We should really stop using the word "increases" to describe a correlation and use "goes hand in hand with" or something like that.
Same for your usage of "decreases". ;)
Respectfully, you seem to be confusing correlation with causation. A very common mistake.
I don't believe that drowning deaths are caused by ice cream sales either, but I can easily prepare the data in a way that makes that appear like a solid assertion.
Edit: I am intrigued by the downvotes. Perhaps I am wrong and ice cream sales actually do cause drowning deaths.
Sick short malnourished male is less likely to find partner, let alone have a children.
Study also excludes divorced men who technically have children, but do not live with them. Divorce decreases life expectancy severely.