Do you mean the risk of developing new-onset diabetes in healthy population or an specific diabetes-related outcome (MI, CHF, Stroke, Neuropathy, Retinopathy, ecc)
1) The authors used data from different studies to model A1c as a function of diabetes incidence. They also plotted a graph of A1C values and diabetes incidence.. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2890379/
2) A post-hoc analysis of the DPP trial showed that "..For every kilogram of weight loss, there was a 16% reduction in risk.." http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1762038/
I suggest reading post-hoc analysis of trials on diabetes prevention> you might find the info you need.