Antimicrobial use (AMU) is often targeted as the primary driver of antimicrobial resistance (AMR), yet population data reveal a far more complex picture. Across Europe, women receive more antibiotics than men but experience fewer drug-resistant bloodstream infections. By age, resistance shows complex and varying patterns across sub-national regions and bacteria–antibiotic combinations, even though antibiotic exposure often increases in older groups. These contrasting trends suggest that selection may not be the main driver and that interventions should consider targeting different sex and age groups differently. Could this also reflect some level of ecological fallacy? To explore this, I will show that individual-level patient data on Staphylococcus aureus reveal similarly weak or indirect associations between antibiotic exposure and resistance acquisition. Even without fully resolving these mechanisms, explicitly modelling demographic structure helps refine baseline AMR burden estimates in ageing populations. Ultimately, capturing this complexity requires rethinking the importance of selection in the real world.