Peter Attia hosts this special episode of The Drive Podcast, focusing entirely on aging clocks - biological tests that claim to measure how fast you're aging beyond chronological age.

The episode examines what aging clocks actually measure, how epigenetic clocks work using DNA methylation patterns, and whether changing these measurements translates to meaningful health improvements.

Attia analyzes two key studies: the DO Health trial testing omega-3, vitamin D, and exercise interventions on four common aging clocks, and research using brain MRI imaging to estimate pace of aging and mortality risk.

The Promise and Problem of Aging Clock Shortcuts

Perfect anti-aging trials would measure hard outcomes like heart attacks, cancer, dementia, and death, but these require 20-year studies with enormous complexity and cost.

Aging clocks represent the most popular proposed shortcut - a proxy marker that moves faster than hard outcomes but still predicts them reliably.

The goal is a single number predicting biological age different from chronological age, or a rate of aging reflecting your current trajectory for precision medicine applications.

How DNA Methylation Powers Epigenetic Clocks

DNA methylation adds methyl groups to cytosine-guanine phosphate bonds (CpG sites), influencing gene expression without changing DNA sequence.

Methylation patterns change predictably with age, affected by smoking, metabolic health, and inflammation, creating a molecular record of body processes.

First-generation clocks like Horvath trained on cross-sectional data to predict chronological age within a few years, proving the concept worked.

Second-generation clocks shifted from predicting age to predicting mortality, physiological decline, and pace of aging using more clinically meaningful outcomes.

Four Major Aging Clocks Tested in DO Health Study

PhenoAge uses 500 CpG sites trained on clinical biomarkers including albumin, glucose, C-reactive protein, kidney function, and white blood cell count.

GrimAge uses 1,000 CpG sites to estimate plasma proteins like GDF15, leptin, PAI1, plus smoking exposure, combined with age and sex to predict time to death.

GrimAge 2 updates the original with C-reactive protein, A1C, and other refinements for improved mortality prediction.

Dunedin PACE uses 173 CpG sites trained on longitudinal New Zealand data to estimate rate of aging rather than biological age.

Mixed Results from Omega-3, Vitamin D, and Exercise Trial

The DO Health study randomized 800 healthy adults aged 70+ to combinations of 2,000 IU vitamin D, 1 gram omega-3, and 90 minutes weekly exercise for 3 years.

Omega-3 supplementation showed significance in three of four clocks (PhenoAge, GrimAge 2, Dunedin PACE), but the effect was only 3 months of reduced aging over 3 years.

Vitamin D at 2,000 IU showed no impact on any clocks, possibly due to insufficient dosing given 30% had baseline levels below 20 ng/dL.

Exercise failed to show independent effects, likely because 88% of participants were already physically active at baseline.

Technical and Biological Noise Complicate Measurements

DNA methylation measurements suffer from biological noise (recent illness, heavy workouts) and technical noise (sample handling, DNA extraction efficiency, batch effects).

Clocks measure hundreds of thousands of methylation sites across the genome, then collapse all information into a single summary score with inherent compression risks.

"BMI is a great example - at the population level it's pretty good, but if I have an individual with BMI 24 versus 29, it's tough to tell which is healthier" - Peter.

Life Insurance Companies Outperform Aging Clocks

Life insurance companies predict mortality with such accuracy that premium payout deviations exceeding 1% would be considered extraordinary.

When asked directly, senior life insurance executives confirmed they don't use any commercially available or research-grade biological clocks in their actuarial models.

Insurance companies rely on traditional biomarkers and factors that have decades of evidence linking them directly to clinical outcomes.

Research Promise vs Consumer Reality

Aging clocks may help researchers detect early biological signals where traditional outcomes would take decades to measure, making 20-year prevention trials more feasible.

Different clocks capture different biology aspects, and small measurement shifts don't necessarily translate to meaningful health improvements.

"All models are wrong, some are useful" - the question remains how useful these models actually are for individual decision-making.

Current evidence doesn't provide clear answers about what to do if your aging clock changes by a few months - should you modify diet, exercise, or medications?