The strangest fact in this month's health news is a page count. The CDC just released its big every-two-years survey of how American teenagers are doing, and the numbers genuinely improved: fewer teens reporting persistent sadness, fewer considering suicide. The report announcing all this is also less than half the length of the last one: 36 pages, down from 88 in 2023.
Both facts point at the same conclusion, and I want to make it before you scroll away: a national average was never going to tell you how you're doing, and now the country is also measuring less of it. The only continuous, detailed record of your own mental health is the one you keep yourself. Everything else in this piece is me arguing for that sentence.
What actually happened
Every two years the CDC runs the Youth Risk Behavior Survey, which asks a huge sample of high schoolers about everything from seatbelts to sadness. The 2025 edition, released in early October, carries real good news. The share of students reporting persistent feelings of sadness or hopelessness fell to 33%, down from 40% in 2023 and 42% in 2021. Students who seriously considered attempting suicide dropped from 22% in 2021 to 14% in 2025; those who made a suicide plan fell from 18% to 12% over the same stretch. The CDC's deputy director, Dr. Jennifer Shuford, pointed to improved student attitudes toward nutrition, physical activity, and mental health.
That's two cycles in a row of the curve bending the right way. After a decade of "teen mental health crisis" headlines, this is the closest thing to an all-clear flare the data has sent.
And yet. The document delivering this news is visibly thinner than its predecessor: 36 pages against 88, title page and references included. Fewer breakdowns, a narrower look. The country's flagship read on its teenagers got better news and a smaller envelope to put it in, at the same time.
One caveat before anyone pops champagne: 33% reporting persistent sadness is still higher than the 30% recorded back in 2015. We've recovered toward the old baseline. We haven't beaten it.
You were in this dataset
If you're in your twenties, this survey isn't abstract. A few years ago you were plausibly one of the kids filling it in during third period, deciding how honest to be on the sadness question with your teacher circling the room. The line that's now ticking down is the line you were on when it peaked.
The parallel data on young adults shows some of the same curve, with a catch. The Jed Foundation's read of the 2025 national survey data shows about 29% of 18-to-25-year-olds met criteria for any mental illness in the past year, down from 33% in 2021, with serious mental illness down from 17% to about 12%. Better. Genuinely better.
The catch is that young adults still carry the highest rates of mental health conditions of any adult age group. A 2025 study in Health Affairs Scholar found adults 18 to 29 reported the highest prevalence of a self-described mental health crisis in the past year: 15.1%, against 2.6% for people over 60. Across all adults, that same nationally representative study found nearly 1 in 10 experienced a crisis in the past year. And Pew found roughly a third of young adults hold negative views of their own mental health, rating it worse than older adults rate theirs.
Hold those two things side by side. The average is improving. Your age group is still the most distressed one in the building. Both are true, and neither one is a statement about your Tuesday.
This is the thing population data structurally cannot do, and it's not a flaw, it's just what an average is. When a headline says "teen sadness fell to 33%," that's a claim about a crowd of millions. It is fully compatible with your specific October being the worst one you've had. I wrote something similar when the record-low mental health self-rating made headlines: the national number is weather, and you are one house. The forecast improving doesn't dry your basement.
My honest take: celebrate the curve, distrust the shrinkage
I believe the improvement is real. Two consecutive cycles, echoed in a separate national survey of young adults, is not noise. Something is moving the needle — post-pandemic recovery, more open conversation, schools taking this seriously, maybe even the wave of phone restrictions, though the evidence there is muddier than its fans admit.
But a surveillance report shrinking by more than half at the exact moment the news turns good makes me nervous, and I think it should make you a little nervous too. Thinner reports mean fewer subgroup breakdowns, fewer questions, less of the texture that tells you who is still struggling while the average floats up. Averages improve fastest when the people doing fine do even better; the people at the bottom can stay exactly where they are and the headline still reads "progress." The less granular the measurement, the easier it is for someone's bad year to disappear into a good mean.
And here's the part that's less about policy and more about you: even the fattest version of that report was never going to track your life. The YRBS asks you once, anonymously, in high school, and never follows up. Nobody calls you at 24 to ask how the sadness question is going. The longitudinal study of you — the one that could actually tell you whether this year is better than last year, whether the dread is seasonal, whether the thing you keep calling "a weird week" has quietly been four months — has exactly one possible researcher.
There's a telling detail in the crisis study, too. When people hit a genuine mental health crisis, most of them turned to health care providers or informal supports like family and friends, not formal crisis systems. The infrastructure people actually use when it matters is personal and close to home. Your own record-keeping is the most personal infrastructure there is.
Run your own study
So here's what I'd actually do with this news, beyond feeling cautiously good about the kids.
Start tracking yourself, loosely and honestly, in whatever medium you'll actually sustain. Not a mood-score spreadsheet. A running account: what was today actually like, what keeps coming up. The CDC checks in on a cohort every two years; you can check in on yourself every few days, and unlike the CDC, you can follow up. This is roughly why we built Joice: talking out loud for five minutes is a survey you'll actually complete, and over months the transcripts become the one dataset about your inner life that no budget cycle can trim. The sample size is one, which statisticians hate and which is exactly the point.
Then, when the next national headline lands (and it will, in two years, possibly in an even slimmer envelope), you'll be able to do something almost nobody can: compare it against your own records. Did your line bend the way the country's did? If you're part of the 29% while the average celebrates, your journal is where that fact refuses to get rounded away. If you've been feeling behind while objectively doing fine, the record is how you catch the gap between the story and the data.
The government just told your generation's story in 36 pages. You were a data point in the last three editions of it, compressed into a percentage, and you'll never see your own row.
Keep a longer report on yourself. Make it the one document about your mental health that only gets more detailed over time.