357 ‒ A new era of longevity science: models of aging, rapamycin trials, biological clocks, & more

Kennedy's lab was among the first to show that baseline mTOR signaling creeps up with age in stem cells, preventing the pathway from turning off properly. This feeds a vicious cycle with chronic inflammation. Rapamycin, discovered as an antifungal on Easter Island, became an immunosuppressant at high continuous doses for organ transplant. However, at low intermittent doses (once weekly), it does not suppress immunity and may even enhance immune function against respiratory infections, as suggested by Mannick and Klickstein's work. Kennedy considers rapamycin the gold-standard small molecule for aging, with the strongest evidence across species. The Singapore trial is testing whether 6 months of 5 mg weekly rapamycin can shift biomarkers in healthy 40–60 year olds. Kennedy is skeptical that functional measures like DEXA or cognition will change in such a short, young cohort, but hopes to see signals in inflammatory cytokines and epigenetic clocks.

Kennedy's group screened natural products for lifespan extension in worms and mice, leading to AKG as a top hit. In mice, time-release AKG gave a 5–10% lifespan increase but a much larger reduction in frailty, effectively squaring the longevity curve. A human trial of time-release AKG alone (without vitamins) has been completed in Singapore, with results pending. An earlier uncontrolled community-based study using methylation clocks suggested a reversal of biological age by a few years, but Kennedy emphasizes the limitations of that data. He personally takes the full Rejuvenate product (AKG + vitamins) and has done so for years. He notes that when combined with sublingual NAD+, he experiences a marked exercise performance boost.

Kennedy was previously skeptical of oral NAD+ precursors (NR, NMN) because mouse studies showed minimal effects and he personally felt nothing from NR. A company (EX BioParma) approached him with a sublingual NAD+ product containing epigenin. He tried it alongside his daily AKG and noticed a dramatic, reproducible improvement in running performance: at a given heart rate, his respiratory rate was lower, perceived exertion dropped, and he could run faster and longer. He cycled on and off multiple times to confirm. He speculates that the combination of two declining metabolites (AKG and NAD+) provides synergistic metabolic flexibility. He has not measured his own NAD+ levels but says the company has unpublished data showing sublingual NAD+ incorporation into RBCs in rats. He now considers this his most exciting personal intervention and wants to do formal lactate testing.

Kennedy agrees with Attia that VO2 max, muscle mass, and strength have hazard ratios for mortality that exceed those of smoking, diabetes, or even cancer. He believes there is causality, not just association. He personally added resistance training because he wasn't getting enough muscle stimulus from running alone. He still runs for mindfulness but now prioritizes lean mass. He acknowledges that animal data suggest exercise primarily squares the curve, but he considers that a major win for healthspan.

Kennedy, who now directs a center for reproductive longevity, was surprised to learn how low HRT use is worldwide, especially in Southeast Asia. He attributes this to lingering fear from the WHI study and deference to outdated US guidelines. He draws a parallel to testosterone replacement in men, where the Traverse trial helped undo damage from earlier flawed studies linking testosterone to prostate cancer. He sees HRT as a low-hanging fruit for improving women's healthspan and is frustrated that it remains underutilized.

Kennedy is a proponent of empowering individuals to make their own health decisions, but he is alarmed by the trend of taking 10+ supplements simultaneously. He likens it to mixing paint colors — you get gray. His own approach is to add one new thing to his baseline (AKG), measure subjective and objective parameters, and then remove it to see if the effect disappears. He did this with the sublingual NAD+ product. He encourages people to educate themselves, understand the risks, and measure even simple things like heart rate, respiratory rate, and perceived exertion.

Kennedy explains that even in his 2014 pillars paper he drew lines connecting all the pillars because he never believed they were independent. He argues that aging is about the network that connects these pathways, and that healthy aging is about maintaining a responsive, malleable network that keeps you in equilibrium. When you slow aging with an intervention like rapamycin, you see improvements across all hallmarks, not just one. He believes the hallmarks are outputs or ways to look at aging, not the root cause. The real driver is a loss of homeostasis and dynamic range, particularly in pathways like mTOR. This view implies that targeting single hallmarks will not lead to radical lifespan extension; instead, we need to understand and preserve the network's resilience.

Kennedy describes working with physicist Peter Fedichev to model aging as a dynamic system. Imagine a ball in a deep valley when young: you can diverge from health but are pulled back. With age, the hills (resilience) come down, making it easier to fall into failure states (chronic disease, frailty). The linear accumulation of damage lowers the hills, while the chance of the ball going over the hill increases exponentially — matching Gompertz mortality. When they decompose biological age from large datasets like UK Biobank, they find one principal component that rises linearly (the damage) and another that oscillates around it (functional status). Most interventions, including rapamycin, seem to affect the oscillating component, suggesting they work 'around the edges' and may add 5–10 years of healthspan but not change maximal lifespan. To bend the slope, entirely new classes of interventions would be needed.

Using NHANES data from 1999–2000 with ~18 years of mortality follow-up, the team used AI to build a linear model from ~50 clinical parameters (CBC, metabolic panel, inflammatory markers, etc.). This clinical chemistry clock predicted mortality better than any other parameter in NHANES, including ASCVD risk scores. When methylation data later became available for the same cohort, they found that some first-generation methylation clocks were worse than chronological age at predicting mortality, while second-generation clocks did better — but the clinical chemistry clock still outperformed them. Crucially, the clock breaks down into principal components that map to smoking, metabolic disease, obesity, etc., allowing clinicians to see which subclinical abnormalities are driving elevated biological age. In one analysis, ~20% of people had parameters out of reference range but were not being treated; those individuals had higher biological age and faster mortality. People whose parameters were well-managed with medication had lower biological age than untreated healthy-seeming individuals, suggesting aggressive early treatment of preconditions is beneficial.

Kennedy states that when his lab tests combinations of interventions in mice, the effects are more often antagonistic than additive. He likens taking 20 pills to mixing 20 colors of paint — you get an ugly gray outcome. Even with well-studied molecules like rapamycin and metformin, the additive effects are small. He believes that without understanding the primary target of each compound, rational combinations are impossible. His lab is now doing proteomics thermal shift assays to identify binding partners for compounds like urolithin A, so they can eventually design synergistic combinations. He personally tries only one or two new things at a time, measures his response, and then adds or removes. He supports self-experimentation but urges caution and measurement.

Kennedy had largely given up on the NAD+ pathway until a company (EX BioParma) approached him to test a sublingual NAD+ product that also contains epigenin, a CD38 inhibitor. CD38 is an enzyme that consumes NAD+, so inhibiting it may further boost NAD+ levels. He began taking it alongside his daily time-release AKG (rejuvenate). Within three days, he noticed on a treadmill run that he kept increasing speed without getting out of breath; he ran 12K instead of his planned 5K and still felt fresh. He observed that his heart rate would rise normally but his respiratory rate stayed lower, and his rate of perceived exertion tracked with respiration, not heart rate. He went off and on the combination multiple times to confirm the effect. He speculates that the two metabolites, both of which decline with age, are providing cellular metabolic flexibility. He has not measured RBC NAD+ levels himself but says the company has unpublished data showing incorporation. He now considers this his most exciting personal finding and wants to do formal lactate testing.