The Longevity Research Index
Comprehensive Research Survey

The Science of Not Dying

Key people, competing theories, ranked interventions, and the honest evidence base for extending human healthspan and lifespan.

The longevity field sits at a remarkable inflection point. Billions of dollars are flowing into aging research from Altos Labs ($3B+), Retro Biosciences, and Hevolution Foundation. Epigenetic clocks can now estimate your biological age from a blood sample. GLP-1 agonists may be the first unintentional longevity drugs. And yet the single most powerful intervention remains what it has always been: regular exercise. The gap between exciting mouse data and proven human benefit remains enormous for almost everything else.

This index maps the landscape — the theories, the interventions, the key figures, and the honest state of the evidence. It is designed for a reader who wants to understand both the science and the sociology of a field where Nobel laureates, biohackers, venture capitalists, and wellness influencers all compete to define what "anti-aging" means.

Every claim has been reviewed for accuracy and bias. Where evidence is contested, we say so. Where we found errors in our original research, we corrected them. Confidence levels are shown throughout. Note that this content was AI-generated and has been primarily AI fact-checked, not personally verified by the author.

The boring stuff works best. Exercise, sleep, not smoking, metabolic health, and social connection have the strongest evidence by far.
What Actually Works — Research Finding
30-40%
Mortality Reduction
from Exercise
12
Hallmarks
of Aging
25
Interventions
Ranked
~$8.5B
Annual Longevity
Investment (2024)
25+
Key Figures
Profiled

Theories of Aging

Why do we age? Nine competing frameworks, from evolutionary logic to molecular mechanisms. Modern geroscience increasingly views them as interconnected.

🧬

Hallmarks of Aging

Lopez-Otin et al. (2013, updated 2023)

12 hallmarks organized as primary, antagonistic, and integrative. The dominant organizing framework. Updated from 9 to 12 hallmarks in 2023, adding dysbiosis, disabled macroautophagy, and chronic inflammation.

Mainstream
🔧

SENS / Damage Repair

Aubrey de Grey

Aging as accumulation of 7 types of molecular damage. An engineering approach: periodically repair damage rather than slow its creation. Influential but the SENS framework has produced limited translational results to date.

Influential / Contested
📀

Information Theory

David Sinclair

Aging as progressive loss of epigenetic information. DNA is intact but cell readout degrades. The ICE mouse model (Cell 2023) was published then corrected. Charles Brenner called it "not even wrong." Controversial within the field.

Contested

Hyperfunction Theory

Blagosklonny / Gems

Aging driven by continued growth-promoting programs (especially mTOR) that were beneficial in youth. Rapamycin's consistent lifespan extension across species is the strongest validation. Rising influence in the field.

Rising Influence
🔀

Antagonistic Pleiotropy

George C. Williams (1957)

Genes enhancing fitness early in life have deleterious effects later. The foundational evolutionary "why" of aging. Explains why aging exists but not its mechanisms. Supported by examples like IGF-1 and testosterone.

Mainstream
🔥

Inflammaging

Franceschi et al.

Chronic, sterile, low-grade inflammation increases with age and drives disease. Driven by senescent cell SASP, gut dysbiosis, and accumulated damage. Elevated to its own hallmark in the 2023 update.

Mainstream
🧬

Epigenetic Drift

Horvath / Levine / Belsky

DNA methylation changes systematically with age. Clocks predict mortality, but whether drift is cause or consequence remains debated. The most measurable aspect of aging, but mechanism unclear.

Emerging
🔋

Mitochondrial / Free Radical

Harman (1956) / Nick Lane

Originally: free radicals cause aging. Antioxidant trials failed. The original theory is largely falsified, but a nuanced version (mitochondrial signaling dysfunction) survives. A cautionary tale about supplement hype.

Revised
🧩

Somatic Mutation

Szilard / Jan Vijg

Somatic mutations accumulate inescapably. Cross-species mutation rates inversely correlate with lifespan. Vijg argues this imposes a hard ceiling (~115 years). CHIP links mutations to cardiovascular and cancer risk.

