Young women who engage in non-suicidal self-injury demonstrate significantly different brain activity when receiving positive and negative social media feedback compared to healthy peers, with the severity of the brain differences mirroring the severity of their condition. This new research was published in Translational Psychiatry.
Self-harm without suicidal intent (known as non-suicidal self-injury, or NSSI) is surprisingly common among young people, affecting an estimated one in five adolescents globally, with higher rates in females. It involves deliberately hurting oneself, such as through cutting or burning, without the intention of ending one’s life. Social media exposure has been associated with a greater risk of self-harm among young people, particularly girls. Until now, the biological reasons for this vulnerability were not well understood.
Social media platforms are specifically designed to trigger reward responses: receiving a like or a positive comment activates the same brain regions involved in processing monetary rewards. Thus, researchers were particularly interested in whether the brain’s reward system—the network of structures that processes pleasure and reinforces behavior—might be altered in young women with NSSI during social media interactions.
Led by Stella Nicolaou of the University of Barcelona, the team recruited 91 young women aged 18 to 30, all of whom had active Instagram accounts. After excluding a few participants due to poor scan data or excessive head movement, the final analysis included 88 participants divided into three groups: a clinical group of 29 women diagnosed with both NSSI and borderline personality disorder, a subclinical group of 27 women who engaged in NSSI but had no other psychiatric diagnoses, and a healthy control group of 32 women with no history of self-harm.
Before the study, researchers followed the participants’ Instagram accounts and selected 15 of their personal photos to use as stimuli. Participants were told other volunteers would be rating their photos, and during the brain scan, they received comments—some positive, some negative—that they believed were genuine. All participants underwent brain imaging using functional MRI while completing this task, which simulated real-life Instagram interactions.
The results revealed a clear pattern linked to severity. The clinical group showed significantly dulled brain responses in key reward regions—including the nucleus accumbens, the caudate, and the medial frontal cortex—when receiving positive versus negative feedback. Strikingly, receiving negative comments actually triggered heightened activity in these same reward regions, suggesting that negative social feedback may feel more engaging for women with more severe self-harm histories.
The brain responses of the subclinical group fell in between those of the healthy controls and the clinical group, suggesting what the researchers describe as a “continuum of severity” mapped onto the reward system. These women responded to positive comments similarly to healthy controls but reacted to negative comments more like the clinical group—showing a selective vulnerability to negative online feedback. Behaviorally, the subclinical group also rated negative comments as significantly more unpleasant than controls did.
Importantly, all three groups reported similar overall levels of Instagram use and addiction, meaning the brain differences cannot simply be explained by how much time the participants spent on social media. However, the researchers discovered a crucial link: in both of the groups that engaged in NSSI, lower brain activity in the reward center was directly correlated with higher scores on the Instagram addiction test. This connection was entirely absent in the healthy control group, suggesting that problematic social media use in those with NSSI is tied to altered neural processing.
As the authors write: “These findings reflect a continuum of severity mapped on the reward system, highlighting potential intervention targets and emphasizing the need to address social media interactions in NSSI treatment.”
Several important limitations should be noted. For instance, the negative comments used in the experiment were intentionally mild due to ethical constraints, so the brain responses observed may underestimate what happens during real-world online cyberbullying. Furthermore, because the study focused exclusively on women and utilized Instagram, the results may not necessarily generalize to men or to users of other platforms, like TikTok.
The study, “Reward-related neural activation during social media exposure in young women with non-suicidal self-injury: evidence for a continuum of severity in the reward network,” was authored by Stella Nicolaou, Anna Julià, Daniela Otero, Carlos Schmidt, Juan Carlos Pascual, Joaquim Soler, Josep Marco-Pallarés, and Daniel Vega.