The Diabetes Control and Complications Trial (DCCT)/Epidemiology of Diabetes Interventions and Complications follow-up study (EDIC) recently reported that after age mean HbA1c was the strongest predictor of MACE over 27 year of follow-up. The DCCT/EDIC cohort had a short T1D duration at baseline (mean 6 yr) and those with high blood pressure or cholesterol were excluded so it is unclear if these results apply to longer T1D durations and without risk factor exclusion. We thus replicated these analyses in the Pittsburgh EDC study a prospective cohort study of childhood onset (<17 yr) T1D with 25 year of follow-up and similar age (mean 27 year in both studies) but longer T1D duration (mean 18 yr) at baseline. EDC participants were diagnosed 1950-1980 (n=658 49% women) and examined at 1986-88 baseline. Follow-up exams occurred at 2, 4, 6, 8, 10, 18 and 25 years. MACE incidence (CVD death, MI, stroke) was ascertained by death certificate or self reported and confirmed with medical records. Participants with prevalent CVD were excluded (n=54). Baseline (BL), time varying updated mean (UM) and time varying most recent (MR) risk factors were assessed in Cox models as in DCCT/EDIC. In EDC 18% (107) developed MACE vs. 6% in DCCT/EDIC. The final multivariable model for MACE comprised BL T1D duration (HR 1.1, p<.001), MR albumin excretion rate (AER) (HR 1.3, p<.001), UM SBP (HR 1.03, p<.001), BL smoking (HR 1.9, p=0.003), UM LDLc (HR 1.01, p=0.03) and UM HbA1c (HR 1.2, p=0.03). In DCCT/EDIC the HbA1c effect size was larger while T1D duration, UM SBP, MR (not UM as in EDC) LDLc and MR (not BL as in EDC) smoking predicted with similar effect sizes. ACE inhibitors (protective) and pulse rate also predicted MACE in DCCT/EDIC. While AER was a strong predictor in EDC no renal markers were significant in DCCT/EDIC. These results suggest that at longer T1D duration much of the HbA1c effect on MACE risk is mediated through other factors as recently reported by DCCT/EDIC.
R.G. Miller: None. T.J. Orchard: None.