Howie et al looked at infection (particularly gastro-intestinal) and found
that 13 weeks of breastfeeding or *mostly* breastfeeding gave the best
outcomes. I recall mostly abm was better than all abm.  This paper was
published in the British Medical Journal in 1990 or 91.  This doesn't
support the all or nothing hypothesis, though perhaps other studies do.
Mary Broadfoot, Paisley, Scotland
[log in to unmask]