Jordan Hale: Ryan, long week — you ready for this one, because I have been sitting with something that I genuinely cannot shake.
Ryan Castillo: Hand it over.
Jordan Hale: Wait, no, picture this: you sign up for a two-year clinical trial, 220 people, you are eating less food every single day for two years, and at the end they show you two results from the same blood draw on the same day. One says you aged slower. One says nothing happened at all.
Ryan Castillo: That's CALERIE — and the clocks that diverged were DunedinPACE versus GrimAge and the Horvath clock, right?
Jordan Hale: Exactly that. Same people, same intervention, same timeframe — DunedinPACE says yes, you slowed your aging, and GrimAge says, I mean, basically nothing. And that is not a fluke in one reading. That is the measurement tool contradicting itself internally.
Ryan Castillo: Which exposes something bigger. The field is trying to answer two completely different questions — are you slowing aging, or are you reversing it — and it's using the same toolkit for both. A toolkit that can't even agree with itself on a single trial.
Jordan Hale: And slowing versus reversing, you know, those aren't just two points on the same dial. They are categorically different claims. Slowing means you're accumulating damage less fast. Reversing means you got something back you already lost. That is a completely different standard of proof.
Ryan Castillo: And right now the field doesn't have a validated way to measure either one cleanly. That's the problem.
Jordan Hale: You know what, the cleaner way to see it — picture a car sitting outside in the rain. Slowing aging is like fixing the leak in the roof so rust builds up less quickly. Reversing aging is like un-rusting metal that's already corroded. Those are not the same repair job.
Ryan Castillo: And the proof you need for each one is completely different.
Jordan Hale: Completely. Like — for slowing, you can plausibly run a two-year trial, get blood draws, look at DunedinPACE or PhenoAge, and say, the rate of damage accumulation went down. That's what CALERIE did with 220 people over exactly two years. You can see it within a funding cycle.
Ryan Castillo: So what's the proof structure for reversal? Because that's where I think the conversation gets... I mean, it stops being about blood tests.
Jordan Hale: Right — and this is the part that doesn't get said clearly enough. Reversal means you have to show someone actually got a function back they had lost. Not a biomarker trending better. Actual capacity. We're talking grip strength improving. Gait speed increasing. A six-minute walk test where someone who used to shuffle is now walking. And you have to rule out a good day.
Ryan Castillo: Hold on — those functional tests exist. Grip strength, the six-minute walk. Why aren't they standard endpoints in aging trials already?
Jordan Hale: They're not standard yet, which is — I mean, that's the structural gap. A functional performance test watching someone's gait speed over years is expensive, it's slow, it doesn't fit a three-year grant. The biomarker is faster, cheaper, and... it gives you something that looks like an answer even when, you know, GrimAge and Horvath are saying the exact opposite of DunedinPACE on the same cohort.
Ryan Castillo: So the field substituted a convenient proxy for an inconvenient measurement. And reversal claims — parabiosis, reprogramming — they're running on that same proxy framework even though the proof standard they actually require is years of functional observation.
Jordan Hale: And that proxy problem — like, it's not abstract. Picture someone at a longevity clinic in, I don't know, Scottsdale, just got their results. The clinic ran DunedinPACE. It moved. The doctor says, your biological age dropped four years. But if that same person had walked in and the clinic happened to run the Horvath clock instead? Nothing. Same blood. Same day. Completely different conversation.
Ryan Castillo: And that's not hypothetical — that's what CALERIE showed. DunedinPACE and PhenoAge moved significantly. Horvath, Hannum, GrimAge — nothing. Same 220 people.
Jordan Hale: Right — but here's the part I want to sit in for a second, because I think it's weirder than it sounds. These aren't just different instruments measuring the same thing with different precision. They're actually measuring different things. Like, the Horvath clock was built to predict chronological age — how old you look on paper. DunedinPACE is measuring the pace of ongoing decline, the rate you're currently falling apart. So of course they don't agree. They were never designed to.
