Your Social Media Posts May Hold Clues to Your Mental Health

By Jenni Deming

Social mediaResearchers are now studying the link between a person’s online posts and their mental health status. In two separate studies — one of Twitter and the other of Instagram — scientists predicted depressive symptoms based on social media activity with about 70 percent accuracy.

That’s promising news, considering a majority of the estimated 300 million people suffering from depression around the world are not diagnosed or receiving treatment.1

Scientists believe that, if they can develop smarter computer models to scan our online interactions, they can also help more people understand their symptoms and seek the help they need.

Twitter Findings

TwitterIn the Twitter study, researchers at Microsoft Research Redmond scanned more than 69,000 posts from 489 total users — 117 of whom had a history of depression. Their computer model looked for factors like word choice, time of day, frequency of posts and overall interaction with other users.2

Twitter users experiencing depression tended to:

  • Use words like “lonely,” “sad,” “blame,” “problems,” “nobody,” “pressure,” “weak,” “uncomfortable,” “escape” and “painful”
  • Engage less with other posts and have an overall lower volume of posts and replies
  • Follow fewer accounts and have fewer followers
  • Show an increase in posts late at night

The same researchers went a step further by studying women who’d recently given birth. This time, they were looking for postpartum depression indicators.3 By using a similar computer model, they were able to predict mothers suffering from depressive symptoms with up to 80 percent accuracy.

Instagram Findings

InstagramIn another study, Harvard researchers focused on Instagram photos.4 They collected more than 43,000 photos from 166 users — 71 of whom had a history of depression.

Instagram users experiencing depression tended to:

  • Use bluer, darker and grayer saturations
  • Receive more comments, but fewer overall likes
  • Have fewer faces in their photos
  • Be less likely to apply Instagram filters
  • Disproportionately favor the Inkwell (colorless, black-and-white) filter

So How Do These Studies Help Us?

While doctors can’t access a patient’s social media feed without permission, Microsoft researchers argue it’s still an extremely valuable resource:

“Social media is a source of population data about behaviors, thoughts, and emotions, and can serve as record and sensor for events in peoples’ (sic) lives. Whether in the form of explicit commentary, patterns of posting, or in the subtleties of language used, social media posts bear the potential to offer evidence as to how a person is affected by life events.”3

In other words, we are constantly self-reporting about how we’re feeling, whether we mean to or not. And if science can eventually harness this information to help us live healthier lives, that’d be a major medical win.

Until then, we can use this information to scan our own social media feeds. Maybe you’ve noticed similar patterns in your updates or a friend or family member’s. If so, don’t jump to conclusions. Always leave the diagnosing to professionals.

But it’s a good idea to check in and see how you’re feeling. Have you been irritable, restless, withdrawn, angry or sad for several weeks or months? Or in the case of a friend or family member, have you noticed these symptoms in them?

If you’re concerned about someone you love, schedule some face-to-face time and see how they’re actually doing. Ask if everything is okay and if there’s anything you can do to relieve some stress.

Remember, the majority of depression cases go unreported and untreated, so don’t be afraid to seek professional guidance if you need it. The good thing about depression is that there are tons of ways to manage it and lead a happy, hopeful life. It just takes a few initial steps to get there.


Sources:

1Depression.” World Health Organization, February 2017.

2 De Choudhury, Munmun, et al. “Social Media as a Measurement Tool of Depression in Populations.” Microsoft Research, May 2013.

3 De Choudhury, Munmun, et al. “Predicting Postpartum Changes in Emotion and Behavior via Social Media.” Microsoft Research, May 2013.

4 Reece, Andrew, and Christopher Danforth. “Instagram Photos Reveal Predictive Markers of Depression.” EPJ Data Science, August 2017.