If someone has threatened you with a gun or you live in an area with increased gun violence, I would expect that greatly increases the chances that you'd want to acquire your own. This cohort would already be at a much higher risk of gun death before they owned a gun as well.

There has never been a placebo controlled double blind study on gun ownership. Probably never will be because of ethical concerns.

It probably does increase risk, but I expect the effect size is much lower than 2-3x.