1. Rats were made oedematous by feeding them low-protein diets (protein: energy (P:E) 0.005) ad lib., and measurements were made of plasma and interstitial fluid colloid osmotic pressures (πp and πi respectively) and interstitial fluid hydrostatic pressure (Pi) before, and at the onset of, oedema formation. Taken together as (πp − πi + Pi) these forces oppose capillary pressure (Pc) and thus determine rates of transcapillary water filtration. Interstitial fluid was sampled, in non-oedematous and oedematous animals, from perforated capsules implanted subcutaneously for the measurement of Pi. Blood, plasma and interstitial fluid volumes were also determined.
2. In Expt I comparisons were made between animals fed on a control
diet (P:E 0.210) and the lowprotein diet. In normal animals the ratio πp:πi was approximately 2, but in protein deficiency it was increased since reductions in the absolute value of πi matched those in πp. These changes were observed 2 weeks after the start of the experiment and became more exaggerated when oedema appeared (weeks 18–22).
3. Pi was normally negative with respect to atmospheric pressure but increased to values close to zero when oedema formation occurred.
4. Despite the reductions in πp that were seen in the protein-deficient animals the sum of the forces opposing filtration (πp − πi + Pi) did not change significantly during the experiment.
5. Plasma and interstitial fluid volumes expressed per kg body-weight (measured using 125I-albumin and 35SO42−) were unchanged as πp initially decreased in the protein-deficient animals but increased markedly with the onset of oedema.
6. In Expt 2 comparisons were made between animals fed the low-protein diet ad lib. and others fed on the control diet in restricted amounts so that weight loss was the same in the two groups of animals.
7. The wasting induced by restriction of the control diet did not produce reductions in πp or πi and values for Pi were normal. Changes in the animals fed on the low-protein diet were similar to those observed in Expt I. By using 51Cr-labelled erythrocytes it was shown that the expansion in plasma volume that occurred when oedema appeared in the protein-deficient animals was mainly due to a reduction in total erythrocyte volume. Blood volume did not increase significantly.
8. It was concluded that in the hypoproteinaemia induced in the experimental animals reductions in the value of πp, which might otherwise result in an imbalance of forces that would produce excessive rates of transcapillary water filtration, were compensated for by reductions in πi. Increases in Pi also compensated but were quantitatively less important.
9. The significance of the results is discussed in terms of the pathogenesis of oedema in kwashiorkor and the concept of an oncotic threshold for oedema formation in hypoproteinaemia.