Media and developing minds header

Social Media, Depression,
and Suicide

October 16, 2018

Paul Weigle, MD


Paul Weigle, MD

Child and Adolescent Psychiatrist; Associate Medical Director, Natchaug Hospital Ambulatory Service, Natchaug Hospital; Chair of the Media Committee, American Academy of Child and Adolescent Psychiatry; Advisory Board Member, Children and Screens

Thomas Joiner, PhD

Thomas Joiner, PhD

Robert O. Lawton Distinguished Professor of Psychology, Director of the Laboratory for the Study and Prevention of Suicide-Related Conditions and Behaviors, Director of the Psychology Clinic, Florida State University; Editor-in-Chief, Suicide & Life-Threatening Behavior

Ethan Kross, PhD

Ethan Kross, PhD

Professor of Psychology, Director of the Emotion and Self Control Laboratory, University of Michigan


Dr. Weigle: Depression is becoming more prevalent among teens, based on statistics from 2005 (8.7%) and 2014 (11.3%). According to the Centers for Disease Control, the rate of teen suicide also has risen, since 2007, making it the second leading cause of death for that age group. These trends have paralleled those regarding teen social media use (which increased by 250% from 2010 to 2015). Based on that three-year-old data, 58% of teens will use social media on a given day, and those who do will spend just over two hours that way. Girls spend more time on social media than boys, blacks do more than whites, and Instagram, Snapchat, and Facebook are the most commonly used platforms.

Social media activities have positive uses. They allow users to maintain “real world” friendships and make new friends. They are associated with increased social support and self-esteem. Most teens believe that social media positively impacts their wellbeing. Minority youths, those struggling with their sexual identity, and those with learning disabilities meet others like themselves online. This allows them to feel less lonely and more confident.

Depressed teens prefer to socialize online (rather than in person). Compared to their non-depressed peers, they are more likely to report that social media is important for feeling less alone and more inspired, and for self-expression. They also tend to look for support or information regarding their depression before they seek professional help.

As with all digital media, social media use raises concerns about the displacement of other, healthy behaviors. These include face-to-face social contact; homework, reading, and other academically valuable activities; physical exercise; and sleep. Insufficient sleep, which is epidemic among teens generally, strongly predicts depression and suicidality. More than 60 studies have found a relationship between screen media and insomnia, and both heavy social media use and phone use in bed predict insomnia.

Depressed teens are three times as likely than others to report negative interactions on social media. There are three possible explanations for this. First, they may be more likely to interpret responses as negative. Second, their social media feedback may be adversely affecting their mood. Third, their depression may be inducing them to post unpopular content. They’re also more likely to use social media to avoid their real world problems (undermining one of the potentially positive effects of this activity). In addition, suicidal teens often engage in online discussions about those feelings, and such discussions amplify their suicidality over time.

Meta-analysis suggests just how complex the relationship is between social media and depression. Of the dozens of studies analyzed, five showed a positive relationship; two showed a negative relationship; and four found none at all. This does not necessarily speak to the particular ways that teens (as opposed to adults) use social media.

There are a number of mediating factors that connect social media and depression. Among girls, unpopularity, lack of self-purpose, and emotional investment are associated with such social media effects. Other mediating factors include: (i) heavy or excessive use (“Internet addiction”); (ii) devotion to impression management (“social comparison”); (iii) and relatively passive use (“lurking”); (iv) fear of missing out (“FOMO”); (v) cyberbullying; and (vi) sexting.  Protective factors include having more online friends and participating more actively in online interactions.

It is unsurprising that an unpopular girl’s passive observations of strangers’ idealized social media personae could contribute to her depression. Depressed teens feel inferior when they look at their peers’ curated, edited, and ultimately unrealistic social media profiles. They are also 18 times more likely to feel left out online (18% vs. 1%). This may compound those feelings of inferiority, and induce compulsive or addictive checking of social media – a potentially uncontrolled, excessive habit that causes dysfunction, whether or not it is formally defined as an addiction. This possibility is of particular concern because of the established bi-directional relationship between depression and addiction. Each is a risk factor for the other.

Most young adults interviewed admitted to sexting in their teens. Such behavior is predicted by unsupervised and unrestricted Internet access. It is associated with depression, and in most (but not all) studies, with a 2-3 times higher risk of a suicide attempt. As for cyberbullying, dozens of studies have established that both the perpetrators and the victims are at increased risk for suicidal thoughts and behaviors. Those who play both roles – as bullies and bullied – are at greatest risk. Cyberbullying has been found to predict worsening depression and suicide attempts over time.

Dr. Kross: The search for answers about Facebook and wellbeing begins with a puzzle. A literature review revealed that some studies show a positive link between Facebook usage and how people feel. Others say the exact opposite. Still others invite more nuanced interpretations. For example, the impact of social media experiences may depend on the subject’s number of friends, how supportive those friends seem, and whether the subject is experiencing depressive symptoms.

One explanation for this confusion is that the literature consists mostly of cross-sectional data from large surveys. The Kross team opted to use experience sampling (also known as ecological momentary assessment) instead, to look at subjects’ Facebook usage (and their feelings) deeply and repeatedly over time.

The first of these experience sampling studies involved sending five questions a day to participating college students, every 90-120 minutes, for 14 consecutive days. The fundamental questions were simple: How do you feel, and how much have you used Facebook since the last time we asked you? The study discovered a robust link, suggesting that using Facebook predicted a subsequent decline in mood. Staying off Facebook for a while predicted a mood improvement.

