Can Facebook Profiles Serve as an Indicator for Medical Conditions?

A recent study published in PLoS One study suggests that Facebook profile updates can indicate over 20 medical conditions, particularly mental health conditions such as anxiety, depression, and psychoses.

Researchers of this study analyzed approximately 20 million words written on the Facebook pages of 999 consenting participants. All participants were drawn from the ongoing Social Mediome Study, which initially began recruitment in March 2014. The participants were invited to share their past social media activity and electronic medical record (EMR) data. In total, they observed 949,530 Facebook status updates and approximately 20 million words from participants whose status updates exceeded 500 words.

Social Media Reveals A Lot

The results of the study showed that all 21 medical condition categories were predictable from Facebook language beyond chance (multi-test corrected p < .05), 2) and 18 categories were better predicted from a combination of demographics and Facebook language than by demographics alone. Moreover, 10 categories were found to better predict Facebook language than by the standard demographic factors (age, sex, and race). The results depicted the accuracy of the three predictive models across all 21 diagnoses categories. Moreover, the medical condition categories for which Facebook statuses displayed the greatest prediction accuracy gains over demographics include diabetes (AUC = .74), pregnancy (AUC = .79; females only) and the mental health categories anxiety (AUC = .66), psychoses (AUC = .58) and depression (AUC = .64).

“Social media, like genomic information, offers enormous promise to personalize health care,” the study authors wrote. “This work is complementary to a growing body of literature using big data analytics for EMR data and provides new insights for applying machine learning to find signal about health in non-healthcare generated data (e.g. social media).”

“People’s personality, mental state, and health behaviors are all reflected in their social media and all have tremendous impact on health,” they added. “This is the first study to show that language on Facebook can predict diagnoses within people’s health record, revealing new opportunities to personalize care and understand how patients’ ordinary daily lives relate to their health.”