Dialysate fluids containing glucose polymers as osmotic agent are different from the conventional solutions, because they are iso-osmotic to plasma and produce transcapillary ultrafiltration (TCUF) by colloid osmosis. To investigate the effects on fluid and solute kinetics, a comparison was made between a 7.5% glucose polymer based dialysate (icodextrin) and 1.36% and 3.86% glucose based dialysate in 10 stable CAPD patients. In each patient three standard peritoneal permeability analyses (SPA) were done with the osmotic agents and concentrations mentioned above. Dextran 70 was added to the glucose solutions to calculate fluid kinetics. In the glucose polymer SPAs fluid kinetics were calculated from the dilution and disappearance of dextrin. The TCUF rate with icodextrin was closer to that obtained with 3.86% glucose than to 1.36% glucose. Extrapolation of the fluid profiles revealed sustained ultrafiltration with icodextrin. TCUF increased linearly in time in the icodextrin tests, whereas a hyperbola best described the glucose profiles. The effective lymphatic absorption rate with icodextrin was similar to the glucose based solutions. Mass transfer area coefficients of low molecular weight solutes with icodextrin were also similar to the values obtained with glucose, as was D/P creatinine. A positive correlation was present between the MTAC creatinine and the TCUF rate with icodextrin (r = 0.66, P = 0.05), which was absent in the glucose SPAs. This suggests that in patients with a larger effective peritoneal surface area, more ultrafiltration can be achieved by glucose polymer solutions. Clearances of beta 2-microglobulin (beta 2m) were higher with icodextrin than with 3.86% glucose and 1.36% glucose dialysate (P < 0.05). No differences were found for the larger serum proteins albumin, IgG and alpha 2-macroglobulin. Initial D/PNa-->was higher (0.96) with icodextrin than with the glucose based solutions (0.92), due to the higher Na+ concentration of icodextrin, and it remained unchanged during the dwell. In contrast, D/PNa+ of 1.36% glucose increased during the dwell, whereas D/PNa+ decreased with 3.86% glucose until 60 minutes, followed by a subsequent increase. The ultrafiltration coefficient (UFC) of the total peritoneal membrane was assessed using 3.86% glucose (0.18 +/- 0.04 ml/min/mm Hg), and the UFC of the small pores was assessed using icodextrin (0.06 +/- 0.008 ml/min/mm Hg). The difference between these represented the UFC through the transcellular pores, which averaged 50.5% of the total UFC, but with a very wide range (0 to 85%). An inverse relation existed between the duration of CAPD treatment and the total ultrafiltration coefficient (r = -0.68, P < 0.04), which could be attributed to a lower UFC of the transcellular pores in long-term patients (r = -0.66, P < 0.05), but not to the UFC of the small pores (r = -0.48, NS). The TCUFRo-60 min through the transcellular pores correlated with the sodium gradient, corrected for diffusion, in the first hour of the dwell (r = 0.69, P < 0.04), indicating that ...