This is prompted by our current discussion about bedsharing, pacifier use and SIDS risk, but it applies in all cases where several factors are observed to occur together. It is of interest to note what common features characterize places with especially high, or especially low, prevalence of health problems. This tells us where we should take a closer look to find out whether the relationship is causal or merely co-incident (which I use here to mean 'occurring together', and not 'coincidental'). An example of just how far off the mark you can get when taking statistical association to mean causality is from the early days of AIDS research, when it was noted that many many AIDS patients had a history of sniffing amyl nitrite ampules. There was a serious theory launched that amyl nitrite caused AIDS. For those of you too young to remember, amys were a fashion intoxicant in the young male homosexual community in the late 1970's and early 1980's because sniffing them enhanced sexual pleasure, and the people who used them were also having a lot of unprotected sex with a lot of different people who were doing the same thing. Amyl nitrites were a confounder: something that goes along with all the other behaviors that predispose a person to something, but neither causes nor prevents it. From when I was about four, I believed that wind was caused by trees waving their branches, because I noticed that every time the wind blew, the trees were waving their branches. I remember very clearly when I learned in school what wind was, and it forced me to seriously revise my entire world view to bring it in line with the cartesian belief system in which I live. The reason researchers are now looking at bedsharing, at sleep position, at tobacco use, and other things, is that it has indeed been observed that in warm places such as Hong Kong where babies bedshare from the start, the SIDS rate is very low, while in New Zealand where babies were kept warm on lambskins in their beds, the incidence was very high. That was perhaps when someone thought to look at position for sleep, since Chinese practice is to place babies on their backs, to prevent suffocation (and, incidentally to prevent the backs of their heads from getting bulbous!) while in the West, parents were being advised to minimize the perceived risk of aspiration of vomitus by placing babies prone. I am conscious that the risk of aspiration would be higher in a non-breastfeeding society, since the vomitus is not benign when babies are getting non-human milk. I wish the scientific community would get its act together and use the same definitions for breastfeeding, bedsharing, pacifier use, and all of it, so we could actually have some hope of finding out something more useful than 'the quality of the research was not high enough to be able to infer causality'. As far as smoking is concerned, the evidence is very clear, but we still don't have a real clue WHY. It looks as though toxins in tobacco smoke, or possibly just the physiological effects of replacing air with smoke, affect CNS development in the fetus, so that the baby has a different threshhold of arousability, but we don't understand it fully nor can we say for sure that this is the ONLY mechanism for tobacco. We know enough now to say categorically that a baby ought to have its own bed and bedclothes when the adults smoke. Apart from the other caveats about safe sleeping for infants, we don't know enough to say categorically that most babies ought, or ought NOT, to bedshare. So, to re-cap: noticing that there are common features in communities with high or low prevalence of an outcome tells you only where to start digging to find the bone. The common features are NOT the bone. Rachel Myr, wearing my student of public health hat now, Kristiansand, Norway *********************************************** To temporarily stop your subscription: set lactnet nomail To start it again: set lactnet mail (or digest) To unsubscribe: unsubscribe lactnet All commands go to [log in to unmask] The LACTNET mailing list is powered by L-Soft's renowned LISTSERV(R) list management software together with L-Soft's LSMTP(R) mailer for lightning fast mail delivery. For more information, go to: http://www.lsoft.com/LISTSERV-powered.html