The most counter-intuitive social media anxiety study 2025 result came not from a brain scanner or a randomised trial, but from a parenting survey: the people most worried about teen feeds are not the teens. Social media anxiety is the cluster of worry, comparison and physiological arousal that builds up around platforms like Instagram, TikTok and Snapchat, and Pew Research Center’s April 2025 report on teens, social media and mental health found that 44% of US parents name social media as the top threat to adolescent mental wellbeing, against only 22% of the teens themselves. The gap is largest for mothers, and largest still for mothers of teen daughters.
That single comparison is one of the most useful data points to come out of 2025, because it reframes who, exactly, is having the bigger feelings. For families that want to navigate this honestly rather than fight about it, our wider anxiety topic hub is a good starting point. This piece sits with the Pew finding and asks: why do mothers worry more than teens, what does that worry do, and how does a family close the gap without anyone losing face?
What the Pew 2025 numbers show
Pew’s 2025 study was a paired parent–teen design, both parties from the same household answered the same questions. That methodology matters, because it strips out the “different populations” excuse: when a parent says “social media is the top threat” and the teen sitting next to them says “actually it’s bullying, or schoolwork, or the future of the planet,” they are not talking past each other. They are sitting in the same kitchen disagreeing.
44% of parents put social media first; 22% of teens did. The gap is more than two-to-one. It is also gendered on both sides: mothers worry more than fathers, and the worry concentrates around daughters. The teens themselves, asked the same question, were more likely than parents to flag schoolwork, bullying or unspecified anxiety as their leading concerns. Both groups acknowledged social media as a problem; the disagreement was over its rank against everything else competing for adolescent attention. That difference of rank is the whole story.
This is not a one-off finding. The 2024 Journal of Adolescent Health systematic review found that across more than thirty studies of social media and adolescent anxiety, effect sizes varied with how exposure was measured, with the strongest correlations clustering around problematic, not total, use. The 2025 Behavioral Sciences meta-analysis of 24 studies and 68 effects reported an aggregate correlation of roughly 0.22 between social media use and adolescent mental health symptoms, moderate, real and well below catastrophic. The evidence supports parental concern; it does not support the rank parents give it.
A second number from the same Pew dataset deserves attention. When parents were asked how often they themselves felt anxious about their teen’s social media use, more than half reported feeling that way at least weekly, and a meaningful minority described daily anxious thoughts. By contrast, when teens were asked how often the platforms made them feel anxious, the weekly figures were lower and the daily figures lower still. The flow of anxiety in the household, in other words, runs partly upstream from teen to parent: the parent’s worry about the feed exceeds what the feed is reliably producing in the teen at any given moment. That asymmetry is the empathy gap in numerical form.
The empathy gap between mothers and teens
The two-to-one gap in worry is what an editor would call an empathy gap: each side, looking at the same screen, sees something the other does not. Mothers see the news stories about teen depression, the clinical caseload reports, the parenting podcasts and the school chats. They are integrating information across every child in the home and across every cautionary tale on their own feed. Teens see DMs, niche communities, group chats with friends they cannot meet in person, edits of their favourite show and, yes, some content that makes them feel worse, but threaded through the same scroll as the content that makes them feel better.
Both views are real. Neither is the whole picture. Mothers are right that the data shows a measurable association between heavy social media use and anxiety symptoms. Teens are right that, on any given evening, the feed is mostly mundane social maintenance rather than a parade of harm. Both can be true simultaneously, and the gap between the two perceptions is the territory where most family conflict about phones lives.
Our piece on the Gen Z gender split covers the related finding that the burden of social media’s effects falls more heavily on adolescent girls than boys. The Pew design hints at why mothers in particular feel it: they are watching daughters who are statistically more vulnerable, even if the daughter herself does not yet experience the platform as the dominant problem in her life.
There is also a temporal element to the gap. Mothers tend to think about the trajectory, where this is going, what could go wrong by the end of secondary school, what a bad relationship with the phone might mean for university, for a first job, for adult mental health. Teens tend to think about today: which group chat is active, what is happening tonight, who has not replied. Both timeframes are legitimate; they just describe different problems. A mother is right that early patterns matter for later years. A teen is right that today’s interactions are the texture of her actual life. When the two timeframes collide in conversation without either party naming the difference, the disagreement looks like values when it is really about clocks.
Why mothers worry more than fathers
Inside the parent half of Pew’s sample, mothers consistently scored higher than fathers on social-media-related concern. Several mechanisms are likely interacting. First, mothers in many households remain the default carrier of emotional triage, the person tracking how each child slept, ate, behaved and felt. That role makes them the household member who notices early shifts in mood and connects them, rightly or wrongly, to the most salient external candidate cause. In 2025, that candidate is the phone.
Second, parenting media is gendered. Mothers’ own feeds, Instagram, Facebook groups, parenting newsletters, push a heavier stream of teen mental health content than fathers’ feeds typically do. The algorithm that mothers fear, in other words, is also showing mothers more reasons to fear it. That is not paranoia; it is platform mechanics. Our piece on the signs your feed is feeding your anxiety walks through the same dynamic from the user side.
Third, mothers are more likely to be in conversation with other mothers about teen wellbeing, peer-to-peer transmission of worry is real and quick. A single story about an adolescent in crisis, shared in a group chat, can spike alarm across a network of dozens of mothers within an hour. The 2024 Journal of Adolescent Health review and the 2025 Behavioral Sciences meta-analysis offer a sober baseline against which to read those stories, but the syntheses do not circulate in mothers’ feeds with anywhere near the same velocity as the cautionary anecdote.
