<International Diabetes>:   Professor Holman， you just mentioned several agents， acarbose and nateglinide.  From the STOP-NIDDM trial it was suggested that acarbose significantly reduces the risk of myocardial infarction， cardiovascular events， and some have even said it reduces hypertension.  However， the NAVIGATOR study with nateglinide it didn’t seem to show an effect in reducing diabetes or improving heart disease.  With these different glycemic drugs we see different results.  Do we need to pay more attention to characteristics and mechanisms of these drugs in patients who also have diabetes?
 
    Prof. Holman:   Absolutely.  We now know from NAVIGATOR that simply reducing postprandial glucose with nateglinide is ineffectual.  There is an interesting study from Japan recently which shows that if you compare nateglinide with acarbose you get the same degree of glucose lowering after a test meal， but with acarbose you see a distinct improvement of flow mediated dilation， which you don’t see with nateglinide.  That is just one example of a class of agents that has an effect that is not directly linked to the glucose.  Agents like acarbose and new classes like DPP-4 inhibitors and GLP-1 agonists with all the incretin biology， do offer a more complex intervention， which may well have additional benefits that we are now testing with the T-COSS and E