Table 2.

Multilevel mixed linear model predicting daily peritoneal protein clearance (square root transformed): base model and model adding dialysate glucose effect

VariableBase ModelModel with Glucose Exposure
CoefficientSEMP ValueCoefficientSEMP Value
Constant9.570.3<0.019.440.34<0.01
PSTR (for each 0.1 increase)0.390.11<0.010.430.11<0.01
Comorbidity score (for each 1 unit increase)−0.130.20.51−0.180.210.39
Sex (if female)−0.520.380.17−0.370.40.35
Age (year)0.0160.0110.140.0190.0110.1
UF capacity on PET (for each 100-ml increase)0.040.030.240.040.040.22
Urine volume (for each 1000-ml increase)0.20.10.20.20.20.15
PSTR×PD duration (year 2 versus year 1 for each 0.1 increase)−0.230.140.11−0.250.140.08
PSTR×PD duration (year 3 versus year 1 for each 0.1 increase)−0.350.14<0.01−0.370.14<0.01
PSTR×PD duration (year 4 versus year 1 for each 0.1 increase)−0.300.14<0.05−0.270.150.07
Icodextrin use (versus not used)0.0070.0190.7
Medium glucose exposure (daily average, 1.36% (1.5%)–2.27% (2.5%) (versus low glucose exposure) (n=51)0.020.020.41
High glucose exposure (daily average≥2.27%) (versus low glucose exposure) (n=13)0.030.030.45
  • See Supplemental Table 1 for calculation of the progressive effect over time on the PSTR β coefficient. UF, ultrafiltration; PET, peritoneal equilibration test; PD, peritoneal dialysis.