Mark Delaney: Okay, rough week but I'm glad we're doing this one because — man, this paper landed and I've been kind of low-key bothered by it ever since.
Juniper Vale: Bothered how — like troubled, or bothered like you can't stop thinking about it?
Mark Delaney: Second one. So Professor Itshak Melzer at Ben-Gurion University of the Negev runs this longitudinal study — 120 older adults, they track them for up to seventeen years — and the thing that predicted survival wasn't age, wasn't how many diseases you had. It was step initiation time. How many milliseconds it took you to start a step.
Juniper Vale: And by how much are we talking?
Mark Delaney: Every hundred milliseconds slower — a tenth of a second — hazard ratio of about 1.28. So 28% higher mortality risk. Per tenth of a second. That's the number that's been sitting in my head.
Juniper Vale: Okay that I did not fully appreciate until you put it that way. That's — yeah, 28% is not nothing.
Mark Delaney: And it cleared out chronological age and chronic illness counts as standalone predictors. Published in the journal Gerontology. So what we're really trying to figure out today is — is this a better window into biological aging than the tools medicine actually uses right now?
Juniper Vale: Which is a genuinely disruptive idea if it holds up.
Mark Delaney: Right — but the part I want to poke at is whether it holds up at scale, because 120 people over seventeen years is detailed work, not a massive sample.
Juniper Vale: But here's what I want to slow down on, because I think there's a piece that makes the whole thing click — it wasn't just step speed on its own. The test that actually sharpened the signal was the dual-task condition. Stepping while your brain is doing something else at the same time.
Mark Delaney: Wait, like — literally stepping and thinking simultaneously?
Juniper Vale: Exactly that. Think of it like driving — anyone can steer on an empty road. Put traffic on it, add a ringing phone, and suddenly you see who's actually a good driver. The BGU test works the same way. You have to start a step while you're answering a question or counting. That layering is what makes it predictive in a way that just timing someone's walk doesn't quite get you.
Mark Delaney: Huh. So the cognitive load is, uh — it's not a distraction from the test, it basically is the test.
Juniper Vale: Right, because real life doesn't give you clean conditions. You're walking to the kitchen while trying to remember something. The dual-task condition stresses motor and cognitive systems at the same time — and that's where frailty and neurological integrity actually show up. The latency they measured, step initiation time, is specifically the gap between when you get the cue and when your body actually starts moving. That gap, under cognitive load, is the variable.
Mark Delaney: And that gap — a tenth of a second of it — maps to 2.9 months of expected survival over a decade? I mean, that's weirdly precise for something so small.
Juniper Vale: It is, and that precision is actually what makes me take it seriously despite the sample size. I want to be honest though — this is a marker. We don't know yet if training someone to step faster actually extends life, or if fast stepping is just reflecting something deeper, some neurological integrity that's doing the real work.
Mark Delaney: Yeah, no — that's the part I keep bumping into too. Like, measuring it is one thing. Knowing what to do about it is a whole other problem.
Juniper Vale: And that gap between measuring and doing — that's actually where the marker-versus-cause question bites hardest. Because picture this: a 78-year-old, annual checkup, doctor knows she has mild hypertension, a little arthritis, tells her she's doing fine for her age. But when she's distracted — say she's standing up to answer the door while she's mid-thought — there's a half-second lag before her body moves. Her chart says fine. That lag is saying something completely different.
Mark Delaney: And the doctor has no idea, because nobody timed her step.
Juniper Vale: Right — but here's what I want to be careful about. Even if someone did time her step, we don't know if fixing the lag would change her outcome. It might just be reflecting something deeper — biological age, neurological reserve, cellular stuff — that a movement drill can't actually touch.
Mark Delaney: Okay but — uh, wait — so where does the hot take actually hold? Because I think there's a real kernel here that I want to give credit to.
