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

What It Covers

Peter Attia examines aging clocks—epigenetic tools that estimate biological age via DNA methylation patterns—analyzing two studies: the DO Health randomized trial testing omega-3, vitamin D, and exercise across four clocks, and a brain MRI study measuring pace of aging to predict dementia risk and mortality.

Key Questions Answered

• •Epigenetic Clock Mechanics: Aging clocks measure methylation at CpG sites—locations where cytosine links to guanine on the DNA backbone—across hundreds of thousands of genomic positions, then compress that data into a single biological age score. These methylation patterns shift predictably with age and are influenced by smoking, inflammation, and metabolic health, making them molecular records of biological wear.
• •Clock Discordance Problem: Four leading clocks—PhenoAge (500 CpG sites), GrimAge (1,000 CpG sites), GrimAge2, and DunedinPACE (173 CpG sites)—produced conflicting results in the DO Health trial. Omega-3 moved three of four clocks; GrimAge showed no effect from any intervention. Before ordering a clock test, identify which specific clock a company uses, since different clocks capture different biological pathways and produce meaningfully different conclusions.
• •Omega-3 Signal Across Clocks: In the DO Health trial, 1 gram daily of EPA/DHA (330mg EPA, 660mg DHA from marine algae) over three years produced statistically significant reductions in biological age across three of four epigenetic clocks. The magnitude translated to roughly three months of slowed aging over three years—a modest but consistent signal suggesting omega-3 status meaningfully influences methylation patterns tied to aging biomarkers.
• •Noise Limitations Before Purchasing Tests: Two distinct noise sources undermine individual-level aging clock reliability: biological noise from transient inflammation, recent illness, or post-exercise recovery that temporarily shifts methylation readings, and technical noise from sample handling, DNA extraction variability, immune cell composition in blood, and batch effects on methylation arrays. The same blood sample sent to multiple labs can return different biological age scores.
• •Validated Biomarkers Still Outperform Clocks: Life insurance actuarial models predict population mortality with under 1% deviation using blood pressure, glucose, lipids, smoking status, and basic fitness metrics—without incorporating any commercial or research-grade aging clocks. Until epigenetic clocks demonstrate direct links to hard outcomes like cancer, cardiovascular disease, or dementia incidence, established clinical biomarkers provide more actionable individual health guidance.