Fluoride-related misinformation analysis on Twitter: Infodemiology study
More details
Hide details
School of Public Health Sciences, University of Waterloo, Waterloo, ON, Canada, Canada
School of Public Health Sciences, University of Waterloo, Waterloo, ON, Canada, India
Department of Pediatric Dentistry, Orthodontics and Public Health, Bauru School of Dentistry, University of São Paulo, Bauru, Brazil, Brazil
Publication date: 2023-04-26
Popul. Med. 2023;5(Supplement):A591
Background & Objectives:
The presence of misinformation on social media significantly harms communities since accessing trustworthy information is a ground principle for people to achieve better health-related outcomes. In this context, antifluoridation content is broadly shared on social media, deceiving people about fluoride’s relevance and safety. Considering the potential negative implications of online fluoride falsehoods, the present study aimed to analyze fluoride-related misinformation on Twitter.

First, 23,436 tweets were collected using the Twitter Application Programming Interface (API) from the keyword “fluoride-free” between May 2016 and May 2022. After data preprocessing, Latent Dirichlet Allocation (LDA) topic modeling techniques were applied with remain 21,169 tweets to determine the salient topics linked to the falsehood. Finally, time series decomposition methods were applied to define fluoride misinformation trends and seasonality over time using pandas, NumPy, and statsmodels.tsa.seasonal libraries in Python 3.

From a coherence score of 0.542, a total of 3 different salient topics emerged from the LDA topic modeling analysis. As a result, fluoride-related misinformation was mainly associated with the people’s perception of a healthy lifestyle (topic 1), followed by the consumption of natural and organic oral care products (topic 2), and recommendations of fluoride-free products and measures (topic 3). Notably, all topics displayed a decreasing trend between 2016-2019 and an increasing trend from 2020 onwards. Moreover, they presented a seasonal pattern of searches with peaks mainly at the beginning and middle of each year.

Fluoride misinformation was primarily related to personal concerns about a healthier lifestyle. The recent increasing trends of falsehoods have probably contributed to the popularization of fluoride-free products and the suspension of fluoridated water community programs. Hence, user-centered digital health strategies are required to control the spread and consumption of fluoride misinformation on social media.