Introduction: autism--the challenges ahead.

There have been many important advances in research into the nature of autism and, as a result, our concepts of autism have undergone a radical change (Rutter 1999). At one time, the prevailing view was that autism was an unusually early variety of schizophrenia that had been caused, in large part, by so-called refrigerator parenting. It became clear that that was a wholly mistaken concept and that, instead, autism constitutes a neurodevelopmental disorder with a rather distinctive pattern of cognitive de¢cits, and that it is strongly genetically in£uenced. Nevertheless, we are a longway fromunderstanding the basic pathophysiology, and numerous puzzles and paradoxes remain. The aim of this symposium is to grapple with these issues, tackling the challenges from a range of di¡erent perspectives in the hope that a coming together of minds, and of di¡erent research strategies, may point the way ahead. My task is to set the scene by outlining some of these challenges in order to provoke us all to abandon the safety of our own research territory, and of the ¢ndings that are well established, in order to focus on the di⁄culties that are inherent in our favoured theories. We need to begin with implications of the huge rise in diagnosed autism (Baird et al 2000, Chakrabarti & Fombonne 2001, Fombonne 1999). To a substantial extent, this rise is a consequence of a major broadening of the concept of autism together with better ascertainment. However, is that all? When like is compared with like, has there been a real rise in the rate of autism? If that should prove to be the case, what is the environmental factor that has brought this about (the rise is unlikely to have been genetically determined)? There have been claims that the rise is due to the use of the combined measles/mumps/rubella (MMR) vaccine but that does not seem very likely. The rise began before the introduction of MMR and it continued to rise, without any plateauing, after MMR was used with the vast majority of the population (Dales et al 2001, Farrington et al 2001, Taylor et al 1999). But, if that is not the cause, what is?