Epic uses cookies to improve your experience on our website.
By continuing to access Epic.com, you will receive all cookies from the Epic website.
You may adjust your browser settings if you do not wish to receive cookies.
Read our privacy policy here.
Menu
New feed articles every Monday

Cool Things

What’s on Your Mind? Social Media Language Predicts Depression Diagnosis

Facebook statuses could help providers identify depressed patients

Most people write their Facebook updates with family and friends in mind, but your statuses might be even more valuable to your doctor. What you post online could one day be an accurate, unobtrusive way for clinicians to identify whether you are at risk for depression, according to a recent study.

“What people write in social media and online captures an aspect of life that’s very hard in medicine and research to access otherwise,” says Dr. H. Andrew Schwartz, senior author of the study. “Depression… really changes people’s use of social media.”

Schwartz and other researchers from the University of Pennsylvania and Stony Brook University designed an algorithm that detects those changes and uses them to predict a diagnosis of depression. Almost 1,200 study participants granted the researchers access to their EHR and Facebook history, data that researchers computationally examined for patterns.

Analyzing more than half a million individual Facebook status updates, they saw that certain linguistic indicators were more common in the statuses of people with a recorded diagnosis of depression. These indicators included specific words, like “miss” and “pain,” and more frequent first-person pronouns, like “I” or “me.” Using these indicators, researchers were able to consistently predict which participants would have a depression diagnosis in their records, even based on data from up to three months before the illness was first documented.

While there’s no clear consensus on whether using social media has a more positive or negative effect on a person’s mental health, these findings show that it “may turn out to be an important tool for diagnosing, monitoring, and eventually treating [depression],” Schwartz says.

Read the complete study in the Proceedings of the National Academy of Sciences.