Table 2

Incidence of further three or more–step progression of retinopathy and new PDR between the end of the DCCT and after 18 years of the EDIC study overall and stratified by the level of retinopathy at the end of DCCT

Retinopathy levels at DCCT closeoutFurther ≥3-step progressionPDR
n at risk*No. with event (%)Adjusted risk reduction (%, CI)P valuen at riskNo. with event (%)Adjusted risk reduction (%, CI)P value
All levels1,35846 (36, 54)<0.00011,31847 (30, 60)<0.0001
 Intensive684267 (39.0%)66886 (12.9%)
 Conventional674380 (56.4%)650172 (26.5%)
Stratified by retinopathy levels at DCCT closeout
 Stratum 1: no retinopathy30 (5, 49)0.0218 (−186, 70)0.89
  Intensive194100 (51.6%)1948 (4.1%)
  Conventional12374 (60.2%)1225 (4.1%)
 Stratum 2: microaneurysm only54 (38, 65)<0.000153 (19, 73)0.007
  Intensive27588 (32.0%)27423 (8.4%)
  Conventional220112 (50.9%)22032 (14.6%)
 Stratum 3: mild nonproliferative retinopathy55 (33, 70)<0.000152 (23, 70)0.002
  Intensive14944 (29.5%)14931 (20.8%)
  Conventional200101 (50.5%)19964 (32.2%)
 Stratum 4: moderate or severe nonproliferative retinopathy45 (17, 63)0.00444 (9, 65)0.018
  Intensive6535 (53.9%)5024 (48.0%)
  Conventional12693 (73.8%)10471 (68.3%)
  • * Analysis includes all subjects who were free of scatter photocoagulation during DCCT, alive at the initiation of EDIC, and had at least one retinal evaluation in EDIC. The stratified analysis was limited to those with retinopathy measurements at DCCT closeout. Analyses were stratified by retinopathy severity at the end of DCCT, defined as no retinopathy (ETDRS grade 10/10), microaneurysms only (grade 20), mild nonproliferative diabetic retinopathy (NPDR) (grade 30), or greater or equal to moderate NPDR (≥grade 40 or scatter laser).

  • A separate Weibull model was performed for each strata and for all levels combined, after adjustment for primary/secondary cohort, HbA1c value at entry to the DCCT, and diabetes duration at DCCT baseline. Analysis of all levels combined was also adjusted for the level of retinopathy at the end of the DCCT. Risk reduction is for intensive therapy as compared with conventional therapy. The P value is obtained from a Wald test of the group coefficient in the model.

  • Analysis includes all subjects who were free of PDR during DCCT, alive at the initiation of EDIC, and either had an EDIC retinal assessment or reported pan-retinal laser treatment for retinopathy during EDIC.