increase in absolute T20 values in all the patients studied after protein repletion. Mean T20 values were 1.79 ml/min before and 2.94 ml/min after correction of the protein deficit. Since GFR increased during protein repletion, values for T20 were corrected for 100 ml of GFR before and after protein repletion. Mean values for T2o for 100 ml of GFR were 3.64 mI/mm in the malnourished state and 4.66 mI/mm after repletion. Hence, the observed increase in T20 values was proportionately greater than the increase in glomerular filtration rate. Effects of urea administration on urine osmolality. In four patients in the malnourished state urine osmolality after 14 hours of fluid deprivation was measured before and after the administration of urea orally in doses of 30 to 45 g daily. Urinary nitrogen excretion and blood urea nitrogen levels were determined daily. In every patient studied there was an increase in urine osmolality concomitant with an elevation in blood urea nitrogen levels and augmented urinary nitrogen excretion. In two patients discontinuation of urea administration resulted in a fall in the values of the three parameters studied. It has also been shown that vasopressin administration after urea loading in malnourished children resulted in increased urine osmolality. On the nature of the concentrating defect. The improvement in the concentrating defect observed in our patients during protein repletion and the dramatic change in concentrating ability observed shortly after the administration of urea argue strongly against the possibility of an anatomic alteration as the basis for the concentrating defect in malnutrition. Rather, a functional abnormality seems to be responsible for the concentrating defect. Impairment of sodium transport out of the loops of Henle seems unlikely since such a condition should affect both the concentrating and the diluting mechanisms. In our patients the diluting capacity was completely normal during malnutrition, indicating normal reabsorption of sodium in the ascending limbs of Henle's loops. Sodiumbalance studies carried out on these patients while in the malnourished state (see above) also showed adequate tubular reabsorption of sodium during periods of low salt intake. It is now known that the formation of a hypertonic milieu in the renal medulla is responsible for the concentration of the urine. The formation of such a hypertonic milieu is thought to depend upon the active transport of sodium ions into the medullary interstitium from the fluid traversing the loops of Henle, this process occurring predominantly in the ascending limbs [45]. It is generally believed that ADH acts by increasing the permeability of the distal nephron to water, thereby permitting osmotic flow of water into the surrounding hypertonic renal interstitium. Part of the hypertonicity of the medulla is the consequence of urea accumulation, resulting from reabsorption and recirculation of this molecule. Experimental studies have shown that urea may accumulate in the interstitial fluid of the renal medulla and thereby increase the maximal osmotic pressure to which the urine can be concentrated [46]. In our patients several days to several weeks of protein repletion were usually required for improvement of the renal concentrating defect. Since in these malnourished patients positive nitrogen balance was quite marked during the first few weeks of protein repletion, urea excretion through the kidney was small initially and increased considerably only after the patients were kept for more or less prolonged periods on a high protein diet. There was a good correlation between the values for urine osmolality after fluid deprivation and the concomitant 24-hour nitrogen excretion via the kidney. These two facts strongly support the idea that decreased urea concentration in the renal medulla is reponsible for the renal concentrating defect of malnutrition. The administration of urea to these malnourished patients resulted in a marked improvement in the concentrating defect similar to the improvement seen with protein repletion. However, the response was much faster, and urine osmolality increased considerably after only four or five days of urea administration. This improvement in concentrating ability occurred pan passu with an increase in blood urea nitrogen levels. Because of the fact that urea diffuses freely across cell membranes, it is very likely that the improvement in the concentrating Uosm T0 mOsm/kg H20 mi/mm N=7 N=4 / V 450 T0 ml/min/100 ml C1,, N=4 7I' + // / // 136 Klahr/Alleyne defect occurred concomitantly with an increased urea concentration in the renal medulla. Again a good correlation between concentrating ability and urinary nitrogen excretion was observed. This evidence strongly supports the hypothesis that decreased urea concentration in the renal medulla is responsible for the concentrating defect present in malnutrition. Improvement in the maximal urine osmolality has been observed during the administration of urea in normal subjects receiving a low protein diet [47]. A markedly reduced concentration of urea in the renal medulla has been found in dogs fed a low protein diet, but medullary sodium concentrations have been shown to remain unchanged [48]. Crawford, Coyle and Probst [49] have clearly demonstrated that in rats placed on a low protein diet and given