The Future of Healthspan Research

The trial-design problem

You cannot run a decades-long randomised trial with mortality as the endpoint for every candidate intervention. That single constraint shapes the whole field, which is why so much effort goes into validating surrogate endpoints that predict outcomes over shorter horizons.

Biological age estimation

Epigenetic clocks built on DNA methylation patterns estimate biological age and correlate with mortality risk in population data. Newer clocks are trained on functional and mortality outcomes rather than chronological age, which improves their predictive value.

The unresolved question is whether these clocks are causal readouts or merely correlated ones. If they only report, moving the number may not move the outcome.

Active intervention classes

Senolytics aim to clear senescent cells; nutrient-sensing modulators target mTOR and AMPK signalling; NAD+ precursors address cofactor decline; partial reprogramming attempts epigenetic restoration without loss of cell identity. Peptide research intersects with several of these, particularly through mitochondrial and metabolic signalling.

What would count as progress

A validated surrogate endpoint accepted by regulators would change the field more than any single compound, because it would make trials of reasonable length possible. Until then, cautious reading of preclinical results is the appropriate default.