Unfortunately, a detailed review by leading science journal Nature found that the diabetes study could not be duplicated with larger, better experiments, while the heart disease study was 'small, short, in an unsuitable animal model and had other design weaknesses' - which is pretty damning even before it fails to be backed up with better experiments. As is often the case with this kind of finding, all too often irresponsibly splashed across the newspapers, the original studies were too small, used bad statistics and simply don't tell us anything useful.
The Nature review calls the A1/A2 hypothesis 'ingenious' - and if it had been backed up it would have resulted in a major transformation of the milk industry. Unfortunately, however, the idea had no credible evidence to support it. Single, small trials are just not enough to do anything more than point out a direction for more in-depth work. In this case, the in-depth work - the stuff we need to listen to - showed that the A2 phenomenon to be unsupported.
Thanks to Professor Etienne van der Poel for bringing this to my attention.