The use of erythrocyte Na+/Li+ countertransport as a gold standard, as in the article by Halkin et al. (Clin. Sci. 1991; 81, 223–32) , should give rise to disquiet because it has been known for a long time that this parameter is influenced by low-density lipoprotein (LDL), by cholesterol and by saturated fatty acids. Recently, it has been clarified that LDL increases the Vmax of this flux . One has to agree that it is interesting that there is an inverse relationship between the ability to dilate blood vessels and Na+/Li+ exchange in the diabetic patient, but actually the basis for this apparent relationship is already known and it is not quite as direct as the authors might wish to imply.
It is not native LDL but oxidized LDL  that stops the formation of endothelium-derived relaxing factor (EDRF) , reduces prostacyclin production and even enhances ‘tissue factor' formation by endothelial cells . In fact, the reduced production of EDRF by diabetic patients that is pertinent to atherogenesis can be correlated with a raised plasma level of cholesterol  and also with the plasma level of lipid peroxides .
In that case there is some basic information missing concerning the effects of lipid peroxides on Na+/Li+ fluxes in erythrocytes. When this is known, it might well explain why this parameter seems to relate to hypertension and/or nephropathy in diabetic patients.
Meantime, in a study of this kind  we should have been told which of the subjects were smokers, especially as there were two outstanding points on each graph that formed the relationship.