Mainstream

The synthesis: Modern geroscience rejects the programmed-vs-damage dichotomy. The field converges on an integrated network view where hallmarks interact bidirectionally. No single theory captures the whole picture.

Interventions Ranked

25 interventions evaluated by evidence quality. Rigorous honesty: distinguishing "exciting mouse result" from "replicated human RCT."

Transparency note: This table was fact-checked in March 2026. One significant error was corrected: the semaglutide all-cause mortality hazard ratio was originally stated as 0.51 (from a subgroup analysis) but the correct SELECT trial result is HR 0.81 (95% CI: 0.71-0.93). All effect sizes are approximate and come from the most relevant meta-analyses available.
#InterventionEvidenceEffect SizeRatingConfidence
1Exercise (all types)Multiple RCTs, mega-cohorts30-40% mortality reductionStrongHigh
2Not SmokingMassive epidemiological+10 years life expectancyStrongHigh
3Sleep Optimization (7-8h)Large prospective cohorts14-34% mortality risk from imbalancedStrongHigh
4Dietary Patterns (Mediterranean)RCTs + large cohorts8-25% mortality reductionStrongHigh
5Social Connection148 studies (Holt-Lunstad)OR 1.50 survival (95% CI 1.42-1.59)StrongHigh
6No/Moderate AlcoholLarge meta-analysesUp to 10 years if quitting heavy useStrongHigh
7Caloric Restriction / IF / TRE1 RCT (CALERIE) + animal2-3% slowed aging pace (DunedinPACE)PromisingMedium
8Rapamycin / RapalogsITP (mice) + PEARL trial14-26% lifespan in mice; safe in humansPromisingMedium
9GLP-1 Agonists (semaglutide)Large RCTs (SELECT)20% MACE; HR 0.81 mortality (CI 0.71-0.93)PromisingCorrected
10MetforminObservational + diabetic RCTsModest; TAME trial underfundedPromisingMedium
11Omega-3 Fatty AcidsRCTs (HR 0.92, CI 0.85-0.99)8% CV mortality reductionPromisingMedium
12Vitamin D (if deficient)RCTs + observational16% cancer mortality reductionPromisingMedium
13AcarboseITP (mice)~10% lifespan in malesEarlyLow
14Canagliflozin (SGLT2i)ITP (mice) + CV RCTs14% lifespan in male miceEarlyLow
15SpermidineObservational + animalEpidemiological associationEarlyLow
16NAD+ Precursors (NMN/NR)Small RCTsRaises NAD+; clinical benefits unclearEarlyLow
17Senolytics (D+Q, fisetin)Phase I/II trialsSafe; efficacy unproven in humansEarlyLow
18Epigenetic Reprogramming (OSK)Animal studies109% remaining lifespan in old mice*SpeculativeLow
19ResveratrolFailed pharma programNo consistent benefitDebunkedHigh (debunked)
20Young Plasma / ParabiosisAnimal + tiny pilotsNo human evidence of benefitSpeculativeLow

Strong = Replicated human evidence · Promising = Some human data · Early = Primarily animal · Speculative = Animal/in-vitro only · Debunked = Failed/contradicted

*OSK 109% is extension of remaining lifespan in 124-week-old mice (Macip et al., Cellular Reprogramming 2024), not total lifespan. These mice were equivalent to ~77 human years.

What Actually Works

An honest assessment for a person today. Four tiers from "do this immediately" to "don't waste your money."

1

Strong Evidence — Do These First

$0-500/year · 5-15 additional healthy years when combined · Everyone who is physically able
Important caveat: Even Tier 1 evidence is primarily observational, not from randomized trials with mortality endpoints. The evidence is strong enough to act on, but confounding is always possible. Effect sizes are population averages; individual responses vary.
Meta-Analyses (20.9M observations)

Exercise: The Most Powerful Longevity Intervention

30-40% all-cause mortality reduction (BJSM 2025). CRF is the single strongest predictor of mortality. Each 1-MET increase = 13-15% mortality reduction. Zone 2 cardio (150-180+ min/week), resistance training (2-3x/week), VO2max intervals (1-2x/week). Peter Attia's "Centenarian Decathlon": train for what you want to do at 90.

