Your Facebook profile can tell if you have a health condition, study finds


Your social media might say a lot more about you than you think.

Say, for example, if you have a medical condition.

In a new study, researchers were able to predict 21 types of medical conditions – ranging from pregnancy to skin disorders – by analyzing people’s Facebook profiles.

Facebook status updates were “particularly effective in predicting diabetes and mental health problems, including anxiety, depression and psychoses”, according to the study.

The study will be published June 19 in PLOS One, a peer-reviewed open-access scientific journal published since 2006.

“People’s personality, mental state and health behaviors are all reflected in their social media and all have a huge impact on health,” the study says.

To conduct the study, the researchers linked consenting patients’ electronic medical records to their social media.

In total, they looked at 949,530 Facebook status updates from 999 participants whose posts were longer than 500 words.

The researchers identified language that likely indicated behavior or symptoms characteristic of certain diagnoses. For example, posts mentioning “drinking”, “drunk”, or “bottle” were flagged for alcohol abuse.

However, the study cautioned that predictive words are not necessarily causal mechanisms. Yet, social media models have predictive value and could open up new possibilities for personalizing healthcare.

“The power of social media language to predict diagnoses raises parallel questions about privacy, informed consent, and data ownership,” the researchers wrote.

In recent years, Facebook has come under fire for its user privacy practices. The tech giant earned $56 billion in 2018 by tracking users and using that data to sell targeted ads.

Facebook CEO Mark Zuckerberg acknowledged the company’s previous privacy issues, including the Cambridge Analytica scandal and a security breach that exposed at least 50 million users.

In March, Zuckerberg revealed plans to reposition the company as a “privacy-centric” platform.