Vaccination rates are declining in many communities due to the spread of misinformation, leading to a resurgence of diseases once thought under control, such as measles, across the United States and Canada. Researchers at the University of Waterloo have developed a new predictive approach that could help public health officials anticipate where outbreaks might emerge. By analyzing social media content, the method detects early signs of rising vaccine skepticism—potential warning signals that could appear well before any disease begins to spread. The study, titled "Forecasting infectious disease outbreak risks from vaccine sentiments on social media: A Data-driven dynamical systems approach," has been published in Mathematical Biosciences and Engineering. Led by Dr. Chris Bauch, professor of Applied Mathematics at the University of Waterloo, the research treats social dynamics like a contagious system, similar to how diseases spread. “In nature, we have contagious systems like diseases,” Bauch explained. “We decided to look at social dynamics as an ecological system and study how misinformation spreads from user to user across social media networks.” The team used machine learning to apply the mathematical concept of a tipping point—the moment when a system rapidly shifts from one state to another. “It doesn’t matter if you're looking at a person’s body having an epileptic seizure, an ecological system like a lake overrun by algae, or the loss of herd immunity in a population,” Bauch said. “Mathematically, there’s a common underlying mechanism.” To test the model, researchers analyzed tens of thousands of public posts from X (formerly Twitter) in California just before the major measles outbreak in 2014. Traditional methods—such as counting the number of vaccine-skeptical tweets—offered little advance warning. “The usual methods of predicting an outbreak by doing a statistical analysis of skeptical tweets don’t provide much lead time,” Bauch said. “By using the mathematical theory of tipping points, we were able to get a much earlier warning and detect subtle patterns in the data more effectively.” The model’s accuracy was confirmed by comparing social media activity in California to that in other similar regions at the same time, where no outbreaks occurred. This research aligns with the University of Waterloo’s broader mission to strengthen evidence-based decision-making and rebuild public trust in science—a key goal of the university’s Societal Futures network and its new TRuST initiative. That initiative brings together philosophers, computer scientists, communicators, and ethicists to explore why trust in science erodes and how it can be restored. While the model was initially tested on X, it can be adapted for platforms like TikTok or Instagram. However, analyzing images and videos from those platforms would require significantly more computing power than text-based analysis. “Ultimately, we would like to turn this into a practical tool for public health officials to monitor which populations are at the highest risk of reaching a tipping point,” Bauch said. “Applied mathematics can be a powerful quantitative tool in predicting, monitoring, and addressing threats to public health.” More information: Zitao He et al, Forecasting infectious disease outbreak risks from vaccine sentiments on social media: A data-driven dynamical systems approach, Mathematical Biosciences and Engineering (2025). DOI: 10.3934/mbe.2025101