Social networks have their share of disadvantages. They would promote suicide, generate stress or even cause body image disorders. And if they revealed for once a positive point for health, by helping to detect depression? A new study from the American universities of Pennsylvania and Stony Brook, published in the PNAS, shows that the data published on Facebook could participate in it. The researchers analyzed statuses shared by users during the months leading up to the diagnosis of the disease. They discovered that their algorithm could accurately predict this future depression. According to a report from the Weekly Epidemiological Bulletin (BEH) published on October 16, nearly one in ten French adults aged 18 to 75 would have experienced a depressive episode in 2017.
Indicators on social networks
The researchers first identified data from people agreeing to share their Facebook statuses, and analyzed them to distinguish individuals who had a definite diagnosis of depression. To do this, they relied on several indicators: signs of hostility or loneliness, words like “tears” Where “feelings” as well as the more pronounced use of first-person pronouns, such as ” I “ Where ” me “. They then combined these results with the information in their electronic medical records.
To build the algorithm, the scientists examined 524,292 data from Facebook. They determined the most frequently used words and phrases, then modeled 200 subjects to explore what they called “language markers associated with depression”. Finally, they compared how and how often depressed participants used them compared to control participants. They were thus able to predict future depression in the three months before the illness was first documented in the medical record.
A tool to diagnose, monitor and treat
For six years, the World Well-Being Project (WWBP) – based at the Center for Positive Psychology in Pennsylvania and the Stony Brook Human Language Analysis Lab – has studied how the words people choose reflect their feelings.
“What people write on social media and online represents an aspect of life that is very difficult to access in medicine and researchExplain on the University of Pennsylvania website H. Andrew Schwartz, author of the article and senior researcher for the WWBP. This is a relatively untapped dimension compared to biophysical markers of disease. Considering conditions such as depression, anxiety and [trouble post-traumatique], for example, you find more signals in the way people express themselves digitally. »
“There is a perception that using social media is not good for mental health », develops the author of the article. Indeed, according to another study, spending too much time on Facebook, Instagram, YouTube, Snapchat and other social networks would promote the feeling of loneliness. “But it can prove to be an important tool to diagnose, monitor and ultimately treat”he continues.
His colleague Johannes Eichstaedt, also a researcher for the WWBP, sees the long-term use of this data as a potential, more discreet form of screening: “We hope that one day these screening systems can be integrated into healthcare systems. This tool raises yellow flags; we hope that [les médecins pourront] refer identified individuals directly to scalable treatment modalities”.
Read also :
- Social networks, a good way to accept yourself as you are?
- Depression at work: the signs that should alert
