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

Episode Summary

Executive Summary: Peter Attia explains what aging clocks are, how epigenetic and pace-of-aging models are built, and why they remain promising but imperfect proxies for longevity. He reviews two studies: the DO-HEALTH trial, where omega-3 modestly shifted several clocks, and a brain-MRI-based pace-of-aging model linked to cognition and mortality. His conclusion: useful research tools, not yet reliable individual decision aids.

Main Topics: Why aging clocks exist (Priority: 5/5): Hard outcomes like heart attacks, cancer, dementia, and death are the real endpoints, but they require very long, expensive trials. Aging clocks are proposed as faster proxy measures that might predict those outcomes sooner. Chronological age vs biological age vs pace of aging (Priority: 5/5): The episode distinguishes simple age from biological age (how old physiology appears) and pace of aging (how quickly deterioration is occurring now), emphasizing that these are related but not identical concepts. How epigenetic clocks work (Priority: 5/5): Attia explains DNA methylation at CpG sites, how methylation patterns shift with age and exposures, and how algorithms use these patterns to estimate age-related outcomes. Limitations: noise, compression, and interpretability (Priority: 5/5): He stresses biological noise, technical measurement noise, and the risk of oversimplifying a multidimensional process into a single score, making individual-level interpretation uncertain. DO-HEALTH randomized trial (Priority: 4/5): A 3-year factorial trial tested vitamin D, omega-3, and exercise in older adults against four clocks. Omega-3 showed the most consistent small effect; vitamin D and exercise were largely null. Brain MRI and pace-of-aging modeling (Priority: 4/5): A second study used structural MRI to estimate pace of aging and linked it to cognitive decline, frailty, dementia risk, and mortality, suggesting clocks can be built from non-blood data too. Clinical utility and consumer marketing (Priority: 5/5): Attia argues that current clocks are not yet actionable for consumers and that established markers like blood pressure, glucose, lipids, fitness, and smoking status remain more reliable for decision-making.

Key Arguments: Aging clocks are valuable because they may provide a faster proxy for long-term outcomes that would otherwise take decades to measure in trials. Chronological age already predicts mortality well at the population level, so any new clock must prove it adds meaningful value beyond age alone. Epigenetic clocks are built from DNA methylation patterns, especially CpG-site changes, which shift predictably with aging and exposures like smoking and inflammation. Second-generation clocks try to predict clinically meaningful outcomes such as mortality or pace of aging rather than just chronological age. Clock outputs can be distorted by biological noise (recent exercise, illness, inflammation) and technical noise (sample handling, assay variability, cell-mixture differences). In the DO-HEALTH trial, omega-3 supplementation was the only intervention with a consistent signal across multiple clocks, but the effect was small. Different clocks disagree with one another, so the choice of clock can change the apparent result of the same intervention. Current aging clocks are better viewed as research tools than as definitive personal health metrics because it is still unclear whether changing a clock changes real outcomes. Established clinical and lifestyle markers remain more actionable and better validated than commercial biological age scores. Life insurance companies already predict mortality extremely well using conventional actuarial data and do not rely on commercial biological clocks.

Data Points: Trial duration: 3 years - DO-HEALTH measured baseline and follow-up methylation after three years. Participants: Nearly 800 - Older adults enrolled in the DO-HEALTH study. Age of participants: 70+ years; mean age 75 - Healthy older adults in the DO-HEALTH trial. Healthy aging subgroup: About 50% - Roughly half met criteria for healthy aging. Regular physical activity: 88% - Most participants reported regular physical activity at baseline. Exercising >3 days/week: 60% - Baseline exercise frequency before enrollment. Vitamin D dose: 2,000 IU/day - Intervention dose used in DO-HEALTH. Omega-3 dose: 1 g/day total (330 mg EPA, 660 mg DHA) - Marine algae-derived EPA/DHA supplementation in DO-HEALTH. Exercise prescription: 30 minutes, 3 times/week - Home-based strength-focused regimen added to baseline activity. PhenoAge CpG sites: ~500 CpG sites - One of the next-generation epigenetic clocks used in the trial. GrimAge CpG sites: ~1,000 CpG sites - Clock estimating plasma proteins and smoking exposure to predict time to death. DunedinPACE CpG sites: 173 CpG sites - Clock designed to estimate pace of aging from longitudinal data. Effect size: 0.2 - Reported significance/effect size for PhenoAge with omega-3-related comparisons. Estimated aging reduction: About 3 months over 3 years - Approximate translation of the omega-3 effect depending on the clock. Baseline vitamin D deficiency: 30% below 20 ng/dL - A subset may have been underdosed relative to their baseline status. MRI study outcome: Structural brain imaging used to estimate pace of aging - Second study discussed as a non-blood-based aging clock approach.

Pivotal Quotes: "all models are wrong, some are useful" — Peter Atiyah: Used to frame aging clocks as imperfect but potentially valuable models. "we need a shortcut. We need a proxy." — Peter Atiyah: Explaining why the field seeks intermediate biomarkers instead of decades-long hard-outcome trials. "it’s not clear that these numbers actually give consumers any actionable information" — Peter Atiyah: His caution about commercial biological age testing and personal decision-making.

Implications: Aging clocks may help researchers screen interventions faster, but consumers should not treat them as definitive health scores. For now, validated basics—fitness, diet, sleep, blood pressure, glucose, lipids, and smoking status—remain the most actionable longevity tools.