People sharing photos on social media are also sharing their feelings—they just don’t know it.
Depressed Instagram users post differently on the platform than other users, a new study suggests. A machine learning algorithm was able to correctly identify markers for depression based on the photos users shared on Instagram.
Researchers say users with a history of depression expressed themselves differently on the platform, from the colors and filters to the faces in photos. According to the New York Times, depressed users tend to share images that were darker as well more gray and blue than healthy users.
Authors of the study Andrew Reece, a postdoctoral researcher at Harvard University and Christopher Danforth, a professor at the University of Vermont, published their findings in the journal EPJ Data Science.
The researchers labeled the 166 participants either “healthy” or “depressed” based on whether or not they said they’d been clinically diagnosed with depression in the past. Using a machine learning algorithm, they found a pattern in the photos posted by both groups, according to the New York Times.