The evidence is no longer ambiguous. Study after study, from governments and independent researchers spanning dozens of countries, arrives at the same conclusion: the world’s most popular social media platforms are corroding the mental health of their youngest users. And the companies behind them have done remarkably little to change course.
The latest salvo comes from the Sapien Labs Global Mind Project, a sweeping annual assessment that surveyed nearly 55,000 young adults aged 18 to 24 across 34 countries. The findings, published in early 2025, are blunt. Platforms built around short-form video and image-based content — Instagram, TikTok, and Snapchat chief among them — are associated with significantly worse mental health outcomes than text-based platforms like X (formerly Twitter) or Reddit. Not marginally worse. Substantially worse, according to Digital Trends, which reported on the study’s release.
This isn’t a single alarming data point from a small sample in one Western country. It’s a global pattern that held up across cultures, languages, and economic conditions. The researchers found that young adults who primarily used image- and video-centric platforms scored lower on measures of emotional well-being, social self-image, and adaptability. Those who favored text-based platforms fared better on virtually every metric of psychological health the study tracked.
The distinction matters enormously.
Visual Feeds, Invisible Damage
What separates Instagram and TikTok from their text-heavy counterparts isn’t just format. It’s the fundamental nature of the interaction. Image and video platforms are built to showcase idealized versions of life — filtered faces, curated vacations, performative vulnerability packaged in 60-second clips. The comparison engine runs constantly. You scroll, you see someone who appears happier, thinner, wealthier, more socially connected. You absorb this passively, often without conscious awareness of the psychological toll.
Text-based platforms aren’t immune from toxicity, of course. Anyone who has spent five minutes in certain corners of X or Reddit can attest to that. But the Sapien Labs data suggests something specific about visual content that hits harder at the level of self-perception. The researchers hypothesize that image-centric feeds activate social comparison mechanisms more intensely than text threads or comment sections. You can argue with a tweet. You can’t argue with a photo that makes you feel inadequate — at least not as easily.
Tara Thiagarajan, founder of Sapien Labs and lead author of the study, told Digital Trends that the consistency of the findings across so many countries was striking. The pattern didn’t shift based on whether participants lived in high-income or low-income nations. It didn’t depend on cultural attitudes toward mental health. Visual platforms correlated with worse outcomes everywhere.
That universality is hard to dismiss.
And it arrives at a moment when regulators around the world are already moving aggressively. Australia passed legislation in late 2024 banning children under 16 from social media entirely, a law that takes effect in 2025. The European Union has opened formal investigations into TikTok under the Digital Services Act, scrutinizing the platform’s algorithmic recommendations and their effects on minors. In the United States, the Supreme Court recently heard arguments about state laws in Texas and Florida that seek to regulate how platforms moderate content, with youth safety concerns looming over the proceedings.
But regulation moves slowly. The platforms move fast. And the gap between what the science says and what the companies do continues to widen.
Meta, which owns Instagram, has introduced features like daily time reminders, hidden like counts, and a “Take a Break” nudge for teens. TikTok has implemented default screen time limits for users under 18. These are real features. They are also, by the account of most independent researchers, grossly insufficient. A screen time reminder that a teenager can dismiss with a single tap does not constitute a meaningful safeguard. It’s a fig leaf — one that allows the company to claim it acted while changing almost nothing about the core product.
Jonathan Haidt, the NYU social psychologist whose 2024 book The Anxious Generation became a bestseller and a rallying point for parents and policymakers, has argued that voluntary industry reforms are structurally incapable of solving the problem. The business model depends on engagement. Engagement depends on emotional arousal. Emotional arousal, particularly among adolescents whose prefrontal cortices are still developing, often means anxiety, envy, and compulsive use. You can’t fix that with a toggle in the settings menu.
The Sapien Labs study reinforces Haidt’s thesis with hard data at global scale. But it also adds nuance that previous research sometimes lacked. By comparing platform types rather than treating “social media” as a monolith, the study suggests that not all digital communication is equally harmful. This matters for policy. A blanket ban on social media for minors — like Australia’s — treats Reddit and TikTok as interchangeable threats. The data says they aren’t.
The Industry’s Defensive Playbook Is Wearing Thin
The tech industry’s standard response to mental health research has followed a predictable script for years. First, question the methodology. Point out that correlation isn’t causation. Emphasize that the study didn’t account for pre-existing mental health conditions. Cite internal research (rarely made public) that supposedly shows benefits. Then announce a new feature — a wellness dashboard, a parental control, a content filter — and move on.
