Using data from the United States Centers for Disease Control and the United States Census Bureau we measure normalised mortality age profiles (NMAPs) of COVID-19 for the 25 worst-affected States. There are clear trends in NMAPS with the start date of the epidemic for each State and its geographical location, with deaths increasingly concentrated in the older age groups for States with later epidemics. Deviations from this trend are correlated with the health index of each State. These findings are predicted outcomes of the recently proposed pre-conditioning field (PCF) hypothesis. A more detailed analysis is performed for the north-eastern States to analyse the effect of the initial epidemic in New York City (NYC) on neighbouring States. Strong geographical trends in NMAPs from west/south-west to east of NYC are highlighted. The NMAPs for the States of New York and Massachusetts show a remarkably precise overlap if the NY data is multiplied by a well-established immune system ageing function, confirming the inter-regional predictions of the PCF hypothesis. We briefly discuss other possible underlying causes for the measured trends, and suggest more sophisticated data analyses to test the PCF hypothesis.