Abstract The colorimetric glucose-oxidase and the Somogyi-Nelson reducing-sugar methods yield the same blood-glucose values when applied to carefully prepared zinc hydroxide filtrates of blood obtained from fasting individuals. The anthrone method yields values only slightly higher than those obtained by the enzymic and the reducing-sugar methods. Use of the enzymic method in conjunction with the reducing-sugar or the carbohydrate method allows the simultaneous accurate determination of other carbohydrates added to the blood, with only microquantities of blood being required for analysis. Possible applications include the galactose and fructose tolerance tests, the inulin clearance test, and the determination of the body water "space" of various carbohydrates. To allow valid glucose determinations by the enzymic method, glutathione, cysteine, and amylase substrates such as glycogen must be absent. The anthrone reaction with glucose is enhanced by chloride, bromide, and iodide. To eliminate interference of chloride with estimation of blood glucose, the samples must be diluted appropriately before analysis.

Fales, F. W., Russell, J. A., & Fain, J. N. (1961). Some Applications and Limitations of the Enzymic, Reducing (Somogyi), and Anthrone Methods for Estimating Sugars. Clinical Chemistry, 7(4), 289–303.