONES.1256
76. Mendiburo-Seguel, A, Vargas, S, Oyanedel, JC, Torres, F, Vergara, E, and Hough, M. Attitudes towards drug policies in Latin America: results from a Latin-American survey. Int J Drug Policy. (2017) 41:8–13. doi: 10.1016/J.DRUGPO.2016.10.001
77. Cáceres Guido, P, Riva, N, Calle, G, Dell’Orso, M, Gatto, M, Sberna, N, et al. Medicinal cannabis in Latin America: history, current state of regulation, and the role of the pharmacist in a new clinical experience with cannabidiol oil. J Am Pharm Assoc (2003). (2020) 60:212–5. doi: 10.1016/J.JAPH.2019.09.012
78. Castaldelli-Maia, JM, Wang, YP, Brunoni, AR, Faro, A, Guimarães, RA, Lucchetti, G, et al. Burden of disease due to amphetamines, cannabis, cocaine, and opioid use disorders in South America, 1990-2019: a systematic analysis of the global burden of disease study 2019. Lancet Psychiatry. (2023) 10:85–97. doi: 10.1016/S2215-0366(22)00339-X
79. Litvinova, O, Baral, B, Wochele-Thoma, T, Matin, M, Tzvetkov, NT, Adamska, O, et al. Efficiency and safety of cannabinoid medical use: an analysis of discussions and observed trends on Instagram. Front Public Health. (2024) 12:1494018. doi: 10.3389/FPUBH.2024.1494018 /XML/NLM
80. Borodovsky, JT, Marsch, LA, and Budney, AJ. Studying cannabis use behaviors with facebook and web surveys: methods and insights. JMIR Public Health Surveill. (2018) 4:e48. doi: 10.2196/PUBLICHEALTH.9408
Keywords: cannabis, marihuana, hachis, Twitter, infodemiology
Citation: Alvarez-Mon MA, Ojeda C, Lara-Abelenda F, Asunsolo del Barco Á, Fraile-Martínez O, García-Montero C, Fernández-Rojo S, Quintero J, Ortega MA, Alvarez-Mon M and Mora F (2025) Understanding the online landscape of cannabis discourse: a Twitter analysis. Front. Public Health. 13:1416171. doi: 10.3389/fpubh.2025.1416171
Edited by:
Eman Leung, The Chinese University of Hong Kong, ChinaReviewed by:
Ismael Hernández Avalos, National Autonomous University of Mexico, MexicoEmmanouil Pasparakis, General Hospital of Agios Nikolaos, Greece
Copyright © 2025 Alvarez-Mon, Ojeda, Lara-Abelenda, Asunsolo del Barco, Fraile-Martínez, García-Montero, Fernández-Rojo, Quintero, Ortega, Alvarez-Mon and Mora. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Cielo García-Montero, Y2llbG8uZ21vbnRlcm9AZ21haWwuY29t