> "These sex differences were larger in countries with a higher level of women's empowerment."

There are confounding factors here, though — it's not like the only variable being changed between two countries is a woman's empowerment.

See also, Simpson's paradox.