Remember that when there is an ASSOCIATION between two things (as in the
Arch Fam Med article quoted), it does not mean there is a cause and effect.
Although it would make sense that if there is a difference in
breastfeeding and non-breastfeeding women, the breastfeeding women would be
less likely to abuse their children..... there is probably no way to PROVE
that it is the breastfeeding that made the difference.  You can't randomly assign women
to breastfeed or not breastfeed and have all other relevant variables the same.

Lucy Towbin, MSW, IBCLC