TABLE 3

Multivariate proportional hazards analysis of ACE genotype and renal outcomes

ACE genotype
Grades of diabetic nephropathy
Persistent microalbuminuria (n = 246)
Severe nephropathy (n = 115)
Single locusGenotype comparisonsHR95% CIWald P valueHR95% CIWald P value
rs1800764CC vs. CT0.980.69–1.380.900.690.38–1.230.20
TT vs. CT0.610.44–0.840.003*0.620.39–0.980.043*
Insertion/deletionDD vs. DI0.980.72–1.330.910.890.56–1.390.60
II vs. DI0.620.43–0.890.009*0.560.32–0.960.033*
rs9896208TT vs. CT1.120.77–1.620.560.950.54–1.650.85
CC vs. CT0.650.49–0.880.005*0.540.34–0.850.008*
Haplotype
    rs1800764, insertion/ deletion, rs9896208
        CDT/CDT0.940.58–1.510.790.790.38–1.640.53
        TIC/TIC0.490.32–0.750.0009*0.410.22–0.780.006*
        CDC/CDT0.700.39–1.250.220.520.19–1.390.19
        CDC/TIC0.680.43–1.060.090.590.29–1.180.14
        CDT/other0.890.57–1.380.600.890.48–1.650.71
        TIC/other0.560.35–0.920.02*0.580.29–1.160.12
        Other/other0.760.36–1.630.480.510.15–1.710.27
  • HRs and Wald tests for ACE genotype association with the time (years) from DCCT entry to the development of persistent microalbuminuria (246 events and 1,053 censored; 66 subjects had AER >20.8 μg/min at DCCT baseline and were therefore excluded from the CoxPH analysis) and severe nephropathy (115 events and 1,250 censored) in multivariate Cox proportional hazards analyses stratified by year of entry into the DCCT. Each locus model was adjusted for cohort, treatment, cohort by treatment interaction, and covariate effects, including linear and quadratic terms for age at diagnosis of type 1 diabetes, duration of type 1 diabetes prior to DCCT entry, sex, HbA1c at DCCT eligibility, and DCCT baseline measures for BMI, mean blood pressure, triglyceride, total cholesterol, HDL cholesterol, and the time-dependent updated HbA1c weighted mean. Indicated haplotypes are compared with the most frequent haplotype pair (CDT/TIC).

  • *

    * P < 0.05.