Table 2.

Multivariable analyses of risk of CKD progression (ESRD or 50% eGFR loss and ESRD), differences in eGFR slope, and mortality (baseline and time-updated covariates)

ModelSurvival ModelsaMixed-Effects ModelsSurvival Modelsa
ESRD or 50% eGFR LossESRDbDifference in eGFR Slope (ml/min per 1.73 m2 per yr)Death
Hispanic versus NHWHispanic versus NHBHispanic versus NHWHispanic versus NHBHispanic versus NHWHispanic versus NHBHispanic versus NHWHispanic versus NHB
HR (95% CI)P ValueHR (95% CI)P ValueHR (95% CI)P ValueHR (95% CI)P ValueMean (95% CI)P ValueMean (95% CI)P valueHR (95% CI)P ValueHR (95% CI)P Value
Model 12.08 (1.65 to 2.62)<0.0011.16 (0.94 to 1.43)0.161.83 (1.40 to 2.38)<0.0011.04 (0.82 to 1.32)0.76−0.45 (−0.66 to −0.23)<0.001−0.07 (−0.29 to 0.14)0.490.96 (0.71 to 1.30)0.800.81 (0.61 to 1.07)0.13
Model 22.22 (1.73 to 2.84)<0.0011.24 (0.99 to 1.54)0.062.28 (1.71 to 3.03)<0.0011.22 (0.95 to 1.58)0.12−0.42 (−0.65 to −0.20)0.001−0.09 (−0.31 to 0.13)0.411.15 (0.78 to 1.69)0.481.06 (0.75 to 1.50)0.74
Model 31.90 (1.46 to 2.48)<0.0011.26 (0.997 to 1.58)0.051.81 (1.34 to 2.45)<0.0011.17 (0.89 to 1.53)0.25−0.41 (−0.66 to −0.17)0.001−0.21 (−0.45 to 0.03)0.081.16 (0.78 to 1.71)0.471.07 (0.75 to 1.53)0.70
Model 41.28 (0.96 to 1.70)0.090.88 (0.68 to 1.13)0.311.32 (0.96 to 1.81)0.080.94 (0.71 to 1.25)0.67−0.09 (−0.33 to 0.16)0.480.09 (−0.15 to 0.33)0.460.89 (0.59 to 1.35)0.580.89 (0.61 to 1.30)0.54
Model 4a1.10 (0.67 to 1.81)0.700.61 (0.39 to 0.95)0.030.92 (0.51 to 1.64)0.770.50 (0.30 to 0.84)0.01−0.06 (−0.41 to 0.30)0.760.15 (−0.20 to 0.50)0.40cc
Model 4b1.31 (0.93 to 1.84)0.121.04 (0.77 to 1.40)0.811.43 (0.97 to 2.10)0.071.16 (0.83 to 1.63)0.38−0.23 (−0.56 to 0.11)0.19−0.05 (−0.37 to 0.27)0.76cc
  • Model 1: age, sex, baseline eGFR, diabetes, clinical site. Model 2: Model 1 + education, health insurance, nephology care, smoking. Model 3: Model 2 + systolic BP, ACEi/ARB use, BMI, HbA1c. Model 4: Model 3 + log (24-hour urine protein + 1). Model 4a: Model 4 participants without diabetes only. Model 4b: Model 4 participants with diabetes only.

  • a All models with time-updated nephrology care, systolic BP, ACEi/ARB use, BMI, and HbA1c.

  • b Death treated as a competing risk using the method of Fine and Gray.43

  • c Interaction between race and diabetes in Model 4 for death is not significant (P>0.05).