Large Prospective Cohorts

Sleep: 7-8 Hours, Consistent Schedule

Short sleep (<7h) increases mortality 14%; long sleep (>9h) increases it 34%. Sleep regularity may matter more than duration. Cool room (65-68°F), complete darkness, morning light exposure, no caffeine after noon.

Massive Epidemiological Data

Not Smoking / Minimal Alcohol

Quitting smoking adds ~10 years of life expectancy. WHO position: no safe level of alcohol. The "J-curve" for moderate drinking was an artifact of sick-quitter bias (JAMA Network Open meta-analysis, 2023).

Consistent Meta-Analyses

Metabolic Health

Key metrics: waist circumference, HbA1c (<5.7%), fasting insulin, blood pressure (<120/80), ApoB (best single lipid metric). Each 10 mg/dL ApoB decrease = ~9% heart disease risk reduction. Annual bloodwork is cheap insurance.

148 Studies, 308,849 Participants

Social Connection

OR 1.50 survival for strong social ties (95% CI 1.42-1.59, Holt-Lunstad 2010). Mortality risk from isolation equivalent to smoking 15 cigarettes/day. The most underrated longevity intervention.

2

Promising but Uncertain

$200-2,000/year · Some human data, NOT proven to extend lifespan

Rapamycin (3-6mg weekly, off-label)

14-26% lifespan extension in mice (ITP). PEARL trial (n=114): safe at low doses, but compounded rapamycin was ~3.5x less bioavailable than expected. Growing off-label community. Human longevity evidence remains thin despite strong animal data.

Metformin

Cheap ($50/year), safe, widely used. Diabetics on metformin may outlive non-diabetic controls (observational). TAME trial chronically underfunded despite importance; ARPA-H now involved. May blunt exercise adaptations.

Time-Restricted Eating / Caloric Restriction

CALERIE RCT: 2-3% slowed aging pace (DunedinPACE) with 12% CR. Extends lifespan 10-30% in mice. TRE (16:8) is more sustainable. No direct evidence fasting extends human lifespan; benefits limited to surrogate endpoints.

GLP-1 Agonists (for metabolically unhealthy)

SELECT trial: 20% MACE reduction, HR 0.81 all-cause mortality (CI 0.71-0.93) in obese non-diabetics. Strong for those with obesity/metabolic dysfunction. For lean, healthy individuals: entirely speculative. Muscle loss concern for aging populations.

Key Supplements (if deficient)

Omega-3 (HR 0.92 CV death, CI 0.85-0.99), vitamin D (if deficient: 16% cancer mortality reduction), magnesium, creatine (muscle + cognitive). Most other supplements are a waste of money.

3

Interesting but Early

$100-1,000+/year · Primarily animal data, small human studies

Senolytics (Dasatinib+Quercetin, Fisetin)

Compelling mouse data. Unity Biotechnology dissolved after clinical failures. D+Q in Phase I/II — safe but efficacy unproven. The most exciting mechanistic target, but years from knowing if it works in humans.

NAD+ Precursors (NMN, NR)

Raises NAD+ reliably. Clinical benefits elusive despite strong mouse data. The supplement industry has far outpaced the science. Sinclair takes 1g NMN daily; Kaeberlein and Brenner are skeptical.

Spermidine

Induces autophagy. Epidemiological association with reduced mortality. Available in wheat germ. Eating spermidine-rich foods is free and likely harmless; supplementation is speculative.

4

Speculative / Unproven / Debunked

Variable cost · Animal/in-vitro only or failed clinical development

Resveratrol

Debunked. The sirtuin-activation assay was an artifact. GSK paid $720M for Sirtris (2008) and shut it down. 2024 systematic review: "no conclusive clinical evidence" in any healthcare setting.