Ryan Castillo: And nobody told the field to treat them as interchangeable?
Jordan Hale: Apparently not, because — I mean, a researcher can look at CALERIE and honestly, not dishonestly, write the headline 'caloric restriction slowed aging' or 'caloric restriction did not slow aging.' Both are defensible. Both pull from the same trial data. That is a credibility problem that goes beyond one study.
Ryan Castillo: So if a clinic reports your biological age dropped five years — what does that actually mean?
Jordan Hale: It means the clock moved. That is it. No epigenetic clock has been formally validated as a regulatory surrogate endpoint for longevity. Not DunedinPACE, not GrimAge, not PhenoAge — none of them. We don't actually know if a four-year reduction in any of these translates to four more years of life.
Ryan Castillo: That's the gap the whole field is standing on.
Jordan Hale: And — wait, actually this gets worse when you bring in reversal claims, because the validation problem for slowing is at least theoretically solvable. For reversal? The proof timescales exceed what any clinical trial can fund, and that's something we need to get into.
Ryan Castillo: And animal data can't bridge it. Parabiosis works in mice — doesn't get you there.
Jordan Hale: And that's the thing about parabiosis — like, the mouse data is genuinely striking. You surgically join an old mouse to a young one, shared circulation, and the older animal's gene expression actually shifts toward a more youthful pattern, tissue regeneration improves. That is real. That is not hype. But we don't even fully know the mechanism yet, and the jump to humans is... I mean, it's not a gap, it's a canyon.
Ryan Castillo: So what does that mean for the proof timeline? Because even if you solved the mechanism tomorrow—
Jordan Hale: You'd need decades of follow-up. That's the structural ceiling. Most trials run two to five years. Reversal isn't a two-to-five-year question.
Ryan Castillo: And no current trial is funded to run that long. That's not a science failure — that's an institutional one.
Jordan Hale: Picture a 68-year-old woman, three years out from a senolytic cocktail — senolytics clear out the cells that stopped dividing and are leaking inflammatory signals into surrounding tissue. Her grip strength is up. Her gait speed improved. Is that reversal? Or did she start physical therapy last spring? You cannot answer that question in year three. You need year thirteen, and nobody is funding year thirteen.
Ryan Castillo: And Buck Institute, Mayo Clinic — these aren't fringe operations. They're running serious programs. But even they can't fund the observation window the claim actually requires.
Jordan Hale: Which is — wait, that's actually the uncomfortable part. It's not that the science is bad. It's that the proof structure the claim demands is longer than science currently runs. So the field fills that gap with biomarkers. A senolytic trial shows a methylation clock ticked down, and that becomes the headline, even though no clock is validated as a surrogate for functional restoration.
Ryan Castillo: So the biomarker isn't evidence of reversal. It's a placeholder where evidence can't exist yet.
Jordan Hale: And that placeholder is getting sold to patients right now. That's the stakes.
Ryan Castillo: The question is a simple one. When you see a headline that says 'we reversed aging' — the thing to ask isn't did the biomarker move. It's did someone who couldn't do something... do it again. Walk faster. Lift more. That's a different question entirely.
Jordan Hale: And it takes longer to answer. Like — grip strength, gait speed, a six-minute walk test, those are real. Those are functional performance tests that actually measure capacity, not a methylation pattern on a clock that GrimAge and Horvath couldn't agree on. But until those are standard endpoints sitting alongside DunedinPACE in every trial, you know, the field is running two separate languages and calling it one conversation. And the headlines will keep outrunning the evidence.
Ryan Castillo: Which brings it back to those 220 people. Two years of eating less, same blood draw, two completely different answers depending on which clock the lab ran. They deserved a better measurement than that.
Jordan Hale: They did. And I think — I mean, that's where I actually land with all of this. The science isn't fraudulent. The gap is structural. But the people inside these trials, and inside these clinics, are real. They deserve the harder question.
Ryan Castillo: Good place to stop.
