#35 – Nir Barzilai, M.D.: How to tame aging

KEYPOINTS

• 01

  Nir was studying metformin's mechanism in 1987—before US approval.

  As an endocrinology/aging researcher he worked the drug's biology early, then watched interest fade until the geroscience wave revived it. That long arc is why he treats metformin as both an old diabetes tool and a candidate aging intervention, not a Twitter novelty. That early unfinished business with metformin is emotional fuel for TAME as much as scientific rationale. Few clinicians pairing bedside endocrinology with centenarian genomics bring that specific scar tissue to the metformin debate.

• 02

  Insulin resistance may be a stress response, not only a villain.

  Animal work he discusses complicates the cartoon: manipulating visceral fat and watching diabetes onset in rodents suggested resistance can be protective modulation under some conditions. Clinically he still fights hyperglycemia; intellectually he wants finer definitions than 'IR bad, always.' Peter presses him on clinical translation: nuance about IR as modulator must not become an excuse to ignore diabetes risk.

• 03

  TAME (Targeting Aging with Metformin) aims to treat aging as an indication.

  Because metformin is pennies, pharma will not fund the usual path—Nir jokes his yearly supply costs pocket change. TAME is designed to show delayed onset of a basket of age-related diseases, creating a regulatory path for aging itself rather than one disease at a time. Regulatory creativity—aging as indication—is half the scientific project. Success would unlock a playbook for testing other aging drugs without forcing every compound into a single-disease silo.

• 04

  Existing human signals: DPP, UKPDS, and observational dementia/cancer hints.

  Diabetes Prevention Program evidence for preventing diabetes, UKPDS cardiovascular signals, and emerging Alzheimer/cancer literature are why he thinks the prior for a broader aging effect is not crazy—even while demanding a proper trial. He is explicit that hints are not TAME; they are why TAME is ethical to run.

• 05

  Centenarians get a ~20–30 year 'bonus' and still die of the same diseases.

  They are not immortal; they postpone heart disease, cancer, and Alzheimer's. End-of-life costs in his cited comparisons can be lower for those dying after 100 than at 70—compression of morbidity as an economic and human argument. Families of centenarians often show the delay pattern across siblings, pointing beyond pure luck. That pattern—same exits, later tickets—is the healthspan definition Nir wants policymakers to fund.

• 06

  Centenarian genetics are not a 'perfect genome.'

  When his group checked ClinVar-style pathogenic mutation lists, centenarians were not mutation-free; many carried plenty. The gift looks more like protective variants and buffering (including pathways around IGF/GH and lipid genes such as CETP patterns he studies) than absence of all risk alleles. CETP VV and related lipid genetics appear in his longevity cohorts as clues, not consumer kit destiny.

• 07

  IGF/GH axis cuts both ways for aging and disease.

  Too much growth signaling links to certain chronic diseases; too little harms resilience. Nir and Peter dig into whether boosting HGH for lifestyle reasons is wise (generally skeptical) versus understanding naturally low-IGF protective genotypes in long-lived kindreds. Peter's clinical instinct against casual HGH aligns with Nir's read of the aging biology.

• 08

  NAD precursors (NR/NMN) get curiosity without blank checks.

  They discuss the hype cycle around NAD boosters frankly: promising mechanistic stories, thinner long-term human outcome data than the marketing. Metformin, for Nir, remains the nearer-term pragmatic test. Both prefer waiting for clearer trials over stacking unproven NAD protocols on healthy middle age.

• 09

  Max human lifespan arguments sit near ~115–122 with statistical caveats.

  Dying before 80 on average leaves decades of theoretical healthspan on the table. TAME's ambition is capturing a slice of that gap with a safe, cheap drug already used by millions. Even a five-year delay in multimorbidity would be civilization-scale if it generalized.

• 10

  Both men hunt their own blind spots out loud.

  Peter worries about over-interpreting rodent metformin data; Nir worries about under-interpreting converging human signals while waiting for perfect evidence. The episode models adversarial collaboration more than guru monologue. Listeners hear two experts willing to say 'I might be wrong' in real time—rare and useful.