Researchers tracked 1,522 teenagers for a year using their phones' own sensors. More hours on Snapchat, TikTok, and Instagram lined up with more drinking, vaping, and cannabis use. More hours on YouTube didn't, if anything the opposite.

I've read plenty of "screen time is bad for kids" studies that dump every app into one bucket. This one didn't. It pulled six apps apart, and they don't behave the same way at all. That's the bit that made me read the methods section twice.

What they did

The data comes from the ABCD study (Adolescent Brain Cognitive Development study), a large, nationally diverse US cohort that has been following kids since childhood. This analysis used the first wave with sensor-based phone data: actual measured hours per app, not a teen's guess at how much they used TikTok that week.

  • Baseline sample: 1,522 kids, ages 13 to 16
  • Follow-up one year later: 1,121 of them
  • Apps measured: Snapchat, TikTok, Instagram, YouTube, and two others
  • Substances tracked: alcohol, nicotine/tobacco, and cannabis

The first pass compared app hours and substance use at the same point in time. Then came the harder question: among teens who already reported using something at baseline, did app hours predict use a year later?

What they found

  • Hours on Snapchat, TikTok, and Instagram were positively associated with alcohol, nicotine, and cannabis use at baseline, across multiple measures.
  • YouTube ran the other way. More YouTube time, less substance use.
  • The pattern held up a year later among teens already using at baseline. Not a same-day coincidence.

So this isn't "social media is bad." One of the biggest platforms in a teen's daily rotation showed a protective pattern. Three others didn't.

Why the split makes sense

The authors' read: it's about content and format, not just hours spent staring at a phone. Snapchat and TikTok run on short, disappearing, personal clips - exactly where a party story or a vape trick gets shared casually and vanishes before anyone catches it. YouTube leans long-form: tutorials, gaming, commentary. A different content mix, and maybe a different reason kids open the app in the first place.

The study is honest about its limits, and I'll be too. Part of it is cross-sectional, so a teen prone to risk-taking might simply prefer Snapchat over YouTube, rather than Snapchat causing anything. Self-reported substance use at that age carries noise. But the association survived a full year among kids already using, which makes "we caught a random blip" harder to argue.

What I take from this, as someone who thinks about drinking behavior for a living: exposure adds up quietly. Nobody decides to normalize a drink or a hit after one clip. It's the accumulated drip of "everyone's doing this and it looks fine" that resets what feels normal. For teens now, and for the adults who grew up on the same feeds. Next time a night out feels like it's following someone else's script instead of your own rhythm, that's worth noticing.

Source: Addictive Behaviors, DOI