PHILADELPHIA — If you are concerned about the mental health of that one friend from high school who just won’t stop posting on Facebook, it turns out you may be onto something.
According to a new study conducted by researchers from Penn Medicine and Stony Brook University, language used in Facebook posts may help identify diabetes and mental health conditions such as anxiety, psychosis, and depression in patients. Researchers are hopeful that, given patient consent, Facebook activity could potentially be monitored just like more traditional physical symptoms.
The Facebook post history of almost 1,000 patient volunteers were analyzed using an automated data collection technique, and linked to each patient’s electronic medical records. After that, researchers constructed three different models to analyze the predictive power of Facebook; one model only analyzed Facebook post history, another only used demographic information such as age and sex, and the last utilized a combination of demographic and Facebook information.
Twenty-one health conditions were investigated, and amazingly, researchers concluded that all 21 were predictable through the use of Facebook alone. Additionally, 10 of those conditions were found to be more accurately predicted using Facebook than demographic information.
Facebook profiles that frequently used the words “drink” and “bottle” predicted alcohol abuse, and users who often used religious verbiage, such as “God” and “pray,” were found to be 15 times more likely to have diabetes than profiles that rarely used religious words. Profiles that typically showed aggressive language such as “dumb,” or any number of curse words, accurately pointed to drug abuse and psychosis.
“This work is early, but our hope is that the insights gleaned from these posts could be used to better inform patients and providers about their health,” says lead author Dr. Raina Merchant, the director of Penn Medicine’s Center for Digital Health and an associate professor of Emergency Medicine, in a statement. “As social media posts are often about someone’s lifestyle choices and experiences or how they’re feeling, this information could provide additional information about disease management and exacerbation.”
Merchant will be conducting another study later this year in which patients will be asked to share their social media profiles with their health care providers. This additional trial will help researchers determine if applying social media data to healthcare is actually helpful in a case by case basis, as well as determining how many people would be comfortable volunteering such information.
“For instance, if someone is trying to lose weight and needs help understanding their food choices and exercise regimens, having a healthcare provider review their social media record might give them more insight into their usual patterns in order to help improve them,” Merchant explains.
This study was published in the scientific journal PLOS ONE.