Can you actually measure how fast you're aging, and can you change it? In this episode of One Step Ahead, Brad Dieter, PhD breaks down the real science behind "biological age" and epigenetic clocks: what they actually measure, how researchers validated them against real disease and mortality outcomes, and what happened when scientists tested whether calorie restriction and diet quality could change them in actual randomized controlled trials. We also get honest about what direct-to-consumer biological age tests can and can't currently tell you. This is the science behind one of the biggest buzzwords in longevity, without the hype. 0:00 Intro 1:27 What Biological Age Actually Measures 5:34 Does It Actually Predict Anything? 8:50 Can You Actually Change It? The CALERIE Trial 12:36 Diet Quality Beyond Calories: The Green Mediterranean Trial 15:49 Buying a Biological Age Test 19:31 What This Means For You 22:08 Takeaway
Brad Dieter (00:00)
Two people can have the exact same birthday, the same age, down to the day, and when researchers look at their DNA, not their genes, but the chemical tags sitting on top of their genes, they often find that one person's cells are functioning like someone five, ten, and sometimes fifteen years younger. That's not a metaphor, that's a real, measurable, published finding. And today, we're breaking down what it actually means, what can change it, and what almost certainly cannot.
Welcome back to One Step Ahead. I'm Brad Dieter, and this show is about the science of getting ahead of your health before something forces you to pay attention to it. Today we're talking about one of the most talked about and most misunderstood ideas in health right now. Biological aging. You've probably seen the headlines. Someone gets their biological age tested and it comes back ten years younger than their birth certificate, and it gets treated like proof they've cracked the code on aging.
We're gonna slow down and actually look at what these tests measure, how we know they mean anything at all, and whether you can actually change the number, and what honest limits of this science are right now. No hype, just what the data actually shows. So let's just jump right into it. The first topic is what biological age actually measures. So let's start with the basic distinction, because this is where most of the confusion starts.
Chronological age is simple. It is the time elapsed since you were born. It's a calendar fact, not a biological one. Biological age is something different. It's a statistical prediction and it comes from something called DNA methylation. Here's what that means in plain simple terms. Your DNA has small chemical tags called methyl groups that attach to specific locations on it, and those tags change in predictable, measurable patterns as you age.
Researchers can look at the patterns of those tags and build a model that estimates how old your cells look biologically compared to someone your same chronological age. I want to make one thing really clear here, because people get this wrong constantly. Methylation changes do not alter your underlying genetic sequence. Your genes themselves are not being rewritten. This is epigenetics, which means changes to how genes are regulated and expressed.
Sitting on top of a genome that stays exactly the same. Nobody's DNA sequence is changing. What's changing is which genes are getting turned up or turned down, turned on, or turned off. Now these tools are called clocks, and there have been a few generations of them. The first generation, built in 2013, includes the Horvath clock and the Hannem clock. The Horvath clock looks at 353 specific locations in the genome called CPG sites.
And it's notable because it works across almost every tissue in the body, not just blood. The Hanum clock is blood-based specifically, and it looks at about 71 sites. Both were built with one explicit goal: predict your chronological age as accurately as possible for methylation data. And they're genuinely good at that. They correlate strongly with your actual calendar age. But here's where it gets interesting. The second generation of clocks.
Things like pheno age and grim age, which we're gonna get to next, were built with a completely different goal. Not to predict your chronological age accurately, to predict your health outcomes, like disease onset and death. That shift turned out to matter enormously because a clock that just tells you your own birthday back doesn't give a doctor anything actionable. A clock that predicts your disease risk does. There's also a newer approach called the Dunedin pace.
Which comes out of the long running Dunedin longitudinal study in New Zealand. This one doesn't give you a single age number at all. Instead, it measures your current pace of aging, meaning how many years of biological aging you're accumulating per one calendar year, right now, today. A Dunedin pace reading of one point means you're aging at the expected rate. Above that means faster than expected, and below it means slower. That's Horvath's original 2013 paper.
Published in Genome Biology. Full citation in the show notes.
