Longevity Biomarker Heat Map

Citation-backed associations between biomarkers and aging-related outcomes, with an intervention-modifiability layer.

Loading data…

How to read this table

Each row is a biomarker, grouped by physiological domain. The five outcome columns show how strongly that biomarker predicts each aging-related outcome — all-cause mortality (the highlighted anchor column), cardiovascular disease, cancer, dementia, and frailty. Every cell is a hazard ratio standardized to a common scale, HR per +1 standard deviation of the biomarker, so markers measured in different units can be compared directly. Cell colour encodes effect size and direction — blue = protective, red = risk — and the cell border encodes evidence quality (A > B > C). Exposures that are inherently categorical (genotypes, calcium scores) cannot be standardized; they are shown with their native HR and visually flagged, never ranked against the per-SD scale. The right-hand Standardization column tags those biomarkers with a short reason — genotype, U-shaped, non-normal, clinical cutpoint — hover it for the full explanation. Hover any cell for the source study; click a biomarker name for its full profile.

The modifiability layer

Prediction alone does not tell you what to do. The rightmost column rates how much each biomarker moves in response to the best-available intervention — high, moderate, low, or fixed (genetically or structurally set). A marker that is both highly predictive and highly modifiable is a priority target (marked ★): worth measuring and worth acting on. High predictive weight with low modifiability is for risk stratification; high modifiability with little predictive weight is largely noise. Sort by Priority to bring the priority targets to the top of each domain.