The relationship between fluoride level and uterine
cancer mortality rate appears to be considerably stronger even
after controlling for the effects of other confounding factors.

See: https://www.jstage.jst.go.jp/article/jea1991/6/4/6_4_184/_pdf