So, the most common misunderstanding people bring to this topic is imagining these clocks work like a bathroom scale for aging. You step on, you get a precise number, and you're done. The reality is messier and honestly more interesting. These are statistical models built from population data, and each one has a different design, philosophy, about what aging even means to measure. A first generation clock and done it in pace can tell you meaningfully different things about the exact same person.
And that's not a flaw in the science. It's more of a reflection of how complicated biological aging actually is. When someone tells me they got one of these tests done, my first question is always which clock? Because the answer changes how much weight I would put on the number.
Let's move on to our second topic. Does it actually predict anything at all? So why do these clocks matter clinically? Instead of just being an interesting novelty number, the answer is that the second generation clocks have actually been validated against real hard outcomes and large groups of people, not just shown to correlate with chronological age. That would be a much lower, much less useful bar to clear. So let's talk about grim age specifically.
From a 2019 paper by Lewin colleagues and large validation cohort, Grimm significantly predicted time to death, time to coronary heart disease, and time to cancer. So let's just sit with that for a second, because it's a categorically different and much stronger claim than this number correlates with age. It's this number measured today predicts who gets seriously sick and who dies sooner, and from which specific conditions.
Per the study's senior author Steve Horvath, who also built the original Horvath clock we just talked about, Grim Age was about 18% more accurate at predicting lifespan than simple calendar age, and about fourteen percent more accurate than early generation epigenetic clocks. That's a meaningful, measurable improvement in predictive power. It's not really just marginal statistical noise. So here's something worth explaining about how Grim Age was actually built, because it's pretty clever.
Instead of being trained directly on age, it was trained to predict several known mortality-associated biomarkers, things like plasma proteins linked to smoking-related inflammatory processes using methylation data. The resulting composite score turned out to predict mortality even better than any of the individual biomarkers it was built from. What this means in practical terms is that these tests are statistically linked to real future health events in very invalidated populations.
They're not just measuring some vague sense of how old your cells look or feel.
So this is a topic where I want to slow down because it's the difference between interesting and clinically meaningful. A test that just rederives your age from a blood sample is a neat party trick. A test that predicts ahead of time who develops heart disease or cancer, validated against people whose outcomes were already known, is a genuinely different category of tool. Grim age is getting close to clearing that bar. It's validated against heart outcomes, actual time to death.
Time to coronary heart disease and time to cancer in large cohorts. And not just a correlation with your birth date. Now, I do want to be careful here though. Statistically predicting population outcomes and telling you your personal fate are not the same thing. Grim being eighteen percent more accurate than calendar age at population level doesn't make your individual result a crystal ball. It's a real validated risk signal worth weighing as one input among many.
It's not a verdict of what your outcome is actually going to be and the exact timeline it's going to occur on. So can you actually change these biological aging scores?
Mm-hmm.
Okay, so if these clocks predict real outcomes, the next question is kind of obvious. Can you actually move the number? This is where the calorie trial comes in, and I think it's one of the more important studies in this whole space. Calorie stands for Comprehensive Assessment of Long-Term Effects of Reducing Calorie Intake of Energy. It's a long acronym. But it's one of the few large multi-year randomized controlled trials that directly tested whether an intervention changes biological aging.
As measured by these clocks. And importantly, it was done in healthy, non-obese adults, not a population that was already sick. That makes the finding more broadly relevant to a general audience. Here's the design. Participants were prescribed a 25% reduction in caloric intake. In practice, like we talked about with the diabetes prevention program last episode.
Sustained adherence in caloric restriction trials is typically lower than the prescribed target. So it's worth keeping in mind as a real-world caveat. These people were followed for two years with methylation-based aging measures taken at the start and after the intervention period. The result was that caloric striction slowed the pace of aging as measured by the Dunedin pace by about 2% to 3%, but it did not produce a statistically significant change in the static.
Pheno age or grim age.
I want to make sure this nuance actually lands because it's important. The effect was real in the sense that it was statistically detected, but it was modest in absolute size and it showed up specifically on the pace measure more than the age measure. Not every clock moves together in response to the same intervention. And this kind of inconsistency across different clocks is common and expected in this field right now. It's not a sign that the whole approach is unreliable.
