The Effects of Social Media on the Brain

Reward circuitry

Why a “Like” Feels Like Winning

The brain's reward circuitry — centered on the ventral striatum, especially the nucleus accumbens (NAcc), together with the ventral tegmental area (VTA) — evolved to reinforce behaviors tied to survival: eating, bonding, mating. It is the same circuitry implicated in gambling and drug reward, and social media engages it directly.

The clearest demonstration comes from a widely cited 2016 study out of UCLA, led by Lauren Sherman with Mirella Dapretto and Patricia Greenfield. The team built a working replica of Instagram and scanned 32 teenagers as they viewed photos — some their own — that appeared to have received varying numbers of likes (the counts were actually assigned by the researchers). Images shown with more likes produced significantly greater activity in reward-related circuitry centered on the nucleus accumbens. Just as notably, the teens were behaviorally more likely to like a photo themselves once it already carried many likes, a conformity effect that held for both neutral images and photos depicting risky behavior such as drinking or smoking. Popularity itself, in other words, became a cue that reshaped both the brain's valuation of an image and adolescents' willingness to endorse it.

Because this particular study tested only adolescents, it cannot on its own tell us how adults would respond in an identical setup — a gap the developmental section below addresses using separate comparative research.

Executive control

The Brakes, and How Reliable They Are

The dorsolateral prefrontal cortex (dlPFC) and anterior cingulate cortex (ACC) support executive functions: planning, inhibiting impulses, weighing long-term consequences against immediate temptation. If reward circuitry is the accelerator, this network is meant to be the brake.

The evidence that heavy social media use wears down this brake is real, but more modest and more indirect than popular narratives often suggest — and it mostly concerns who uses social media heavily, not what a single scrolling session does to any given brain. Individual differences in smartphone and social media use have been linked to steeper delay discounting — a tendency to devalue future rewards in favor of smaller immediate ones — in work including studies by Wilmer and Chein. A related line of research found that reduced volume in the posterior insula predicts steeper delay discounting, which in turn predicts more severe social-media-use symptoms — a mediation chain rather than a direct link. These are trait-level, cross-sectional associations, not demonstrations that browsing itself degrades self-control in the moment, and the causal arrow could just as easily run in the other direction: people who are already more impulsive may simply be drawn to highly engaging feeds.

Larger studies complicate the picture further. One prospective cohort study tracking adolescent brain structure over several years found associations between social media use and cortical thickness in the lateral and medial prefrontal cortex — but those particular associations did not survive statistical correction for multiple comparisons, even though a separate, independent association between general wellbeing and brain structure in the same study did. That is an important, and rarely reported, asterisk on this part of the story.

Threat response

Bracing for Disapproval

The amygdala and insula are central to detecting threat and emotionally salient information, including social threats like exclusion or disapproval. Because social media delivers peer feedback unpredictably and continuously, researchers have asked whether habitual use recalibrates this system over time — and the best evidence says yes, in a more interesting way than “heightened sensitivity” alone suggests.

The most rigorous data come from a three-year longitudinal study by Eva Telzer's lab at the University of North Carolina at Chapel Hill, published in JAMA Pediatrics in 2023. Researchers followed 169 sixth- and seventh-graders from rural North Carolina, scanning them annually while they completed a task measuring brain responses to anticipated social reward and punishment from peers. At the study's outset, adolescents who checked social media habitually — more than fifteen times a day — actually showed lower neural sensitivity in the amygdala, insula, and ventral striatum than infrequent checkers. But the two groups' trajectories diverged sharply over the following three years: habitual checkers showed rising sensitivity in those same regions, plus the dorsolateral prefrontal cortex, while infrequent checkers' sensitivity in all of those regions declined.

By the study's end, habitual checking was associated with a brain that had become progressively more attuned — not less — to the possibility of social reward or rejection.

Telzer lab, JAMA Pediatrics, 2023

Notably, the study's own authors were careful to note that this growing sensitivity isn't necessarily damage; it could also reflect an adaptive recalibration to a genuinely more socially demanding digital environment.

