Your Instagram photos might say much more about you than what you had for lunch or where you’re vacationing this summer. They might also hold clues about whether you are depressed.
New research shows that people with a history of depression share more photos, use fewer filters, and post more images that are darker and grayer in color. The findings, published this week in the journal EPJ Data Science, suggest that Instagram and other social media sites are more tools that can be used to screen people for mental illnesses.
Previous studies have shown that how you word your tweets or your Facebook status updates might indicate whether or not you’re depressed. Instead of mining social media data for words, this latest study looked at visual cues. Researchers at Harvard and the University of Vermont looked at more than 43,000 photos from 166 users, 71 of whom had a history of depression.
They used machine learning to analyze the photos’ hues, brightness, filters, and whether they featured faces. Their findings: depressed users posted more photos, and photos that tended to be “bluer, darker, and grayer.” They also used fewer Instagram filters; when they did use a filter, they tended to choose “Inkwell,” which makes the photo black and white. Healthier users favored the “Valencia” filter instead, which lightens the tint of photos. Depressed users were also more likely to post photos with faces, but compared to healthier users, those photos contained fewer faces per image. That could be a sign that people with depression “interact in smaller social settings,” the study says.
The results obviously don’t apply to all Instagram users: the researchers analyzed thousands of photos, but they were from a relatively small number of people. The participants also had to meet certain criteria to be included in the study: they had to be active on Instagram, willing to share their entire Instagram posting history, as well as whether they had received an actual diagnosis of depression. Users were classified as depressed based on their responses to a standard clinical depression survey.
But the technique used in this study “may serve as a blueprint for effective mental health screening in an increasingly digitalized society,” the study concludes.