The researchers' own interpretation, and I think this is the right way to frame it, dynamic pace of aging, measures like Dunadin pace may simply be more sensitive to picking up real intervention effects over a two year window than static biological age estimates. Those may need longer follow up or a bigger effect size to move detectably.
Now, I wanna be disciplined about not overselling this result because this field definitely has a height problem and I don't wanna add to it. A two to three percent slowing and pace of aging picked up on one clock but not others over two years of a genuinely hard intervention, because twenty-five percent caloric restriction is not trivial to sustain, is a real scientific finding. And it's not necessarily that caloric restriction or versus aging, which is kind of the headline version that tends to circulate.
I really want to be precise about what it was actually measuring here. Dunned in Pace is a validated epigenetic biomarker. It's not a hard clinical outcome. The calorie trial didn't show fewer heart attacks or longer lifespans because it wasn't designed or powered to show that. What makes it interesting isn't the size of the effect, it's that a rigorous randomized trial in healthy people detected a change in validated aging biomarkers at all.
That's a meaningfully different bar than most of what gets marketed in the longevity space. And it deserves to be reported at the size it actually is, not inflated for a better headline.
So let's talk about something beyond just calories. Because the calorie study asked whether eating less slows aging, but there's a separate trial that asked a different question entirely. Does eating specifically in a polyphenol rich way do something caloric restriction alone doesn't? And often polyphenol rich diets are a good marker for higher diet quality in general. This is the Direct Plus trial, also known as the Green Mediterranean Diet Study. It ran for 18 months with the
Completely different design from calorie. Instead of testing pure caloric restriction, it tested diet quality and polyphenol content specifically, using epigenetic clocks as one of several outcome measures. 256 participants, all with abdominal obesity or dislipidemia, meaning elevated blood lipids, were randomized into one of three arms. One got general healthy dietary guidelines, one got a standard Mediterranean diet.
And one got a polyphenol-rich green Mediterranean diet, which added green tea and something called manchi, a high polyphenol aquatic plant sometimes called duckweed, used here as a nutritional supplement, on top of standard Mediterranean diet pattern. Researchers calculated multiple generations of epigenetic clocks, Horvath, Hanum, Pheno Age, Grim Age, and done it in pace before and after the intervention, which gives a much more complete picture across different clock generations.
Than a single clock study would.
And here was the primary finding. The polyphenol rich green Mediterranean diet showed a distinct epigenetic signature and greater biological age attenuation compared to both the standard Mediterranean diet and the general guidelines arm. Since all three arms involve some degree of dietary change, this appears to be a diet quality and specific compound effect, not simply a caloric restriction effect. This ties directly back to something we said earlier this season.
It's not just about how much you eat, it's also about the quality and specific consumption of what you eat. And here, the idea is showing up that the level at the level of gene regulation itself, not just body weight or blood markers.
What I like about pairing these two trials in the same episode is that they're testing genuinely different hypotheses, and I want to walk away from this episode, seeing the distinction clearly. Calorie asks whether eating less slows aging. Direct Plus asks whether eating better, specifically in a polyphenol-rich way, does something caloric restriction alone doesn't. The honest answer emerging from the field is that both levers seem to matter: quantity and quality, and probably through different mechanisms. And I'll be front.
Up front with you, I can't really nail down the exact quantitative size of how these different effects work. We just don't really know. These are very early in our studies, kind of early in our understanding. And I want to make sure that we kind of view this as improving the overall kind of energy balance we're in and our diet quality both probably contribute to slower biological aging. So now let's talk about buying a biological aging test. So with all of this getting real mainstream attention.
Direct consumer biological age tests are now everywhere and they're heavily marketed. I wanna be honest with you about what you can actually say about what they can and can't deliver, especially given how much money has changed hands in this space. What they can give
So what they can do.
So what they can do, they can give you a snapshot estimate that's correlated with population level health outcomes and validated research cohorts.
