From my post on previous thread:
Been reading about this. Along with viral inoculation amounts, I suspect it is the main source of variability of attack rates among affected individuals. Basically - type O is protective. Type A is susceptible. Type O blood is seen in 59-85% of Mexico and lower equatorial countries which have much lower cases of COVID-19.
Very complicated Sars Cov-2 apparently can't bind well to type O RBC surface proteins and when it tries it creates a highly antigenic protein bond structure that is easily cleared by native immunity.
Type A pts, on the other hand, apparently have different cell membrane serine protease activity, which interacts with serine residues on the viruses S spike protein. This is also governed by ACE-2 receptors, which have differing phenotypes in ABO variability. The cell glycan/S spike binding structure is affected by differing serine protease activity making it more easily penetrated. This all translates to easier sars cov2 human cell entry into type A individuals.
My read of complex topic.