Earlier quoted context omitted.
The actual paradox is that, in the US, simply having a firearm in your home increases your risk of death to gun violence.
Simply owning or encountering a seat belt also 'increases' your risk of dying in a car crash. This is the kind of nonsense causation-correlation mix-up statistics you are operating on.
It's much more likely that you shoot yourself or your kid shoots you or your husband/brother/other-troubled-man has a bad day and shoots you than a criminal shooting you.
The relationship between gun deaths and guns is not correlative, it's causative. Because, surprise! Guns cause gun death.
Generally, less guns = less gun death. Which might seem like such a simple understanding that it must be naive or stupid. But no, it's actually just that simple.
On a related note, less automobiles = less automobile deaths.