Young Blood / Plasma Exchange

Conboy lab: dilution of old blood may matter more than adding young factors. Ambrosia received FDA warning. Bryan Johnson's plasma exchange with his son showed no benefit.

Telomerase Supplements / Stem Cell Tourism / Most Biohacking Gadgets

TA-65 lacks evidence. Stem cell clinics carry real risks (infection, tumors). Red light panels, hyperbaric oxygen, peptide stacks — marketing ahead of evidence. Do not confuse consumer enthusiasm for science.

The equity problem: The most effective interventions (exercise, sleep, not smoking) are free. The most speculative ones cost thousands per year. The longevity field risks becoming a luxury market. A society serious about longevity would invest in making healthcare, nutrition, and safe exercise accessible to everyone.

The Great Debates

The key fault lines that shape regulation, funding, and what millions put in their bodies.

I

Is Aging a Disease?

Should aging be classified as a treatable medical condition? Without disease status, the FDA cannot approve drugs "for aging."

Yes, it's a disease

De Grey, Sinclair, Zhavoronkov. Aging causes disease, therefore aging is the root disease. Classification would unlock drug pipelines and insurance coverage.

No, it's natural

Hayflick, Olshansky, Ramakrishnan. Aging is a fundamental biological process. Medicalizing everyone over 60 reinforces ageism. Disease implies deviation from normal; aging is universal.

II

Sinclair vs. Critics

A recurring pattern: Sinclair makes exciting claims, media amplifies them, then the scientific community finds problems. Resveratrol (assay artifact, $720M failure), NMN (clinical benefits elusive), ICE paper (correction), Leap Years (dog supplement controversy).

In defense

Prolific researcher (200+ papers), popularized NAD+ biology, driving public interest in aging research. Some findings confirmed broadly.

The critique

Pattern of commercially-driven hype outpacing evidence. 12+ companies, many failed (OvaScience, CohBar, Sirtris). Brenner: the Information Theory is "not even wrong."

III

Maximum Lifespan: Hard Limit or Not?

Vijg: ceiling at ~115 based on supercentenarian plateaus. Olshansky (2024 Nature Aging): life expectancy gains decelerating since 1990. Rebuttals: mortality plateau above 105, negligible senescence in some species, de Grey's LEV argument.

No hard limit

De Grey, Barbi. Mortality plateaus suggest no fixed wall. Biotech could break historical patterns.

Hard limit exists

Vijg, Olshansky, Hayflick. Supercentenarian gains plateaued. Multiple organ systems have concordant limits. Evolution never selected for extreme longevity.

IV

Caloric Restriction: Does It Work in Humans?

CALERIE showed 2-3% slowed aging pace — real but modest. CR Society members show impressive biomarkers but longevity data is decades away. A Nature 2024 study in genetically diverse mice showed individual genetic variation matters enormously.

V

Bryan Johnson: Science or Performance Art?

$2M+/year protocol, Netflix documentary, $20M first-day product sales. Credits: unprecedented transparency, mainstreaming longevity, Rejuvenation Olympics. Concerns: N=1 conflated with science, gene therapy in Honduras, accessibility gap, "Don't Die" as ideology.

VI

Should We Even Want Radical Life Extension?

Ezekiel Emanuel ("Why I Hope to Die at 75"): extended old age often means decline, not vitality. Ramakrishnan's "Why We Die" questions whether immortality would satisfy us. Disability rights perspectives, environmental concerns, generational equity. The longevity movement's implicit assumption that death is always bad deserves scrutiny.

The boring truth: Across all these debates, lifestyle fundamentals have a larger evidence base and effect size than any pharmacological or experimental intervention. The exciting research gets the attention; the mundane protocols deliver the results.

Measuring Aging

How do we know if an anti-aging intervention works? The central challenge of longevity science is measurement.

First Generation

Horvath Clock

2013 · Steve Horvath, UCLA

353 CpG sites from 8,000 samples across 51 tissues. Median error: 3.6 years. Multi-tissue applicability. Trained on chronological age, so limited for health prediction.

