New biomarkers promise early disease detection, but doctors using them daily struggle to weigh the value of these warnings.

There’s a kind of frustration that drives people out of conventional medicine and into prevention. For Dr Alexandru Ardelean, a specialist in internal medicine at Years, it was a pattern, rather than a single moment. Working in the ER, he kept seeing the same cardiology patients cycle back through, and he started asking a question that traditional medicine isn’t built to answer: when did this actually start?

That question is the thread running through this week’s episode of Longevity Technology UNLOCKED, where hosts Dr Nina Patrick and Longevity.Technology CEO Phil Newman sit down with Ardelean and Jürgen Mehrtens, the clinician and CEO, respectively, behind Years, a Berlin-based longevity clinic betting its entire model on one idea: that the body sends out distress signals long before disease shows up on a chart.

Two signals, in particular, are getting a lot of attention right now: GDF-15 and P-tau217. Neither is a household name yet. Both might be soon.

Think of GDF-15 less as a diagnosis and more as your body’s smoke detector; it doesn’t tell you what’s burning, only that something is. Ardelean describes it as a marker of generalized cellular stress, produced across a huge range of tissues in response to everything from mitochondrial dysfunction to chronic inflammation. It’s already used clinically to help predict heart failure outcomes and, notably, in oncology, where it’s linked to the appetite loss seen in cancer cachexia – one reason a pharmaceutical company has developed an antibody specifically targeting it.

But its pull into longevity medicine came from a different body of research: studies showing GDF-15 tracks closely with chronological age, and that its trend over time, not a single snapshot, predicts mortality and illness down the line.

As Ardelean put it: “we do not use these biomarkers as of yet in preventative medicine to build the therapeutic strategy. We use them in order to measure the therapeutic success,” meaning a high reading doesn’t hand you a treatment plan. It tells you whether the plan you’re already on is working.

What made this conversation refreshing is that it didn’t oversell the science. Pressed on whether these markers throw off false positives, Ardelean’s answer wasn’t reassuring in the way marketing usually is; it was more useful than that.

“My concern is not necessarily about false positives, it’s about false negatives,” he said, describing patients whose P-tau217, a marker tied to early Alzheimer’s risk, came back elevated but who showed no cognitive symptoms, no genetic risk and scored well on testing. In those cases, the clinic leans on repeated measurement over time rather than reacting to one number.

There’s no settled reference range for GDF-15 in preventive use either; it can swing from 300 to 600 in a perfectly healthy 40-year-old. Ardelean even floated a working theory mid-interview, coining the idea of a “cellular stress reserve,” the notion that some people’s healthy habits may be buying them a buffer that keeps the number looking normal even as underlying strain builds. It’s a hypothesis, not a finding, and he said as much. That kind of transparency is rare, and it’s worth more to patients than false certainty.

The business model behind Years is arguably the more provocative part of this story. Mehrtens set out to build a longitudinal dataset, and realized the only way to guarantee data quality was to own the clinics generating it. Every client runs through an identical diagnostic protocol, repeated annually, no menu nor shortcuts. The pitch to future partners, including pharmaceutical companies, is a dataset clean enough to accelerate drug development.

It’s a bet that consistency beats convenience and a quiet rebuke of a healthcare system Mehrtens doesn’t hesitate to criticize.

“Our healthcare system is broken,” he said flatly, adding that it’s likely true across most Western countries, Germany and the UK included.

What stood out beyond the biomarkers was how much of Years’ actual retention strategy is built on community – running clubs, book readings, patients sharing their own turnaround stories with each other. It’s an implicit acknowledgment that data alone doesn’t change behavior; belonging might. As competitors in the functional medicine space struggle with client retention because, in Mehrtens’ words, patients often “don’t see or feel a difference,” Years is betting that visible, trackable progress – paired with people cheering each other on – is what keeps someone coming back for their next blood draw.

Predictive biomarkers won’t replace the boring fundamentals – sleep, movement, nutrition – anytime soon. But they might finally give people a reason to take them seriously before the ER visit, not after.

Want to stay ahead of what’s next in longevity? New episodes of Longevity.Technology UNLOCKED land every Monday, followed by our weekly news roundup each Friday. Find us on  Apple Podcasts, Spotify and YouTube.