Work with Franco Bottazzo on the Bart's-Windsor Family Study founded by Andrew Cudworth, led to the establishment of the population-based Bart?s-Oxford Study in 1985. This led to the first survival analysis of type 1 diabetes prediction.
In 1997 I moved to Bristol with Polly Bingley, my closest collaborator, and many other members of the team. In collaboration with many other groups, my team helped to develop the current model used to predict future onset of diabetes in close relatives of an affected child. This led on to ENDIT (1994-2003), a multinational intervention trial.
Another major interest is in the epidemiology of type 1 diabetes, and I played a prominent role in the early stages of the EURODIAB ACE network, which has greatly clarified trends in the epidemiology of childhood onset diabetes in Europe. I believe that the human brain is an under-utilized resource in diabetes research.