# Understanding the online landscape of cannabis discourse: a Twitter analysis

**Authors:** Miguel Angel Alvarez-Mon, Carla Ojeda, Francisco Lara-Abelenda, Ángel Asunsolo del Barco, Oscar Fraile-Martínez, Cielo García-Montero, Sonia Fernández-Rojo, Javier Quintero, Miguel Angel Ortega, Melchor Alvarez-Mon, Fernando Mora

PMC · DOI: 10.3389/fpubh.2025.1416171 · Frontiers in Public Health · 2025-07-17

## TL;DR

This study analyzes Spanish-language tweets about cannabis to understand public opinions on its regulation and health effects.

## Contribution

The novel aspect is the use of machine learning to analyze a large Spanish-language Twitter dataset on cannabis discourse.

## Key findings

- 73.2% of tweets supported cannabis regulation, while only 3.5% opposed it.
- 20.4% of tweets discussed negative health effects of cannabis.
- 30.1% of tweets supported therapeutic use of cannabis.

## Abstract

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.

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.

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.

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.

## Full-text entities

- **Diseases:** epilepsy (MESH:D004827), psychiatric (MESH:D001523), CUD (MESH:D002189), MS (MESH:D009103), anorexia (MESH:D000855), AIDS (MESH:D000163), hallucination (MESH:D006212), cognitive dysfunction (MESH:D003072), nausea and vomiting (MESH:D020250), mental or substance abuse disorders (MESH:D019966), spasticity (MESH:D009128), mental health disorders (OMIM:603663), weight loss (MESH:D015431), schizophrenia (MESH:D012559), hyperemesis syndrome (MESH:D006939), mood disorders (MESH:D019964), pain (MESH:D010146), paranoia (MESH:D010259), depressive symptoms (MESH:D003866), overdose (MESH:D062787), post-traumatic stress disorder (MESH:D013313), anxiety (MESH:D001007), chronic pain (MESH:D059350), Lennox-Gastaut syndrome (MESH:D065768), Dravet syndrome (MESH:D004831), COPD (MESH:D029424), cancer (MESH:D009369), psychosis (MESH:D011618), insomnia (MESH:D007319), inflammation (MESH:D007249)
- **Chemicals:** cannabinoid (MESH:D002186), opiates (MESH:D053610), THC (MESH:D013759), flavonoids (MESH:D005419), nabilone (MESH:C011941), terpenes (MESH:D013729), cannabis oil (-), endocannabinoid (MESH:D063388), CBD (MESH:D002185), aripiprazole (MESH:D000068180), alkaloids (MESH:D000470), cocaine (MESH:D003042)
- **Species:** Homo sapiens (human, species) [taxon 9606], Cannabis sativa (species) [taxon 3483]

## Full text

_Full body text omitted from this summary view._ Fetch the complete paper as Markdown: https://tomesphere.com/paper/PMC12310699/full.md

## Figures

4 figures with captions in the complete paper: https://tomesphere.com/paper/PMC12310699/full.md

## References

81 references — full list in the complete paper: https://tomesphere.com/paper/PMC12310699/full.md

---
Source: https://tomesphere.com/paper/PMC12310699