None of this makes maternal worry irrational. It contextualises it. Mothers are responding to a real signal, amplified by an information environment that selects for that signal. Naming the amplification is part of how a family closes the gap.
Does parental worry help or hurt the teen?
This is the question that matters most to the teenager, and the answer is more nuanced than either camp prefers. Parental worry that translates into engaged, informed conversation, co-viewing a feed, asking which accounts make the teen feel worse, agreeing on small experiments together, is associated with better outcomes in the existing parenting literature. Parental worry that translates into surveillance, conflict or unilateral device confiscation is associated with worse outcomes: more secrecy, more alternate accounts, higher baseline arousal, and in clinically anxious teens, a measurable worsening of symptoms.
The 2025 evidence base does not give us a clean randomised trial of parenting style vs. teen anxiety outcomes, that study has not been done. But the broader adolescent literature on autonomy-supportive parenting is consistent: when teens experience parental concern as collaboration, they integrate it; when they experience it as control, they resist it, and the resistance carries physiological cost.
In practice, that means a worried mother has more leverage than she thinks, and is more likely to spend it badly than well. The same anxiety that drives her to act is also driving her to act in the form that is least likely to help. Our piece on the therapist 5-rule protocol covers the small, structured habits that families can adopt instead of the big confrontation.
Worth flagging too: parental anxiety has its own cost on the parent, not only on the teen. Chronic worry about something the parent cannot fully control feeds insomnia, low-grade physiological arousal, and the kind of background dread that erodes a household’s general mood. Mothers describing this in clinic often say something like “I am exhausted by being scared for her all the time.” That is not a small problem. Looking after the maternal anxiety in its own right, through structured information diet, peer support and, where relevant, the parent’s own therapy, is often what frees up the cognitive bandwidth to have the patient, curious conversations the teen actually needs.
How families talk about this without losing each other
The most useful concrete advice from the 2025 evidence is unglamorous. Three things land repeatedly across the parenting research, the adolescent mental health literature and clinicians’ own descriptions of what they recommend.
First, ask the teen what they actually see. Not “are you on TikTok too much”, they will say no, but “show me three accounts you follow, and tell me which one makes you feel best.” The question respects the teen as the expert on her own feed and gives the mother data she did not have. It also reframes the conversation from prosecution to curiosity.
Second, share the evidence honestly. Tell the teen that the 2025 syntheses do find a moderate but real association between heavy use and anxiety, and that problematic use, not screen time per se, is what tracks symptoms most tightly. Teens can hold complexity. Adolescent reasoning is not the limiting factor; adult willingness to acknowledge nuance usually is.
Third, agree on small experiments. A one-week break, a no-phones-at-meals rule, no devices in the bedroom overnight, a weekly check-in. The 2025 detox literature suggests these structured interventions have small but real effects on wellbeing, and crucially, they are joint experiments rather than parental imposition. The mother gets to act on her worry, the teen gets to keep her agency, and the family generates its own data instead of arguing about somebody else’s.
Where a teen is already showing signs of clinically significant anxiety, sleep disturbance, school avoidance, panic episodes, the conversation needs to widen beyond the phone. Our overview of the overdiagnosis debate covers the case for not blaming everything on a single platform, and the wider biggest findings piece ranks the 2025 evidence on what actually moves the dial.
What the evidence supports, and what it doesn’t
The 2025 Pew finding is durable: parents are more worried than teens, mothers most of all. The 2024 Journal of Adolescent Health review and the 2025 Behavioral Sciences meta-analysis support the rough shape of parental concern, a real, moderate association exists between social media exposure and adolescent anxiety symptoms, especially in problematic users and especially in girls.
What the evidence does not support is the implicit assumption inside the worry: that solving social media solves the teen’s anxiety. The effect sizes are not large enough for that. A daughter whose anxiety is rooted in academic pressure, family conflict, identity questions or a clinical predisposition will not be transformed by deleting Instagram. She may improve, modestly, on the dimensions the platform was amplifying. That is worth doing. It is not a complete intervention.
The honest position is that mothers are right to be worried and that the way the worry is currently directed often overshoots its evidence base. The 2025 data is an invitation to recalibrate, to take the concern seriously without pretending the phone is the whole story.
There is also something the evidence does support that parents often miss: protective factors. The same studies that find a moderate harm signal also find that adolescents with strong offline friendships, regular physical activity, predictable sleep schedules and at least one trusted adult to talk to are buffered against the worst of the platform’s effects. None of those protective factors require a confrontation about the phone. They are built quietly, over years, in the texture of family life. Mothers who direct some of their worry-energy toward building those protective factors, instead of toward policing the device, often find both their own anxiety and their teen’s improving on the same timeline.
Practical implications
If you are a parent reading this, the move is not to worry less. It is to convert the worry into shared infrastructure with the teen. Ask, listen, share the evidence as it actually reads, run small joint experiments, and revisit. If you are a teen reading this, perhaps because a parent sent you the link, the move is not to dismiss the worry. It is to engage it with concrete data from inside your own feed, so the worry has somewhere honest to go.
Both moves require a degree of humility that does not come naturally to either side in a charged conversation. But the alternative, phones-as-permanent-battleground, is exhausting and, on the 2025 evidence, ineffective. Families that close the empathy gap do better. That is most of what the data is asking for.
Frequently asked questions
The FAQ above this section is the structured-data version Google reads. The body section ends here. If you want to keep reading in the same direction, the most natural next stops are our biggest 2025 findings ranked, the therapist 5-rule protocol, and the overdiagnosis debate.