And if you retest after a real sustained intervention over a meaningful stretch of time, they may be able to show you directional change, especially with a pace of aging test like Dunn at in pace, which, as we just covered, may be more sensitive to picking up real effects than the static age estimates. What they can't reliably do yet? Give you a single number that's precise enough to drive an individual clinical decision from in isolation, or definitely prove that a change in your number came from any one specific thing you did. Real daily.
Life has too many simultaneous variables: stress, sleep, illness, medication changes, to isolate a single cause that can control the way a trial can. The test, retest, reliability, and clock choice also matter here. Horvath, Grim Age, and Dunedin Pace can give meaningfully different readings on the exact same person from the exact same blood draw. That's why part of researchers in this field still actively debate which clock is best suited for which purpose.
There isn't one settled gold standard yet. So here's the honest framing I want to leave you with. This is a genuinely promising, rapidly developing area of legitimate science. It's not quite pseudoscience and it's not a scam, but is also not yet a mature, individually precise diagnostic tool, the way, say, a fasting glucose test is. Treat it as a result.
As one data point worth revisiting over time and in context, not a verdict on your health and not something to be anxious about. So one practical note is if you do decide to get one of these tests, use the same tests, the same company, for repeat measures rather than switching providers between tests. At minimum, that controls for some of the clock choice variability we've just talked about.
So I get asked about these tests fairly frequently, and my honest answer doesn't always satisfy people looking for a simple yes or no.
I don't necessarily think they're a scam, but I also don't think the science is mature enough yet for a test result to be the primary basis for a major health decision on its own.
Where I find them genuinely useful is a motivational tool and a tracking tool. If getting a number back makes someone change their sleep, diet, or activity more seriously, and the retest with the same provider a year later after real sustained changes, that's a reasonable use case. Where I'd push back is treating a single result as a diagnostic truth or kind of panicking over one test result that comes back older than expected. The field itself is still working out clock reliability. Hold your own results.
With the same uncertainty the researchers hold their results in. So what does this mean for you? Let's try to bring this all together. Biological age is a real, quasi-measurable, and increasingly validated scientific concept. That's not pseudoscientists.
So what does this all mean for you? Let's bring it all together. Biological age is a real, measurable, and increasingly validated scientific concept. It's not quite pseudoscience, but it's also not a precise individual diagnostic tool, the way established decades-old clinical tests are.
The interventions with the clearest randomized trial evidence so far for slowing the pace of aging is moderate sustained caloric moderation. And the honest effect size was real but modest at about two to three percent on one specific measure, not a dramatic reversal of aging. So this is worth remembering the next time a supplement or protocol claims otherwise. Diet quality sp specifically, things like a very polyphenol or nutrient-rich Mediterranean style way of eating.
has shown its own distinct epigenetic effect in a separate trial, independent of just cutting calories.
That suggests these may be two separate additive layers rather than the same lever measured twice. So don't expect a single test of any kind to tell the whole story on its own. Like most of what we've covered this season, biological age is most useful as a trend to track over time along a real, sustained intervention, not a one-time verdict to be anxious about or treat as final.
And no coaching pitch here. I'm just naming honestly what the current evidence actually supports and being equally honest and specific about what it doesn't yet support. I think that's the more useful and more scientifically accurate thing to do for an audience that's genuinely trying to make good decisions with limited time and information.
If this episode accomplishes one thing, I want it to be a framework for evaluating the next big biological age claim you see online. Not just a verdict on today's specific claims. The pattern to watch for is consistent. Is this backed by a validated clock tested against real hard outcomes in a randomized trial with an honestly reported effect size, or is it a compelling mechanism?
Dressed up as a proven result. Everything we covered today that holds up past the harder bar. That's the lens I want you carrying into the next longevity headline you see. Long after you've forgotten this specific episode.
So here's the full takeaway. So here's the full takeaway. The intervention with the strongest trial evidence so far for slowing your pace of aging is moderate, sustained caloric moderation combined with a nutrient dense diet, a high quality diet, not a dramatic biohack. If you want to try it, give it months and not days before expecting anything to move.
That's it for this episode of One Step Ahead. I will see you guys next week. Thank you so much for tuning in.