EDIC
PubMed • Full text • PDF
Clinical Question
In patients with T1DM, what is the benefit of intensive therapy in reducing the long-term incidence of a CV event?
Bottom Line
The EDIC long term follow-up to the DCCT demonstrated that a period of intensive glycemic control reduced the subsequent risk of a CV event by 42% and severe CV events by 57% at 11 years among patients with T1DM.
Major Points
The earlier Diabetes Control and Complications Trial (DCCT) investigated CV outcomes among T1DM patients, but failed to demonstrate a reduction in CV events, likely because the study population was relatively young at the time. The follow-up Epidemiology of Diabetes Interventions and Complications (EDIC) cohort study did demonstrate such benefits.
Long-term observational data from EDIC demonstrated that, compared with conventional therapy, intensive glycemic control reduced the risk of a CV event by 42% and reduced the risk of severe CV events, including nonfatal MI, stroke, or CV death by 57%.
Subgroup analyses demonstrated that the differences in the mean HbA1c levels at the end of DCCT (7.4% vs. 9.1%) explained the majority of the treatment effect of intensive therapy on the risk of CV events. A history of microalbuminuria or albuminuria also independently increased risk of CV events, but differences in outcome remained after correction for these factors.
The differences in HbA1c narrowed by the end of EDIC (7.9% vs. 7.8%) under the same intensive treatment regimen. However, the risk of retinopathy and nephropathy in the intensive therapy during DCCT group continued to be reduced over the next 10 years, despite comparable HbA1c values. This phenomenon is commonly referred to as metabolic memory, and suggests that a sustained period of intensive therapy (approximately 6.5 years during DCCT) can have a long-term effect on the subsequent risk of microvascular complications as well as CV morbidity and mortality in T1DM.
Design
- Prospective, observational cohort study
- N=1,394 patients after completion of DCCT in 1993 (97% from both arms) continued with intensive insulin therapy and joined EDIC in 1994.
- Intensive therapy (n=698)
- Conventional therapy (n=723)
- Mean follow-up: 11 years
Population
Inclusion Criteria
- Participation in DCCT.
- No new inclusion criteria than those in the DCCT were identified.
Exclusion Criteria
- No new exclusion criteria than those in the DCCT were identified.
Baseline Characteristics
Comparisons are intensive vs. conventional therapy at the end of DCCT (1993).
- Mean age: 34 vs. 33 years
- Females: 49% vs. 46%
- BMI: 26.6 vs. 25.1 (P<0.01)
- Duration of diabetes: 12 years
- Current smoker: 20%
- SBP: 117 mmHg
- DBP: 75 mmHg
- HTN: 3% vs. 4%
- HL: 26% vs. 30%
- LDL: 112 vs. 115 mg/dL
- HDL: 51 vs 52 mg/dL
- Total cholesterol: 180 vs. 184 mg/dL
- Triglycerides: 84 vs. 88 mg/dL (P<0.05)
- Albumin excretion rate: 29.8 vs. 75.4 mg/24 hr (P<0.01)
- ≥40 mg/24 hr: 7% vs. 13% (P<0.01)
- ≥300 mg/24 hr: 1% vs. 3% (P<0.05)
- HbA1c: 7.4% vs. 9.1% (P<0.01)
- Intensive diabetes management: 98% vs. 10% (P<0.01)
Interventions
- Interventions in DCCT were continued
- Annual measurements of HbA1c, EKGs
- Biennial measurements of fasting lipid levels, serum creatinine
- Diabetes management per PCP
Outcomes
Characteristics
Comparisons are intensive (n=593) vs. conventional therapy (n=589) at the end of DCCT (1993).
