Sir.—We wish to comment on the article by Drs Ginsberg-Fellner and Primack that appeared in the February issue of the Journal (129:240-243, 1975). We read with interest the clinical description of this very exceptional case of recurrent hyperosmolar nonketotic diabetes.
Unfortunately, the article does not contain values of blood ketones. However, they probably were not very high since the blood pH was normal. On the other hand in two thirds of the cases described in the literature, the pH values are in the acid range. Since these observations do not mention the levels of blood ketones and lactic acid, it is therefore impossible to determine the origin of the acidosis.
In the article we published,1 we found the acidosis to be due to an important accumulation of acetoacetate and β-hydroxybutyrate. In our case, the absence of urinary ketone bodies during the phase of oliguria can only be explained by