A quick note before we start: this is a guide to reading science news, not medical advice. Nothing here tells you what to take, stop, or try. For decisions about your own health, talk to a clinician and look at the primary studies yourself.
You see the headline: "Scientists reverse aging in landmark study." Your pulse ticks up. Then some quieter voice asks: is this real, or is it the same breathless story you fell for last year?
Both feelings are reasonable. Real biotech progress is happening, and most of what reaches your feed is compressed, sped up, or sold. The good news is that you don't need a PhD to tell the two apart. You need a short, repeatable reading process. Here it is.
Start At The End: What Did The Study Actually Measure?
Headlines describe conclusions. Studies measure specific things. The gap between those two is where most hype lives.
So work backward. Before you react to "reverses aging," ask: what did the researchers actually measure? Aging is not a single number. A study might have measured a marker in cells, lifespan in a small group of animals, grip strength, a blood test, or a scan. Any of those can be interesting. None of them is "aging, reversed."
A fast way to find the real claim: skip the headline and the opening paragraph, and look for the study's stated outcome. In a paper this is usually called the "primary endpoint" or "primary outcome" — the one specific result the study was designed to detect. If the article never names what was measured, that absence is itself a signal.
Translate the finding into a plain sentence:
"In this study, [who or what] showed [specific measured change] over [time period]."
If you can't fill that sentence in from the article, the article hasn't told you what happened. It has told you how to feel about it.
Mice, Cells, Or Humans? Why The Model Changes Everything
The single most useful question you can ask about a biotech story: what was this done in?
There's a rough ladder of how much a result should move you:
- Cells in a dish (in vitro): earliest signal. Useful for scientists, weak evidence for you. A compound that does something to cells in a lab often does nothing useful in a living body.
- Animals (in vivo), usually mice: a real step up, and the source of a huge share of exciting headlines. But mice are not tiny humans.
- Humans: the only model that directly tells you about people, and even then only the specific people who were studied.
The mouse-to-human gap is the one to internalize, because so many "breakthrough" stories are mouse studies. Most therapies that work in animals do not go on to succeed in humans. Estimates of that drop-off vary by field and by how you count, but the attrition from promising animal results to approved human treatments is large. One 2024 systematic analysis found that only about 5% of therapies tested in animals go on to win regulatory approval for people (PLOS Biology, 2024) — and even that varies widely by disease area. The point isn't that mouse work is worthless. It's foundational. The point is that "cured in mice" is a beginning, not an ending, and the road from there to a person like you is long and mostly ends in failure.
So when a headline omits the species, go find it. If the answer is "mice" or "in a dish," you can stay curious without rearranging your hopes.
What Phase, How Many People, And For How Long
If the study was in humans, three numbers tell you most of what you need.
Phase. Human drug trials run in stages, and the stage tells you what question was being asked. Roughly: early phases check basic safety in small groups (the FDA describes Phase 1 as typically 20–80 people); middle phases start to look at whether the thing works (Phase 2, up to a few hundred); later phases test it in much larger groups against a real comparison (Phase 3, often hundreds to thousands) (FDA: Step 3, Clinical Research). A dramatic result from an early, small trial is far more fragile than the same result from a large, late-stage one. This is worth understanding in its own right — see what the study's trial phase tells you for the full breakdown of each phase.
Sample size. How many people? A result in 12 participants and a result in 12,000 are not the same kind of evidence, even if the percentages look similar. Small studies swing wildly by chance. They're how science starts, not how it settles.
Duration. For how long? A marker that improved over six weeks tells you little about whether people live longer, healthier lives. Short studies measure short things. Claims about aging, prevention, or survival need long follow-up, and short trials can only hint at them.
None of these numbers has to be huge for a study to be worth reporting. But a story that hides all three is a story betting you won't ask.
Preprint Vs. Peer-Reviewed Vs. Press Release
Where a result was published tells you how much scrutiny it has survived.
- Press release: written by the institution or company that did the work, to promote it. Often accurate, always motivated. Treat it as the beginning of your reading, never the end.
