Published this week in the American Journal of Public Health, the study compared Twitter posts containing cannabis-related terms between May 1 and Dec. 31, 2018. The review considered those using “methods to distinguish between posts from social bots and non-bots,” and using text classifiers to identify post topics.

The tweets were then coded into a dozen categories, such as health and medical, legal, first-time consumption, processed products like edibles and beverages, and polysubstance use — consuming cannabis alongside other substances such as painkillers or alcohol.

“What we found was that the proportion of bot posts that talked about health claims was larger than the proportion among non-bot accounts,” doctoral student and study co-author Patricia Escobedo. Escobedo also noted the researchers did not encounter a single reference to scientifically backed medical applications for cannabis, such as childhood epilepsy or seizures, in their findings.

Researchers found posts from bots suggesting that cannabis could help a wide variety of health issues, including cancer, Crohn’s disease, depression, trauma and plantar fasciitis beyond symptomatic relief.