It's hard to draw that conclusion without knowing what's in either data set. I'm fairly certain that the obesity data would have gender as one of its attribute

It's hard to draw that conclusion without knowing what's in either data set. I'm fairly certain that the obesity data would have gender as one of its attributes, and it's also probable that any death statistics would include other vital stats such as height and weight at time of death. But you're right, it's certainly possible that all of the women in those states are of normal weight and it's only the men who are obese. I'm being a little glib though, it's hard to know without seeing the data.