At first, mental-health videos on TikTok made Amy Russell feel understood. The 35-year-old saw creators joking about ADHD and sharing tips that resembled her own struggles. After two years of watching such clips, she sought a medical assessment. The diagnosis — and subsequent treatment — changed her life for the better. TikTok, she says, played a role in that transformation.
But now she can’t get the mental-health content to go away. The more she scrolls, the stranger the videos become, drifting into unreliable territory. Her efforts to avoid the genre — skipping videos, refusing to engage — seem powerless. “You just keep finding more tunnels and it gets harder to find your way out,” she said.
Her experience tracks with an extensive Washington Post analysis of viewing histories from nearly 900 U.S. TikTok users. The study found that mental-health content is “stickier” than videos about politics, cats, sports or even Taylor Swift. Watching one mental-health video requires an average of 2.2 skips to counteract — far more than most other categories. In other words, once the algorithm decides you’re interested, it clings.
TikTok uses simple signals: Watch, or skip. Over time, those milliseconds of behavior shape entire feeds. While only about 3 percent of a typical user’s feed contains mental-health content, a few watched videos can quickly tilt the balance. According to the Post’s analysis, two watched clips about ADHD can result in more than ten additional mental-health videos appearing in a user’s next 500 posts.
TikTok disputes the methodology, calling it incomplete. But the platform’s patterns are striking — and they reflect a broader shift in how Americans learn about mental health online. With shortages of therapists, high costs of care and persistent stigma, millions of users turn to social media for self-diagnosis and guidance. Mental-health tags on TikTok are as common as sports tags, according to analytics cited by the Post.
For Russell, the numbers feel all too real. During the period the Post studied her feed, roughly one in every eleven videos she saw was tied to mental health — far above average. She said the longer she watched, the more she encountered clips from non-professionals that seemed designed for engagement, not accuracy.
Experts say the pattern is familiar. University of British Columbia psychiatrist Anthony Yeung found that about half of widely viewed ADHD videos on TikTok were misleading. The algorithm, he said, intensifies the issue: “It becomes this very vicious feedback loop of misinformation.”
Clinicians across the United States say they’re seeing the effects. Therapists report Gen Z patients arriving with self-diagnoses shaped by TikTok creators, sometimes complicating treatment. Some videos portray everyday quirks as symptoms of neurological disorders, muddying the line between relatable content and medical criteria.
Creators acknowledge the tension. One popular ADHD account with millions of views, @lifeactuator, said that short-form video simply can’t provide full context. His clips are broad, he said — and not meant to be diagnostic. But viewers often treat them as such.
For some users, the algorithm’s persistence becomes distressing. Utah teenager Ace Bannon said curiosity about autism spiraled into feeds full of traumatic personal stories he didn’t want to see. Others report that the app worsens their mental-health symptoms. Kailey Stephen-Lane, who has OCD, said TikTok bombarded her with videos reinforcing the very fears her therapist was helping her confront.
Researchers say it’s unclear why mental-health topics stick so strongly. TikTok offers little transparency about how its recommendation engine works. Some experts suspect the app’s incentives play a role: Users who watch emotional content may spend more time on the platform or engage more heavily. Or it may be an accidental byproduct of a black-box algorithm optimizing for attention.
TikTok has added tools to help users reset or filter their feeds, including a “not interested” button and options to block certain keywords. But mental health is not among the categories users can manually dial up or down.
Despite everything, Russell doesn’t regret her TikTok-inspired path to diagnosis. She just wishes the algorithm would show more of what she actually wants now.
“I want like 10 to 20 percent cute cat videos — maybe even 30 percent,” she said. “But those disappear really quickly.”
