I. Experiments with Rats

IN current nutritional research it is a matter of importance that methods should be made available for detecting the presence of partial or latent deficiencies in the human subject-in other words, methods for assessing the "level of nutrition". Several such tests have been elaborated during the past few years, including the urinary excretion ("saturation") test for vitamin C and the darkadaptation test for vitamin A. For vitamin B1 the only method so far at the disposal of the investigator has been the test for urinary excretion [Harris & Leong, 1936; Harris et al. 1938]. Using this method it was shown that in normal subjects significant amounts of vitamin B1 are excreted in the urine (the output being related quantitatively to the past intake), that in beri-beri the excretion is negligible, that in " conditioned " polyneuritis (as associated for example with alcoholism, pregnancy, gastro-intestinal obstruction or faulty absorption) there is a similar deficit in the excretion of vitamin B1, and that in hospitalized patients the excretion is generally low, as a result no doubt of such factors as anorexia, restricted diets, increased metabolism resulting from pyrexia, and other causes [Harris et al. 1938]. Although it seems established that the urine test gives a reliable indication of the reserves of a subject in vitamin B1, and a helpful simplification of the technique is now offered by the proposed substitution of chemical procedures in place of the original biological method of assay [cf. Westenbrink & Goudsmit, 1937, 1, 2; Wang & Harris, 1939], it is still desirable to explore the feasibility of devising alternative methods which might be used alongside the urine test, either for purposes of comparison or as an additional check. The most likely line of approach seemed to be in the direction of carbohydrate metabolism. It is well known that in experimental animals suffering from beriberi abnormal quantities of pyruvic and lactic acid accumulate in various tissues and body fluids (for bibliography, see below). Platt & Lu [1936] have reported that in advanced beri-beri in human beings an increased concentration of pyruvic acid is present in the blood and urine; unfortunately, however, the fact that in mild degrees of deficiency little pyruvic acid could be detected seemed to limit the possibility of applying this finding in assessing the level of nutrition. However, we thought it worth while to ascertain whether with intakes of vitamin B1 only slightly below the optimum there might be some latent diminution in carbohydrate tolerance which could be made apparent by " loading tests" with the products of intermediate metabolism.