Why the Age You Start Social Media Matters Most

Parents have been fighting about the wrong thing. Not fighting badly — fighting about the wrong lever entirely. The argument in most households is “does my kid get an account, yes or no,” as if that’s the switch that matters. A study published in JAMA Network Open in late August 2026 just handed us the first hard data saying that’s not the switch. The age a kid first touches social media predicts anxiety and depression more strongly than whether they currently own an account at all.

Let that sit for a second, because it reorganizes the whole debate. Two kids can both have zero accounts right now. If one of them first logged into TikTok at 9 and the other at 14, the data says their mental health risk isn’t the same. Account status is the thing parents can see and control today. Age of first exposure is the thing that already happened and can’t be undone. That’s a much harder conversation, and it’s the one this study forces.

The short version

What the study found	What it means for your family
Researchers surveyed 365 Australian adolescents ages 10-15 and one parent each, in the six weeks before Australia’s under-16 social media ban took effect	This is a real-time snapshot of a generation right before a landmark law changed their access
Average age of first social media use in the sample was 11 years old	Most of these kids started years before any account was “official” or parent-approved
70% of adolescents had used social media, but only about half currently held an account	Ownership and actual use are different things. A kid without an account has still probably been on the apps
Age of first exposure correlated with anxiety and depression symptoms more strongly than current account ownership	The earlier the first login, the higher the symptom scores — regardless of whether an account exists today
Three months after Australia’s ban took effect, roughly 8 in 10 teens were still using social media, just with fewer formal accounts	Banning accounts doesn’t erase use. It mostly just erases the account, which this study says was never the real variable

What the study actually measured

The team, led by Associate Professor Susanne Schweizer of UNSW and the Black Dog Institute, fielded the survey between late October and early December 2025 — timed deliberately to land right before Australia’s Social Media Minimum Age Restrictions kicked in on December 10. They asked each of the 365 adolescents three separate questions: had they ever used social media, how old were they the first time, and did they currently hold an account. Then they measured anxiety with the GAD-7 scale and depression with the PHQ-8, the same instruments doctors actually use in clinics, not a homemade quiz.

Splitting those three questions apart is the whole contribution here. Most public debate — and most parenting decisions — collapse “has used it,” “first used it at what age,” and “owns an account” into one question: does my kid have social media or not. This study shows those are three different facts with three different relationships to a teenager’s mental health. Only one of them turned out to carry real weight.

Why age of first use beats account ownership

Here’s the actual finding, stripped of hedging. Earlier age of first use showed a statistically strong relationship with both anxiety and depression symptoms. Current account ownership showed a weaker relationship with anxiety and, once you account for other factors, wasn’t a reliable predictor of depression at all. I’ve written about what 4+ hours of daily screen time does to anxiety and depression risk — this study adds a piece that changes how you should read that number. It’s not just how much a teenager scrolls right now. It’s how young they were when the scrolling started.

Think about what that means practically. A parent who deletes their 15-year-old’s Instagram app today, believing they’ve fixed the problem, hasn’t touched the variable this study says matters most — the age that kid first opened a social app, which already happened, years ago. Meanwhile a parent who let their 9-year-old sign up “just to watch videos with no posting” has already set the variable that predicts risk, even if that child never technically “owns an account” today.

I know that’s an uncomfortable thing to read if your kid already started young. You can’t rewind a first login. What you can do is treat every year you delay the next kid’s first exposure as protective, and stop assuming that account deletion is the same thing as prevention.

There’s no single magic number backed by a controlled trial — nobody’s run that experiment, and nobody should, ethically. But the JAMA data, combined with what Australia’s under-16 law and the broader research base already show, points to a few concrete guardrails:

Treat “watching without an account” as exposure, not a loophole. 70% of the study’s adolescents had used social media, but only half held accounts — meaning a lot of first exposure happens through a parent’s phone, a shared device, or a friend’s account before any sign-up ever occurs.
The earlier number matters more than the current number. If your kid is 13 and just got their first account, that’s a very different risk profile than a 13-year-old who first logged in at 9, even if both have “an account” today.
Delay beats deletion. Removing an app after years of use doesn’t erase the age-of-first-use variable. Preventing that first exposure in the first place is the only lever that actually moves it.
Match access to the platform’s own design, not just an age gate. Some apps are built around passive video feeds; others are built around DMs with strangers. The mechanism behind the harm matters as much as the age — passive scrolling and active, unsupervised contact carry different risks.
Don’t assume a ban solves what a conversation should. Three months after Australia’s law took effect, most teens were still finding their way back onto these platforms through borrowed logins and workaround accounts. Legislation changes access. It doesn’t change a kid’s relationship to the apps by itself.

The timing here isn’t an accident

This study landed in the middle of the most consequential natural experiment social media policy has ever run. Australia became the first country in the world to bar under-16s from major platforms, and the law took effect December 10, 2025. U.S. states have been watching closely, several drafting their own versions of age minimums and verification requirements. Most of that debate — in Australia and here — has centered on account access: can a 14-year-old have Instagram, yes or no.

This study is the first real evidence that the debate has been aimed at the wrong target. If age of first use is the stronger predictor, then a law that only blocks new accounts for kids who are already fifteen and have been scrolling since age nine is closing a door that’s already been walked through. The early results back that up in an ugly way — eSafety’s own tracking found roughly 8 in 10 Australian teens still using social media three months after the ban, with account ownership dropping but daily use barely moving. The account disappeared. The exposure didn’t.

That’s not a reason to give up on age laws. It’s a reason to stop treating “no account” as the finish line, for lawmakers and for parents both.

What this looks like on a Tuesday

Your 8-year-old asks to watch videos on your phone during a car ride, and you hand it over because it’s easier than the alternative. That’s not nothing. That moment, repeated across a few years, is exactly the kind of early exposure this study is measuring — not the day you finally let them create an “official” account at 13, but the hundred small handoffs before that.

I’m not telling you to panic about a car ride. I’m telling you the age question isn’t really about a birthday and a sign-up form. It’s about how many years a kid’s nervous system spends absorbing a feed built by people whose job is to keep them scrolling, before anyone in the house calls it “social media” out loud.

If you’ve got a younger kid who hasn’t started yet, that’s the leverage you actually have. This is one of the harder conversations in the Health & Fitness section of the field guide, and it doesn’t get easier by waiting for a law to make the decision for you.

The takeaway

The account was never the real variable. The age your kid first logged in was. You can’t move that number for the exposure that already happened — but for whichever kid in your house hasn’t started yet, you still can, and that’s worth more than any ban.