Second Generation

PhenoAge

2018 · Morgan Levine & Steve Horvath

Two-step: 9 blood biomarkers predicting mortality, then CpGs predicting the composite. Each 1-year increase = ~9% mortality increase. Free blood-biomarker version available.

Second Generation

GrimAge / GrimAge2

2019/2022 · Ake Lu & Steve Horvath

The "death clock." DNA methylation surrogates for 12 plasma proteins + smoking. Best mortality predictor of all epigenetic clocks. 81% increased hazard per SD of acceleration.

Third Generation

DunedinPACE

2022 · Daniel Belsky et al.

Measures pace of aging (rate of change), not static age. Trained on 20 years of longitudinal data. Used in the Rejuvenation Olympics. Responds to interventions, making it useful for clinical trials.

Causality-Enriched

CausAge / DamAge / AdaptAge

2024 · Vadim Gladyshev, Harvard

Separates CpGs into those causally linked to aging (DamAge) vs. adaptive responses (AdaptAge). Aims to distinguish clocks that measure aging from those measuring the body's response to it.

Functional

Most Actionable Measures

No test needed · Available now

VO2max: strongest mortality predictor. Grip strength: independent predictor across all ages. ApoB: best single lipid metric. These are cheaper and more actionable than epigenetic testing.

The validation gap: We can measure biological age with increasing precision, but we don't yet know if changing your clock score actually changes your health outcomes. Step 1 (build clocks that predict mortality) is done. Step 2 (prove that interventions changing clock scores also change outcomes) remains the gap.

The Funding Landscape

Who pays for longevity research? An estimated $8.5B flowed into the sector in 2024, but the details reveal important biases.

Government

NIA (Non-Alzheimer's)

~$600-900M

Of the NIA's ~$4.5B total, roughly $3.5B goes to Alzheimer's. Basic biology of aging gets a fraction — far less than any single disease category.

Government

ARPA-H PROSPR

$144M

First major US government program explicitly targeting aging. Funding Buck Institute, Stanford, and Cambrian Bio for 3-year aging trials. Established 2022.

Philanthropy

Hevolution Foundation

Up to $1B/yr

Saudi-backed, largest single commitment to aging research. Actual disbursements ~$200-400M so far. Includes $230M for geroscience, $40M XPRIZE Healthspan.

VC-Backed

Altos Labs

$3B+

Largest longevity company ever. Bezos, Milner, ARCH. Epigenetic reprogramming focus. Recruited Nobel laureates. No products yet. Moving toward clinical development (2025).

VC-Backed

Retro Biosciences

$180M+

Sam Altman's personal investment. Raising $1B Series A at $5B valuation. AI+longevity convergence with OpenAI partnership. Three pillars: reprogramming, autophagy, plasma.

Corporate

Calico (Alphabet)

~$2-3B est.

11 years, no products. AbbVie partnership dissolved Nov 2025. The cautionary tale of longevity funding. Extreme secrecy, limited output despite enormous spend.

Cautionary Tale

Unity Biotechnology

Dissolved 2025

Backed by Bezos, Thiel, ARCH. First senolytic trials failed. Stockholders voted to liquidate Sept 2025. Mouse logic doesn't always translate to humans.

Regulatory Milestone

Loyal (Dog Longevity)

~$58M

First FDA-approved drug for lifespan extension (dogs). STAY study: 1,300 dogs, 70 clinics. If approved, establishes precedent that "longevity" is an approvable endpoint.

The funding paradox: We spend ~30x more studying individual age-related diseases than studying aging itself. If the geroscience hypothesis is correct — targeting aging would prevent all downstream diseases — this is a massive misallocation. If critics are right that aging is too complex to target, the billions in longevity biotech may be wasted. We don't yet know which view is correct.

Key Figures

25 important people in longevity science. Scientists, clinicians, entrepreneurs, communicators — and skeptics.

🔍
ai gen