History of the measurement of glucose in urine: a cautionary tale.

In a recent study" of glucose reabsorption by the kidney it became clear that enzymatic methods of determining glucose in blood and urine produced a picture totally at variance with classical teaching, but traditional results could be obtained with the same material ifthe traditional copper reduction methods of estimation were employed. The major difference lay in the gross underestimation of glucose in urine by copper reduction techniques, the converse being true for plasma but to a lesser extent; the compound effects ofthe errors exaggerating the kidney's ability to reabsorb glucose. Glucose determinations based on the reduction of copper have been used since the nineteenth century and are still commonly employed in clinical laboratories. They were used in the classical work on renal physiology undertaken thirty to forty years ago, and which is only now being questioned. It seemed worth going back to the original biochemical literature to find out whether doubts had been expressed previously and, if so, why they nevertheless failed to cause persistent misgivings and subsequent search for better methods. Our review of the literature has shown that doubts indeed arose but for some reason they only led to a polemical situation which eventually died away during the 1920s without resolution. Despite the biochemists' failure to agree, clinicians and physiologists appear to have accepted "authoritative" methods without recognizing their limitations. As a result, edifices of clinical and physiological theory have persisted despite their weak technical foundations.