9. Aging Biomarkers as Surrogate Endpoints

Aging biomarkers may be extremely useful, but require validation for the intended use (11, 68). Surrogate endpoints should capture enough of the causal path from intervention to outcome; otherwise, they encourage optimizing the marker rather than improving health. This is a version of Goodhart’s law: when an aging clock becomes a target, it can cease to be a good measure of biological age. A pragmatic route is to start with short-term studies where biomarker shifts are linked to intermediate functional measures (e.g., recovery from stress, physiologic reserve) and then to hard outcomes in longer term studies. Different interventions may require different surrogates, e.g., a senolytic, a reprogramming therapy, and an exercise program may move different layers of the deleteriome. A preferred measurement system would track biological age longitudinally, at organ, system, and cell-type resolution. That requires minimally invasive sampling (e.g., liquid biopsies), multimodality (e.g., omics, imaging), robust computational models, and anchoring to physiology and outcomes.