"Doctor, please see my Facebook posts on God. Am I likely to be a diabetic?"

Suffering from medical conditions such as diabetes, anxiety, depression and psychosis? Worry not. Dr Facebook is here to diagnose and rescue. But do you want to be?

Have you been posting about religion and using religious language like "God" or "pray" in your Facebook posts recently? And do you think it is sweet of you to do so? Maybe not. Maybe it is time to go off real sweets for a bit and get your blood glucose levels checked as you might be 15 times more likely to have diabetes than your friends who have not posted about God in a while now.

Or so says a medical research by Penn Medicine and Stony Brook University scientists that was published in PLOS ONE science journal. According to them, language in Facebook posts may help identify medical conditions — such as diabetes, anxiety, depression and psychosis in patients —as it is believed that language in social media posts could be indicators of diseases.

Your Facebook status may be revealing more than you think. New research uses social media posts to predict health conditions like diabetes, anxiety and STDs. https://t.co/gsrNQYCGvk #Facebook #socialmedia pic.twitter.com/B5RtJ5oQdb

— PLOS ONE (@PLOSONE) June 17, 2019

The researchers used an automated data collection technique and analysed the Facebook post history of nearly 1,000 users. These users agreed to have their electronic medical record data linked to their profiles.

It was found that all the 21 different medical conditions that the researchers were looking to determine, could be predicted from Facebook posts alone, as opposed to demographic data (age, location, gender etc). In fact, 10 of the conditions were better predicted through the use of Facebook data instead of demographic information, the report says.

Some of the Facebook data that was found to be more predictive seemed intuitive.

"For example, alcohol abuse was marked by a topic mentioning drink, drunk, bottle. Topics expressing hostility (e.g. people, dumb, bulls**t, b**ches) were the predominant marker of drug abuse and also marked psychoses. Topics most associated with depression suggested somatization (e.g. stomach, head, hurt) and emotional distress (e.g. pain, crying, tears),” the research paper states.  

Meet your new doctor: Dr Facebook. (Photo: Getty Images)

However, other markers may not be so direct. For example, “diabetes was predicted by religious language (e.g. god, family, pray) even when controlling for demographic variables. This does not mean that everyone mentioning these topics has the condition, but just that those mentioning it are more likely to have it. For example, those mentioning the topic were 15 times more likely to have been diagnosed with diabetes than those less likely to mention it,” the paper states.

This association (of religion and diabetes) suggests the potential to explore the role of religion in diabetes management or control, the authors opine. "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," says lead author Raina Merchant, an associate professor at the University of Pennsylvania. "Our digital language captures powerful aspects of our lives that are likely quite different from what is captured through traditional medical data," adds the study's senior author Andrew Schwartz, who is also an assistant professor at Stony Brook University.

While the results seem promising, the privacy risks of allowing doctors access to social media pose a serious concern in the scope of the study.

There have been enough and more debates on the breaching of data security by Facebook. Not one shred of information, it seems, is private — even if you opt-in for the information to be confidential. While the “terms and conditions” may specify that the data could be shared even if you have asked for it not to be, let’s face it — how many of us read the fine print?

Would you trust Facebook with your health profile despite the security breaches in the past? (Photo: Reuters)

In this scenario, what is the guarantee of sensitive information like health diagnoses staying private? Further, if it is diagnosed by posts alone, how accurate will this data be in determining graver and more chronic illnesses? And if it is inaccurate even by 0.01%, it should be scrapped at the bud stage.

This is because if social media posts suddenly become part of the user’s health records, it would give access to the insurance companies too. If the health prediction is off even by a tiny chance, it would mean that the user will have to pay a higher premium — and unfairly so. This is because insurance companies can use your social media posts to decide on the premium. It is perfectly legal in cities like New York, for instance, where the insurance companies can use data from social media — as long as they can prove that they are not unfairly discriminating against certain groups.

More importantly, should the prerogative to share, be diagnosed and treated for medical conditions and/or addictions not be a choice of the individual? Is it fair that a social media platform that is meant for networking be determining our health choices?

Unless these questions are addressed, it is not only unsafe but positively dangerous to even explore possibilities unearthed by this research.

Also read: How much does Facebook know about you?