- Mean age: 45 years
- Females: 48% vs. 46%
- BMI: 28.4 vs. 27.6
- Duration of diabetes: 24 years
- Current smoker: 14% vs. 11%
- SBP: 120 mmHg
- DBP: 75 mmHg
- HTN: 38% vs. 41%
- HL: 52% vs. 48%
- LDL: 112 vs. 109 mg/dL
- HDL: 55 mg/dL
- Total cholesterol: 186 vs. 181 mg/dL (P<0.05)
- Triglycerides: 93 vs. 86 mg/dL (P<0.05)
- Albumin excretion rate: 54.2 vs. 116.4 mg/24 hr (P<0.01)
- ≥40 mg/24 hr: 9% vs. 17% (P<0.01)
- ≥300 mg/24 hr: 2% vs. 6% (P<0.01)
- Serum creatinine ≥2 mg/dL: 0% vs. 2% (P<0.05)
- HbA1c: 7.9% vs. 7.8% (P=0.38)
- Intensive diabetes management: 97% vs. 96%
- Medications
- ACE-inhibitors or ARBs: 38% vs. 43%
- Hormone-replacement therapy: 6% vs. 4%
- ≥14 aspirin tablets/month: 37% vs. 40%
- Beta-blocker: 3% vs. 7% (P<0.01)
- Statin: 34% vs. 33%
Comparisons are intensive vs. conventional therapy at year 11 of EDIC (2004).
Primary Outcomes
- CV event (CV death, nonfatal MI or stroke, silent MI, angina, revascularization)
- First: 31 vs. 52 events (RR 0.58, 95% CI 0.37-0.91, P=0.02); 0.38 vs. 0.80 events/100 patient years (P=0.007)
- Any: 46 vs. 98 events
Secondary Outcomes
- Nonfatal MI, stroke, or CV death
- (RR 0.43, 95% CI 0.21-0.88, p=0.02)
- First: 11 vs. 25 events
- Any: 11 vs. 30 events
- Mean HbA1c
- 8.0% vs. 8.2% (RR 0.98; P=0.03)
- CV death
- First: 3 vs. 4 events
- Any: 3 vs. 9 events
- Nonfatal acute MI
- First: 6 vs. 11 events
- Any: 7 vs. 16 events
- Silent MI
- First: 7 vs. 13 events
- Any: 7 vs. 21 events
- Revascularization (CABG or PCI)
- First: 4 vs. 6 events
- Any: 17 vs. 25 events
- Angina
- First: 10 vs. 13 events
- Any: 11 v. 22 events
Subgroup Analysis
- Risk of CV disease before adjusting for time-dependent covariate
- Renal disease (yes vs. no): HR 2.99; 95% CI 0.72-12.5; P=0.20
- Microalbuminuria (yes vs. no): HR 2.93; 95% CI 1.85-4.65; P<0.001
- Albuminuria (yes vs. no): HR 2.57; 95% CI 1.36-4.88; P=0.009
- Mean HbA1c
- per 10% increase: HR 1.25; 95% CI 1.10-1.43; P<0.001
- per 10% decrease: HR 0.80; 95% CI 0.70-0.91; P<0.001
- Risk of CV disease after adjusting for time-dependent covariate
- None: HR 0.53; 95% CI 0.34-0.83; P=0.005
- Microalbuminuria (yes vs. no): HR 0.62; 95% CI 0.39-0.97; P=0.04
- Albuminuria (yes vs. no): HR 0.58; 95% CI 0.37-0.91; P=0.02
- Mean HbA1c
- per 10% increase: HR 0.84; 95% CI 0.43-1.64; P=0.61
At baseline, older age (31 vs. 27 years), longer duration of diabetes (7 vs. 6 years), presence of retinopathy, current smoking, higher BMI (24.0 vs. 23.3), higher total and LDL cholesterol levels (194 v. 175 mg/dL and 127 vs. 109 mg/dL, respectively), higher HbA1c (9.5% vs. 9.0%), and higher albumin excretion rate (19.3 vs. 15.7 mg/24 hours), and assignment to conventional treatment were all associated with development of CV disease.
Criticisms
- No definitive analysis of treatment effects on the individual types of CVD because of the low amount of events
- The need for revascularization was a subjective outcome dependent on a clinician's judgment
- There was a higher-than-expected amount of silent MIs
- Interventions during the original trial were not blinded
Funding
Supported by National Institute of Diabetes, Digestive, and Kidney Diseases division of Diabetes, Endocrinology and Metabolic Disease, National Center for Research Resources General Clinical Research Center Program