I just had to add my .02 on this issue. I agree with Marsha that the
numbers are rarely what they seem and that we have to be very careful about
research that is published by the formula industry. They have an
interest in asking the questions in a way that will give the results
they are looking for. Would anyone rely on smoking data published
by the American Tobacco Institute? They have an interest in making it
look like fewer kids are smoking etc. etc. Same is true of breastfeeding.
  The question is why is this the only information on this? THis is
something that the government should have good data on. I don't know
what is the right office, perhaps CDC, or one of the many offices in
the Department of Heatlth and Human Services that deals with maternal
child issues. Any insights from lactnetters outside the U.S. Who
compiles this data in other countries?
  The MArch 1`997 issue of JHL has an article byKaty Kennedy and Cynthia
Visness comparing the Roos study with the National Maternal and INfant
Health Survey from 1988 and 1989. They conclude that the numbers are
about the same and that becasue the Ross study is ongoing, it is
valuable to use. They also conclude that the govt study is a better
representation of the US population. THe study looked at demographic
charactertistics of breastfeeders but did not really look at issues
like breastfeeding defitions and who is breastfeeding how much and for
how long.  IN general, I think that we need to see that these data are collected
by objective sources like the government so that we do not need to rely
on Ross for our bf statistics. That's what we pay taxes for, isn't it?
  Naomi Bar-Yam PhD