People struggling with their mental health are more likely to browse negative content online, and in turn, that negative content makes their symptoms worse, according to a series of studies by researchers at MIT.

The group behind the research has developed a web plug-in tool to help those looking to protect their mental health make more informed decisions about the content they view.

The findings were outlined in a paper by Tali Sharot, an affiliate professor of cognitive neurosciences at MIT and professor at University College London, and Christopher A. Kelly, who was a member of Sharot's Affective Brain Lab when the studies were conducted, and now a postdoctoral fellow at Stanford University’s Institute for Human Centered AI. The findings were published in the most recent edition of the journal Nature Human Behavior.

"Our study shows a causal, bidirectional relationship between health and what you do online. We found that people who already have mental health symptoms are more likely to go online and more likely to browse for information that ends up being negative or fearful," Sharot says. "After browsing this content, their symptoms become worse. It is a feedback loop."

The studies analyzed the web browsing habits of more than 1,000 participants by using natural language processing to calculate a negative score and a positive score for each web page visited, as well as scores for anger, fear, anticipation, trust, surprise, sadness, joy, and disgust. Participants also completed questionnaires to assess their mental health and indicated their mood directly before and after web-browsing sessions. The researchers found that participants expressed better moods after browsing less negative web pages, and participants with worse pre-browsing moods tended to browse more negative web pages.

In a subsequent study, participants were asked to read information from two web pages randomly selected from either six negative webpages or six neutral pages. They then indicated their mood levels both before and after viewing the pages. An analysis found that participants exposed to negative web pages reported to be in a worse mood than those who viewed neutral pages, and then subsequently visited more negative pages when asked to browse the internet for 10 minutes.

"The results contribute to the ongoing debate regarding the relationship between mental health and online behavior," the authors wrote. "Most research addressing this relationship has focused on the quantity of use, such as screen time or frequency of social media use, which has led to mixed conclusions. Here, instead, we focus on the type of content browsed and find that its affective properties are causally and bidirectionally related to mental health and mood."

To test whether intervention could alter web-browsing choices and improve mood, the researchers provided participants with search engine results pages with three search results for each of several queries. Some participants were provided labels for each search result on a scale of “feel better” to “feel worse.” Other participants were not provided with any labels. Those who were provided with labels were less likely to choose negative content and more likely to choose positive content. A follow-up study found that those who viewed more positive content reported a significantly better mood.

Based on these findings, Sharot and Kelly created a downloadable plug-in tool that offers scores for Google search results in three categories: emotion (whether people find the content positive or negative, on average), knowledge (to what extent information on a webpage helps people understand a topic, on average), and actionability (to what extent information on a webpage is useful on average). Jonatan Fontanez, a former undergraduate researcher from MIT in Sharot's lab, also contributed to the development of the tool. He is now an MIT computer science graduate student. The tool was introduced publicly this week, along with the publication of the paper in Nature Human Behavior.

“People with worse mental health tend to seek out more negative and fear-inducing content, which in turn exacerbates their symptoms, creating a vicious feedback loop,” Kelly says. “It is our hope that this tool can help them gain greater autonomy over what enters their minds and break negative cycles.”