- Preprint: a full paper the authors have posted publicly before peer review. Preprints are genuinely valuable — they make science faster and more open — but the key caveat has to be spelled out: a preprint has not yet been peer reviewed. No independent experts have vetted it. Some preprints are excellent; some contain errors that review would catch; a portion never get published in final form at all. Read them as "here is what one team is claiming," not "here is what the field has confirmed."
- Peer-reviewed paper: other researchers in the field checked the methods and reasoning before publication. This is not a guarantee of truth — peer review misses things, and papers get corrected or retracted — but it's a meaningfully higher bar.
A quick tell: if the article links only to a company announcement and names no journal or preprint, you're reading marketing about science, not science.
Relative Vs. Absolute: The Percentage Trick
This one is small and it will save you again and again.
"Cuts risk by 50%" sounds enormous. But 50% of what? If a condition affects 2 in 1,000 people and a treatment brings that to 1 in 1,000, that's a 50% relative reduction — and a 0.1 percentage point absolute reduction. Both numbers are true. Only one of them is honest about what changed for a typical person.
Headlines almost always quote the relative number, because it's bigger. When you see a dramatic percentage, ask: percentage of what baseline? If the article doesn't give you the absolute numbers — how many people out of how many, before and after — you can't judge how much the finding actually matters.
A Worked Example
Here's a made-up headline, clearly hypothetical, invented only to show the method:
"New Compound Reverses Aging, Extends Lifespan 30% — Scientists Hail Breakthrough"
Run the process:
- What was measured? The article, three paragraphs down, says median lifespan increased in a group of mice given the compound. So the real finding is about lifespan in mice, not "aging reversed" in anyone.
- What model? Mice. Immediately, this is early. Remember the attrition on the road from mouse to human.
- Phase, size, duration? No human phase — it never reached humans. The article doesn't state how many mice. No comparison to an approved treatment.
- Where published? A preprint linked at the bottom, plus a university press release. Not yet peer reviewed.
- Relative or absolute? "30%" is a relative extension of median lifespan in that mouse group. Interesting to researchers; not a promise about human years.
Decoded, the honest version reads: "In a not-yet-peer-reviewed mouse study, a compound was associated with longer median lifespan; whether it does anything in humans is unknown." Still genuinely worth being curious about. Just not worth rearranging your life over. That gap — between the two sentences — is the whole skill.
Red Flags That A Headline Is Running Ahead Of The Science
You'll start to spot these fast:
- Absolute language — "cures," "reverses," "proves," "will end." Careful science almost never talks this way.
- No species named, when the study was in animals or cells.
- No sample size or duration, especially for human claims.
- A percentage with no baseline (the relative-risk trick above).
- Only a press release cited, with no journal or preprint.
- A product you can buy attached to a "breakthrough," where the story functions as a sales page.
- "Scientists say" with no scientist named or paper linked.
Seeing these doesn't mean the underlying work is bad. Often the research is careful and the coverage got ahead of it. But each red flag is a reason to slow down and read for what was measured.
In Plain English
What this is: a short, repeatable way to read a sensational biotech headline down to the actual study underneath it — what was measured, in what species, at what phase, with how many people, and whether anyone has checked the work yet.
Who it helps: anyone who wants to stay hopeful about biotech without getting whipsawed by hype, and who's tired of feeling either fooled or cynical.
When to use it: any time a health or longevity headline makes your pulse jump. Run the questions before you share it, believe it, or worry about it.
What to do next: find the primary source. Read past the headline to the endpoint. Note the species and the numbers. If you can't answer "what did they actually measure?", you don't yet have a story — you have a feeling someone wanted you to have.
This decoder pairs with the grounded frame for reading biotech news, which is the mindset behind the method: progress is real, timelines are long, and no one honest is promising a cure. The same discipline is worth applying to longevity claims specifically, where the marketing runs especially hot.
Grab The Review Checklist
Want the questions in one place? Save the checklist version of this decoder — the endpoint, species, phase, sample, source, and percentage questions — and keep it handy for the next headline that spikes your pulse. You can also browse the Bio, Health & Longevity hub for more grounded reading, or see dated, honestly framed entries in the Directory when you want the substance without the spin.
Stay curious. Read for what was measured. Hope is a lot more durable when it's built on what the study actually says.