Life-shortening injuries aren't counted as fatalities, so few fatalities doesn't necessarily negate the argument.
We're talking about life expectancy, not deaths.
Workplace injuries and stress don't kill you, but they make you die faster.
Yes, musculoskeletal injury is very common in manual jobs. But whether it's opiates today or whiskey in the past, the means to combat pain from work related injuries must have a negative effect on lifespan.
I doubt it’s relevant. In 2011, there were ~5k fatal work injuries, i.e. 0.2% of the 2,5M deaths in the US that year.
https://en.wikipedia.org/wiki/Occupational_fatality#/media/F...
https://www.infoplease.com/us/mortality/deaths-united-states...
Update: I’m not saying your job has no effect on your lifespan. It does, but mostly because of the (lack of) money you get from it rather than its dangerosity. Job-related deaths are probably too low to be the cause of the men/women lifespan difference.