It is a mistake to conclude that if there was a big difference, we would have found it with largely observational data. There are many counterexamples to this i

It is a mistake to conclude that if there was a big difference, we would have found it with largely observational data. There are many counterexamples to this idea. I’m not so sure that the data proves that there isn’t much difference. The randomized PROBIT study showed less eczema (3 v 6%) and less gastrointestinal infections (9% v 13%). That is with 43% of mothers breastfeeding in the intervention arm versus 6% in the control arm at 3 months. You might not care about this at all, but clearly breastfeeding is doing something significant. It also shows, inescapably, that some kids will get eczema because they were not breastfed. This was my point that breastfeeding probably does make a big difference to some kids. Finally, this study showed that encouraging and supporting breastfeeding really made a lot more people breastfeed. Again, it seems very reasonable to encourage breastfeeding. You might not like the way it is done, but this is a public health type intervention, and it is impossible to please everyone. Most of the population isn’t looking up the evidence themselves, so I guess you aren’t really the target audience. I mean I could whinge about anti-smoking ads on TV because I would never smoke, or the extra time it takes people in wheelchairs to get on the bus because I’m not in a wheelchair, but on the population level these things are very desirable.