04-08-2026Price:

The Frontier

Your signal. Your price.

SCIENCE

Aging clocks fail insurance and clinical tests

Wednesday, April 8, 2026 · from 1 podcast
  • Life insurers ignore epigenetic clocks, sticking to blood pressure and cholesterol.
  • A major trial found vitamin D and exercise barely move clock readings.
  • Technical and biological noise makes individual results unreliable.

The data-driven longevity industry is facing a hard reality check. Despite public fascination with epigenetic clocks like GrimAge and PhenoAge, experts argue these tools currently offer more research promise than personal utility.

Peter Attia, on his podcast The Peter Attia Drive, detailed the core problem: noise. A single measurement can be skewed by a recent workout or illness, while lab-to-lab variation often masks any real biological signal. “These models attempt to compress complex, multi-dimensional systems into one summary number,” he noted, a compression that loses the nuance needed for individual health decisions.

Peter Attia, The Peter Attia Drive:

- While aging clocks are a promising research tool for compressing multidimensional aging into a single metric, they currently lack proven clinical utility for individual health decisions.

- Proven biomarkers like blood pressure, glucose, and lipid levels have decades of evidence linking them directly to clinical outcomes, unlike current aging clock scores.

The disconnect is stark in the data. The DO-HEALTH trial, a large randomized controlled study, tested vitamin D, omega-3s, and exercise in healthy adults over 70 against four epigenetic clocks. The results were inconsistent and underwhelming: omega-3s showed a tiny benefit, roughly three months of "youth" gained over three years. Vitamin D and exercise had almost no measurable effect.

The ultimate litmus test is the insurance industry. Life insurers, whose business depends on precise mortality prediction, ignore epigenetic clocks. They still bet billions on blood pressure, smoking status, and lipid panels. Their payouts rarely deviate from internal models by more than 1%. If clocks offered a superior edge, the industry would have adopted them.

For now, the flashy summary number remains an experimental proxy. We don’t yet know if “turning back the clock” actually prevents heart attacks or dementia.

By the Numbers

  • 800participant countmetric
  • 3trial duration in yearsmetric
  • 70minimum participant agemetric
  • 500CpG sites measuredmetric
  • 1000CpG sites measuredmetric
  • 173CpG sites measuredmetric

Entities Mentioned

DO-HEALTHConcept
DunedinPACETool
GrimAgeTool
PhenoAgeTool

Source Intelligence

What each podcast actually said

#386 - Aging clocks—what they measure, how they work, and their clinical and real-world relevanceApr 6

  • The DO-HEALTH trial tested vitamin D, omega-3, and exercise on epigenetic clocks in 800 generally healthy adults aged 70 and older over three years.
  • The PhenoAge epigenetic clock uses methylation at 500 CpG sites and incorporates clinical biomarkers like albumin and CRP to predict mortality risk.
  • GrimAge uses methylation at 1,000 CpG sites to estimate levels of plasma proteins linked to aging and smoking exposure, combining them with age and sex to predict time to death.
  • The DunedinPACE clock uses 173 CpG sites in a longitudinal model trained on the Dunedin cohort to estimate an individual's pace of aging, rather than a static biological age.
  • In the DO-HEALTH trial, omega-3 supplementation showed a significant but small effect on three of the four epigenetic clocks, translating roughly to three months of reduced aging over the three-year study.
  • The vitamin D intervention in DO-HEALTH used 2,000 IU daily, which Attia considered modest; 30% of participants had baseline levels below 20 ng/dL.
  • Life insurance companies use proprietary actuarial models to predict mortality with extreme accuracy but do not currently incorporate commercially available biological aging clocks.
  • Peter Attia argues that while aging clocks are a promising research tool for compressing multidimensional aging into a single metric, they currently lack proven clinical utility for individual health decisions.
  • Aging clocks are susceptible to both biological noise, like transient inflammation from a workout, and technical measurement noise from sample handling and lab processing.
  • Attia notes that proven biomarkers like blood pressure, glucose, and lipid levels have decades of evidence linking them directly to clinical outcomes, unlike current aging clock scores.