How social media may help prevent self-harm among teens victimized by parental abuses

Once again, social-media critic Jean Twenge raises an important issue in the teen mental health discussion – then engages in speculations that get things backwards.

Psychologist Jean Twenge, a leader in the movement claiming social media seriously harms young people. offers a popular speculation as to why a growing number of teenagers, particularly girls age 10-14, are harming themselves: “Because social media provides an endless way for other kids to be cruel, (girls) can never achieve the perfect bodies they see on Instagram, they are constantly judged for their appearance in the endless selfies they are compelled to post, unknown adults can sexualize them, they are continually stressed about how many likes they’re going to get, and some social media accounts glorify (and even instruct about) self-harm.”

Twenge’s claim that social media use causes or promotes teen girls’ self-harm relies not on science, but on “correlation equals causation” logic (social media use and self-harm both rose during the 2010s) along with demeaning stereotypes of teenaged girls as fragile and shallow – their distressed fraction can’t possibly be concerned about anything important; just prettiness, popularity, and snarkiness.

So, it comes as a surprise – shock, actually – that those who claim social media is driving teens’ troubles universally ignore the biggest factors in girls’ self harm: parents’ and household adults’ bullying, violence, and troubled behaviors.

Numbers from the Centers for Disease Control’s latest, massive survey of 20,000 teenagers show medical treatments for self-inflicted (suicidal) injuries among girls under age 16 are associated with the following factors:

   Source: CDC, 2024.

That is, girls who use social media more are less likely to self-harm. This was also suggested in the CDC’s 2021 survey, which, using a different measure, found that girls under age 16 who never used screens were much more likely to attempt suicide and injure themselves (9.0% and 8.1%, respectively) than those who used screens 3 or more hours a day (6.1% and 2.6%), with those using screens 1-2 hours per day the least at risk.

Not only are younger teen girls who use social media less likely to harm themselves than those who don’t, the best information in the 2023 CDC survey (which finally asked the right questions), as well as the 2021 survey and other analyses, suggests that social media may well help depressed teens to deal with family issues and avoid sinking into more drastic measures, like suicide attempts and self-harm.

In that regard, the numbers in both the 2023 and 2021 CDC surveys are compelling. For frequently depressed girls under age 16, here are the 2023 proportions who self-harm or attempt suicide by social media use:

— Never use social media:  53% attempt suicide, 23% harm themselves

— Occasionally use social media: 43% attempt suicide, 14% harm themselves

— Use social media daily or less: 37% attempt suicide, 13% harm themselves

— Use social media many times a day: 34% attempt suicide, 9% harm themselves

The above figures require clarification, starting with a pattern that at first looks like it challenges my claim that social media use deters girls’ self-destructive acts. Girls who never use social media are the least likely to be frequently depressed (29%), compared to 35% for girls who occasionally use social media, 35% for those who use social media once a day, and 43% for those who use social media many times a day.

Looking just at the numbers in the previous paragraph and going no further, Twenge and other social-media-blamers declare that proves social media use makes girls more depressed.

Yet, digging deeper, we can see from the table above that the pattern largely reverses for suicide attempts and completely reverses for self-harm. Girls who never or occasionally use social media (despite reporting less depression) are the most likely to attempt suicide and harm themselves; girls who use social media daily or more (despite reporting more depression), the least likely to be suicidal or self-harming.

The most logical resolution of this seemingly contradictory pattern is that Twenge and others are misinterpreting the direction of a reverse correlation: social media doesn’t drive girls’ depression; rather, bullied and depressed girls use social media more. Here are the proportions of girls under age 16 reporting frequent daily social media use by level of parents’/adults bullying (emotional abuse):

—  Never/rarely abused: 38% of girls use social media several times per day

— Sometimes abused: 45%

— Frequently abused: 51%

Encouragingly, depressed girls who use social media daily or more are less likely to engage in rash self-destructive acts. This is consistent with Pew Research’s 2022 finding that two-thirds of teens report using social media to connect with “people who can support them through tough times.”

Twenge cites only one study, from Norway, that finds the associations of exposure to violence and bullying are higher than for social media use in teens’ self-harm. When co-founding variables are included, the association between heavy social media use and self harm falls to barely significant levels (OR = 1.39, Cohen’s d = 0.18, less than small). Unfortunately, that study does not include abuses and troubled behaviors by parents/adults and thus is incomplete regarding key factors, as its authors acknowledge.

Coincidence (mere correlation) is easy to show, which is why so many spurious popular claims dominate media banter. Causality is difficult. More refined questions will further help pin down causality and how social media can be of more help in preventing harm and enhancing physical life.

Note: my analyses have tracked girls, since girls (especially those 15 and younger) are being relentlessly attacked and demeaned by those proposing draconian restrictions on teens’ use of smartphones and social media. Non-binary teens also deserve a lot more analysis. Boys seem to occupy a different universe and will be discussed more in later posts.