Eek, I just reread my earlier post and I was being misleading - in that study I mentioned, 1.2% were exclusively breast fed to 6 months. Some others would have been mix fed or exclusively formula fed to 6 months - but I gather, not many. So sorry for not making that clear. That said, it really is fabulously rare to wean at 6 months.
niecie,it's a good point about why there isn't a good study we can point to, comparing babies who were weaned at different ages.
There have been a few studies like that, but the results are inconclusive. The main difficulty is with confounding. What this means is that babies who wean early and late are different for other reasons than simply the age they were weaned.
You mentioned studying the babies weaned at 6 months, comparing them with babies weaned earlier. But those 6 month babies aren't a random section of the population.
It's been established in other studies, for example, that the more educated parents are, the later they tend to wean. The parents of the 6 monthers, in particular, are likely to be well educated and have read around the topic in order to have resisted the crowd pressure to wean earlier.
Many of the babies in that group will come from allergic families, who tend to wean their babies later. Babies weaned at 6 months are more likely to be breastfed than the general population (and they'll be quite a special case of breastfed babies too, as most have stopped by 6 months). And some babies who wean later might be refusing food because they're intolerant to something and it's making them feel unwell.
So you see, these babies could be different from the general population in various ways. If you then compare their health with other babies, you're comparing apples and pears. You might find differences between the groups but it'd be hard to know whether the diferences were due to weaning, or some other factor.
The only way to be sure of eliminating the confounding effects would be to randomise (ie tell parents "You will wean your baby at X weeks") but most parents wouldn't stand for it (I wouldn't).
My job is in research (nothing to do with infant nutrition!) and it strikes me every day how easy it is to ask questions - good, relevant, practical questions - but how difficult it is to design a study that will clearly answer that question. Very frustrating!