Mammalian H5N1 cases predict online discourse volume and emotion during the 2024–2025 H5N1 outbreak
This study analyzes 42,957 Reddit comments from the 2024–2025 H5N1 outbreak to demonstrate that reported mammalian cases, particularly in domestic cats and dairy cattle, serve as early-warning predictors of online discourse volume and emotional intensity, offering a one-week window for proactive risk communication.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
When a new disease threat emerges, the way people talk about it can be just as important as the disease itself. In the field of public health, understanding the public's emotional reaction helps officials decide how to communicate risks effectively. If people are terrified, they might panic; if they are angry, they might demand action; if they are sad, they might feel helpless. Social media has become a massive, real-time window into these collective feelings. Platforms like Reddit, where users gather in specific communities to discuss niche topics, offer a unique record of how ordinary people process complex scientific news. By listening to these conversations, researchers can see not just what people are worried about, but what specific events trigger those worries. This approach, known as infoveillance, treats online chatter as a signal that can reveal the pulse of public concern during a health crisis.
In the years 2024 and 2025, a highly contagious bird flu virus known as H5N1 began spreading rapidly among animals across the United States. It jumped from wild birds into poultry, dairy cattle, and even domestic cats and wild mammals. While human infections remained rare, the sheer scale of the animal outbreak sparked intense public interest and anxiety. A team of researchers at Cornell University decided to investigate how this unfolding crisis played out in the minds of the online public. They turned to three specific Reddit communities dedicated to bird flu, where users had posted nearly 43,000 comments over the two-year period. Their goal was to map the landscape of these conversations: what were people discussing, how were they feeling, and most importantly, could the number of infected animals predict the volume and emotion of the online discussion?
The researchers first looked at the content of the comments to understand the main themes. They found that the public was naturally weaving together human health, animal health, and food safety into a single, cohesive narrative, even without using the formal scientific term for this connection. The most frequent topics included the safety of the food supply, comparisons to the recent pandemic, and the specific details of the virus. A significant portion of the discussion centered on "raw milk," reflecting a deep concern about whether the virus could be transmitted through unpasteurized dairy products. The emotional tone of these conversations was dominated by fear, which was the most common feeling expressed. However, a second emotion, anger, was the most reactive one, spiking sharply in response to specific news events.
To determine what actually drove these conversations, the team built a statistical model to see if the number of reported animal cases could predict the number of comments posted each week. They tested various factors, including the price of eggs, the volume of news articles, and confirmed cases in poultry, dairy cattle, wild mammals, and domestic cats. The results showed that the number of infected animals was a powerful predictor of online activity, but not all animals triggered the same response. Specifically, reports of H5N1 in domestic cats and dairy cattle were the strongest signals. When the number of infected cats went up, the number of comments increased significantly the following week. Similarly, reports of infected dairy cattle were closely linked to the volume of discussion. In contrast, the number of infected wild birds or poultry did not reliably predict spikes in conversation volume, suggesting that the public was less reactive to those specific reports.
The study also revealed that different animals triggered different emotional responses. When the news involved domestic cats, the online discourse was marked by a surge in anger. This reaction suggests that people view their pets as family members, and an infection in a cat feels like a direct, personal threat that demands a strong emotional response. On the other hand, news about dairy cattle was associated with a dominant feeling of sadness. This emotional signature aligns with a more sustained, lingering concern about the food supply and the integrity of the agricultural system, rather than an immediate, explosive reaction. The researchers found that for every additional case of H5N1 in a cat, the public discussion grew disproportionately larger, indicating that each new case felt like a major event. In contrast, as the number of cattle cases accumulated, the public attention grew more slowly, suggesting a sense of saturation where the news became a steady background worry rather than a breaking story.
One of the most practical findings of the study was the timing of these reactions. The data showed that reports of infected cats provided a one-week early warning signal for a spike in public engagement. This means that health officials could potentially use reports of animal infections to anticipate when the public would become most active and emotional online. By recognizing that a new case in a cat is likely to trigger anger and a surge in discussion, authorities could prepare proactive risk communication strategies. They could address the specific fears about pet safety and food security before misinformation or panic takes hold. The study suggests that the public is already thinking in a connected way about human, animal, and environmental health, and that effective communication should build on this existing understanding rather than trying to teach it from scratch.
The researchers were careful to note that their findings apply specifically to the active, engaged community on Reddit, which may not represent the entire US population. They also emphasized that while their models showed a strong predictive link, they could not prove that the animal cases directly caused the comments; both could be driven by a third factor, such as a major news broadcast. However, the consistency of the patterns across the two-year period offers a clear picture of how the public processes zoonotic threats. The study concludes that by monitoring specific animal hosts, particularly pets and livestock, health communicators can better understand the emotional landscape of a crisis. This allows them to tailor their messages to the specific fears and concerns of the moment, turning a reactive public discourse into a more informed and manageable conversation.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.