Analysis of 501,946 UK Biobank participants reveals that blood samples collected between 2006-2010 predicted COVID-19 hospitalization (68% accuracy) and mortality (73% accuracy) using inflammatory and metabolic markers. The IL-1 pathway index predicted hospitalization specifically, while IL-6 trans-signaling predicted mortality—remarkably mirroring the later clinical success of tocilizumab (anti-IL-6R) versus limited efficacy of anakinra (anti-IL-1R) in treating severe COVID. Central adiposity, respiratory compromise, and cardiovascular markers further enhanced mortality prediction. This represents a paradigm shift in understanding pandemic vulnerability. Rather than COVID severity stemming purely from acute viral factors or healthcare access, it appears largely predetermined by measurable inflammatory and metabolic states existing years before exposure. The findings suggest we could identify high-risk populations before future pandemics emerge, enabling targeted interventions and resource allocation. However, this preprint awaits peer review, and the retrospective design cannot definitively establish causation versus correlation. While compelling, the real test will be whether similar predictive power holds for entirely novel pathogens in prospective studies.