Out of 400 Drugs, Only These Might Help You Live Longer - Dr. Kevin Perez and Siim Land

The 14 drugs that increased lifespan:
Cardiovascular / metabolic

  • Atorvastatin (statin)

PDE5 inhibitor

  • Sildenafil (Viagra)

Anti-inflammatory

  • Naproxen

Estrogen / hormone-related (multiple hits)

  • Estradiol
  • Estriol
  • Estraderm
  • Vagifem
  • Marvelon (ethinylestradiol + desogestrel contraceptive)

Other pharmaceuticals

  • Lymecycline (antibiotic)
  • Otomize (ear drop: steroid + antibiotic combo)

Vaccines

  • Avaxim (hepatitis A vaccine)
  • Revaxis (diphtheria, tetanus, polio booster)

ChatGPT summary:

Study Overview

  • Researchers analyzed ~500,000 participants from the UK Biobank with up to ~20 years of follow-up.
  • Looked at 400+ medications and compared users vs matched controls.
  • Adjusted for confounders: age, sex, disease status, BMI, smoking, socioeconomic status.
  • Goal: identify drugs associated with lower all-cause mortality (proxy for lifespan extension).

Core Finding

  • The majority of drugs were linked to higher mortality.
  • This is expected due to confounding by indication (people taking meds are often already sicker).
  • After filtering for strong statistical signals → only 14 drugs showed lifespan-associated benefits.

Top Longevity-Associated Drug Classes

  • Statins (e.g., atorvastatin)
    → Likely via cardiovascular risk reduction over decades
  • PDE5 inhibitors (e.g., sildenafil/tadalafil)
    → Possible mechanisms:
    • Improved vascular function
    • Lower blood pressure
    • Potential neuroprotective effects
    • Or indirect: higher sexual activity / overall health marker
  • Hormone Replacement Therapy (HRT) (estrogen/estradiol)
    → Strong signal in women
    → Supported by known sex differences in lifespan
  • SGLT2 inhibitors (newer diabetes drugs)
    → Among the strongest signals observed
    → Effects likely go beyond glucose control (multi-organ benefits)

Metformin Reality Check

  • Despite its reputation, Metformin did NOT show a lifespan benefit in this dataset.
  • Earlier studies suggesting benefit (even vs non-diabetics) have not consistently replicated.
  • Mechanism still unclear and possibly too “broad/dirty” compared to newer drugs.

Important Nuance

  • Not all “beneficial” drugs are overcoming disease risk.
  • Some (like HRT or PDE5 inhibitors) are not strictly disease-driven, which reduces confounding.
  • Socioeconomic factors were controlled for, but never perfectly.

Mechanistic Themes Emerging
Across both human data + animal work:

  • Insulin sensitivity / glucose handling (SGLT2, diabetes pathways)
  • Growth signaling (PI3K, mTOR, IGF-1)
  • Cardiovascular function
  • Hormonal environment
  • Systemic damage response

Animal + Translational Work

  • Follow-up studies are testing these drugs in:
    • Worms (C. elegans)
    • Flies
    • Fish
    • Mice
  • Goal: find conserved lifespan effects across species → higher chance of human relevance.

Biomarkers Angle

  • Shift from just lifespan → biological age + mortality prediction
  • Proteomics emerging as key tool
  • One standout marker: GDF15
    • Strongly correlated with age and mortality
    • Possibly reflects damage or senescence burden
    • Role (causal vs protective) still unclear

Limitations

  • Observational study → cannot prove causation
  • Confounding still present (even after adjustments)
  • Only includes prescription drugs (no supplements, diet, etc.)
  • Newer drugs (like SGLT2 inhibitors) have shorter follow-up

Big Picture Takeaways

  • True human longevity signals are rare when rigorously analyzed
  • Some drug classes consistently stand out:
    • SGLT2 inhibitors
    • PDE5 inhibitors
    • Estrogen therapies
    • Statins (context-dependent)
  • Metformin is not as strong as often claimed
  • Future direction:
    • Cross-species validation
    • Mechanism-first targeting (PI3K > mTOR may be underappreciated)
    • Better biomarkers tied directly to mortality

Bottom Line
We’re moving from hype to signal. Out of 400+ drugs, only a handful show real-world associations with longer life. The next step is figuring out which of these actually cause the effect and why.

1 Like