It has been demonstrated that the generally accepted view of an equal or nearly equal distribution of sugar between corpuscles and plasma in human blood is erroneous.1 The evidence was based on true sugar values, obtained as the difference between apparent sugar (total reduction) and nonsugar-reducing substances.
It has long been known that nonfermentable substances constitute a part of the sugar values estimated by the usual reduction methods, but—owing to reasons set forth elsewhere2—these substances eluded quantitative determination to such an extent that some investigators asserted that almost half of the reduction of protein-free filtrates of normal blood is due to nonfermentable constituents, while others denied altogether the existence of such substances. A simple procedure devised by me, involving fermentation by means of thoroughly washed yeast, permits the rapid and complete removal of fermentable sugars from blood and other biologic fluids, and thus the accurate determination of nonfermentable reducing substances
SOMOGYI M. DISTRIBUTION OF BLOOD SUGAR BETWEEN CORPUSCLES AND PLASMA IN DIABETIC AND IN ALIMENTARY HYPERGLYCEMIA. Arch Intern Med (Chic). 1928;42(6):931-938. doi:10.1001/archinte.1928.00130230133012