Meta has been particularly aggressive in this regard. After internal research leaked in 2021 through whistleblower Frances Haugen — showing that Instagram’s own researchers found the platform made body image issues worse for one in three teen girls — the company disputed the characterization of its findings, rebranded to Meta, and launched a campaign emphasizing parental tools. But the core product didn’t change. The algorithm still rewards content that generates strong emotional reactions. The Explore page still surfaces appearance-focused content to young users who haven’t asked for it.
TikTok’s defense has leaned on its investments in content moderation and its partnerships with mental health organizations. The company points to its restricted mode, family pairing features, and content warnings as evidence of good faith. But the platform’s fundamental design — an endless, algorithmically curated feed of short videos optimized for maximum watch time — remains untouched. The algorithm is the product. Changing it would mean changing everything.
So the status quo holds.
Meanwhile, the numbers keep getting worse. The Sapien Labs study found that the mental health scores of 18-to-24-year-olds have declined globally over the past several years, with the sharpest drops in English-speaking countries where smartphone adoption and social media use among young people are highest. The United States, the United Kingdom, Australia, and Canada all showed particularly pronounced declines in what the researchers call the “Mental Health Quotient” — a composite score measuring mood, outlook, social connections, and cognitive function.
These aren’t abstract metrics. They translate into rising rates of anxiety disorders, depression diagnoses, self-harm, and, in the most extreme cases, suicide among young people. The CDC reported in 2023 that nearly three in five teenage girls in the U.S. experienced persistent feelings of sadness or hopelessness — the highest level recorded in a decade. The U.K.’s National Health Service has seen referrals for child and adolescent mental health services more than double since 2019.
Not all of this can be attributed to social media. The pandemic, economic uncertainty, climate anxiety, and political polarization all play roles. But the temporal correlation between smartphone-based social media adoption and the deterioration of youth mental health has become difficult for even skeptics to wave away. The Sapien Labs data, with its massive sample size and cross-cultural consistency, makes the case harder to ignore than ever.
What comes next is less clear. Legislative efforts remain fragmented. In the U.S., the Kids Online Safety Act passed the Senate in 2024 with overwhelming bipartisan support but stalled in the House. State-level laws have proliferated — Utah, Arkansas, Ohio, and others have passed or proposed restrictions on minors’ social media access — but face legal challenges on First Amendment grounds. The patchwork approach means a teenager in one state may be protected by law while one in a neighboring state is not.
Europe’s approach is more unified but slower to enforce. The Digital Services Act gives regulators significant power to demand transparency about algorithmic systems and to penalize platforms that fail to protect minors. But enforcement actions take years. TikTok, which faces a formal DSA investigation, has contested the proceedings. Meta has cooperated selectively. The gap between regulatory ambition and practical impact remains wide.
A Question the Platforms Won’t Answer
There’s a simple question at the center of this debate that Instagram, TikTok, and Snapchat have never directly answered: If your product is safe for young people, why won’t you share the data proving it?
Meta’s internal research, revealed by Haugen, showed the company knew about Instagram’s harmful effects and chose not to act. TikTok has resisted calls from U.S. legislators to open its recommendation algorithm to independent audit. Snapchat, which the Sapien Labs study also flagged as associated with worse mental health outcomes, has released limited data about its Discover feature’s impact on younger users.
The opacity is telling. Companies that believe their products are safe tend to prove it. Companies that don’t tend to hire lobbyists.
And the lobbying has been extraordinary. According to OpenSecrets, Meta spent over $19 million on federal lobbying in 2024. TikTok’s parent company, ByteDance, spent more than $8 million. These figures don’t include state-level lobbying, trade association spending, or the extensive public relations campaigns both companies have waged to shape the narrative around youth safety.
The Sapien Labs study won’t end this fight. No single study could. But its scale and its specificity — distinguishing between platform types, spanning 34 countries, surveying tens of thousands of young adults — make it one of the strongest pieces of evidence yet assembled. It tells us something we already suspected but can now say with greater confidence: the way Instagram and TikTok are designed is making young people measurably less well. And the companies behind these platforms have the data, the engineers, and the resources to build something better. They’ve simply chosen not to.
The question for regulators, parents, and the platforms themselves is how much more evidence they need before something actually changes. The young adults in this study are already living with the consequences. The next generation is already online.
Time isn’t on anyone’s side.
WebProNews is an iEntry Publication