JournalIdentity
Social Media and Body Image: What We Actually Know
The evidence here is genuinely mixed, and the honest version of the story is more useful than the alarming one.
Sian Trombley
18 MAY 2026 · 4 MIN READ
Somewhere between "social media is destroying an entire generation's self-image" and "it's basically fine, people have always compared themselves to each other" sits the actual research — messier, more interesting, and harder to put on a headline.
It's worth starting there, because most conversations about social media and body image skip straight to conclusions. This one is going to move more slowly, because the honesty the topic deserves requires it.
It isn't the mirror that's the problem. It's which mirrors get held up most often.
What we're fairly confident about
The clearest and most consistent finding in this area isn't about time spent on platforms — it's about the type of content. Appearance-focused content, particularly content involving idealised or edited bodies, shows a more reliable association with body dissatisfaction than general social media use does. Scrolling through a feed of friends' holiday photos and scrolling through a feed of fitness influencers and filtered selfies are not the same activity, even if both count identically in a screen time report. This distinction — content over duration — is one of the more robust patterns across a fairly noisy body of research.
There's also reasonably solid evidence around the direction of comparison. Upward comparison — measuring yourself against people who appear to be doing better, looking better, living better — tends to correlate with worse mood and body image than comparisons with peers who feel more attainable. Feeds are structured, whether by design or by algorithmic drift, to surface content that performs well, and content that performs well tends to be aspirational. That's not a neutral backdrop. It's a comparison target that's been selected, in part, for its ability to make you look for a moment longer.
Where it gets genuinely contested
What the evidence says
Effect sizes in this field are a real point of disagreement among serious researchers. Jean Twenge has pointed to correlational data linking the rise of smartphone-era social media with increases in body dissatisfaction and related distress among teenage girls in particular. Amy Orben and Andrew Przybylski's analyses of large-scale datasets have found associations that are statistically real but small on average, and have cautioned against treating population-level averages as if they predict any individual's experience. Both positions are held by credible researchers working with real data. The disagreement is less about whether an effect exists and more about how large it is, and for whom.
What this does not mean is that body image concerns can be waved away as statistically minor. Averages hide enormous variation. A small average effect across a population of millions can still coincide with a serious, specific effect for a smaller group of people who are more vulnerable to it — often people already prone to social comparison, already dealing with body image concerns from other sources, or going through a developmental period (like early adolescence) when identity and appearance feel unusually fused.
The mechanism, not just the outcome
It helps to understand roughly why appearance-focused content lands harder than other content. Social comparison theory, going back to Leon Festinger's original work in the 1950s, holds that people evaluate themselves partly by comparing to others — and that this is a normal, largely unavoidable part of how self-assessment works, not a modern pathology. What's changed is the volume and selection of comparison targets available at any moment. In an ordinary day, a person might once have compared their appearance to a handful of people they actually saw. Now the comparison set can run into the hundreds, most of them presenting a version of themselves optimised — through angle, lighting, editing, or simply careful selection — for exactly the kind of comparison being made.
There's also a passive-versus-active distinction that shows up repeatedly in this research. Passive scrolling — consuming content without interacting — tends to correlate more strongly with negative mood and comparison than active use, like messaging friends or posting your own content. This matches a broader pattern in social media research: it's often not the platform itself but the specific way it's being used that predicts the outcome.
What actually helps
Given the mechanism, the useful interventions aren't about willpower or overall reduction so much as about editing the specific inputs.
Try this.
- Do a quick audit of the accounts that reliably leave you feeling worse about your body after viewing them, and unfollow or mute them — not as a moral stance, just as basic input management.
- Notice the difference between scrolling passively and actually engaging (commenting, messaging, posting); if a session is purely passive and appearance-heavy, that's the pattern worth interrupting first.
- Follow at least a few accounts that show bodies and lives in an unfiltered, ordinary way — not as a corrective fantasy, just to widen the comparison set back toward reality.
None of this requires deciding social media is good or bad. It requires treating your feed the way you'd treat a diet: not by refusing food, but by noticing which foods reliably make you feel worse, and eating less of those.
8 minutes
Comparison Reset
A short written exercise for the days the feed makes your life look small.
Open the tool →A thought to keep.
Comparison isn't a modern flaw in human psychology — it's an old one, running on new and much faster machinery. The old advice still works. You just have to apply it to a much bigger, much better-lit room than the one it was designed for.