Juniper Vale: It holds in the biomarker claim. Gait speed, sit-to-stand tests, handgrip strength, step initiation time — they're all converging on the same thing. Physical performance battery as a read on biological age, not just chronological age. That's the kernel. These tests are picking up something called frailty — it's a clinical syndrome, decreased physiological reserve — and they're catching it before it shows up anywhere else on a chart.
Mark Delaney: So the hot take wins on the detection side. It's the intervention side that's, uh — that's still open.
Juniper Vale: Exactly. And that's not a small concession — the BGU findings cover 10 to 17 years on 120 people. That longitudinal depth is genuinely unusual. The sample is small, yes, but you don't get that kind of temporal signal from a short study.
Mark Delaney: Huh, no — I'll give it that. What I keep wondering though is whether any of this ever gets in front of the actual 78-year-old at that checkup. Because the evidence is one thing, and then there's what actually happens in a clinic room, and — honestly, that part might be its own whole problem.
Juniper Vale: This is structural, and we should get into it — because the gap between what the evidence says and what routine practice does is not an accident.
Mark Delaney: And that gap — it's not like doctors don't care, it's that the infrastructure to actually do anything with a step-speed number just isn't there. Like, who runs the test? Who interprets it?
Juniper Vale: That's exactly the structural piece. The APTA Academy of Geriatric Physical Therapy has been pushing annual mobility screening as a primary prevention standard — not just rehab after a fall, but catch it before anything goes wrong. Preclinical. That's the model.
Mark Delaney: Wait, they've been pushing for this for a while?
Juniper Vale: In 2023, Michelle Lusardi gave a call to action at the APTA Carole B. Lewis Lecture — named lecture, formal venue — and that directly prompted the formation of an Annual Mobility Assessment Task Force. That's not a fringe idea, that's the professional body organizing around it.
Mark Delaney: Okay but — uh, task force is one thing, and then there's what happens in an actual clinic. Like, the BGU study had Soroka University Medical Center, University of Minnesota behind it. Real institutions. Published in Gerontology. And the clinical adoption still hasn't followed. I can't get past that.
Juniper Vale: No, you're right to flag that. Because replacing age and disease counts with timed physical tests — you need trained assessors, you need the equipment, and honestly you need patients who'll cooperate with something that feels unfamiliar. Those aren't small asks in a fifteen-minute appointment.
Mark Delaney: And if your doctor did flag a delay tomorrow — say your step initiation's slow under dual-task — what actually happens next? Does insurance cover the PT? Does the doctor even know the protocol to fix it? I mean, that's not rhetorical, I genuinely don't know the answer.
Juniper Vale: Honestly? That's where the evidence goes quiet. The BGU finding — 120 people, up to seventeen years, Melzer's group plus Soroka and Minnesota — it's a genuinely strong signal for detection. But the implementation gap is documented across multiple systems. It's structural, not just a lag in awareness.
Mark Delaney: So the defensible version is: the research is solid, the professional bodies are aligned, and the infrastructure to act on it in routine care isn't there yet. That's the honest verdict.
Juniper Vale: Which circles us back to where you started, actually. You said this paper had been bothering you. And I think what was bothering you was exactly that — a tenth of a second predicted seventeen years of survival better than someone's age or their diagnosis list. The BGU finding is real. Melzer's group earned that. It's the distance between that number and your doctor's appointment that's the uncomfortable part.
Mark Delaney: Yeah. I mean — okay, I'll walk back the hot take a little. 'Step speed makes your doctor obsolete' was, uh, maybe a bit much. But the tenth-of-a-second thing? The hazard ratio? I'm keeping that. That part I still stand by.
Juniper Vale: You should. The research from Ben-Gurion University of the Negev is genuine. Knowing your step initiation speed under dual-task conditions — that's solved. Knowing what clinical infrastructure then does with that number — that's, right now, mostly not solved.
Mark Delaney: Two very different problems wearing the same lab coat. Thanks for working through it with me.