A separate, smaller line of structural MRI research has looked for physical correlates of this pattern. Several voxel-based-morphometry studies — typically with twenty to ninety participants — have found that heavier or more compulsive social media use correlates with reduced gray matter volume in the bilateral amygdala, and in some studies the ventral striatum, subgenual anterior cingulate cortex, orbitofrontal cortex, and posterior insula as well. These structural patterns echo those reported in substance-use and gambling-addiction research. They also rest on small, cross-sectional samples that cannot establish whether heavy use shrinks these regions over time or whether people with smaller amygdalae to begin with are simply more prone to compulsive checking.

Self-reference

The Default Network on Overdrive

Scrolling a curated feed is, cognitively, an extended exercise in imagining other people's lives and measuring your own against them. Both activities recruit heavily overlapping brain systems: the default mode network (DMN), engaged during self-referential thought, and the mentalizing network — sometimes called the “social brain” — engaged when inferring what others think, feel, or value. The two networks share key hubs: the medial prefrontal cortex (mPFC) and the precuneus / posterior cingulate cortex (PCC), along with the temporoparietal junction (TPJ).

Neuroimaging work on Facebook and Instagram use has repeatedly implicated these regions during tasks involving self-presentation, receiving feedback, and evaluating peers' social value — consistent with theoretical models proposing that self-referential thought, mentalizing, and reward processing are the three systems most relevant to why social media is psychologically compelling. The practical upshot is that a feed is not processed as neutral information; it runs through the same machinery the brain uses to build and continuously update its model of the self relative to others. That may be exactly why curated, high-comparison content — a fitness transformation, a vacation, someone else's highlight reel — tends to land with disproportionate weight.

Embodiment

The “Digital Erosion of Bodily Identity” Hypothesis

The newest and most speculative thread in this research concerns something more basic than mood or self-esteem: the brain's underlying sense of which body is “mine.”

A 2026 study led by Maria Sansoni and Giuseppe Riva's team at the Università Cattolica del Sacro Cuore in Milan, published in Computers in Human Behavior, used virtual-reality body-illusion techniques — methods long used in neuroscience to study how the brain integrates internal bodily signals, like heartbeat and posture, with external visual information to construct a stable sense of embodiment. The researchers recruited 95 young adults, averaging 26 years old, who had used Instagram for roughly 7.7 years on average and about an hour a day. Longer Instagram histories were associated with greater susceptibility to a VR illusion in which a stranger's face could be experienced as one's own — a dose-response pattern the researchers had not anticipated. Notably, the effect was specific to the face illusion rather than showing up uniformly across every bodily-illusion measure they tested, which the authors argue makes sense given how central the face is to identity.

The team proposed calling this the “Digital Erosion of Bodily Identity Hypothesis”: the idea that years of viewing, editing, and comparing filtered self-images may gradually shift the brain's balance away from internal bodily signals and toward external image as the dominant source of identity information, potentially loosening rather than sharpening the perceptual boundary between self and other. It's a striking idea, but it rests on one study, in adults rather than adolescents, with a small, cross-sectional, and demographically narrow sample (mostly white, European, and university-educated). The authors themselves are explicit that it does not establish that Instagram use causes any mental health outcome. It is included here because it shows where this research is heading — toward the basic architecture of embodied selfhood — not because it is settled fact.

Development

Adolescents vs. Adults: A Developmental Mismatch

Most of the sharpest effects described above concentrate in youth, and there is a well-established developmental reason why. Four dimensions, compared:

Evidence quality

Reading the Evidence Responsibly

Two things are true at once, and a fair account of this research has to hold both.

First, the convergent picture across independent labs, methods, and platforms is genuinely consistent: social media engages reward, threat, and self-referential systems in ways that parallel other powerful reinforcers, and does so more intensely in the still-maturing adolescent brain. That is a real, replicated pattern, not a single overhyped study.

None of this means the neuroimaging findings are wrong. It means they are best read as an early, credible map of a real phenomenon — not a finished verdict on what social media does to any individual brain.

Conclusion

Taken together, the neuroimaging evidence describes a coherent story: social media hooks into ancient reward and threat circuitry, borrows the brain's machinery for tracking the self relative to others, and does all of this with particular force during a developmental window when the reward and threat systems are already running ahead of the prefrontal brakes. That is a biologically sensible, well-supported account of why social media feels so compelling, especially to teenagers. What it is not, yet, is proof of lasting harm in any specific person — the studies behind it are still small, mostly correlational, and only beginning to be tested at the scale needed to separate real signal from noise.