Study subjects also were asked about their face-to-phone or phone interactions between inquiries. These also predicted mood changes, but the opposite of those associated with Facebook. More extensive direct social interaction has mood-lifting benefits, and less of such interaction brings mood down. This confirms many other studies.

The next step in this research was to consider the differential impact of using Facebook in different ways. A study was designed to test whether the potential harm of Facebook was particularly associated with “voyeuristically browsing” (passive exposure to others’ curated, idealized versions of themselves), as opposed to more actively producing information and engaging with others online. This research found evidence that people are using Facebook more passively than actively, and that such passive usage is driving the negative emotional effect. “So, in other words, the more you use Facebook passively, the more it elicits feelings of envy, which in turn explains how passively using Facebook leads to a decline in how good people feel.” Surprisingly, no link was found between active usage and positive emotional effects.

Follow-up laboratory studies replicated aspects of the field research, to search for a causal connection. Again, active Facebook use produced no change in how good people felt. Neither was the difference between how the subjects felt before and after they scrolled through Facebook for ten minutes. An emotional decline did appear, however, by the end of the day. This suggests a “ruminative component” – that as you reflect on what you saw on Facebook and how well others seemed to be doing compared to you, your mood declines. There is substantial support in the literature for this link between passive social media usage, envy, and feeling worse. The data on positive effects from more active social media engagement is conflicting, suggesting the need for further investigation.

In the offline world, depressed people talk a lot about what’s bothering them. This drives away people who might otherwise provide support. A study about whether that’s also true on Facebook yielded surprising findings. Study participants were clinically depressed. Researchers took all the posts, responses, and network responses to responses on the study subjects’ Facebook pages; de-identified that content; and coded it for valence and supportiveness. On Facebook, as in the real world, depressed people were found to self-disclose much more negative information than their healthier counterparts. On the other hand, while such self-disclosure offline drives people away, on Facebook, it actually attracts more supportive responses. This suggests that social media may be employed, strategically, to improve wellbeing.

In sum, as others have observed, the effects of social media usage depend on context.

Dr. Joiner: Two drivers of suicide are a feeling of hopeless alienation (“thwarted belongingness”) and the sense of being an intractable burden to others (“perceived burdensomeness”). The desire to die derives from the coincidence of these two feelings, each of which has both self-referential and interpersonal components. What leads from the desire to die (passive suicidal ideation), to the desire to kill one’s self (suicidal intent), to the ability to do so, is a third element: fearlessness about death. This is a simplified version of the Interpersonal Theory of Suicide.

Comparing how much time people invest in social media, and how happy they are with the amount of that investment, reveals something striking. Except for Facetime, among users of 16 popular applications, those who use them the most are saddest about how much time they spend doing so. In the case of Reddit, Facebook, Snapchat, Twitter, YouTube, WhatsApp, and Messenger, the unhappiest users average more than twice as much daily time on those apps as the happiest ones.

The US Department of Defense is the world’s largest funder of suicide prevention research. The Military Suicide Research Consortium funded the Military Continuity Project (originally known as the Caring Text Study), which was directed by Kate Comtois, PhD, MPH. That study compared two cohorts of distressed soldiers and Marines. One group received robust, but conventional, treatment. The other received the same treatment, plus an occasional text message expressing care and support. Over the course of three months, that single, simple intervention reduced subsequent suicide attempts by 45%.

Data from 2009 to 2015, on 500,000 adolescents, shows an association (more pronounced for girls than boys) between daily hours of device use and the presence of at least one suicide risk factor. The relationship is not linear, however. Those doing best are moderate users. Extreme immersion in digital life may come with the greatest risks, but digital isolation also may be associated with mental health hazards.

Questions and Answers

Audience Question: (Vicky Rideout, MA) How big were the positive and negative mood changes due to either Facebook or face-to-face interactions?

Dr. Kross: Small to medium. In layman’s terms, they are larger than the effect size of the link between asbestos and cancer, and other kinds or risk factors for serious physical disease.

Audience Question: How can natural language processing and image analysis be used to understand how people are presenting themselves online, and their suicide ideation? How might gender identity and racial identity influence online self-presentation (and its association with suicide ideation)?

Dr. Kross: All this is rife with complexity. Studies have tried to draw inferences about people’s feelings by counting the emotional words in Facebook responses. You can’t just count words and come up with something that’s psychologically meaningful. With real-time data on subjects’ emotional states as they post, and synchronized data on other people’s assessments of those posters’ emotions, it’s possible to establish correlations between self-reports and third-party observations. Whether or not this might lead to the creation of more sophisticated algorithms – more capable of predicting emotional states – is an open question. This has implications for other psychological constructs, too, including identity.

Audience Question: Looking at the association between electronic device use and suicide risks, is there an analogy to alcohol’s dual character as a coping mechanism and regulator? That is, could media use be regulating those emotions rather than just generating them? This relates to recent research findings that depressed mothers are more likely (than their non-depressed peers) to provide educational media to their children, and gave their children more time with tablets containing educational content.

Dr. Joiner: Substituting something like alcohol for Internet use in the chart seems likely to produce the same U-shaped curve. This speaks to the broadly applicable concept of moderation.

Dr. Kross: The researched presented here statistically accounted for those factors as much as possible, but they matter.

Audience Question: Thinking about attention restorative therapy, could we teach people to deal with a negative experience on Facebook by taking a walk outside to reset their brain?

Dr. Kross: That’s a great idea, and a viable way to combat rumination. There are a lot of tools for dampening rumination and helping people to reflect on their lives more adaptively. Connecting with nature is one of them. Comparing one’s own life to others’ is driving the pernicious effects of passive Facebook use. Knowing that opens the door to think about this and other potential interventions.

Session Materials